threads / discuss / 18885
Subject: "git cat-file <sha1>" shows usage but should not?
Here is what I did:
$ git cat-file -t b8c11fbbffbe171c4960058fa2d96f708e57a80c commit
$ git cat-file b8c11fbbffbe171c4960058fa2d96f708e57a80c usage: git cat-file [-t|-s|-e|-p|<type>] <sha1> or: git cat-file [--batch|--batch-check] < <list_of_sha1s> [..]
$ git --version git version 1.6.3.rc0.1.gf800.dirty
Sebastian
Jeff KingRe: "git cat-file <sha1>" shows usage but should not?
On Wed, Apr 15, 2009 at 06:50:47PM +0200, Sebastian Pipping wrote:
> Here is what I did: > > $ git cat-file -t b8c11fbbffbe171c4960058fa2d96f708e57a80c > commit > > $ git cat-file b8c11fbbffbe171c4960058fa2d96f708e57a80c > usage: git cat-file [-t|-s|-e|-p|<type>] <sha1> > or: git cat-file [--batch|--batch-check] < <list_of_sha1s> > [..]
That usage message is misleading. You need to use one of -t, -s, -e, -p, or <type>. So it should perhaps just be:
git cat-file -t|-s|-e|-p|<type> <sha1>
though that looks terribly ugly. Suggestions welcome.
-Peff
P.S. Depending on what you're trying to accomplish, "git show <sha1>"
may be a better choice.Re: "git cat-file <sha1>" shows usage but should not?
Jeff King wrote:
> That usage message is misleading. You need to use one of -t, -s, -e, -p, > or <type>. So it should perhaps just be: > > git cat-file -t|-s|-e|-p|<type> <sha1> > > though that looks terribly ugly. Suggestions welcome.
Thanks for making that clear. How about round or curly brackets?
Sebastian
Jeff KingRe: "git cat-file <sha1>" shows usage but should not?
On Wed, Apr 15, 2009 at 06:56:05PM +0200, Sebastian Pipping wrote:
> Jeff King wrote: > > That usage message is misleading. You need to use one of -t, -s, -e, -p, > > or <type>. So it should perhaps just be: > > > > git cat-file -t|-s|-e|-p|<type> <sha1> > > > > though that looks terribly ugly. Suggestions welcome. > > Thanks for making that clear. How about round or curly brackets?
I think curly makes the most sense for grouping, though I still think it is a bit ugly. Patch is below.
-- >8 -- Subject: [PATCH] fix cat-file usage message and documentation
cat-file with an object on the command line requires an option to tell it what to output (type, size, pretty-print, etc). However, the square brackets in the usage imply that those options are not required. This patch switches them to curly braces to indicate "required but grouped-OR".
While we're at it, let's change the <sha1> specifier in the usage to <object>. That's what the documentation uses, and it does actually use the regular object lookup. --- The curly braces seem to render fine via asciidoc (presumably because there is no entity with that name), but I'm not sure if that will work with all asciidoc versions.
Documentation/git-cat-file.txt | 2 +- builtin-cat-file.c | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/Documentation/git-cat-file.txt b/Documentation/git-cat-file.txt index b191276..0856568 100644 --- a/Documentation/git-cat-file.txt +++ b/Documentation/git-cat-file.txt @@ -9,7 +9,7 @@ git-cat-file - Provide content or type and size information for repository objec SYNOPSIS -------- [verse] -'git cat-file' [-t | -s | -e | -p | <type>] <object> +'git cat-file' {-t | -s | -e | -p | <type>} <object> 'git cat-file' [--batch | --batch-check] < <list-of-objects> DESCRIPTION diff --git a/builtin-cat-file.c b/builtin-cat-file.c index 8fad19d..4e6ad73 100644 --- a/builtin-cat-file.c +++ b/builtin-cat-file.c @@ -201,8 +201,8 @@ static int batch_objects(int print_contents) } static const char * const cat_file_usage[] = { - "git cat-file [-t|-s|-e|-p|<type>] <sha1>", - "git cat-file [--batch|--batch-check] < <list_of_sha1s>", + "git cat-file {-t|-s|-e|-p|<type>} <object>", + "git cat-file [--batch|--batch-check] < <list_of_objects>", NULL };
-- 1.6.3.rc0.156.g19ef8
Junio C HamanoRe: "git cat-file <sha1>" shows usage but should not?
Jeff King <peff@peff.net> writes:
> On Wed, Apr 15, 2009 at 06:56:05PM +0200, Sebastian Pipping wrote: > >> Jeff King wrote: >> > That usage message is misleading. You need to use one of -t, -s, -e, -p, >> > or <type>. So it should perhaps just be: >> > >> > git cat-file -t|-s|-e|-p|<type> <sha1> >> > >> > though that looks terribly ugly. Suggestions welcome. >> >> Thanks for making that clear. How about round or curly brackets? > > I think curly makes the most sense for grouping, though I still think it > is a bit ugly. Patch is below.
Thanks. Doesn't this remind us of $gmane/72243, by the way?
Jeff KingRe: "git cat-file <sha1>" shows usage but should not?
On Wed, Apr 15, 2009 at 10:18:40AM -0700, Junio C Hamano wrote:
> > I think curly makes the most sense for grouping, though I still think it > > is a bit ugly. Patch is below. > > Thanks. Doesn't this remind us of $gmane/72243, by the way?
Ah, yes. I felt like we had dealt with this before, but I couldn't remember for what.
That message indicates that {} is preferred to (). Interestingly, though, the stash usage and documentation still uses ().
-Peff
Julian PhillipsRe: "git cat-file <sha1>" shows usage but should not?
On Wed, 15 Apr 2009, Jeff King wrote:
> static const char * const cat_file_usage[] = {
> - "git cat-file [-t|-s|-e|-p|<type>] <sha1>",
> - "git cat-file [--batch|--batch-check] < <list_of_sha1s>",
> + "git cat-file {-t|-s|-e|-p|<type>} <object>",
> + "git cat-file [--batch|--batch-check] < <list_of_objects>",Shouldn't that read:
+ "git cat-file {--batch|--batch-check} < <list_of_objects>",since, like the previous line, you are required to supply one of the two options?
--
Julian
---
DPRINTK("FAILURE, CAPUT\n");
linux-2.6.6/drivers/net/tokenring/ibmtr.cJeff KingRe: "git cat-file <sha1>" shows usage but should not?
On Wed, Apr 15, 2009 at 06:33:10PM +0100, Julian Phillips wrote:
> On Wed, 15 Apr 2009, Jeff King wrote:
>
>> static const char * const cat_file_usage[] = {
>> - "git cat-file [-t|-s|-e|-p|<type>] <sha1>",
>> - "git cat-file [--batch|--batch-check] < <list_of_sha1s>",
>> + "git cat-file {-t|-s|-e|-p|<type>} <object>",
>> + "git cat-file [--batch|--batch-check] < <list_of_objects>",
>
> Shouldn't that read:
>
> + "git cat-file {--batch|--batch-check} < <list_of_objects>",
>
> since, like the previous line, you are required to supply one of the two
> options?Oops, yes, it should.
-Peff