Replace std::map, std::set by WTF::HashMap and WTF::HashSet....

Replace std::map, std::set by WTF::HashMap and WTF::HashSet.

WebKit gurus, please review the change carefully, WTF::HashMap interface is not well documented, so I have implement
customized hash functions and HashTraits in order to use StringKey as the hash key.


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=14995

[Feng Qian, 2009-04-27 22:16:22]

Replace stl classes, map and set by WTF classes, HashMap and HashSet. WebKit
Guru, please review the change carefully. WTF::HashMap does not have clear
documentation and is not generic enough. I have to implement our own hash
functions and traits.

[Feng Qian, 2009-04-27 22:29:50]

Please hold off the review. I found some issues when running test. Fixing....

[Feng Qian, 2009-04-30 16:11:25]

Hi,

I think it is ready for review. I am going to fix lint issues.

Is 'gcl upload change_name' to update the patch? It seems lost the history.

Thanks
Feng

[dglazkov, 2009-04-30 21:14:04]

I think this is fine. Even though there are multiple WebKit style issues and
ideally this should be moved to WebCore/bindings/v8, this is probably out of
scope for this CL.

LGTM.

[Mike Belshe, 2009-04-30 21:37:52]

lgtm

http://codereview.chromium.org/100069/diff/2003/4
File webkit/port/bindings/v8/npruntime.cpp (right):

http://codereview.chromium.org/100069/diff/2003/4#newcode51
Line 51: StringKey(const char* str) : _string(str), _length(strlen(str)) {}
style wise, these should be m_string and m_length?  (for webkit style)

http://codereview.chromium.org/100069/diff/2003/4#newcode88
Line 88: // Use the same string hash function as in V8.
This comment might make the reader think this needs to match the v8 algorithm,
but I don't think it does?

[Feng Qian, 2009-04-30 23:04:13]

I already check it in, let's fix style issue when upstreaming the code.

On 2009/04/30 21:37:52, Mike Belshe wrote:
> lgtm
> 
> http://codereview.chromium.org/100069/diff/2003/4
> File webkit/port/bindings/v8/npruntime.cpp (right):
> 
> http://codereview.chromium.org/100069/diff/2003/4#newcode51
> Line 51: StringKey(const char* str) : _string(str), _length(strlen(str)) {}
> style wise, these should be m_string and m_length?  (for webkit style)
> 
> http://codereview.chromium.org/100069/diff/2003/4#newcode88
> Line 88: // Use the same string hash function as in V8.
> This comment might make the reader think this needs to match the v8 algorithm,
> but I don't think it does?