If a caller asks the http_get_* functions to go to a
particular URL and we end up elsewhere due to a redirect,
the effective_url field can tell us where we went.
It would be nice to remember this redirect and short-cut
further requests for two reasons:
1. It's more efficient. Otherwise we spend an extra http
round-trip to the server for each subsequent request,
just to get redirected.
2. If we end up with an http 401 and are going to ask for
credentials, it is to feed them to the redirect target.
If the redirect is an http->https upgrade, this means
our credentials may be provided on the http leg, just
to end up redirected to https. And if the redirect
crosses server boundaries, then curl will drop the
credentials entirely as it follows the redirect.
However, it, it is not enough to simply record the effective
URL we saw and use that for subsequent requests. We were
originally fed a "base" url like:
http://example.com/foo.git
and we want to figure out what the new base is, even though
the URLs we see may be:
original: http://example.com/foo.git/info/refs
effective: http://example.com/bar.git/info/refs
Subsequent requests will not be for "info/refs", but for
other paths relative to the base. We must ask the caller to
pass in the original base, and we must pass the redirected
base back to the caller (so that it can generate more URLs
from it). Furthermore, we need to feed the new base to the
credential code, so that requests to credential helpers (or
to the user) match the URL we will be requesting.
This patch teaches http_request_reauth to do this munging.
Since it is the caller who cares about making more URLs, it
seems at first glance that callers could simply check
effective_url themselves and handle it. However, since we
need to update the credential struct before the second
re-auth request, we have to do it inside http_request_reauth.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
When we ask curl to access a URL, it may follow one or more
redirects to reach the final location. We have no idea
this has happened, as curl takes care of the details and
simply returns the final content to us.
The final URL that we ended up with can be accessed via
CURLINFO_EFFECTIVE_URL. Let's make that optionally available
to callers of http_get_*, so that they can make further
decisions based on the redirection.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
When we are handling a curl response code in http_request or
in the remote-curl RPC code, we use the handle_curl_result
helper to translate curl's response into an easy-to-use
code. When we see an HTTP 401, we do one of two things:
1. If we already had a filled-in credential, we mark it as
rejected, and then return HTTP_NOAUTH to indicate to
the caller that we failed.
2. If we didn't, then we ask for a new credential and tell
the caller HTTP_REAUTH to indicate that they may want
to try again.
Rejecting in the first case makes sense; it is the natural
result of the request we just made. However, prompting for
more credentials in the second step does not always make
sense. We do not know for sure that the caller is going to
make a second request, and nor are we sure that it will be
to the same URL. Logically, the prompt belongs not to the
request we just finished, but to the request we are (maybe)
about to make.
In practice, it is very hard to trigger any bad behavior.
Currently, if we make a second request, it will always be to
the same URL (even in the face of redirects, because curl
handles the redirects internally). And we almost always
retry on HTTP_REAUTH these days. The one exception is if we
are streaming a large RPC request to the server (e.g., a
pushed packfile), in which case we cannot restart. It's
extremely unlikely to see a 401 response at this stage,
though, as we would typically have seen it when we sent a
probe request, before streaming the data.
This patch drops the automatic prompt out of case 2, and
instead requires the caller to do it. This is a few extra
lines of code, and the bug it fixes is unlikely to come up
in practice. But it is conceptually cleaner, and paves the
way for better handling of credentials across redirects.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
The help text for the `tool` flag should mention:
--tool=<tool>
instead of:
--tool-<tool>
Signed-off-by: Stefan Saasen <ssaasen@atlassian.com>
Reviewed-by: David Aguilar <davvid@gmail.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Sparse issues an "using sizeof on a function" warning for each
call to curl_easy_setopt() which sets an option that takes a
function pointer parameter. (currently 12 such warnings over 4
files.)
The warnings relate to the use of the "typecheck-gcc.h" header
file which adds a layer of type-checking macros to the curl
function invocations (for gcc >= 4.3 and !__cplusplus). As part
of the type-checking layer, 'sizeof' is applied to the function
parameter of curl_easy_setopt(). Note that, in the context of
sizeof, the function to function pointer conversion is not
performed and that sizeof(f) != sizeof(&f).
