| Commit message (Collapse) | Author | Age |
|---|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Take for example a repository name like: `some/organisation/project_a/team_c/repo_b`.
Setting the setting to `-1` means that cgit traverses the path from left
to right, until it has found 1 element (and `section-from-path` (or n
for short)is 0, because n is incremented after each iteration). E.g. :
~ [n=-1]
starting point: `some/organisation/project_a/team_c/repo_b`
~ [n=0]
after the first iteration: `some/organisation/project_a/team_c/repo_b`
Now `some/organisation/project_a/team_c` becomes the section, whilst
`repo_b` becomes the repo name.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Cgit effectively splits the repo path on '/' and then takes
`section-from-path` segments, which form the section. A negative value
here results in cgit traversing the path from left to right instead of
right to left. Beware that cgit only sets the section, if the path
contains `section-from-path` or more slashes in it (thus rendering this
setting defunct with the previous value of 1000).
There seems to be no way to tell cgit to always use all components up-to
the second to last for the section name, thus requiring all projects
that need a longer than 1 section length to set the `cgit.section`
git config variable via gitolite.
|
| |
|
|
|
|
|
|
| |
The previously set variable is only used in the gitolite.conf file for
the `config` specifications on each repo. We can't use that because we
use "wild-repos". Thus we need to add the `user-configs` option to each
repo, allow users to change the git settings specified there with a
simple `ssh git@git.vhack.eu config <repo> --set cgit.owner <name>`.
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
| |
A lot of the added settings here have been tested. They will get
get tests to ensure they stay working, in later commits.
|
| | |
|
| | |
|
| |
|
