Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 1 | /* |
| 2 | * Copyright (C) 2013 Google Inc. |
| 3 | * |
| 4 | * Licensed under the Apache License, Version 2.0 (the "License"); |
| 5 | * you may not use this file except in compliance with the License. |
| 6 | * You may obtain a copy of the License at |
| 7 | * |
| 8 | * http://www.apache.org/licenses/LICENSE-2.0 |
| 9 | * |
| 10 | * Unless required by applicable law or agreed to in writing, software |
| 11 | * distributed under the License is distributed on an "AS IS" BASIS, |
| 12 | * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 13 | * See the License for the specific language governing permissions and |
| 14 | * limitations under the License. |
| 15 | */ |
| 16 | |
| 17 | package com.google.doclava; |
| 18 | |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 19 | import java.io.*; |
| 20 | import java.text.BreakIterator; |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 21 | import java.util.ArrayList; |
| 22 | import java.util.Collections; |
| 23 | import java.util.Comparator; |
| 24 | import java.util.List; |
| 25 | import java.util.regex.Pattern; |
| 26 | import java.util.regex.Matcher; |
| 27 | import java.io.File; |
| 28 | |
| 29 | import com.google.clearsilver.jsilver.data.Data; |
| 30 | |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 31 | import org.ccil.cowan.tagsoup.*; |
| 32 | import org.xml.sax.XMLReader; |
| 33 | import org.xml.sax.InputSource; |
| 34 | import org.xml.sax.Attributes; |
| 35 | import org.xml.sax.helpers.DefaultHandler; |
| 36 | |
| 37 | import org.w3c.dom.Node; |
| 38 | import org.w3c.dom.NodeList; |
| 39 | |
| 40 | import javax.xml.transform.dom.DOMResult; |
| 41 | import javax.xml.transform.sax.SAXSource; |
| 42 | import javax.xml.transform.Transformer; |
| 43 | import javax.xml.transform.TransformerFactory; |
| 44 | import javax.xml.xpath.XPath; |
| 45 | import javax.xml.xpath.XPathConstants; |
| 46 | import javax.xml.xpath.XPathExpression; |
| 47 | import javax.xml.xpath.XPathFactory; |
| 48 | |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 49 | /** |
| 50 | * Metadata associated with a specific documentation page. Extracts |
| 51 | * metadata based on the page's declared hdf vars (meta.tags and others) |
| 52 | * as well as implicit data relating to the page, such as url, type, etc. |
| 53 | * Includes a Node class that represents the metadata and lets it attach |
| 54 | * to parent/child elements in the tree metadata nodes for all pages. |
| 55 | * Node also includes methods for rendering the node tree to a json file |
| 56 | * in docs output, which is then used by JavaScript to load metadata |
| 57 | * objects into html pages. |
| 58 | */ |
| 59 | |
| 60 | public class PageMetadata { |
| 61 | File mSource; |
| 62 | String mDest; |
| 63 | String mTagList; |
| 64 | static boolean sLowercaseTags = true; |
| 65 | static boolean sLowercaseKeywords = true; |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 66 | //static String linkPrefix = (Doclava.META_DBG) ? "/" : "http://developer.android.com/"; |
| 67 | /** |
| 68 | * regex pattern to match javadoc @link and similar tags. Extracts |
| 69 | * root symbol to $1. |
| 70 | */ |
| 71 | private static final Pattern JD_TAG_PATTERN = |
| 72 | Pattern.compile("\\{@.*?[\\s\\.\\#]([A-Za-z\\(\\)\\d_]+)(?=\u007D)\u007D"); |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 73 | |
| 74 | public PageMetadata(File source, String dest, List<Node> taglist) { |
| 75 | mSource = source; |
| 76 | mDest = dest; |
| 77 | |
| 78 | if (dest != null) { |
| 79 | int len = dest.length(); |
| 80 | if (len > 1 && dest.charAt(len - 1) != '/') { |
| 81 | mDest = dest + '/'; |
| 82 | } else { |
| 83 | mDest = dest; |
| 84 | } |
| 85 | } |
| 86 | } |
| 87 | |
| 88 | /** |
| 89 | * Given a list of metadata nodes organized by type, sort the |
| 90 | * root nodes by type name and render the types and their child |
| 91 | * metadata nodes to a json file in the out dir. |
| 92 | * |
| 93 | * @param rootTypeNodesList A list of root metadata nodes, each |
| 94 | * representing a type and it's member child pages. |
| 95 | */ |
| 96 | public static void WriteList(List<Node> rootTypeNodesList) { |
| 97 | |
| 98 | Collections.sort(rootTypeNodesList, BY_TYPE_NAME); |
| 99 | Node pageMeta = new Node.Builder().setLabel("TOP").setChildren(rootTypeNodesList).build(); |
| 100 | |
| 101 | StringBuilder buf = new StringBuilder(); |
| 102 | // write the taglist to string format |
| 103 | pageMeta.renderTypeResources(buf); |
| 104 | pageMeta.renderTypesByTag(buf); |
| 105 | // write the taglist to js file |
| 106 | Data data = Doclava.makeHDF(); |
| 107 | data.setValue("reference_tree", buf.toString()); |
