123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382 |
- var Tokenizer = require("./Tokenizer.js");
- /*
- Options:
- xmlMode: Disables the special behavior for script/style tags (false by default)
- lowerCaseAttributeNames: call .toLowerCase for each attribute name (true if xmlMode is `false`)
- lowerCaseTags: call .toLowerCase for each tag name (true if xmlMode is `false`)
- */
- /*
- Callbacks:
- oncdataend,
- oncdatastart,
- onclosetag,
- oncomment,
- oncommentend,
- onerror,
- onopentag,
- onprocessinginstruction,
- onreset,
- ontext
- */
- var formTags = {
- input: true,
- option: true,
- optgroup: true,
- select: true,
- button: true,
- datalist: true,
- textarea: true
- };
- var openImpliesClose = {
- tr: { tr: true, th: true, td: true },
- th: { th: true },
- td: { thead: true, th: true, td: true },
- body: { head: true, link: true, script: true },
- li: { li: true },
- p: { p: true },
- h1: { p: true },
- h2: { p: true },
- h3: { p: true },
- h4: { p: true },
- h5: { p: true },
- h6: { p: true },
- select: formTags,
- input: formTags,
- output: formTags,
- button: formTags,
- datalist: formTags,
- textarea: formTags,
- option: { option: true },
- optgroup: { optgroup: true }
- };
- var voidElements = {
- __proto__: null,
- area: true,
- base: true,
- basefont: true,
- br: true,
- col: true,
- command: true,
- embed: true,
- frame: true,
- hr: true,
- img: true,
- input: true,
- isindex: true,
- keygen: true,
- link: true,
- meta: true,
- param: true,
- source: true,
- track: true,
- wbr: true
- };
- var foreignContextElements = {
- __proto__: null,
- math: true,
- svg: true
- };
- var htmlIntegrationElements = {
- __proto__: null,
- mi: true,
- mo: true,
- mn: true,
- ms: true,
- mtext: true,
- "annotation-xml": true,
- foreignObject: true,
- desc: true,
- title: true
- };
- var re_nameEnd = /\s|\//;
- function Parser(cbs, options) {
- this._options = options || {};
- this._cbs = cbs || {};
- this._tagname = "";
- this._attribname = "";
- this._attribvalue = "";
- this._attribs = null;
- this._stack = [];
- this._foreignContext = [];
- this.startIndex = 0;
- this.endIndex = null;
- this._lowerCaseTagNames =
- "lowerCaseTags" in this._options
- ? !!this._options.lowerCaseTags
- : !this._options.xmlMode;
- this._lowerCaseAttributeNames =
- "lowerCaseAttributeNames" in this._options
- ? !!this._options.lowerCaseAttributeNames
- : !this._options.xmlMode;
- if (this._options.Tokenizer) {
- Tokenizer = this._options.Tokenizer;
- }
- this._tokenizer = new Tokenizer(this._options, this);
- if (this._cbs.onparserinit) this._cbs.onparserinit(this);
- }
- require("inherits")(Parser, require("events").EventEmitter);
- Parser.prototype._updatePosition = function(initialOffset) {
- if (this.endIndex === null) {
- if (this._tokenizer._sectionStart <= initialOffset) {
- this.startIndex = 0;
- } else {
- this.startIndex = this._tokenizer._sectionStart - initialOffset;
- }
- } else this.startIndex = this.endIndex + 1;
- this.endIndex = this._tokenizer.getAbsoluteIndex();
- };
- //Tokenizer event handlers
- Parser.prototype.ontext = function(data) {
- this._updatePosition(1);
- this.endIndex--;
- if (this._cbs.ontext) this._cbs.ontext(data);
- };
- Parser.prototype.onopentagname = function(name) {
- if (this._lowerCaseTagNames) {
- name = name.toLowerCase();
- }
- this._tagname = name;
- if (!this._options.xmlMode && name in openImpliesClose) {
- for (
- var el;
- (el = this._stack[this._stack.length - 1]) in
- openImpliesClose[name];
- this.onclosetag(el)
- );
- }
- if (this._options.xmlMode || !(name in voidElements)) {
- this._stack.push(name);
- if (name in foreignContextElements) this._foreignContext.push(true);
- else if (name in htmlIntegrationElements)
- this._foreignContext.push(false);
- }
- if (this._cbs.onopentagname) this._cbs.onopentagname(name);
- if (this._cbs.onopentag) this._attribs = {};
- };
- Parser.prototype.onopentagend = function() {
- this._updatePosition(1);
- if (this._attribs) {
- if (this._cbs.onopentag)
- this._cbs.onopentag(this._tagname, this._attribs);
- this._attribs = null;
- }
- if (
- !this._options.xmlMode &&
- this._cbs.onclosetag &&
- this._tagname in voidElements
- ) {
- this._cbs.onclosetag(this._tagname);
- }
- this._tagname = "";
- };
- Parser.prototype.onclosetag = function(name) {
- this._updatePosition(1);
- if (this._lowerCaseTagNames) {
- name = name.toLowerCase();
