Dan Brook [Mon, 1 Apr 2013 21:53:59 +0000]
Bump version to 0.004002 and update Changes.
Dan Brook [Mon, 1 Apr 2013 21:44:36 +0000]
Update gitalist_app.psgi with working copy.
As t0m added and I've now put into place.
Dan Brook [Mon, 1 Apr 2013 21:39:29 +0000]
Merge branch 'branch-selector-fix'
* branch-selector-fix:
The branch selector now sends you to the correct url
Dennis Kaarsemaker [Fri, 3 Aug 2012 17:53:02 +0000]
The branch selector now sends you to the correct url
It used to not escape the branch name, so branches like
seveas/branch_selector_fix were not really reachable via the selector
Dan Brook [Mon, 1 Apr 2013 21:28:44 +0000]
Merge remote branch 'seveas/seveas/syntax_highlight'
* seveas/seveas/syntax_highlight:
Have linenumbers in blob displays
Syntax higlighting improvements
Dan Brook [Mon, 1 Apr 2013 21:20:24 +0000]
Merge remote branch 'seveas/seveas/tree_speedup'
* seveas/seveas/tree_speedup:
Massive tree action speedup
Dan Brook [Mon, 1 Apr 2013 17:46:16 +0000]
Fix the bare.git test repo.
In it's previous state recent git versions didn't consider it a git repo which
resulted in noisy (but passing) tests. Now the tests are less noisy and still
passing.
Dan Brook [Fri, 15 Mar 2013 21:32:28 +0000]
Bump version to 0.004001 and update Changes.
Dan Brook [Fri, 15 Mar 2013 21:32:13 +0000]
Add dep that was previously implicit.
Not sure what used to pull in this dep either but now the app should
function and pass its tests again.
Dan Brook [Thu, 18 Oct 2012 17:11:47 +0000]
Make test for diff output more flexible.
The diff output changed (for the better) in more recent git versions
so have the relevant test deal with it.
Dan Brook [Tue, 21 Aug 2012 19:28:46 +0000]
Merge pull request #43 from seveas/seveas/tag_fix
Stop calling tags branches
Dennis Kaarsemaker [Fri, 3 Aug 2012 09:55:49 +0000]
Massive tree action speedup
The tree action used to use $tree_obj->tree to get all tree entries.
This is ridiculously slow as it loads all blobs via git cat-file.
Instead, expose directory_entries from the underlying Git::PurePerl
object and use it instead.
This also required moving _mode_str (now mode_string for consistency) to
Gitalist::Utils so it could be used in the template.
Dennis Kaarsemaker [Sun, 29 Jul 2012 17:08:26 +0000]
Stop breaking trees when submodules are encountered
For projects that contain submodules, ls-tree also lists the commits of
these, like:
dennis@lightning:~/code/Diamond$ git ls-tree HEAD
100644 blob
76548fb89c2c220d43759b40848600a16cfe65ea .gitignore
100644 blob
c2c5e7f9ad41d2a6bf49dcfe2a18c67f1b7264d3 .gitmodules
100644 blob
159d72c1edf6e0e227b0c1dda86aa08c307468df LICENSE
100644 blob
3ce75ca4a98adbc6b1cdf68797b6b9328bf112cc MANIFEST.in
100644 blob
588e2ba64d1968ecf949325cab4b1bfaf60caf75 Makefile
100644 blob
974a51cd30c73dc4fa5083252902b5596cbdeec8 README.md
040000 tree
e69b70cc3f14491f167b3edd1561279344267b89 bin
040000 tree
10b76bbf7f6900e53916bf52c651b048569de46c conf
040000 tree
90be53cbef1f3b6167a2b24f59365e94ce6cd989 debian
160000 commit
69c721d6d71b1a471e945bf680773467ee2c551e docs
040000 tree
daa6b06e1917938c2e939bbcd1d7b451e290b680 rpm
100644 blob
9046e2354860cb14b2ade3159c20e5245013a238 setup.cfg
100755 blob
0a4a2748cc2bb61685d5535d5a6557854153e7c0 setup.py
040000 tree
4b95cf8a722b456382abbd025476c87f87a97043 src
100644 blob
a375d7dbdb98e15dc5021c0c5e4cb9757fcc0273 test.py
100644 blob
13f3e59423a6f49d89527b9e9a2050fb79d0025f test.watchr
Gitalist would choke on those because these commits are commits from the
submodule's repository, not itself. So we should ignore them.
