* [PATCH] cg-admin-rewritehist: fix reappearing files with --filter-tree.
From: Johannes Sixt @ 2006-04-29 21:45 UTC (permalink / raw)
With --filter-tree a working copy is checked out for each commit.
However, if a file is removed by a commit, the file is _not_ removed
from the working copy by git-checkout-index. This must be done explicitly,
otherwise the file becomes added back again.
Signed-off-by: Johannes Sixt <johannes.sixt@telecom.at>
---
cg-admin-rewritehist | 5 ++++-
1 files changed, 4 insertions(+), 1 deletions(-)
26bb71a2d3d583d9eee10f4e950ff1b7d400e975
diff --git a/cg-admin-rewritehist b/cg-admin-rewritehist
index 7dd83cf..13ffb5d 100755
--- a/cg-admin-rewritehist
+++ b/cg-admin-rewritehist
@@ -213,10 +213,13 @@ while read commit; do
if [ "$filter_tree" ]; then
git-checkout-index -f -u -a
+ # files that $commit removed are now still in the working tree;
+ # remove them, else they would be added again
+ git-ls-files -z --others | xargs -0 rm -f
eval "$filter_tree"
git-diff-index -r $commit | cut -f 2- | tr '\n' '\0' | \
xargs -0 git-update-index --add --replace --remove
- git-ls-files --others | tr '\n' '\0' | \
+ git-ls-files -z --others | \
xargs -0 git-update-index --add --replace --remove
fi
--
1.3.1.gaa6b
^ permalink raw reply related
* Re: cg-clone not fetching all tags?
From: Wolfgang Denk @ 2006-04-29 22:11 UTC (permalink / raw)
To: Petr Baudis; +Cc: Git Mailing List
In-Reply-To: <20060429170542.GJ27689@pasky.or.cz>
Dear Petr,
in message <20060429170542.GJ27689@pasky.or.cz> you wrote:
>
> you need to run git-update-server-info every time you add or update a
> tag (or best every time you push). See the NOTES section of
I do this. Also, this does not seem to explain why I can only see
recent tags, but not old ones?
> cg-admin-setuprepo documentation for details on how to set it up to be
> called automagically at every push.
Actually I don't use push, but "cg-update" on the receiving end. Is
this a problem?
Best regards,
Wolfgang Denk
--
Software Engineering: Embedded and Realtime Systems, Embedded Linux
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd@denx.de
There's an old proverb that says just about whatever you want it to.
^ permalink raw reply
* Re: cg-clone not fetching all tags?
From: Wolfgang Denk @ 2006-04-29 22:11 UTC (permalink / raw)
To: Johannes Sixt; +Cc: git, Petr Baudis
In-Reply-To: <200604292342.16306.johannes.sixt@telecom.at>
In message <200604292342.16306.johannes.sixt@telecom.at> you wrote:
>
> There are two types of tags: They can point to
> 1. a commit object
> 2. a proper tag object (which in turn references the commit)
>
> git-update-server-info seems to generate info only for case 2, and so are the
> only ones that http can fetch.
And git-cvsimport (at least older versions of it) imports CVS tags
only as type 1 ?
That would perfectly explain the situation. How can this be fixed?
Best regards,
Viele Grüße,
Wolfgang Denk
--
Software Engineering: Embedded and Realtime Systems, Embedded Linux
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd@denx.de
There is a multi-legged creature crawling on your shoulder.
-- Spock, "A Taste of Armageddon", stardate 3193.9
^ permalink raw reply
* Re: cg-clone not fetching all tags?
From: Johannes Sixt @ 2006-04-29 22:54 UTC (permalink / raw)
To: git
In-Reply-To: <20060429221114.557FC35288F@atlas.denx.de>
On Sunday 30 April 2006 00:11, Wolfgang Denk wrote:
> In message <200604292342.16306.johannes.sixt@telecom.at> you wrote:
> > There are two types of tags: They can point to
> > 1. a commit object
> > 2. a proper tag object (which in turn references the commit)
> >
> > git-update-server-info seems to generate info only for case 2, and so are
> > the only ones that http can fetch.
>
> And git-cvsimport (at least older versions of it) imports CVS tags
> only as type 1 ?
>
> That would perfectly explain the situation. How can this be fixed?
Something like this tells you which one points to a commit or a tag:
for i in `find .git/refs/tags -type f`; do
echo -n "$i "; git-cat-file -t $(cat $i);
done
Then for each commit-type $tag you do
git-tag -m ' ' $tag-new $tag
mv .git/refs/tags/$tag-new .git/refs/tags/$tag
(git-tag does not accept an empty tag message - if you dislike the
single-blank-message, you have to manufacture your tags with git-mktag
manually.)
-- Hannes
^ permalink raw reply
* Re: cg-clone not fetching all tags?
From: Junio C Hamano @ 2006-04-30 4:11 UTC (permalink / raw)
To: Wolfgang Denk; +Cc: git, pasky
In-Reply-To: <20060429221114.557FC35288F@atlas.denx.de>
Wolfgang Denk <wd@denx.de> writes:
> In message <200604292342.16306.johannes.sixt@telecom.at> you wrote:
>>
>> There are two types of tags: They can point to
>> 1. a commit object
>> 2. a proper tag object (which in turn references the commit)
>>
>> git-update-server-info seems to generate info only for case 2, and so are the
>> only ones that http can fetch.
>
> And git-cvsimport (at least older versions of it) imports CVS tags
> only as type 1 ?
>
> That would perfectly explain the situation. How can this be fixed?
I suspect there is no need to manufacture the tag. Something
like this should do.
*WARNING* Since I do not do Porcelain, and I am not a Cogito
user, this is obviously untested. In addition, I am seriously
drunk right now...
-- >8 --
[PATCH] (cogito) Auto-follow lightweight tags as well.
---
diff --git a/cg-fetch b/cg-fetch
index 687e6b9..39764c7 100755
--- a/cg-fetch
+++ b/cg-fetch
@@ -218,7 +218,8 @@ fetch_tags()
git-ls-remote --tags "$uri" |
# SHA1 refs/tags/v0.99.8^{} --> SHA1 tags/v0.99.8
# where SHA1 is the object v0.99.8 tag points at.
- sed -ne 's:\([^ ]\) refs/\(tags/.*\)^{}$:\1 \2:p' |
+ sed -n -e 's:\([^ ]\) refs/\(tags/.*\)^{}$:\1 \2:p' \
+ -e 's:\([^ ]\) refs/\(tags/.*\)$:\1 \2:p|' \
while read sha1 tagname; do
# Do we have the tag itself?
[ -s "$_git/refs/$tagname" ] && continue
^ permalink raw reply related
* [PATCH] git builtin "push"
From: Linus Torvalds @ 2006-04-30 4:22 UTC (permalink / raw)
To: Junio C Hamano, Git Mailing List
This adds a builtin "push" command, which is largely just a C'ification of
the "git-push.sh" script.