- }
-
- if (name in foreignContextElements || name in htmlIntegrationElements) {
- this._foreignContext.pop();
- }
- if (
- this._stack.length &&
- (!(name in voidElements) || this._options.xmlMode)
- ) {
- var pos = this._stack.lastIndexOf(name);
- if (pos !== -1) {
- if (this._cbs.onclosetag) {
- pos = this._stack.length - pos;
- while (pos--) this._cbs.onclosetag(this._stack.pop());
- } else this._stack.length = pos;
- } else if (name === "p" && !this._options.xmlMode) {
- this.onopentagname(name);
- this._closeCurrentTag();
- }
- } else if (!this._options.xmlMode && (name === "br" || name === "p")) {
- this.onopentagname(name);
- this._closeCurrentTag();
- }
- };
- Parser.prototype.onselfclosingtag = function() {
- if (
- this._options.xmlMode ||
- this._options.recognizeSelfClosing ||
- this._foreignContext[this._foreignContext.length - 1]
- ) {
- this._closeCurrentTag();
- } else {
- this.onopentagend();
- }
- };
- Parser.prototype._closeCurrentTag = function() {
- var name = this._tagname;
- this.onopentagend();
- //self-closing tags will be on the top of the stack
- //(cheaper check than in onclosetag)
- if (this._stack[this._stack.length - 1] === name) {
- if (this._cbs.onclosetag) {
- this._cbs.onclosetag(name);
- }
- this._stack.pop();
-
- }
- };
- Parser.prototype.onattribname = function(name) {
- if (this._lowerCaseAttributeNames) {
- name = name.toLowerCase();
- }
- this._attribname = name;
- };
- Parser.prototype.onattribdata = function(value) {
- this._attribvalue += value;
- };
- Parser.prototype.onattribend = function() {
- if (this._cbs.onattribute)
- this._cbs.onattribute(this._attribname, this._attribvalue);
- if (
- this._attribs &&
- !Object.prototype.hasOwnProperty.call(this._attribs, this._attribname)
- ) {
- this._attribs[this._attribname] = this._attribvalue;
- }
- this._attribname = "";
- this._attribvalue = "";
- };
- Parser.prototype._getInstructionName = function(value) {
- var idx = value.search(re_nameEnd),
- name = idx < 0 ? value : value.substr(0, idx);
- if (this._lowerCaseTagNames) {
- name = name.toLowerCase();
- }
- return name;
- };
- Parser.prototype.ondeclaration = function(value) {
- if (this._cbs.onprocessinginstruction) {
- var name = this._getInstructionName(value);
- this._cbs.onprocessinginstruction("!" + name, "!" + value);
- }
- };
- Parser.prototype.onprocessinginstruction = function(value) {
- if (this._cbs.onprocessinginstruction) {
- var name = this._getInstructionName(value);
- this._cbs.onprocessinginstruction("?" + name, "?" + value);
- }
- };
- Parser.prototype.oncomment = function(value) {
- this._updatePosition(4);
- if (this._cbs.oncomment) this._cbs.oncomment(value);
- if (this._cbs.oncommentend) this._cbs.oncommentend();
- };
- Parser.prototype.oncdata = function(value) {
- this._updatePosition(1);
- if (this._options.xmlMode || this._options.recognizeCDATA) {
- if (this._cbs.oncdatastart) this._cbs.oncdatastart();
- if (this._cbs.ontext) this._cbs.ontext(value);
- if (this._cbs.oncdataend) this._cbs.oncdataend();
- } else {
- this.oncomment("[CDATA[" + value + "]]");
- }
- };
- Parser.prototype.onerror = function(err) {
- if (this._cbs.onerror) this._cbs.onerror(err);
- };
- Parser.prototype.onend = function() {
- if (this._cbs.onclosetag) {
- for (
- var i = this._stack.length;
- i > 0;
- this._cbs.onclosetag(this._stack[--i])
- );
- }
- if (this._cbs.onend) this._cbs.onend();
- };
- //Resets the parser to a blank state, ready to parse a new HTML document
- Parser.prototype.reset = function() {
- if (this._cbs.onreset) this._cbs.onreset();
- this._tokenizer.reset();
- this._tagname = "";
- this._attribname = "";
- this._attribs = null;
- this._stack = [];
- if (this._cbs.onparserinit) this._cbs.onparserinit(this);
- };
- //Parses a complete HTML document and pushes it to the handler
- Parser.prototype.parseComplete = function(data) {
- this.reset();
- this.end(data);
- };
- Parser.prototype.write = function(chunk) {
- this._tokenizer.write(chunk);
- };
- Parser.prototype.end = function(chunk) {
- this._tokenizer.end(chunk);
- };
- Parser.prototype.pause = function() {
- this._tokenizer.pause();
- };
- Parser.prototype.resume = function() {
- this._tokenizer.resume();
- };
- //alias for backwards compat
- Parser.prototype.parseChunk = Parser.prototype.write;
- Parser.prototype.done = Parser.prototype.end;
- module.exports = Parser;
|