Dennis Kaarsemaker [Fri, 20 Jul 2012 20:30:20 +0000]
Make Git/Repository.pm correctly handle utf-8 encoded .git/description files
Dennis Kaarsemaker [Sun, 29 Jul 2012 18:12:02 +0000]
Have linenumbers in blob displays
Steal some code from the blob template and adapt the syntax highlighting
code to cope with it.
Dennis Kaarsemaker [Sun, 29 Jul 2012 16:53:51 +0000]
Syntax higlighting improvements
- Support all languages Syntax::Highlight::Engine::Kate can detect
- Detect filetype based on #! lines in scripts
- Consider .t files to be perl code
Dennis Kaarsemaker [Sun, 29 Jul 2012 13:28:19 +0000]
Stop calling tags branches
In the repository overview tags were called branches, that's not right :)
Tomas Doran [Sat, 21 Jul 2012 13:56:09 +0000]
Controller::ActionRole is dead
Tomas Doran [Thu, 14 Jun 2012 03:34:24 +0000]
Add gitalist.psgi
Dan Brook [Sun, 18 Mar 2012 22:01:33 +0000]
Bump version to 0.003009 and update Changes.
Dan Brook [Sun, 18 Mar 2012 22:02:59 +0000]
Drop MooseX::Types::ISO8601 dependency.
Dan Brook [Sun, 18 Mar 2012 21:51:30 +0000]
Cope with changes in Catalyst 5.90007 in t/model_collectionofrepos.t.
Dan Brook [Fri, 9 Mar 2012 22:23:01 +0000]
Use MooseX::Types::ISO8601 coercion behaviour without the coercion.
As of 0.09 the type constraints from MooseX::Types::ISO8601 are stricter (and
more correct AIUI) so we just use the coercion code without attempting a
coercion directly.
Dan Brook [Sun, 29 Jan 2012 00:26:41 +0000]
Drop redundant requires in G::G::COR::Role::Context.
Not much point in requiring methods be implemented if the role already
provides concrete implementations.
Dan Brook [Sun, 29 Jan 2012 00:02:11 +0000]
Bump version number to 0.003008 and add to Changes.
While 0.003007 hit the CPAN it was missing Changes hence 0.003008.
Dan Brook [Sat, 28 Jan 2012 23:16:01 +0000]
Align Vhost code with reality.
Now that we potentially have a different COR per hit (depending on config) the
COR model has changed a lot. Now its COMPONENT will return a G::G::COR-based
class which in turn provides an ACCEPT_CONTEXT method that should return a
further G::G::COR object based on the request. In the simple case, as with the
default config, the same G::G::COR that was built for the COMPONENT is also
returned by ACCEPT_CONTEXT. With the Vhost COR a different G::G::COR will be
returned depending on the related configuration.
Dan Brook [Mon, 23 Jan 2012 23:39:14 +0000]
Provide required method to test model.
Now its test no longer blows up as it is a valid COR again.
Dan Brook [Mon, 23 Jan 2012 23:23:43 +0000]
Changed repos_dir to repo_dir in the COR model.
Unfortunately the name "repo_dir" is in the config files and the rest
of the code so the more sensible "repos_dir" (a directory containing
repositories) must step aside for now. Also the majority tests are
passing again which makes me happy.
Tomas Doran [Tue, 22 Nov 2011 21:44:25 +0000]
Revert "Change to visit"
This reverts commit
0321f9d9dffcc9f595caba3d1dc2201dc6351c54.