Now, the reason I did it as a built-in is partly because it's yet another
step on relying less on shell, but it's actually mostly because I've
wanted to be able to push to _multiple_ repositories, and the most obvious
and simplest interface for that would seem be to just have a "remotes"
file that has multiple URL entries.
(For "pull", having multiple entries should either just select the first
one, or you could fall back on the others on failure - your choice).
And quite frankly, it just became too damn messy to do that in shell.
Besides, we actually have a fair amount of infrastructure in C, so it just
wasn't that hard to do.
Of course, this is almost totally untested. It probably doesn't work for
anything but the one trial I threw at it. "Simple" doesn't necessarily
mean "obviously correct".
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
---
Comments? I wrote it so that it _should_ be fairly easy to re-use at
least the branches/remotes helper functions for a built-in "git fetch" as
well. But I didn't have the multi-URI issue with anything but pushing.
Junio - I currently have a "push-all" script in each of the repos I
maintain. That has worked, but I'd just rather do something like
git push all
instead, and have it pick up the list of URI's from .git/remotes/all.
How do you do multi-targeted pushes?
diff --git a/Makefile b/Makefile
index 8ce27a6..9919992 100644
--- a/Makefile
+++ b/Makefile
@@ -214,7 +214,7 @@ LIB_OBJS = \
$(DIFF_OBJS)
BUILTIN_OBJS = \
- builtin-log.o builtin-help.o
+ builtin-log.o builtin-help.o builtin-push.o
GITLIBS = $(LIB_FILE) $(XDIFF_LIB)
LIBS = $(GITLIBS) -lz
diff --git a/builtin-push.c b/builtin-push.c
new file mode 100644
index 0000000..a0c1caa
--- /dev/null
+++ b/builtin-push.c
@@ -0,0 +1,255 @@
+/*
+ * "git push"
+ */
+#include "cache.h"
+#include "refs.h"
+#include "run-command.h"
+#include "builtin.h"
+
+#define MAX_URI (16)
+
+static const char push_usage[] = "git push [--all] [--tags] [--force] <repository> [<refspec>...]";
+
+static int all = 0, tags = 0, force = 0, thin = 1;
+static const char *execute = NULL;
+
+#define BUF_SIZE (2084)
+static char buffer[BUF_SIZE];
+
+static const char **refspec = NULL;
+static int refspec_nr = 0;
+
+static void add_refspec(const char *ref)
+{
+ int nr = refspec_nr + 1;
+ refspec = xrealloc(refspec, nr * sizeof(char *));
+ refspec[nr-1] = ref;
+ refspec_nr = nr;
+}
+
+static int expand_one_ref(const char *ref, const unsigned char *sha1)
+{
+ /* Ignore the "refs/" at the beginning of the refname */
+ ref += 5;
+
+ if (strncmp(ref, "tags/", 5))
+ return 0;
+
+ add_refspec(strdup(ref));
+ return 0;
+}
+
+static void expand_refspecs(void)
+{
+ if (all) {
+ if (refspec_nr)
+ die("cannot mix '--all' and a refspec");
+
+ /*
+ * No need to expand "--all" - we'll just use
+ * the "--all" flag to send-pack
+ */
+ return;
+ }
+ if (!tags)
+ return;
+ for_each_ref(expand_one_ref);
+}
+
+static void set_refspecs(const char **refs, int nr)
+{
+ if (nr) {
+ size_t bytes = nr * sizeof(char *);
+
+ refspec = xrealloc(refspec, bytes);
+ memcpy(refspec, refs, bytes);
+ refspec_nr = nr;
+ }
+ expand_refspecs();
+}
+
+static int get_remotes_uri(const char *repo, const char *uri[MAX_URI])
+{
+ int n = 0;
+ FILE *f = fopen(git_path("remotes/%s", repo), "r");
+
+ if (!f)
+ return -1;
+ while (fgets(buffer, BUF_SIZE, f)) {
+ char *s, *p;
+
+ if (strncmp("URL: ", buffer, 5))
+ continue;
+ s = buffer + 5;
+
+ /* Remove whitespace at the head.. */
+ while (isspace(*s))
+ s++;
+ if (!*s)
+ continue;
+
+ /* ..and at the end */
+ p = s + strlen(s);
+ while (isspace(p[-1]))
+ *--p = 0;
+
+ uri[n++] = strdup(s);
+ if (n == MAX_URI)
+ break;
+ }
+ fclose(f);
+ if (!n)
+ die("remote '%s' has no URL", repo);
+ return n;
+}
+
+static int get_branches_uri(const char *repo, const char *uri[MAX_URI])
+{
+ const char *slash = strchr(repo, '/');
+ int n = slash ? slash - repo : 1000;
+ FILE *f = fopen(git_path("branches/%.*s", n, repo), "r");
+ char *s, *p;
+ int len;
+
+ if (!f)
+ return 0;
+ s = fgets(buffer, BUF_SIZE, f);
+ fclose(f);
+ if (!s)
+ return 0;
+ while (isspace(*s))
+ s++;
+ if (!*s)
+ return 0;
+ p = s + strlen(s);
+ while (isspace(p[-1]))
+ *--p = 0;
+ len = p - s;
+ if (slash)
+ len += strlen(slash);
+ p = xmalloc(len + 1);
+ strcpy(p, s);
+ if (slash)
+ strcat(p, slash);
+ uri[0] = p;
+ return 1;
+}
+
+static int get_uri(const char *repo, const char *uri[MAX_URI])
+{
+ int n;
+
+ if (*repo != '/') {
+ n = get_remotes_uri(repo, uri);
+ if (n > 0)
+ return n;
+
+ n = get_branches_uri(repo, uri);
+ if (n > 0)
+ return n;
+ }
+
+ uri[0] = repo;
+ return 1;
+}
+
+static int do_push(const char *repo)
+{
+ const char *uri[MAX_URI];
+ int i, n = get_uri(repo, uri);
+ int remote;
+ const char **argv;
+ int argc;
+
+ n = get_uri(repo, uri);
+ if (n <= 0)
+ die("bad repository '%s'", repo);
+
+ argv = xmalloc((refspec_nr + 10) * sizeof(char *));
+ argv[0] = "dummy-send-pack";
+ argc = 1;
+ if (all)
+ argv[argc++] = "--all";
+ if (force)
+ argv[argc++] = "--force";
+ if (execute)
+ argv[argc++] = execute;
+ if (thin)
+ argv[argc++] = "--thin";
+ remote = argc;
+ argv[argc++] = "dummy-remote";
+ while (refspec_nr--)
+ argv[argc++] = *refspec++;
+ argv[argc] = NULL;
+
+ for (i = 0; i < n; i++) {
+ int error;
+ const char *dest = uri[i];
+ const char *sender = "git-send-pack";
+ if (!strncmp(dest, "http://", 7) ||
+ !strncmp(dest, "https://", 8))
+ sender = "git-http-push";
+ argv[0] = sender;
+ argv[remote] = dest;
+ error = run_command_v(argc, argv);
+ if (!error)
+ continue;
+ switch (error) {
+ case -ERR_RUN_COMMAND_FORK:
+ die("unable to fork for %s", sender);
+ case -ERR_RUN_COMMAND_EXEC:
+ die("unable to exec %s", sender);