A simple solution, therefore, would be to replace the function
name in each such call to curl_easy_setopt() with an explicit
function pointer expression (i.e. replace f with &f).
However, the "typecheck-gcc.h" header file is only conditionally
included, in addition to the gcc and C++ checks mentioned above,
depending on the CURL_DISABLE_TYPECHECK preprocessor variable.
In order to suppress the warnings, we use target-specific variable
assignments to add -DCURL_DISABLE_TYPECHECK to SPARSE_FLAGS for
each file affected (http-push.c, http.c, http-walker.c and
remote-curl.c).
Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
The Thunderbird section of the 'MUA-specific hints' contains three
different approaches to setting up the mail client to leave patch
emails unmolested. The second approach (configuration) has a step
missing when configuring the composition window not to wrap. In
particular, the "mailnews.wraplength" configuration variable needs
to be set to zero. Update the documentation to add the missing
setting.
Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Linux Mint has an implementation of the highlight command (unrelated
to the one from http://www.andre-simon.de) that works as a simple
filter. The script uses 'sed' to add terminal colour escape codes
around text matching a regular expression. When t9500-*.sh attempts
to run "highlight --version", the script simply hangs waiting for
input. (See https://bugs.launchpad.net/linuxmint/+bug/815005).
The tool required by gitweb can be installed from the 'highlight'
package. Unfortunately, given the default $PATH, this leads to the
tool having lower precedence than the script.
In order to avoid hanging the test, add '</dev/null' to the command
line of the highlight invocation. Also, since the 'highlight' tool
requred by gitweb produces '--version' output (and the script does
not), saving the command output allows a simple check for the wrong
'highlight'.
Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
When the NO_MKSTEMPS build variable is not set, the gitmkstemps
function is dead code. Use a preprocessor conditional to only include
the definition when needed.
Noticed by sparse. ("'gitmkstemps' was not declared. Should it be
static?")
Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
A call to update_ref_lock() passes '0' to the 'int *type_p' parameter.
Noticed by sparse. ("Using plain integer as NULL pointer")
Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Commit 7192777 refactors git_parse_ulong, which is public, into a more
generic function. But since we kept the git_parse_ulong wrapper, only
that part needs to be public; nobody outside the file calls the
lower-level git_parse_unsigned.
Noticed with sparse. ("'git_parse_unsigned' was not declared. Should
it be static?")
Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk>
Explained-by: Jeff King <peff@peff.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
The style for multi-line comments is often mentioned and should be documented
for clarity.
Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
The description of the user.signingkey option only mentioned its use
when creating a signed tag. Make it clear that is is also used when
creating signed commits.
Signed-off-by: Nicolas Vigier <boklm@mars-attacks.org>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Since 920b691 (clone: refuse to clone if --branch
points to bogus ref) we refuse to clone with option
"-b" if the specified branch does not exist in the
(non-empty) upstream. If the upstream repository is empty,
the branch doesn't exist, either. So refuse the clone too.
Reported-by: Robert Mitwicki <robert.mitwicki@opensoftware.pl>
Signed-off-by: Ralf Thielow <ralf.thielow@gmail.com>
Acked-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
When working with multiple remotes, it is common to switch the upstream
from a remote to another. Doing so, the prompt may not be the expected
one. Providing an option to display tracking information sounds useful.
Add a "name" option to GIT_PS1_SHOWUPSTREAM which will show the upstream
abbrev name. This option is ignored if "verbose" is false.
Signed-off-by: Julien Carsique <julien.carsique@gmail.com>
Improved-by: SZEDER Gábor <szeder@ira.uka.de>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
mingw_path was introduced in abd4284 to output a mangled path as it is
passed as an argument to main(). But the name is misleading because
mangling does not come from MinGW, but from MSYS [1]. As abd4284 does not
introduce any MSYS or MinGW specific code but just prints out argv[2] as
it is passed to main(), give the function the more generic and less
confusing name "print_path".
