Restructure handling of list type extended attributes

Restructure handling of list type extended attributes

This patch does a number of things:

- Drops the [Attr=Some|Other] syntax, in favor of [Attr=(Some,Other)]
  which WebIDL defines, and converts all uses of the former in our IDL
  files. Also drops related grammar overrides from blink_idl_parser.py.

- Simplifies [Conditional] to only support single identifier form, which
  is the only way it's used in current (real) IDL files.

- Adds [Attr="value"] and [Attr=("some","other")] instead of the previous
  [Attr="some"|"other"] syntax, to better match the syntax of identifier
  lists.

- Changes the output from the parser to be Python lists for list type
  attributes, so that the rest of the code doesn't need to do string
  splitting operations. In the case of string literal values means we can
  now handle string literals containing ',' and '|' correctly, should
  that ever become an issue.

None of these changes affect code generation.

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=180852

[Jens Widell, 2014-08-25 10:47:14]

WDYT? Worth splitting up and landing?

These changes do not cause any CG changes.

[haraken, 2014-08-25 11:35:16]

LGTM. I'm fine with landing this CL in one go.

Just to confirm: Is [Attr=(A ,    B)] allowed? (Are spaces around ',' allowed?)

[Jens Widell, 2014-08-25 11:47:28]

On 2014/08/25 11:35:16, haraken wrote:
> LGTM. I'm fine with landing this CL in one go.

Okay, then I'll rewrite the description and do that.

> Just to confirm: Is [Attr=(A ,    B)] allowed? (Are spaces around ','
allowed?)

Yes, they're allowed. We were always parsing these compound values as sequences
of tokens from the lexer, which ignores all spaces (and tabs) for us, so nothing
changes in that regard. The old way of concatenating the tokens into a "A,B"
string had the spaces stripped by the same mechanism.

In general, we only need to worry about spaces in the code that uses regexps to
extract stuff from idl files, like in aggregate_generated_bindings.py.

[Jens Widell, 2014-08-25 13:55:47]

The CQ bit was checked by jl@opera.com

[commit-bot: I haz the power, 2014-08-25 13:56:45]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/jl@opera.com/464273003/20001

[commit-bot: I haz the power, 2014-08-25 13:59:19]

Committed patchset #2 (20001) as 180852