+ case -ERR_RUN_COMMAND_WAITPID:
+ case -ERR_RUN_COMMAND_WAITPID_WRONG_PID:
+ case -ERR_RUN_COMMAND_WAITPID_SIGNAL:
+ case -ERR_RUN_COMMAND_WAITPID_NOEXIT:
+ die("%s died with strange error", sender);
+ default:
+ return -error;
+ }
+ }
+ return 0;
+}
+
+int cmd_push(int argc, const char **argv, char **envp)
+{
+ int i;
+ const char *repo = "origin"; // default repository
+
+ for (i = 1; i < argc; i++) {
+ const char *arg = argv[i];
+
+ if (arg[0] != '-') {
+ repo = arg;
+ i++;
+ break;
+ }
+ if (!strcmp(arg, "--all")) {
+ all = 1;
+ continue;
+ }
+ if (!strcmp(arg, "--tags")) {
+ tags = 1;
+ continue;
+ }
+ if (!strcmp(arg, "--force")) {
+ force = 1;
+ continue;
+ }
+ if (!strcmp(arg, "--thin")) {
+ thin = 1;
+ continue;
+ }
+ if (!strcmp(arg, "--no-thin")) {
+ thin = 0;
+ continue;
+ }
+ if (!strncmp(arg, "--exec=", 7)) {
+ execute = arg;
+ continue;
+ }
+ usage(push_usage);
+ }
+ set_refspecs(argv + i, argc - i);
+ return do_push(repo);
+}
diff --git a/builtin.h b/builtin.h
index 47408a0..94fa9b5 100644
--- a/builtin.h
+++ b/builtin.h
@@ -20,4 +20,6 @@ extern int cmd_whatchanged(int argc, con
extern int cmd_show(int argc, const char **argv, char **envp);
extern int cmd_log(int argc, const char **argv, char **envp);
+extern int cmd_push(int argc, const char **argv, char **envp);
+
#endif
diff --git a/git.c b/git.c
index 01b7e28..fd479e9 100644
--- a/git.c
+++ b/git.c
@@ -46,6 +46,7 @@ static void handle_internal_command(int
{ "log", cmd_log },
{ "whatchanged", cmd_whatchanged },
{ "show", cmd_show },
+ { "push", cmd_push },
};
int i;
^ permalink raw reply related
* [RFC] 'Download' stat
From: Marco Costalba @ 2006-04-30 4:59 UTC (permalink / raw)
To: git
Hi all,
none is interested in (writing ;-) ) a kind of 'download' summary stat?
If git-daemon after a connection could save a record with fields like:
- date
- command requested (clone, pull, etc..)
- current HEAD
Perhaps would be possible to show nice stat/graph about repository
activity and most downloaded repository content tags.
Something like to know how many people has cloned/pulled v2.6.17-rc2
in date 2006/4/19
Perhaps is totally useless, just an idea.
Marco
^ permalink raw reply
* Re: [PATCH] git builtin "push"
From: Junio C Hamano @ 2006-04-30 7:08 UTC (permalink / raw)
To: Linus Torvalds; +Cc: git
In-Reply-To: <Pine.LNX.4.64.0604292111570.9901@g5.osdl.org>
Linus Torvalds <torvalds@osdl.org> writes:
> Junio - I currently have a "push-all" script in each of the repos I
> maintain. That has worked, but I'd just rather do something like
>
> git push all
>
> instead, and have it pick up the list of URI's from .git/remotes/all.
Makes sense. I am still somewhat drunk (cachaça -- agua de
beber -- pretty good but strong stuff), so I will look at the
patch tomorrow, but...
> How do you do multi-targeted pushes?
I do not. At least not that often. I merge things up, test all
of "master", "maint", "next", and "pu" locally (the Meta/Doit
script, available after checking out "todo" branch in Meta/
subdirectory), then "git push ko-private" to push into ~/git of
the kernel.org machine and do "all branches" test again
(Meta/DoKernelOrg script). Only after all that passes, I do a
"git push ko" from my local machine to push into the public
area, /pub/scm/git/git.git, of the kernel.org machine.
For Solaris and Cygwin testing, only occasionally, I do not push
but pull from the machines at work. This is because I want to
catch breakage of pull side myself if there is one.
The only time I do multi-target push is just before cutting a
release. In addition to "git push ko-private" to push into the
primary Opteron machine, I push to "git push ko-old-private" to
push into the i386 machine at kernel.org, "Meta/DoKernelOrg
master" (or "maint") to prepare RPMs for i386 there, and another
"Meta/DoKernelOrg master" (or "maint") on the Opteron to
prepare RPMs for x86-64. So personally I never felt the need
for a multi-target push.
^ permalink raw reply
* Re: [RFC] 'Download' stat
From: Junio C Hamano @ 2006-04-30 7:11 UTC (permalink / raw)
To: Marco Costalba; +Cc: git
In-Reply-To: <e5bfff550604292159t5fac5436q73aecd5dbd56f516@mail.gmail.com>
"Marco Costalba" <mcostalba@gmail.com> writes:
> Hi all,
>
> none is interested in (writing ;-) ) a kind of 'download' summary stat?
>
> If git-daemon after a connection could save a record with fields like:
>
> - date
> - command requested (clone, pull, etc..)
> - current HEAD
>
> Perhaps would be possible to show nice stat/graph about repository
> activity and most downloaded repository content tags.
Perhaps "git-daemon --verbose --syslog". Grep for loginfo in
daemon.c for details.
^ permalink raw reply
* Fwd: More qgit defects 1/2
From: Marco Costalba @ 2006-04-30 8:19 UTC (permalink / raw)
To: git
I would like to forward this thread to the list in case someone find
it interesting and/or suggests improvements.
---------- Forwarded message ----------
From: Pavel Roskin <proski@gnu.org>
Date: Apr 30, 2006 9:50 AM
Subject: Re: More qgit defects
To: Marco Costalba <mcostalba@gmail.com>
Cc: ydirson@altern.org
Hello, Marco!
On Fri, 2006-04-28 at 19:21 +0200, Marco Costalba wrote:
> I have just pushed some patches according to your suggestions.
Thank you very much.
> If we don't find any issue/annoyance I would like to release 1.2 in
> the next couple of days, so to have a broader test base, and also
> because currently released 1.2rc2 has a bad crash bug.
I think it's good for the 1.2 release.
> So if you have any concerns or you would like to see something more
> added please let me know.
I think the list items in the file window need a pop-up menu. Without
it, the users would not be able to discover what e.g. the double click
would do.