| 108 | ClearPage.write(data, "jd_lists_unified.cs", "jd_lists_unified.js"); |
| 109 | } |
| 110 | |
| 111 | /** |
| 112 | * Extract supported metadata values from a page and add them as |
| 113 | * a child node of a root node based on type. Some metadata values |
| 114 | * are normalized. Unsupported metadata fields are ignored. See |
| 115 | * Node for supported metadata fields and methods for accessing values. |
| 116 | * |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 117 | * @param docfile The file from which to extract metadata. |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 118 | * @param dest The output path for the file, used to set link to page. |
| 119 | * @param filename The file from which to extract metadata. |
| 120 | * @param hdf Data object in which to store the metadata values. |
| 121 | * @param tagList The file from which to extract metadata. |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 122 | */ |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 123 | public static void setPageMetadata(String docfile, String dest, String filename, |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 124 | Data hdf, List<Node> tagList) { |
| 125 | //exclude this page if author does not want it included |
| 126 | boolean excludeNode = "true".equals(hdf.getValue("excludeFromSuggestions","")); |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 127 | |
| 128 | //check whether summary and image exist and if not, get them from itemprop/markup |
| 129 | Boolean needsSummary = "".equals(hdf.getValue("page.metaDescription", "")); |
| 130 | Boolean needsImage = "".equals(hdf.getValue("page.image", "")); |
| 131 | if ((needsSummary) || (needsImage)) { |
| 132 | //try to extract the metadata from itemprop and markup |
| 133 | inferMetadata(docfile, hdf, needsSummary, needsImage); |
| 134 | } |
| 135 | |
| 136 | //extract available metadata and set it in a node |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 137 | if (!excludeNode) { |
| 138 | Node pageMeta = new Node.Builder().build(); |
| 139 | pageMeta.setLabel(getTitleNormalized(hdf, "page.title")); |
| 140 | pageMeta.setTitleFriendly(hdf.getValue("page.titleFriendly","")); |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 141 | pageMeta.setSummary(hdf.getValue("page.metaDescription","")); |
| 142 | pageMeta.setLink(getPageUrlNormalized(filename)); |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 143 | pageMeta.setGroup(getStringValueNormalized(hdf,"sample.group")); |
| 144 | pageMeta.setKeywords(getPageTagsNormalized(hdf, "page.tags")); |
| 145 | pageMeta.setTags(getPageTagsNormalized(hdf, "meta.tags")); |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 146 | pageMeta.setImage(getImageUrlNormalized(hdf.getValue("page.image", ""))); |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 147 | pageMeta.setLang(getLangStringNormalized(filename)); |
| 148 | pageMeta.setType(getStringValueNormalized(hdf, "page.type")); |
| 149 | appendMetaNodeByType(pageMeta, tagList); |
| 150 | } |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 151 | } |
| 152 | |
| 153 | /** |
| 154 | * Attempt to infer page metadata based on the contents of the |
| 155 | * file. Load and parse the file as a dom tree. Select values |
| 156 | * in this order: 1. dom node specifically tagged with |
| 157 | * microdata (itemprop). 2. first qualitifed p or img node. |
| 158 | * |
| 159 | * @param docfile The file from which to extract metadata. |
| 160 | * @param hdf Data object in which to store the metadata values. |
| 161 | * @param needsSummary Whether to extract summary metadata. |
| 162 | * @param needsImage Whether to extract image metadata. |
| 163 | */ |
| 164 | public static void inferMetadata(String docfile, Data hdf, |
| 165 | Boolean needsSummary, Boolean needsImage) { |
| 166 | String sum = ""; |
| 167 | String imageUrl = ""; |
| 168 | String sumFrom = needsSummary ? "none" : "hdf"; |
| 169 | String imgFrom = needsImage ? "none" : "hdf"; |
| 170 | String filedata = hdf.getValue("commentText", ""); |
| 171 | if (Doclava.META_DBG) System.out.println("----- " + docfile + "\n"); |
| 172 | |
| 173 | try { |
| 174 | XPathFactory xpathFac = XPathFactory.newInstance(); |
| 175 | XPath xpath = xpathFac.newXPath(); |
| 176 | InputStream inputStream = new ByteArrayInputStream(filedata.getBytes()); |
| 177 | XMLReader reader = new Parser(); |
| 178 | reader.setFeature(Parser.namespacesFeature, false); |
| 179 | reader.setFeature(Parser.namespacePrefixesFeature, false); |
| 180 | reader.setFeature(Parser.ignoreBogonsFeature, true); |
| 181 | |
| 182 | Transformer transformer = TransformerFactory.newInstance().newTransformer(); |
| 183 | DOMResult result = new DOMResult(); |
