IDL parser: fix rebuilding of (stale) cached lexer tables

IDL parser: fix rebuilding of (stale) cached lexer tables

The PLY lexer caches its table as a Python module (lextab.py). The
cache is loaded by attempting to import the module, which succeeds
if there's a lextab.py{,c,o} anywhere in the Python path. If the
import succeeds, the cache is assumed by PLY to be up-to-date, so
to update the cache, we need to remove those files before calling
PLY to initialize the lexer.

This patch makes blink_idl_lexer.main() do just that. It also
extends blink_idl_parser.main() to call the former, since the GYP
(and GN) build systems only actually call the latter script.

BUG=397909
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=179129

[Jens Widell, 2014-07-29 09:42:59]

PTAL

Similar to https://codereview.chromium.org/332683004, but handles the other half
of the IDL parser (i.e. the lexer.)

This fixes both the issue discovered by Nico (that having lextab.pyc but no
lextab.py caused infinite rebuilding) and that the cache wasn't updated properly
when the lexer definitions changed.

[haraken, 2014-07-29 10:03:51]

LGTM

[Jens Widell, 2014-07-29 10:54:04]

The CQ bit was checked by jl@opera.com

[commit-bot: I haz the power, 2014-07-29 10:55:10]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/jl@opera.com/425953002/1

[commit-bot: I haz the power, 2014-07-29 11:36:08]

FYI, CQ is re-trying this CL (attempt #1).
The failing builders are:
  win_blink_rel on tryserver.blink
(http://build.chromium.org/p/tryserver.blink/builders/win_blink_rel/builds/19434)

[commit-bot: I haz the power, 2014-07-29 12:17:48]

Change committed as 179129

[Nils Barth (inactive), 2014-07-29 16:00:25]

Quick style note.

https://codereview.chromium.org/425953002/diff/1/Source/bindings/scripts/blin...
File Source/bindings/scripts/blink_idl_lexer.py (right):

https://codereview.chromium.org/425953002/diff/1/Source/bindings/scripts/blin...
Source/bindings/scripts/blink_idl_lexer.py:111: unlink(tablefile)
FWIW, cleaner would be a loop through suffixes, which avoids needing a helper
function:

tablefile_root = os.path.join(outputdir, 'lextab')
for ext in ('.py', '.pyc', '.pyo'):
    try:
        os.unlink(tablefile_root + ext)
    except OSError:
        pass

...and to be really formal, we'd want
LEXTAB = 'lextab'
...and to use that both places.
...but that's a *bit* picky.