Re: [PATCH] request-pull: resurrect for-linus - tags/for-linus DWIM
On Fri, May 16, 2014 at 10:57:50AM -0700, Junio C Hamano wrote: Older versions of Git before v1.7.10 did not DWIM $ git pull $URL for-linus to the tag tags/for-linus and the users were required to say $ git pull $URL tags/for-linus instead. Because newer versions of Git works either way, request-pull used to show tags/for-linus when asked $ git request-pull origin/master $URL for-linus The recent updates broke this and in the output we see for-linus without the tags/ prefix. As v1.7.10 is more than 2 years old, this should matter very little in practice, but resurrecting it is very simple. Well RHEL6 apparently comes with git 1.7.1, there are probably others. The problem isn't theorectical actually, the reason I noticed the change in behaviour is because one of my pull requests got bounced because of it: http://article.gmane.org/gmane.comp.emulators.qemu/273121 Signed-off-by: Junio C Hamano gits...@pobox.com Tested-by: Michael S. Tsirkin m...@redhat.com --- * I _think_ the fix, without breaking the spirit of Linus's I do not want the thing DWIM based on what the remote end has original, would be as simple as this patch. We can queue it as a regression fix and do another round of -rc4 if those who depend on request-pull heavily feel strongly about it. git-request-pull.sh | 6 ++ t/t5150-request-pull.sh | 8 +++- 2 files changed, 13 insertions(+), 1 deletion(-) diff --git a/git-request-pull.sh b/git-request-pull.sh index b67513a..d6648b2 100755 --- a/git-request-pull.sh +++ b/git-request-pull.sh @@ -118,6 +118,12 @@ then status=1 fi +# Special case: turn for_linus to tags/for_linus when it is correct +if test $ref = refs/tags/$pretty_remote +then + pretty_remote=tags/$pretty_remote +fi + url=$(git ls-remote --get-url $url) git show -s --format='The following changes since commit %H: diff --git a/t/t5150-request-pull.sh b/t/t5150-request-pull.sh index 75d6b38..93e2c65 100755 --- a/t/t5150-request-pull.sh +++ b/t/t5150-request-pull.sh @@ -223,7 +223,13 @@ test_expect_success 'pull request format' ' git request-pull initial $downstream_url tags/full:refs/tags/full ) request sed -nf fuzz.sed request request.fuzzy - test_i18ncmp expect request.fuzzy + test_i18ncmp expect request.fuzzy + + ( + cd local + git request-pull initial $downstream_url full + ) request + grep ' tags/full$' ' test_expect_success 'request-pull ignores OPTIONS_KEEPDASHDASH poison' ' -- 2.0.0-rc3-434-g1ba2fe8 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH v3] format-patch --signature-file file
On Sat, May 17, 2014 at 09:02:22AM -0700, Jeremiah Mahler wrote:
Added feature that allows a signature file to be used with format-patch.
$ git format-patch --signature-file ~/.signature -1
Now signatures with newlines and other special characters can be
easily included.
Signed-off-by: Jeremiah Mahler jmmah...@gmail.com
This is looking much better. I have a few comments still, inline below.
By the way, did you want to add a format.signaturefile config, too? I do
not care myself, but I would imagine most workflows would end up
specifying it every time.
@@ -1447,6 +1450,16 @@ int cmd_format_patch(int argc, const char **argv,
const char *prefix)
cover_letter = (config_cover_letter == COVER_ON);
}
+ if (signature_file) {
+ if (signature signature != git_version_string)
+ die(_(--signature and --signature-file are mutually
exclusive));
Technically signature might have come from config, not --signature
on the command-line. But I don't know if that's even worth worrying
about; presumably the user can figure it out if they set the config.
+ struct strbuf buf = STRBUF_INIT;
+
+ strbuf_read_file(buf, signature_file, 128);
+ signature = strbuf_detach(buf, NULL);
Your cover letter mentioned generating an error here. Did you want:
if (strbuf_read_file(buf, signature_file, 128) 0)
die_errno(unable to read signature file '%s', signature_file);
or similar?
diff --git a/t/t4014-format-patch.sh b/t/t4014-format-patch.sh
index 9c80633..fb3dc1b 100755
--- a/t/t4014-format-patch.sh
+++ b/t/t4014-format-patch.sh
@@ -762,6 +762,34 @@ test_expect_success 'format-patch --signature=
suppresses signatures' '
! grep ^-- \$ output
'
+cat expect EOF
Minor style nits, but we usually omit the whitespace between redirection
operations, and we always quote our here-doc endings unless they
explicitly want to interpolate. So:
cat expect \EOF
(we also tend to use -\EOF to drop leading tabs, and then include
them inside the test_expect_success properly indented, but as this
expectation is used in multiple places, it's not unreasonable to keep it
separate).
+test_expect_success 'format-patch --signature-file file' '
+ git format-patch --stdout --signature-file expect -1 output
+ check_patch output
+ fgrep -x -f output expect output2
Both of these fgrep options are in POSIX, but it looks like this will be
the first use for either of them. I'm not sure if they will give us any
portability problems.
We could probably do something like:
sed -n '/^-- $/,$p'
if we have to.
+ diff expect output2
Please use test_cmp here, which adjusts automatically for less-abled
systems where diff is not available.
+test_expect_success 'format-patch --signature-file=file' '
+ git format-patch --stdout --signature-file=expect -1 output
+ check_patch output
+ fgrep -x -f output expect output2
+ diff expect output2
+'
Same comments as above; I note that this is just checking --foo=bar
rather than --foo bar. I don't mind being thorough, but for the most
part we just assume this is tested as part of the parse-options tests,
and don't check it explicitly for each option.
+test_expect_success 'format-patch --signature and --signature-file die' '
+ ! git format-patch --stdout --signature=foo --signature-file=expect
-1 output
+'
Please use test_must_fail instead of ! here; it is more thorough in
checking that we exited with a non-zero error code (and didn't die from
signal death, for example).
-Peff
--
To unsubscribe from this list: send the line unsubscribe git in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] remote-helpers: point at their upstream repositories
Felipe Contreras felipe.contre...@gmail.com writes:
% git fetch
WARNING: git-remote-hg is now maintained independently.
WARNING: For more information visit https://github.com/felipec/git-remote-hg
searching for changes
no changes found
I don't think the situation is as simple as you claim. In many cases,
the first step before the ones you are mentionning are:
cd $git/contrib/remote-helpers
cp git-remote-{hg,bzr} somewhere/in/path
They produces no warning if git-remote-{hg,bzr} exist with the warning,
but no such file or directory: contrib/remote-helpers if the directory
has been renamed or removed.
When git-remote-{hg,bzr} are installed with a package manager, the fact
that they are part of Git's core or not is often irrelevant. For
example, Debian splits the git.git source into many packages, so a
Debian user will not see any difference between helpers included in
git.git or outside (e.g. I have to install the package git-svn if I want
to use git-svn).
That said, I'm fine with the add a warning option too.
--
Matthieu Moy
http://www-verimag.imag.fr/~moy/
--
To unsubscribe from this list: send the line unsubscribe git in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH v3] format-patch --signature-file file
Peff, Lots of good suggestions. I had to expand upon the signature pattern match to get to work. On Sun, May 18, 2014 at 07:20:20AM -0400, Jeff King wrote: On Sat, May 17, 2014 at 09:02:22AM -0700, Jeremiah Mahler wrote: ... +test_expect_success 'format-patch --signature-file file' ' + git format-patch --stdout --signature-file expect -1 output + check_patch output + fgrep -x -f output expect output2 Both of these fgrep options are in POSIX, but it looks like this will be the first use for either of them. I'm not sure if they will give us any portability problems. We could probably do something like: sed -n '/^-- $/,$p' This gets the signature out but it will have '--' and some trailing blank lines which were not in the original signature. So then test_cmp won't work directly. What I came up with was to use head and tail to remove the first line and the last two lines. Then test_cmp can be used normally. test_expect_success 'format-patch --signature-file=file' ' git format-patch --stdout --signature-file=expect -1 output check_patch_output output sed -n /^-- $/,\$p output | head --lines=-2 | tail --lines=+2 output2 test_cmp expect output2 ' -- Jeremiah Mahler jmmah...@gmail.com http://github.com/jmahler -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[RFC/PATCH] replace: add --graft option
The usage string for this option is:
git replace [-f] --graft commit [parent...]
First we create a new commit that is the same as commit
except that its parents are [parents...]
Then we create a replace ref that replace commit with
the commit we just created.
With this new option, it should be straightforward to
convert grafts to replace refs, with something like:
cat .git/info/grafts | while read line
do git replace --graft $line; done
Signed-off-by: Christian Couder chrisc...@tuxfamily.org
---
builtin/replace.c | 79 ++-
1 file changed, 78 insertions(+), 1 deletion(-)
diff --git a/builtin/replace.c b/builtin/replace.c
index 1bb491d..4b3705d 100644
--- a/builtin/replace.c
+++ b/builtin/replace.c
@@ -17,6 +17,7 @@
static const char * const git_replace_usage[] = {
N_(git replace [-f] object replacement),
N_(git replace [-f] --edit object),
+ N_(git replace [-f] --graft commit [parent...]),
N_(git replace -d object...),
N_(git replace [--format=format] [-l [pattern]]),
NULL
@@ -294,6 +295,71 @@ static int edit_and_replace(const char *object_ref, int
force)
return replace_object_sha1(object_ref, old, replacement, new, force);
}
+static int read_sha1_commit(const unsigned char *sha1, struct strbuf *dst)
+{
+ void *buf;
+ enum object_type type;
+ unsigned long size;
+
+ buf = read_sha1_file(sha1, type, size);
+ if (!buf)
+ return error(_(cannot read object %s), sha1_to_hex(sha1));
+ if (type != OBJ_COMMIT) {
+ free(buf);
+ return error(_(object %s is not a commit), sha1_to_hex(sha1));
+ }
+ strbuf_attach(dst, buf, size, size + 1);
+ return 0;
+}
+
+static int create_graft(int argc, const char **argv, int force)
+{
+ unsigned char old[20], new[20];
+ const char *old_ref = argv[0];
+ struct strbuf buf = STRBUF_INIT;
+ struct strbuf new_parents = STRBUF_INIT;
+ const char *parent_start, *parent_end;
+ int i;
+
+ if (get_sha1(old_ref, old) 0)
+ die(_(Not a valid object name: '%s'), old_ref);
+ lookup_commit_or_die(old, old_ref);
+ if (read_sha1_commit(old, buf))
+ die(_(Invalid commit: '%s'), old_ref);
+
+ /* find existing parents */
+ parent_start = buf.buf;
+ parent_start += 46; /* tree + hex sha1 + \n */
+ parent_end = parent_start;
+
+ while (starts_with(parent_end, parent ))
+ parent_end += 48; /* parent + hex sha1 + \n */
+
+ /* prepare new parents */
+ for (i = 1; i argc; i++) {
+ unsigned char sha1[20];
+ if (get_sha1(argv[i], sha1) 0)
+ die(_(Not a valid object name: '%s'), argv[i]);
+ lookup_commit_or_die(sha1, argv[i]);
+ strbuf_addf(new_parents, parent %s\n, sha1_to_hex(sha1));
+ }
+
+ /* replace existing parents with new ones */
+ strbuf_splice(buf, parent_start - buf.buf, parent_end - parent_start,
+ new_parents.buf, new_parents.len);
+
+ if (write_sha1_file(buf.buf, buf.len, commit_type, new))
+ die(_(could not write replacement commit for: '%s'), old_ref);
+
+ strbuf_release(new_parents);
+ strbuf_release(buf);
+
+ if (!hashcmp(old, new))
+ return error(new commit is the same as the old one: '%s',
sha1_to_hex(old));
+
+ return replace_object_sha1(old_ref, old, replacement, new, force);
+}
+
int cmd_replace(int argc, const char **argv, const char *prefix)
{
int force = 0;
@@ -303,12 +369,14 @@ int cmd_replace(int argc, const char **argv, const char
*prefix)
MODE_LIST,
MODE_DELETE,
MODE_EDIT,
+ MODE_GRAFT,
MODE_REPLACE
} cmdmode = MODE_UNSPECIFIED;
struct option options[] = {
OPT_CMDMODE('l', list, cmdmode, N_(list replace refs),
MODE_LIST),
OPT_CMDMODE('d', delete, cmdmode, N_(delete replace refs),
MODE_DELETE),
OPT_CMDMODE('e', edit, cmdmode, N_(edit existing object),
MODE_EDIT),
+ OPT_CMDMODE('g', graft, cmdmode, N_(change a commit's
parents), MODE_GRAFT),
OPT_BOOL('f', force, force, N_(replace the ref if it
exists)),
OPT_STRING(0, format, format, N_(format), N_(use this
format)),
OPT_END()
@@ -325,7 +393,10 @@ int cmd_replace(int argc, const char **argv, const char
*prefix)
usage_msg_opt(--format cannot be used when not listing,
git_replace_usage, options);
- if (force cmdmode != MODE_REPLACE cmdmode != MODE_EDIT)
+ if (force
+ cmdmode != MODE_REPLACE
+ cmdmode != MODE_EDIT
+ cmdmode != MODE_GRAFT)
usage_msg_opt(-f only makes sense when writing a
Re: [GUILT v3 14/31] Use git check-ref-format to validate patch names.
On Fri, May 16, 2014 at 5:20 PM, Jeff Sipek jef...@josefsipek.net wrote:
On Fri, May 16, 2014 at 04:46:01PM +0200, Per Cederqvist wrote:
The valid_patchname now lets git check-ref-format do its job instead
of trying (and failing) to implement the same rules. See
git-check-ref-format(1) for a list of the rules.
Refer to the git-check-ref-format(1) man page in the error messages
produced when valid_patchname indicates that the name is bad.
Added testcases that breaks most of the rules in that man-page.
Git version 1.8.5 no longer allows the single character @ as a
branch name. Guilt always rejects that name, for increased
compatibility.
Signed-off-by: Per Cederqvist ced...@opera.com
Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net
---
guilt| 21 ++-
guilt-fork | 2 +-
guilt-import | 2 +-
guilt-new| 2 +-
regression/t-025.out | 426
+--
regression/t-025.sh | 12 +-
regression/t-032.out | 4 +-
7 files changed, 446 insertions(+), 23 deletions(-)
diff --git a/guilt b/guilt
index 3fc524e..23cc2da 100755
--- a/guilt
+++ b/guilt
@@ -132,14 +132,19 @@ fi
# usage: valid_patchname patchname
valid_patchname()
{
- case $1 in
- /*|./*|../*|*/./*|*/../*|*/.|*/..|*/|*\ *|*\*)
- return 1;;
- *:*)
- return 1;;
- *)
- return 0;;
- esac
+ if git check-ref-format --allow-onelevel $1; then
I know I already signed off on this, but I just tried to run the regression
suite with your changes on 1.7.3.2. It blows up because --allow-onelevel is
not recognized. Yes, 1.7 is a bit on the old side but I think it's
reasonable to still support it. Thoughts?
Oh. I was not aware that --allow-onelevel was so new (it was added in
Git 1.7.8).
I'll rewrite this using git check-ref-format a/$1 instead.
+ # Starting with Git version 1.8.5, a branch cannot be
+ # the single character @. Make sure guilt rejects
+ # that name even if we are currently using an older
+ # version of Git. This ensures that the test suite
+ # runs fine using any version of Git.
+ if [ $1 = @ ]; then
+ return 1
+ fi
+ return 0
There were other changes in Git version 1.7.6.4 and 1.7.8. I'll
add similar code for those as well in the upcoming v4 of the series.
With those changes, the test suite works with Git 1.5.0 and newer
(I've checked all versions made of three numbers, but not those
made of four numbers.)
Jeff, I'll remove your signed-off-by so that you can check the
new code that I will post within a couple of days.
/ceder
+ else
+ return 1
+ fi
}
get_branch()
diff --git a/guilt-fork b/guilt-fork
index a85d391..6447e55 100755
--- a/guilt-fork
+++ b/guilt-fork
@@ -37,7 +37,7 @@ else
fi
if ! valid_patchname $newpatch; then
- die The specified patch name contains invalid characters (:).
+ die The specified patch name is invalid according to
git-check-ref-format(1).
fi
if [ -e $GUILT_DIR/$branch/$newpatch ]; then
diff --git a/guilt-import b/guilt-import
index 3e9b3bb..928e325 100755
--- a/guilt-import
+++ b/guilt-import
@@ -40,7 +40,7 @@ if [ -e $GUILT_DIR/$branch/$newname ]; then
fi
if ! valid_patchname $newname; then
- die The specified patch name contains invalid characters (:).
+ die The specified patch name is invalid according to
git-check-ref-format(1).
fi
# create any directories as needed
diff --git a/guilt-new b/guilt-new
index 9528438..9f7fa44 100755
--- a/guilt-new
+++ b/guilt-new
@@ -64,7 +64,7 @@ fi
if ! valid_patchname $patch; then
disp Patchname is invalid. 2
- die it cannot begin with '/', './' or '../', or contain /./, /../, or
whitespace
+ die It must follow the rules in git-check-ref-format(1).
fi
# create any directories as needed
diff --git a/regression/t-025.out b/regression/t-025.out
index 7811ab1..01bc406 100644
--- a/regression/t-025.out
+++ b/regression/t-025.out
@@ -141,7 +141,7 @@ f da39a3ee5e6b4b0d3255bfef95601890afd80709
.git/patches/master/prepend
r acdeef96ee30eb34bbbf65d11de5cf7da4b5fee8 .git/refs/patches/master/prepend
% guilt new white space
Patchname is invalid.
-it cannot begin with '/', './' or '../', or contain /./, /../, or whitespace
+It must follow the rules in git-check-ref-format(1).
% list_files
d .git/patches
d .git/patches/master
@@ -211,7 +211,7 @@ f da39a3ee5e6b4b0d3255bfef95601890afd80709
.git/patches/master/prepend
r acdeef96ee30eb34bbbf65d11de5cf7da4b5fee8 .git/refs/patches/master/prepend
% guilt new /abc
Patchname is invalid.
-it cannot begin with '/', './' or '../', or contain /./, /../, or whitespace
+It must follow the rules in git-check-ref-format(1).
