Package Details: ctags-git 6.0.0.r525.gc480d71e1-1

Git Clone URL: https://aur.archlinux.org/ctags-git.git (read-only, click to copy)
Package Base: ctags-git
Description: Generates an index file of language objects found in source files
Upstream URL: https://ctags.io/
Licenses: GPL
Conflicts: ctags
Provides: ctags
Submitter: polyzen
Maintainer: Chocobo1
Last Packager: Chocobo1
Votes: 60
Popularity: 0.000000
First Submitted: 2021-01-14 17:35 (UTC)
Last Updated: 2023-11-23 19:27 (UTC)

Required by (59)

Sources (1)

Latest Comments

« First ‹ Previous 1 2 3 4 5 6 Next › Last »

SolarAquarion commented on 2020-10-13 02:28 (UTC)

there's only one tag https://github.com/universal-ctags/ctags/releases/tag/p5.9.20201011.0

pepper_chico commented on 2020-10-13 01:52 (UTC) (edited on 2020-10-13 01:53 (UTC) by pepper_chico)

@polyzen just not sure what's happening, except that the git describe/sed result in your machine is one character longer than ours.

polyzen commented on 2020-10-13 00:38 (UTC) 

If tag contains a prefix, like v or project name then it should be cut off:

    pkgver() {
      cd "$pkgname"
      # cutting off 'foo-' prefix that presents in the git tag
      git describe --long | sed 's/^foo-//;s/\([^-]*-g\)/r\1/;s/-/./g'
    }

    6.1.r3.gd77e105

pepper_chico commented on 2020-10-12 18:20 (UTC) (edited on 2020-10-12 18:21 (UTC) by pepper_chico)

Same issue here:

ctags on HEAD (01b9fc8)
❯ git describe --long --tags | sed 's/^p//;s/\([^-]*-g\)/r\1/;s/-/./g'
5.9.20201011.0.r12.g01b9fc86

It seems this git describe/sed command is giving one character longer in the package maintainer's machine. Notice that this sed regex isn't listed in the recommended pkgver guidelines: https://wiki.archlinux.org/index.php/VCS_package_guidelines.

kseistrup commented on 2020-10-12 09:49 (UTC)

That's interesting, because if I do the “git describe” dance I get

5.9.20201011.0.r12.g01b9fc86

and

$ ls -l *.zst
.rw-r--r-- 589.583 kas 12 Oct 11:44 universal-ctags-git-5.9.20201011.0.r12.g01b9fc86-1-x86_64.pkg.tar.zst

which is also the package name I end up with if I install it with yay.

I wonder what is happening here. Any ideas?

polyzen commented on 2020-10-12 09:17 (UTC) 

~ > cd pkg/build/universal-ctags-git/src/ctags
~/p/b/un/s/ctags > git describe --long --tags | sed 's/^p//;s/\([^-]*-g\)/r\1/;s/-/./g'
5.9.20201011.0.r12.g01b9fc866

==> Finished making: universal-ctags-git 5.9.20201011.0.r12.g01b9fc866-1 (Mon 12 Oct 2020 05:12:13 AM EDT)

Producing universal-ctags-git-5.9.20201011.0.r12.g01b9fc866-1-x86_64.pkg.tar.zst

kseistrup commented on 2020-10-12 08:06 (UTC)

There's a ‘6’ too much at the end of

pkgver=5.9.20201011.0.r12.g01b9fc866

When the package is built, you end up with

universal-ctags-git-5.9.20201011.0.r12.g01b9fc86-1

and next time you run yay it thinks there's a newer version.

Maelan commented on 2018-08-25 03:18 (UTC)

Does not build for me. Compiling the man pages fail, even with python{,2}-docutils installed:

  CCLD     readtags
RST2MAN   man/ctags.1
Traceback (most recent call last):
  File "/bin/rst2man", line 21, in <module>
    from docutils.core import publish_cmdline, default_description
ModuleNotFoundError: No module named 'docutils'
make[1]: *** [Makefile:7357: man/ctags.1] Error 1

lenormf commented on 2018-07-27 13:59 (UTC)

I think bash should be removed from the list of runtime dependencies.

cbb commented on 2018-06-22 03:55 (UTC) (edited on 2018-06-22 05:41 (UTC) by cbb)

@J5lx Yeah, I understand your line of thinking. My line of thinking is rather that AUR is an absolute wild west anyway, so a little bit of ill-considered patching can't hurt the standards of quality much.

I've pushed uctags-git for anyone else who may want a universal-ctags package that can be installed in parallel with ctags.

« First ‹ Previous 1 2 3 4 5 6 Next › Last »