| 1 | define(["dojo/_base/kernel", "./entities", "dojo/_base/array", "dojo/_base/window", "dojo/_base/sniff"], |
|---|
| 2 | function(lang, Entities, ArrayUtil, Window, has) { |
|---|
| 3 | var dhf = lang.getObject("dojox.html.format",true); |
|---|
| 4 | |
|---|
| 5 | dhf.prettyPrint = function(html/*String*/, indentBy /*Integer?*/, maxLineLength /*Integer?*/, map/*Array?*/, /*boolean*/ xhtml){ |
|---|
| 6 | // summary: |
|---|
| 7 | // Function for providing a 'pretty print' version of HTML content from |
|---|
| 8 | // the provided string. It's nor perfect by any means, but it does |
|---|
| 9 | // a 'reasonable job'. |
|---|
| 10 | // html: String |
|---|
| 11 | // The string of HTML to try and generate a 'pretty' formatting. |
|---|
| 12 | // indentBy: Integer |
|---|
| 13 | // Optional input for the number of spaces to use when indenting. |
|---|
| 14 | // If not defined, zero, negative, or greater than 10, will just use tab |
|---|
| 15 | // as the indent. |
|---|
| 16 | // maxLineLength: Integer |
|---|
| 17 | // Optional input for the number of characters a text line should use in |
|---|
| 18 | // the document, including the indent if possible. |
|---|
| 19 | // map: Array |
|---|
| 20 | // Optional array of entity mapping characters to use when processing the |
|---|
| 21 | // HTML Text content. By default it uses the default set used by the |
|---|
| 22 | // dojox.html.entities.encode function. |
|---|
| 23 | // xhtml: boolean |
|---|
| 24 | // Optional parameter that declares that the returned HTML should try to be 'xhtml' compatible. |
|---|
| 25 | // This means normally unclosed tags are terminated with /> instead of >. Example: `<hr>` -> `<hr />` |
|---|
| 26 | var content = []; |
|---|
| 27 | var indentDepth = 0; |
|---|
| 28 | var closeTags = []; |
|---|
| 29 | var iTxt = "\t"; |
|---|
| 30 | var textContent = ""; |
|---|
| 31 | var inlineStyle = []; |
|---|
| 32 | var i; |
|---|
| 33 | |
|---|
| 34 | // Compile regexps once for this call. |
|---|
| 35 | var rgxp_fixIEAttrs = /[=]([^"']+?)(\s|>)/g; |
|---|
| 36 | var rgxp_styleMatch = /style=("[^"]*"|'[^']*'|\S*)/gi; |
|---|
| 37 | var rgxp_attrsMatch = /[\w-]+=("[^"]*"|'[^']*'|\S*)/gi; |
|---|
| 38 | |
|---|
| 39 | // Check to see if we want to use spaces for indent instead |
|---|
| 40 | // of tab. |
|---|
| 41 | if(indentBy && indentBy > 0 && indentBy < 10){ |
|---|
| 42 | iTxt = ""; |
|---|
| 43 | for(i = 0; i < indentBy; i++){ |
|---|
| 44 | iTxt += " "; |
|---|
| 45 | } |
|---|
| 46 | } |
|---|
| 47 | |
|---|
| 48 | //Build the content outside of the editor so we can walk |
|---|
| 49 | //via DOM and build a 'pretty' output. |
|---|
| 50 | var contentDiv = Window.doc.createElement("div"); |
|---|
| 51 | contentDiv.innerHTML = html; |
|---|
| 52 | |
|---|
| 53 | // Use the entity encode/decode functions, they cache on the map, |
|---|
| 54 | // so it won't multiprocess a map. |
|---|
| 55 | var encode = Entities.encode; |
|---|
| 56 | var decode = Entities.decode; |
|---|
| 57 | |
|---|
| 58 | /** Define a bunch of formatters to format the output. **/ |
|---|
| 59 | var isInlineFormat = function(tag){ |
|---|
| 60 | // summary: |
|---|
