* Re: problem with BOINC repository and CR/LF
From: Torsten Bögershausen @ 2012-12-18 12:15 UTC (permalink / raw)
To: Toralf Förster; +Cc: Andrew Ardill, git@vger.kernel.org
In-Reply-To: <50D03D80.3090005@gmx.de>
On 18.12.12 10:55, Toralf Förster wrote:
> On 12/18/2012 02:56 AM, Andrew Ardill wrote:
>> On 18 December 2012 03:01, Toralf Förster <toralf.foerster@gmx.de> wrote:
>>> On 12/17/2012 12:38 PM, Andrew Ardill wrote:
>>>> On 17 December 2012 21:23, Toralf Förster <toralf.foerster@gmx.de> wrote:
>>>>> Hello,
>>>>>
>>>>> I'm faced with this situation :
>>>>> http://lists.ssl.berkeley.edu/mailman/private/boinc_alpha/2012-December/017371.html
>>>>> and even a "git stash" doesn't help.
>>>>
>>>> Hi Toralf,
>>>>
>>>> That list is private and not visible without an account. Can you
>>>> transcribe the relevant parts?
>>>>
>>>> Regards,
>>>>
>>>> Andrew Ardill
>>>>
>>> Oh of course :
>>>
>>>
>>> On 12/17/2012 12:03 AM, Gianfranco Costamagna wrote:
>>>> So if you have further issues with boinc feel free to look in our debian
>>>> git and feel free to download appropriate patches :-)
>>>>
>>>> Gianfranco
>>> thx
>>>
>>> Currently I'm struggling with a git problem of the boinc repository
>>> itself and b/c I'm using git for the linux kernel tree w/o any problems
>>> since eons /me wonders whether this is a BOINC-repository specific problem :
>>>
>>>
>>> After doing the following sequence with git 1.8.0.2 :
>>>
>>> $> git clone git://boinc.berkeley.edu/boinc.git
>>> $> cd boinc
>>> $> git checkout client_release_7.0.39
>>> $> git checkout master
>>> (sometimes I've to repeat this :
>>> $> git checkout client_release_7.0.39
>>> $> git checkout master
>>> )
>>> I'm faced with this situation :
>>>
>>> $ git status
>>> # On branch master
>>> # Changes not staged for commit:
>>> # (use "git add <file>..." to update what will be committed)
>>> # (use "git checkout -- <file>..." to discard changes in working
>>> directory)
>>> #
>>> # modified: clientgui/AsyncRPC.cpp
>>> # modified: clientgui/sg_BoincSimpleFrame.cpp
>>> #
>>> no changes added to commit (use "git add" and/or "git commit -a")
>>>
>>> (sometimes only clientgui/sg_BoincSimpleFrame.cpp is mentioned)
>>>
>>> Now these commands
>>>
>>> $ git checkout -- clientgui/AsyncRPC.cpp
>>> $ git checkout -- clientgui/sg_BoincSimpleFrame.cpp
>>>
>>> doesn't help - the status is still the same (and ofc now I'm no longer
>>> allowed to make a "git checkout" - due to un-commited changes).
>>>
>>> Now I'm wondering where to start to investigate this issue ...
>>
>> Hi Toralf,
>>
>> That does look like a weird issue. What operating system are you on?
>
> I'm running a stable Gentoo Linux x86, 32bit with gcc 4.6.3 and current
> stable kernel 3.6.1x and 3.7.1, file system is ext4 at an external USB
> 2.0 drive.
>
> FWIW from the boinc maintainer I know that all tags till 7.0.3X are
> imported from svn.
>
> I upgraded to git 1.8.0.2 from 1.7.8.6 - situation is the same. The
> emerge gave :
>
> failed test(s): t3600 t7508
>
> fixed 0
> success 8342
> failed 8
> broken 56
> total 8528
>
> Ok, now answering your other questions:
>
>
> $> git stash
> warning: CRLF will be replaced by LF in clientgui/AsyncRPC.cpp.
> The file will have its original line endings in your working directory.
> warning: CRLF will be replaced by LF in clientgui/sg_BoincSimpleFrame.cpp.
> The file will have its original line endings in your working directory.
> warning: CRLF will be replaced by LF in clientgui/AsyncRPC.cpp.
> The file will have its original line endings in your working directory.
> warning: CRLF will be replaced by LF in clientgui/sg_BoincSimpleFrame.cpp.
> The file will have its original line endings in your working directory.
> Saved working directory and index state WIP on master: 4a296dc - client
> simulator: fix build errors
> HEAD is now at 4a296dc - client simulator: fix build errors
>
> After that the situation is unchanged.
>
>> What happens if you do a hard reset to the branch?
>
> $> git reset --hard HEAD~1
>
> not better.
>
>
>> What is the ouptut of git diff --cached ?
>
> The output is empty but "git status" shows still modified files.
>
>
>
> FWIW there's a related issue I'm wondering about which might help:
>
> $> git clone git://boinc.berkeley.edu/boinc.git
> $> tar -cpf boinc.tar boinc/
> $> rm -rf boinc/
> $> tar -xpf boinc.tar
> $> cd boinc/
> $> git status
> # On branch master
> # Changes not staged for commit:
> # (use "git add <file>..." to update what will be committed)
> # (use "git checkout -- <file>..." to discard changes in working
> directory)
> #
> # modified: client/win/boinc_log.h
> # modified: client/win/boinc_log.rc
> # modified: clientctrl/boincsvcctrl.cpp
> # modified: clientctrl/boincsvcctrl.h
> # modified: clientctrl/boincsvcctrl.rc
> # modified: clientgui/AsyncRPC.cpp
> # modified: clientgui/DlgEventLog.cpp
> # modified: clientgui/DlgEventLog.h
> # modified: clientgui/DlgEventLogListCtrl.cpp
> # modified: clientgui/DlgEventLogListCtrl.h
> # modified: clientgui/DlgExitMessage.h
> # modified: clientgui/DlgItemProperties.h
> # modified: clientgui/TermsOfUsePage.cpp
> # modified: clientgui/TermsOfUsePage.h
> # modified: clientgui/ViewNotices.cpp
> # modified: clientgui/ViewNotices.h
> # modified: clientgui/sg_BoincSimpleFrame.cpp
> # modified: clientscr/boinc_ss_opengl.h
> # modified: clientscr/boinc_ss_opengl.rc
> # modified: clientscr/screensaver.cpp
> # modified: clienttray/boinc_tray.h
> # modified: clienttray/boinc_tray.rc
> # modified: clienttray/tray_win.cpp
> # modified: clienttray/tray_win.h
> # modified: coprocs/NVIDIA/include/nvapi.h
> #
> no changes added to commit (use "git add" and/or "git commit -a")
> $> git diff --cached
> $>
>
>
>
> Meaning, w/o any other interaction a tar'ed archive has modified files -
> and the diff is empty...
>
Hej,
I could re-produce the problem here:
git version 1.8.0.197.g5a90748
Mac OS X (that what I had at hands fastest)
After doing
git checkout 5db4a05b5c8f9c420fc418727cafbb58e6051f1e
(same as master ?)