[1] http://www.mingw.org/wiki/Posix_path_conversion
Signed-off-by: Sebastian Schuberth <sschuberth@gmail.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Several uses of the '^' operator are being interpreted by asciidoc
as requests to show the following text as a superscript. In order
to fix this problem, use backticks (`) to quote the text of the
affected git command invocations.
Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
The text contains two 'grep' invocations which include the 'start
of line' regular expression character '^'. Asciidoc mis-interprets
this use of '^' as a superscript request. In order to fix this
formatting problem, use backticks (`) to quote the text of the
affected 'grep' command invocations.
Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
This is a follow up to commit e47a8583 (enable SO_KEEPALIVE for
connected TCP sockets, 2011-12-06).
Sockets may never receive notification of some link errors,
causing "git fetch" or similar processes to hang forever.
Enabling keepalive messages allows hung processes to error out
after a few minutes/hours depending on the keepalive settings of
the system.
I noticed this problem with some non-interactive cronjobs getting
hung when talking to HTTP servers.
Signed-off-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Using a relative_path as git_dir first appears in v1.5.6-1-g044bbbc.
It will make git_dir shorter only if git_dir is inside work_tree,
and this will increase performance. But my last refactor effort on
relative_path function (commit v1.8.3-rc2-12-ge02ca72) changed that.
Always use relative_path as git_dir may bring troubles like
$gmane/234434.
Because new relative_path is a combination of original relative_path
from path.c and original path_relative from quote.c, so in order to
restore the origin implementation, save the original relative_path
as remove_leading_path, and call it in setup.c.
Suggested-by: Karsten Blees <karsten.blees@gmail.com>
Signed-off-by: Jiang Xin <worldhello.net@gmail.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Tvangeste found that the "relative_path" function could not work
properly on Windows if "in" and "prefix" have DOS drive prefix
(such as "C:/windows"). ($gmane/234434)
E.g., When execute: test-path-utils relative_path "C:/a/b" "D:/x/y",
should return "C:/a/b", but returns "../../C:/a/b", which is wrong.
So make relative_path honor DOS drive prefix, and add test cases
for it in t0060.
Reported-by: Tvangeste <i.4m.l33t@yandex.ru>
Helped-by: Johannes Sixt <j6t@kdbg.org>
Signed-off-by: Jiang Xin <worldhello.net@gmail.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
In test cases for relative_path, path with one leading character
(such as /a, /x) may be recogonized as "a:/" or "x:/" if there is
such DOS drive on MSYS platform. Use an umambigous leading path
"/foo" instead.
Also change two leading slashes (//) to three leading slashes (///),
otherwize it will be recognized as UNC name on MSYS platform.
Signed-off-by: Jiang Xin <worldhello.net@gmail.com>
Acked-by: Sebastian Schuberth <sschuberth@gmail.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
In commit 0656781fa "git mv" learned to update the submodule path in the
.gitmodules file when moving a submodule in the work tree. But since that
commit update_path_in_gitmodules() gets called no matter if we moved a
submodule or a regular file, which is wrong and leads to a bogus warning
when moving a regular file in a repo containing a .gitmodules file:
warning: Could not find section in .gitmodules where path=<filename>
Fix that by only calling update_path_in_gitmodules() when moving a
submodule. To achieve that, we introduce the special SUBMODULE_WITH_GITDIR
define to distinguish the cases where we also have to connect work tree
and git directory from those where we only need to update the .gitmodules
setting.
A test for submodules using a .git directory together with a .gitmodules
file has been added to t7001. Even though newer git versions will always
use a gitfile when cloning submodules, repositories cloned with older git
versions will still use this layout.
Reported-by: Matthieu Moy <Matthieu.Moy@grenoble-inp.fr>
Signed-off-by: Jens Lehmann <Jens.Lehmann@web.de>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
DiffMerge is a non-free (but gratis) tool that supports OS X, Windows and Linux.