Tomas Doran [Tue, 22 Nov 2011 21:42:30 +0000]
A basic --repo_dir now works again
Tomas Doran [Tue, 22 Nov 2011 21:36:37 +0000]
Better debugging
Tomas Doran [Tue, 22 Nov 2011 21:30:42 +0000]
Make die string in controller much better
Tomas Doran [Mon, 14 Nov 2011 23:18:42 +0000]
Get sane debugging info for the setup of everything
Tomas Doran [Mon, 14 Nov 2011 23:11:52 +0000]
Change to visit
Tomas Doran [Tue, 25 Oct 2011 02:05:37 +0000]
Use new Catalyst features
Gun.io Whitespace Robot [Fri, 28 Oct 2011 00:20:42 +0000]
Remove whitespace [Gun.io WhitespaceBot]
Conflicts:
t/model_collectionofrepos.t
Tomas Doran [Mon, 14 Nov 2011 09:00:15 +0000]
Don't blow up so much, but still not working
Tomas Doran [Sun, 13 Nov 2011 23:18:29 +0000]
Add tests
Tomas Doran [Sun, 13 Nov 2011 23:17:29 +0000]
Get the vhost based repository displatcher for shadowcat working, I think
Tomas Doran [Sun, 13 Nov 2011 22:16:39 +0000]
Pass the vhost in as a parameter
Tomas Doran [Wed, 20 Jul 2011 11:17:27 +0000]
Vhost based repository chooser.
Dan Brook [Sun, 13 Nov 2011 22:40:53 +0000]
Ressurect repo_dir as it is still in use.
May be worth deprecating or at least documenting clearly. It's usually
not healthy to have singular & plural options do slightly different things.
Dan Brook [Sun, 13 Nov 2011 22:36:19 +0000]
Drop noisy debug call.
Tomas Doran [Sun, 13 Nov 2011 14:11:45 +0000]
Make the tests pass again
Tomas Doran [Sun, 13 Nov 2011 13:03:54 +0000]
Better document and clean up the factory for collections of repos
Tomas Doran [Sun, 13 Nov 2011 12:48:29 +0000]
Fix README
Tomas Doran [Sun, 13 Nov 2011 12:44:18 +0000]
Merge branch 'git_in_path_at_startup'
* git_in_path_at_startup:
Make sure we find git at app start time
Tomas Doran [Sun, 13 Nov 2011 12:43:58 +0000]
Merge branch 'run_options'
* run_options:
Less madness with the environment. Not tested
Unfuck DateTime type with newer MX::Types
Fix trailing hatespace
Skip MYMETA
Allow the model to be user defined.
Use appropriate prototype depending on perl version in test.
Fix a few test annoyances.
Tomas Doran [Tue, 25 Oct 2011 02:18:50 +0000]
Unfuck DateTime type with newer MX::Types
Tomas Doran [Tue, 25 Oct 2011 02:06:28 +0000]
Fix trailing hatespace
Tomas Doran [Tue, 25 Oct 2011 02:05:54 +0000]
Skip MYMETA
Dan Brook [Sat, 29 Oct 2011 22:42:38 +0000]
Allow the model to be user defined.
Given --model-class that class will be used instead of the whatever Gitalist
reasons from the model options. If --model-args is defined then they will be
used when constructing the model object in addition to the default arguments.
Dan Brook [Sun, 30 Oct 2011 22:22:52 +0000]
Use appropriate prototype depending on perl version in test.
The _ prototype didn't exist before-5.10 and since Gitalist runs under 5.8.9
at least it seemed only proper to have this test not blow up horribly.
Dan Brook [Sun, 30 Oct 2011 20:19:46 +0000]
Fix a few test annoyances.
Tomas Doran [Sat, 12 Nov 2011 01:04:47 +0000]
Make sure we find git at app start time
Tomas Doran [Wed, 9 Nov 2011 23:39:58 +0000]
Less madness with the environment. Not tested
Tomas Doran [Wed, 9 Nov 2011 23:30:24 +0000]
Merge remote branch 'origin/model-from-args'
* origin/model-from-args:
Allow the model to be user defined.