We see that
clientgui/AsyncRPC.cpp
is full of CRLF and as it seems only CRLF, no single LF.
The file is classified as text:
$git check-attr text clientgui/AsyncRPC.cpp
$clientgui/AsyncRPC.cpp: text: set
(And we can see this in .gitattributes as well)
(And there are more files affected, but I will only look at one of them)
If we remove the text attribute like this:
$mv .gitattributes .gitattributes.sav
we see
$git status
deleted: .gitattributes
# modified: clientgui/sg_BoincSimpleFrame.cpp
If we dig into the file:
$git ls-files -s clientgui/AsyncRPC.cpp
100644 6832333ad133181986ada54fe0229b45a30c614a 0 clientgui/AsyncRPC.cpp
We see that it is recorded under 6832333ad1
And if we look into it:
$git show 6832333 | od -c
[snip]
we can see that the file has CRLF in the repo.
So my conclusion is:
The file has CRLF in the repo, but should have LF.
This is not a good thing, and the files need to be normalized.
A very good instruction how to do this, is found here:
http://kernel.org/pub/software/scm/git/docs/gitattributes.html
(You may want to search for "End-of-line conversion" or "core.autocrlf")
HTH
/Torsten
^ permalink raw reply
* Re: [BUG] Cannot push some grafted branches
From: Johannes Sixt @ 2012-12-18 12:03 UTC (permalink / raw)
To: Yann Dirson
Cc: Junio C Hamano, Andreas Schwab, Christian Couder, Thomas Rast,
git list
In-Reply-To: <20121218120058.0c558ba5@chalon.bertin.fr>
Am 12/18/2012 12:00, schrieb Yann Dirson:
> On Mon, 17 Dec 2012 13:14:56 -0800
> Junio C Hamano <gitster@pobox.com> wrote:
>
>> Andreas Schwab <schwab@linux-m68k.org> writes:
>>
>>> Christian Couder <christian.couder@gmail.com> writes:
>>>
>>>> Yeah, at one point I wanted to have a command that created to craft a
>>>> new commit based on an existing one.
>>>
>>> This isn't hard to do, you only have to resort to plumbing:
>>>
>>> $ git cat-file commit fef11965da875c105c40f1a9550af1f5e34a6e62 | sed s/bfae342c973b0be3c9e99d3d86ed2e6b152b4a6b/790c83cda92f95f1b4b91e2ddc056a52a99a055d/ | git hash-object -t commit --stdin -w
>>> bb45cc6356eac6c7fa432965090045306dab7026
>>
>> Good. I do not think an extra special-purpose command is welcome
>> here.
>
> Well, I'm not sure this is intuitive enough to be useful to the average user :)
When I played with git-replace in the past, I imagined that it could be
git replace <object> --commit ...commit options...
that would do the trick.
We could implement it with a git-replace--commit helper script that
generates the replacement commit using the ...commit options... (to be
defined what this should be), and git-replace would just pick its output
(the SHA1 of the generated commit) as a substitute for the <replacement>
argument that would have to be given without the --commit option.
-- Hannes
^ permalink raw reply
* Re: [BUG] Cannot push some grafted branches
From: Yann Dirson @ 2012-12-18 11:00 UTC (permalink / raw)
To: Junio C Hamano; +Cc: Andreas Schwab, Christian Couder, Thomas Rast, git list
In-Reply-To: <7vwqwgjs8f.fsf@alter.siamese.dyndns.org>
On Mon, 17 Dec 2012 13:14:56 -0800
Junio C Hamano <gitster@pobox.com> wrote:
> Andreas Schwab <schwab@linux-m68k.org> writes:
>
> > Christian Couder <christian.couder@gmail.com> writes:
> >
> >> Yeah, at one point I wanted to have a command that created to craft a
> >> new commit based on an existing one.
> >
> > This isn't hard to do, you only have to resort to plumbing:
> >
> > $ git cat-file commit fef11965da875c105c40f1a9550af1f5e34a6e62 | sed s/bfae342c973b0be3c9e99d3d86ed2e6b152b4a6b/790c83cda92f95f1b4b91e2ddc056a52a99a055d/ | git hash-object -t commit --stdin -w
> > bb45cc6356eac6c7fa432965090045306dab7026
>
> Good. I do not think an extra special-purpose command is welcome
> here.
Well, I'm not sure this is intuitive enough to be useful to the average user :)
Adding git-rev-parse calls for convenience, and calling git-replace, would make it
a more complete recipe, and we could suggest that as an alias in the collection that's
in the wiki (which is not even linked any more from git-scm.com btw), but imho that
would be hiding valuable information in a dark corner.
Anyway, in this form it will only replace a parent with another, whereas a full
graft replacement should allow to write a different number of new parents instead.
That is, instead of this simple sed, something like:
(NEWPARENTS='parent xxx\nparent yyy\nparent zzz\n; git cat-file commit master | perl -ne 'BEGIN { $state=0 }; if ($state eq 0) { if (/^parent/) { $state=1 } else { print } } elsif ($state eq 1) { if (/^author/) { print "'"$NEWPARENTS"'"; print; $state=2 } } else { print }')
Well, a short bash script should be more readable and possibly faster, but that's the
idea. Such a script could be a candidate for contrib ?
--
Yann Dirson - Bertin Technologies
^ permalink raw reply
* Re: problem with BOINC repository and CR/LF
From: Toralf Förster @ 2012-12-18 9:55 UTC (permalink / raw)
To: Andrew Ardill; +Cc: git@vger.kernel.org
In-Reply-To: <CAH5451=xiipSKrAb_DFXCW=+NAn+mnSm1zPzjhEVc8fZ2KGcnw@mail.gmail.com>
On 12/18/2012 02:56 AM, Andrew Ardill wrote:
> On 18 December 2012 03:01, Toralf Förster <toralf.foerster@gmx.de> wrote:
>> On 12/17/2012 12:38 PM, Andrew Ardill wrote:
>>> On 17 December 2012 21:23, Toralf Förster <toralf.foerster@gmx.de> wrote:
>>>> Hello,
>>>>
>>>> I'm faced with this situation :
>>>> http://lists.ssl.berkeley.edu/mailman/private/boinc_alpha/2012-December/017371.html
>>>> and even a "git stash" doesn't help.
>>>
>>> Hi Toralf,
>>>
>>> That list is private and not visible without an account. Can you
>>> transcribe the relevant parts?