| 184 | transformer.transform(new SAXSource(reader, new InputSource(inputStream)), result); |
| 185 | org.w3c.dom.Node htmlNode = result.getNode(); |
| 186 | |
| 187 | if (needsSummary) { |
| 188 | StringBuilder sumStrings = new StringBuilder(); |
| 189 | XPathExpression ItempropDescExpr = xpath.compile("/descendant-or-self::*" |
| 190 | + "[@itemprop='description'][1]//text()[string(.)]"); |
| 191 | org.w3c.dom.NodeList nodes = (org.w3c.dom.NodeList) ItempropDescExpr.evaluate(htmlNode, |
| 192 | XPathConstants.NODESET); |
| 193 | if (nodes.getLength() > 0) { |
| 194 | for (int i = 0; i < nodes.getLength(); i++) { |
| 195 | String tx = nodes.item(i).getNodeValue(); |
| 196 | sumStrings.append(tx); |
| 197 | sumFrom = "itemprop"; |
| 198 | } |
| 199 | } else { |
| 200 | XPathExpression FirstParaExpr = xpath.compile("//p[not(../../../" |
| 201 | + "@class='notice-developers') and not(../@class='sidebox')" |
| 202 | + "and not(@class)]//text()"); |
| 203 | nodes = (org.w3c.dom.NodeList) FirstParaExpr.evaluate(htmlNode, XPathConstants.NODESET); |
| 204 | if (nodes.getLength() > 0) { |
| 205 | for (int i = 0; i < nodes.getLength(); i++) { |
| 206 | String tx = nodes.item(i).getNodeValue(); |
| 207 | sumStrings.append(tx + " "); |
| 208 | sumFrom = "markup"; |
| 209 | } |
| 210 | } |
| 211 | } |
| 212 | //found a summary string, now normalize it |
| 213 | sum = sumStrings.toString().trim(); |
| 214 | if ((sum != null) && (!"".equals(sum))) { |
| 215 | sum = getSummaryNormalized(sum); |
| 216 | } |
| 217 | //normalized summary ended up being too short to be meaningful |
| 218 | if ("".equals(sum)) { |
| 219 | if (Doclava.META_DBG) System.out.println("Warning: description too short! (" |
| 220 | + sum.length() + "chars) ...\n\n"); |
| 221 | } |
| 222 | //summary looks good, store it to the file hdf data |
| 223 | hdf.setValue("page.metaDescription", sum); |
| 224 | } |
| 225 | if (needsImage) { |
| 226 | XPathExpression ItempropImageExpr = xpath.compile("//*[@itemprop='image']/@src"); |
| 227 | org.w3c.dom.NodeList imgNodes = (org.w3c.dom.NodeList) ItempropImageExpr.evaluate(htmlNode, |
| 228 | XPathConstants.NODESET); |
| 229 | if (imgNodes.getLength() > 0) { |
| 230 | imageUrl = imgNodes.item(0).getNodeValue(); |
| 231 | imgFrom = "itemprop"; |
| 232 | } else { |
| 233 | XPathExpression FirstImgExpr = xpath.compile("//img/@src"); |
| 234 | imgNodes = (org.w3c.dom.NodeList) FirstImgExpr.evaluate(htmlNode, XPathConstants.NODESET); |
| 235 | if (imgNodes.getLength() > 0) { |
| 236 | //iterate nodes looking for valid image url and normalize. |
| 237 | for (int i = 0; i < imgNodes.getLength(); i++) { |
| 238 | String tx = imgNodes.item(i).getNodeValue(); |
| 239 | //qualify and normalize the image |
| 240 | imageUrl = getImageUrlNormalized(tx); |
| 241 | //this img src did not qualify, keep looking... |
| 242 | if ("".equals(imageUrl)) { |
| 243 | if (Doclava.META_DBG) System.out.println(" >>>>> Discarded image: " + tx); |
| 244 | continue; |
| 245 | } else { |
| 246 | imgFrom = "markup"; |
| 247 | break; |
| 248 | } |
| 249 | } |
| 250 | } |
| 251 | } |
| 252 | //img src url looks good, store it to the file hdf data |
| 253 | hdf.setValue("page.image", imageUrl); |
| 254 | } |
| 255 | if (Doclava.META_DBG) System.out.println("Image (" + imgFrom + "): " + imageUrl); |
| 256 | if (Doclava.META_DBG) System.out.println("Summary (" + sumFrom + "): " + sum.length() |
| 257 | + " chars\n\n" + sum + "\n"); |
| 258 | return; |
| 259 | |
| 260 | } catch (Exception e) { |
| 261 | if (Doclava.META_DBG) System.out.println(" >>>>> Exception: " + e + "\n"); |
| 262 | } |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 263 | } |
| 264 | |
| 265 | /** |
| 266 | * Normalize a comma-delimited, multi-string value. Split on commas, remove |
| 267 | * quotes, trim whitespace, optionally make keywords/tags lowercase for |
| 268 | * easier matching. |
| 269 | * |
| 270 | * @param hdf Data object in which the metadata values are stored. |
| 271 | * @param tag The hdf var from which the metadata was extracted. |
| 272 | * @return A normalized string value for the specified tag. |
| 273 | */ |
| 274 | public static String getPageTagsNormalized(Data hdf, String tag) { |
| 275 | |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 276 | String normTags = ""; |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 277 | StringBuilder tags = new StringBuilder(); |
| 278 | String tagList = hdf.getValue(tag, ""); |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 279 | if (tag.equals("meta.tags") && (tagList.equals(""))) { |
| 280 | //use keywords as tags if no meta tags are available |
| 281 | tagList = hdf.getValue("page.tags", ""); |
| 282 | } |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 283 | if (!tagList.equals("")) { |
