XML file parser using C language

gopi_raghu · 08-21-2007, 07:36 AM

I am new to parsing XML file using C

I need to parse the XML file and get the values between tags(start and end).I am using expat library(expat.h) and am having C API as fallows

XML_SetElementHandler(parser, startElement, endElement);

static void XMLCALL
startElement(void *userData, const char *name, const char **atts)
{
int i;
int natts;
int *depthPtr = (int *)userData;
const XML_Char **p;

for (i = 0; i < global_depth; i++)
printf(" ");

printf("%s\n", name);
LINE:8 for (i = 0; atts[i]; i += 2)
printf("%s = %s", atts[i], atts[i + 1]);

printf("\n");
// printf("%d\n", global_depth);
global_depth++;

// printf("%s = %s\n", atts[0], atts[1]);
// puts(name);
// *depthPtr += 1;
}

I am trying to print the values between start and end tag but is not printing (LINE no 8 in code).

Any code to get values between start and end tag of XML file ?

Thanks in advance

95se · 08-21-2007, 11:11 AM

The atts argument passed in is the element attributes. So, if you have something like this:

Code:

<some-element id="blah" name="Joe Blow">
Some text.
</some-element>

The atts will hold something like this:

Code:

atts[0] = "id"
atts[1] = "blah"
atts[2] = "name"
atts[3] = "Joe Blow"

To get "Some text.", you have to set up a CharacterData (NOT Cdata) handler. This will get the text passed into it. A warning though, most XML parsers, expat included, may split up the text. You can't count on having the entire string "Some text." given to you on a single call of your CharacterData handler, it may be split up across 2 or more. So, you would have to have some function like:

Code:

/* ... */
        XML_SetCharacterDataHandler(parser, text_handler);
/* ... */
void text_handler(void * user_data, const XML_Char *s, int len)
{
        fwrite(s, 1, len, stdout);

        /* this may be put in the middle of the text */
        putchar('\n'); 

        /* perhaps something to keep track of all the text */
        /* buffer_append(((buffer *) user_data), s, len); */
}

With expat, it REALLY pays off to read the header file (expat.h).

Edit: Note, the, XML_Char *, text passed in is not NULL terminated.

gopi_raghu · 08-22-2007, 01:06 AM

Thanks a lot.
Got the output.

bye

T.Raghavendra

Quote:

Originally Posted by 95se

The atts argument passed in is the element attributes. So, if you have something like this:

Code:

<some-element id="blah" name="Joe Blow">
Some text.
</some-element>

The atts will hold something like this:

Code:

atts[0] = "id"
atts[1] = "blah"
atts[2] = "name"
atts[3] = "Joe Blow"

To get "Some text.", you have to set up a CharacterData (NOT Cdata) handler. This will get the text passed into it. A warning though, most XML parsers, expat included, may split up the text. You can't count on having the entire string "Some text." given to you on a single call of your CharacterData handler, it may be split up across 2 or more. So, you would have to have some function like:

Code:

/* ... */
        XML_SetCharacterDataHandler(parser, text_handler);
/* ... */
void text_handler(void * user_data, const XML_Char *s, int len)
{
        fwrite(s, 1, len, stdout);

        /* this may be put in the middle of the text */
        putchar('\n'); 

        /* perhaps something to keep track of all the text */
        /* buffer_append(((buffer *) user_data), s, len); */
}

With expat, it REALLY pays off to read the header file (expat.h).

Edit: Note, the, XML_Char *, text passed in is not NULL terminated.