Age | Commit message (Collapse) | Author |
|
The parser is actually really strict, and maybe we should back off on
enforcing some of the stricter rules.
|
|
|
|
|
|
This now supports most of the tree building. It has limited support for
insertion modes, as well as some syntax correction.
|
|
|
|
|
|
Minor bugs fixed, too.
|
|
|
|
|
|
|
|
|
|
|
|
then the regular expressions and twice as fast as the next fastest method I tested.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
for it.
|
|
subdirectory.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Neither iconv nor mb seem to be able to convert UTF-8 surrogates
into UTF-8. As I understand it, this is an extreme edge case.
Still, the behavior in both cases is that the surrogates are
stripped from the string. We test for that condition, now.
|
|
|
|
|