| 284 | tagList = tagList.replaceAll("\"", ""); |
Dirk Dougherty | 5400437 | 2015-07-02 13:10:45 -0700 | [diff] [blame] | 285 | |
| 286 | String[] tagParts = tagList.split("[,\u3001]"); |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 287 | for (int iter = 0; iter < tagParts.length; iter++) { |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 288 | tags.append("\""); |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 289 | if (tag.equals("meta.tags") && sLowercaseTags) { |
| 290 | tagParts[iter] = tagParts[iter].toLowerCase(); |
| 291 | } else if (tag.equals("page.tags") && sLowercaseKeywords) { |
| 292 | tagParts[iter] = tagParts[iter].toLowerCase(); |
| 293 | } |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 294 | if (tag.equals("meta.tags")) { |
| 295 | //tags.append("#"); //to match hashtag format used with yt/blogger resources |
| 296 | tagParts[iter] = tagParts[iter].replaceAll(" ",""); |
| 297 | } |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 298 | tags.append(tagParts[iter].trim()); |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 299 | tags.append("\""); |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 300 | if (iter < tagParts.length - 1) { |
| 301 | tags.append(","); |
| 302 | } |
| 303 | } |
| 304 | } |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 305 | //write this back to hdf to expose through js |
| 306 | if (tag.equals("meta.tags")) { |
| 307 | hdf.setValue(tag, tags.toString()); |
| 308 | } |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 309 | return tags.toString(); |
| 310 | } |
| 311 | |
| 312 | /** |
| 313 | * Normalize a string for which only a single value is supported. |
| 314 | * Extract the string up to the first comma, remove quotes, remove |
| 315 | * any forward-slash prefix, trim any whitespace, optionally make |
| 316 | * lowercase for easier matching. |
| 317 | * |
| 318 | * @param hdf Data object in which the metadata values are stored. |
| 319 | * @param tag The hdf var from which the metadata should be extracted. |
| 320 | * @return A normalized string value for the specified tag. |
| 321 | */ |
| 322 | public static String getStringValueNormalized(Data hdf, String tag) { |
| 323 | StringBuilder outString = new StringBuilder(); |
| 324 | String tagList = hdf.getValue(tag, ""); |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 325 | tagList.replaceAll("\"", ""); |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 326 | if (!tagList.isEmpty()) { |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 327 | int end = tagList.indexOf(","); |
| 328 | if (end != -1) { |
| 329 | tagList = tagList.substring(0,end); |
| 330 | } |
| 331 | tagList = tagList.startsWith("/") ? tagList.substring(1) : tagList; |
| 332 | if ("sample.group".equals(tag) && sLowercaseTags) { |
| 333 | tagList = tagList.toLowerCase(); |
| 334 | } |
| 335 | outString.append(tagList.trim()); |
| 336 | } |
| 337 | return outString.toString(); |
| 338 | } |
| 339 | |
| 340 | /** |
| 341 | * Normalize a page title. Extract the string, remove quotes, remove |
| 342 | * markup, and trim any whitespace. |
| 343 | * |
| 344 | * @param hdf Data object in which the metadata values are stored. |
| 345 | * @param tag The hdf var from which the metadata should be extracted. |
| 346 | * @return A normalized string value for the specified tag. |
| 347 | */ |
| 348 | public static String getTitleNormalized(Data hdf, String tag) { |
| 349 | StringBuilder outTitle = new StringBuilder(); |
| 350 | String title = hdf.getValue(tag, ""); |
| 351 | if (!title.isEmpty()) { |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 352 | title = escapeString(title); |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 353 | if (title.indexOf("<span") != -1) { |
| 354 | String[] splitTitle = title.split("<span(.*?)</span>"); |
| 355 | title = splitTitle[0]; |
| 356 | for (int j = 1; j < splitTitle.length; j++) { |
| 357 | title.concat(splitTitle[j]); |
| 358 | } |
| 359 | } |
| 360 | outTitle.append(title.trim()); |
| 361 | } |
| 362 | return outTitle.toString(); |
| 363 | } |
| 364 | |
| 365 | /** |
| 366 | * Extract and normalize a page's language string based on the |
| 367 | * lowercased dir path. Non-supported langs are ignored and assigned |
| 368 | * the default lang string of "en". |
| 369 | * |
| 370 | * @param filename A path string to the file relative to root. |
| 371 | * @return A normalized lang value. |
| 372 | */ |
| 373 | public static String getLangStringNormalized(String filename) { |
| 374 | String[] stripStr = filename.toLowerCase().split("\\/"); |
| 375 | String outFrag = "en"; |
| 376 | if (stripStr.length > 0) { |
| 377 | for (String t : DocFile.DEVSITE_VALID_LANGS) { |
| 378 | if ("intl".equals(stripStr[0])) { |