Re: [GUILT v3 09/31] Test suite: properly check the exit status of commands.
On Fri, May 16, 2014 at 5:45 PM, Jeff Sipek jef...@josefsipek.net wrote:
On Fri, May 16, 2014 at 04:45:56PM +0200, Per Cederqvist wrote:
The cmd and shouldfail functions checked the exit status of the
replace_path function instead of the actual command that was running.
(The $? construct checks the exit status of the last command in a
pipeline, not the first command.)
Updated t-032.sh, which used shouldfail instead of cmd in one
place. (The comment in the script makes it clear that the command is
expected to succeed.)
Signed-off-by: Per Cederqvist ced...@opera.com
---
regression/scaffold | 17 +++--
regression/t-032.sh | 2 +-
2 files changed, 12 insertions(+), 7 deletions(-)
diff --git a/regression/scaffold b/regression/scaffold
index 5c8b73e..e4d7487 100644
--- a/regression/scaffold
+++ b/regression/scaffold
@@ -51,18 +51,23 @@ function filter_dd
function cmd
{
echo % $@
- $@ 21 | replace_path return 0
- return 1
+ (
+ exec 31
+ rv=`(($@ 21; echo $? 4) | replace_path 3 ) 41`
Wow. This took a while to decipher :)
Ancien wisdom from the Csh Programming Considered Harmful
article: http://www.faqs.org/faqs/unix-faq/shell/csh-whynot/
These functions work only because of the set -e earlier in scaffold.
The final return statements are not actually reached. I don't like that.
So the next version of the patch series will print an explicit message
like % FAIL: The above command should succeed but failed. or
% FAIL: The above command should fail but succeeded. and do
an explicit exit 1 on failure. I think it makes it easier to debug issues.
(I recently spent a few hours trying to figure our why the test just silently
exited. Turns out the old git version I was running didn't like my .gitconfig,
so it exited with a non-zero exit code...)
Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net
I'll let you re-check the next version of the code.
/ceder
+ exit $rv
+ )
+ return $?
}
# usage: shouldfail cmd..
function shouldfail
{
echo % $@
- (
- $@ 21 || return 0
- return 1
- ) | replace_path
+ ! (
+ exec 31
+ rv=`(($@ 21; echo $? 4) | replace_path 3 ) 41`
+ exit $rv
+ )
return $?
}
diff --git a/regression/t-032.sh b/regression/t-032.sh
index b1d5f19..bba401e 100755
--- a/regression/t-032.sh
+++ b/regression/t-032.sh
@@ -28,7 +28,7 @@ shouldfail guilt import -P foo3 foo
cmd guilt import -P foo2 foo
# ok
-shouldfail guilt import foo
+cmd guilt import foo
# duplicate patch name (implicit)
shouldfail guilt import foo
--
1.8.3.1
--
Fact: 28.1% of all statistics are generated randomly.
--
To unsubscribe from this list: send the line unsubscribe git in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH v4] format-patch --signature-file file
Added option that allows a signature file to be used with format-patch
so that signatures with newlines and other special characters can be
easily included.
$ git format-patch --signature-file ~/.signature -1
The config variable format.signaturefile is also provided so that it
can be added by default.
$ git config format.signaturefile ~/.signature
$ git format-patch -1
Signed-off-by: Jeremiah Mahler jmmah...@gmail.com
---
Documentation/config.txt | 4
Documentation/git-format-patch.txt | 4
builtin/log.c | 16
t/t4014-format-patch.sh| 33 +
4 files changed, 57 insertions(+)
diff --git a/Documentation/config.txt b/Documentation/config.txt
index 1932e9b..140ed77 100644
--- a/Documentation/config.txt
+++ b/Documentation/config.txt
@@ -1114,6 +1114,10 @@ format.signature::
Set this variable to the empty string () to suppress
signature generation.
+format.signaturefile::
+ Works just like format.signature except the contents of the
+ file specified by this variable will be used as the signature.
+
format.suffix::
The default for format-patch is to output files with the suffix
`.patch`. Use this variable to change that suffix (make sure to
diff --git a/Documentation/git-format-patch.txt
b/Documentation/git-format-patch.txt
index 5c0a4ab..c0fd470 100644
--- a/Documentation/git-format-patch.txt
+++ b/Documentation/git-format-patch.txt
@@ -14,6 +14,7 @@ SYNOPSIS
[(--attach|--inline)[=boundary] | --no-attach]
[-s | --signoff]
[--signature=signature | --no-signature]
+ [--signature-file=file]
[-n | --numbered | -N | --no-numbered]
[--start-number n] [--numbered-files]
[--in-reply-to=Message-Id] [--suffix=.sfx]
@@ -233,6 +234,9 @@ configuration options in linkgit:git-notes[1] to use this
workflow).
signature option is omitted the signature defaults to the Git version
number.
+--signature-file=file::
+ Works just like --signature except the signature is read from a file.
+
--suffix=.sfx::
Instead of using `.patch` as the suffix for generated
filenames, use specified suffix. A common alternative is
diff --git a/builtin/log.c b/builtin/log.c
index 39e8836..0a8f417 100644
--- a/builtin/log.c
+++ b/builtin/log.c
@@ -673,6 +673,7 @@ static void add_header(const char *value)
static int thread;
static int do_signoff;
static const char *signature = git_version_string;
+static const char *signature_file;
static int config_cover_letter;
enum {
@@ -742,6 +743,8 @@ static int git_format_config(const char *var, const char
*value, void *cb)
}
if (!strcmp(var, format.signature))
return git_config_string(signature, var, value);
+ if (!strcmp(var, format.signaturefile))
+ return git_config_string(signature_file, var, value);
if (!strcmp(var, format.coverletter)) {
if (value !strcasecmp(value, auto)) {
config_cover_letter = COVER_AUTO;
@@ -1230,6 +1233,8 @@ int cmd_format_patch(int argc, const char **argv, const
char *prefix)
PARSE_OPT_OPTARG, thread_callback },
OPT_STRING(0, signature, signature, N_(signature),
N_(add a signature)),
+ OPT_FILENAME(0, signature-file, signature_file,
+ N_(add a signature from a file)),
OPT__QUIET(quiet, N_(don't print the patch filenames)),
OPT_END()
};
@@ -1447,6 +1452,17 @@ int cmd_format_patch(int argc, const char **argv, const
char *prefix)
cover_letter = (config_cover_letter == COVER_ON);
}
+ if (signature_file) {
+ if (signature signature != git_version_string)
+ die(_(cannot specify both signature and
signature-file));
+
+ struct strbuf buf = STRBUF_INIT;
+
+ if (strbuf_read_file(buf, signature_file, 128) 0)
+ die_errno(_(unable to read signature file '%s'),
signature_file);
+ signature = strbuf_detach(buf, NULL);
+ }
+
if (in_reply_to || thread || cover_letter)
rev.ref_message_ids = xcalloc(1, sizeof(struct string_list));
if (in_reply_to) {
diff --git a/t/t4014-format-patch.sh b/t/t4014-format-patch.sh
index 9c80633..c6d44b8 100755
--- a/t/t4014-format-patch.sh
+++ b/t/t4014-format-patch.sh
@@ -762,6 +762,39 @@ test_expect_success 'format-patch --signature=
suppresses signatures' '
! grep ^-- \$ output
'
+cat expect -\EOF
+
+Test User test.em...@kernel.org
+http://git.kernel.org/cgit/git/git.git
+
+git.kernel.org/?p=git/git.git;a=summary
+
+EOF
+
+test_expect_success 'format-patch --signature-file=file' '
+
[PATCH v4] format-patch --signature-file file
v4 of patch to add format-patch --signature-file file option. This revision includes more suggestions from Jeff King: - Added a format.signaturefile config option. - Fixed error messages, --signature and format.signature - Fixed error checking, dies if strbuf_read_file fails. - style: cat expect EOF to cat expect -\EOF - Eliminated use of fgrep, which is not used anywhere else in the project. Possibility of portability problems. Replaced with sed which is commonly used. - Eliminated use of diff in favor of test_cmp. - Removed redundant test, --foo=bar and --foo bar - Changed to use test_must_fail instead of ! Jeremiah Mahler (1): format-patch --signature-file file Documentation/config.txt | 4 Documentation/git-format-patch.txt | 4 builtin/log.c | 16 t/t4014-format-patch.sh| 33 + 4 files changed, 57 insertions(+) -- Jeremiah Mahler jmmah...@gmail.com http://github.com/jmahler -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH] rebase --keep-empty -i: add test
There's some special code in rebase -i to deal with --keep-empty. Add test for this combination. Signed-off-by: Michael S. Tsirkin m...@redhat.com --- t/t3404-rebase-interactive.sh | 9 + 1 file changed, 9 insertions(+) diff --git a/t/t3404-rebase-interactive.sh b/t/t3404-rebase-interactive.sh index c0023a5..3b1b863 100755 --- a/t/t3404-rebase-interactive.sh +++ b/t/t3404-rebase-interactive.sh @@ -67,6 +67,14 @@ test_expect_success 'setup' ' SHELL= export SHELL +test_expect_success 'rebase --keep-empty' ' + git checkout -b emptybranch master + git commit --allow-empty -m empty + git rebase --keep-empty -i HEAD~2 + git log --oneline actual + test_line_count = 6 actual +' + test_expect_success 'rebase -i with the exec command' ' git checkout master ( -- MST -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH 2/4] git-rebase: document ack
document ack! behaviour and use Signed-off-by: Michael S. Tsirkin m...@redhat.com --- Documentation/git-rebase.txt | 45 +++- 1 file changed, 40 insertions(+), 5 deletions(-) diff --git a/Documentation/git-rebase.txt b/Documentation/git-rebase.txt index 2a93c64..c27aef4 100644 --- a/Documentation/git-rebase.txt +++ b/Documentation/git-rebase.txt @@ -384,7 +384,7 @@ or by giving more than one `--exec`: + If `--autosquash` is used, exec lines will not be appended for the intermediate commits, and will only appear at the end of each -squash/fixup series. +squash/fixup/ack series. --root:: Rebase all commits reachable from branch, instead of @@ -398,13 +398,13 @@ squash/fixup series. --autosquash:: --no-autosquash:: - When the commit log message begins with squash! ... (or - fixup! ...), and there is a commit whose title begins with + When the commit log message begins with squash! ... (fixup! ... + or ack! ...), and there is a commit whose title begins with the same ..., automatically modify the todo list of rebase -i so that the commit marked for squashing comes right after the commit to be modified, and change the action of the moved - commit from `pick` to `squash` (or `fixup`). Ignores subsequent - fixup! or squash! after the first, in case you referred to an + commit from `pick` to `squash` (`fixup` or `ack`). Ignores subsequent + ack! , fixup! or squash! after the first, in case you referred to an earlier fixup/squash with `git commit --fixup/--squash`. + This option is only valid when the '--interactive' option is used. @@ -624,6 +624,41 @@ consistent (they compile, pass the testsuite, etc.) you should use 'git stash' to stash away the not-yet-committed changes after each commit, test, and amend the commit if fixes are necessary. + +RECORDING ACKS + + +Interactive mode with --autosquash can be used to concatenate +commit log for several commits, which is useful to record +extra information about the commit, such as ack signatures. +This allows, for example, the following workflow: + +1. receive patches by mail and commit +2. receive by mail ack signatures for the patches +3. prepare a series for submission +4. submit + +where point 2. consists of several instances of + i) create a (possibly empty) commit with signature + in the commit message + +Sometimes the ack signature added in i. cannot be amended to the +commit it acks, because that commit is buried deeply in a +patch series. That is exactly what rebase --autosquash +option is for: use it +after plenty of is, to automaticlly rearrange +commits, and squashing multiple sign-off commits into +the commit that is signed. + +Start it with the last commit you want to retain as-is: + + git rebase --autosquash -i after-this-commit + +An editor will be fired up with all the commits in your current branch +which come after the given commit. Ack commits will be +re-arranged to come after the commit that is acked, +and the action will be utomticlly changed from `pick` to `ack` +to cause them to be squashed into the acked commit. RECOVERING FROM UPSTREAM REBASE --- -- MST -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH 3/4] rebase: test ack
test ack! handling Signed-off-by: Michael S. Tsirkin m...@redhat.com --- t/t3415-rebase-autosquash.sh | 15 +++ 1 file changed, 15 insertions(+) diff --git a/t/t3415-rebase-autosquash.sh b/t/t3415-rebase-autosquash.sh index 41370ab..9d7db13 100755 --- a/t/t3415-rebase-autosquash.sh +++ b/t/t3415-rebase-autosquash.sh @@ -74,6 +74,21 @@ test_expect_success 'auto squash (option)' ' test_auto_squash final-squash --autosquash ' +test_expect_success 'auto ack' ' + ack=Acked-by: xyz + msg=$(test_write_lines ack! first commit $ack) + git reset --hard base + git commit --allow-empty -m $msg -- + git tag ack + test_tick + git rebase --autosquash -i HEAD^^^ + git log --oneline actual + git show -s first-commit | grep -v ^commit expected-msg + echo $ack expected-msg + git show -s HEAD^ | grep -v ^commit actual-msg + diff actual-msg expected-msg +' + test_expect_success 'auto squash (config)' ' git config rebase.autosquash true test_auto_squash final-squash-config-true -- MST -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH 1/4] rebase -i: add ack action
This implements a new ack! action for git rebase -i
It is essentially a middle ground between fixup! and squash!:
- commits are squashed silently without editor being started
- commit logs are concatenated (with action line being discarded)
- because of the above, empty commits aren't discarded,
their log is also included.
I am using it as follows:
git am -s mailbox #creates first commit
hack ...
get mail with Ack
git commit --allow-empty -m `cat -EOF
ack! first
Acked-by: maintainer
EOF`
repeat cycle
git rebase --autosquash -i origin/master
before public branch push
The cat command above is actually a script that
parses the Ack mail to create the empty commit -
to be submitted separately.
Signed-off-by: Michael S. Tsirkin m...@redhat.com
---
git-rebase--interactive.sh | 34 +++---
1 file changed, 27 insertions(+), 7 deletions(-)
diff --git a/git-rebase--interactive.sh b/git-rebase--interactive.sh
index 6ec9d3c..821872c 100644
--- a/git-rebase--interactive.sh
+++ b/git-rebase--interactive.sh
@@ -140,6 +140,7 @@ Commands:
r, reword = use commit, but edit the commit message
e, edit = use commit, but stop for amending
s, squash = use commit, but meld into previous commit
+ a, ack = like squash, but append commit body only to previous commit
f, fixup = like squash, but discard this commit's log message
x, exec = run command (the rest of the line) using shell
@@ -412,6 +413,15 @@ update_squash_messages () {
echo
commit_message $2
;;
+ ack)
+ if test -f $fixup_msg
+ then
+ commit_message $2 | git stripspace --strip-comments |
sed -e '1,2d' $fixup_msg
+ fi
+ printf '%s\n' $comment_char This is the $(nth_string $count)
commit message:
+ echo
+ commit_message $2
+ ;;
fixup)
echo
printf '%s\n' $comment_char The $(nth_string $count) commit
message will be skipped:
@@ -453,7 +463,7 @@ record_in_rewritten() {
echo $oldsha1 $rewritten_pending
case $(peek_next_command) in
- squash|s|fixup|f)
+ squash|s|fixup|f|ack|a)
;;
*)
flush_rewritten_pending
@@ -521,8 +531,11 @@ do_next () {
warn Stopped at $sha1... $rest
exit_with_patch $sha1 0
;;
- squash|s|fixup|f)
+ squash|s|fixup|f|ack|a)
case $command in
+ ack|a)
+ squash_style=ack
+ ;;
squash|s)
squash_style=squash
;;
@@ -546,7 +559,7 @@ do_next () {
die_failed_squash $sha1 $rest
fi
case $(peek_next_command) in
- squash|s|fixup|f)
+ squash|s|fixup|f|ack|a)
# This is an intermediate commit; its message will only
be
# used in case of trouble. So use the long version:
do_with_author output git commit --amend --no-verify -F
$squash_msg \
@@ -557,7 +570,7 @@ do_next () {
# This is the final command of this squash/fixup group
if test -f $fixup_msg
then
- do_with_author git commit --amend --no-verify
-F $fixup_msg \
+ do_with_author git commit --quiet --amend
--no-verify -F $fixup_msg \
${gpg_sign_opt:+$gpg_sign_opt} ||
die_failed_squash $sha1 $rest
else
@@ -690,7 +703,7 @@ skip_unnecessary_picks () {
done $todo $todo.new 3$done
mv -f $todo.new $todo
case $(peek_next_command) in
- squash|s|fixup|f)
+ squash|s|fixup|f|ack|a)
record_in_rewritten $onto
;;
esac ||
@@ -732,7 +745,7 @@ rearrange_squash () {
while read -r pick sha1 message
do
case $message in
- squash! *|fixup! *)
+ squash! *|fixup! *|ack! *)
action=${message%%!*}
rest=$message
prefix=
@@ -740,7 +753,7 @@ rearrange_squash () {
while :
do
case $rest in
- squash! *|fixup! *)
+ squash! *|fixup! * |ack! *)
prefix=$prefix${rest%%!*},
rest=${rest#*! }
;;
@@ -975,6 +988,13 @@ do
comment_out=
fi
+ # keep empty ack! commits around: useful to add
[PATCH 0/4] ack recoding in commit log
As a maintainer, I often get patches by mail, then acked-by,reviewed-by etc responses are sent by separate mail. I like making acks commits, this way they are easy to keep track of as part of git history. Since response mail happens to have appropriate subject matching the patch, it's a natural fit to then use git rebase mechanics if we want to smash these acks into the original commit. I have been using these patches without any problems for a while now, and find the approach very convenient. Included: rebase: new ack! action to handle ack commits this part seems ready for merge to me, please review and comment git-ack: new tool to record an ack this does not have proper documentation and tests yet, I definitely intend to do this but wanted to see whether people like the UI first. posting for early review and feedback Note: yes, I think notes+git replace could be used for this too, my workflow always includes a rebase with --autosquash before publishing anyway, so dealing with ack as with any other commit is nicer. Michael S. Tsirkin (4): rebase -i: add ack action git-rebase: document ack rebase: test ack git-ack: record an ack Documentation/git-rebase.txt | 45 ++--- contrib/git-ack | 79 git-rebase--interactive.sh | 34 +++ t/t3415-rebase-autosquash.sh | 15 + 4 files changed, 161 insertions(+), 12 deletions(-) create mode 100755 contrib/git-ack -- MST -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[PATCH 4/4] git-ack: record an ack
This is a simple script that I use by piping
incoming mail with an ack to it.
