IDL: various adjustments to match Web IDL specification better

IDL: various adjustments to match Web IDL specification better

Notably adds support for maplike<type, type>, setlike<type>, iterable<type>
and legacyiterable<type> declarations.

Also changes some interface member grammar productions around a bit, and
drops the deprecated Attribute production, which Blink's parser no longer
references.

Committed: https://crrev.com/9016ef32769c396deb32c17dc5f3ff4a6d8acdd4
Cr-Commit-Position: refs/heads/master@{#308584}

[Jens Widell, 2014-12-15 11:35:15]

jl@opera.com changed reviewers:
+ haraken@chromium.org, noelallen@chromium.org

[Jens Widell, 2014-12-15 11:35:16]

PTAL

There are no (relevant) changes in parser's output. Blink's IDL compiler is
unaffected.

https://codereview.chromium.org/803933003/diff/20001/tools/idl_parser/idl_par...
File tools/idl_parser/idl_parser.py (right):

https://codereview.chromium.org/803933003/diff/20001/tools/idl_parser/idl_par...
tools/idl_parser/idl_parser.py:530: # [45] Unreferenced in the specification
This is:

[45] Inherit → "inherit"
               | ε

The productions that include the inherit keyword however all use "inherit"
directly instead of referencing this.

[haraken, 2014-12-15 15:11:13]

LGTM

https://codereview.chromium.org/803933003/diff/20001/tools/idl_parser/idl_par...
File tools/idl_parser/idl_parser.py (right):

https://codereview.chromium.org/803933003/diff/20001/tools/idl_parser/idl_par...
tools/idl_parser/idl_parser.py:528: p[0] = p[1]

p[0] = self.BuildTrue('REQUIRED') ?

https://codereview.chromium.org/803933003/diff/20001/tools/idl_parser/idl_par...
tools/idl_parser/idl_parser.py:671: if len(p) > 1:

if len(p) > 2:

[Jens Widell, 2014-12-15 15:30:32]

Thanks for review!

https://codereview.chromium.org/803933003/diff/20001/tools/idl_parser/idl_par...
File tools/idl_parser/idl_parser.py (right):

https://codereview.chromium.org/803933003/diff/20001/tools/idl_parser/idl_par...
tools/idl_parser/idl_parser.py:528: p[0] = p[1]
On 2014/12/15 15:11:13, haraken wrote:
> 
> p[0] = self.BuildTrue('REQUIRED') ?

No, this is simply a grammar workaround to allow attributes named "required"
despite that (now) being a keyword, so the output should be a string.

https://codereview.chromium.org/803933003/diff/20001/tools/idl_parser/idl_par...
tools/idl_parser/idl_parser.py:671: if len(p) > 1:
On 2014/12/15 15:11:13, haraken wrote:
> 
> if len(p) > 2:

The pattern in cases like this is "len(p) > 1". See for instance [50] Specials()
or [52] OptionalIdentifier() above.

[Jens Widell, 2014-12-16 15:01:46]

The CQ bit was checked by jl@opera.com

[commit-bot: I haz the power, 2014-12-16 15:02:26]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/803933003/20001

[commit-bot: I haz the power, 2014-12-16 15:05:20]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2014-12-16 15:06:14]

Patchset 2 (id:??) landed as
https://crrev.com/9016ef32769c396deb32c17dc5f3ff4a6d8acdd4
Cr-Commit-Position: refs/heads/master@{#308584}