jw-pkg supports more than RPM-based package managers, but for historic reasons, lots of its Makefile variables still have "RPM" in their names. This is misleading. Replace "RPM" in variable names by the more generic "PKG" where appropriate.
Maintainer scripts often mess with systemd services via systemctl. In Docker containers, chroot environments or other environments not governed by Systemd, systemctl will not exist or complain. This is a frequent use case, worthy of providing a wrapper to catch and ignore these cases conveniently.
pkg.run is not evaluated on Debian, fix that. For now it's hacked into pkg.sh, which is bound to be replaced by Python. The limited hassle is still worth the detour.
Rename the --from-user option to --from-owner. Forgejo supports users and organizations under the more general term "owner", so that's the better term.
PGIT_SH_PROJECTS currently only sets the projects to operate on for the get command. Extend this to all commands. And replace the environment variable by a command-line option, likely after this script has been ported to Python.
"clone" in the Git sense means to copy a remote project over from scratch. pgit.sh clone has come from that, but has since evolved into something different, a mixture of clone, pull and fetch, so find a different name. "get" seems generic enough and doesn't clash with a Git meaning. Adapt variable names accordingly across the project.
Continue to name variables in pgit.sh somewhat more consistently, notably turn somevar into some_var. Plus some additional cleanup. Still not a beauty.
If "current-branch" is specified within --refspec, either as from-ref or as to-ref, expand that to the branch the working directory has currently checked out.
dc945537 (pkg.sh log-install: Log %attr(0777, ...) for links) fixed packaging symlink for Debian-based distros, but produces a warning on RPM based distros: Links may not have explicit attributes, so go back to not specifying them at all for RPM.
This commit allows pgit.sh to target not only multiple projects below a projects-directory, but also one single project. If invoked from the toplevel directory of a project, it uses that as the only project it should deal with. This is meant to facilitate running the same VCS abstraction logic for one project as for many projects. The project or projects to deal with should probably be specified on the command line, but changing the auto-detection mechanism buys us what we want for now with low hassle.
Some variable names are too short for global scope ($p, $pdir, $pdirs), among others. For those mentioned: Make them longer and more descriptive.
Also add a variable project_name, which denotes what a project is in a remote repository, and which is currently but not necessarily always the same as the project directory.
Not logging any attribute for links, as it's now, breaks Debian's parser. So, log %attr(0777, $owner, $mode). This fixes the parser on the Debian side and hopefully leaves the RPM side intact.
Calling make git-pull-xxx from a projects directory stops iterating projects if one has a dirty workspace. Calling --autostash fixes that.
With this in place, a failed rebase leaves the local changes behind stashed. So, after manually fixing the rebase, the stash needs to be manually reapplied. The commands that led up to the failure are logged right before, so I have hope that this is learnable, and not too much of a footgun.
Before merging the remote branch, do a rebase. This may fail and prompt conflict resolution, but that seems the canonical outcome for the common use case "interactive make git pull-xxx" with master out-of-sync.
Retire pkg-manager.sh and replace it by the cleaner "jw-pkg.sh distro" command, essentially providing the same functionality and nearly the same command-line interface.
400 LOC more. That's what the move from a shell script to the more maintainable Python versions costs. Still a good idea, and the enhanced extensibility might pay off in terms of LOC with other shell scripts in the future.
Reorganize the Python module structure. Placing the command classes under jw.cmds.projects instead of jw.build.cmds will allow to add a nested command structure, with the current commands, being mostly related to building software, found below a "projects" toplevel command.
Other conceivable commands could be "package" for packaging, or "distro" for commands wrapping the distribution's package manager.
$(TOPDIR)/scripts contains a symbolic link jw -> ../src/python/jw, to allow jw-python.py access to the in-repo module jw.pkg. Should be fine now even on Windows, OTOH, it's also solvable via __init__.py, so do that.
To support minimal environments, notably minimal Docker containers which don't have /usr/bin/sudo by the time pkg-manager.sh is invoked (possibly to install sudo), support running all commands without invoking sudo first. Of course this only works if invoked as root.
Note that this is still somewhat hacky, command-line parsing needs to be cleaned up.
"git fetch $remote $fromref:$toref" fails if the $fromref is behind $toref.
Unrolling the syntax into "git fetch" followed by "git merge --ff-only $remote/$fromref $toref" is accepted, though, and saves some otherwise necessary case distinction code around it.
jw-build doesn't stop at building software, packaging it afterwards is also a core feature, so this commit gives the package a better name.
The commit replaces strings s/jw-build/jw-pkg/ in text files and file names. Fallout to the functionality is fixed, variable names are left as they are, though. To be adjusted by later commits.
--login is not understood by pgit.sh push. Solve that by allowing all commands a --login option. This addresses our use case, but isn't ideal of course. Will be finally fixed by moving pgit.sh's functionality into Python code.
pgit.sh clone --login <username> fails to insert said username into a remotes url while adding it: Of ssh://<username>@git.janware.com/srv/git, only ssh://@git.janware.com/srv/git makes it into the config. Fix that.