| 61 | // Function to determine if the current tag is an inline |
|---|
| 62 | // element that does formatting, as we don't want to |
|---|
| 63 | // break/indent around it, as it can screw up text. |
|---|
| 64 | // tag: |
|---|
| 65 | // The tag to examine |
|---|
| 66 | switch(tag){ |
|---|
| 67 | case "a": |
|---|
| 68 | case "b": |
|---|
| 69 | case "strong": |
|---|
| 70 | case "s": |
|---|
| 71 | case "strike": |
|---|
| 72 | case "i": |
|---|
| 73 | case "u": |
|---|
| 74 | case "em": |
|---|
| 75 | case "sup": |
|---|
| 76 | case "sub": |
|---|
| 77 | case "span": |
|---|
| 78 | case "font": |
|---|
| 79 | case "big": |
|---|
| 80 | case "cite": |
|---|
| 81 | case "q": |
|---|
| 82 | case "small": |
|---|
| 83 | return true; |
|---|
| 84 | default: |
|---|
| 85 | return false; |
|---|
| 86 | } |
|---|
| 87 | }; |
|---|
| 88 | |
|---|
| 89 | //Create less divs. |
|---|
| 90 | var div = contentDiv.ownerDocument.createElement("div"); |
|---|
| 91 | var outerHTML = function(node){ |
|---|
| 92 | // summary: |
|---|
| 93 | // Function to return the outer HTML of a node. |
|---|
| 94 | // Yes, IE has a function like this, but using cloneNode |
|---|
| 95 | // allows avoiding looking at any child nodes, because in this |
|---|
| 96 | // case, we don't want them. |
|---|
| 97 | var clone = node.cloneNode(false); |
|---|
| 98 | div.appendChild(clone); |
|---|
| 99 | var html = div.innerHTML; |
|---|
| 100 | div.innerHTML = ""; |
|---|
| 101 | return html; |
|---|
| 102 | }; |
|---|
| 103 | |
|---|
| 104 | var sizeIndent = function(){ |
|---|
| 105 | var i, txt = ""; |
|---|
| 106 | for(i = 0; i < indentDepth; i++){ |
|---|
| 107 | txt += iTxt; |
|---|
| 108 | } |
|---|
| 109 | return txt.length; |
|---|
| 110 | } |
|---|
| 111 | |
|---|
| 112 | var indent = function(){ |
|---|
| 113 | // summary: |
|---|
| 114 | // Function to handle indent depth. |
|---|
| 115 | var i; |
|---|
| 116 | for(i = 0; i < indentDepth; i++){ |
|---|
| 117 | content.push(iTxt); |
|---|
| 118 | } |
|---|
| 119 | }; |
|---|
| 120 | var newline = function(){ |
|---|
| 121 | // summary: |
|---|
| 122 | // Function to handle newlining. |
|---|
| 123 | content.push("\n"); |
|---|
| 124 | }; |
|---|
| 125 | |
|---|
| 126 | var processTextNode = function(n){ |
|---|
| 127 | // summary: |
|---|
| 128 | // Function to process the text content for doc |
|---|
| 129 | // insertion |
|---|
| 130 | // n: |
|---|
| 131 | // The text node to process. |
|---|
| 132 | textContent += encode(n.nodeValue, map); |
|---|
| 133 | }; |
|---|
| 134 | |
|---|
| 135 | var formatText = function(txt){ |
|---|
| 136 | // summary: |
|---|
| 137 | // Function for processing the text content encountered up to a |
|---|
| 138 | // point and inserting it into the formatted document output. |
|---|
| 139 | // txt: |
|---|
| 140 | // The text to format. |
|---|
| 141 | var i; |
|---|
| 142 | var _iTxt; |
|---|
| 143 | |
|---|
| 144 | // Clean up any indention organization since we're going to rework it |
|---|
| 145 | // anyway. |
|---|
| 146 | var _lines = txt.split("\n"); |
|---|
| 147 | for(i = 0; i < _lines.length; i++){ |
|---|
| 148 | _lines[i] = lang.trim(_lines[i]); |
|---|
| 149 | } |
|---|
| 150 | txt = _lines.join(" "); |