>>>
>>> Regards,
>>>
>>> Andrew Ardill
>>>
>> Oh of course :
>>
>>
>> On 12/17/2012 12:03 AM, Gianfranco Costamagna wrote:
>>> So if you have further issues with boinc feel free to look in our debian
>>> git and feel free to download appropriate patches :-)
>>>
>>> Gianfranco
>> thx
>>
>> Currently I'm struggling with a git problem of the boinc repository
>> itself and b/c I'm using git for the linux kernel tree w/o any problems
>> since eons /me wonders whether this is a BOINC-repository specific problem :
>>
>>
>> After doing the following sequence with git 1.8.0.2 :
>>
>> $> git clone git://boinc.berkeley.edu/boinc.git
>> $> cd boinc
>> $> git checkout client_release_7.0.39
>> $> git checkout master
>> (sometimes I've to repeat this :
>> $> git checkout client_release_7.0.39
>> $> git checkout master
>> )
>> I'm faced with this situation :
>>
>> $ git status
>> # On branch master
>> # Changes not staged for commit:
>> # (use "git add <file>..." to update what will be committed)
>> # (use "git checkout -- <file>..." to discard changes in working
>> directory)
>> #
>> # modified: clientgui/AsyncRPC.cpp
>> # modified: clientgui/sg_BoincSimpleFrame.cpp
>> #
>> no changes added to commit (use "git add" and/or "git commit -a")
>>
>> (sometimes only clientgui/sg_BoincSimpleFrame.cpp is mentioned)
>>
>> Now these commands
>>
>> $ git checkout -- clientgui/AsyncRPC.cpp
>> $ git checkout -- clientgui/sg_BoincSimpleFrame.cpp
>>
>> doesn't help - the status is still the same (and ofc now I'm no longer
>> allowed to make a "git checkout" - due to un-commited changes).
>>
>> Now I'm wondering where to start to investigate this issue ...
>
> Hi Toralf,
>
> That does look like a weird issue. What operating system are you on?
I'm running a stable Gentoo Linux x86, 32bit with gcc 4.6.3 and current
stable kernel 3.6.1x and 3.7.1, file system is ext4 at an external USB
2.0 drive.
FWIW from the boinc maintainer I know that all tags till 7.0.3X are
imported from svn.
I upgraded to git 1.8.0.2 from 1.7.8.6 - situation is the same. The
emerge gave :
failed test(s): t3600 t7508
fixed 0
success 8342
failed 8
broken 56
total 8528
Ok, now answering your other questions:
$> git stash
warning: CRLF will be replaced by LF in clientgui/AsyncRPC.cpp.
The file will have its original line endings in your working directory.
warning: CRLF will be replaced by LF in clientgui/sg_BoincSimpleFrame.cpp.
The file will have its original line endings in your working directory.
warning: CRLF will be replaced by LF in clientgui/AsyncRPC.cpp.
The file will have its original line endings in your working directory.
warning: CRLF will be replaced by LF in clientgui/sg_BoincSimpleFrame.cpp.
The file will have its original line endings in your working directory.
Saved working directory and index state WIP on master: 4a296dc - client
simulator: fix build errors
HEAD is now at 4a296dc - client simulator: fix build errors
After that the situation is unchanged.
> What happens if you do a hard reset to the branch?
$> git reset --hard HEAD~1
not better.
> What is the ouptut of git diff --cached ?
The output is empty but "git status" shows still modified files.
FWIW there's a related issue I'm wondering about which might help:
$> git clone git://boinc.berkeley.edu/boinc.git
$> tar -cpf boinc.tar boinc/
$> rm -rf boinc/
$> tar -xpf boinc.tar
$> cd boinc/
$> git status
# On branch master
# Changes not staged for commit:
# (use "git add <file>..." to update what will be committed)
# (use "git checkout -- <file>..." to discard changes in working
directory)
#
# modified: client/win/boinc_log.h
# modified: client/win/boinc_log.rc
# modified: clientctrl/boincsvcctrl.cpp
# modified: clientctrl/boincsvcctrl.h
# modified: clientctrl/boincsvcctrl.rc
# modified: clientgui/AsyncRPC.cpp
# modified: clientgui/DlgEventLog.cpp
# modified: clientgui/DlgEventLog.h
# modified: clientgui/DlgEventLogListCtrl.cpp
# modified: clientgui/DlgEventLogListCtrl.h
# modified: clientgui/DlgExitMessage.h
# modified: clientgui/DlgItemProperties.h
# modified: clientgui/TermsOfUsePage.cpp
# modified: clientgui/TermsOfUsePage.h
# modified: clientgui/ViewNotices.cpp
# modified: clientgui/ViewNotices.h
# modified: clientgui/sg_BoincSimpleFrame.cpp
# modified: clientscr/boinc_ss_opengl.h
# modified: clientscr/boinc_ss_opengl.rc
# modified: clientscr/screensaver.cpp
# modified: clienttray/boinc_tray.h
# modified: clienttray/boinc_tray.rc
# modified: clienttray/tray_win.cpp
# modified: clienttray/tray_win.h
# modified: coprocs/NVIDIA/include/nvapi.h
#
no changes added to commit (use "git add" and/or "git commit -a")
$> git diff --cached
$>
Meaning, w/o any other interaction a tar'ed archive has modified files -
and the diff is empty...
--
MfG/Sincerely
Toralf Förster
pgp finger print: 7B1A 07F4 EC82 0F90 D4C2 8936 872A E508 7DB6 9DA3
^ permalink raw reply
* Re: problem with BOINC repository and CR/LF
From: Toralf Förster @ 2012-12-18 9:58 UTC (permalink / raw)
To: Andrew Ardill; +Cc: git@vger.kernel.org
In-Reply-To: <50D03D80.3090005@gmx.de>
On 12/18/2012 10:55 AM, Toralf Förster wrote:
> failed test(s): t3600 t7508
>
> fixed 0
> success 8342
> failed 8
> broken 56
> total 8528
>
ick forgot these :
n22 /usr/portage/dev-vcs/git # grep -i "^not ok" /tmp/git.log | grep -v TODO
not ok - 15 Test that "git rm -f" fails if its rm fails
not ok - 16 When the rm in "git rm -f" fails, it should not remove the file from the index
not ok - 20 Re-add foo and baz
not ok - 21 Modify foo -- rm should refuse
not ok - 22 Modified foo -- rm -f should work
not ok - 23 Re-add foo and baz for HEAD tests
not ok - 24 foo is different in index from HEAD -- rm should refuse
not ok - 55 status succeeds in a read-only repository
--
MfG/Sincerely
Toralf Förster
pgp finger print: 7B1A 07F4 EC82 0F90 D4C2 8936 872A E508 7DB6 9DA3
^ permalink raw reply
* Re: What's cooking in git.git (Dec 2012, #04; Sun, 16)
From: Johannes Sixt @ 2012-12-18 8:38 UTC (permalink / raw)
To: Junio C Hamano; +Cc: git
In-Reply-To: <20121217213730.GA17212@ftbfs.org>
> Junio C Hamano wrote:
>> It could turn out that we may be able to get rid of sys/param.h
>> altogether, but one step at a time. Inputs from people on minority
>> platforms are very much appreciated---does your platform build fine
>> when the inclusion of the file is removed from git-compat-util.h?
MinGW works fine with sys/param.h removed from git-compat-util.h.
-- Hannes
^ permalink raw reply
* Re: What's cooking in git.git (Dec 2012, #04; Sun, 16)
From: Joachim Schmitz @ 2012-12-18 8:28 UTC (permalink / raw)
To: git
In-Reply-To: <20121217213730.GA17212@ftbfs.org>
Matt Kraai wrote:
> Junio C Hamano wrote:
>> It could turn out that we may be able to get rid of sys/param.h
>> altogether, but one step at a time. Inputs from people on minority
>> platforms are very much appreciated---does your platform build fine
>> when the inclusion of the file is removed from git-compat-util.h?
