[Webkit-unassigned] [Bug 70107] Custom written CSS lexer

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=70107





--- Comment #13 from Oliver Hunt <oliver at apple.com>  2011-12-12 13:28:17 PST ---
(In reply to comment #12)
> Hey Oliver,
> 
> actually the CSS parser working in an internal buffer, which is terminated by two '\0'-s, see the following code:
> 
> void CSSParser::setupParser(const char* prefix, const String& string, const char* suffix)
> {
>     int length = string.length() + strlen(prefix) + strlen(suffix) + 2;
> 
>     fastFree(m_data);
>     m_data = static_cast<UChar*>(fastMalloc(length * sizeof(UChar)));
>     for (unsigned i = 0; i < strlen(prefix); i++)
>         m_data[i] = prefix[i];
> 
>     memcpy(m_data + strlen(prefix), string.characters(), string.length() * sizeof(UChar));
> 
>     unsigned start = strlen(prefix) + string.length();
>     unsigned end = start + strlen(suffix);
>     for (unsigned i = start; i < end; i++)
>         m_data[i] = suffix[i - start];
> 
>     m_data[length - 1] = 0;
>     m_data[length - 2] = 0;
> 
>     yy_hold_char = 0;
>     yyleng = 0;
>     yytext = yy_c_buf_p = m_data;
>     yy_hold_char = *yy_c_buf_p;
>     resetRuleBodyMarks();
> }
> 
> Flex needs this, and I actually liked this approach, since it requires less buffer range checks (you know that any non \0 character must be followed at least two \0-s). CSS actually designed that way that '\0' is not part of any token.

On the downside it requires a copy of the css source -- is a well predicted branch really more expensive than a copy?

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.