|---|
| 151 | txt = lang.trim(txt); |
|---|
| 152 | if(txt !== ""){ |
|---|
| 153 | var lines = []; |
|---|
| 154 | if(maxLineLength && maxLineLength > 0){ |
|---|
| 155 | var indentSize = sizeIndent(); |
|---|
| 156 | var maxLine = maxLineLength; |
|---|
| 157 | if(maxLineLength > indentSize){ |
|---|
| 158 | maxLine -= indentSize; |
|---|
| 159 | } |
|---|
| 160 | while(txt){ |
|---|
| 161 | if(txt.length > maxLineLength){ |
|---|
| 162 | for(i = maxLine; (i > 0 && txt.charAt(i) !== " "); i--){ |
|---|
| 163 | // Do nothing, we're just looking for a space to split at. |
|---|
| 164 | } |
|---|
| 165 | if(!i){ |
|---|
| 166 | // Couldn't find a split going back, so go forward. |
|---|
| 167 | for(i = maxLine; (i < txt.length && txt.charAt(i) !== " "); i++){ |
|---|
| 168 | // Do nothing, we're just looking for a space to split at. |
|---|
| 169 | } |
|---|
| 170 | } |
|---|
| 171 | var line = txt.substring(0, i); |
|---|
| 172 | line = lang.trim(line); |
|---|
| 173 | // Shift up the text string to the next chunk. |
|---|
| 174 | txt = lang.trim(txt.substring((i == txt.length)?txt.length:i + 1, txt.length)); |
|---|
| 175 | if(line){ |
|---|
| 176 | _iTxt = ""; |
|---|
| 177 | for(i = 0; i < indentDepth; i++){ |
|---|
| 178 | _iTxt += iTxt; |
|---|
| 179 | } |
|---|
| 180 | line = _iTxt + line + "\n"; |
|---|
| 181 | } |
|---|
| 182 | lines.push(line); |
|---|
| 183 | }else{ |
|---|
| 184 | // Line is shorter than out desired length, so use it. |
|---|
| 185 | // as/is |
|---|
| 186 | _iTxt = ""; |
|---|
| 187 | for(i = 0; i < indentDepth; i++){ |
|---|
| 188 | _iTxt += iTxt; |
|---|
| 189 | } |
|---|
| 190 | txt = _iTxt + txt + "\n"; |
|---|
| 191 | lines.push(txt); |
|---|
| 192 | txt = null; |
|---|
| 193 | } |
|---|
| 194 | } |
|---|
| 195 | return lines.join(""); |
|---|
| 196 | }else{ |
|---|
| 197 | _iTxt = ""; |
|---|
| 198 | for(i = 0; i < indentDepth; i++){ |
|---|
| 199 | _iTxt += iTxt; |
|---|
| 200 | } |
|---|
| 201 | txt = _iTxt + txt + "\n"; |
|---|
| 202 | return txt; |
|---|
| 203 | } |
|---|
| 204 | }else{ |
|---|
| 205 | return ""; |
|---|
| 206 | } |
|---|
| 207 | }; |
|---|
| 208 | |
|---|
| 209 | var processScriptText = function(txt){ |
|---|
| 210 | // summary: |
|---|
| 211 | // Function to clean up potential escapes in the script code. |
|---|
| 212 | if(txt){ |
|---|
| 213 | txt = txt.replace(/"/gi, "\""); |
|---|
| 214 | txt = txt.replace(/>/gi, ">"); |
|---|
| 215 | txt = txt.replace(/</gi, "<"); |
|---|
| 216 | txt = txt.replace(/&/gi, "&"); |
|---|
| 217 | } |
|---|
| 218 | return txt; |
|---|
| 219 | }; |
|---|
| 220 | |
|---|
| 221 | var formatScript = function(txt){ |
|---|
| 222 | // summary: |
|---|
| 223 | // Function to rudimentary formatting of script text. |
|---|
| 224 | // Not perfect, but it helps get some level of organization |
|---|
| 225 | // in there. |
|---|
| 226 | // txt: |
|---|
| 227 | // The script text to try to format a bit. |
|---|
| 228 | if(txt){ |
|---|
| 229 | txt = processScriptText(txt); |
|---|
| 230 | var i, t, c, _iTxt; |
|---|
| 231 | var indent = 0; |
|---|
| 232 | var scriptLines = txt.split("\n"); |
|---|
| 233 | var newLines = []; |
|---|
