htmlparser.js 4.1 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126
  1. const startTag = /^<([-A-Za-z0-9_]+)((?:\s+[a-zA-Z0-9_:][-a-zA-Z0-9_:.]*(?:\s*=\s*(?:(?:"[^"]*")|(?:'[^']*')|[^>\s]+))?)*)\s*(\/?)>/;
  2. const endTag = /^<\/([-A-Za-z0-9_]+)[^>]*>/;
  3. const attr = /([a-zA-Z0-9_:][-a-zA-Z0-9_:.]*)(?:\s*=\s*(?:(?:"((?:\\.|[^"])*)")|(?:'((?:\\.|[^'])*)')|([^>\s]+)))?/g;
  4. function makeMap(str) {
  5. const obj = {};
  6. const items = str.split(',');
  7. for (let i = 0; i < items.length; i += 1) obj[items[i]] = true;
  8. return obj;
  9. }
  10. const empty = makeMap('area,base,basefont,br,col,frame,hr,img,input,link,meta,param,embed,command,keygen,source,track,wbr');
  11. const block = makeMap('address,code,article,applet,aside,audio,blockquote,button,canvas,center,dd,del,dir,div,dl,dt,fieldset,figcaption,figure,footer,form,frameset,h1,h2,h3,h4,h5,h6,header,hgroup,hr,iframe,ins,isindex,li,map,menu,noframes,noscript,object,ol,output,p,pre,section,script,table,tbody,td,tfoot,th,thead,tr,ul,video');
  12. const inline = makeMap('a,abbr,acronym,applet,b,basefont,bdo,big,br,button,cite,del,dfn,em,font,i,iframe,img,input,ins,kbd,label,map,object,q,s,samp,script,select,small,span,strike,strong,sub,sup,textarea,tt,u,var');
  13. const closeSelf = makeMap('colgroup,dd,dt,li,options,p,td,tfoot,th,thead,tr');
  14. const fillAttrs = makeMap('checked,compact,declare,defer,disabled,ismap,multiple,nohref,noresize,noshade,nowrap,readonly,selected');
  15. function HTMLParser(html, handler) {
  16. let index;
  17. let chars;
  18. let match;
  19. let last = html;
  20. const stack = [];
  21. stack.last = () => stack[stack.length - 1];
  22. function parseEndTag(tag, tagName) {
  23. let pos;
  24. if (!tagName) {
  25. pos = 0;
  26. } else {
  27. tagName = tagName.toLowerCase();
  28. for (pos = stack.length - 1; pos >= 0; pos -= 1) {
  29. if (stack[pos] === tagName) break;
  30. }
  31. }
  32. if (pos >= 0) {
  33. for (let i = stack.length - 1; i >= pos; i -= 1) {
  34. if (handler.end) handler.end(stack[i]);
  35. }
  36. stack.length = pos;
  37. }
  38. }
  39. function parseStartTag(tag, tagName, rest, unary) {
  40. tagName = tagName.toLowerCase();
  41. if (block[tagName]) {
  42. while (stack.last() && inline[stack.last()]) {
  43. parseEndTag('', stack.last());
  44. }
  45. }
  46. if (closeSelf[tagName] && stack.last() === tagName) {
  47. parseEndTag('', tagName);
  48. }
  49. unary = empty[tagName] || !!unary;
  50. if (!unary) stack.push(tagName);
  51. if (handler.start) {
  52. const attrs = [];
  53. rest.replace(attr, function genAttr(matches, name) {
  54. const value = arguments[2] || arguments[3] || arguments[4] || (fillAttrs[name] ? name : '');
  55. attrs.push({
  56. name,
  57. value,
  58. escaped: value.replace(/(^|[^\\])"/g, '$1\\"'), // "
  59. });
  60. });
  61. if (handler.start) {
  62. handler.start(tagName, attrs, unary);
  63. }
  64. }
  65. }
  66. while (html) {
  67. chars = true;
  68. if (html.indexOf('</') === 0) {
  69. match = html.match(endTag);
  70. if (match) {
  71. html = html.substring(match[0].length);
  72. match[0].replace(endTag, parseEndTag);
  73. chars = false;
  74. }
  75. } else if (html.indexOf('<') === 0) {
  76. match = html.match(startTag);
  77. if (match) {
  78. html = html.substring(match[0].length);
  79. match[0].replace(startTag, parseStartTag);
  80. chars = false;
  81. }
  82. }
  83. if (chars) {
  84. index = html.indexOf('<');
  85. let text = '';
  86. while (index === 0) {
  87. text += '<';
  88. html = html.substring(1);
  89. index = html.indexOf('<');
  90. }
  91. text += index < 0 ? html : html.substring(0, index);
  92. html = index < 0 ? '' : html.substring(index);
  93. if (handler.chars) handler.chars(text);
  94. }
  95. if (html === last) throw new Error(`Parse Error: ${html}`);
  96. last = html;
  97. }
  98. parseEndTag();
  99. }
  100. export default HTMLParser;