More generally, I feel that something is wrong with the user interface.
I think gitk is doing it better with a single pane that has the comments
and the diff. Splitting the window horizontally in three panes seems to
be too much, especially when one of the panes is used for diffs.
Actually, qgit may be better for serious work, such as comparing the
description with the actual text. But still, the panes are too small in
vertical direction, and if I tear off the diff, it will obscure
something.
The jumpiness of qgit has been largely fixed, but there are still cases
when mere selecting an item affect another window.
Is it really necessary to have a separate file window? How different is
it from the main window with the "filter by tree" option on? Can we
switch between diff and annotation? More generally, what are the things
that the users are not normally viewing at the same time?
I'm not asking to fix anything, but qgit still feels odd to me. I know
I'm not qualified to ask you to rearrange everything, because I'm not a
GUI specialist, and I don't see how qgit should work.
Maybe you could ask in the git list how to improve qgit? I cannot find
any suitable forum about HIG (human interface guidelines), but if you
ask in the git list, maybe somebody will give you an idea where to ask.
--
Regards,
Pavel Roskin
^ permalink raw reply
* Fwd: More qgit defects 2/2
From: Marco Costalba @ 2006-04-30 8:19 UTC (permalink / raw)
To: git
I would like to forward this thread to the list in case someone find
it interesting and/or suggests improvements.
---------- Forwarded message ----------
From: Marco Costalba <mcostalba@gmail.com>
Date: Apr 30, 2006 10:11 AM
Subject: Re: More qgit defects
To: Pavel Roskin <proski@gnu.org>
Cc: ydirson@altern.org
Hi Pavel,
>
> I think the list items in the file window need a pop-up menu. Without
> it, the users would not be able to discover what e.g. the double click
> would do.
>
yes, I agree.
> More generally, I feel that something is wrong with the user interface.
> I think gitk is doing it better with a single pane that has the comments
> and the diff. Splitting the window horizontally in three panes seems to
> be too much, especially when one of the panes is used for diffs.
>
Please Pavel pull from latest qgit repo. I just pushed a patch that,
at least for my
browsing style, fixes this and IMHO is better then gitk.
Patch description is as follow:
"Usability enhancement: use 'P' key to quick toggle patch viewer
One of the limitation of patch browsing in qgit is the use of two
frames, one for revision logs and the other for patch viewer.
If patch viewer is docked the space is normally limited by revision
log frame. If patch viewer is undocked it is always at top level and
hides main view and revision logs.
So this patch adds a new key binding 'P' that hides/unhides diff
viewer window.
The intended use is simple: open patch viewer and undock it so to cover
a good chunk of screen space, then press 'P' key and diff viewer will
disappear.
Now you can browse the revisions as usual and toggle diff viewer with 'P'
key so to easily view both logs and patches without any space constrains.
"
After a little bit of using I have found this way very natural, quick
and, especially important for me, do not compromise patch view size.
What I do is to size the undocked patch viewer on my left screen, and
toggle the P key, but other adjustment are possible, see:
http://digilander.libero.it/mcostalba/qgit_P_1.png
http://digilander.libero.it/mcostalba/qgit_P_2.png
Tip: let your mouse be always over main list view so that patch view
does not steal the focus and you can browse only with keyboard arrows
as well with new key bindings (see F1).
Let me know what you think.
Marco
^ permalink raw reply
* Re: [PATCH] git builtin "push"
From: sean @ 2006-04-30 9:40 UTC (permalink / raw)
To: Linus Torvalds; +Cc: junkio, git
In-Reply-To: <Pine.LNX.4.64.0604292111570.9901@g5.osdl.org>
On Sat, 29 Apr 2006 21:22:49 -0700 (PDT)
Linus Torvalds <torvalds@osdl.org> wrote:
> Comments? I wrote it so that it _should_ be fairly easy to re-use at
> least the branches/remotes helper functions for a built-in "git fetch" as
> well. But I didn't have the multi-URI issue with anything but pushing.
> + if (strncmp(ref, "tags/", 5))
> + return 0;
[...]
> + for_each_ref(expand_one_ref);
How about a for_each_tag() function?
> + int i, n = get_uri(repo, uri);
[...]
> + n = get_uri(repo, uri);
> + if (n <= 0)
> + die("bad repository '%s'", repo);
get_uri() called twice.
The patch looks good and is easy to read, but wouldn't it be better
to require new c code to be thread safe and not leak memory? Assuming
run-once-and-exit is making it difficult to push into a library.
Sean
builtin-push.c | 11 +++--------
refs.c | 5 +++++
refs.h | 1 +
3 files changed, 9 insertions(+), 8 deletions(-)
0e0e3cff71d65ac4cdc560ae143aded03acb4ceb
diff --git a/builtin-push.c b/builtin-push.c
index a0c1caa..0d74ed1 100644
--- a/builtin-push.c
+++ b/builtin-push.c
@@ -31,10 +31,6 @@ static int expand_one_ref(const char *re
{
/* Ignore the "refs/" at the beginning of the refname */
ref += 5;
-
- if (strncmp(ref, "tags/", 5))
- return 0;
-
add_refspec(strdup(ref));
return 0;
}
@@ -51,9 +47,8 @@ static void expand_refspecs(void)
*/
return;
}
- if (!tags)
- return;
- for_each_ref(expand_one_ref);
+ if (tags)
+ for_each_tag(expand_one_ref);
}
static void set_refspecs(const char **refs, int nr)
@@ -156,7 +151,7 @@ static int get_uri(const char *repo, con
static int do_push(const char *repo)
{
const char *uri[MAX_URI];
- int i, n = get_uri(repo, uri);
+ int i, n;
int remote;
const char **argv;
int argc;
diff --git a/refs.c b/refs.c
index 03398cc..c5a2dd0 100644
--- a/refs.c
+++ b/refs.c
@@ -178,6 +178,11 @@ int head_ref(int (*fn)(const char *path,
return 0;
}
+int for_each_tag(int (*fn)(const char *path, const unsigned char *sha1))
+{
+ return do_for_each_ref("refs/tags", fn);
+}
+
int for_each_ref(int (*fn)(const char *path, const unsigned char *sha1))
{
return do_for_each_ref("refs", fn);
diff --git a/refs.h b/refs.h
index 2625596..b74cd4d 100644
--- a/refs.h
+++ b/refs.h
@@ -7,6 +7,7 @@ #define REFS_H
*/
extern int head_ref(int (*fn)(const char *path, const unsigned char *sha1));
extern int for_each_ref(int (*fn)(const char *path, const unsigned char *sha1));
+extern int for_each_tag(int (*fn)(const char *path, const unsigned char *sha1));
/** Reads the refs file specified into sha1 **/
extern int get_ref_sha1(const char *ref, unsigned char *sha1);
--
1.3.1.g9c203
^ permalink raw reply related
* Re: More qgit defects
From: Marco Costalba @ 2006-04-30 10:13 UTC (permalink / raw)
To: Pavel Roskin; +Cc: ydirson, git
In-Reply-To: <1146390144.13634.9.camel@dv>
On 4/30/06, Pavel Roskin <proski@gnu.org> wrote:
> On Sun, 2006-04-30 at 05:26 -0400, Pavel Roskin wrote:
> > No, something still feels wrong. I think even the gurus of GUI cannot
> > decide what to do if many frames need to be on screen. Do you know that
> > many graphic designers hate GIMP for the overuse of dockable toplevel
> > windows? Krita prefers dockable frames. Photoshop uses non-dockable
> > child windows, I believe.
