Find issue number from comment in status file.

Find issue number from comment in status file.

R=kustermann@google.com
BUG=

Committed: https://code.google.com/p/dart/source/detail?r=26360

[Søren Gjesse, 2013-08-20 13:48:04]

This information is not used yet. There is not test as the testing of the status
file parser are skipped as it is not running with the current VM.

[kustermann, 2013-08-20 14:29:55]

lgtm

We should probably enforce a certain way of specifying issue numbers (and make
sure every entry has an issue number).

https://codereview.chromium.org/23172010/diff/1/tools/testing/dart/status_fil...
File tools/testing/dart/status_file_parser.dart (right):

https://codereview.chromium.org/23172010/diff/1/tools/testing/dart/status_fil...
tools/testing/dart/status_file_parser.dart:27: final RegExp SplitComment = new
RegExp("(^[^#]*)(:?#(.*))?");
"(^[^#]*)(:?#(.*))?"  -->  "^([^#]*)(#.*)?$" ?

This is maybe a bit easier to read.
The first one greedily matches all non '#' characters from the beginning of the
line, the second one will match '#' and everything else until the end of the
line.

https://codereview.chromium.org/23172010/diff/1/tools/testing/dart/status_fil...
tools/testing/dart/status_file_parser.dart:31: new RegExp("Issue
([0-9]+)|dartbug.com/([0-9]+)", caseSensitive: false);
I think there are more patterns, like:
 "Issue: 123"
 "http://dartbug.com/123"
 ...

Maybe we should just have a list of regular expressions, it would definitely be
easier to read.

https://codereview.chromium.org/23172010/diff/1/tools/testing/dart/status_fil...
tools/testing/dart/status_file_parser.dart:116: if (issueString == null)
issueString = match[2];
Are you sure that if a optional group was not matched it's null and not ""?

[Søren Gjesse, 2013-08-20 14:55:48]

Yes, I will add the requirement of having issue numbers on all lines to the
presubmit script in a way so we can enable it on a file-by-file basis.

https://codereview.chromium.org/23172010/diff/1/tools/testing/dart/status_fil...
File tools/testing/dart/status_file_parser.dart (right):

https://codereview.chromium.org/23172010/diff/1/tools/testing/dart/status_fil...
tools/testing/dart/status_file_parser.dart:27: final RegExp SplitComment = new
RegExp("(^[^#]*)(:?#(.*))?");
On 2013/08/20 14:29:56, kustermann wrote:
> "(^[^#]*)(:?#(.*))?"  -->  "^([^#]*)(#.*)?$" ?
> 
> This is maybe a bit easier to read.
> The first one greedily matches all non '#' characters from the beginning of
the
> line, the second one will match '#' and everything else until the end of the
> line.

Done. I only did it the other way to avoid having the # as part of the comment,
but it does not matter.

https://codereview.chromium.org/23172010/diff/1/tools/testing/dart/status_fil...
tools/testing/dart/status_file_parser.dart:31: new RegExp("Issue
([0-9]+)|dartbug.com/([0-9]+)", caseSensitive: false);
On 2013/08/20 14:29:56, kustermann wrote:
> I think there are more patterns, like:
>  "Issue: 123"
>  "http://dartbug.com/123"
>  ...
> 
> Maybe we should just have a list of regular expressions, it would definitely
be
> easier to read.

I know, but I think we should stick to these two, and maybe in the end just have
one. http:// in front of dartbug.com/ should be found.

https://codereview.chromium.org/23172010/diff/1/tools/testing/dart/status_fil...
tools/testing/dart/status_file_parser.dart:116: if (issueString == null)
issueString = match[2];
On 2013/08/20 14:29:56, kustermann wrote:
> Are you sure that if a optional group was not matched it's null and not ""?

Yes - I tested it manually.

[Søren Gjesse, 2013-08-20 15:01:07]

Committed patchset #1 manually as r26360 (presubmit successful).