See http://www.sourcegear.com/diffmerge/
DiffMerge includes a script `/usr/bin/diffmerge` that can be used to launch the
graphical compare tool.
This change adds mergetool support for DiffMerge and adds 'diffmerge' as an
option to the mergetool help.
Signed-off-by: Stefan Saasen <ssaasen@atlassian.com>
Acked-by: David Aguilar <davvid@gmail.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Normalize to my personal address, as my ETH addresses will expire
soon. Also add my new corp account to be somewhat futureproof.
Note that despite the private address being first, Google owns the
copyright as long as I am employed there.
Signed-off-by: Thomas Rast <trast@inf.ethz.ch>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
In Git v2.0, we will change the default --prefix for init/clone from
none/empty to "origin/" (which causes SVN-tracking branches to be
placed at refs/remotes/origin/* instead of refs/remotes/*).
This patch warns users about the upcoming change, both in the git-svn
manual page, and on stderr when running init/clone in the "multi-mode"
without providing a --prefix.
Cc: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Johan Herland <johan@herland.net>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
Currently, the git-svn defaults to using an empty prefix, which ends
up placing the SVN-tracking refs directly in refs/remotes/*. This
placement runs counter to Git's convention of placing remote-tracking
branches in refs/remotes/$remote/*.
Furthermore, combining git-svn with "regular" Git remotes run the risk
of clobbering refs under refs/remotes (e.g. if you have a git remote
called "tags" with a "v1" branch, it will overlap with the git-svn's
tracking branch for the "v1" tag from Subversion.
Even though the git-svn refs stored in refs/remotes/* are not "proper"
remote-tracking branches (since they are not covered by a proper git
remote's refspec), they clearly represent a similar concept, and would
benefit from following the same convention.
For example, if git-svn tracks Subversion branch "foo" at
refs/remotes/foo, and you create a local branch refs/heads/foo to add
some commits to be pushed back to Subversion (using "git svn dcommit),
then it is clearly unhelpful of Git to throw
warning: refname 'foo' is ambiguous.
every time you checkout, rebase, or otherwise interact with the branch.
At this time, the user is better off using the --prefix=foo/ (the
trailing slash is important) to git svn init/clone, to cause the
SVN-tracking refs to be placed at refs/remotes/foo/* instead of
refs/remotes/*. This patch updates the documentation to encourage
use of --prefix.
This is also in preparation for changing the default value of --prefix
at some point in the future.
Cc: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Johan Herland <johan@herland.net>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
When a test forgets to include && after each command, it is possible
for an early command to succeed but the test to fail, which can hide
bugs.
Checked using the following patch to the test harness:
--- a/t/test-lib.sh
+++ b/t/test-lib.sh
@@ -425,7 +425,17 @@ test_eval_ () {
eval </dev/null >&3 2>&4 "$*"
}
+check_command_chaining_ () {
+ eval >&3 2>&4 "(exit 189) && $*"
+ eval_chain_ret=$?
+ if test "$eval_chain_ret" != 189
+ then
+ error 'bug in test script: missing "&&" in test commands'
+ fi
+}
+
test_run_ () {
+ check_command_chaining_ "$1"
test_cleanup=:
expecting_failure=$2
setup_malloc_check
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Currently, we only try converting argv[1] from "-" into "@{-1}". This
means we do not notice "-" when used together with an option. Worse,
when "git cherry-pick" is run with no options, we segfault. Fix this
by doing the substitution after we have checked that there is
something in argv to cherry-pick and know any remaining options are
meant for the revision-listing machinery.
This still does not handle "-" after the first non-cherry-pick option.
For example,
git cherry-pick foo~2 - bar~5
and
git cherry-pick --no-merges -
will still dump usage.
Reported-by: Stefan Beller <stefanbeller@googlemail.com>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
The man page for `git svn` describes a situation in which "'git svn'
will not be able to rebuild" your $GIT_DIR/svn/**/.rev_map* files, but
no mention is made of in what circumstances `git svn` *will* be able to
do so, how to get `git svn` to do so, or even what these files are.