> >
> > The difference for qgit is that is generally wants bigger windows.
> > Whether the revision tree or the patch, having more space allows the
> > frame to present a better picture to the users.
>
> Replying to myself, sorry. How about tabs?
>
> One tab for the main view. Basically what we have now.
>
> Then tabs for revisions. We can have more than one revision open, with
> the comment and with the patch, and and with affected files. They will
> have the GUI centered on the change made by the revision. StGIT commits
> would have an editable comment.
>
> Then tabs for files. Again, possibly more than one. Each tab about a
> specific file. The file history, annotations, maybe even an editor for
> the file.
>
> The idea was inspired by Azureus.
>
Throwing in the tabs is a *very* big change, but, just to discuss....I
agree on the note that in qgit we have three different approaches:
fixed frames (revisions, file tree, affected files), detachable frames
(patch) and separate windows (annotations).
This is a bit strange and could give an odd GUI feeling.
I like the tab idea because it's clear and simple and fixes the 'many
approaches' problem. What I would suggest is, at least at first step,
do not change the main view and have only three tabs:
Tab1: Main view with revisions, file tree (hide able), affected files.
Tab2: Patch view with patch stat and diffs
Tab3: File history + file content/annotation view
In other words just put the frames/windows as are now in browse able
tabs. In this way main view still gives a good amount of information
without requiring changing the tab and the tabs are reserved for 'big
space' needed infos only.
Marco
^ permalink raw reply
* [PATCH] builtin-push: resurrect parsing of Push: lines
From: Johannes Schindelin @ 2006-04-30 12:05 UTC (permalink / raw)
To: git
The C'ification of push left these behind.
Signed-off-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
---
builtin-push.c | 26 +++++++++++++++++++++-----
1 files changed, 21 insertions(+), 5 deletions(-)
diff --git a/builtin-push.c b/builtin-push.c
index a0c1caa..4e659f0 100644
--- a/builtin-push.c
+++ b/builtin-push.c
@@ -68,6 +68,10 @@ static void set_refspecs(const char **re
expand_refspecs();
}
+#define MAX_REFSPECS 10
+static int current_refspec = 0;
+static char *refspecs_[MAX_REFSPECS];
+
static int get_remotes_uri(const char *repo, const char *uri[MAX_URI])
{
int n = 0;
@@ -76,11 +80,17 @@ static int get_remotes_uri(const char *r
if (!f)
return -1;
while (fgets(buffer, BUF_SIZE, f)) {
+ int is_refspec;
char *s, *p;
- if (strncmp("URL: ", buffer, 5))
+ if (!strncmp("URL: ", buffer, 5)) {
+ is_refspec = 0;
+ s = buffer + 5;
+ } else if (!strncmp("Push: ", buffer, 6)) {
+ is_refspec = 1;
+ s = buffer + 6;
+ } else
continue;
- s = buffer + 5;
/* Remove whitespace at the head.. */
while (isspace(*s))
@@ -93,9 +103,10 @@ static int get_remotes_uri(const char *r
while (isspace(p[-1]))
*--p = 0;
- uri[n++] = strdup(s);
- if (n == MAX_URI)
- break;
+ if (!is_refspec && n < MAX_URI)
+ uri[n++] = strdup(s);
+ else if (is_refspec && current_refspec < MAX_REFSPECS)
+ refspecs_[current_refspec++] = strdup(s);
}
fclose(f);
if (!n)
@@ -140,6 +151,8 @@ static int get_uri(const char *repo, con
int n;
if (*repo != '/') {
+ current_refspec = 0;
+
n = get_remotes_uri(repo, uri);
if (n > 0)
return n;
@@ -165,6 +178,9 @@ static int do_push(const char *repo)
if (n <= 0)
die("bad repository '%s'", repo);
+ if (refspec_nr == 0)
+ set_refspecs((const char**)refspecs_, current_refspec);
+
argv = xmalloc((refspec_nr + 10) * sizeof(char *));
argv[0] = "dummy-send-pack";
argc = 1;
^ permalink raw reply related
* [PATCH 1/3] builtin-push: also ask config for remote information
From: Johannes Schindelin @ 2006-04-30 13:23 UTC (permalink / raw)
To: git
Now you can store your remote information in the config file like this:
[remote.upstream]
url = me@company.com:the-project
push = master:iceballs
Signed-off-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
---
Obviously, this is on top of the patch to parse Pull: lines, too.
builtin-push.c | 33 +++++++++++++++++++++++++++++++++
1 files changed, 33 insertions(+), 0 deletions(-)
e55eb69f6332087c45082e16ccbf4e510d721e29
diff --git a/builtin-push.c b/builtin-push.c
index 4e659f0..e3131ed 100644
--- a/builtin-push.c
+++ b/builtin-push.c
@@ -72,6 +72,36 @@ #define MAX_REFSPECS 10
static int current_refspec = 0;
static char *refspecs_[MAX_REFSPECS];
+static int repo_len = 0;
+static const char *repo_ = NULL;
+static int current_uri = 0;
+static const char **uri_;
+
+static int get_value(const char* key, const char* value)
+{
+ if (!strncmp(key, "remote.", 7) && !strncmp(key + 7, repo_, repo_len)) {
+ if (!strcmp(key + 7 + repo_len, ".url"))
+ uri_[current_uri++] = strdup(value);
+ else if (!strcmp(key + 7 + repo_len, ".push")
+ && current_refspec < MAX_REFSPECS)
+ refspecs_[current_refspec++] = strdup(value);
+ }
+
+ return 0;
+}
+
+static int get_config_remotes_uri(const char *repo, const char *uri[MAX_URI])
+{
+ repo_len = strlen(repo);
+ repo_ = repo;
+ current_uri = 0;
+ uri_ = uri;
+
+ git_config(get_value);
+
+ return current_uri;
+}
+
static int get_remotes_uri(const char *repo, const char *uri[MAX_URI])
{
int n = 0;
@@ -153,6 +183,9 @@ static int get_uri(const char *repo, con
if (*repo != '/') {
current_refspec = 0;
+ n = get_config_remotes_uri(repo, uri);
+ if (n > 0)
+ return n;
n = get_remotes_uri(repo, uri);
if (n > 0)
return n;
--
1.3.1.g38c00-dirty
^ permalink raw reply related
* [PATCH 2/3] fetch, pull: ask config for remote information
From: Johannes Schindelin @ 2006-04-30 13:24 UTC (permalink / raw)
To: git
Now you can say
[remote.junio]
url = git://git.kernel.org/pub/scm/git/git.git
pull = next:next
in your .git/config.
