parser.c File Reference

#include <assert.h>
#include <ctype.h>
#include <stdarg.h>
#include <stdlib.h>
#include <string.h>
#include <strings.h>
#include "attribute.h"
#include "error.h"
#include "gumbo.h"
#include "insertion_mode.h"
#include "parser.h"
#include "tokenizer.h"
#include "tokenizer_states.h"
#include "utf8.h"
#include "util.h"
#include "vector.h"

Classes
struct	_ReplacementEntry
struct	_NamespacedAttributeReplacement
struct	_TextNodeBufferState
struct	GumboInternalParserState
struct	InsertionLocation

Macros
#define	AVOID_UNUSED_VARIABLE_WARNING(i)   (void)(i)
#define	GUMBO_STRING(literal)    { literal, sizeof(literal) - 1 }
#define	TERMINATOR    { "", 0 }
#define	TAG(tag)   [GUMBO_TAG_##tag] = (1 << GUMBO_NAMESPACE_HTML)
#define	TAG_SVG(tag)   [GUMBO_TAG_##tag] = (1 << GUMBO_NAMESPACE_SVG)
#define	TAG_MATHML(tag)   [GUMBO_TAG_##tag] = (1 << GUMBO_NAMESPACE_MATHML)
#define	TAGSET_INCLUDES(tagset, namespace, tag)    (tag < GUMBO_TAG_LAST && tagset[(int) tag] & (1 << (int) namespace))
#define	REPLACEMENT_ENTRY(from, to)    { GUMBO_STRING(from), GUMBO_STRING(to) }

Typedefs
typedef char	gumbo_tagset[GUMBO_TAG_LAST]
typedef struct _ReplacementEntry	ReplacementEntry
typedef struct _NamespacedAttributeReplacement	NamespacedAttributeReplacement
typedef struct _TextNodeBufferState	TextNodeBufferState
typedef struct GumboInternalParserState	GumboParserState
typedef bool(*	TokenHandler) (GumboParser *parser, GumboToken *token)

Functions
InsertionLocation	get_appropriate_insertion_location (GumboParser *parser, GumboNode *override_target)
void	clear_stack_to_table_body_context (GumboParser *parser)
GumboNode *	clone_node (GumboParser *parser, GumboNode *node, GumboParseFlags reason)
const char *	gumbo_normalize_svg_tagname (const GumboStringPiece *tag)
GumboOutput *	gumbo_parse (const char *buffer)
GumboOutput *	gumbo_parse_with_options (const GumboOptions *options, const char *buffer, size_t length)
void	gumbo_destroy_node (GumboOptions *options, GumboNode *node)
void	gumbo_destroy_output (const GumboOptions *options, GumboOutput *output)

Variables
const GumboOptions	kGumboDefaultOptions

Macro Definition Documentation

AVOID_UNUSED_VARIABLE_WARNING

#define AVOID_UNUSED_VARIABLE_WARNING

(

i

)

(void)(i)

GUMBO_STRING

#define GUMBO_STRING

(

literal

)

{ literal, sizeof(literal) - 1 }

REPLACEMENT_ENTRY

#define REPLACEMENT_ENTRY	(		from,
			to
	)		{ GUMBO_STRING(from), GUMBO_STRING(to) }

TAG

#define TAG

(

tag

)

[GUMBO_TAG_##tag] = (1 << GUMBO_NAMESPACE_HTML)

TAG_MATHML

#define TAG_MATHML

(

tag

)

[GUMBO_TAG_##tag] = (1 << GUMBO_NAMESPACE_MATHML)

TAG_SVG

#define TAG_SVG

(

tag

)

[GUMBO_TAG_##tag] = (1 << GUMBO_NAMESPACE_SVG)

TAGSET_INCLUDES

#define TAGSET_INCLUDES	(		tagset,
			namespace,
			tag
	)		(tag < GUMBO_TAG_LAST && tagset[(int) tag] & (1 << (int) namespace))

TERMINATOR

#define TERMINATOR    { "", 0 }

Typedef Documentation

gumbo_tagset

typedef char gumbo_tagset[GUMBO_TAG_LAST]

GumboParserState

typedef struct GumboInternalParserState GumboParserState

NamespacedAttributeReplacement

typedef struct _NamespacedAttributeReplacement NamespacedAttributeReplacement

ReplacementEntry

typedef struct _ReplacementEntry ReplacementEntry

TextNodeBufferState

typedef struct _TextNodeBufferState TextNodeBufferState

TokenHandler

typedef bool(* TokenHandler) (GumboParser *parser, GumboToken *token)

Function Documentation

clear_stack_to_table_body_context()

void clear_stack_to_table_body_context

(

GumboParser *

parser

)

clone_node()

GumboNode * clone_node	(	GumboParser *	parser,
		GumboNode *	node,
		GumboParseFlags	reason
	)

get_appropriate_insertion_location()

InsertionLocation get_appropriate_insertion_location	(	GumboParser *	parser,
		GumboNode *	override_target
	)

gumbo_destroy_node()

void gumbo_destroy_node	(	GumboOptions *	options,
		GumboNode *	node
	)

gumbo_destroy_output()

void gumbo_destroy_output	(	const GumboOptions *	options,
		GumboOutput *	output
	)

Release the memory used for the parse tree & parse errors.

gumbo_normalize_svg_tagname()

const char * gumbo_normalize_svg_tagname

(

const GumboStringPiece *

tagname

)

Fixes the case of SVG elements that are not all lowercase. http://www.whatwg.org/specs/web-apps/current-work/multipage/tree-construction.html#parsing-main-inforeign This is not done at parse time because there's no place to store a mutated tag name. tag_name is an enum (which will be TAG_UNKNOWN for most SVG tags without special handling), while original_tag_name is a pointer into the original buffer. Instead, we provide this helper function that clients can use to rename SVG tags as appropriate. Returns the case-normalized SVG tagname if a replacement is found, or NULL if no normalization is called for. The return value is static data and owned by the library.

gumbo_parse()

GumboOutput * gumbo_parse

(

const char *

buffer

)

Parses a buffer of UTF8 text into an GumboNode parse tree. The buffer must live at least as long as the parse tree, as some fields (eg. original_text) point directly into the original buffer.

This doesn't support buffers longer than 4 gigabytes.

gumbo_parse_with_options()

GumboOutput * gumbo_parse_with_options	(	const GumboOptions *	options,
		const char *	buffer,
		size_t	buffer_length
	)

Extended version of gumbo_parse that takes an explicit options structure, buffer, and length.

Variable Documentation

kGumboDefaultOptions

const GumboOptions kGumboDefaultOptions

Initial value:

= {&malloc_wrapper, &free_wrapper, NULL,
    8, false, -1, GUMBO_TAG_LAST, GUMBO_NAMESPACE_HTML}

Default options struct; use this with gumbo_parse_with_options.