jw-pkg/src/python/jw/pkg/App.py

666 lines
22 KiB
Python
Raw Normal View History

#
# This source code file is a merge of various build tools and a horrible mess.
#
from __future__ import annotations
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
import os
import pwd
import re
import sys
from enum import Enum, auto
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
from functools import lru_cache
from typing import TYPE_CHECKING
from .lib.App import App as Base
from .lib.Distro import Distro
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
from .lib.log import DEBUG, ERR, log
if TYPE_CHECKING:
import argparse
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
from typing import TypeAlias
from .lib.ExecContext import ExecContext
from .lib.PackageFilter import PackageFilter
# Meaning of pkg.requires.xxx variables
# build: needs to be built and installed before this can be built
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
# devel: needs to be installed before this-devel can be installed,
# i.e. before _other_ packages can be built against this
# run: needs to be installed before this-run can be installed,
# i.e. before this and other packages can run with this
# --------------------------------------------------------------------- Helpers
class ResultCache(object):
def __init__(self):
self.__cache = {}
def run(self, func, args):
d = self.__cache
depth = 0
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
keys = [func.__name__] + args
sz = len(keys)
for k in keys:
if k is None:
k = 'None'
else:
k = str(k)
depth += 1
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
# log(DEBUG, 'depth = ', depth, 'key = ', k, 'd = ', str(d))
if k in d:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
if sz == depth:
return d[k]
d = d[k]
continue
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
if sz == depth:
r = func(*args)
d[k] = r
return r
d = d[k] = {}
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
# d = d[k]
raise Exception(
'cache algorithm failed for function', func.__name__, 'in depth', depth
)
class Scope(Enum):
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
Self = auto()
One = auto()
Subtree = auto()
Graph: TypeAlias = dict[str, set[str]]
# ----------------------------------------------------------------- class App
class App(Base):
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
def __format_topdir(self, path: None | str, fmt: str) -> str | None:
if path is None:
return None
match fmt:
case 'unaltered':
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
return path
case None | 'absolute':
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
return os.path.abspath(path)
case _:
m = re.search(r'^make:(\S+)$', fmt)
if m is None:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
raise Exception(
f'Can\'t interpret "{fmt}" as valid topdir reference, '
'expecting "unaltered", "absolute", or "make:<variable-name>"'
)
return '$(' + m.group(1) + ')'
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
@property
def __topdir(self) -> str:
if self.___topdir is None:
raise Exception('Tried to access undefined top directory')
return self.___topdir
@property
def __pretty_topdir(self) -> str:
if self.___pretty_topdir is None:
raise Exception('Tried to access undefined pretty top directory')
return self.___pretty_topdir
def __proj_dir(self, name: str, pretty: bool) -> str | None:
if name == self.__top_name:
if pretty:
return self.__pretty_topdir
return self.__topdir
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
for d in [self.__projs_root, '/opt']:
ret = d + '/' + name
if os.path.exists(ret):
return ret
if os.path.exists(f'/usr/share/doc/packages/{name}/VERSION'):
# The package exists but does not have a dedicated project directory
return None
raise Exception('No project path found for module "{}"'.format(name))
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
def __find_dir(
self,
name: str,
search_subdirs: list[str] = [],
search_absdirs: list[str] = [],
pretty: bool = True,
) -> str | None:
def format_pd(name: str, pd: str, pretty: bool):
if not pretty:
return pd
if self.__topdir_fmt == 'absolute':
return os.path.abspath(pd)
if self.__topdir_fmt == 'unaltered':
return pd
if name == self.__top_name:
return self.__pretty_topdir
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
raise NotImplementedError(
f'Tried to pretty-format directory {pd}, not implemented'
)
pd = self.__proj_dir(name, False)
if pd is None:
return None
if not search_subdirs and not search_absdirs:
return format_pd(name, pd, pretty)
for sd in search_subdirs:
path = pd + '/' + sd
if os.path.isdir(path):
ret = format_pd(name, pd, pretty)
if sd and sd[0] != '/':
ret += '/'
ret += sd
return ret
for ret in search_absdirs:
if os.path.isdir(ret):
return ret
return None
