const char* csXmlReadDocument::Parse (iFile* file, bool collapse)
{
size_t want_size = file->GetSize ();
char *data = (char*)cs_malloc (want_size + 1);
char* parse_data = data;
if (want_size >= 3)
{
if (file->Read (data, 3) != 3)
{
cs_free (parse_data);
return "Unexpected EOF encountered";
}
want_size -= 3;
if (!HasUTF8Bom (data))
// Not a BOM - keep the read sata
data += 3;
}
size_t real_size = file->Read (data, want_size);
if (want_size != real_size)
{
cs_free (parse_data);
return "Unexpected EOF encountered";
}
data[real_size] = '\0';
#ifdef CS_DEBUG
if (strlen (data) != real_size)
{
cs_free (parse_data);
return "File contains one or more null characters";
}
#endif
const char* b = parse_data;
while ((*b == ' ') || (*b == '\n') || (*b == '\t') ||
(*b == '\r')) b++;
if (*b != '<')
{
cs_free (parse_data);
return "Data does not seem to be XML.";
}
/* Note: it's okay if want_size is a bit too small;
* it's not used as the actual data size in ParseInPlace,
* only to choose between 'small' and 'large' docs. */
return ParseInPlace (parse_data, want_size, collapse);
}
const char* csXmlReadDocument::Parse (const char* buf, size_t bufSize, bool collapse)
{
// Skip any UTF8 BOM
if ((bufSize >= 3) && HasUTF8Bom (buf))
{
buf += 3;
bufSize -= 3;
}
const char* b = buf;
while ((*b == ' ') || (*b == '\n') || (*b == '\t') ||
(*b == '\r')) b++;
if (*b != '<')
{
return "Data does not seem to be XML.";
}
size_t want_size = bufSize;
char *data = (char*)cs_malloc (want_size + 1);
memcpy (data, buf, bufSize);
data[bufSize] = 0;
return ParseInPlace (data, bufSize, collapse);
}
HtmlElement *HtmlParser::Parse(const char *s, UINT codepage)
{
HtmlElement *root = ParseInPlace(str::Dup(s), codepage);
freeHtml = true;
return root;
}
