[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: Genx

• To: <xml-dev@l...>
• Subject: Re: Genx
• From: "Karl Waclawek" <karl@w...>
• Date: Tue, 20 Jan 2004 10:22:14 -0500
• References: <200401201509.i0KF9v8P088186@a...>

> Joe Gregorio:
> > David Tolpin wrote: 
> > > passing data in buffers oftern requires number of bytes or characters to be passed;
> > > consider getting data from Expat and writing it to the output.
> > > 
> > > int genxText(genxWriter w, const utf8Byte * text, int nbytes);
> > > int genxTextW(genxWriter w, const codePoint * text, int nchars);
> > 
> > This is a C interface and null terminated strings are
> > the standard way of handling this, the nbytes would
> > just be redundant information. If the implmentation of
> > genxText() needs the length it can call strlen(text), 
> > if it doesn't need the length then why force it to 
> > be passed in?
> 
> Have you looked at the API of James Clark's Expat?

The reason there is that (especially) character data are passed
through pointers into an internal working buffer. Having a null
terminator in the API would effectively force Expat to allocate
a new copy of the string.

Could there be a similar siutation for a writer API?
Possibly yes - as it is conceivable that the data to be
written out are part of some larger text/structure and
there is no reason to force a new string allocation on the caller.

Karl

• Follow-Ups:
  • Re: Genx
    • From: David Tolpin <dvd@d...>

• References:
  • Re: Genx
    • From: David Tolpin <dvd@d...>

• Prev by Date: Re: Genx
• Next by Date: Re: Genx
• Previous by thread: Re: Genx
• Next by thread: Re: Genx
• Index(es):
  • Date
  • Thread