>
> QNX builds fine when sys/param.h is not included.
HP-NonStop build fine too without it.
Bye, Jojo
^ permalink raw reply
* Re: Proposal: create meaningful aliases for git reset's hard/soft/mixed
From: Martin von Zweigbergk @ 2012-12-18 6:34 UTC (permalink / raw)
To: Junio C Hamano; +Cc: Philippe Vaucher, git, Christian Couder
In-Reply-To: <7vlir6brjw.fsf@alter.siamese.dyndns.org>
On Wed, Nov 23, 2011 at 10:51 AM, Junio C Hamano <gitster@pobox.com> wrote:
>
> I am guilty of introducing "git reset --soft HEAD^" before I invented
> "commit --amend" during v1.3.0 timeframe to solve the issue "soft" reset
> originally wanted to.
I do use "commit --amend" a lot, but I still appreciate having "reset
--soft". For example, to squash the last few commits:
git reset --soft HEAD^^^ && git commit --amend
or undo "commit --amend":
git reset --soft HEAD@{1} && git commit --amend
Maybe there's a better way of doing that?
^ permalink raw reply
* Re: Proposal: create meaningful aliases for git reset's hard/soft/mixed
From: Martin von Zweigbergk @ 2012-12-18 6:24 UTC (permalink / raw)
To: Matthieu Moy; +Cc: Philippe Vaucher, git
In-Reply-To: <vpq4nxvusty.fsf@bauges.imag.fr>
On Wed, Nov 23, 2011 at 12:49 AM, Matthieu Moy
<Matthieu.Moy@grenoble-inp.fr> wrote:
> Philippe Vaucher <philippe.vaucher@gmail.com> writes:
>
>> Optional: a new mode would be introduced for consistency:
>> --worktree (or maybe --tree): only updates the worktree but not the index
>
> That would be an alias for "git checkout <rev> -- path", right?
Not quite, in two ways, I think. First, it _would_ update the index,
wouldn't it? Second, "git checkout <rev> -- path" doesn't delete files
that are deleted in <rev> as compared to head.
I'm considering implementing support for an operation that would do
what I expected "git checkout <rev> -- <path>" and "git reset --hard
<rev> -- <path>" to do. I'm currently planning for it to be exactly
"git reset --hard <rev> -- <path>" (which is currently simply not
allowed), but perhaps it would be more natural as an option to
checkout (--also-deleted or something)?
^ permalink raw reply
* Re: bug? 'git log -M100' is different from 'git log -M100%'
From: Sitaram Chamarty @ 2012-12-18 4:17 UTC (permalink / raw)
To: Junio C Hamano; +Cc: git
In-Reply-To: <7vfw34jgmv.fsf@alter.siamese.dyndns.org>
On Tue, Dec 18, 2012 at 6:55 AM, Junio C Hamano <gitster@pobox.com> wrote:
> Sitaram Chamarty <sitaramc@gmail.com> writes:
>
>> When using -M with a number to act as a threshold for declaring
>> a change as being a rename, I found a... quirk. Any 2-digit
>> number after the M will work,...
>
> That is not 2-digit number.
>
> A few historical trivia may help.
>
> Originally we said "you can use -M2 to choose 2/10" (like "gzip"
> taking compression levels between "-0" to "-9"). Then Linus came up
> with a clever idea to let people specify arbitrary precision by
> letting you say "-M25" to mean 25/100 and "-M254" to mean 254/1000.
>
> Read the numbers without per-cent as if it has decimal point before
> it (i.e. -M005 is talking about 0.005 which is 0.5%). Full hundred
> per-cent has to be spelled with per-cent sign for obvious reasons
> with this scheme but that cannot be avoided. It is a special case.
Oh nice. Makes sense; thanks!
I submitted a patch to diff-options.txt (separately).
regards
sitaram
^ permalink raw reply
* [PATCH] clarify -M without % symbol in diff-options
From: Sitaram Chamarty @ 2012-12-18 4:15 UTC (permalink / raw)
To: Junio C Hamano, git
---
Documentation/diff-options.txt | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/Documentation/diff-options.txt b/Documentation/diff-options.txt
index f4f7e25..39f2c50 100644
--- a/Documentation/diff-options.txt
+++ b/Documentation/diff-options.txt
@@ -309,7 +309,11 @@ endif::git-log[]
index (i.e. amount of addition/deletions compared to the
file's size). For example, `-M90%` means git should consider a
delete/add pair to be a rename if more than 90% of the file
- hasn't changed.
+ hasn't changed. Without a `%` sign, the number is to be read as
+ a fraction, with a decimal point before it. I.e., `-M5` becomes
+ 0.5, and is thus the same as `-M50%`. Similarly, `-M05` is
+ the same as `-M5%`. To limit detection to exact renames, use
+ `-M100%`.
-C[<n>]::
--find-copies[=<n>]::
--
1.7.11.7
^ permalink raw reply related
* Re: [PATCH 1/2] Documentation/git-checkout.txt: clarify usage
From: Chris Rorvick @ 2012-12-18 3:33 UTC (permalink / raw)
To: Junio C Hamano
Cc: Andrew Ardill, Philip Oakley, Git List, Tomas Carnecky, Woody Wu,
Johannes Sixt
In-Reply-To: <7vvcc0i0rz.fsf@alter.siamese.dyndns.org>
On Mon, Dec 17, 2012 at 7:53 PM, Junio C Hamano <gitster@pobox.com> wrote:
> Here is a work-in-progress relative to Chris's 83c9989
> (Documentation/git-checkout.txt: document 70c9ac2 behavior,
> 2012-12-17).
It sounds pretty good to me.
> @@ -54,12 +61,17 @@ $ git checkout <branch>
> that is to say, the branch is not reset/created unless "git checkout" is
> successful.
>
> -'git checkout' [--detach] [<commit>]::
> +'git checkout' --detach [<commit>]::
> +'git checkout' <commit>::
>
> - Update the index and working tree to reflect the specified
> - commit and set HEAD to point directly to <commit> (see
> - "DETACHED HEAD" section.) Passing `--detach` forces this
> - behavior even if <commit> is a branch.
> + Prepare to work on top of <commit>, by detaching HEAD at it
> + (see "DETACHED HEAD" section), and updating the index and the
> + files in the working tree. Local modifications to the files
> + in the working tree are kept, so that the resulting working
> + tree will be the state recorded in the commit plus the local
> + modifications.
> ++
> +Passing `--detach` forces this behavior even if <commit> is a branch.
>
> 'git checkout' [-p|--patch] [<tree-ish>] [--] <pathspec>...::
>
I like Johannes' suggestion of using "<branch>" in the --detach case
instead of "<commit>" as I think it makes the reason for the
separation more obvious at a glance. On top of your changes, maybe
something like:
--->8---
diff --git a/Documentation/git-checkout.txt b/Documentation/git-checkout.txt
index dcf1a32..4fdf41a 100644
--- a/Documentation/git-checkout.txt
+++ b/Documentation/git-checkout.txt
@@ -61,8 +61,8 @@ $ git checkout <branch>
that is to say, the branch is not reset/created unless "git checkout" is
successful.
