Re: [xml] Build DOM tree manually from ASCII file and getting data into xmlChar

[Steve Williams <swilliams rinax com>]

Hi,

Thanks for the assistance to date, it is much appreciated.  One of the 
joys of having the source is that I can look and see exactly how the 
routine works.  

As it stands now, isolat1ToUTF8 will never return an error.  (there's 
only one return statement, and it's a "0").

So, I figure if I check the number of bytes processed in the input 
stream against the length of the input stream would work fine as a 
verification that the entire input string was processed.

Here's a sample of code I threw togeter....  does anything glaring wrong 
jump out at anyone??

#include <tree.h>

/* 80 character input buffer */
#define ISOLAT1BUFF 80

/* isolat1ToUTF8 shouldn't more than double (worst case) */
#define UTF8DEFAULT 2*ISOLAT1BUFF

main()
{
 int tmp;

 char * isolat1Tag="test conversion";
 int isolat1TagLength;

 xmlChar *utf8Tag = xmlMalloc( UTF8DEFAULT * sizeof(utf8Tag));
 int utf8TagLength = UTF8DEFAULT;

 /* Let's take the input tag and convert from an assumed format of
  * ISO-8559-1 (iso-latin-1) to the internal representation UTF-8
  */
 tmp = isolat1TagLength = strlen(isolat1Tag);

 if (!isolat1ToUTF8(utf8Tag, &utf8TagLength, isolat1Tag, 
&isolat1TagLength)){
   /* It worked, let's make sure that it processed the whole input 
string */
   if (isolat1TagLength != tmp){
     fprintf(stderr, "isolat1ToUTF8 could not process entire TAG\n");
     exit(1);
   }
 } else {
   fprintf(stderr, "isolat1ToUTF8 returned an error!!\n");
   exit(1);
 }

 /* null terminate the UTF8 string */
 *(utf8Tag+utf8TagLength)='\0';

 exit(0);
}



Daniel Veillard wrote:

> On Thu, Aug 01, 2002 at 08:53:47PM +0200, Hannu Krosing wrote:
>  
>
> > On Thu, 2002-08-01 at 10:49, Daniel Veillard wrote:
> >    
> >
> > >  French characters (well accented ones for example) are not in ASCII
> > > but probably in ISO Latin 1, use the following function to convert them
> > > before passing the strings to the API:
> > >  int
> > >  isolat1ToUTF8(unsigned char* out, int *outlen,
> > >                const unsigned char* in, int *inlen) {
> > >      
> > Why is it "unsigned char* out" and not "xmlChar* out" ?
> >    
>
>  for conformance with iconv() interfaces
>
>  
>
> > Or are they actually made the same by some macro ?
> >    
>
>  No need for macros in both case it's a 0 terminated sequence of bytes
> you just need to be sure it's in the UTF8 encoding to use it for xmlChar *
>
> Daniel
>
>