[webkit-reviews] review requested: [Bug 34019] Custom-written JavaScript parser : [Attachment 56704] latest version of my recursive descent parser

[bugzilla-daemon at webkit.org]

Zoltan Herczeg <zherczeg at webkit.org> has asked	for review:
Bug 34019: Custom-written JavaScript parser
https://bugs.webkit.org/show_bug.cgi?id=34019

Attachment 56704: latest version of my recursive descent parser
https://bugs.webkit.org/attachment.cgi?id=56704&action=review

------- Additional Comments from Zoltan Herczeg <zherczeg at webkit.org>
Here comes an updated version of our recursive descent JS parser. It's twice as
fast as the bison-generated one. Please, have a look.

TEST			       COMPARISON	     FROM		  TO   
	 DETAILS

=============================================================================

** TOTAL **:		       2.17x as fast	 75.4ms +/- 0.9%   34.7ms +/-
1.4%	 significant

=============================================================================

  jquery:		       2.29x as fast	 11.9ms +/- 1.9%    5.2ms +/-
5.8%	 significant
    1.3.2:		       2.29x as fast	 11.9ms +/- 1.9%    5.2ms +/-
5.8%	 significant

  mootools:		       2.10x as fast	 12.4ms +/- 3.0%    5.9ms +/-
3.8%	 significant
    1.2.2-core-nc:	       2.10x as fast	 12.4ms +/- 3.0%    5.9ms +/-
3.8%	 significant

  prototype:		       2.12x as fast	 13.8ms +/- 2.2%    6.5ms +/-
5.8%	 significant
    1.6.0.3:		       2.12x as fast	 13.8ms +/- 2.2%    6.5ms +/-
5.8%	 significant

  concat:		       2.18x as fast	 37.3ms +/- 0.9%   17.1ms +/-
1.3%	 significant
    jquery-mootools-prototype: 2.18x as fast	 37.3ms +/- 0.9%   17.1ms +/-
1.3%	 significant

(Oliver mentioned on IRC that there are plans for writing a recursive descent
parser. Why to implement another one if here is an already working solution?)