Re[2]: Related: xml2po (was Re: HTML2POT and PO2HTML tools)

[Funda Wang <fundawang linux net cn>]

I'm the developer Jack mentioned.

liu> One of our developers is doing the same work. He use xmllint and php. 
No, I'm just using php without libxml. 

liu> You can only provide a 
liu> mechanism to process XHTML, so it's features are too limited.
No, the script Danilo write could phrase XML quite well, but only for
the well-formed ones.

In some projects(as Mandrake doc project), the source is stored
separated in many XML files. The entities are stored in an another
common file. The result(HTML/PDF) are generated by Borges which is a
program Mandrake doc team wrote for content management.

Thus, those XML files are not well-formed, as the entities are not
declared. That's why I must throw the libxml away and turned to regular
expression matching, which is not much easy.