-'git checkout' --detach [<commit>]::
'git checkout' <commit>::
+'git checkout' --detach [<branch>]::
Prepare to work on top of <commit>, by detaching HEAD at it
(see "DETACHED HEAD" section), and updating the index and the
@@ -71,7 +71,8 @@ successful.
tree will be the state recorded in the commit plus the local
modifications.
+
-Passing `--detach` forces this behavior even if <commit> is a branch.
+Passing `--detach` forces this behavior in the case of a <branch>, or
+the current branch if one is not specified.
'git checkout' [-p|--patch] [<tree-ish>] [--] <pathspec>...::
^ permalink raw reply related
* Re: [PATCH 1/2] Documentation/git-checkout.txt: clarify usage
From: Chris Rorvick @ 2012-12-18 2:55 UTC (permalink / raw)
To: Johannes Sixt
Cc: Junio C Hamano, git, Andrew Ardill, Tomas Carnecky, Woody Wu
In-Reply-To: <50CED5D4.5040705@viscovery.net>
On Mon, Dec 17, 2012 at 2:20 AM, Johannes Sixt <j.sixt@viscovery.net> wrote:
>>> +'git checkout' [--detach] [<commit>]::
>
> The title here is better spelled as two lines:
>
> 'git checkout' <commit>::
> 'git checkout' --detach <branch>::
AsciiDoc renders these horizontally separated by a comma when
formatted as a man page instead of vertically as written (and as
rendered by the HTML documentation.) I think this makes this
separation into two less effective, but at least it doesn't line wrap
on an 80-wide terminal like the previous title did.
Chris
^ permalink raw reply
* Re: [PATCH 1/2] Documentation/git-checkout.txt: clarify usage
From: Andrew Ardill @ 2012-12-18 2:12 UTC (permalink / raw)
To: Junio C Hamano
Cc: Philip Oakley, Chris Rorvick, Git List, Tomas Carnecky, Woody Wu
In-Reply-To: <7vvcc0i0rz.fsf@alter.siamese.dyndns.org>
I like these, and I think they are conveying the right amount of
information. There is a slight discrepancy between the <branch> and
<commit> versions, where it seems we are assuming that by checking out
a commit you are intending to work 'on top of' it. This could be
avoided by using the term 'with' in both cases. Also, they are in
different tenses, but I'm not sure which is preferred ('Prepare to' vs
'To prepare for').
In the second tense, these opening lines might look like this:
+ To prepare for working with <branch>, switch to it by updating
+ To prepare for working with <commit>, detach HEAD at it
+ (see "DETACHED HEAD" section), and update the index and the
+ files in the working tree.
Regards,
Andrew Ardill
^ permalink raw reply
* Re: problem with BOINC repository and CR/LF
From: Andrew Ardill @ 2012-12-18 1:56 UTC (permalink / raw)
To: Toralf Förster; +Cc: git@vger.kernel.org
In-Reply-To: <50CF41EB.1060402@gmx.de>
On 18 December 2012 03:01, Toralf Förster <toralf.foerster@gmx.de> wrote:
> On 12/17/2012 12:38 PM, Andrew Ardill wrote:
>> On 17 December 2012 21:23, Toralf Förster <toralf.foerster@gmx.de> wrote:
>>> Hello,
>>>
>>> I'm faced with this situation :
>>> http://lists.ssl.berkeley.edu/mailman/private/boinc_alpha/2012-December/017371.html
>>> and even a "git stash" doesn't help.
>>
>> Hi Toralf,
>>
>> That list is private and not visible without an account. Can you
>> transcribe the relevant parts?
>>
>> Regards,
>>
>> Andrew Ardill
>>
> Oh of course :
>
>
> On 12/17/2012 12:03 AM, Gianfranco Costamagna wrote:
>> So if you have further issues with boinc feel free to look in our debian
>> git and feel free to download appropriate patches :-)
>>
>> Gianfranco
> thx
>
> Currently I'm struggling with a git problem of the boinc repository
> itself and b/c I'm using git for the linux kernel tree w/o any problems
> since eons /me wonders whether this is a BOINC-repository specific problem :
>
>
> After doing the following sequence with git 1.8.0.2 :
>
> $> git clone git://boinc.berkeley.edu/boinc.git
> $> cd boinc
> $> git checkout client_release_7.0.39
> $> git checkout master
> (sometimes I've to repeat this :
> $> git checkout client_release_7.0.39
> $> git checkout master
> )
> I'm faced with this situation :
>
> $ git status
> # On branch master
> # Changes not staged for commit:
> # (use "git add <file>..." to update what will be committed)
> # (use "git checkout -- <file>..." to discard changes in working
> directory)
> #
> # modified: clientgui/AsyncRPC.cpp
> # modified: clientgui/sg_BoincSimpleFrame.cpp
> #
> no changes added to commit (use "git add" and/or "git commit -a")
>
> (sometimes only clientgui/sg_BoincSimpleFrame.cpp is mentioned)
>
> Now these commands
>
> $ git checkout -- clientgui/AsyncRPC.cpp
> $ git checkout -- clientgui/sg_BoincSimpleFrame.cpp
>
> doesn't help - the status is still the same (and ofc now I'm no longer
> allowed to make a "git checkout" - due to un-commited changes).
>
> Now I'm wondering where to start to investigate this issue ...
Hi Toralf,
That does look like a weird issue. What operating system are you on?
What happens if you do a hard reset to the branch?
What is the ouptut of git diff --cached ?
Regards,
Andrew Ardill
^ permalink raw reply
* Re: [PATCH 1/2] Documentation/git-checkout.txt: clarify usage
From: Junio C Hamano @ 2012-12-18 1:53 UTC (permalink / raw)
To: Andrew Ardill
Cc: Philip Oakley, Chris Rorvick, Git List, Tomas Carnecky, Woody Wu
In-Reply-To: <7vobhsjq6a.fsf@alter.siamese.dyndns.org>
Junio C Hamano <gitster@pobox.com> writes:
> I agree with you that sightseeing use case where you do not intend
> to make any commit is also important. That is exactly why I said
> "further work is done on that branch" not "to that branch" in the
> message you are responding to.
Here is a work-in-progress relative to Chris's 83c9989
(Documentation/git-checkout.txt: document 70c9ac2 behavior,
2012-12-17).
Even though "switch to that specific branch" may be easy to grasp as
a concept, I do not think "switch to detached HEAD" makes much
sense, so I ended up with "switch" for the <branch> case, and
"detach" for the "--detach" one, at least for now.
diff --git c/Documentation/git-checkout.txt w/Documentation/git-checkout.txt
index db89cf7..dcf1a32 100644
--- c/Documentation/git-checkout.txt
+++ w/Documentation/git-checkout.txt
@@ -21,10 +21,12 @@ or the specified tree. If no paths are given, 'git checkout' will
also update `HEAD` to set the specified branch as the current
branch.
-'git checkout' [<branch>]::
-
- Update the index, working tree, and HEAD to reflect the
- specified branch.
+'git checkout' <branch>::
+ To prepare for working on <branch>, switch to it by updating
+ the index and the files in the working tree, and by pointing
+ HEAD at the branch. Local modifications to the files in the
+ working tree are kept, so that they can be committed to the
+ <branch>.