| 379 | if (t.equals(stripStr[1])) { |
| 380 | outFrag = stripStr[1]; |
| 381 | break; |
| 382 | } |
| 383 | } |
| 384 | } |
| 385 | } |
| 386 | return outFrag; |
| 387 | } |
| 388 | |
| 389 | /** |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 390 | * Normalize a page summary string and truncate as needed. Strings |
| 391 | * exceeding max_chars are truncated at the first word boundary |
| 392 | * following the max_size marker. Strings smaller than min_chars |
| 393 | * are discarded (as they are assumed to be too little context). |
| 394 | * |
| 395 | * @param s String extracted from the page as it's summary. |
| 396 | * @return A normalized string value. |
| 397 | */ |
| 398 | public static String getSummaryNormalized(String s) { |
| 399 | String str = ""; |
| 400 | int max_chars = 250; |
| 401 | int min_chars = 50; |
| 402 | int marker = 0; |
| 403 | if (s.length() < min_chars) { |
| 404 | return str; |
| 405 | } else { |
| 406 | str = s.replaceAll("^\"|\"$", ""); |
| 407 | str = str.replaceAll("\\s+", " "); |
| 408 | str = JD_TAG_PATTERN.matcher(str).replaceAll("$1"); |
| 409 | str = escapeString(str); |
| 410 | BreakIterator bi = BreakIterator.getWordInstance(); |
| 411 | bi.setText(str); |
| 412 | if (str.length() > max_chars) { |
| 413 | marker = bi.following(max_chars); |
| 414 | } else { |
| 415 | marker = bi.last(); |
| 416 | } |
| 417 | str = str.substring(0, marker); |
| 418 | str = str.concat("\u2026" ); |
| 419 | } |
| 420 | return str; |
| 421 | } |
| 422 | |
| 423 | public static String escapeString(String s) { |
| 424 | s = s.replaceAll("\"", """); |
| 425 | s = s.replaceAll("\'", "'"); |
| 426 | s = s.replaceAll("<", "<"); |
| 427 | s = s.replaceAll(">", ">"); |
| 428 | s = s.replaceAll("/", "/"); |
| 429 | return s; |
| 430 | } |
| 431 | |
| 432 | //Disqualify img src urls that include these substrings |
| 433 | public static String[] IMAGE_EXCLUDE = {"/triangle-", "favicon","android-logo", |
| 434 | "icon_play.png", "robot-tiny"}; |
| 435 | |
| 436 | public static boolean inList(String s, String[] list) { |
| 437 | for (String t : list) { |
| 438 | if (s.contains(t)) { |
| 439 | return true; |
| 440 | } |
| 441 | } |
| 442 | return false; |
| 443 | } |
| 444 | |
| 445 | /** |
| 446 | * Normalize an img src url by removing docRoot and leading |
| 447 | * slash for local image references. These are added later |
| 448 | * in js to support offline mode and keep path reference |
| 449 | * format consistent with hrefs. |
| 450 | * |
| 451 | * @param url Abs or rel url sourced from img src. |
| 452 | * @return Normalized url if qualified, else empty |
| 453 | */ |
| 454 | public static String getImageUrlNormalized(String url) { |
| 455 | String absUrl = ""; |
| 456 | // validate to avoid choosing using specific images |
| 457 | if ((url != null) && (!url.equals("")) && (!inList(url, IMAGE_EXCLUDE))) { |
| 458 | absUrl = url.replace("{@docRoot}", ""); |
| 459 | absUrl = absUrl.replaceFirst("^/(?!/)", ""); |
| 460 | } |
| 461 | return absUrl; |
| 462 | } |
| 463 | |
| 464 | /** |
| 465 | * Normalize an href url by removing docRoot and leading |
| 466 | * slash for local image references. These are added later |
| 467 | * in js to support offline mode and keep path reference |
| 468 | * format consistent with hrefs. |
| 469 | * |
| 470 | * @param url Abs or rel page url sourced from href |
| 471 | * @return Normalized url, either abs or rel to root |
| 472 | */ |
| 473 | public static String getPageUrlNormalized(String url) { |
| 474 | String absUrl = ""; |
| 475 | if ((url !=null) && (!url.equals(""))) { |
| 476 | absUrl = url.replace("{@docRoot}", ""); |
| 477 | absUrl = absUrl.replaceFirst("^/(?!/)", ""); |
| 478 | } |
| 479 | return absUrl; |
| 480 | } |
| 481 | |
| 482 | /** |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 483 | * Given a metadata node, add it as a child of a root node based on its |
| 484 | * type. If there is no root node that matches the node's type, create one |
| 485 | * and add the metadata node as a child node. |
| 486 | * |
| 487 | * @param gNode The node to attach to a root node or add as a new root node. |
| 488 | * @param rootList The current list of root nodes. |
| 489 | * @return The updated list of root nodes. |
| 490 | */ |
| 491 | public static List<Node> appendMetaNodeByType(Node gNode, List<Node> rootList) { |
| 492 | |
| 493 | String nodeTags = gNode.getType(); |
| 494 | boolean matched = false; |
| 495 | for (Node n : rootList) { |
| 496 | if (n.getType().equals(nodeTags)) { //find any matching type node |
| 497 | n.getChildren().add(gNode); |
| 498 | matched = true; |
| 499 | break; // add to the first root node only |
| 500 | } // tag did not match |
| 501 | } // end rootnodes matching iterator |