This patch adds a FILES section to the man page with a description of
what $GIT_DIR/svn/**/.rev_map* files are and how they are (re)built, and
links to this description from various other parts of the man page.
Signed-off-by: Keshav Kini <keshav.kini@gmail.com>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
As $GIT_DIR may not equal '.git', it's usually more generally correct to
refer to files in $GIT_DIR rather than in .git .
This will also allow me to link some of the occurrences of '.git' in
git-svn.txt to a new reference target inside this file in an upcoming
commit, because in AsciiDoc definitions apparently can't start with
a '.' character.
Signed-off-by: Keshav Kini <keshav.kini@gmail.com>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
It's redundant to say that $GIT_DIR/svn/<refname>/unhandled.log or
$GIT_DIR/svn/<refname>/index is in .git/svn when $GIT_DIR is '.git', and
is wrong when $GIT_DIR is not '.git'
Also, a '/' was missing from the pathname $GIT_DIR/svn/<refname>/index .
Signed-off-by: Keshav Kini <keshav.kini@gmail.com>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
As asterisks are used to indicate bold text in AsciiDoc, shell glob
expressions must be escaped appropriately.
Signed-off-by: Keshav Kini <keshav.kini@gmail.com>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
When parsing a commit object, git-svn wrongly think that a line
containing spaces means the end of headers and the start of the commit
message. In case of signed commit, the gpgsig entry contains a line with
one space, so "git svn dcommit" will include part of the signature in
the commit message.
An example of such problem :
http://svnweb.mageia.org/treasurer?view=revision&revision=86
This commit changes the regex to only match an empty line as separator
between the headers and the commit message.
Signed-off-by: Nicolas Vigier <boklm@mars-attacks.org>
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Eric Wong <normalperson@yhbt.net>
This test was added, commented out, in fed1b5ca (git-checkout: Test
for relative path use, 2007-11-09). Later git's path handling was
improved (d089ebaa, setup: sanitize absolute and funny paths in
get_pathspec(), 2008-01-28) but we forgot to enable the now-working
test.
This test expects to run from a subdirectory, so add a 'cd'. While
we're here, examine the content of the checked-out file instead of
just checking that it exists. The other checkout tests already do the
same.
Signed-off-by: Stefan Beller <stefanbeller@googlemail.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Simple patch to avoid unitialized warning and log what we'll do.
Signed-off-by: Ted Zlatanov <tzz@lifelogs.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
On my system this is in /usr/share/asciidoc/dblatex not
/etc/asciidoc/dblatex. Extract this portion of the path to a variable
so that is can be set in config.mak.
Signed-off-by: John Keeping <john@keeping.me.uk>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Over time, the http_get_strbuf function has grown several
optional parameters. We now have a bitfield with multiple
boolean options, as well as an optional strbuf for returning
the content-type of the response. And a future patch in this
series is going to add another strbuf option.
Treating these as separate arguments has a few downsides:
1. Most call sites need to add extra NULLs and 0s for the
options they aren't interested in.
2. The http_get_* functions are actually wrappers around
2 layers of low-level implementation functions. We have
to pass these options through individually.
3. The http_get_strbuf wrapper learned these options, but
nobody bothered to do so for http_get_file, even though
it is backed by the same function that does understand
the options.
Let's consolidate the options into a single struct. For the
common case of the default options, we'll allow callers to
simply pass a NULL for the options struct.
The resulting code is often a few lines longer, but it ends
up being easier to read (and to change as we add new
options, since we do not need to update each call site).
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
collapse_todo_ids() uses `git rev-parse --short=7' to abbreviate
commit ids before showing them to the user in a text editor. Let's
drop argument from --short to the configured value instead (still
defaulting to 7).
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Eric Sunshine <sunshine@sunshineco.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
When we retrieve the content-type of an http response, curl
gives us a pointer to internal storage, which we then copy
into a strbuf. Let's factor out the get-and-copy routine,
which can be used for getting other curl info.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