Signed-off-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
---
git-parse-remote.sh | 55 +++++++++++++++++++++++++++++++++++++++++++++++----
1 files changed, 51 insertions(+), 4 deletions(-)
5a1cf349639823183fce287be9f809d797f2cd84
diff --git a/git-parse-remote.sh b/git-parse-remote.sh
index c9b899e..8ce57c8 100755
--- a/git-parse-remote.sh
+++ b/git-parse-remote.sh
@@ -4,13 +4,43 @@ # git-ls-remote could be called from out
# this would fail in that case and would issue an error message.
GIT_DIR=$(git-rev-parse --git-dir 2>/dev/null) || :;
+if [ -d "$GIT_DIR"/remotes -a "$GIT_REWRITE_REMOTES" = true ]; then
+ echo "Rewriting $GIT_DIR/remotes" >&2
+ error=0
+ # rewrite into config
+ {
+ cd "$GIT_DIR"/remotes
+ ls | while read f; do
+ name=$(echo -n "$f" | tr -c "A-Za-z0-9" ".")
+ sed -n \
+ -e "s/^URL: /remote.$name.url . /p" \
+ -e "s/^Pull: /remote.$name.pull ^$ /p" \
+ -e "s/^Push: /remote.$name.push ^$ /p" \
+ < "$f"
+ done
+ echo done
+ } | while read key regex value; do
+ case $key in
+ done)
+ if [ $error = 0 ]; then
+ mv "$GIT_DIR"/remotes "$GIT_DIR"/remotes.old
+ fi ;;
+ *)
+ git-repo-config $key "$value" $regex || error=1 ;;
+ esac
+ done
+fi
+
get_data_source () {
case "$1" in
*/*)
# Not so fast. This could be the partial URL shorthand...
token=$(expr "z$1" : 'z\([^/]*\)/')
remainder=$(expr "z$1" : 'z[^/]*/\(.*\)')
- if test -f "$GIT_DIR/branches/$token"
+ if test "$(git-repo-config --get "remote.$token.url")"
+ then
+ echo config-partial
+ elif test -f "$GIT_DIR/branches/$token"
then
echo branches-partial
else
@@ -18,7 +48,10 @@ get_data_source () {
fi
;;
*)
- if test -f "$GIT_DIR/remotes/$1"
+ if test "$(git-repo-config --get "remote.$1.url")"
+ then
+ echo config
+ elif test -f "$GIT_DIR/remotes/$1"
then
echo remotes
elif test -f "$GIT_DIR/branches/$1"
@@ -35,6 +68,15 @@ get_remote_url () {
case "$data_source" in
'')
echo "$1" ;;
+ config-partial)
+ token=$(expr "z$1" : '\([^/]*\)/')
+ remainder=$(expr "z$1" : '[^/]*/\(.*\)')
+ url=$(git-repo-config --get "remote.$token.url")
+ echo "$url/$remainder"
+ ;;
+ config)
+ git-repo-config --get "remote.$1.url"
+ ;;
remotes)
sed -ne '/^URL: */{
s///p
@@ -56,8 +98,10 @@ get_remote_url () {
get_remote_default_refs_for_push () {
data_source=$(get_data_source "$1")
case "$data_source" in
- '' | branches | branches-partial)
+ '' | config-partial | branches | branches-partial)
;; # no default push mapping, just send matching refs.
+ config)
+ git-repo-config --get-all "remote.$1.push" ;;
remotes)
sed -ne '/^Push: */{
s///p
@@ -111,8 +155,11 @@ # Returns list of src: (no store), or sr
get_remote_default_refs_for_fetch () {
data_source=$(get_data_source "$1")
case "$data_source" in
- '' | branches-partial)
+ '' | config-partial | branches-partial)
echo "HEAD:" ;;
+ config)
+ canon_refs_list_for_fetch \
+ $(git-repo-config --get-all "remote.$1.pull") ;;
branches)
remote_branch=$(sed -ne '/#/s/.*#//p' "$GIT_DIR/branches/$1")
case "$remote_branch" in '') remote_branch=master ;; esac
--
1.3.1.g38c00-dirty
^ permalink raw reply related
* [PATCH 3/3] fetch: optionally store the current remote information in the config
From: Johannes Schindelin @ 2006-04-30 13:24 UTC (permalink / raw)
To: git
Instead of editing files, you can now say
git pull --store junio \
git://git.kernel.org/pub/scm/git/git.git next:next
and next time, just
git pull junio
Signed-off-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
---
This is what the patch series is all about.
If there is no interest in a feature like this, let's just forget
about the whole "remote info in config" thing.
If there is interest, I could add the same functionality to
builtin-push.
Documentation/fetch-options.txt | 6 ++++++
git-fetch.sh | 19 +++++++++++++++++++
git-pull.sh | 8 ++++++--
3 files changed, 31 insertions(+), 2 deletions(-)
6bd937b0de211465e9664f8dc890fc5066617b73
diff --git a/Documentation/fetch-options.txt b/Documentation/fetch-options.txt
index 13f34d3..caf98de 100644
--- a/Documentation/fetch-options.txt
+++ b/Documentation/fetch-options.txt
@@ -16,6 +16,12 @@
fetches is a descendant of `<lbranch>`. This option
overrides that check.
+-S, \--store <nick>::
+ Store the URL and the refnames in the config file so that
+ `git fetch <nick>` repeats the exercise.
+ If the nick exists already, edit the URL, but append the
+ refnames.
+
\--no-tags::
By default, `git-fetch` fetches tags that point at
objects that are downloaded from the remote repository
diff --git a/git-fetch.sh b/git-fetch.sh
index 280f62e..ac122da 100755
--- a/git-fetch.sh
+++ b/git-fetch.sh
@@ -15,8 +15,10 @@ no_tags=
tags=
append=
force=
+keep=
verbose=
update_head_ok=
+store=
exec=
upload_pack=
while case "$#" in 0) break ;; esac
@@ -34,6 +36,10 @@ do
-f|--f|--fo|--for|--forc|--force)
force=t
;;
+ -S|--s|--st|--sto|--stor|--store)
+ store="$2"
+ shift
+ ;;
-t|--t|--ta|--tag|--tags)
tags=t
;;
@@ -235,6 +241,12 @@ then
fi
fi
+if test "$store"
+then
+ git-repo-config remote."$store".url $remote ||
+ die "Could not store into $store"
+fi
+
fetch_main () {
reflist="$1"
refs=
@@ -243,6 +255,11 @@ fetch_main () {
do
refs="$refs$LF$ref"
+ if test "$store"
+ then
+ git-repo-config remote."$store".pull "$ref" '^$'
+ fi
+
# These are relative path from $GIT_DIR, typically starting at refs/
# but may be HEAD
if expr "z$ref" : 'z\.' >/dev/null
@@ -381,6 +398,8 @@ fetch_main () {
fetch_main "$reflist"
+store=
+
# automated tag following
case "$no_tags$tags" in
'')
diff --git a/git-pull.sh b/git-pull.sh
index 4611ae6..ab0fba3 100755
--- a/git-pull.sh
+++ b/git-pull.sh
@@ -8,7 +8,7 @@ USAGE='[-n | --no-summary] [--no-commit]
LONG_USAGE='Fetch one or more remote refs and merge it/them into the current HEAD.'