It produces an empty ack commit suitable for
squshing with git rebase -i -autosquash.
Works best if people ack individual commits: you simply
pipe each ack to git ack, before pushing your branch,
rebase.
Some people ack series by responding to cover letter
or to commit 1.
To address this usecase, there are two additional
flags: -s saves the ack signature in a file (you can
save several in a row), -r creates an ack for
a given patch using the saved signature.
Thus: pipe ack(s) to git ack -s, then select and pipe
each individual patch to git ack -r.
I don't use these flags much so they likely work
less well.
If it's found useful, this script can either
become a first-class command (with documentation
and tests) or be integrated as a flag into git am.
Limitations: requires that index is clean, this is
so we can create an empty commit recording the ack.
Signed-off-by: Michael S. Tsirkin m...@redhat.com
---
contrib/git-ack | 84 +
1 file changed, 84 insertions(+)
create mode 100755 contrib/git-ack
diff --git a/contrib/git-ack b/contrib/git-ack
new file mode 100755
index 000..4aeb16a
--- /dev/null
+++ b/contrib/git-ack
@@ -0,0 +1,84 @@
+msg=`mktemp`
+patch=`mktemp`
+info=`git mailinfo $msg $patch`
+subj=`echo $info|sed -n 's/^Subject: //p'`
+author=`echo $info|sed -n 's/^Author: //p'`
+email=`echo $info|sed -n 's/^Email: //p'`
+auth=$author $email
+date=`echo $info|sed -n 's/^Date: //p'`
+sign=`mktemp`
+echo ack! $subj $sign
+echo $sign
+if
+git diff --cached HEAD
+then
+nop /dev/null
+else
+echo DIFF in cache. Not acked, reset or commit!
+exit 1
+fi
+GIT_DIR=`pwd`/${GIT_DIR}
+
+usage () {
+ echo Usage: git ack \
+[-s|--save|-a|--append|-r|--restore |-c|--clear]\n 2;
+exit 1;
+}
+
+append=
+save=
+clear=
+
+while test $# != 0
+do
+ case $1 in
+ -a|--append)
+ append=y
+ ;;
+ -s|--s)
+ save=y
+ ;;
+ -r|--restore)
+ restore=y
+ ;;
+ -c|--clear)
+ clear=y
+;;
+ *)
+ usage ;;
+ esac
+ shift
+done
+
+if
+test $clear
+then
+rm -f ${GIT_DIR}/ACKS
+fi
+
+if
+test $save
+then
+if
+test $append
+then
+cat $msg ${GIT_DIR}/ACKS
+else
+cat $msg ${GIT_DIR}/ACKS
+fi
+fi
+
+if
+test $restore
+then
+msg = ${GIT_DIR}/ACKS
+fi
+
+if
+grep '^[A-Z][A-Za-z-]*-by:' $msg $sign
+then
+git commit --allow-empty -F $sign --author=$auth --date=$date
+else
+echo No signature found!
+exit 2
+fi
--
MST
--
To unsubscribe from this list: send the line unsubscribe git in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 00/33] Teach guilt import-commit how to create legal patch names, and more
This is version 4 of the patch series I sent back on 21 Mar 2014. I have addressed all feedback to date, update the documentation where needed, improved the error handling in the cmd and shouldfail functions in the test suite, fixed compatibility with older Git versions, and added signed-off-by lines from Jeff Sipek for those commits that I have received an explicit approval from him (but only if I have not made any other change to that particular commit). I have added two new patches: * Patch 32/33 adds documentation and test cases for the new regexp functionality of guilt header. This could be folded into patch 12. * Patch 33/33 adds documentation for the exit value of guilt push and guilt pop. This could be folded into patches 25-26. To see how the patches have evolved, you might find http://www.lysator.liu.se/~ceder/guilt-oslo-2014-v4/ useful. It displays diffs of all the patches between v3 and v4, in pdiffdiff output format. Here is the original blurb for the series, slightly edited: I recently found myself sitting on a train with a computer in front of me. I tried to use guilt import-commit, which seemed to work, but when I tried to guilt push the commits I had just imported I got some errors. It turned out that guilt import-commit had generated invalid patch names. I decided to fix the issue, and write a test case that demonstrated the problem. One thing led to another, and here I am, a few late nights at a hotel and a return trip on the train later, submitting a patch series in 28 parts. Sorry about the number of patches, but this is what happens when you uncover a bug while writing a test case for the bug you uncovered while writing a test case for your original problem. The patch series consists of: - A number of fixes to the test suite. - A number of bug fixes which I hope are non-controversial. Most of the fixes have test cases. - Changed behavior: guilt push when there is nothing more to push now uses exit status 1. This makes it possible to write shell loops such as while guilt push; do make test||break; done. Also, guilt pop when no patches are applied also uses exit status 1. (This aligns guilt push and guilt pop with how hg qpush and hg qpop has worked for several years.) - Changed behavior: by default, guilt no longer changes branch when you push a patch. You need to do git config guilt.reusebranch false to re-enable that. This patch sets the default value of guilt.reusebranch to true; it should in my opinion change to false a year or two after the next release. - The humble beginnings of a coding style guide. A more detailed overview of the patches: 1. Some tests fail if git config guilt.diffstat true is in effect. 2-4. Some commands fail if run from a directory with spaces in its name. 5. guilt new had an overly restrictive argument parser. 6-8. guilt.diffstat could break guilt fold and guilt new. 9-10. The test suite did not test exit status properly. 11. Remove pointless redirections in the test suite. 12-13. guilt header tried to check if a patch existed, but the check was broken. 14-16. Various parts of guilt tried to ensure that patch names were legal git branch names, but failed. 17-20. guilt graph failed when no patch was applied, and when a branch name contained a comma or a quote. 21-23. git config log.decorate short caused guilt import-commit, guilt patchbomb and guilt rebase to fail in various ways. 24. Patches may contain backslashes, but various informative messages from guilt did backslash processing. 25-26. guilt push and guilt pop should fail when there is nothing to do. 28. Fix coding style problems in a test case. 27, 29. These two commits were things I intended to contribute a while back, when contributing the Change git branch when patches are applied change (commit 67d3af63f422). These commits makes that behavior optional, and it defaults to being disabled, so that you can use both Guilt v0.35 (and earlier) and the current Guilt code against the same repo. These commits add some code complexity, and you might want to skip them if you think the current behavior is better. 30. A coding style guide, with Emacs support. 31. Avoid using git log -p. 32. Test case + doc for guilt header. 33. Doc for guilt push and guilt pop. This patch series is also available on http://repo.or.cz/w/guilt/ceder.git in the oslo-2014-v4 branch. If you already have a copy of guilt, you should be able to fetch that branch with something like: git remote add ceder http://repo.or.cz/r/guilt/ceder.git git fetch ceder refs/heads/oslo-2014-v4:refs/remotes/ceder/oslo-2014-v4 A few of the regression/t-*.out files contain non-ASCII characters. I hope they survive the mail transfer; if not, please use the repo above to fetch the commits. Per Cederqvist (33): The tests should not fail if guilt.diffstat is set.
[GUILT v4 01/33] The tests should not fail if guilt.diffstat is set.
Explicitly set guilt.diffstat to its default value. Without this, the 027 test (and possibly others) fail if guilt.diffstat is set to true in ~/.gitconfig. Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- regression/scaffold | 1 + 1 file changed, 1 insertion(+) diff --git a/regression/scaffold b/regression/scaffold index 546d8c6..5c8b73e 100644 --- a/regression/scaffold +++ b/regression/scaffold @@ -87,6 +87,7 @@ function setup_git_repo # Explicitly set config that the tests rely on. git config log.date default git config log.decorate no + git config guilt.diffstat false } function setup_guilt_repo -- 1.8.3.1 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 02/33] Allow guilt delete -f to run from a dir which contains spaces.
Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- guilt-delete | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/guilt-delete b/guilt-delete index 3e394f8..967ac10 100755 --- a/guilt-delete +++ b/guilt-delete @@ -49,7 +49,7 @@ series_remove_patch $patch guilt_hook delete $patch -[ ! -z $force ] rm -f $GUILT_DIR/$branch/$patch +[ ! -z $force ] rm -f $GUILT_DIR/$branch/$patch exit 0 -- 1.8.3.1 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 04/33] Allow guilt import-commit to run from a dir which contains spaces.
Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- guilt-import-commit | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/guilt-import-commit b/guilt-import-commit index 20dcee2..f14647c 100755 --- a/guilt-import-commit +++ b/guilt-import-commit @@ -23,7 +23,7 @@ if ! must_commit_first; then fi disp About to begin conversion... 2 -disp Current head: `cat $GIT_DIR/refs/heads/\`git_branch\`` 2 +disp Current head: `git rev-parse \`git_branch\`` 2 for rev in `git rev-list $rhash`; do s=`git log --pretty=oneline -1 $rev | cut -c 42-` @@ -46,7 +46,7 @@ for rev in `git rev-list $rhash`; do do_make_header $rev echo git diff --binary $rev^..$rev - ) $GUILT_DIR/$branch/$fname + ) $GUILT_DIR/$branch/$fname # FIXME: grab the GIT_AUTHOR_DATE from the commit object and set the # timestamp on the patch @@ -68,6 +68,6 @@ for rev in `git rev-list $rhash`; do done disp Done. 2 -disp Current head: `cat $GIT_DIR/refs/heads/\`git_branch\`` 2 +disp Current head: `git rev-parse \`git_branch\`` 2 } -- 1.8.3.1 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 03/33] Added test case for guilt delete -f.
Ensure that the file really is deleted. Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- regression/t-026.out | 15 +++ regression/t-026.sh | 5 - 2 files changed, 19 insertions(+), 1 deletion(-) diff --git a/regression/t-026.out b/regression/t-026.out index 3b9fb14..be50b48 100644 --- a/regression/t-026.out +++ b/regression/t-026.out @@ -29,3 +29,18 @@ f 7848194fd2e9ee0eb6589482900687d799d60a12 .git/patches/master/series f 9c18cc7abe6b87f18503714a80a677b4094eb457 .git/patches/master/add f bc9ab2e0f5db99d483961e956e814d963f0309f8 .git/patches/master/modify f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/status +% guilt new delete-me +% guilt pop +All patches popped. +% guilt delete -f delete-me +% list_files +d .git/patches +d .git/patches/master +d .git/refs/patches +d .git/refs/patches/master +f 22930c6d1f1938f298a4fca51c57e4b47171db21 .git/patches/master/mode +f 413390f3906f16f30b054a4fb86c1e014b964504 .git/patches/master/remove +f 7848194fd2e9ee0eb6589482900687d799d60a12 .git/patches/master/series +f 9c18cc7abe6b87f18503714a80a677b4094eb457 .git/patches/master/add +f bc9ab2e0f5db99d483961e956e814d963f0309f8 .git/patches/master/modify +f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/status diff --git a/regression/t-026.sh b/regression/t-026.sh index 0ccdf85..e25d0df 100755 --- a/regression/t-026.sh +++ b/regression/t-026.sh @@ -20,4 +20,7 @@ cmd guilt delete add cmd list_files -# FIXME: test delete -f +cmd guilt new delete-me +cmd guilt pop +cmd guilt delete -f delete-me +cmd list_files -- 1.8.3.1 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 06/33] Fix the do_get_patch function.
A patch file consists of:
(1) the description
(2) optional diffstat
(3) the patches
When extracting the patch, we only want part 3. The do_get_patch used
to give us part 2 and part 3. That made for instance this series of
operations fail if guilt.diffstat is true:
$ guilt new empty-1
$ guilt new empty-2
$ guilt pop
Now at empty-1
$ guilt fold empty-2
$ guilt pop
All patches popped.
$ guilt push
Applying patch..empty-1
fatal: unrecognized input
To force apply this patch, use 'guilt push -f'
Signed-off-by: Per Cederqvist ced...@opera.com
Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net
---
guilt | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/guilt b/guilt
index 8701481..3fc524e 100755
--- a/guilt
+++ b/guilt
@@ -334,7 +334,7 @@ do_get_patch()
{
awk '
BEGIN{}
-/^(diff |---$|--- )/ {patch = 1}
+/^(diff |--- )/ {patch = 1}
patch == 1 {print $0}
END{}
' $1
--
1.8.3.1
--
To unsubscribe from this list: send the line unsubscribe git in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 07/33] Added test cases for guilt fold.
Test that we can combine any combination of patches with empty and non-empty messages, both with and without guilt.diffstat. (All patches are empty.) Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- regression/t-035.out | 467 +++ regression/t-035.sh | 61 +++ 2 files changed, 528 insertions(+) create mode 100644 regression/t-035.out create mode 100755 regression/t-035.sh diff --git a/regression/t-035.out b/regression/t-035.out new file mode 100644 index 000..cc16fb4 --- /dev/null +++ b/regression/t-035.out @@ -0,0 +1,467 @@ +% setup_repo +% git config guilt.diffstat true +%% empty + empty (diffstat=true) +% guilt new empty-1 +% guilt pop +All patches popped. +% guilt push +Applying patch..empty-1 +Patch applied. +% guilt new empty-2 +% guilt pop +Now at empty-1. +% guilt push +Applying patch..empty-2 +Patch applied. +% guilt pop +Now at empty-1. +% guilt fold empty-2 +% guilt pop +All patches popped. +% guilt push +Applying patch..empty-1 +Patch applied. +% list_files +d .git/patches +d .git/patches/master +d .git/refs/patches +d .git/refs/patches/master +f 22930c6d1f1938f298a4fca51c57e4b47171db21 .git/patches/master/mode +f 413390f3906f16f30b054a4fb86c1e014b964504 .git/patches/master/remove +f 4ea806e306f0228a8ef41f186035e7b04097f1f2 .git/patches/master/status +f 7d261b8caad0f161c21daf5de65eeb521ff8c067 .git/patches/master/empty-1 +f 9c18cc7abe6b87f18503714a80a677b4094eb457 .git/patches/master/add +f bc9ab2e0f5db99d483961e956e814d963f0309f8 .git/patches/master/modify +f d28d87b88c1e24d637e390dc3603cfa7c1715711 .git/patches/master/series +f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/empty-1~ +f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/empty-2~ +r bde3d337af70f36836ad606c800d194006f883b3 .git/refs/patches/master/empty-1 +% guilt pop +All patches popped. +% guilt delete -f empty-1 +% list_files +d .git/patches +d .git/patches/master +d .git/refs/patches +d .git/refs/patches/master +f 22930c6d1f1938f298a4fca51c57e4b47171db21 .git/patches/master/mode +f 413390f3906f16f30b054a4fb86c1e014b964504 .git/patches/master/remove +f 9c18cc7abe6b87f18503714a80a677b4094eb457 .git/patches/master/add +f bacb4aad8a55fe4e7aa58a9ae169990bb764069f .git/patches/master/series +f bc9ab2e0f5db99d483961e956e814d963f0309f8 .git/patches/master/modify +f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/empty-1~ +f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/empty-2~ +f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/status +%% empty + nonempty (diffstat=true) +% guilt new empty +% guilt pop +All patches popped. +% guilt push +Applying patch..empty +Patch applied. +% guilt new -f -s -m A commit message. nonempty +% guilt pop +Now at empty. +% guilt push +Applying patch..nonempty +Patch applied. +% guilt pop +Now at empty. +% guilt fold nonempty +% guilt pop +All patches popped. +% guilt push +Applying patch..empty +Patch applied. +% list_files +d .git/patches +d .git/patches/master +d .git/refs/patches +d .git/refs/patches/master +f 15aab0fd8b937eb3bb01841693f35dcb75da2faf .git/patches/master/status +f 22930c6d1f1938f298a4fca51c57e4b47171db21 .git/patches/master/mode +f 413390f3906f16f30b054a4fb86c1e014b964504 .git/patches/master/remove +f 51fcfcf16db2903f19ab4a4a3caacd297ea9f6cd .git/patches/master/empty~ +f 51fcfcf16db2903f19ab4a4a3caacd297ea9f6cd .git/patches/master/nonempty~ +f 683678040eef9334d6329e00d5b9babda3e65b57 .git/patches/master/empty +f 9c18cc7abe6b87f18503714a80a677b4094eb457 .git/patches/master/add +f a26a22287b500a2a372e42c2bab03599bbe37cdf .git/patches/master/series +f bc9ab2e0f5db99d483961e956e814d963f0309f8 .git/patches/master/modify +f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/empty-1~ +f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/empty-2~ +r 4eedaa32894fc07af3298d8c1178052942a3ca6a .git/refs/patches/master/empty +% guilt pop +All patches popped. +% guilt delete -f empty +% list_files +d .git/patches +d .git/patches/master +d .git/refs/patches +d .git/refs/patches/master +f 22930c6d1f1938f298a4fca51c57e4b47171db21 .git/patches/master/mode +f 413390f3906f16f30b054a4fb86c1e014b964504 .git/patches/master/remove +f 51fcfcf16db2903f19ab4a4a3caacd297ea9f6cd .git/patches/master/empty~ +f 51fcfcf16db2903f19ab4a4a3caacd297ea9f6cd .git/patches/master/nonempty~ +f 9c18cc7abe6b87f18503714a80a677b4094eb457 .git/patches/master/add +f bacb4aad8a55fe4e7aa58a9ae169990bb764069f .git/patches/master/series +f bc9ab2e0f5db99d483961e956e814d963f0309f8 .git/patches/master/modify +f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/empty-1~ +f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/empty-2~ +f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/status +%% nonempty + empty (diffstat=true) +% guilt new -f -s -m A commit
[GUILT v4 05/33] guilt new: Accept more than 4 arguments.