| 502 | if (!matched) { |
| 503 | List<Node> mtaglist = new ArrayList<Node>(); // list of file objects that have a given type |
| 504 | mtaglist.add(gNode); |
| 505 | Node tnode = new Node.Builder().setChildren(mtaglist).setType(nodeTags).build(); |
| 506 | rootList.add(tnode); |
| 507 | } |
| 508 | return rootList; |
| 509 | } |
| 510 | |
| 511 | /** |
| 512 | * Given a metadata node, add it as a child of a root node based on its |
| 513 | * tag. If there is no root node matching the tag, create one for it |
| 514 | * and add the metadata node as a child node. |
| 515 | * |
| 516 | * @param gNode The node to attach to a root node or add as a new root node. |
| 517 | * @param rootTagNodesList The current list of root nodes. |
| 518 | * @return The updated list of root nodes. |
| 519 | */ |
| 520 | public static List<Node> appendMetaNodeByTagIndex(Node gNode, List<Node> rootTagNodesList) { |
| 521 | |
| 522 | for (int iter = 0; iter < gNode.getChildren().size(); iter++) { |
| 523 | if (gNode.getChildren().get(iter).getTags() != null) { |
| 524 | List<String> nodeTags = gNode.getChildren().get(iter).getTags(); |
| 525 | boolean matched = false; |
| 526 | for (String t : nodeTags) { //process each of the meta.tags |
| 527 | for (Node n : rootTagNodesList) { |
| 528 | if (n.getLabel().equals(t.toString())) { |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 529 | n.getTags().add(String.valueOf(iter)); |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 530 | matched = true; |
| 531 | break; // add to the first root node only |
| 532 | } // tag did not match |
| 533 | } // end rootnodes matching iterator |
| 534 | if (!matched) { |
| 535 | List<String> mtaglist = new ArrayList<String>(); // list of objects with a given tag |
| 536 | mtaglist.add(String.valueOf(iter)); |
| 537 | Node tnode = new Node.Builder().setLabel(t.toString()).setTags(mtaglist).build(); |
| 538 | rootTagNodesList.add(tnode); |
| 539 | } |
| 540 | } |
| 541 | } |
| 542 | } |
| 543 | return rootTagNodesList; |
| 544 | } |
| 545 | |
| 546 | public static final Comparator<Node> BY_TAG_NAME = new Comparator<Node>() { |
| 547 | public int compare (Node one, Node other) { |
| 548 | return one.getLabel().compareTo(other.getLabel()); |
| 549 | } |
| 550 | }; |
| 551 | |
| 552 | public static final Comparator<Node> BY_TYPE_NAME = new Comparator<Node>() { |
| 553 | public int compare (Node one, Node other) { |
| 554 | return one.getType().compareTo(other.getType()); |
| 555 | } |
| 556 | }; |
| 557 | |
| 558 | /** |
| 559 | * A node for storing page metadata. Use Builder.build() to instantiate. |
| 560 | */ |
| 561 | public static class Node { |
| 562 | |
| 563 | private String mLabel; // holds page.title or similar identifier |
| 564 | private String mTitleFriendly; // title for card or similar use |
| 565 | private String mSummary; // Summary for card or similar use |
| 566 | private String mLink; //link href for item click |
| 567 | private String mGroup; // from sample.group in _index.jd |
| 568 | private List<String> mKeywords; // from page.tags |
| 569 | private List<String> mTags; // from meta.tags |
| 570 | private String mImage; // holds an href, fully qualified or relative to root |
| 571 | private List<Node> mChildren; |
| 572 | private String mLang; |
| 573 | private String mType; // can be file, dir, video show, announcement, etc. |
| 574 | |
| 575 | private Node(Builder builder) { |
| 576 | mLabel = builder.mLabel; |
| 577 | mTitleFriendly = builder.mTitleFriendly; |
| 578 | mSummary = builder.mSummary; |
| 579 | mLink = builder.mLink; |
| 580 | mGroup = builder.mGroup; |
| 581 | mKeywords = builder.mKeywords; |
| 582 | mTags = builder.mTags; |
| 583 | mImage = builder.mImage; |
| 584 | mChildren = builder.mChildren; |
| 585 | mLang = builder.mLang; |
| 586 | mType = builder.mType; |
| 587 | } |
| 588 | |
| 589 | private static class Builder { |
| 590 | private String mLabel, mTitleFriendly, mSummary, mLink, mGroup, mImage, mLang, mType; |
| 591 | private List<String> mKeywords = null; |
| 592 | private List<String> mTags = null; |
| 593 | private List<Node> mChildren = null; |
| 594 | public Builder setLabel(String mLabel) { this.mLabel = mLabel; return this;} |
| 595 | public Builder setTitleFriendly(String mTitleFriendly) { |
| 596 | this.mTitleFriendly = mTitleFriendly; return this; |
| 597 | } |
| 598 | public Builder setSummary(String mSummary) {this.mSummary = mSummary; return this;} |
| 599 | public Builder setLink(String mLink) {this.mLink = mLink; return this;} |
| 600 | public Builder setGroup(String mGroup) {this.mGroup = mGroup; return this;} |
| 601 | public Builder setKeywords(List<String> mKeywords) { |
| 602 | this.mKeywords = mKeywords; return this; |