. git-sh-setup
-strategy_args= no_summary= no_commit=
+strategy_args= no_summary= no_commit= store=
while case "$#,$1" in 0) break ;; *,-*) ;; *) break ;; esac
do
case "$1" in
@@ -31,6 +31,10 @@ do
esac
strategy_args="${strategy_args}-s $strategy "
;;
+ -S|--store)
+ store="-S $2"
+ shift
+ ;;
-h|--h|--he|--hel|--help)
usage
;;
@@ -43,7 +47,7 @@ do
done
orig_head=$(git-rev-parse --verify HEAD) || die "Pulling into a black hole?"
-git-fetch --update-head-ok "$@" || exit 1
+git-fetch --update-head-ok $store "$@" || exit 1
curr_head=$(git-rev-parse --verify HEAD)
if test "$curr_head" != "$orig_head"
--
1.3.1.g38c00-dirty
^ permalink raw reply related
* Re: cg-clone not fetching all tags?
From: Wolfgang Denk @ 2006-04-30 13:48 UTC (permalink / raw)
To: Junio C Hamano; +Cc: git, pasky
In-Reply-To: <7vaca31z50.fsf@assigned-by-dhcp.cox.net>
Dear Junio,
in message <7vaca31z50.fsf@assigned-by-dhcp.cox.net> you wrote:
>
> I suspect there is no need to manufacture the tag. Something
> like this should do.
Great! Thanks a lot.
> *WARNING* Since I do not do Porcelain, and I am not a Cogito
> user, this is obviously untested. In addition, I am seriously
> drunk right now...
Good enough. This one actually works:
> -- >8 --
> [PATCH] (cogito) Auto-follow lightweight tags as well.
--- a/cg-fetchG 2006-04-10 19:14:14.000000000 +0200
+++ b/cg-fetch 2006-04-30 14:35:03.000000000 +0200
@@ -211,7 +211,8 @@
git-ls-remote --tags "$uri" |
# SHA1 refs/tags/v0.99.8^{} --> SHA1 tags/v0.99.8
# where SHA1 is the object v0.99.8 tag points at.
- sed -ne 's:\([^ ]\) refs/\(tags/.*\)^{}$:\1 \2:p' |
+ sed -n -e 's:\([^ ]\) refs/\(tags/.*\)^{}$:\1 \2:p' \
+ -e 's:\([^ ]\) refs/\(tags/.*\)$:\1 \2:p' | \
while read sha1 tagname; do
# Do we have the tag itself?
[ -s "$_git/refs/$tagname" ] && continue
Best regards,
Wolfgang Denk
--
Software Engineering: Embedded and Realtime Systems, Embedded Linux
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd@denx.de
It's hard to think of you as the end result of millions of years of
evolution.
^ permalink raw reply
* Re: [PATCH 3/3] fetch: optionally store the current remote information in the config
From: Jakub Narebski @ 2006-04-30 14:19 UTC (permalink / raw)
To: git
In-Reply-To: <Pine.LNX.4.63.0604301524080.2646@wbgn013.biozentrum.uni-wuerzburg.de>
There was thread about storing somewhere default branch we merge to during
pull, instead of using always surrent one. Different schemes were proposed,
most of them depending on the remotes configuration being available [also]
in config file.
Perhaps it would be easiest to extend existing notation in the following
way:
<from>:<to>[:<merge>]
By the way: it would be nice to have command/script to trasform freely
between 'remotes/' and config file.
P.S. I wonder if it would be difficult to implement 'include <file>' for
config file...
--
Jakub Narebski
Warsaw, Poland
^ permalink raw reply
* Re: [PATCH 3/3] fetch: optionally store the current remote information in the config
From: sean @ 2006-04-30 14:30 UTC (permalink / raw)
To: Johannes Schindelin; +Cc: git
In-Reply-To: <Pine.LNX.4.63.0604301524080.2646@wbgn013.biozentrum.uni-wuerzburg.de>
On Sun, 30 Apr 2006 15:24:22 +0200 (CEST)
Johannes Schindelin <Johannes.Schindelin@gmx.de> wrote:
> Instead of editing files, you can now say
>
> git pull --store junio \
> git://git.kernel.org/pub/scm/git/git.git next:next
>
> and next time, just
>
> git pull junio
>
> Signed-off-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
>
> ---
>
> This is what the patch series is all about.
>
> If there is no interest in a feature like this, let's just forget
> about the whole "remote info in config" thing.
>
> If there is interest, I could add the same functionality to
> builtin-push.
Well I agree with you that doing something like this is important. We
should take this moment of moving things to the config file to correct
the terminology and help make things clear. We're not storing "Pull:"
information, we're storing config/remote.$NICK.fetch data. It's really
used just by fetch, pull just happens to call fetch.
Along that same line of reasoning, it seems more appropriate to use
git fetch --store ... rather than git pull --store ... to set this
information. And there needs to be a way to change and delete the
nick information, perhaps git fetch store junio "" would delete the
entry. Or maybe people should just be instructed to use git-repo-config
for setting, changing and deleting?
Pull needs additional logic that allows it to merge from the proper
local branch after it calls fetch. Right now it just uses whatever
fetch sets as FETCH_HEAD. It's not clear to me what is set as
FETCH_HEAD when multiple refs are fetched from the remote. It'll
be even more confusing once it's possible to fetch from multiple
remotes at once.
As for these specific patches, it doesn't appear that your change to
builtin-push allows the push variable to hold more than one remote
repo URI or even more than one refspec, or did I misread that?
Also it seems that the refspec is used from the config file even if
the user tries to override it by specifying an alternative on the
command line.
Sean
^ permalink raw reply
* Re: [RFC] [PATCH 0/5] Implement 'prior' commit object links (and
From: Jakub Narebski @ 2006-04-30 15:21 UTC (permalink / raw)
To: git
In-Reply-To: <e30k0n$ij5$1@sea.gmane.org>
Jakub Narebski wrote:
> Junio C Hamano wrote:
>
>> Jakub Narebski <jnareb@gmail.com> writes:
>>
>>> * "prior" - heads that represent topic branch merges
>>
>> This is not any different from usual "parent" at all (but you
>> have to think about it a bit to realize it).