| 234 | for (i = 0; i < scriptLines.length; i++){ |
|---|
| 235 | var line = scriptLines[i]; |
|---|
| 236 | var hasNewlines = (line.indexOf("\n") > -1); |
|---|
| 237 | line = lang.trim(line); |
|---|
| 238 | if(line){ |
|---|
| 239 | var iLevel = indent; |
|---|
| 240 | // Not all blank, so we need to process. |
|---|
| 241 | for(c = 0; c < line.length; c++){ |
|---|
| 242 | var ch = line.charAt(c); |
|---|
| 243 | if(ch === "{"){ |
|---|
| 244 | indent++; |
|---|
| 245 | }else if(ch === "}"){ |
|---|
| 246 | indent--; |
|---|
| 247 | // We want to back up a bit before the |
|---|
| 248 | // line is written. |
|---|
| 249 | iLevel = indent; |
|---|
| 250 | } |
|---|
| 251 | } |
|---|
| 252 | _iTxt = ""; |
|---|
| 253 | for(t = 0; t < indentDepth + iLevel; t++){ |
|---|
| 254 | _iTxt += iTxt; |
|---|
| 255 | } |
|---|
| 256 | newLines.push(_iTxt + line + "\n"); |
|---|
| 257 | }else if(hasNewlines && i === 0){ |
|---|
| 258 | // Just insert a newline for blank lines as |
|---|
| 259 | // long as it's not the first newline (we |
|---|
| 260 | // already inserted that in the openTag handler) |
|---|
| 261 | newLines.push("\n"); |
|---|
| 262 | } |
|---|
| 263 | |
|---|
| 264 | } |
|---|
| 265 | // Okay, create the script text, hopefully reasonably |
|---|
| 266 | // formatted. |
|---|
| 267 | txt = newLines.join(""); |
|---|
| 268 | } |
|---|
| 269 | return txt; |
|---|
| 270 | }; |
|---|
| 271 | |
|---|
| 272 | var openTag = function(node){ |
|---|
| 273 | // summary: |
|---|
| 274 | // Function to open a new tag for writing content. |
|---|
| 275 | var name = node.nodeName.toLowerCase(); |
|---|
| 276 | // Generate the outer node content (tag with attrs) |
|---|
| 277 | var nText = lang.trim(outerHTML(node)); |
|---|
| 278 | var tag = nText.substring(0, nText.indexOf(">") + 1); |
|---|
| 279 | |
|---|
| 280 | // Also thanks to IE, we need to check for quotes around |
|---|
| 281 | // attributes and insert if missing. |
|---|
| 282 | tag = tag.replace(rgxp_fixIEAttrs,'="$1"$2'); |
|---|
| 283 | |
|---|
| 284 | // And lastly, thanks IE for changing style casing and end |
|---|
| 285 | // semi-colon and webkit adds spaces, so lets clean it up by |
|---|
| 286 | // sorting, etc, while we're at it. |
|---|
| 287 | tag = tag.replace(rgxp_styleMatch, function(match){ |
|---|
| 288 | var sL = match.substring(0,6); |
|---|
| 289 | var style = match.substring(6, match.length); |
|---|
| 290 | var closure = style.charAt(0); |
|---|
| 291 | style = lang.trim(style.substring(1,style.length -1)); |
|---|
| 292 | style = style.split(";"); |
|---|
| 293 | var trimmedStyles = []; |
|---|
| 294 | ArrayUtil.forEach(style, function(s){ |
|---|
| 295 | s = lang.trim(s); |
|---|
| 296 | if(s){ |
|---|
| 297 | // Lower case the style name, leave the value alone. Mainly a fixup for IE. |
|---|
| 298 | s = s.substring(0, s.indexOf(":")).toLowerCase() + s.substring(s.indexOf(":"), s.length); |
|---|
| 299 | trimmedStyles.push(s); |
|---|
| 300 | } |
|---|
| 301 | }); |
|---|
| 302 | trimmedStyles = trimmedStyles.sort(); |
|---|
| 303 | |
|---|
| 304 | // Reassemble and return the styles in sorted order. |
|---|
| 305 | style = trimmedStyles.join("; "); |
|---|
| 306 | var ts = lang.trim(style); |