Use appropriate prototype depending on perl version in test.
Fix a few test annoyances.
Tomas Doran [Tue, 25 Oct 2011 02:18:50 +0000]
Unfuck DateTime type with newer MX::Types
Tomas Doran [Tue, 25 Oct 2011 02:06:28 +0000]
Fix trailing hatespace
Tomas Doran [Tue, 25 Oct 2011 02:05:54 +0000]
Skip MYMETA
Dan Brook [Sat, 29 Oct 2011 22:42:38 +0000]
Allow the model to be user defined.
Given --model-class that class will be used instead of the whatever Gitalist
reasons from the model options. If --model-args is defined then they will be
used when constructing the model object in addition to the default arguments.
Dan Brook [Sun, 30 Oct 2011 22:22:52 +0000]
Use appropriate prototype depending on perl version in test.
The _ prototype didn't exist before-5.10 and since Gitalist runs under 5.8.9
at least it seemed only proper to have this test not blow up horribly.
Dan Brook [Sun, 30 Oct 2011 20:19:46 +0000]
Fix a few test annoyances.
Dan Brook [Wed, 19 Oct 2011 18:14:27 +0000]
Bump version to 0.003006, update Changes and README.
Also update MANIFEST.SKIP to ignore more editor droppings.
Dan Brook [Wed, 19 Oct 2011 18:13:18 +0000]
Fix failures in recursive repo tests.
Since a new repo was added the tests needed updating.
Tomas Doran [Thu, 15 Sep 2011 12:03:29 +0000]
Fix deprecated qw
Tomas Doran [Thu, 15 Sep 2011 12:01:19 +0000]
This setting isn't actually used
Tomas Doran [Tue, 11 Jan 2011 12:46:56 +0000]
Update config to make this more obvious
Tomas Doran [Thu, 19 Aug 2010 08:34:04 +0000]
Make another bare repos to testing recursive
Dan Brook [Mon, 17 Oct 2011 13:15:05 +0000]
Merge pull request #26 from gisle/patch-1
Use of qw(...) as parentheses is deprecated
Gisle Aas [Sun, 16 Oct 2011 15:17:21 +0000]
Avoid "Use of qw(...) as parentheses is deprecated" warning with perl-5.14
Dan Brook [Wed, 13 Jul 2011 17:54:45 +0000]
Bump version to 0.003005, update Changes and README.
Dan Brook [Wed, 13 Jul 2011 17:27:42 +0000]
Provide option to disable gravatars.
Apparently this is of concern to folk who fear their stuff being leaked in
browser referer headers. Fair enough. Also centralized gravatar template bits
while I was at it for simplicity's sake.
Dan Brook [Wed, 13 Jul 2011 13:17:02 +0000]
Drop the extraneous root/static/i/icons/Thumbs.db
Dan Brook [Sat, 11 Jun 2011 21:48:15 +0000]
Update Changes with latest fixes.
Christian Walde [Sun, 29 May 2011 21:37:21 +0000]
Win32 does not have a proper owner for repos, so to keep the tests untouched, we substitute the expected one
Christian Walde [Sun, 29 May 2011 21:36:44 +0000]
Win32 can have git binaries with extensions, and non-/ path separators
Christian Walde [Sun, 29 May 2011 21:34:58 +0000]
paths need to be nativized uniformly
Christian Walde [Sun, 29 May 2011 21:34:10 +0000]
Win32 does not have reliable symlinks
Christian Walde [Sun, 29 May 2011 21:32:49 +0000]
I18N::Langinfo is not available on Win32
Dan Brook [Mon, 6 Jun 2011 17:49:00 +0000]
Add Apache specific section to docs.
The AllowEncodedSlashes was lost in the FCGI docs and not specific to
FCGI. Also autarch kindly provided a bit of config to have Apache
serve the static content.
