html2json.js 4.7 KB

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697
  1. function e(e) {
  2. for (var t = {}, r = e.split(","), s = 0; s < r.length; s++) t[r[s]] = !0;
  3. return t;
  4. }
  5. function t(e) {
  6. return e.replace(/<\?xml.*\?>\n/, "").replace(/<.*!doctype.*\>\n/, "").replace(/<.*!DOCTYPE.*\>\n/, "");
  7. }
  8. function r(e) {
  9. var t = [];
  10. if (0 == a.length || !n) return (d = {}).node = "text", d.text = e, s = [ d ];
  11. e = e.replace(/\[([^\[\]]+)\]/g, ":$1:");
  12. for (var r = new RegExp("[:]"), s = e.split(r), i = 0; i < s.length; i++) {
  13. var l = s[i], d = {};
  14. n[l] ? (d.node = "element", d.tag = "emoji", d.text = n[l], d.baseSrc = o) : (d.node = "text",
  15. d.text = l), t.push(d);
  16. }
  17. return t;
  18. }
  19. var s = "https", a = "", o = "", n = {}, i = require("./wxDiscode.js"), l = require("./htmlparser.js"), d = (e("area,base,basefont,br,col,frame,hr,img,input,link,meta,param,embed,command,keygen,source,track,wbr"),
  20. e("br,a,code,address,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")), c = e("abbr,acronym,applet,b,basefont,bdo,big,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"), u = e("colgroup,dd,dt,li,options,p,td,tfoot,th,thead,tr");
  21. e("checked,compact,declare,defer,disabled,ismap,multiple,nohref,noresize,noshade,nowrap,readonly,selected"),
  22. e("wxxxcode-style,script,style,view,scroll-view,block");
  23. module.exports = {
  24. html2json: function(e, a) {
  25. e = t(e), e = i.strDiscode(e);
  26. var o = [], n = {
  27. node: a,
  28. nodes: [],
  29. images: [],
  30. imageUrls: []
  31. };
  32. return l(e, {
  33. start: function(e, t, r) {
  34. var l = {
  35. node: "element",
  36. tag: e
  37. };
  38. if (d[e] ? l.tagType = "block" : c[e] ? l.tagType = "inline" : u[e] && (l.tagType = "closeSelf"),
  39. 0 !== t.length && (l.attr = t.reduce(function(e, t) {
  40. var r = t.name, s = t.value;
  41. return "class" == r && (l.classStr = s), "style" == r && (l.styleStr = s),
  42. s.match(/ /) && (s = s.split(" ")), e[r] ? Array.isArray(e[r]) ? e[r].push(s) : e[r] = [ e[r], s ] : e[r] = s,
  43. e;
  44. }, {})), "img" === l.tag) {
  45. l.imgIndex = n.images.length;
  46. var p = l.attr.src;
  47. p = i.urlToHttpUrl(p, s), l.attr.src = p, l.from = a, n.images.push(l), n.imageUrls.push(p);
  48. }
  49. if ("font" === l.tag) {
  50. var m = [ "x-small", "small", "medium", "large", "x-large", "xx-large", "-webkit-xxx-large" ], f = {
  51. color: "color",
  52. face: "font-family",
  53. size: "font-size"
  54. };
  55. l.attr.style || (l.attr.style = []), l.styleStr || (l.styleStr = "");
  56. for (var h in f) if (l.attr[h]) {
  57. var g = "size" === h ? m[l.attr[h] - 1] : l.attr[h];
  58. l.attr.style.push(f[h]), l.attr.style.push(g), l.styleStr += f[h] + ": " + g + ";";
  59. }
  60. }
  61. if ("source" === l.tag && (n.source = l.attr.src), r) {
  62. var v = o[0] || n;
  63. void 0 === v.nodes && (v.nodes = []), v.nodes.push(l);
  64. } else o.unshift(l);
  65. },
  66. end: function(e) {
  67. var t = o.shift();
  68. if (t.tag !== e && console.error("invalid state: mismatch end tag"), "video" === t.tag && n.source && (t.attr.src = n.source,
  69. delete result.source), 0 === o.length) n.nodes.push(t); else {
  70. var r = o[0];
  71. void 0 === r.nodes && (r.nodes = []), r.nodes.push(t);
  72. }
  73. },
  74. chars: function(e) {
  75. var t = {
  76. node: "text",
  77. text: e,
  78. textArray: r(e)
  79. };
  80. if (0 === o.length) n.nodes.push(t); else {
  81. var s = o[0];
  82. void 0 === s.nodes && (s.nodes = []), s.nodes.push(t);
  83. }
  84. },
  85. comment: function(e) {}
  86. }), n;
  87. },
  88. emojisInit: function() {
  89. var e = arguments.length > 0 && void 0 !== arguments[0] ? arguments[0] : "", t = arguments.length > 1 && void 0 !== arguments[1] ? arguments[1] : "/wxParse/emojis/", r = arguments[2];
  90. a = e, o = t, n = r;
  91. }
  92. };