| 603 | } |
| 604 | public Builder setTags(List<String> mTags) {this.mTags = mTags; return this;} |
| 605 | public Builder setImage(String mImage) {this.mImage = mImage; return this;} |
| 606 | public Builder setChildren(List<Node> mChildren) {this.mChildren = mChildren; return this;} |
| 607 | public Builder setLang(String mLang) {this.mLang = mLang; return this;} |
| 608 | public Builder setType(String mType) {this.mType = mType; return this;} |
| 609 | public Node build() {return new Node(this);} |
| 610 | } |
| 611 | |
| 612 | /** |
| 613 | * Render a tree of metadata nodes organized by type. |
| 614 | * @param buf Output buffer to render to. |
| 615 | */ |
| 616 | void renderTypeResources(StringBuilder buf) { |
| 617 | List<Node> list = mChildren; //list of type rootnodes |
| 618 | if (list == null || list.size() == 0) { |
| 619 | buf.append("null"); |
| 620 | } else { |
| 621 | final int n = list.size(); |
| 622 | for (int i = 0; i < n; i++) { |
| 623 | buf.append("var " + list.get(i).mType.toUpperCase() + "_RESOURCES = ["); |
| 624 | list.get(i).renderTypes(buf); //render this type's children |
| 625 | buf.append("\n];\n\n"); |
| 626 | } |
| 627 | } |
| 628 | } |
| 629 | /** |
| 630 | * Render all metadata nodes for a specific type. |
| 631 | * @param buf Output buffer to render to. |
| 632 | */ |
| 633 | void renderTypes(StringBuilder buf) { |
| 634 | List<Node> list = mChildren; |
| 635 | if (list == null || list.size() == 0) { |
| 636 | buf.append("nulltype"); |
| 637 | } else { |
| 638 | final int n = list.size(); |
| 639 | for (int i = 0; i < n; i++) { |
| 640 | buf.append("\n {\n"); |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 641 | buf.append(" \"title\":\"" + list.get(i).mLabel + "\",\n" ); |
| 642 | buf.append(" \"titleFriendly\":\"" + list.get(i).mTitleFriendly + "\",\n" ); |
| 643 | buf.append(" \"summary\":\"" + list.get(i).mSummary + "\",\n" ); |
| 644 | buf.append(" \"url\":\"" + list.get(i).mLink + "\",\n" ); |
| 645 | buf.append(" \"group\":\"" + list.get(i).mGroup + "\",\n" ); |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 646 | list.get(i).renderArrayType(buf, list.get(i).mKeywords, "keywords"); |
| 647 | list.get(i).renderArrayType(buf, list.get(i).mTags, "tags"); |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 648 | buf.append(" \"image\":\"" + list.get(i).mImage + "\",\n" ); |
| 649 | buf.append(" \"lang\":\"" + list.get(i).mLang + "\",\n" ); |
| 650 | buf.append(" \"type\":\"" + list.get(i).mType + "\""); |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 651 | buf.append("\n }"); |
| 652 | if (i != n - 1) { |
| 653 | buf.append(", "); |
| 654 | } |
| 655 | } |
| 656 | } |
| 657 | } |
| 658 | |
| 659 | /** |
| 660 | * Build and render a list of tags associated with each type. |
| 661 | * @param buf Output buffer to render to. |
| 662 | */ |
| 663 | void renderTypesByTag(StringBuilder buf) { |
| 664 | List<Node> list = mChildren; //list of rootnodes |
| 665 | if (list == null || list.size() == 0) { |
| 666 | buf.append("null"); |
| 667 | } else { |
| 668 | final int n = list.size(); |
| 669 | for (int i = 0; i < n; i++) { |
| 670 | buf.append("var " + list.get(i).mType.toUpperCase() + "_BY_TAG = {"); |
| 671 | List<Node> mTagList = new ArrayList(); //list of rootnodes |
| 672 | mTagList = appendMetaNodeByTagIndex(list.get(i), mTagList); |
| 673 | list.get(i).renderTagIndices(buf, mTagList); |
| 674 | buf.append("\n};\n\n"); |
| 675 | } |
| 676 | } |
| 677 | } |
| 678 | |
| 679 | /** |
| 680 | * Render a list of tags associated with a type, including the |
| 681 | * tag's indices in the type array. |
| 682 | * @param buf Output buffer to render to. |
| 683 | * @param tagList Node tree of types to render. |
| 684 | */ |
| 685 | void renderTagIndices(StringBuilder buf, List<Node> tagList) { |
| 686 | List<Node> list = tagList; |
| 687 | if (list == null || list.size() == 0) { |
Dirk Dougherty | 20afdc1 | 2013-12-09 13:09:22 -0800 | [diff] [blame] | 688 | buf.append(""); |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 689 | } else { |
| 690 | final int n = list.size(); |
| 691 | for (int i = 0; i < n; i++) { |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 692 | buf.append("\n " + list.get(i).mLabel + ":["); |
| 693 | renderArrayValue(buf, list.get(i).mTags); |
| 694 | buf.append("]"); |
| 695 | if (i != n - 1) { |
| 696 | buf.append(", "); |
| 697 | } |
| 698 | } |
| 699 | } |
| 700 | } |
| 701 | |
| 702 | /** |
| 703 | * Render key:arrayvalue pair. |
| 704 | * @param buf Output buffer to render to. |
| 705 | * @param type The list value to render as an arrayvalue. |
| 706 | * @param key The key for the pair. |
| 707 | */ |
| 708 | void renderArrayType(StringBuilder buf, List<String> type, String key) { |