|---|
| 307 | if(!ts || ts === ";"){ |
|---|
| 308 | // Just remove any style attrs that are empty. |
|---|
| 309 | return ""; |
|---|
| 310 | }else{ |
|---|
| 311 | style += ";"; |
|---|
| 312 | return sL + closure + style + closure; |
|---|
| 313 | } |
|---|
| 314 | }); |
|---|
| 315 | |
|---|
| 316 | // Try and sort the attributes while we're at it. |
|---|
| 317 | var attrs = []; |
|---|
| 318 | tag = tag.replace(rgxp_attrsMatch, function(attr){ |
|---|
| 319 | attrs.push(lang.trim(attr)); |
|---|
| 320 | return ""; |
|---|
| 321 | }); |
|---|
| 322 | attrs = attrs.sort(); |
|---|
| 323 | |
|---|
| 324 | // Reassemble the tag with sorted attributes! |
|---|
| 325 | tag = "<" + name; |
|---|
| 326 | if(attrs.length){ |
|---|
| 327 | tag += " " + attrs.join(" "); |
|---|
| 328 | } |
|---|
| 329 | |
|---|
| 330 | // Determine closure status. If xhtml, |
|---|
| 331 | // then close the tag properly as needed. |
|---|
| 332 | if(nText.indexOf("</") != -1){ |
|---|
| 333 | closeTags.push(name); |
|---|
| 334 | tag += ">"; |
|---|
| 335 | }else{ |
|---|
| 336 | if(xhtml){ |
|---|
| 337 | tag += " />"; |
|---|
| 338 | }else{ |
|---|
| 339 | tag += ">"; |
|---|
| 340 | } |
|---|
| 341 | closeTags.push(false); |
|---|
| 342 | } |
|---|
| 343 | |
|---|
| 344 | var inline = isInlineFormat(name); |
|---|
| 345 | inlineStyle.push(inline); |
|---|
| 346 | if(textContent && !inline){ |
|---|
| 347 | // Process any text content we have that occurred |
|---|
| 348 | // before the open tag of a non-inline. |
|---|
| 349 | content.push(formatText(textContent)); |
|---|
| 350 | textContent = ""; |
|---|
| 351 | } |
|---|
| 352 | |
|---|
| 353 | // Determine if this has a closing tag or not! |
|---|
| 354 | if(!inline){ |
|---|
| 355 | indent(); |
|---|
| 356 | content.push(tag); |
|---|
| 357 | newline(); |
|---|
| 358 | indentDepth++; |
|---|
| 359 | }else{ |
|---|
| 360 | textContent += tag; |
|---|
| 361 | } |
|---|
| 362 | |
|---|
| 363 | }; |
|---|
| 364 | |
|---|
| 365 | var closeTag = function(){ |
|---|
| 366 | // summary: |
|---|
| 367 | // Function to close out a tag if necessary. |
|---|
| 368 | var inline = inlineStyle.pop(); |
|---|
| 369 | if(textContent && !inline){ |
|---|
| 370 | // Process any text content we have that occurred |
|---|
| 371 | // before the close tag. |
|---|
| 372 | content.push(formatText(textContent)); |
|---|
| 373 | textContent = ""; |
|---|
| 374 | } |
|---|
| 375 | var ct = closeTags.pop(); |
|---|
| 376 | if(ct){ |
|---|
| 377 | ct = "</" + ct + ">"; |
|---|
| 378 | if(!inline){ |
|---|
| 379 | indentDepth--; |
|---|
| 380 | indent(); |
|---|
| 381 | content.push(ct); |
|---|
| 382 | newline(); |
|---|
| 383 | }else{ |
|---|
| 384 | textContent += ct; |
|---|
| 385 | } |
|---|
| 386 | }else{ |
|---|
| 387 | indentDepth--; |
|---|
| 388 | } |
|---|
| 389 | }; |
|---|
| 390 | |
|---|
| 391 | var processCommentNode = function(n){ |
|---|
| 392 | // summary: |
|---|
| 393 | // Function to handle processing a comment node. |
|---|
| 394 | // n: |
|---|
| 395 | // The comment node to process. |
|---|
| 396 | |
|---|
| 397 | //Make sure contents aren't double-encoded. |
|---|