The argument parser arbitrarily refused to accept more than 4
arguments. That made it impossible to run guilt new -f -s -m msg
patch. Removed the checks for the number of arguments from the
guilt new parser -- the other checks that are already there are
enough to catch all errors.
Give a better error message if -m isn't followed by a message
argument.
Signed-off-by: Per Cederqvist ced...@opera.com
Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net
---
guilt-new | 7 +++
1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/guilt-new b/guilt-new
index bb68924..9528438 100755
--- a/guilt-new
+++ b/guilt-new
@@ -11,10 +11,6 @@ fi
_main() {
-if [ $# -lt 1 ] || [ $# -gt 4 ]; then
- usage
-fi
-
while [ $# -gt 0 ] ; do
case $1 in
-f)
@@ -31,6 +27,9 @@ while [ $# -gt 0 ] ; do
fi
;;
-m)
+ if [ $# -eq 1 ]; then
+ usage
+ fi
msg=$2
shift
--
1.8.3.1
--
To unsubscribe from this list: send the line unsubscribe git in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 08/33] Added more test cases for guilt new: empty patches.
Test that empty patches are handled correctly, both with and without the guilt.diffstat configuration option. Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- regression/t-020.out | 269 +++ regression/t-020.sh | 60 2 files changed, 329 insertions(+) diff --git a/regression/t-020.out b/regression/t-020.out index af45734..42433dc 100644 --- a/regression/t-020.out +++ b/regression/t-020.out @@ -1128,3 +1128,272 @@ f 9c18cc7abe6b87f18503714a80a677b4094eb457 .git/patches/master/add f bacb4aad8a55fe4e7aa58a9ae169990bb764069f .git/patches/master/series f bc9ab2e0f5db99d483961e956e814d963f0309f8 .git/patches/master/modify f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/status +% guilt new empty.patch +% guilt pop +All patches popped. +% guilt push +Applying patch..empty.patch +Patch applied. +% list_files +d .git/patches +d .git/patches/master +d .git/refs/patches +d .git/refs/patches/master +f 22930c6d1f1938f298a4fca51c57e4b47171db21 .git/patches/master/mode +f 413390f3906f16f30b054a4fb86c1e014b964504 .git/patches/master/remove +f 9c18cc7abe6b87f18503714a80a677b4094eb457 .git/patches/master/add +f bc9ab2e0f5db99d483961e956e814d963f0309f8 .git/patches/master/modify +f d15a1d2d34493f790c78ddacb8815b0b9536ee2b .git/patches/master/series +f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/empty.patch +f e90b964f01cbef60bbe00c38c55d9ea86618a66a .git/patches/master/status +r c7a139f532a43c3c8b0e068cac04f8f6af0f94e1 .git/refs/patches/master/empty.patch +% git log -p +commit c7a139f532a43c3c8b0e068cac04f8f6af0f94e1 +Author: Author Name author@email +Date: Mon Jan 1 00:00:00 2007 + + +patch empty.patch + +commit d4850419ccc1146c7169f500725ce504b9774ed0 +Author: Author Name author@email +Date: Mon Jan 1 00:00:00 2007 + + +initial + +Signed-off-by: Commiter Name commiter@email + +diff --git a/def b/def +new file mode 100644 +index 000..8baef1b +--- /dev/null b/def +@@ -0,0 +1 @@ ++abc +% git config guilt.diffstat true +% guilt refresh +Patch empty.patch refreshed +% guilt pop +All patches popped. +% guilt push +Applying patch..empty.patch +Patch applied. +% list_files +d .git/patches +d .git/patches/master +d .git/refs/patches +d .git/refs/patches/master +f 22930c6d1f1938f298a4fca51c57e4b47171db21 .git/patches/master/mode +f 413390f3906f16f30b054a4fb86c1e014b964504 .git/patches/master/remove +f 7d261b8caad0f161c21daf5de65eeb521ff8c067 .git/patches/master/empty.patch +f 9c18cc7abe6b87f18503714a80a677b4094eb457 .git/patches/master/add +f bc9ab2e0f5db99d483961e956e814d963f0309f8 .git/patches/master/modify +f d15a1d2d34493f790c78ddacb8815b0b9536ee2b .git/patches/master/series +f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/empty.patch~ +f e90b964f01cbef60bbe00c38c55d9ea86618a66a .git/patches/master/status +r c7a139f532a43c3c8b0e068cac04f8f6af0f94e1 .git/refs/patches/master/empty.patch +% git log -p +commit c7a139f532a43c3c8b0e068cac04f8f6af0f94e1 +Author: Author Name author@email +Date: Mon Jan 1 00:00:00 2007 + + +patch empty.patch + +commit d4850419ccc1146c7169f500725ce504b9774ed0 +Author: Author Name author@email +Date: Mon Jan 1 00:00:00 2007 + + +initial + +Signed-off-by: Commiter Name commiter@email + +diff --git a/def b/def +new file mode 100644 +index 000..8baef1b +--- /dev/null b/def +@@ -0,0 +1 @@ ++abc +% git config guilt.diffstat false +% guilt pop +All patches popped. +% guilt push +Applying patch..empty.patch +Patch applied. +% list_files +d .git/patches +d .git/patches/master +d .git/refs/patches +d .git/refs/patches/master +f 22930c6d1f1938f298a4fca51c57e4b47171db21 .git/patches/master/mode +f 413390f3906f16f30b054a4fb86c1e014b964504 .git/patches/master/remove +f 7ad87a0bdb8cf0a57cfc384633edabbb9c2bfa1b .git/patches/master/empty.patch +f 9c18cc7abe6b87f18503714a80a677b4094eb457 .git/patches/master/add +f bc9ab2e0f5db99d483961e956e814d963f0309f8 .git/patches/master/modify +f d15a1d2d34493f790c78ddacb8815b0b9536ee2b .git/patches/master/series +f da39a3ee5e6b4b0d3255bfef95601890afd80709 .git/patches/master/empty.patch~ +f e90b964f01cbef60bbe00c38c55d9ea86618a66a .git/patches/master/status +r 8ed27228b117c0c88abf3d586bcc43c68e975cea .git/refs/patches/master/empty.patch +% git log -p +commit 8ed27228b117c0c88abf3d586bcc43c68e975cea +Author: Per Cederqvist ce...@lysator.liu.se +Date: Mon Jan 1 00:00:00 2007 + + +Fix a bug. + +This commit fixes a serious bug. + +FIXME: +- add a test case +- track down the bug +- actually fix it + +commit d4850419ccc1146c7169f500725ce504b9774ed0 +Author: Author Name author@email +Date: Mon Jan 1 00:00:00 2007 + + +initial + +Signed-off-by: Commiter Name commiter@email + +diff --git a/def b/def +new file mode 100644 +index 000..8baef1b +---
[GUILT v4 09/33] Test suite: properly check the exit status of commands.
The cmd and shouldfail functions checked the exit status of the
replace_path function instead of the actual command that was running.
(The $? construct checks the exit status of the last command in a
pipeline, not the first command.)
Print an explicit error message if a command that should fail actually
succeeds.
Updated t-032.sh, which used shouldfail instead of cmd in one
place. (The comment in the script makes it clear that the command is
expected to succeed.)
Signed-off-by: Per Cederqvist ced...@opera.com
---
regression/scaffold | 23 ---
regression/t-032.sh | 2 +-
2 files changed, 17 insertions(+), 8 deletions(-)
diff --git a/regression/scaffold b/regression/scaffold
index 5c8b73e..2e04c83 100644
--- a/regression/scaffold
+++ b/regression/scaffold
@@ -51,19 +51,28 @@ function filter_dd
function cmd
{
echo % $@
- $@ 21 | replace_path return 0
- return 1
+ if ! (
+ exec 31
+ rv=`(($@ 21; echo $? 4) | replace_path 3 ) 41`
+ exit $rv
+ ) ; then
+ echo % FAIL: The above command should succeed but failed.
+ exit 1
+ fi
}
# usage: shouldfail cmd..
function shouldfail
{
echo % $@
- (
- $@ 21 || return 0
- return 1
- ) | replace_path
- return $?
+ if (
+ exec 31
+ rv=`(($@ 21; echo $? 4) | replace_path 3 ) 41`
+ exit $rv
+ ) ; then
+ echo % FAIL: The above command should fail but succeeded.
+ exit 1
+ fi
}
# usage: list_files
diff --git a/regression/t-032.sh b/regression/t-032.sh
index b1d5f19..bba401e 100755
--- a/regression/t-032.sh
+++ b/regression/t-032.sh
@@ -28,7 +28,7 @@ shouldfail guilt import -P foo3 foo
cmd guilt import -P foo2 foo
# ok
-shouldfail guilt import foo
+cmd guilt import foo
# duplicate patch name (implicit)
shouldfail guilt import foo
--
1.8.3.1
--
To unsubscribe from this list: send the line unsubscribe git in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 10/33] Run test_failed if the exit status of a test script is bad.
There were two problems with the old code: - Since set -e is in effect (that is set in scaffold) the run-test script exited immediately if a t-*.sh script failed. This is not nice, as we want the error report that test_failed prints. - The code ran cd - between running the t-*.sh script and checking the exit status, so the exit status was lost. (Actually, the exit status was saved in $ERR, but nothing ever looked at $ERR.) Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- regression/run-tests | 10 +++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/regression/run-tests b/regression/run-tests index a10e796..8e0af9f 100755 --- a/regression/run-tests +++ b/regression/run-tests @@ -55,11 +55,15 @@ function run_test # run the test cd $REPODIR /dev/null - $REG_DIR/t-$1.sh 21 $LOGFILE - ERR=$? + if $REG_DIR/t-$1.sh 21 $LOGFILE; then + ERR=false + else + ERR=true + fi + cd - /dev/null - [ $? -ne 0 ] test_failed + $ERR test_failed diff -u t-$1.out $LOGFILE || test_failed echo done. -- 1.8.3.1 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 13/33] Check that guilt header '.*' fails.
Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- regression/t-028.out | 7 +++ regression/t-028.sh | 4 2 files changed, 11 insertions(+) diff --git a/regression/t-028.out b/regression/t-028.out index 1564c09..ea72a3a 100644 --- a/regression/t-028.out +++ b/regression/t-028.out @@ -49,3 +49,10 @@ Signed-off-by: Commiter Name commiter@email % guilt header non-existant Patch non-existant is not in the series +% guilt header .* +.* does not uniquely identify a patch. Did you mean any of these? + modify + add + remove + mode + patch-with-some-desc diff --git a/regression/t-028.sh b/regression/t-028.sh index 88e9adb..2ce0378 100755 --- a/regression/t-028.sh +++ b/regression/t-028.sh @@ -31,4 +31,8 @@ done shouldfail guilt header non-existant +# This is an evil variant of a non-existant patch. However, this +# patch name is a regexp that just happens to match an existing patch. +shouldfail guilt header '.*' + # FIXME: how do we check that -e works? -- 1.8.3.1 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 12/33] guilt header: more robust header selection.
If you run something like guilt header '.*' the command would crash, because the grep comand that tries to ensure that the patch exist would detect a match, but the later code expected the match to be exact. Fixed by comparing exact strings. And as a creeping feature guilt header will now try to use the supplied patch name as an unanchored regexp if no exact match was found. If the regexp yields a unique match, it is used; if more than one patch matches, the names of all patches are listed and the command fails. (Exercise left to the reader: generalized this so that guilt push also accepts a unique regular expression.) Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- guilt-header | 29 ++--- 1 file changed, 26 insertions(+), 3 deletions(-) diff --git a/guilt-header b/guilt-header index 41e00cc..c3d24f9 100755 --- a/guilt-header +++ b/guilt-header @@ -45,10 +45,33 @@ esac [ -z $patch ] die No patches applied. # check that patch exists in the series -ret=`get_full_series | grep -e ^$patch\$ | wc -l` -if [ $ret -eq 0 ]; then - die Patch $patch is not in the series +TMP_FULL_SERIES=`get_tmp_file series` +get_full_series $TMP_FULL_SERIES +found_patch= +while read x; do + if [ $x = $patch ]; then + found_patch=$patch + break + fi +done $TMP_FULL_SERIES +if [ -z $found_patch ]; then + TMP_MATCHES=`get_tmp_file series` + grep $patch $TMP_FULL_SERIES $TMP_MATCHES + nr=`wc -l $TMP_MATCHES` + if [ $nr -gt 1 ]; then + echo $patch does not uniquely identify a patch. Did you mean any of these? 2 + sed 's/^/ /' $TMP_MATCHES 2 + rm -f $TMP_MATCHES $TMP_FULL_SERIES + exit 1 + elif [ $nr -eq 0 ]; then + rm -f $TMP_MATCHES $TMP_FULL_SERIES + die Patch $patch is not in the series + fi + found_patch=`cat $TMP_MATCHES` + rm -f $TMP_MATCHES fi +rm -f $TMP_FULL_SERIES +patch=$found_patch # FIXME: warn if we're editing an applied patch -- 1.8.3.1 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 14/33] Use git check-ref-format to validate patch names.
The valid_patchname now lets git check-ref-format do its job instead
of trying (and failing) to implement the same rules. See
git-check-ref-format(1) for a list of the rules. Re-implement rules
added to git check-ref-format after Git 1.5.0, so that guilt rejects
the same names no matter what version of Git we are using (but
versions prior to 1.5.0 are not supported).
Refer to the git-check-ref-format(1) man page in the error messages
produced when valid_patchname indicates that the name is bad.
Added testcases that breaks most of the rules in that man-page.
Signed-off-by: Per Cederqvist ced...@opera.com
---
guilt| 46 +-
guilt-fork | 2 +-
guilt-import | 2 +-
guilt-new| 2 +-
regression/t-025.out | 426 +--
regression/t-025.sh | 12 +-
regression/t-032.out | 4 +-
7 files changed, 474 insertions(+), 20 deletions(-)
diff --git a/guilt b/guilt
index 3fc524e..9567a78 100755
--- a/guilt
+++ b/guilt
@@ -132,14 +132,50 @@ fi
# usage: valid_patchname patchname
valid_patchname()
{
+ # Once we only support Git 1.7.8 and newer, the command below
+ # could be replaced with:
+ #
+ # git check-ref-format --allow-onelevel $1
+ #
+ # Instead, we arbitrarily prepend one level. The result
+ # should be the same, and this is portable to all existing
+ # versions of Git.
+ git check-ref-format a/$1
+ if [ $? -ne 0 ]; then
+ return 1
+ fi
+
+ # We want to reject all names that Git 2.0.0 rejects. In case
+ # we are running an older version, we explicitly check some
+ # cases that were added to Git after version 1.5.0. This code
+ # aims to support Git version 1.5.0 and newer.
+
+ # Git 1.7.6.4 and newer rejects the DEL character.
+ if [ `echo $1|tr -d '\177'` != $1 ]; then
+ return 1
+ fi
+
+ # Git 1.7.8 and newer rejects refs that start or end with
+ # slash or contain multiple adjacent slashes.
case $1 in
- /*|./*|../*|*/./*|*/../*|*/.|*/..|*/|*\ *|*\*)
+ /*|*/|*//*)
return 1;;
- *:*)
- return 1;;
- *)
- return 0;;
esac
+
+ # Git 1.7.8 and newer rejects refname components that end in
+ # .lock.
+ case $1 in
+ *.lock/*|*.lock)
+ return 1;;
+ esac
+
+ # Git 1.8.5 and newer rejects refnames that are made up of the
+ # single character @.
+ if [ $1 = @ ]; then
+ return 1
+ fi
+
+ return 0
}
get_branch()
diff --git a/guilt-fork b/guilt-fork
index a85d391..6447e55 100755
--- a/guilt-fork
+++ b/guilt-fork
@@ -37,7 +37,7 @@ else
fi
if ! valid_patchname $newpatch; then
- die The specified patch name contains invalid characters (:).
+ die The specified patch name is invalid according to
git-check-ref-format(1).
fi
if [ -e $GUILT_DIR/$branch/$newpatch ]; then
diff --git a/guilt-import b/guilt-import
index 3e9b3bb..928e325 100755
--- a/guilt-import
+++ b/guilt-import
@@ -40,7 +40,7 @@ if [ -e $GUILT_DIR/$branch/$newname ]; then
fi
if ! valid_patchname $newname; then
- die The specified patch name contains invalid characters (:).
+ die The specified patch name is invalid according to
git-check-ref-format(1).
fi
# create any directories as needed
diff --git a/guilt-new b/guilt-new
index 9528438..9f7fa44 100755
--- a/guilt-new
+++ b/guilt-new
@@ -64,7 +64,7 @@ fi
if ! valid_patchname $patch; then
disp Patchname is invalid. 2
- die it cannot begin with '/', './' or '../', or contain /./, /../, or
whitespace
+ die It must follow the rules in git-check-ref-format(1).
fi
# create any directories as needed
diff --git a/regression/t-025.out b/regression/t-025.out
index 7811ab1..01bc406 100644
--- a/regression/t-025.out
+++ b/regression/t-025.out
@@ -141,7 +141,7 @@ f da39a3ee5e6b4b0d3255bfef95601890afd80709
.git/patches/master/prepend
r acdeef96ee30eb34bbbf65d11de5cf7da4b5fee8 .git/refs/patches/master/prepend
% guilt new white space
Patchname is invalid.
-it cannot begin with '/', './' or '../', or contain /./, /../, or whitespace
+It must follow the rules in git-check-ref-format(1).
% list_files
d .git/patches
d .git/patches/master
@@ -211,7 +211,7 @@ f da39a3ee5e6b4b0d3255bfef95601890afd80709
.git/patches/master/prepend
r acdeef96ee30eb34bbbf65d11de5cf7da4b5fee8 .git/refs/patches/master/prepend
% guilt new /abc
Patchname is invalid.
-it cannot begin with '/', './' or '../', or contain /./, /../, or whitespace
+It must follow the rules in git-check-ref-format(1).
% list_files
d .git/patches
d .git/patches/master
@@ -235,7 +235,7 @@ f da39a3ee5e6b4b0d3255bfef95601890afd80709
.git/patches/master/prepend
r
[GUILT v4 11/33] test suite: remove pointless redirection.