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
def __get_project_refs_cached(
self, buf, visited, spec, section, key, add_self, scope, names_only
):
return self.__res_cache.run(
self.__get_project_refs,
[buf, visited, spec, section, key, add_self, scope, names_only],
)
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
def __get_project_refs(
self,
buf: list[str],
visited: set[str],
spec: str,
section: str,
key: str,
add_self: bool,
scope: Scope,
names_only: bool,
) -> None:
name = self.strip_module_from_spec(spec)
if names_only:
spec = name
if spec in buf:
return
if spec in visited:
if add_self:
buf.append(spec)
return
visited.add(spec)
deps = self.get_value(name, section, key)
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
log(
DEBUG,
(
f'name={name}, section={section}, key={key}, deps={deps}, '
f'scope={scope.name}, visited={visited}'
),
)
if deps and scope != Scope.Self:
if scope == Scope.One:
subscope = Scope.Self
else:
subscope = Scope.Subtree
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
for dep in deps.split(','):
dep = dep.strip()
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
if not (len(dep)):
continue
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
self.__get_project_refs_cached(
buf,
visited,
dep,
section,
key,
add_self = True,
scope = subscope,
names_only = names_only,
)
if add_self:
buf.append(spec)
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
def __read_dep_graph(
self,
projects: list[str],
sections: str | list[str],
graph: Graph,
) -> None:
if isinstance(sections, str):
sections = [sections]
for project in projects:
if project in graph:
continue
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
for section in sections:
deps = self.get_project_refs(
[project],
['pkg.requires.jw'],
sections,
scope = Scope.One,
add_self = False,
names_only = True,
)
if deps is None:
continue
graph[project] = set(deps)
for dep in deps:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
self.__read_dep_graph([dep], sections, graph)
def __flip_dep_graph(self, graph: Graph):
ret: Graph = {}
for project, deps in graph.items():
for d in deps:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
if d not in ret:
ret[d] = set()
ret[d].add(project)
return ret
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
def __find_circular_deps_recursive(
self,
project: str,
graph: Graph,
unvisited: list[str],
temp: set[str],
path: list[str],
) -> str | None:
if project in temp:
log(DEBUG, 'found circular dependency at project', project)
return project
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
if project not in unvisited:
return None
temp.add(project)
if project in graph:
for dep in graph[project]:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
last = self.__find_circular_deps_recursive(
dep, graph, unvisited, temp, path
)
if last is not None:
path.insert(0, dep)
return last
unvisited.remove(project)
temp.remove(project)
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
return None
def __find_circular_deps(self, projects: list[str], flavours: list[str]) -> bool:
graph: Graph = {}
ret: list[str] = []
self.__read_dep_graph(projects, flavours, graph)
unvisited = list(graph.keys())
temp: set[str] = set()
while unvisited:
project = unvisited[0]
log(DEBUG, 'Checking circular dependency of', project)
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
last = self.__find_circular_deps_recursive(
project, self.__flip_dep_graph(graph), unvisited, temp, ret
)
if last is not None:
log(DEBUG, f'Found circular dependency below {project}, last is {last}')
return True
return False
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
def __init__(self, distro: Distro | None = None) -> None:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
super().__init__('jw-pkg swiss army knife', modules = ['jw.pkg.cmds'])
# -- Members without default values
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
self.__opt_interactive: bool | None = None
self.__opt_verbose: bool | None = None
self.__top_name: str | None = None
self.__distro = distro
self.__res_cache = ResultCache()
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
self.___topdir: str | None = None
self.___pretty_topdir: str | None = None
self.__exec_context: ExecContext | None = None
# -- Members with default values
self.__topdir_fmt = 'absolute'
self.__projs_root = pwd.getpwuid(os.getuid()).pw_dir + '/local/src/jw.dev/proj'
self.__pretty_projs_root = None
async def __init_async(self) -> None:
if self.__distro is None:
pkg_filter_str = self.args.pkg_filter
if pkg_filter_str is None:
pkg_filter_str = os.getenv('JW_DEFAULT_PKG_FILTER')
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
pkg_filter: PackageFilter | None = None
if pkg_filter_str is not None:
from .lib.PackageFilter import PackageFilterString
