On Fri Mar 17, 2017 at 07:43:10PM +0100, Jeremie Courreges-Anglas wrote: > Alexandr Shadchin <[email protected]> writes: > > > On Wed, Mar 15, 2017 at 1:34 PM, Jeremie Courreges-Anglas <[email protected]> > > wrote: > > > >> > >> (Cc'ing Alexandr, the maintainer of devel/ectags.) > >> > >> Rafael Sadowski <[email protected]> writes: > >> > >> > Hi all, > >> > > >> > please find attach a maintained ctags implementation. > >> > > >> > $ pkg/DESCR > >> > universal-ctags has the objective of continuing the development from what > >> > existed in the Sourceforge area. The goal of the project is preparing and > >> > maintaining common/unified space where people interested in making ctags > >> better > >> > can work together. > >> > >> There are a few things that caught my eye, but first: shouldn't this > >> replace devel/ectags? (IIUC ectags is exuberant ctags, the sourceforge > >> project mentioned in DESCR, with no release since 2009). > >> > >> -- > >> jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE > >> > > > > It would be nice. ectags is old and contain bugs. > > upstream had a discussion about this: > > https://github.com/universal-ctags/ctags/issues/446 > > The tags file format seems compatible with ectags, but I don't know what > are the differences between ectags and uctags. I'll leave the question > to ectags users. > > The updated port contains the following changes: > - move to devel, like ectags. I don't think it fits in sysutils or > textproc
Okay for me.
> - afaik CONFIGURE_ARGS --prefix=${LOCALBASE} --sysconfdir=${PREFIX}/etc
> were no-op. Is there a reason to specify them?
Maybe from an old version. universal-ctags has been living for a long
time in openbsd-wip.
> - embed the git commit in the build (after all, it's not a release, and
> it could help upstream bug reports)
Nice!
> - enable regress tests (one test is failing)
... also with me.
> - don't clobber LDFLAGS
>
> The regress tests show that multibyte support is not built. (by
> default?) Should it be enabled? In a FLAVOR maybe? No multibyte = no
> need for iconv = no need to tweak CPPFLAGS/LDFLAGS.
>
Nice idea, please find attached an extended version of your last
tarball:
- add multibyte FLAVOR (If you don't like the name, feel free to rename
it)
- add gsort in pre-test
Runtime test on amd64 in:
- /usr/ports/pobj/firefox-52.0/firefox-52.0
- /usr/ports/pobj/chromium-57.0.2987.110/chromium-57.0.2987.110
> The patch for regress tests could probably be pushed upstream.
>
I'll push it upstream.
As always thank you jca!
Rafael Sadowksi
universal-ctags.tar.gz
Description: application/tar-gz