Dirk Dougherty | c770a6e | 2014-02-07 20:04:27 -0800 | [diff] [blame] | 709 | buf.append(" \"" + key + "\": ["); |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 710 | renderArrayValue(buf, type); |
| 711 | buf.append("],\n"); |
| 712 | } |
| 713 | |
| 714 | /** |
| 715 | * Render an array value to buf, with special handling of unicode characters. |
| 716 | * @param buf Output buffer to render to. |
| 717 | * @param type The list value to render as an arrayvalue. |
| 718 | */ |
| 719 | void renderArrayValue(StringBuilder buf, List<String> type) { |
| 720 | List<String> list = type; |
| 721 | if (list != null) { |
| 722 | final int n = list.size(); |
| 723 | for (int i = 0; i < n; i++) { |
| 724 | String tagval = list.get(i).toString(); |
| 725 | final int L = tagval.length(); |
| 726 | for (int t = 0; t < L; t++) { |
| 727 | char c = tagval.charAt(t); |
Dirk Dougherty | 5400437 | 2015-07-02 13:10:45 -0700 | [diff] [blame] | 728 | if (c >= Character.MIN_HIGH_SURROGATE && c <= Character.MAX_HIGH_SURROGATE ) { |
| 729 | // we have a UTF-16 multi-byte character |
| 730 | int codePoint = tagval.codePointAt(t); |
| 731 | int charSize = Character.charCount(codePoint); |
| 732 | t += charSize - 1; |
| 733 | buf.append(String.format("\\u%04x",codePoint)); |
| 734 | } else if (c >= ' ' && c <= '~' && c != '\\') { |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 735 | buf.append(c); |
Dirk Dougherty | 5400437 | 2015-07-02 13:10:45 -0700 | [diff] [blame] | 736 | } else { |
| 737 | // we are encoding a two byte character |
| 738 | buf.append(String.format("\\u%04x", (int) c)); |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 739 | } |
| 740 | } |
| 741 | if (i != n - 1) { |
| 742 | buf.append(","); |
| 743 | } |
| 744 | } |
| 745 | } |
| 746 | } |
| 747 | |
| 748 | public String getLabel() { |
| 749 | return mLabel; |
| 750 | } |
| 751 | |
| 752 | public void setLabel(String label) { |
| 753 | mLabel = label; |
| 754 | } |
| 755 | |
| 756 | public String getTitleFriendly() { |
| 757 | return mTitleFriendly; |
| 758 | } |
| 759 | |
| 760 | public void setTitleFriendly(String title) { |
| 761 | mTitleFriendly = title; |
| 762 | } |
| 763 | |
| 764 | public String getSummary() { |
| 765 | return mSummary; |
| 766 | } |
| 767 | |
| 768 | public void setSummary(String summary) { |
| 769 | mSummary = summary; |
| 770 | } |
| 771 | |
| 772 | public String getLink() { |
| 773 | return mLink; |
| 774 | } |
| 775 | |
| 776 | public void setLink(String ref) { |
| 777 | mLink = ref; |
| 778 | } |
| 779 | |
| 780 | public String getGroup() { |
| 781 | return mGroup; |
| 782 | } |
| 783 | |
| 784 | public void setGroup(String group) { |
| 785 | mGroup = group; |
| 786 | } |
| 787 | |
| 788 | public List<String> getTags() { |
| 789 | return mTags; |
| 790 | } |
| 791 | |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 792 | public void setTags(String tags) { |
Dirk Dougherty | 8a8f791 | 2013-11-26 18:45:31 -0800 | [diff] [blame] | 793 | if ("".equals(tags)) { |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 794 | mTags = null; |
| 795 | } else { |
| 796 | List<String> tagList = new ArrayList(); |
| 797 | String[] tagParts = tags.split(","); |
| 798 | |
| 799 | for (String t : tagParts) { |
| 800 | tagList.add(t); |
| 801 | } |
| 802 | mTags = tagList; |
| 803 | } |
| 804 | } |
| 805 | |
Dirk Dougherty | 8a8f791 | 2013-11-26 18:45:31 -0800 | [diff] [blame] | 806 | public List<String> getKeywords() { |
| 807 | return mKeywords; |
| 808 | } |
| 809 | |
| 810 | public void setKeywords(String keywords) { |
| 811 | if ("".equals(keywords)) { |
| 812 | mKeywords = null; |
| 813 | } else { |
| 814 | List<String> keywordList = new ArrayList(); |
| 815 | String[] keywordParts = keywords.split(","); |
| 816 | |
| 817 | for (String k : keywordParts) { |
| 818 | keywordList.add(k); |
| 819 | } |
| 820 | mKeywords = keywordList; |
| 821 | } |
| 822 | } |
| 823 | |
Dirk Dougherty | f9d9cb0 | 2013-09-13 19:45:17 -0700 | [diff] [blame] | 824 | public String getImage() { |
| 825 | return mImage; |
| 826 | } |
| 827 | |
| 828 | public void setImage(String ref) { |
| 829 | mImage = ref; |
| 830 | } |
| 831 | |
| 832 | public List<Node> getChildren() { |
| 833 | return mChildren; |
| 834 | } |
| 835 | |
| 836 | public void setChildren(List<Node> node) { |
| 837 | mChildren = node; |
| 838 | } |
| 839 | |
| 840 | public String getLang() { |
| 841 | return mLang; |
| 842 | } |
| 843 | |
| 844 | public void setLang(String lang) { |
| 845 | mLang = lang; |
| 846 | } |
| 847 | |
| 848 | public String getType() { |
| 849 | return mType; |
| 850 | } |
| 851 | |
| 852 | public void setType(String type) { |
| 853 | mType = type; |
| 854 | } |
| 855 | } |
| 856 | } |