Dan Brook [Thu, 7 Jul 2011 17:03:46 +0000]
Merge pull request #21 from wchristian/css_fix
css fix
Dan Brook [Thu, 7 Jul 2011 17:03:27 +0000]
Merge pull request #23 from hggh/mainifest_fix
added missing MANIFEST.SKIP so file will not be installed
Dan Brook [Thu, 7 Jul 2011 17:03:03 +0000]
Merge pull request #22 from hggh/pod_fix
fixing inline pod documentation for the debian package. (missing whatis e
Jonas Genannt [Wed, 6 Jul 2011 20:18:37 +0000]
added missing MANIFEST.SKIP so file will not be installed
Jonas Genannt [Wed, 6 Jul 2011 19:35:39 +0000]
fixing inline pod documentation for the debian package. (missing whatis entry)
Dan Brook [Tue, 7 Jun 2011 20:23:14 +0000]
Bump version to 0.003004, update Changes.
Dan Brook [Tue, 7 Jun 2011 20:16:45 +0000]
Fix dependency typo in Makefile.PL.
Dan Brook [Mon, 6 Jun 2011 15:06:07 +0000]
Bump version to 0.003003, update Changes.
Dan Brook [Mon, 6 Jun 2011 11:29:48 +0000]
Add editor temp files to MANIFEST.SKIP.
Dan Brook [Mon, 6 Jun 2011 11:15:30 +0000]
Change default local config to not search recursively.
This is non-optimal default behaviour, if repo_dir isn't just a bunch
of repos this could potentially slow things down a lot.
Dan Brook [Sun, 5 Jun 2011 21:20:06 +0000]
Don't attempt to syntax highlight anything big.
Basically Syntax::Highlight::Kate is slow and trying to prettify
anything relative big s really slow. So just don't even try. Sad face.
Dan Brook [Sun, 5 Jun 2011 21:18:22 +0000]
Fix for infinite loop bug in blame.
Previously blame would get stuck in an infinite loop when processing
an empty file. Conceivably this a bug in git-blame as file content should
begin with a tab, even if it's empty.
Also adding a commit to repo1 involves fiddling lots tests so I won't
be doing that again.
Thanks to autarch for flagging this up in the first place!
Christian Walde [Sun, 5 Jun 2011 15:27:25 +0000]
Enabled content type css for css files served by template, previous implementation was non-functional, but nobody noticed since some browsers paper over such messes.
Dan Brook [Sun, 29 May 2011 17:41:15 +0000]
Give favicon.ico proper transparency.
Previously the icon only looked sane on a white background, anything else and
you could see ugly white splotches. Now the background is fully transparent.
Also cleaned up unused favicon.ico paths.
Christian Walde [Sun, 29 May 2011 11:26:45 +0000]
Changed the owner check to only check that *something came back.
Since the owner name is retrieved from the system user table and thus can
end up being anything.
Dan Brook [Mon, 23 May 2011 14:09:20 +0000]
Bump version to 0.003002, update Changes and README.
Looks like the latest pod2text now includes the links from the POD in
the resulting text, which is nice.
Dan Brook [Sun, 22 May 2011 17:07:23 +0000]
Updated Changes and made CPAN::Changes compliant.
Will also integrates Test::CPAN::Changes when it is AuthorRequires friendly.
Dan Brook [Sun, 22 May 2011 16:22:01 +0000]
Provide .psgi for Plack goodness.
The .psgi was written by franky (Francoise Dehinbo) and gist-ed by
Ranguard (Leo Lapworth). I've tweaked it to play nice with checkouts
and "make install".
Dan Brook [Sat, 7 May 2011 17:02:57 +0000]
Expose serialization via JSON.
The last of Tom's work is now integrated with all URLs that can will
respond with JSON. Unfortuantely browsers (I'm looking at you Chrome)
will cache per URL rather than URL + Content-Type so if one requests
JSON for /repo/sha1/commit then navigates to said commit the browser
will served the cached JSON rather than HTML. As such JSON is also
exposed with the .json extension allowing JS to talk JSON without
messing up certain browsers' (i.e Chrome's) caching.