This patch sets the directory prefix in archives to be the filename,
excluding the suffix (.tar.gz, .tar.bz2 etc).
The patch also removes the prefix parameter in cgit_print_snapshot()
as the prefix might differ.
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
When the agefile was empty the old code would happily reuse the static
buffer filled by a previous call to read_agefile().
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
The system shell (/bin/sh) on Ubuntu is dash, which aims to be a
POSIX standard shell. In particular, dash does not implement any
of the common extensions to the standard that, say, bash and ksh
do.
Replace some non-POSIX constructs in setup.sh with more portable
and mundane code.
Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk>
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
The type used to declare the st_size field of a 'struct stat' can
be a 32- or 64-bit sized type, which can vary from one platform to
another, or even from one compilation to another. In particular,
on linux, if you include the following define:
#define _FILE_OFFSET_BITS 64
prior to including certain system header files, then the type used
for the st_size field will be __off64_t, otherwise it will be an
__off_t. Note that the above define is included at the top of
git-compat-util.h.
In cache.c, the "%zd" format specifier expects a "signed size_t",
another type which can vary, when an __off64_t or a __off_t is
provided. To supress the warning, use the PRIuMAX format specifier
and cast the st_size field to uintmax_t. This should work an any
platform for which git currently compiles.
In ui-plain.c, the size parameter of sha1_object_info() and
read_sha1_file() is defined to be "unsigned long *" not "size_t *".
So, to supress the warning, simply declare size with the correct type.
Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk>
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
When downloading a snapshot, the snapshot name will often contain the repo
name combined with a tag. This patch tries to exploit this so that the
correct revision is downloaded even if no specific revision is specified.
PS: this only occurs if neither 'h' nor 'id' is specified in the query-
string.
PPS: this also fixes a bug which occurs when trying to download a filename
with an unsupported suffix: it used to try to print an error message to
the user but failed since it didn't prepare the output properly.
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
This flag is set when no HEAD is specified in the querystring. Currently
it has no users, but it will be used by ui-snapshot to invoke a DWIM-mode
where the revision is extracted from the snapshot name.
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
When a repo uses an url with e.g. '#' or '?' characters this needs to be
properly escaped when used as action in a form tag.
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
The new file describes all cgitrc options in a more structured manner then
the cgitrc example file and it might also work as the source for a cgitrc
man page.
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
This makes is possible to use cgit with repository urls containing special
url characters like '#' and '?'.
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
The struct member was introduces in git commit d53fe8187c38, but the cgit
testsuite failed to detect that cgit always generated archives without
prefixes, i.e. the result from cgit_repobasename was ignored.
This fixes the bug and the testsuite.
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
The link-generating functions are updated to use the new html_url_arg
function, thereby fixing links to strange repos, branches and files.
Also, the test-suite is updated to verify some cases of strange urls.
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
When printing a path-filtered diff it wasn't obvious how to get back to
the full diff (clicking the 'diff' tab would do this). Making the diffstat
heading into a link seems to improve the usability.
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
The links in the diffstat is supposed to work as a filter for the diff,
but this only worked when a single rev was supplied, i.e. the filtered
diff was always against the parent of the specified rev.
With this patch it is now possible to use the diffstat as a 'filter menu'
for urls like http://hjemli.net/git/cgit/diff/?id=v0.7.2&id2=v0.7.1
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
This will allow for creating static builds which is useful for chrooted
environments.
Signed-off-by: Harley Laue <losinggeneration@gmail.com>
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
If a repo url is specified but no exact match is found in the list of
repos the url will now be used as a prefix-filter.
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
This option makes cgit scan a directory tree looking for git repositories,
generating suitable definitions for a cgitrc file on stdout.
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
This commit makes cgit use the cgi variables SCRIPT_NAME and PATH_INFO
when virtual-root is unspecified in cgitrc and no url-parameter is
specified on the querystring. This has two nice effects:
* Virtual urls works out of the box, no more need for rewrite-rules in httpd.
* Virtual urls with special querystring characters are handled correctly.
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
When parsing commits and tags cgit made too many assumptions about the
formatting of said objects. This patch tries to make the code be more
prepared to handle 'malformed' objects.
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
I run an instance of lighttpd for cgit behind nginx (nginx
doesn't execute CGI). So the port (SERVER_PORT=33333) that
lighttpd runs on sends to cgit is different from the standard
port 80 that public clients connect to (via nginx).
This was causing the Atom feed URL to show the private port
number that lighttpd was running on.
Since the HTTP/1.1 "Host" header includes the port number if
running on a non-standard port, it allows non-client-facing HTTP
servers to transparently generate public URLs that clients can
see.
So use the "Host" header if it is available and fall back to
SERVER_NAME/SERVER_PORT for some clients that don't set
HTTP_HOST.
Signed-off-by: Eric Wong <normalperson@yhbt.net>
The callback from read_tree_recursive just needs to check the type of
each tree entry; if it's a dir we want to continue scanning, if it's a
regular file we'll assume it's the one we requested.
And while at it, remove some stray fprintfs.
Signed-off-by: Lars Hjemli <hjemli@gmail.com>
Natanael Copa