The shouldfail function already redirects stderr to stdout, so there is no need to do the same in t-028.sh and t-021.sh. Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- regression/t-021.sh | 2 +- regression/t-025.sh | 2 +- regression/t-028.sh | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/regression/t-021.sh b/regression/t-021.sh index 6337d7b..614e870 100755 --- a/regression/t-021.sh +++ b/regression/t-021.sh @@ -61,7 +61,7 @@ for n in `_seq -2 $npatches`; do if [ $n -gt 0 ]; then cmd guilt pop -n $n else - shouldfail guilt pop -n $n 21 + shouldfail guilt pop -n $n fi cmd list_files diff --git a/regression/t-025.sh b/regression/t-025.sh index 3824608..985fed4 100755 --- a/regression/t-025.sh +++ b/regression/t-025.sh @@ -44,7 +44,7 @@ shouldfail guilt new white space cmd list_files for pname in prepend mode /abc ./blah ../blah abc/./blah abc/../blah abc/. abc/.. abc/ ; do - shouldfail guilt new $pname 21 + shouldfail guilt new $pname cmd list_files done diff --git a/regression/t-028.sh b/regression/t-028.sh index 8480100..88e9adb 100755 --- a/regression/t-028.sh +++ b/regression/t-028.sh @@ -29,6 +29,6 @@ guilt series | while read n; do cmd guilt header $n done -shouldfail guilt header non-existant 21 +shouldfail guilt header non-existant # FIXME: how do we check that -e works? -- 1.8.3.1 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 15/33] Produce legal patch names in guilt-import-commit.
Try harder to create patch names that adhere to the rules in
git-check-ref-format(1) when deriving a patch name from the commit
message. Verify that the derived name using git check-ref-format,
and as a final fallback simply use the patch name x (to ensure that
the code is future-proof in case new rules are added in the future).
Always append a .patch suffix to the patch name.
Added test cases.
Signed-off-by: Per Cederqvist ced...@opera.com
Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net
---
guilt-import-commit | 20 +-
regression/t-034.out | 567 +++
regression/t-034.sh | 71 +++
3 files changed, 656 insertions(+), 2 deletions(-)
create mode 100644 regression/t-034.out
create mode 100755 regression/t-034.sh
diff --git a/guilt-import-commit b/guilt-import-commit
index f14647c..6260c56 100755
--- a/guilt-import-commit
+++ b/guilt-import-commit
@@ -28,19 +28,35 @@ disp Current head: `git rev-parse \`git_branch\`` 2
for rev in `git rev-list $rhash`; do
s=`git log --pretty=oneline -1 $rev | cut -c 42-`
+ # Try to convert the first line of the commit message to a
+ # valid patch name.
fname=`echo $s | sed -e s//and/g -e s/[ :]/_/g -e s,[/\\],-,g \
-e s/['\\[{}]//g -e 's/]//g' -e 's/\*/-/g' \
- -e 's/\?/-/g' | tr A-Z a-z`
+ -e 's/\?/-/g' -e 's/\.\.\.*/./g' -e 's/^\.//' \
+ -e 's/\.patch$//' -e 's/\.$//' | tr A-Z a-z`
+
+ if ! valid_patchname $fname; then
+ # Try harder to make it a legal commit name by
+ # removing all but a few safe characters.
+ fname=`echo $fname|tr -d -c _a-zA-Z0-9---/\\n`
+ fi
+ if ! valid_patchname $fname; then
+ # If we failed to derive a legal patch name, use the
+ # name x. (If this happens, we likely have to
+ # append a suffix to make the name unique.)
+ fname=x
+ fi
disp Converting `echo $rev | cut -c 1-8` as $fname
mangle_prefix=1
fname_base=$fname
- while [ -f $GUILT_DIR/$branch/$fname ]; do
+ while [ -f $GUILT_DIR/$branch/$fname.patch ]; do
fname=$fname_base-$mangle_prefix
mangle_prefix=`expr $mangle_prefix + 1`
disp Patch under that name exists...trying '$fname'
done
+ fname=$fname.patch
(
do_make_header $rev
diff --git a/regression/t-034.out b/regression/t-034.out
new file mode 100644
index 000..7bc9459
--- /dev/null
+++ b/regression/t-034.out
@@ -0,0 +1,567 @@
+% setup_git_repo
+% git tag base
+% create_commit a The sequence /. is forbidden.
+[master eebb76e] The sequence /. is forbidden.
+ Author: Author Name author@email
+ 1 file changed, 1 insertion(+)
+ create mode 100644 a
+% create_commit a The sequence .lock/ is forbidden.
+[master 45e81b5] The sequence .lock/ is forbidden.
+ Author: Author Name author@email
+ 1 file changed, 1 insertion(+)
+% create_commit a A/component/may/not/end/in/foo.lock
+[master bbf3f59] A/component/may/not/end/in/foo.lock
+ Author: Author Name author@email
+ 1 file changed, 1 insertion(+)
+% create_commit a Two consecutive dots (..) is forbidden.
+[master 1535e67] Two consecutive dots (..) is forbidden.
+ Author: Author Name author@email
+ 1 file changed, 1 insertion(+)
+% create_commit a Check/multiple/../dots/./foo..patch
+[master 48eb60c] Check/multiple/../dots/./foo..patch
+ Author: Author Name author@email
+ 1 file changed, 1 insertion(+)
+% create_commit a Space is forbidden.
+[master 10dea83] Space is forbidden.
+ Author: Author Name author@email
+ 1 file changed, 1 insertion(+)
+% create_commit a Tilde~is~forbidden.
+[master 70a83b7] Tilde~is~forbidden.
+ Author: Author Name author@email
+ 1 file changed, 1 insertion(+)
+% create_commit a Caret^is^forbidden.
+[master ee6ef2c] Caret^is^forbidden.
+ Author: Author Name author@email
+ 1 file changed, 1 insertion(+)
+% create_commit a Colon:is:forbidden.
+[master c077fe2] Colon:is:forbidden.
+ Author: Author Name author@email
+ 1 file changed, 1 insertion(+)
+% create_commit a Delisforbidden.
+[master 589ee30] Delisforbidden.
+ Author: Author Name author@email
+ 1 file changed, 1 insertion(+)
+% git branch some-branch
+% git tag some-tag
+% create_commit a Ctrlisforbidden.
+[master e63cdde] Ctrlisforbidden.
+ Author: Author Name author@email
+ 1 file changed, 1 insertion(+)
+% create_commit a CR
is
also
forbidden.
+[master 21ad093] CR
is
also
forbidden.
+ Author: Author Name author@email
+ 1 file changed, 1 insertion(+)
+% create_commit a Question-mark?is?forbidden.
+[master be2fa9b] Question-mark?is?forbidden.
+ Author: Author Name author@email
+ 1 file changed, 1 insertion(+)
+% create_commit a Asterisk*is*forbidden.
+[master af7b50f] Asterisk*is*forbidden.
+ Author: Author Name author@email
+ 1 file changed, 1 insertion(+)
+%
[GUILT v4 17/33] guilt graph no longer loops when no patches are applied.
Give an error message if no patches are applied. Added a test case
that never terminates unless this fix is applied.
Signed-off-by: Per Cederqvist ced...@opera.com
Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net
---
guilt-graph | 9 +++--
regression/t-033.out | 3 +++
regression/t-033.sh | 13 +
3 files changed, 23 insertions(+), 2 deletions(-)
create mode 100644 regression/t-033.out
create mode 100755 regression/t-033.sh
diff --git a/guilt-graph b/guilt-graph
index b3469dc..56d0e77 100755
--- a/guilt-graph
+++ b/guilt-graph
@@ -17,8 +17,13 @@ fi
patchname=$1
-bottom=`git rev-parse refs/patches/$branch/$(head_n 1 $applied)`
-base=`git rev-parse $bottom^`
+bottompatch=$(head_n 1 $applied)
+if [ -z $bottompatch ]; then
+ echo No patch applied. 2
+ exit 1
+fi
+
+base=`git rev-parse refs/patches/${branch}/${bottompatch}^`
if [ -z $patchname ]; then
top=`git rev-parse HEAD`
diff --git a/regression/t-033.out b/regression/t-033.out
new file mode 100644
index 000..76613f9
--- /dev/null
+++ b/regression/t-033.out
@@ -0,0 +1,3 @@
+% setup_repo
+% guilt graph
+No patch applied.
diff --git a/regression/t-033.sh b/regression/t-033.sh
new file mode 100755
index 000..a3a8981
--- /dev/null
+++ b/regression/t-033.sh
@@ -0,0 +1,13 @@
+#!/bin/bash
+#
+# Test the graph code
+#
+
+source $REG_DIR/scaffold
+
+cmd setup_repo
+
+# Check that guilt graph gives a proper No patch applied error
+# message when no patches are applied. (An older version of guilt
+# used to enter an endless loop in this situation.)
+shouldfail guilt graph
--
1.8.3.1
--
To unsubscribe from this list: send the line unsubscribe git in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 16/33] Fix backslash handling when creating names of imported patches.
The 'echo $s' construct sometimes processes escape sequences. (This
happens, for instance, under Ubuntu 14.04 when /bin/sh is actually
dash.) We don't want that to happen when we are importing commits, so
use 'printf %s $s' instead.
(The -E option of bash that explicitly disables backslash expansion is
not portable; it is not supported by dash.)
Signed-off-by: Per Cederqvist ced...@opera.com
Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net
---
guilt-import-commit | 2 +-
regression/t-034.out | 14 +++---
2 files changed, 8 insertions(+), 8 deletions(-)
diff --git a/guilt-import-commit b/guilt-import-commit
index 6260c56..45f2404 100755
--- a/guilt-import-commit
+++ b/guilt-import-commit
@@ -30,7 +30,7 @@ for rev in `git rev-list $rhash`; do
# Try to convert the first line of the commit message to a
# valid patch name.
- fname=`echo $s | sed -e s//and/g -e s/[ :]/_/g -e s,[/\\],-,g \
+ fname=`printf %s $s | sed -e s//and/g -e s/[ :]/_/g -e
s,[/\\],-,g \
-e s/['\\[{}]//g -e 's/]//g' -e 's/\*/-/g' \
-e 's/\?/-/g' -e 's/\.\.\.*/./g' -e 's/^\.//' \
-e 's/\.patch$//' -e 's/\.$//' | tr A-Z a-z`
diff --git a/regression/t-034.out b/regression/t-034.out
index 7bc9459..bda4399 100644
--- a/regression/t-034.out
+++ b/regression/t-034.out
@@ -236,7 +236,7 @@ Date: Mon Jan 1 00:00:00 2007 +
About to begin conversion...
Current head: 2a8b1889aa5066193bac978e6bf5073ffcfa6541
Converting 2a8b1889 as can-have-embedded-single-slashes
-Converting 0a46f8fa as backslash-isorbidden
+Converting 0a46f8fa as backslash-is-forbidden
Converting aedb74fd as x
Converting 30187ed0 as cannot@have@the@sequence@at-brace
Converting 106e8e5a as cannot_end_in_
@@ -300,7 +300,7 @@ Applying patch..cannot@have@the@sequence@at-brace.patch
Patch applied.
Applying patch..x.patch
Patch applied.
-Applying patch..backslash-isorbidden.patch
+Applying patch..backslash-is-forbidden.patch
Patch applied.
Applying patch..can-have-embedded-single-slashes.patch
Patch applied.
@@ -311,7 +311,7 @@ Date: Mon Jan 1 00:00:00 2007 +
Can/have/embedded/single/slashes
-commit 7c3ffa4f940c862e9f11f5d4a5ae421f7a8d3141
(refs/patches/master/backslash-isorbidden.patch)
+commit 7c3ffa4f940c862e9f11f5d4a5ae421f7a8d3141
(refs/patches/master/backslash-is-forbidden.patch)
Author: Author Name author@email
Date: Mon Jan 1 00:00:00 2007 +
@@ -518,8 +518,6 @@ d .git/patches/master
d .git/refs/patches
d .git/refs/patches/master
f 06beca7069b9e576bd431f65d13862ed5d3e2a0f
.git/patches/master/ctrlisforbidden.patch
-f 08267ec6783ea9d1adae55b275198f7594764ed0 .git/patches/master/series
-f 08267ec6783ea9d1adae55b275198f7594764ed0 .git/patches/master/status
f 09b7e9be44ae5ec3a4bb30f5ee9d4ebc2c042f64
.git/patches/master/two_consecutive_dots_(.)_is_forbidden.patch
f 0b971c9a17aeca2319c93d700ffd98acc2a93451
.git/patches/master/question-mark-is-forbidden.patch
f 2b8392f63d61efc12add554555adae30883993cc
.git/patches/master/cannot-end-in-slash-.patch
@@ -529,7 +527,7 @@ f 34e07c584032df137f19bdb66d93f316f00a5ac8
.git/patches/master/tildeisforbidden
f 49bab499826b63deb2bd704629d60c7268c57aee
.git/patches/master/the_sequence_-._is_forbidden.patch
f 5bcddb8ccb6e6e5e8a61e9e56cb2e0f70cbab2f5
.git/patches/master/cannot@have@the@sequence@at-brace.patch
f 637b982fe14a240de181ae63226b27e0c406b3dc
.git/patches/master/asterisk-is-forbidden.patch
-f 698f8a7d41a64e3b6be1a3eba86574078b22a5f3
.git/patches/master/backslash-isorbidden.patch
+f 698f8a7d41a64e3b6be1a3eba86574078b22a5f3
.git/patches/master/backslash-is-forbidden.patch
f 7b103c3c7ae298cd2334f6f49da48bae1424f77b
.git/patches/master/crisalsoforbidden.patch
f 9b810b8c63779c51d2e7f61ab59cd49835041563 .git/patches/master/x.patch
f a22958d9ae9976fd7b2b5a9d0bcd44bf7ad9b08a
.git/patches/master/caretisforbidden.patch
@@ -537,6 +535,8 @@ f ab325bf5a432937fc6f231d3e8a773a62d53952b
.git/patches/master/multiple-slashes
f cb9cffbd4465bddee266c20ccebd14eb687eaa89
.git/patches/master/delisforbidden.patch
f d0885a1a1fdee0fd1e4fedce3f7acd3100540bc4
.git/patches/master/openbracketisforbidden.patch
f d2903523fb66a346596eabbdd1bda4e52b266440
.git/patches/master/check-multiple-.-dots-.-foo.patch
+f da90de1c84138194524994e0bc3bc4ca8189c999 .git/patches/master/series
+f da90de1c84138194524994e0bc3bc4ca8189c999 .git/patches/master/status
f dfc11f76394059909671af036598c5fbe33001ba
.git/patches/master/space_is_forbidden.patch
f e47474c52d6c893f36d0457f885a6dd1267742bb
.git/patches/master/colon_is_forbidden.patch
f e7a5f8912592d9891e6159f5827c8b4f372cc406
.git/patches/master/the_sequence_.lock-_is_forbidden.patch
@@ -548,7 +548,7 @@ r 1626a11d979a1e9e775c766484172212277153df
.git/refs/patches/master/asterisk-is
r 3a0d5ccef0359004fcaa9cee98fbd6a2c4432e74
.git/refs/patches/master/tildeisforbidden.patch
r 434e07cacdd8e7eb4723e67cb2d100b3a4121a3a
[GUILT v4 19/33] Check that guilt graph works when working on a branch with a comma.
git branch names can contain commas. Check that guilt graph works
even in that case.
Signed-off-by: Per Cederqvist ced...@opera.com
Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net
---
regression/t-033.out | 65
regression/t-033.sh | 39 +++
2 files changed, 104 insertions(+)
diff --git a/regression/t-033.out b/regression/t-033.out
index 76613f9..3d1c61f 100644
--- a/regression/t-033.out
+++ b/regression/t-033.out
@@ -1,3 +1,68 @@
% setup_repo
% guilt graph
No patch applied.
+%% Testing branch a,graph
+% git checkout -b a,graph master
+Switched to a new branch 'a,graph'
+% guilt init
+% guilt new a.patch
+% guilt pop
+All patches popped.
+% guilt push
+Applying patch..a.patch
+Patch applied.
+% guilt graph
+digraph G {
+# checking rev 95275d7c05c6a6176d3941374115b91272877f6c
+ 95275d7c05c6a6176d3941374115b91272877f6c [label=a.patch]
+}
+% git add file.txt
+% guilt refresh
+Patch a.patch refreshed
+% guilt pop
+All patches popped.
+% guilt push
+Applying patch..a.patch
+Patch applied.
+% guilt graph
+digraph G {
+# checking rev ff2775f8d1dc753f635830adcc3a067e0b681e2d
+ ff2775f8d1dc753f635830adcc3a067e0b681e2d [label=a.patch]
+}
+%% Adding an unrelated file in a new patch. No deps.
+% guilt new b.patch
+% git add file2.txt
+% guilt refresh
+Patch b.patch refreshed
+% guilt pop
+Now at a.patch.
+% guilt push
+Applying patch..b.patch
+Patch applied.
+% guilt graph
+digraph G {
+# checking rev c7014443c33d2b0237293687ceb9cbd38313df65
+ c7014443c33d2b0237293687ceb9cbd38313df65 [label=b.patch]
+# checking rev ff2775f8d1dc753f635830adcc3a067e0b681e2d
+ ff2775f8d1dc753f635830adcc3a067e0b681e2d [label=a.patch]
+}
+%% Changing a file already changed in the first patch adds a dependency.
+% guilt new c.patch
+% git add file.txt
+% guilt refresh
+Patch c.patch refreshed
+% guilt pop
+Now at b.patch.
+% guilt push
+Applying patch..c.patch
+Patch applied.
+% guilt graph
+digraph G {
+# checking rev 891bc14b5603474c9743fd04f3da888644413dc5
+ 891bc14b5603474c9743fd04f3da888644413dc5 [label=c.patch]
+# checking rev c7014443c33d2b0237293687ceb9cbd38313df65
+ c7014443c33d2b0237293687ceb9cbd38313df65 [label=b.patch]
+# checking rev ff2775f8d1dc753f635830adcc3a067e0b681e2d
+ ff2775f8d1dc753f635830adcc3a067e0b681e2d [label=a.patch]
+ 891bc14b5603474c9743fd04f3da888644413dc5 -
ff2775f8d1dc753f635830adcc3a067e0b681e2d; // ?
