Hubbub $Id$
detect.h File Reference
#include <inttypes.h>
#include <parserutils/errors.h>

Go to the source code of this file.

Functions

parserutils_error hubbub_charset_extract (const uint8_t *data, size_t len, uint16_t *mibenum, uint32_t *source)
 Extract a charset from a chunk of data.
uint16_t hubbub_charset_parse_content (const uint8_t *value, uint32_t valuelen)
 Parse a content= attribute's value.
void hubbub_charset_fix_charset (uint16_t *charset)
 Fix charsets, according to the override table in HTML5, section 8.2.2.2.

Function Documentation

◆ hubbub_charset_extract()

parserutils_error hubbub_charset_extract ( const uint8_t * data,
size_t len,
uint16_t * mibenum,
uint32_t * source )

Extract a charset from a chunk of data.

Parameters
dataPointer to buffer containing data
lenBuffer length
mibenumPointer to location containing current MIB enum
sourcePointer to location containint current charset source
Returns
PARSERUTILS_OK on success, appropriate error otherwise

mibenum and source will be updated on exit

The larger a chunk of data fed to this routine, the better, as it allows charset autodetection access to a larger dataset for analysis.

Meaning of *source on entry:

CONFIDENT - Do not pass Go, do not attempt auto-detection. TENTATIVE - We've tried to autodetect already, but subsequently discovered that we don't actually support the detected charset. Thus, we've defaulted to Windows-1252. Don't perform auto-detection again, as it would be futile. (This bit diverges from the spec) UNKNOWN - No autodetection performed yet. Get on with it.

Todo
We probably want to wait for ~512 bytes of data / 500ms here
Todo
Charset autodetection

◆ hubbub_charset_fix_charset()

void hubbub_charset_fix_charset ( uint16_t * charset)

Fix charsets, according to the override table in HTML5, section 8.2.2.2.

Character encoding requirements http://www.whatwg.org/specs/web-apps/current-work/#character0

Parameters
charsetPointer to charset value to fix

◆ hubbub_charset_parse_content()

uint16_t hubbub_charset_parse_content ( const uint8_t * value,
uint32_t valuelen )

Parse a content= attribute's value.

Parameters
valueAttribute's value
valuelenLength of value
Returns
MIB enum of detected encoding, or 0 if none found