Summary: | Invalid repoman warning about quoting with custom ${dir} unquoted and used in e.g. exeinto | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Samuli Suominen (RETIRED) <ssuominen> |
Component: | Current packages | Assignee: | Portage team <dev-portage> |
Status: | RESOLVED FIXED | ||
Severity: | trivial | Keywords: | InVCS |
Priority: | High | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Bug Depends on: | |||
Bug Blocks: | 210077, 288499 | ||
Attachments: | funky ebuild |
Description
Samuli Suominen (RETIRED)
![]() Created attachment 195204 [details]
funky ebuild
I'll likely end up closing this after sorting it out in tree, let's see..
Renamed the ${dir} to ${dest} and declared it local in src_install() and the warnings are gone. Beats me?! pym/repoman/checks.py contains: class EbuildQuote(LineCheck): """Ensure ebuilds have valid quoting around things like D,FILESDIR, etc...""" repoman_check_name = 'ebuild.minorsyn' _message_commands = ["die", "echo", "eerror", "einfo", "elog", "eqawarn", "ewarn"] _message_re = re.compile(r'\s(' + "|".join(_message_commands) + \ r')\s+"[^"]*"\s*$') _ignored_commands = ["local", "export"] + _message_commands ignore_line = re.compile(r'(^$)|(^\s*#.*)|(^\s*\w+=.*)' + \ r'|(^\s*(' + "|".join(_ignored_commands) + r')\s+)') var_names = ["D", "DISTDIR", "FILESDIR", "S", "T", "ROOT", "WORKDIR"] # variables for games.eclass var_names += ["Ddir", "dir", "GAMES_PREFIX_OPT", "GAMES_DATADIR", "GAMES_DATADIR_BASE", "GAMES_SYSCONFDIR", "GAMES_STATEDIR", "GAMES_LOGDIR", "GAMES_BINDIR"] (In reply to comment #3) > pym/repoman/checks.py contains: > var_names += ["Ddir", "dir", I've removed locally the dir and Ddir from the checks.py and the problem is gone, do we really need to check quoting for such generic variables? (In reply to comment #4) > do we really need to check quoting for such generic variables? No. I've removed that one in svn r13659. (In reply to comment #5) > (In reply to comment #4) > > do we really need to check quoting for such generic variables? > > No. I've removed that one in svn r13659. > Thanks This is fixed in 2.2_rc34. This is fixed in 2.1.7. |