+}
diff --git a/regression/t-033.sh b/regression/t-033.sh
index a3a8981..fac081e 100755
--- a/regression/t-033.sh
+++ b/regression/t-033.sh
@@ -3,6 +3,13 @@
# Test the graph code
#
+function fixup_time_info
+{
+ cmd guilt pop
+ touch -a -m -t $TOUCH_DATE .git/patches/a,graph/$1
+ cmd guilt push
+}
+
source $REG_DIR/scaffold
cmd setup_repo
@@ -11,3 +18,35 @@ cmd setup_repo
# message when no patches are applied. (An older version of guilt
# used to enter an endless loop in this situation.)
shouldfail guilt graph
+
+echo %% Testing branch a,graph
+cmd git checkout -b a,graph master
+
+cmd guilt init
+
+cmd guilt new a.patch
+
+fixup_time_info a.patch
+cmd guilt graph
+
+cmd echo a file.txt
+cmd git add file.txt
+cmd guilt refresh
+fixup_time_info a.patch
+cmd guilt graph
+
+echo %% Adding an unrelated file in a new patch. No deps.
+cmd guilt new b.patch
+cmd echo b file2.txt
+cmd git add file2.txt
+cmd guilt refresh
+fixup_time_info b.patch
+cmd guilt graph
+
+echo %% Changing a file already changed in the first patch adds a dependency.
+cmd guilt new c.patch
+cmd echo c file.txt
+cmd git add file.txt
+cmd guilt refresh
+fixup_time_info c.patch
+cmd guilt graph
--
1.8.3.1
--
To unsubscribe from this list: send the line unsubscribe git in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 18/33] guilt-graph: Handle commas in branch names.
This fix relies on the fact that git branch names can not contain :.
Signed-off-by: Per Cederqvist ced...@opera.com
Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net
---
guilt-graph | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/guilt-graph b/guilt-graph
index 56d0e77..924a63e 100755
--- a/guilt-graph
+++ b/guilt-graph
@@ -51,7 +51,7 @@ safebranch=`echo $branch|sed 's%/%/%g'`
while [ $current != $base ]; do
pname=`git show-ref | sed -n -e
/^$current refs\/patches\/$safebranch/ {
- s,^$current refs/patches/$branch/,,
+ s:^$current refs/patches/$branch/::
p
q
}`
--
1.8.3.1
--
To unsubscribe from this list: send the line unsubscribe git in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 20/33] guilt graph: Handle patch names containing quotes.
Quote quotes with a backslash in the guilt graph output. Otherwise,
the dot file could contain syntax errors.
Added a test case.
Signed-off-by: Per Cederqvist ced...@opera.com
Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net
---
guilt-graph | 2 ++
regression/t-033.out | 22 ++
regression/t-033.sh | 9 +
3 files changed, 33 insertions(+)
diff --git a/guilt-graph b/guilt-graph
index 924a63e..0857e0d 100755
--- a/guilt-graph
+++ b/guilt-graph
@@ -57,6 +57,8 @@ while [ $current != $base ]; do
}`
[ -z $pname ] pname=?
+ pname=`printf \%s\ \$pname\ | sed 's/\/\/g'`
+
disp # checking rev $current
disp \$current\ [label=\$pname\]
diff --git a/regression/t-033.out b/regression/t-033.out
index 3d1c61f..c120d4f 100644
--- a/regression/t-033.out
+++ b/regression/t-033.out
@@ -66,3 +66,25 @@ digraph G {
ff2775f8d1dc753f635830adcc3a067e0b681e2d [label=a.patch]
891bc14b5603474c9743fd04f3da888644413dc5 -
ff2775f8d1dc753f635830adcc3a067e0b681e2d; // ?
}
+% guilt new a-betterquicker'-patch.patch
+% git add file.txt
+% guilt refresh
+Patch a-betterquicker'-patch.patch refreshed
+% guilt pop
+Now at c.patch.
+% guilt push
+Applying patch..a-betterquicker'-patch.patch
+Patch applied.
+% guilt graph
+digraph G {
+# checking rev bc7df666a646739eaf559af23cab72f2bfd01f0e
+ bc7df666a646739eaf559af23cab72f2bfd01f0e
[label=a-\betterquicker'-patch.patch]
+# checking rev 891bc14b5603474c9743fd04f3da888644413dc5
+ 891bc14b5603474c9743fd04f3da888644413dc5 [label=c.patch]
+ bc7df666a646739eaf559af23cab72f2bfd01f0e -
891bc14b5603474c9743fd04f3da888644413dc5; // ?
+# checking rev c7014443c33d2b0237293687ceb9cbd38313df65
+ c7014443c33d2b0237293687ceb9cbd38313df65 [label=b.patch]
+# checking rev ff2775f8d1dc753f635830adcc3a067e0b681e2d
+ ff2775f8d1dc753f635830adcc3a067e0b681e2d [label=a.patch]
+ 891bc14b5603474c9743fd04f3da888644413dc5 -
ff2775f8d1dc753f635830adcc3a067e0b681e2d; // ?
+}
diff --git a/regression/t-033.sh b/regression/t-033.sh
index fac081e..9fe1827 100755
--- a/regression/t-033.sh
+++ b/regression/t-033.sh
@@ -50,3 +50,12 @@ cmd git add file.txt
cmd guilt refresh
fixup_time_info c.patch
cmd guilt graph
+
+# A patch name that contains funky characters, including unbalanced
+# quotes.
+cmd guilt new a-\betterquicker'-patch.patch
+cmd echo d file.txt
+cmd git add file.txt
+cmd guilt refresh
+fixup_time_info a-\betterquicker'-patch.patch
+cmd guilt graph
--
1.8.3.1
--
To unsubscribe from this list: send the line unsubscribe git in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 21/33] The log.decorate setting should not influence import-commit.
Use --no-decorate in the call to git log that tries to read the commit message to produce patch names. Otherwise, if the user has set log.decorate to short or full, the patch name will be less useful. Modify the t-034.sh test case to demonstrate that this is needed. Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- guilt-import-commit | 2 +- regression/t-034.out | 2 ++ regression/t-034.sh | 2 ++ 3 files changed, 5 insertions(+), 1 deletion(-) diff --git a/guilt-import-commit b/guilt-import-commit index 45f2404..1da7c8e 100755 --- a/guilt-import-commit +++ b/guilt-import-commit @@ -26,7 +26,7 @@ disp About to begin conversion... 2 disp Current head: `git rev-parse \`git_branch\`` 2 for rev in `git rev-list $rhash`; do - s=`git log --pretty=oneline -1 $rev | cut -c 42-` + s=`git log --no-decorate --pretty=oneline -1 $rev | cut -c 42-` # Try to convert the first line of the commit message to a # valid patch name. diff --git a/regression/t-034.out b/regression/t-034.out index bda4399..5d81bd4 100644 --- a/regression/t-034.out +++ b/regression/t-034.out @@ -232,6 +232,7 @@ Date: Mon Jan 1 00:00:00 2007 + Signed-off-by: Commiter Name commiter@email % guilt init +% git config log.decorate short % guilt import-commit base..HEAD About to begin conversion... Current head: 2a8b1889aa5066193bac978e6bf5073ffcfa6541 @@ -259,6 +260,7 @@ Converting 45e81b51 as the_sequence_.lock-_is_forbidden Converting eebb76e9 as the_sequence_-._is_forbidden Done. Current head: d4850419ccc1146c7169f500725ce504b9774ed0 +% git config log.decorate no % guilt push -a Applying patch..the_sequence_-._is_forbidden.patch Patch applied. diff --git a/regression/t-034.sh b/regression/t-034.sh index f41f958..648d009 100755 --- a/regression/t-034.sh +++ b/regression/t-034.sh @@ -57,7 +57,9 @@ cmd git log # Import all the commits to guilt. cmd guilt init +cmd git config log.decorate short cmd guilt import-commit base..HEAD +cmd git config log.decorate no for patch in .git/patches/master/*.patch; do touch -a -m -t $TOUCH_DATE $patch -- 1.8.3.1 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 22/33] The log.decorate setting should not influence patchbomb.
Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- guilt-patchbomb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/guilt-patchbomb b/guilt-patchbomb index 1231418..164b10c 100755 --- a/guilt-patchbomb +++ b/guilt-patchbomb @@ -47,7 +47,7 @@ if [ $? -ne 0 ]; then fi # display the list of commits to be sent as patches -git log --pretty=oneline $r | cut -c 1-8,41- | $pager +git log --no-decorate --pretty=oneline $r | cut -c 1-8,41- | $pager _disp Are these what you want to send? [Y/n] read n -- 1.8.3.1 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 23/33] The log.decorate setting should not influence guilt rebase.
Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- guilt-rebase | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/guilt-rebase b/guilt-rebase index fd28e48..a1714a0 100755 --- a/guilt-rebase +++ b/guilt-rebase @@ -66,7 +66,7 @@ pop_all_patches git merge --no-commit $upstream /dev/null 2 /dev/null disp -log=`git log -1 --pretty=oneline` +log=`git log -1 --no-decorate --pretty=oneline` disp HEAD is now at `echo $log | cut -c 1-7`... `echo $log | cut -c 41-` # -- 1.8.3.1 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 25/33] guilt push now fails when there are no more patches to push.
This makes it easier to script operations on the entire queue, for
example run the test suite on each patch in the queue:
guilt pop -a;while guilt push; do make test||break; done
This brings guilt push in line with the push operation in Mercurial
Queues (hg qpush), which fails when there are no patches to apply.
Updated the test suite.
guilt push -a still does not fail. (It successfully manages to
ensure that all patches are pushed, even if it did not have to do
anything to make it so.)
Signed-off-by: Per Cederqvist ced...@opera.com
Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net
---
guilt-push | 19 ++-
regression/t-020.out | 89
regression/t-020.sh | 13 +++-
3 files changed, 113 insertions(+), 8 deletions(-)
diff --git a/guilt-push b/guilt-push
index 67687e7..39c125e 100755
--- a/guilt-push
+++ b/guilt-push
@@ -56,6 +56,12 @@ fi
patch=$1
[ ! -z $all ] patch=-a
+# Treat guilt push as guilt push -n 1.
+if [ -z $patch ]; then
+ patch=1
+ num=t
+fi
+
if [ $patch = -a ]; then
# we are supposed to push all patches, get the last one out of
# series
@@ -65,7 +71,7 @@ if [ $patch = -a ]; then
die There are no patches to push.
fi
elif [ ! -z $num ]; then
- # we are supposed to pop a set number of patches
+ # we are supposed to push a set number of patches
[ $patch -lt 0 ] die Invalid number of patches to push.
@@ -78,11 +84,6 @@ elif [ ! -z $num ]; then
# clamp to minimum
[ $tidx -lt $eidx ] eidx=$tidx
-elif [ -z $patch ]; then
- # we are supposed to push only the next patch onto the stack
-
- eidx=`wc -l $applied`
- eidx=`expr $eidx + 1`
else
# we're supposed to push only up to a patch, make sure the patch is
# in the series
@@ -109,7 +110,11 @@ if [ $sidx -gt $eidx ]; then
else
disp File series fully applied, ends at patch `get_series |
tail -n 1`
fi
- exit 0
+ if [ -n $all ]; then
+ exit 0
+ else
+ exit 1
+ fi
fi
get_series | sed -n -e ${sidx},${eidx}p | while read p
diff --git a/regression/t-020.out b/regression/t-020.out
index 42433dc..bcb8797 100644
--- a/regression/t-020.out
+++ b/regression/t-020.out
@@ -270,6 +270,95 @@ index 000..8baef1b
+++ b/def
@@ -0,0 +1 @@
+abc
+% guilt push
+File series fully applied, ends at patch mode
+% guilt push -a
+File series fully applied, ends at patch mode
+% list_files
+d .git/patches
+d .git/patches/master
+d .git/refs/patches
+d .git/refs/patches/master
+f 22930c6d1f1938f298a4fca51c57e4b47171db21 .git/patches/master/mode
+f 413390f3906f16f30b054a4fb86c1e014b964504 .git/patches/master/remove
+f 71596bf71b72c2717e1aee378aabefbfa19ab7c8 .git/patches/master/status
+f 9c18cc7abe6b87f18503714a80a677b4094eb457 .git/patches/master/add
+f bacb4aad8a55fe4e7aa58a9ae169990bb764069f .git/patches/master/series
+f bc9ab2e0f5db99d483961e956e814d963f0309f8 .git/patches/master/modify
+r 33633e7a1aa31972f125878baf7807be57b1672d .git/refs/patches/master/modify
+r 37d588cc39848368810e88332bd03b083f2ce3ac .git/refs/patches/master/add
+r ccd56089d1b5305a9d35617cb7f6f4b06ffa68ba .git/refs/patches/master/mode
+r ffb7faa126a6d91bcdd44a494f76b96dd860b8b9 .git/refs/patches/master/remove
+% git log -p
+commit ccd56089d1b5305a9d35617cb7f6f4b06ffa68ba
+Author: Author Name author@email
+Date: Mon Jan 1 00:00:00 2007 +
+
+patch mode
+
+diff --git a/def b/def
+old mode 100644
+new mode 100755
+
+commit ffb7faa126a6d91bcdd44a494f76b96dd860b8b9
+Author: Author Name author@email
+Date: Mon Jan 1 00:00:00 2007 +
+
+patch remove
+
+diff --git a/abd b/abd
+deleted file mode 100644
+index fd3896d..000
+--- a/abd
/dev/null
+@@ -1 +0,0 @@
+-öuؽáZâñeÏÈE£ WÀV¼/U?Ú|¢@6¤8'H¸1G_ͧ*·ðRÒ¤
ªÂ~·
+\ No newline at end of file
+
+commit 37d588cc39848368810e88332bd03b083f2ce3ac
+Author: Author Name author@email
+Date: Mon Jan 1 00:00:00 2007 +
+
+patch add
+
+diff --git a/abd b/abd
+new file mode 100644
+index 000..fd3896d
+--- /dev/null
b/abd
+@@ -0,0 +1 @@
++öuؽáZâñeÏÈE£ WÀV¼/U?Ú|¢@6¤8'H¸1G_ͧ*·ðRÒ¤
ªÂ~·
+\ No newline at end of file
+
+commit 33633e7a1aa31972f125878baf7807be57b1672d
+Author: Author Name author@email
+Date: Mon Jan 1 00:00:00 2007 +
+
+patch modify
+
+diff --git a/def b/def
+index 8baef1b..7d69c2f 100644
+--- a/def
b/def
+@@ -1 +1,2 @@
+ abc
++asjhfksad
+
+commit d4850419ccc1146c7169f500725ce504b9774ed0
+Author: Author Name author@email
+Date: Mon Jan 1 00:00:00 2007 +
+
+initial
+
+Signed-off-by: Commiter Name commiter@email
+
+diff --git a/def b/def
+new file mode 100644
+index 000..8baef1b
+--- /dev/null
b/def
+@@ -0,0 +1 @@
++abc
% guilt pop --all
All patches popped.
% guilt push
diff --git a/regression/t-020.sh b/regression/t-020.sh
index
[GUILT v4 26/33] guilt pop now fails when there are no more patches to pop.
This is analogous to how guilt push now fails when there are no more patches to push. Like push, the --all argument still succeeds even if there was no need to pop anything. Updated the test suite. Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- guilt-pop| 17 +++-- regression/t-021.out | 2 ++ regression/t-021.sh | 6 ++ regression/t-061.sh | 6 +- 4 files changed, 24 insertions(+), 7 deletions(-) diff --git a/guilt-pop b/guilt-pop index f0e647f..191313e 100755 --- a/guilt-pop +++ b/guilt-pop @@ -49,9 +49,19 @@ fi patch=$1 [ ! -z $all ] patch=-a +# Treat guilt pop as guilt pop -n 1. +if [ -z $patch ]; then + patch=1 + num=t +fi + if [ ! -s $applied ]; then disp No patches applied. - exit 0 + if [ $patch = -a ]; then + exit 0 + else + exit 1 + fi elif [ $patch = -a ]; then # we are supposed to pop all patches @@ -68,11 +78,6 @@ elif [ ! -z $num ]; then # catch underflow [ $eidx -lt 0 ] eidx=0 [ $eidx -eq $sidx ] die No patches requested to be removed. -elif [ -z $patch ]; then - # we are supposed to pop only the current patch on the stack - - sidx=`wc -l $applied` - eidx=`expr $sidx - 1` else # we're supposed to pop only up to a patch, make sure the patch is # in the series diff --git a/regression/t-021.out b/regression/t-021.out index 9b42d9c..58be12f 100644 --- a/regression/t-021.out +++ b/regression/t-021.out @@ -287,6 +287,8 @@ index 000..8baef1b +++ b/def @@ -0,0 +1 @@ +abc +% guilt pop +No patches applied. % guilt push --all Applying patch..modify Patch applied. diff --git a/regression/t-021.sh b/regression/t-021.sh index 614e870..e0d2dc1 100755 --- a/regression/t-021.sh +++ b/regression/t-021.sh @@ -23,6 +23,12 @@ guilt series | _tac | while read n ; do done # +# pop when there is nothing to pop +# + +shouldfail guilt pop + +# # push all # cmd guilt push --all diff --git a/regression/t-061.sh b/regression/t-061.sh index 1411baa..6192f1b 100755 --- a/regression/t-061.sh +++ b/regression/t-061.sh @@ -48,7 +48,11 @@ cmd list_files for i in `seq 5` do - cmd guilt pop + if [ $i -ge 5 ]; then + shouldfail guilt pop + else + cmd guilt pop + fi cmd git for-each-ref cmd guilt push cmd git for-each-ref -- 1.8.3.1 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 27/33] Minor testsuite fix.
Fix remove_topic() in t-061.sh so that it doesn't print a git hash.