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
pkg_filter = PackageFilterString(pkg_filter_str)
self.__distro = await Distro.instantiate(
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
ec = self.exec_context,
id = self.args.distro_id,
default_pkg_filter = pkg_filter,
)
async def __aexit__(self, exc_type, exc, tb) -> None:
if self.__exec_context is not None:
await self.__exec_context.close()
self.__exec_context = None
def _add_arguments(self, parser) -> None:
super()._add_arguments(parser)
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
parser.add_argument('-t', '--topdir', default = None, help = 'Project Path')
parser.add_argument(
'--topdir-format',
default = 'absolute',
help = (
'Output references to topdir as one of "make:<var-name>", '
'"unaltered", "absolute". Absolute topdir by default'
),
)
parser.add_argument(
'-p',
'--prefix',
default = None,
help = 'Parent directory of project source directories',
)
parser.add_argument(
'--distro-id',
default = None,
help = 'Distribution ID (default is taken from /etc/os-release)',
)
parser.add_argument(
'--interactive',
choices = ['true', 'false', 'auto'],
default = 'true',
help = 'Wait for user input or try to proceed unattended',
)
parser.add_argument(
'--verbose',
action = 'store_true',
default = False,
help = "Be verbose on stderr about what's being done on the distro level",
)
parser.add_argument(
'--target', default = 'local', help = 'Run commands on this host'
)
parser.add_argument(
'--pkg-filter',
help = 'Default filter for all distribution package-related operations',
)
async def _run(self, args: argparse.Namespace) -> None:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
self.___topdir = args.topdir
self.___pretty_topdir = self.__format_topdir(self.___topdir, args.topdir_format)
self.__topdir_fmt = args.topdir_format
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
if self.___topdir is not None:
self.__top_name = self.read_value(
self.___topdir + '/make/project.conf', 'build', 'name'
)
if not self.__top_name:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
self.__top_name = re.sub(
'-[0-9.-]*$',
'',
os.path.basename(os.path.realpath(self.___topdir))
)
if args.prefix is not None:
self.__projs_root = args.prefix
self.__pretty_projs_root = args.prefix
await self.__init_async()
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
await super()._run(args)
@property
def interactive(self) -> bool:
if self.__opt_interactive is None:
match self.args.interactive:
case 'true':
self.__opt_interactive = True
case 'false':
self.__opt_interactive = False
case 'auto':
self.__opt_interactive = sys.stdin.isatty()
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
case _:
raise ValueError(
f'Unknown --interactive value: {self.args.interactive}'
)
# Not logically possible to fail, but this keeps pyright happy
assert self.__opt_interactive is not None
return self.__opt_interactive
@property
def verbose(self) -> bool:
if self.__opt_verbose is None:
self.__opt_verbose = self.args.verbose
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
# Not logically possible to fail, but this keeps pyright happy
assert self.__opt_verbose is not None
return self.__opt_verbose
@property
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
def exec_context(self) -> ExecContext:
if self.__exec_context is None:
from .lib.ExecContext import ExecContext
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
self.__exec_context = ExecContext.create(
self.args.target,
interactive = self.interactive,
verbose_default = self.verbose,
)
return self.__exec_context
@property
def top_name(self):
return self.__top_name
@property
def projs_root(self):
return self.__projs_root
@property
def distro(self) -> Distro:
if self.__distro is None:
raise Exception('No distro object')
return self.__distro
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
def find_dir(
self,
name: str,
search_subdirs: list[str] = [],
search_absdirs: list[str] = [],
pretty: bool = True,
) -> str:
ret = self.__find_dir(name, search_subdirs, search_absdirs, pretty)
if ret is None:
msg = f'Failed to find directory for "{name}"'
log(ERR, msg)
for search_name, search in [
('subdirs', search_subdirs),
('absdirs', search_absdirs),
]:
if search:
log(ERR, f' Searched {search_name} in:')
for d in search:
log(ERR, f' - {d}')
raise FileNotFoundError(msg)
return ret
# TODO: add support for customizing this in project.conf
def htdocs_dir(self, project: str) -> str:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
return self.find_dir(
project,
['/src/html/htdocs', '/tools/html/htdocs', '/htdocs'],
['/srv/www/proj/' + project],
)
# TODO: add support for customizing this in project.conf
def tmpl_dir(self, name: str) -> str:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
return self.find_dir(name, ['/tmpl'], ['/opt/' + name + '/share/tmpl'])