+
If <branch> is not found but there does exist a tracking branch in
exactly one remote (call it <remote>) with a matching name, treat as
@@ -33,6 +35,11 @@ equivalent to
------------
$ git checkout -b <branch> --track <remote>/<branch>
------------
++
+You could omit <branch>, in which case the command degenerates to
+"check out the current branch", which is a glorified no-op with a
+rather expensive side-effects to show only the tracking information,
+if exists, for the current branch.
'git checkout' -b|-B <new_branch> [<start point>]::
@@ -54,12 +61,17 @@ $ git checkout <branch>
that is to say, the branch is not reset/created unless "git checkout" is
successful.
-'git checkout' [--detach] [<commit>]::
+'git checkout' --detach [<commit>]::
+'git checkout' <commit>::
- Update the index and working tree to reflect the specified
- commit and set HEAD to point directly to <commit> (see
- "DETACHED HEAD" section.) Passing `--detach` forces this
- behavior even if <commit> is a branch.
+ Prepare to work on top of <commit>, by detaching HEAD at it
+ (see "DETACHED HEAD" section), and updating the index and the
+ files in the working tree. Local modifications to the files
+ in the working tree are kept, so that the resulting working
+ tree will be the state recorded in the commit plus the local
+ modifications.
++
+Passing `--detach` forces this behavior even if <commit> is a branch.
'git checkout' [-p|--patch] [<tree-ish>] [--] <pathspec>...::
^ permalink raw reply related
* Re: [PATCH 1/1] tests: Allow customization of how say_color() prints
From: Junio C Hamano @ 2012-12-18 1:42 UTC (permalink / raw)
To: Ramsay Jones; +Cc: GIT Mailing-list
In-Reply-To: <50CF9D4C.9080706@ramsay1.demon.co.uk>
Ramsay Jones <ramsay@ramsay1.demon.co.uk> writes:
> Junio C Hamano wrote:
> ...
>> Why does your printf die in the first place???
>
> I really don't know. ...
>
> Sorry for wasting your time.
Not a waste. I was hoping somebody (not necessarily you) may be able
to come up with a cleaner solution. Unfortunately it hasn't
happened (yet), but discussing issues on the list is often not a
waste.
We could introduce git_test_print and git_test_println shell
functions that default to the current "printf", and let the users
override these by including a custom scriptllet from t/test-lib.sh,
or something.
^ permalink raw reply
* Re: [PATCH] git-completion.bash: update obsolete code.
From: Junio C Hamano @ 2012-12-18 1:37 UTC (permalink / raw)
To: Manlio Perillo; +Cc: git
In-Reply-To: <50CF4E05.7050103@gmail.com>
Manlio Perillo <manlio.perillo@gmail.com> writes:
> Il 17/12/2012 05:54, Junio C Hamano ha scritto:
>> Manlio Perillo <manlio.perillo@gmail.com> writes:
>>
>>> The git-completion.bash script was using the git ls-tree command
>>> without the --name-only option, with a sed filter to parse path names;
>>> use the --name-only option, instead.
>>>
>>> Signed-off-by: Manlio Perillo <manlio.perillo@gmail.com>
>>> ---
>>
>> Did you miss the different handling between blobs and trees the
>> latter gets trailing slash in the completion)?
>>
>
> Yes, I totally missed it, sorry.
> I was assuming the bash completion script was written before --name-only
> option was added to ls-tree.
There is no need to say "sorry" here; catching this kind of mistake
is what we have the patch review process for, after all. If
anything, "thanks" is more appropriate ;-)
And the way you stated the reasoning behind this change in the
proposed commit log message was really good. It showed clearly that
the patch was an attempt to clean up the code, not was an attempt to
change the behaviour to drop the trailing slash or space. If it
said "let's use ls-tree --name-only" without such an explanation on
"why", future readers of "git log" would have been left scratching
their heads, wondering "why" the change was made.
Thanks.
^ permalink raw reply
* Re: [PATCH 1/2] Documentation/git-checkout.txt: clarify usage
From: Andrew Ardill @ 2012-12-18 1:29 UTC (permalink / raw)
To: Junio C Hamano
Cc: Philip Oakley, Chris Rorvick, Git List, Tomas Carnecky, Woody Wu
In-Reply-To: <7vobhsjq6a.fsf@alter.siamese.dyndns.org>
On 18 December 2012 08:59, Junio C Hamano <gitster@pobox.com> wrote:
> Andrew Ardill <andrew.ardill@gmail.com> writes:
>> Even if the primary purpose of "git checkout <branch>" is to "check
>> out the branch so that further work is done on that branch", I don't
>> believe that means it has to be stated first. In fact, I would say
>> that there are enough other use cases that the language should be
>> slightly more use-case agnostic in the first situation. For example,
>> someone might switch to another branch or commit simply to see what
>> state the tree was in at that point.
>
> I've been deliberately avoiding the term "switch", actually. I
> agree that it may be familiar to people with prior exposure to
> subversion, but that is not the primary audience of the manual.
I don't have much experience with svn, so I didn't make that
connection. Independent of svn usage, what is wrong with the term
'switch'?
I would be interested to hear how translators communicate the checkout
concept, as I assume the word checkout doesn't exist in many
languages. For me, switching between revisions is a natural way of
phrasing the action, but perhaps there is a better way of saying the
same thing?
>> Some people use checkout to
>> deploy a tag of the working tree onto a production server. The first
>> example in particular is, I think, a common enough operation that
>> restricting the opening lines of documentation to talking about
>> building further work is misleading.
>
> I agree with you that sightseeing use case where you do not intend
> to make any commit is also important. That is exactly why I said
> "further work is done on that branch" not "to that branch" in the
> message you are responding to.
Ah ok, I didn't pick up on that nuance. Your suggestion from earlier
has, for example, "Prepare to work on building new history on
<branch>" which *is* excluding that use case. Perhaps modifying
similar lines to something like "Prepare to work with the
repository/history/something from <branch>" or maybe just "Prepare to
work with <branch>" would better encapsulate those use cases.
Following lines would expand on what it means to work with a branch or
commit, and the technical details of updates to the repositories
current state.
Regards,
Andrew Ardill
^ permalink raw reply
* Re: bug? 'git log -M100' is different from 'git log -M100%'
From: Junio C Hamano @ 2012-12-18 1:25 UTC (permalink / raw)
To: Sitaram Chamarty; +Cc: git
In-Reply-To: <20121217233446.GA5987@sita-lt.atc.tcs.com>
Sitaram Chamarty <sitaramc@gmail.com> writes:
> When using -M with a number to act as a threshold for declaring
> a change as being a rename, I found a... quirk. Any 2-digit
> number after the M will work,...
That is not 2-digit number.
A few historical trivia may help.
Originally we said "you can use -M2 to choose 2/10" (like "gzip"
taking compression levels between "-0" to "-9"). Then Linus came up
with a clever idea to let people specify arbitrary precision by
letting you say "-M25" to mean 25/100 and "-M254" to mean 254/1000.