Signed-off-by: Per Cederqvist ced...@opera.com
Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net
---
regression/t-061.out | 1 -
regression/t-061.sh | 2 +-
2 files changed, 1 insertion(+), 2 deletions(-)
diff --git a/regression/t-061.out b/regression/t-061.out
index ef0f335..60ad56d 100644
--- a/regression/t-061.out
+++ b/regression/t-061.out
@@ -381,7 +381,6 @@ ccd56089d1b5305a9d35617cb7f6f4b06ffa68ba commit
refs/patches/master/mode
ffb7faa126a6d91bcdd44a494f76b96dd860b8b9 commit
refs/patches/master/remove
% guilt pop -a
No patches applied.
-ccd56089d1b5305a9d35617cb7f6f4b06ffa68ba
% git checkout guilt/master
Switched to branch guilt/master
% guilt pop -a
diff --git a/regression/t-061.sh b/regression/t-061.sh
index 6192f1b..db26e12 100755
--- a/regression/t-061.sh
+++ b/regression/t-061.sh
@@ -15,7 +15,7 @@ old_style_branch() {
remove_topic() {
cmd guilt pop -a
- if git rev-parse --verify --quiet guilt/master
+ if git rev-parse --verify --quiet guilt/master /dev/null
then
cmd git checkout guilt/master
else
--
1.8.3.1
--
To unsubscribe from this list: send the line unsubscribe git in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 28/33] Fix coding style errors in t-061.sh.
Signed-off-by: Per Cederqvist ced...@opera.com
Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net
---
regression/t-061.sh | 6 ++
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/regression/t-061.sh b/regression/t-061.sh
index db26e12..bda50c7 100755
--- a/regression/t-061.sh
+++ b/regression/t-061.sh
@@ -15,8 +15,7 @@ old_style_branch() {
remove_topic() {
cmd guilt pop -a
- if git rev-parse --verify --quiet guilt/master /dev/null
- then
+ if git rev-parse --verify --quiet guilt/master /dev/null; then
cmd git checkout guilt/master
else
cmd git checkout master
@@ -46,8 +45,7 @@ cmd git for-each-ref
cmd list_files
-for i in `seq 5`
-do
+for i in `seq 5`; do
if [ $i -ge 5 ]; then
shouldfail guilt pop
else
--
1.8.3.1
--
To unsubscribe from this list: send the line unsubscribe git in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 29/33] Added guilt.reusebranch configuration option.
When the option is true, Guilt does not create a new Git branch when patches are applied. This way, you can switch between Guilt 0.35 and the current version of Guilt with no issues. By default, the option is false, so that all users will immediately get the added protection against pushing a branch upstream with a patch applied. While this might break guilt if a user is running both version 0.35 and the current version against the same local repository, it will not lead to data loss, and that situation is probably rare. Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- guilt| 24 ++- regression/scaffold | 1 + regression/t-062.out | 420 +++ regression/t-062.sh | 130 4 files changed, 569 insertions(+), 6 deletions(-) create mode 100644 regression/t-062.out create mode 100755 regression/t-062.sh diff --git a/guilt b/guilt index 0500500..ea9d131 100755 --- a/guilt +++ b/guilt @@ -884,6 +884,9 @@ guilt_push_diff_context=1 # default diffstat value: true or false DIFFSTAT_DEFAULT=false +# default guilt.reusebranch value: true or false +REUSE_BRANCH_DEFAULT=false + # Prefix for guilt branches. GUILT_PREFIX=guilt/ @@ -895,6 +898,10 @@ GUILT_PREFIX=guilt/ diffstat=`git config --bool guilt.diffstat` [ -z $diffstat ] diffstat=$DIFFSTAT_DEFAULT +# reuse Git branch? +reuse_branch=`git config --bool guilt.reusebranch` +[ -z $reuse_branch ] reuse_branch=$REUSE_BRANCH_DEFAULT + # # The following gets run every time this file is source'd # @@ -959,13 +966,18 @@ else die Unsupported operating system: $UNAME_S fi -if [ $branch = $raw_git_branch ] [ -n `get_top 2/dev/null` ] -then -# This is for compat with old repositories that still have a -# pushed patch without the new-style branch prefix. -old_style_prefix=true +if [ -n `get_top 2/dev/null` ]; then + # If there is at least one pushed patch, we set + # old_style_prefix according to how it was pushed. It is only + # possible to change the prefix style while no patches are + # applied. + if [ $branch = $raw_git_branch ]; then + old_style_prefix=true + else + old_style_prefix=false + fi else -old_style_prefix=false + old_style_prefix=$reuse_branch fi _main $@ diff --git a/regression/scaffold b/regression/scaffold index 2e04c83..b3135be 100644 --- a/regression/scaffold +++ b/regression/scaffold @@ -97,6 +97,7 @@ function setup_git_repo git config log.date default git config log.decorate no git config guilt.diffstat false + git config guilt.reusebranch false } function setup_guilt_repo diff --git a/regression/t-062.out b/regression/t-062.out new file mode 100644 index 000..ad5c081 --- /dev/null +++ b/regression/t-062.out @@ -0,0 +1,420 @@ +% setup_repo +% git config guilt.reusebranch true +% guilt push -a +Applying patch..modify +Patch applied. +Applying patch..add +Patch applied. +Applying patch..remove +Patch applied. +Applying patch..mode +Patch applied. +% list_files +d .git/patches +d .git/patches/master +d .git/refs/patches +d .git/refs/patches/master +f 22930c6d1f1938f298a4fca51c57e4b47171db21 .git/patches/master/mode +f 413390f3906f16f30b054a4fb86c1e014b964504 .git/patches/master/remove +f 71596bf71b72c2717e1aee378aabefbfa19ab7c8 .git/patches/master/status +f 9c18cc7abe6b87f18503714a80a677b4094eb457 .git/patches/master/add +f bacb4aad8a55fe4e7aa58a9ae169990bb764069f .git/patches/master/series +f bc9ab2e0f5db99d483961e956e814d963f0309f8 .git/patches/master/modify +r 33633e7a1aa31972f125878baf7807be57b1672d .git/refs/patches/master/modify +r 37d588cc39848368810e88332bd03b083f2ce3ac .git/refs/patches/master/add +r ccd56089d1b5305a9d35617cb7f6f4b06ffa68ba .git/refs/patches/master/mode +r ffb7faa126a6d91bcdd44a494f76b96dd860b8b9 .git/refs/patches/master/remove +% git for-each-ref +ccd56089d1b5305a9d35617cb7f6f4b06ffa68ba commitrefs/heads/master +37d588cc39848368810e88332bd03b083f2ce3ac commitrefs/patches/master/add +ccd56089d1b5305a9d35617cb7f6f4b06ffa68ba commitrefs/patches/master/mode +33633e7a1aa31972f125878baf7807be57b1672d commit refs/patches/master/modify +ffb7faa126a6d91bcdd44a494f76b96dd860b8b9 commit refs/patches/master/remove +% guilt pop +Now at remove. +% git for-each-ref +ffb7faa126a6d91bcdd44a494f76b96dd860b8b9 commitrefs/heads/master +37d588cc39848368810e88332bd03b083f2ce3ac commitrefs/patches/master/add +33633e7a1aa31972f125878baf7807be57b1672d commit refs/patches/master/modify +ffb7faa126a6d91bcdd44a494f76b96dd860b8b9 commit refs/patches/master/remove +% guilt push +Applying patch..mode +Patch applied. +% git for-each-ref +ccd56089d1b5305a9d35617cb7f6f4b06ffa68ba commitrefs/heads/master +37d588cc39848368810e88332bd03b083f2ce3ac commit
[GUILT v4 30/33] Added a short style guide, and Emacs settings.
Signed-off-by: Per Cederqvist ced...@opera.com Signed-off-by: Josef 'Jeff' Sipek jef...@josefsipek.net --- .dir-locals.el | 3 +++ Documentation/Contributing | 15 +++ 2 files changed, 18 insertions(+) create mode 100644 .dir-locals.el diff --git a/.dir-locals.el b/.dir-locals.el new file mode 100644 index 000..50ef2b7 --- /dev/null +++ b/.dir-locals.el @@ -0,0 +1,3 @@ +((nil . ((indent-tabs-mode . t) +(tab-width . 8))) + (sh-mode . ((sh-basic-offset . 8 diff --git a/Documentation/Contributing b/Documentation/Contributing index abf3480..0da49d6 100644 --- a/Documentation/Contributing +++ b/Documentation/Contributing @@ -4,6 +4,21 @@ Documentation/SubmittingPatches file. :) 1) Hack on the code a bit +Please follow this style guide: + + - Use tabs for indentation. + + - Put then on the same line as if. + + - Follow the style of the existing code, except if it breaks the + above guidlines. + + - If you change the code to conform to the style guide, please do so + in a separate commit that does not change anything else. + +Please check that you change does not break make test. Please add +new testcases for any new functionality, and if you fix a bug. + 2) Make a patch: Use diff -up or diff -uprN to create patches. Or simply use git to -- 1.8.3.1 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 32/33] Improved doc and tests for guilt header.
--- Documentation/guilt-header.txt | 5 - regression/t-028.out | 9 + regression/t-028.sh| 3 +++ 3 files changed, 16 insertions(+), 1 deletion(-) diff --git a/Documentation/guilt-header.txt b/Documentation/guilt-header.txt index 870bfaf..71b2e66 100644 --- a/Documentation/guilt-header.txt +++ b/Documentation/guilt-header.txt @@ -18,7 +18,10 @@ Prints either the topmost patch's header or the header of a specified patch. -E:: Open the raw patch in an editor, instead of printing it. patchname:: - Name of the patch. + Name of the patch. If a patch with exactly this name exists, + use it. Otherwise, treat the name as a regexp; if the regexp + matches a single patch, use it. Otherwise, list all matching + patch names to stderr and fail. Author -- diff --git a/regression/t-028.out b/regression/t-028.out index ea72a3a..39ac900 100644 --- a/regression/t-028.out +++ b/regression/t-028.out @@ -56,3 +56,12 @@ Patch non-existant is not in the series remove mode patch-with-some-desc +% guilt header de +de does not uniquely identify a patch. Did you mean any of these? + mode + patch-with-some-desc +% guilt header des +blah blah blah + +Signed-off-by: Commiter Name commiter@email + diff --git a/regression/t-028.sh b/regression/t-028.sh index 2ce0378..cd3088c 100755 --- a/regression/t-028.sh +++ b/regression/t-028.sh @@ -35,4 +35,7 @@ shouldfail guilt header non-existant # patch name is a regexp that just happens to match an existing patch. shouldfail guilt header '.*' +shouldfail guilt header de +cmd guilt header des + # FIXME: how do we check that -e works? -- 1.8.3.1 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[GUILT v4 33/33] Document the exit status of guilt push and guilt pop.
--- Documentation/guilt-pop.txt | 3 +++ Documentation/guilt-push.txt | 3 +++ 2 files changed, 6 insertions(+) diff --git a/Documentation/guilt-pop.txt b/Documentation/guilt-pop.txt index 36fea9e..b0b89cc 100644 --- a/Documentation/guilt-pop.txt +++ b/Documentation/guilt-pop.txt @@ -26,6 +26,9 @@ OPTIONS If no patchname is given, pop the top-most patch. +Exit with a non-zero exit status if requested to pop more patches +than there are on the stack. + Author -- Written by Josef Jeff Sipek jef...@josefsipek.net diff --git a/Documentation/guilt-push.txt b/Documentation/guilt-push.txt index 6ee86b9..6439f21 100644 --- a/Documentation/guilt-push.txt +++ b/Documentation/guilt-push.txt @@ -26,6 +26,9 @@ OPTIONS If no patchname is given, push the next patch in the series onto the tree. +Exit with a non-zero exit status if requested to push more patches +than there are in the series. + Author -- Written by Josef Jeff Sipek jef...@josefsipek.net -- 1.8.3.1 -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] remote-helpers: point at their upstream repositories
Matthieu Moy wrote:
Felipe Contreras felipe.contre...@gmail.com writes:
% git fetch
WARNING: git-remote-hg is now maintained independently.
WARNING: For more information visit
https://github.com/felipec/git-remote-hg
searching for changes
no changes found
I don't think the situation is as simple as you claim. In many cases,
the first step before the ones you are mentionning are:
cd $git/contrib/remote-helpers
cp git-remote-{hg,bzr} somewhere/in/path
In many cases, but not all cases. In other cases they are:
ln -s $git/contrib/remote-helpers/git-remote-hg \
somewhere/in/path/git-remote-hg
Which has to be done only once, and not every time Git is updated.
They produces no warning if git-remote-{hg,bzr} exist with the warning,
but no such file or directory: contrib/remote-helpers if the directory
has been renamed or removed.
So? They'll get the warning the next time they try to use it, and their
workflow won't be interrupted, and the warning will be more useful than
No such file or directory.
When git-remote-{hg,bzr} are installed with a package manager, the fact
that they are part of Git's core or not is often irrelevant. For
example, Debian splits the git.git source into many packages, so a
Debian user will not see any difference between helpers included in
git.git or outside (e.g. I have to install the package git-svn if I want
to use git-svn).
Yet there is no git-hg. And I doubt Jonathan Nieder is going to package
the out-of-tree git-bzr, which as far as I know, is the only out-of-tree
package of these remote helpers in any distro.