| 398 | var commentText = decode(n.nodeValue, map); |
|---|
| 399 | indent(); |
|---|
| 400 | content.push("<!--"); |
|---|
| 401 | newline(); |
|---|
| 402 | indentDepth++; |
|---|
| 403 | content.push(formatText(commentText)); |
|---|
| 404 | indentDepth--; |
|---|
| 405 | indent(); |
|---|
| 406 | content.push("-->"); |
|---|
| 407 | newline(); |
|---|
| 408 | }; |
|---|
| 409 | |
|---|
| 410 | var processNode = function(node) { |
|---|
| 411 | // summary: |
|---|
| 412 | // Entrypoint for processing all the text! |
|---|
| 413 | var children = node.childNodes; |
|---|
| 414 | if(children){ |
|---|
| 415 | var i; |
|---|
| 416 | for(i = 0; i < children.length; i++){ |
|---|
| 417 | var n = children[i]; |
|---|
| 418 | if(n.nodeType === 1){ |
|---|
| 419 | var tg = lang.trim(n.tagName.toLowerCase()); |
|---|
| 420 | if(has("ie") && n.parentNode != node){ |
|---|
| 421 | // IE is broken. DOMs are supposed to be a tree. |
|---|
| 422 | // But in the case of malformed HTML, IE generates a graph |
|---|
| 423 | // meaning one node ends up with multiple references |
|---|
| 424 | // (multiple parents). This is totally wrong and invalid, but |
|---|
| 425 | // such is what it is. We have to keep track and check for |
|---|
| 426 | // this because otherwise the source output HTML will have dups. |
|---|
| 427 | continue; |
|---|
| 428 | } |
|---|
| 429 | if(tg && tg.charAt(0) === "/"){ |
|---|
| 430 | // IE oddity. Malformed HTML can put in odd tags like: |
|---|
| 431 | // </ >, </span>. It treats a mismatched closure as a new |
|---|
| 432 | // start tag. So, remove them. |
|---|
| 433 | continue; |
|---|
| 434 | }else{ |
|---|
| 435 | //Process non-dup, seemingly wellformed elements! |
|---|
| 436 | openTag(n); |
|---|
| 437 | if(tg === "script"){ |
|---|
| 438 | content.push(formatScript(n.innerHTML)); |
|---|
| 439 | }else if(tg === "pre"){ |
|---|
| 440 | var preTxt = n.innerHTML; |
|---|
| 441 | if(has("mozilla")){ |
|---|
| 442 | //Mozilla screws this up, so fix it up. |
|---|
| 443 | preTxt = preTxt.replace("<br>", "\n"); |
|---|
| 444 | preTxt = preTxt.replace("<pre>", ""); |
|---|
| 445 | preTxt = preTxt.replace("</pre>", ""); |
|---|
| 446 | } |
|---|
| 447 | // Add ending newline, if needed. |
|---|
| 448 | if(preTxt.charAt(preTxt.length - 1) !== "\n"){ |
|---|
| 449 | preTxt += "\n"; |
|---|
| 450 | } |
|---|
| 451 | content.push(preTxt); |
|---|
| 452 | }else{ |
|---|
| 453 | processNode(n); |
|---|
| 454 | } |
|---|
| 455 | closeTag(); |
|---|
| 456 | } |
|---|
| 457 | }else if(n.nodeType === 3 || n.nodeType === 4){ |
|---|
| 458 | processTextNode(n); |
|---|
| 459 | }else if(n.nodeType === 8){ |
|---|
| 460 | processCommentNode(n); |
|---|
| 461 | } |
|---|
| 462 | } |
|---|
| 463 | } |
|---|
| 464 | }; |
|---|
| 465 | |
|---|
| 466 | //Okay, finally process the input string. |
|---|
| 467 | processNode(contentDiv); |
|---|
| 468 | if(textContent){ |
|---|
| 469 | // Insert any trailing text. See: #10854 |
|---|
| 470 | content.push(formatText(textContent)); |
|---|
| 471 | textContent = ""; |
|---|
| 472 | } |
|---|
| 473 | return content.join(""); //String |
|---|
| 474 | }; |
|---|
| 475 | return dhf; |
|---|
| 476 | }); |
|---|
| 477 | |
|---|