Read the numbers without per-cent as if it has decimal point before
it (i.e. -M005 is talking about 0.005 which is 0.5%). Full hundred
per-cent has to be spelled with per-cent sign for obvious reasons
with this scheme but that cannot be avoided. It is a special case.
^ permalink raw reply
* Re: [PATCH 2/3] perl/Git.pm: Honor SSH_ASKPASS as fallback if GIT_ASKPASS is not set
From: Jeff King @ 2012-12-18 0:57 UTC (permalink / raw)
To: Sven Strickroth; +Cc: Junio C Hamano, git, Jakub Narebski, Eric Wong
In-Reply-To: <50CFB8BF.4000405@tu-clausthal.de>
On Tue, Dec 18, 2012 at 01:28:47AM +0100, Sven Strickroth wrote:
> If GIT_ASKPASS environment variable is not set, git-svn does not try to use
> SSH_ASKPASS as git-core does. This change adds a fallback to SSH_ASKPASS.
>
> Signed-off-by: Sven Strickroth <email@cs-ware.de>
> ---
Thanks, this series looks fine to me.
I skimmed through the original thread. It looks like we got bogged down
in discussion of whether GIT_ASKPASS should fall back on error, and
whether SSH_ASKPASS should come _after_ the terminal has been tried and
failed. This version of the series brings the behavior of git-svn in
line with the rest of git, which I think is a good first step.
I don't know whether we want to take a second step and tweak the order
in both perl and C code. Since nobody has mentioned it in the interim
months, I'm assuming it's not a big deal, and people are happy enough
with the current ordering. So unless somebody feels strongly about it,
it's not worth bothering.
-Peff
^ permalink raw reply
* bug? 'git log -M100' is different from 'git log -M100%'
From: Sitaram Chamarty @ 2012-12-17 23:34 UTC (permalink / raw)
To: git
Hi,
When using -M with a number to act as a threshold for declaring
a change as being a rename, I found a... quirk. Any 2-digit
number after the M will work, but if the number is 100, it will
require a % to be appended to be effective.
Here's a transcript that will demonstrate the problem when run
in an empty directory.
# setup
git init
seq 1 100 > f1
git add f1
git commit -m f1
git rm f1
( seq 1 45; seq 1001 1010; seq 56 100 ) > f2
git add f2
git commit -m f2
# here's the buglet
git log -1 --stat --raw -M
# this tells you the files are 83% similar
git log -1 --stat --raw -M82
# this shows it like a rename
git log -1 --stat --raw -M83
# this also
git log -1 --stat --raw -M84
# this shows two separate files
git log -1 --stat --raw -M99
# this also
# so far so good...
git log -1 --stat --raw -M100
# but this shows it like a rename
git log -1 --stat --raw -M100%
# adding a percent sign fixes it, now they're two separate
# files. It seems to be required only when you ask for 100%
--
Sitaram Chamarty
^ permalink raw reply
* [PATCH 3/3] git-svn, perl/Git.pm: extend and use Git->prompt method for querying users
From: Sven Strickroth @ 2012-12-18 0:28 UTC (permalink / raw)
To: Junio C Hamano; +Cc: git, Jeff King, Jakub Narebski, Eric Wong
In-Reply-To: <7vehiol9w2.fsf@alter.siamese.dyndns.org>
git-svn reads usernames and other user queries from an interactive
terminal. This cause GUIs (w/o STDIN connected) to hang waiting forever
for git-svn to complete (http://code.google.com/p/tortoisegit/issues/detail?id=967).
This change extends the Git::prompt helper, so that it can also be used
for non password queries, and makes use of it instead of using
hand-rolled prompt-response code that only works with the interactive
terminal.
Signed-off-by: Sven Strickroth <email@cs-ware.de>
---
perl/Git.pm | 28 +++++++++++++++++-----------
perl/Git/SVN/Prompt.pm | 16 +++++++---------
2 files changed, 24 insertions(+), 20 deletions(-)
diff --git a/perl/Git.pm b/perl/Git.pm
index 8dfca65..931047c 100644
--- a/perl/Git.pm
+++ b/perl/Git.pm
@@ -511,18 +511,19 @@ C<git --html-path>). Useful mostly only internally.
sub html_path { command_oneline('--html-path') }
-=item prompt ( PROMPT )
+=item prompt ( PROMPT , ISPASSWORD )
Query user C<PROMPT> and return answer from user.
Honours GIT_ASKPASS and SSH_ASKPASS environment variables for querying
the user. If no *_ASKPASS variable is set or an error occoured,
the terminal is tried as a fallback.
+If C<ISPASSWORD> is set and true, the terminal disables echo.
=cut
sub prompt {
- my ($prompt) = @_;
+ my ($prompt, $isPassword) = @_;
my $ret;
if (exists $ENV{'GIT_ASKPASS'}) {
$ret = _prompt($ENV{'GIT_ASKPASS'}, $prompt);
@@ -533,16 +534,20 @@ sub prompt {
if (!defined $ret) {
print STDERR $prompt;
STDERR->flush;
- require Term::ReadKey;
- Term::ReadKey::ReadMode('noecho');
- $ret = '';
- while (defined(my $key = Term::ReadKey::ReadKey(0))) {
- last if $key =~ /[\012\015]/; # \n\r
- $ret .= $key;
+ if (defined $isPassword && $isPassword) {
+ require Term::ReadKey;
+ Term::ReadKey::ReadMode('noecho');
+ $ret = '';
+ while (defined(my $key = Term::ReadKey::ReadKey(0))) {
+ last if $key =~ /[\012\015]/; # \n\r
+ $ret .= $key;
+ }
+ Term::ReadKey::ReadMode('restore');
+ print STDERR "\n";
+ STDERR->flush;
+ } else {
+ chomp($ret = <STDIN>);
}
- Term::ReadKey::ReadMode('restore');
- print STDERR "\n";
- STDERR->flush;
}
return $ret;
}
@@ -550,6 +555,7 @@ sub prompt {
sub _prompt {
my ($askpass, $prompt) = @_;
return unless length $askpass;
+ $prompt =~ s/\n/ /g;
my $ret;
open my $fh, "-|", $askpass, $prompt or return;
$ret = <$fh>;
diff --git a/perl/Git/SVN/Prompt.pm b/perl/Git/SVN/Prompt.pm
index a2cbcc8..74daa7a 100644
--- a/perl/Git/SVN/Prompt.pm
+++ b/perl/Git/SVN/Prompt.pm
@@ -62,16 +62,16 @@ sub ssl_server_trust {
issuer_dname fingerprint);
my $choice;
prompt:
- print STDERR $may_save ?
+ my $options = $may_save ?
"(R)eject, accept (t)emporarily or accept (p)ermanently? " :
"(R)eject or accept (t)emporarily? ";
STDERR->flush;
- $choice = lc(substr(<STDIN> || 'R', 0, 1));
- if ($choice =~ /^t$/i) {
+ $choice = lc(substr(Git::prompt("Certificate problem.\n" . $options) || 'R', 0, 1));
+ if ($choice eq 't') {
$cred->may_save(undef);
- } elsif ($choice =~ /^r$/i) {
+ } elsif ($choice eq 'r') {
return -1;
- } elsif ($may_save && $choice =~ /^p$/i) {
+ } elsif ($may_save && $choice eq 'p') {
$cred->may_save($may_save);
} else {
goto prompt;
@@ -109,9 +109,7 @@ sub username {
if (defined $_username) {
$username = $_username;
} else {
- print STDERR "Username: ";
- STDERR->flush;
- chomp($username = <STDIN>);
+ $username = Git::prompt("Username: ");
}
$cred->username($username);
$cred->may_save($may_save);
@@ -120,7 +118,7 @@ sub username {
sub _read_password {
my ($prompt, $realm) = @_;
- my $password = Git::prompt($prompt);
+ my $password = Git::prompt($prompt, 1);
$password;
}
--
Best regards,
Sven Strickroth
PGP key id F5A9D4C4 @ any key-server
^ permalink raw reply related
* [PATCH 1/3] git-svn, perl/Git.pm: add central method for prompting passwords
From: Sven Strickroth @ 2012-12-18 0:28 UTC (permalink / raw)
To: Junio C Hamano; +Cc: git, Jeff King, Jakub Narebski, Eric Wong
In-Reply-To: <7vehiol9w2.fsf@alter.siamese.dyndns.org>
git-svn reads passwords from an interactive terminal or by using
GIT_ASKPASS helper tool. This cause GUIs (w/o STDIN connected) to hang
waiting forever for git-svn to complete
(http://code.google.com/p/tortoisegit/issues/detail?id=967).
Commit 56a853b62c0ae7ebaad0a7a0a704f5ef561eb795 also tried to solve
this issue, but was incomplete as described above.
Instead of using hand-rolled prompt-response code that only works with the
interactive terminal, a reusable prompt() method is introduced in this commit.
Signed-off-by: Sven Strickroth <email@cs-ware.de>
---
perl/Git.pm | 45 ++++++++++++++++++++++++++++++++++++++++++++-
perl/Git/SVN/Prompt.pm | 20 +-------------------
2 files changed, 45 insertions(+), 20 deletions(-)
diff --git a/perl/Git.pm b/perl/Git.pm
index 497f420..72e93c7 100644
--- a/perl/Git.pm
+++ b/perl/Git.pm
@@ -58,7 +58,7 @@ require Exporter;
command_output_pipe command_input_pipe command_close_pipe
command_bidi_pipe command_close_bidi_pipe
version exec_path html_path hash_object git_cmd_try
- remote_refs
+ remote_refs prompt
temp_acquire temp_release temp_reset temp_path);
@@ -511,6 +511,49 @@ C<git --html-path>). Useful mostly only internally.
sub html_path { command_oneline('--html-path') }
+=item prompt ( PROMPT )
+
+Query user C<PROMPT> and return answer from user.
+
+Honours GIT_ASKPASS environment variable for querying
+the user. If no GIT_ASKPASS variable is set or an error occoured,
+the terminal is tried as a fallback.
+
+=cut
+
+sub prompt {
+ my ($prompt) = @_;
+ my $ret;
+ if (exists $ENV{'GIT_ASKPASS'}) {
+ $ret = _prompt($ENV{'GIT_ASKPASS'}, $prompt);
+ }
+ if (!defined $ret) {
+ print STDERR $prompt;
+ STDERR->flush;
+ require Term::ReadKey;
+ Term::ReadKey::ReadMode('noecho');
+ $ret = '';
+ while (defined(my $key = Term::ReadKey::ReadKey(0))) {
+ last if $key =~ /[\012\015]/; # \n\r
+ $ret .= $key;
+ }
+ Term::ReadKey::ReadMode('restore');
+ print STDERR "\n";
+ STDERR->flush;
+ }
+ return $ret;
+}
+
+sub _prompt {
+ my ($askpass, $prompt) = @_;
+ return unless length $askpass;
+ my $ret;
+ open my $fh, "-|", $askpass, $prompt or return;
+ $ret = <$fh>;
+ $ret =~ s/[\015\012]//g; # strip \r\n, chomp does not work on all systems (i.e. windows) as expected
+ close ($fh);
+ return $ret;
+}
=item repo_path ()
diff --git a/perl/Git/SVN/Prompt.pm b/perl/Git/SVN/Prompt.pm
index 3a6f8af..a2cbcc8 100644
--- a/perl/Git/SVN/Prompt.pm
+++ b/perl/Git/SVN/Prompt.pm
@@ -120,25 +120,7 @@ sub username {
sub _read_password {
my ($prompt, $realm) = @_;
- my $password = '';
- if (exists $ENV{GIT_ASKPASS}) {
- open(PH, "-|", $ENV{GIT_ASKPASS}, $prompt);
- $password = <PH>;
- $password =~ s/[\012\015]//; # \n\r
- close(PH);
- } else {
- print STDERR $prompt;
- STDERR->flush;
- require Term::ReadKey;
- Term::ReadKey::ReadMode('noecho');
- while (defined(my $key = Term::ReadKey::ReadKey(0))) {
- last if $key =~ /[\012\015]/; # \n\r
- $password .= $key;
- }
- Term::ReadKey::ReadMode('restore');
- print STDERR "\n";
- STDERR->flush;
- }
+ my $password = Git::prompt($prompt);
$password;
}
--
Best regards,
Sven Strickroth
PGP key id F5A9D4C4 @ any key-server
^ permalink raw reply related
* [PATCH 2/3] perl/Git.pm: Honor SSH_ASKPASS as fallback if GIT_ASKPASS is not set
From: Sven Strickroth @ 2012-12-18 0:28 UTC (permalink / raw)
To: Junio C Hamano; +Cc: git, Jeff King, Jakub Narebski, Eric Wong
In-Reply-To: <7vehiol9w2.fsf@alter.siamese.dyndns.org>
If GIT_ASKPASS environment variable is not set, git-svn does not try to use
SSH_ASKPASS as git-core does. This change adds a fallback to SSH_ASKPASS.
Signed-off-by: Sven Strickroth <email@cs-ware.de>
---
perl/Git.pm | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/perl/Git.pm b/perl/Git.pm
index 72e93c7..8dfca65 100644
--- a/perl/Git.pm
+++ b/perl/Git.pm
@@ -515,8 +515,8 @@ sub html_path { command_oneline('--html-path') }
Query user C<PROMPT> and return answer from user.
-Honours GIT_ASKPASS environment variable for querying
-the user. If no GIT_ASKPASS variable is set or an error occoured,
+Honours GIT_ASKPASS and SSH_ASKPASS environment variables for querying
+the user. If no *_ASKPASS variable is set or an error occoured,
the terminal is tried as a fallback.
=cut
@@ -527,6 +527,9 @@ sub prompt {
if (exists $ENV{'GIT_ASKPASS'}) {
$ret = _prompt($ENV{'GIT_ASKPASS'}, $prompt);
}
+ if (!defined $ret && exists $ENV{'SSH_ASKPASS'}) {
+ $ret = _prompt($ENV{'SSH_ASKPASS'}, $prompt);
+ }
if (!defined $ret) {
print STDERR $prompt;
STDERR->flush;
--
Best regards,
Sven Strickroth
PGP key id F5A9D4C4 @ any key-server
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox