Add a "gn path" command for finding deps.

Add a "gn path" command for finding deps.

This command will find an annotate dependencies between two targets in
the tree. I wanted this when tracking down some dependency issues so
wrote a tool for it.

Committed: https://crrev.com/14769fd8e3bd31bd5f5f59e6d0feb7eaefd51e34
Cr-Commit-Position: refs/heads/master@{#330632}

[brettw, 2015-05-18 20:32:09]

brettw@chromium.org changed reviewers:
+ dpranke@chromium.org

[brettw, 2015-05-18 20:32:14]

The CQ bit was checked by brettw@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2015-05-18 20:32:55]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1130423006/1

[commit-bot: I haz the power, 2015-05-18 21:06:45]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-05-18 21:06:46]

Dry run: This issue passed the CQ dry run.

[Dirk Pranke, 2015-05-18 21:22:46]

This basically looks good. 

The bazel query language is here:

http://bazel.io/docs/query.html

It is substantially more complicated in approach, but one insight I took away
from it is that they have the option to say "somepath" or "allpaths" (where the
former is one randomly chosen thing).

Since I think we agree that the most likely case for running gn path is that
you're looking for at least one path because you expect there to be no paths,
perhaps we should default to just showing one path, and then add an "--all" flag
for the others?

[brettw, 2015-05-19 18:37:04]

New patch prints the first by default with a --all option as suggested. I also
sped it up by about 30% by not pushing and popping from the vector as much.

[brettw, 2015-05-19 18:37:10]

The CQ bit was checked by brettw@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2015-05-19 18:37:43]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1130423006/40001

[commit-bot: I haz the power, 2015-05-19 19:18:57]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-05-19 19:18:58]

Dry run: This issue passed the CQ dry run.

[Dirk Pranke, 2015-05-19 19:43:29]

https://codereview.chromium.org/1130423006/diff/40001/tools/gn/command_path.cc
File tools/gn/command_path.cc (right):

https://codereview.chromium.org/1130423006/diff/40001/tools/gn/command_path.c...
tools/gn/command_path.cc:59: (*found_count)++;
If I'm reading this correctly, this routine will still search for and find all
paths, but only print one of them. 

Am I right? If so, wouldn't it make sense to break out of the loops in lines
68-80 if print_all is false and found_count > 0?

[brettw, 2015-05-19 21:09:55]

https://codereview.chromium.org/1130423006/diff/40001/tools/gn/command_path.cc
File tools/gn/command_path.cc (right):

https://codereview.chromium.org/1130423006/diff/40001/tools/gn/command_path.c...
tools/gn/command_path.cc:59: (*found_count)++;
On 2015/05/19 19:43:29, Dirk Pranke wrote:
> If I'm reading this correctly, this routine will still search for and find all
> paths, but only print one of them. 
> 
> Am I right? If so, wouldn't it make sense to break out of the loops in lines
> 68-80 if print_all is false and found_count > 0?

Correct. I want to find the number of items which I print out at the end.
Currently it prints out that it found one but there are X other ones, which I
find very helpful.

[Dirk Pranke, 2015-05-19 22:00:48]

makes sense. I doubt it's all that slow either way.

lgtm.

[brettw, 2015-05-19 22:15:23]

The CQ bit was checked by brettw@chromium.org

[commit-bot: I haz the power, 2015-05-19 22:16:22]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1130423006/40001

[commit-bot: I haz the power, 2015-05-19 22:22:36]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-05-19 22:24:19]

Patchset 3 (id:??) landed as
https://crrev.com/14769fd8e3bd31bd5f5f59e6d0feb7eaefd51e34
Cr-Commit-Position: refs/heads/master@{#330632}