> [cut]
> Thanks for an explanation.
>
> I would say that "prior" is not THAT different from usual "parent",
> rather than it is not ANY different.
>
> My doubts about recording previous head of a "union" (pu-like) branch
> is that for merge (e.g. 'pu' to 'next', cherrypick to/from 'pu', 'pu'
> rebase) is that for merge algorithm all parents are equivalent, with
> eventual exception of first which can be treated special ('ours').
Additionally with "prior" (or at least some convention on which of parents
is to prior head of "union (pu-like) branch) I think we could fast-forward
such branches...
--
Jakub Narebski
Warsaw, Poland
^ permalink raw reply
* Re: [PATCH 3/3] fetch: optionally store the current remote information in the config
From: Johannes Schindelin @ 2006-04-30 15:49 UTC (permalink / raw)
To: sean; +Cc: git
In-Reply-To: <BAYC1-PASMTP08069B2CE6005391A1AFF9AEB00@CEZ.ICE>
Hi,
On Sun, 30 Apr 2006, sean wrote:
> On Sun, 30 Apr 2006 15:24:22 +0200 (CEST)
> Johannes Schindelin <Johannes.Schindelin@gmx.de> wrote:
>
> > Instead of editing files, you can now say
> >
> > git pull --store junio \
> > git://git.kernel.org/pub/scm/git/git.git next:next
> >
> > and next time, just
> >
> > git pull junio
> >
> > Signed-off-by: Johannes Schindelin <Johannes.Schindelin@gmx.de>
> >
> > ---
> >
> > This is what the patch series is all about.
> >
> > If there is no interest in a feature like this, let's just forget
> > about the whole "remote info in config" thing.
> >
> > If there is interest, I could add the same functionality to
> > builtin-push.
>
> Well I agree with you that doing something like this is important. We
> should take this moment of moving things to the config file to correct
> the terminology and help make things clear. We're not storing "Pull:"
> information, we're storing config/remote.$NICK.fetch data. It's really
> used just by fetch, pull just happens to call fetch.
I have no strong feelings either way.
> Along that same line of reasoning, it seems more appropriate to use
> git fetch --store ... rather than git pull --store ... to set this
> information.
Both works.
> And there needs to be a way to change and delete the nick information,
> perhaps git fetch store junio "" would delete the entry. Or maybe
> people should just be instructed to use git-repo-config for setting,
> changing and deleting?
The latter should be done, because "git fetch" really is about fetching,
not playing games with the config.
> Pull needs additional logic that allows it to merge from the proper
> local branch after it calls fetch. Right now it just uses whatever
> fetch sets as FETCH_HEAD. It's not clear to me what is set as
> FETCH_HEAD when multiple refs are fetched from the remote. It'll
> be even more confusing once it's possible to fetch from multiple
> remotes at once.
FETCH_HEAD can contain multiple refs. And I don't get the part about
fetching from multiple remotes: my patch does not allow for that.
> As for these specific patches, it doesn't appear that your change to
> builtin-push allows the push variable to hold more than one remote
> repo URI or even more than one refspec, or did I misread that?
But it does! Note the "uri_[current_uri++]" part of the patch.
> Also it seems that the refspec is used from the config file even if
> the user tries to override it by specifying an alternative on the
> command line.
No. It is only used when there were no refspecs specified on the command
line:
if (refspec_nr == 0)
set_refspecs((const char**)refspecs_, current_refspec);
Ciao,
Dscho
^ permalink raw reply
* Re: [PATCH 3/3] fetch: optionally store the current remote information in the config
From: Johannes Schindelin @ 2006-04-30 15:52 UTC (permalink / raw)
To: Jakub Narebski; +Cc: git
In-Reply-To: <e32h0o$15n$1@sea.gmane.org>
Hi,
On Sun, 30 Apr 2006, Jakub Narebski wrote:
> There was thread about storing somewhere default branch we merge to during
> pull, instead of using always surrent one. Different schemes were proposed,
> most of them depending on the remotes configuration being available [also]
> in config file.
I was not following that thread closely, since it became too confusing for
me. However, I think that my patch could be a start in that direction.
> By the way: it would be nice to have command/script to trasform freely
> between 'remotes/' and config file.
If you set the environment variable GIT_REWRITE_REMOTES to "true", and
call git-parse-remotes.sh, it will do the rewriting to the config file.
Obviously, I did not test that part of the patch all that well.
> P.S. I wonder if it would be difficult to implement 'include <file>' for
> config file...
You really need that?
Ciao,
Dscho
^ permalink raw reply
* Re: [PATCH 3/3] fetch: optionally store the current remote information in the config
From: Jakub Narebski @ 2006-04-30 16:07 UTC (permalink / raw)
To: git
In-Reply-To: <Pine.LNX.4.63.0604301749130.3641@wbgn013.biozentrum.uni-wuerzburg.de>
Johannes Schindelin wrote:
> On Sun, 30 Apr 2006, Jakub Narebski wrote:
>
>> P.S. I wonder if it would be difficult to implement 'include <file>' for
>> config file...
>
> You really need that?
Need? Not exactly. I don't think git ever reach complexity of Apache or
Samba configuration files, and _need_ for includes. Still dividing separate
areas of configuration (core, user, default commands options, remotes) has
it's merits.
--
Jakub Narebski
Warsaw, Poland
^ permalink raw reply
* Re: [PATCH 3/3] fetch: optionally store the current remote information in the config
From: sean @ 2006-04-30 16:37 UTC (permalink / raw)
To: Johannes Schindelin; +Cc: git
In-Reply-To: <Pine.LNX.4.63.0604301743370.3641@wbgn013.biozentrum.uni-wuerzburg.de>
On Sun, 30 Apr 2006 17:49:06 +0200 (CEST)
Johannes Schindelin <Johannes.Schindelin@gmx.de> wrote:
> > the terminology and help make things clear. We're not storing "Pull:"
> > information, we're storing config/remote.$NICK.fetch data. It's really
> > used just by fetch, pull just happens to call fetch.
>
> I have no strong feelings either way.
Yeah, once you "get" it, it's not a problem; but it's not easy when you're
just learning git to separate fetch and pull. It's made harder if git
can't even keep them straight internally. :o/
[...]
> The latter should be done, because "git fetch" really is about fetching,
> not playing games with the config.
Then we should also remove the --store option from pull and fetch. It
can be set with git-repo-config.
> FETCH_HEAD can contain multiple refs.
Which head does git-pull then use to merge, all of them?
> And I don't get the part about fetching from multiple remotes:
> my patch does not allow for that.
Actually it does :o) User just needs multiple remote.$nick.url entries
in his config.
> But it does! Note the "uri_[current_uri++]" part of the patch.
[...]
> No. It is only used when there were no refspecs specified on the command
> line:
>
> if (refspec_nr == 0)
> set_refspecs((const char**)refspecs_, current_refspec);
Right you are, on both counts.
Sean
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox