[PATCH] grep: -L should show empty files

The -L (--files-without-match) option is supposed to show paths that
produced no matches.  When running the internal grep on work tree files,
however, we had an optimization to just return on zero-sized files,
without doing anything.

This optimization doesn't matter too much in practice (a tracked empty
file must be rare, or there is something wrong with your project); to
produce results consistent with GNU grep, we should stop the optimization
and show empty files as not having the given pattern.

Signed-off-by: Junio C Hamano <redacted>
---

 * Fix for a longstanding bug meant for maint.

 builtin-grep.c |    2 --
 1 files changed, 0 insertions(+), 2 deletions(-)

diff --git a/builtin-grep.c b/builtin-grep.c
index fd450bc..84a5af3 100644
--- a/builtin-grep.c
+++ b/builtin-grep.c

@@ -159,8 +159,6 @@ static int grep_file(struct grep_opt *opt, const char *filename)
 			error("'%s': %s", filename, strerror(errno));
 		return 0;
 	}
-	if (!st.st_size)
-		return 0; /* empty file -- no grep hit */
 	if (!S_ISREG(st.st_mode))
 		return 0;
 	sz = xsize_t(st.st_size);

-- 
1.6.6.280.ge295b7.dirty

Heya,

On Tue, Jan 12, 2010 at 09:32, Junio C Hamano [off-list ref] wrote:

This optimization doesn't matter too much in practice (a tracked empty
file must be rare, or there is something wrong with your project);

How about python projects, where there's an __init__.py file
everywhere you turn your head? ;)

-- 
Cheers,

Sverre Rabbelier

Sverre Rabbelier [off-list ref] writes:

Heya,

On Tue, Jan 12, 2010 at 09:32, Junio C Hamano [off-list ref] wrote:

quoted

This optimization doesn't matter too much in practice (a tracked empty
file must be rare, or there is something wrong with your project);

How about python projects, where there's an __init__.py file
everywhere you turn your head? ;)

It's Ok as the price we pay for producing correct result is to open those
empty files, read them, and look for matches which we will never find ;-)

Heya,

On Wed, Jan 13, 2010 at 07:56, Junio C Hamano [off-list ref] wrote:

It's Ok as the price we pay for producing correct result is to open those
empty files, read them, and look for matches which we will never find ;-)

I'm not that familiar with the code, but wouldn't it be possible to
keep the early abort, but make it dependent on not using the '-L'
flag?

-- 
Cheers,

Sverre Rabbelier

Sverre Rabbelier [off-list ref] writes:

I'm not that familiar with the code, but wouldn't it be possible to
keep the early abort, but make it dependent on not using the '-L'
flag?

Anything codable is possible to code, but my point was I don't think such
an optimization to avoid reading empty files is worth the time to write
and maintain extra code necessary for it.