def strip_module_from_spec(self, mod):
return re.sub(r'-dev$|-devel$|-run$', '', re.split('([=><]+)', mod)[0].strip())
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
@lru_cache(maxsize = None)
def get_section(self, path: str, section: str) -> str:
ret = ''
pat = '[' + section + ']'
in_section = False
file = open(path)
for line in file:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
if line.rstrip() == pat:
in_section = True
continue
if in_section:
if len(line) and line[0] == '[':
break
ret += line
file.close()
return ret.rstrip()
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
@lru_cache(maxsize = None)
def read_value(self, path: str, section: str, key: str) -> str | None:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
def scan_section(f, key: str) -> str | None:
if key is None:
ret = ''
for line in f:
if len(line) and line[0] == '[':
break
ret += line
return ret if len(ret) else None
lines: list[str] = []
cont_line = ''
for line in f:
if len(line) and line[0] == '[':
break
cont_line += line.rstrip()
if len(cont_line) and cont_line[-1] == '\\':
cont_line = cont_line[0:-1]
continue
lines.append(cont_line)
cont_line = ''
rx = re.compile(r'^\s*' + key + r'\s*=\s*(.*)\s*$')
for line in lines:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
# log(DEBUG, ' looking for "%s" in line="%s"' % (key, line))
m = re.search(rx, line)
if m is not None:
return m.group(1)
return None
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
def scan_section_debug(f, key: str) -> str | None:
ret = scan_section(f, key)
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
# log(DEBUG, ' returning', rr)
return ret
try:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
# log(DEBUG, 'looking for {}::[{}].{}'.format(path, section, key))
# TODO: Parse
with open(path, 'r') as f:
if not len(section):
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
return scan_section(f, key)
pat = '[' + section + ']'
for line in f:
if line.rstrip() == pat:
return scan_section(f, key)
return None
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
except Exception:
log(DEBUG, f'Not found: {path}')
# TODO: handle this special case cleaner somewhere up the stack
if section == 'build' and key == 'libname':
return 'none'
return None
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
@lru_cache(maxsize = None)
def get_value(self, project: str, section: str, key: str) -> str | None:
ret: str | None
proj_dir = self.__proj_dir(project, pretty = False)
if proj_dir is None:
raise Exception(f"Can't get project directory for {project}")
if section == 'version':
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
proj_version_dirs = [proj_dir]
if proj_dir != self.___topdir:
proj_version_dirs.append('/usr/share/doc/packages/' + project)
for d in proj_version_dirs:
version_path = d + '/VERSION'
try:
with open(version_path) as fd:
ret = fd.read().replace('\n', '').replace('-dev', '')
fd.close()
return ret
except EnvironmentError:
log(DEBUG, f'Ignoring unreadable file "{version_path}"')
continue
raise Exception(f'No version file found for project "{project}"')
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
path = proj_dir + '/make/project.conf'
ret = self.read_value(path, section, key)
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
log(
DEBUG,
'Lookup %s -> %s / [%s%s] -> "%s"' %
(self.__top_name, project, section, '.' + key if key else '', ret),
)
return ret
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
@lru_cache(maxsize = None)
def get_version(self, project) -> str:
ret = self.get_value(project, 'version', '')
if ret is None:
raise Exception(f"Can't get version of project {project}")
return ret
def get_values(self, projects: list[str], sections: list[str],
keys: list[str]) -> list[str]:
"""
Collect a list of values from a list of given projects, sections and
keys, maintaining order
"""
ret: list[str] = []
for p in projects:
for section in sections:
for key in keys:
vals = self.get_value(p, section, key)
if vals:
ret += [val.strip() for val in vals.split(',')]
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
return list(dict.fromkeys(ret)) # Remove duplicates, keep ordering
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
def get_project_refs(
self,
projects: list[str],
sections: list[str],
keys: str | list[str],
add_self: bool,
scope: Scope,
names_only = True,
) -> list[str]:
if isinstance(keys, str):
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
keys = [keys]
ret: list[str] = []
for section in sections:
for key in keys:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
visited: set[str] = set()
for name in projects:
rr: list[str] = []
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
self.__get_project_refs_cached(
rr, visited, name, section, key, add_self, scope, names_only
)
# TODO: this looks like a performance hogger
for m in rr:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
if m not in ret:
ret.append(m)
return ret
def get_libname(self, projects) -> str:
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
vals = self.get_project_refs(
projects,
['build'],
'libname',
scope = Scope.One,
add_self = False,
names_only = True,
)
if not vals:
return ' '.join(projects)
if 'none' in vals:
vals.remove('none')
return ' '.join(reversed(vals))
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
def is_excluded_from_build(self, project: str) -> str | None:
log(DEBUG, 'checking if project ' + project + ' is excluded from build')
exclude = self.get_value(project, 'build', 'exclude')
if exclude is None:
return None
exclude_arr = re.split(r'[, ]+', exclude)
jw.pkg: Fix "make check" static code check fallout The previous commits have put rules for linting and formatting via ruff, yapf, mypy and pyright into place. They are checked with the make check target, and this commit adds the fixes for the target to succeed. It does some refactoring where type checking dug up dirty bits, and also adds lots of churn in the Python code. To a good deal, that's owed to mere formatting changes. It would have been better to seperate those from syntax and refactoring fixes into multiple commits, so that the interesting changes don't drown in the formatting nose. However, that would have been a lot of additional work only to be thrown away by later commits, hence this commit has a big diff in one piece. The size of the diff is regrettable but hopefully a one-off: What it buys is automatic format checking for CI and predictble formats for smaller diffs in the future. Rules that "make check" enforces are, in the following order - Syntax checkers: - ruff check . - mypy . - pyright - Format check: - yapf --diff --recursive . The refactoring includes: - Turn the Result class into a more elaborate object, capable of doing more heavy lifting around stderr and stdout decoding, summarizing outcome, and matching error strings. Aside from fixing broken type checks, this also removes lots of boilerplate calling code which is currently used for handling possible call outcome scenarios. Trying to access an inexistent, decoded string should raise a meaningful exception by itself now, which removes lots of code with case distinctions. - Fix Cmd type hierarchy: - Add the AbstractCmd class above Cmd. This is necessary because the checker rightfully complains it can't instantiate a Cmd instance where constructor arguments were needed. They never were, but the type used at the instantiating code's location in jw.pkg.App so claims. - Lots of sub- and sub-subcommands are derived from the base class of the invoking command. That provides some properties shared across the ancestor hierarchy of a command, but is semantically unsound. Fix that by introducing jw.pkg.BaseCmd class as a place to provide basic helpers shared across all commands used in a jw.pkg.App's context, and derive all command classes from that afresh. The parent command is still reachable via a common parent property. Formatting changes are conforming to PEP-8, mostly, with minor tweaks. All in all they include the following changes. - Remove # -*- coding: utf-8 -*- The line was needed by Python 2 which is not supported anylonger. For Python 3, the default encoding is UTF-8, anyway. - Allow to run "make py-format" without having it produce any changes. It's basically "yapf --in-place --recursive ." with some code style settings, see conf/topdir/pyproject.toml. The settings may be debatable. I've had custom tweaks in place on that target, too, but then again, IDEs would have more hassle to integrate that. - Introduce a 88 character line length limit - One import per line, reshuffle them semantically, see [tool.isort] in pyproject.toml. - Hide imports needed for type-checking only behind if TYPE_CHECKING - Spaces around assignments accounts for much churn. Having having no spaces in inline parameter list assignments and default parameter values would arguably be more compact where it's useful. On the other hand, I have not found a code formatter which allows spaces around assignments in parameter lists broken into one per line and that's often better than a wall of text. - Add two spaces before # export, as this seems to be mandated by PEP-8 - Use single quotes by default Signed-off-by: Jan Lindemann <jan@janware.com>
2026-05-27 07:16:05 +02:00
cascade = self.distro.os_cascade + ['all']
intersection = [x for x in cascade if x in set(exclude_arr)]
if intersection:
return ', '.join(intersection)
return None
def find_circular_deps(self, projects: list[str], flavours: list[str]) -> bool:
return self.__find_circular_deps(projects, flavours)