--
Felipe Contreras
--
To unsubscribe from this list: send the line unsubscribe git in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] remote-helpers: point at their upstream repositories
Felipe Contreras felipe.contre...@gmail.com writes: But that being said, this is Felipe's code. While we have a legal right to distribute it in v2.0, if he would really prefer it out for v2.0, I would respect that. I am fine with that. Are you? Because in two of the three options you list below you wouldn't be doing that. that does not refer to remove them at v2.0 (unconditional). It refers to If Felipe really wants for the removal for v2.0, I would respect that. And I saw you said you did not want to disrupt v2.0. If the options I listed all meant removal at v2.0, then I would understand your complaints, but that is not the case, so I am not sure what to make of that. The fact of the matter is that users cannot depend on packages any more. Maybe they'll be packaged, maybe not. If they are it will take a long time before they do. In the meantime they'll have to manually install them all all out-of-tree tools. I have always thought that distro packagers are the biggest ally us project leaders have. They locate useful pieces of software, massage them into a shape that fit their distro well and deliver what we write to their audience. Packaging stuff that are useful to their end-users is what they do best, and not leaving useful stuff unpackaged is in their best interest. Your statement makes it sound like they are incompetent lazy fools who do not know what is useful for their users. I find it disturbing to see such a distrust. Or am I being too naive to have too much faith in packaging folks? I checked the list of packages that depend on git on one of my boxes (it is a bit old Ubuntu). I of course expected that many of them are what comes from our tree split into their own niche tool packages (e.g. git-svn, git-gui, gitweb...), but I was pleasantly surprised to see many that I haven't even aware of being packaged. Of course, tig is among the packages that depend on us which I am happy to see. There are things of somewhat questionable value I saw in the list, of course. It is already 2014, and I feel fairly safe to feel that I can say without offending too many people that I doubt git-arch would be on such a list of packages distros offer to their users, if it were written as a third-party plug-in today. It is an (odd) example of a package that is still there mostly by inertia at this point, and that inertia comes from many things. It is in our tree outside contrib/, it was found useful once in the past and was packaged, the packager already has infrastructure to cut a separate package out of our tree, and it is more trouble to retire it and risk breaking minority users than just keep shipping it. But hg is not in a situation similar to tla, is it? I simply cannot imagine there is no history worthwhile to salvage out of Mercurial repositories coming anytime in the near future. After looking at the reverse-depends list of packages, my faith is strengthened in that the Git ecosystem is truly maturing and useful third-party plug-ins will be picked up by distro packagers. Am I delusional? -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] remote-helpers: point at their upstream repositories
Jeff King p...@peff.net writes: My concerns were with people not noticing the README. Removing the code entirely is the way I thought of to address that. Junio suggested another way, which I would also be fine with. And it seems like a friendlier way than removal to handle it for v2.0, if we are going to remove the code entirely post-v2.0. As before, if your desire is to have the code out for v2.0, then say so. I think we should respect such a request. OK. After thinking about it a bit more, I think renaming the directory at this point is not a clearly superiour option and the judgment largely depends on your philosophy on transition. Let me explain, backwards from the desired endgame. I think all of us (including Felipe) agree that in some future time [*1*], we won't have either of these two scripts in contrib/, but just like contrib/vim, we will leave a README to help those who read a stale page on the Web that says remote-hg in contrib/ is the officially recommended tool to interact with Mercurial. The README will tell them that they read a stale page that is no longer true, and direct them to where they can grab remote-hg/bzr. We will need to keep contrib/remote-helpers in that endgame state for quite some time. If a user of 1.9.x updates to such an endgame version and then runs the same fetch from Hg, he will not just notice the breakage but is forced at that point to go to the GitHub URL and switch, but it may not be a convenient time for him to go through that process. To help these users, a step that ships the stale but working scripts that give warning is necessary before the endgame state, and Felipe's 77621193 (contrib: remote-helpers: add move warnings (v2.0), 2014-05-13) is such a change. My suggestion to rename the directory without smudging the scripts was meant to be a step that can come before that step, and I think its necessity is debatable. It depends on how gradual a transition you want to give, and being always the more cautious type, I think having such a step will give packagers who pay attention to what they package and users who pay attention to what they install without packaging an early chance to notice and prepare. - The endgame will force the user to update at the point when the user needs to use it for his real work, when he may not have time for sysadmin. - The always warn does not force update at the point of use, but it still does not help them to notice well before they try to use it for the first time after update; - Break the build attempts to help them notice when they try to update, not when they need to use the updated one right at this moment. But I am fine with an expedited transition schedule without the break the build step. That was an optional first step, because warn but still work state we must have before the endgame will give the users the choice of when to adjust anyway. I also thought about adding an extra step to have even more gradual transition, by the way. A step before the endgame will ship these scripts without anything but instruct and fail (this is not warn and fail, as it is too late warn, as the scripts are crippled not to work at this point). That will still force the user to update at the point when the user needs to use it, but seeing the instruction (e.g. run this curl command to fetch from this URL and store it in a file called git-remote-xx on your $PATH) that is easy to follow immediately would be better than seeing only a failure (i.e. remote-hg not found), having to go fish the README, visiting the GitHub pages and figuring out how to fetch and install the script, which would be what the user will get with README only, no scripts endgame. For that matter, the warning message given by 77621193, also README added by f000c4e6 (remote-helpers: point at their upstream repositories, 2014-05-15) may want to mention not just the GitHub URL for the repository as the whole, but the URL to fetch the latest blob with curl/wget, if we really want to help users go back to their tasks at hand with minimum interruption. I know how to construct a URL to follow the tip of a specific branch and obtain a full .zip from a GitHub repository, but I do not know offhand if you can do the same for a single blob. If we can come up with one, we should add such a instruction to the warning message and README, as that instruction will be the only thing to help the users in the endgame state anyway. So to summarize, the following timeline is a full possibility: 1. (optional) break the build by renaming directory and add README. Include not just the repository URL but a blob URL and instruction to download via wget/curl. 2. add warning that is given every time the scripts are run and give the same instruction as in README. 3. (optional) cripple the script to make them always fail after showing the same warning as above. 4. Keep README and retire everything else. but I am perfectly
Re: [PATCH] rebase --keep-empty -i: add test
On Sun, May 18, 2014 at 11:28:39PM +0300, Michael S. Tsirkin wrote: There's some special code in rebase -i to deal with --keep-empty. Add test for this combination. Signed-off-by: Michael S. Tsirkin m...@redhat.com Acked-by: Neil Horman nhor...@tuxdriver.com --- t/t3404-rebase-interactive.sh | 9 + 1 file changed, 9 insertions(+) diff --git a/t/t3404-rebase-interactive.sh b/t/t3404-rebase-interactive.sh index c0023a5..3b1b863 100755 --- a/t/t3404-rebase-interactive.sh +++ b/t/t3404-rebase-interactive.sh @@ -67,6 +67,14 @@ test_expect_success 'setup' ' SHELL= export SHELL +test_expect_success 'rebase --keep-empty' ' + git checkout -b emptybranch master + git commit --allow-empty -m empty + git rebase --keep-empty -i HEAD~2 + git log --oneline actual + test_line_count = 6 actual +' + test_expect_success 'rebase -i with the exec command' ' git checkout master ( -- MST -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 2/4] git-rebase: document ack
On Sun, May 18, 2014 at 5:17 PM, Michael S. Tsirkin m...@redhat.com wrote: document ack! behaviour and use Signed-off-by: Michael S. Tsirkin m...@redhat.com --- Documentation/git-rebase.txt | 45 +++- 1 file changed, 40 insertions(+), 5 deletions(-) diff --git a/Documentation/git-rebase.txt b/Documentation/git-rebase.txt index 2a93c64..c27aef4 100644 --- a/Documentation/git-rebase.txt +++ b/Documentation/git-rebase.txt @@ -384,7 +384,7 @@ or by giving more than one `--exec`: + If `--autosquash` is used, exec lines will not be appended for the intermediate commits, and will only appear at the end of each -squash/fixup series. +squash/fixup/ack series. --root:: Rebase all commits reachable from branch, instead of @@ -398,13 +398,13 @@ squash/fixup series. --autosquash:: --no-autosquash:: - When the commit log message begins with squash! ... (or - fixup! ...), and there is a commit whose title begins with + When the commit log message begins with squash! ... (fixup! ... + or ack! ...), and there is a commit whose title begins with the same ..., automatically modify the todo list of rebase -i so that the commit marked for squashing comes right after the commit to be modified, and change the action of the moved - commit from `pick` to `squash` (or `fixup`). Ignores subsequent - fixup! or squash! after the first, in case you referred to an + commit from `pick` to `squash` (`fixup` or `ack`). Ignores subsequent + ack! , fixup! or squash! after the first, in case you referred to an earlier fixup/squash with `git commit --fixup/--squash`. + This option is only valid when the '--interactive' option is used. @@ -624,6 +624,41 @@ consistent (they compile, pass the testsuite, etc.) you should use 'git stash' to stash away the not-yet-committed changes after each commit, test, and amend the commit if fixes are necessary. + +RECORDING ACKS + + +Interactive mode with --autosquash can be used to concatenate +commit log for several commits, which is useful to record +extra information about the commit, such as ack signatures. +This allows, for example, the following workflow: + +1. receive patches by mail and commit +2. receive by mail ack signatures for the patches +3. prepare a series for submission +4. submit + +where point 2. consists of several instances of + i) create a (possibly empty) commit with signature + in the commit message + +Sometimes the ack signature added in i. cannot be amended to the +commit it acks, because that commit is buried deeply in a +patch series. That is exactly what rebase --autosquash +option is for: use it +after plenty of is, to automaticlly rearrange +commits, and squashing multiple sign-off commits into +the commit that is signed. + +Start it with the last commit you want to retain as-is: + + git rebase --autosquash -i after-this-commit + +An editor will be fired up with all the commits in your current branch +which come after the given commit. Ack commits will be +re-arranged to come after the commit that is acked, +and the action will be utomticlly changed from `pick` to `ack` s/utomticlly/automatically/ +to cause them to be squashed into the acked commit. RECOVERING FROM UPSTREAM REBASE --- -- MST -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[ANNOUNCE] git reintegrate v0.3; manager of integration branches
Hi, git reintegrate is a helper tool to manage integration branches, it has all the options of other known tools. This is a rewrite of John Keeping's git-integration in Ruby, it has essentially the same features and passes all the git-integration tests, but it has more features. One feature that is missing from git-integration is the ability to parse existing integration branches. To give a try you can do: git clone https://github.com/gitster/git/ cd git git fetch -u origin 'refs/tags/*:refs/tags/*' 'refs/heads/*:refs/heads/*' git checkout pu git reintegrate --generate pu master Which will generate the integration instructions for you: % git reintegrate --cat base master merge jl/submodule-mv Moving a regular file in a repository with a .gitmodules file was producing a warning 'Could not find section in .gitmodules where path=filename'. merge ap/remote-hg-unquote-cquote A fast-import stream expresses a pathname with funny characters by quoting them in C style; remote-hg remote helper forgot to unquote such a path. merge jk/for-each-ref-skip-parsing merge jk/pack-corruption-post-mortem merge jk/reset-p-current-head-fix git reset -p HEAD has codepath to special case it from resetting to contents of other commits, but recent change broke it. ... It also has support for evil merges, so it should be perfectly usable for git.git maintenance. You can edit the instructions with `git reintegrate --edit`. The simplest way to begin an integration branch is with: git reintegrate --create pu master git reintegrate --add=branch1 --add=branch2 --add=branch3 To generate the integration branch run `git reintegrate --rebuild`, if there are merge conflicts, solve them and continue with `git reintegrate --continue`. Despite having more features, the code is actually smaller thanks to Ruby awesomeness. Enjoy. https://github.com/felipec/git-reintegrate Changes since v0.1: * Add support for empty commits * Add support for pause command * Update manpage * Add bash completion Felipe Contreras (26): Add copyright and license notices Fix EDITOR support with arguments Trivial style cleanup Improve command regex Add support for empty commits test: improve check_int() Add support for 'pause' command doc: rename manpage file doc: update options doc: add description doc: add missing instruction commands doc: cleanup . command Verify branches after parsing test: fix test names Remove unused statements Update README test: cleanup instruction sheets Update copyright notices Add bash completion build: add installation stuff readme: add installation instructions Add gitignore for documentation trvis: initial configuration travis: add verbosity test: add test-lib helper travis: remove Ruby 1.8 -- Felipe Contreras -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[ANNOUNCE] git related v0.3
Hi, This tool finds people that might be interested in a patch, by going back through the history for each single hunk modified, and finding people that reviewed, acknowledged, signed, or authored the code the patch is modifying. It does this by running `git blame` incrementally on each hunk, and finding the relevant commit message. After gathering all the relevant people, it groups them to show what exactly was their role when the participated in the development of the relevant commit, and on how many relevant commits they participated. They are only displayed if they pass a minimum threshold of participation. It is similar the the `git contacts` tool in the contrib area, which is a rewrite of this tool, except that `git contacts` does the absolute minimum; `git related` is way superior in every way. For example: % git related master..fc/transport/improv Junio C Hamano gits...@pobox.com (signer: 90%, author: 5%) Felipe Contreras felipe.contre...@gmail.com (author: 25%, reviewer: 2%) Sverre Rabbelier srabbel...@gmail.com (author: 17%, acker: 2%, signer: 7%) Jeff King p...@peff.net (acker: 17%, author: 10%) Shawn O. Pearce spea...@spearce.org (author: 5%, signer: 2%, cced: 2%) Elijah Newren new...@gmail.com (author: 10%) In addition, it has an option to output the list of commits, instead of the contributors, which allows you to easily find out the previous changes to the lines your patches modify. % git related -c master..fc/transport/improv 99d9ec0 Merge branch 'fc/transport-helper-no-refspec' 67c9c78 transport-helper: barf when user tries old:new 0460ed2 documentation: trivial style cleanups 126aac5 transport-helper: fix remote helper namespace regression 21610d8 transport-helper: clarify pushing without refspecs a93b4a0 transport-helper: warn when refspec is not used 664059f transport-helper: update remote helper namespace c4458ec fast-export: Allow pruned-references in mark file ... Moreover, when sending patches for review, you can configure `git send-email` to use `git related` to find relevant people that should be Cc'ed: % git send-email --cc-cmd='git related' *.patch Compared to `git related`, `git contacts` has the following limitations: 1) Doesn't show the amount of involvement 2) Doesn't show the kind of involvement (reviewer, author) nor does it group people by their email address 3) Doesn't have the option to show the commit themselves 4) Doesn't have any options at all (--since, --min-percent) Cheers. https://github.com/felipec/git-related Changes since v0.1: * Fix compatibility with older versions * Add -clong option * Add manpage * Improve performance by grouping line ranges Felipe Contreras (12): Fix compatibility with Ruby 1.9 Add support for Ruby 1.8 Fix popen workaround Refactor blame parsing Pass multiple ranges to `git blame` test: add gitingnore file Add -clong option Add manpage build: add installation stuff readme: trivial updates test: add test-lib helper travis: initial configuration -- Felipe Contreras -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[ANNOUNCE] git send-series v0.1
Hi, This tool allows you to maintain a patch series more easily. You can store the cover letter, the cc list, version of the series, and other metadata. By default it will use the currently checked out branch, and show you a template like this: version: cc: Subject Content. The first part is a YAML document with all optional fields, such as 'version', 'cc', 'to'. You can store any information you want using the YAML syntax. The second part, delimited by a blank line, is the cover letter. The first like will be used as the subject of the cover letter email, and the rest as the content of the email. Then standard tools will be used to send the emails: `git format-patch` and `git send-email`, which you must have configured before using this helper. In the end a custom ref will be created to save the current state of the branch. For example, if you want to see what was the status of your series of the branch 'feature-a' in version 2, you can use 'sent/feature-a/v2', for example to create an interdiff to see what changed between one version an the other. Enjoy. https://github.com/felipec/git-send-series -- Felipe Contreras -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] remote-helpers: point at their upstream repositories
Junio C Hamano wrote: Felipe Contreras felipe.contre...@gmail.com writes: But that being said, this is Felipe's code. While we have a legal right to distribute it in v2.0, if he would really prefer it out for v2.0, I would respect that. I am fine with that. Are you? Because in two of the three options you list below you wouldn't be doing that. that does not refer to remove them at v2.0 (unconditional). It refers to If Felipe really wants for the removal for v2.0, I would respect that. And I saw you said you did not want to disrupt v2.0. If the options I listed all meant removal at v2.0, then I would understand your complaints, but that is not the case, so I am not sure what to make of that. It is a weird choice of semantics then. You said you would respect my wish, but your proposals did not follow my wish. The fact of the matter is that users cannot depend on packages any more. Maybe they'll be packaged, maybe not. If they are it will take a long time before they do. In the meantime they'll have to manually install them all all out-of-tree tools. I have always thought that distro packagers are the biggest ally us project leaders have. They locate useful pieces of software, massage them into a shape that fit their distro well and deliver what we write to their audience. Packaging stuff that are useful to their end-users is what they do best, and not leaving useful stuff unpackaged is in their best interest. Yet I bet a lot of open-source software is not actually packaged. That's the reason there's Python's pip, and Ruby gems. *If* your software is popular enough, then yes, packagers are your biggest allys, but if not, they aren't. Your statement makes it sound like they are incompetent lazy fools who do not know what is useful for their users. This sentence proves you have no idea how packaging is done. There is no comittee that hunts down packages that are useful for their users and assigns available packagers to those projects. Exactly the same way you don't assign Git developers tasks based on what is useful for our users. Each packager decides what project they package, just like every Git developer decides on what feature they work on. An obscure package might be packaged because a prominent Debian package maintainer likes it, and a much more useful and popular project might not be packaged simply because no package maintainer is interested in it. Exactly the same happens in Git; people work on relatively obsucre features such as ref transactions, because they are interested in them, and features much more useful for our users get ignored, because there's nobody (of relevance) championing them. When a popular project that is useful for the users is neglected for too long, what usually happens is that an outsider steps up and does the packaging, which then goes through a review process, and that outsider might become an official maintainer, and maybe start to package other things too. That's how packagers join the project. But nothing gets done if no ousider steps up. Excatly the same happens in Git; when a feature has been neglected for too long, an outsider comes and tries to implement it, go through a review process, and eventually start fixing other things too. So no, there's no comittee that decides what should be packaged, just like there's no committe that decides what Git features should be developed. It's incredibly alarming that you would think packagers in open source distributions would work any other way. And it's incredibly funny that you would label people working on such model as incompetent lazy fools for not knowing what is useful for their users, when it is *EXACTLY* the same thing you do in Git; you do not know what is useful for our users; you don't actually care; you just work on whatever you like to work on. It's even worst than that, because if somebody steps up to package a popular project, the package goes in, but when somebody steps up to implement a feature that improves our user-interface in Git; they get their knees shot. I find it disturbing to see such a distrust. Or am I being too naive to have too much faith in packaging folks? There is so much wrong with your mode of thinking and the blindness that you don't see what you yourself do that I don't even know where to begin. Yes you are too naive, on many levels. I checked the list of packages that depend on git on one of my boxes (it is a bit old Ubuntu). I of course expected that many of them are what comes from our tree split into their own niche tool packages (e.g. git-svn, git-gui, gitweb...), but I was pleasantly surprised to see many that I haven't even aware of being packaged. Of course, tig is among the packages that depend on us which I am happy to see. There are things of somewhat questionable value I saw in the list, of course. It is already 2014, and I feel fairly safe to feel that I can say without offending too
bug: autostash is lost after aborted rebase
scenario: - edit some tracked files; do not add them to the index - git config rebase.autostash true - git rebase -i HEAD~3 (an autostash will be created) - delete the entire buffer and save/exit the editor - this will abort the rebase poof, the autostash is gone (it is not reapplied) -- it must be explicitly applied again via the SHA that was printed earlier. (please cc me; I am not subscribed to the list.) -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] remote-helpers: point at their upstream repositories
Junio C Hamano wrote: My suggestion to rename the directory without smudging the scripts was meant to be a step that can come before that step, and I think its necessity is debatable. It depends on how gradual a transition you want to give, and being always the more cautious type, I think having such a step will give packagers who pay attention to what they package and users who pay attention to what they install without packaging an early chance to notice and prepare. Immaginary packagers. - The always warn does not force update at the point of use, but it still does not help them to notice well before they try to use it for the first time after update; I don't understand this sentence. They will see a big fat warning every time they run the tool, of course they'll notice. - Break the build attempts to help them notice when they try to update, not when they need to use the updated one right at this moment. This cannot be done. But I am fine with an expedited transition schedule without the break the build step. That was an optional first step, because warn but still work state we must have before the endgame will give the users the choice of when to adjust anyway. I also thought about adding an extra step to have even more gradual transition, by the way. A step before the endgame will ship these scripts without anything but instruct and fail (this is not warn and fail, as it is too late warn, as the scripts are crippled not to work at this point). That will still force the user to update at the point when the user needs to use it, but seeing the instruction (e.g. run this curl command to fetch from this URL and store it in a file called git-remote-xx on your $PATH) that is easy to follow immediately would be better than seeing only a failure (i.e. remote-hg not found), having to go fish the README, visiting the GitHub pages and figuring out how to fetch and install the script, which would be what the user will get with README only, no scripts endgame. I don't see what's so complicated about this: WARNING: git-remote-hg is now maintained independently. WARNING: For more information visit https://github.com/felipec/git-remote-hg They click that URL, and the are immediately greated with this: To enable this, simply add the git-remote-hg script anywhere in your $PATH: wget https://raw.github.com/felipec/git-remote-hg/master/git-remote-hg -O ~/bin/git-remote-hg chmod +x ~/bin/git-remote-hg Clearly you haven't even bothered to visit the home pages of the projects you threw to the wolves. So to summarize, the following timeline is a full possibility: 1. (optional) break the build by renaming directory and add README. Include not just the repository URL but a blob URL and instruction to download via wget/curl. That won't break the build. 2. add warning that is given every time the scripts are run and give the same instruction as in README. 3. (optional) cripple the script to make them always fail after showing the same warning as above. This is what I want, and I already sent the patches for; the scripts will be stubs. At this point you would have effectively removed the code, which what I want. 4. Keep README and retire everything else. After you've removed the code, I don't care what you do, but I'd say you should remove the stubs after a long period of time. -- Felipe Contreras -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
RE: bug: autostash is lost after aborted rebase
Karen Etheridge wrote: scenario: - edit some tracked files; do not add them to the index - git config rebase.autostash true - git rebase -i HEAD~3 (an autostash will be created) - delete the entire buffer and save/exit the editor - this will abort the rebase poof, the autostash is gone (it is not reapplied) -- it must be explicitly applied again via the SHA that was printed earlier. Yeah, I noticed this issue while rewriting `git rebase` in Ruby. I also noticed many other issues and inconsistencies that happen depending on what arguments you pass to `git rebase`. I'm in the process of writing a much more improved and consistent `git rebase` in Ruby with a different interface. Sadly it's still not ready. (please cc me; I am not subscribed to the list.) No need to say that, only on mailing lists that do Reply-To munging, and Git is not one of them. Cheers. -- Felipe Contreras -- To unsubscribe from this list: send the line unsubscribe git in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
