You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

PDFStructureTreeBuilder.java 15KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362
  1. /*
  2. * Licensed to the Apache Software Foundation (ASF) under one or more
  3. * contributor license agreements. See the NOTICE file distributed with
  4. * this work for additional information regarding copyright ownership.
  5. * The ASF licenses this file to You under the Apache License, Version 2.0
  6. * (the "License"); you may not use this file except in compliance with
  7. * the License. You may obtain a copy of the License at
  8. *
  9. * http://www.apache.org/licenses/LICENSE-2.0
  10. *
  11. * Unless required by applicable law or agreed to in writing, software
  12. * distributed under the License is distributed on an "AS IS" BASIS,
  13. * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  14. * See the License for the specific language governing permissions and
  15. * limitations under the License.
  16. */
  17. /* $Id$ */
  18. package org.apache.fop.render.pdf;
  19. import java.util.LinkedList;
  20. import java.util.Locale;
  21. import java.util.Map;
  22. import org.xml.sax.Attributes;
  23. import org.xml.sax.helpers.AttributesImpl;
  24. import org.apache.fop.accessibility.StructureTreeElement;
  25. import org.apache.fop.accessibility.StructureTreeEventHandler;
  26. import org.apache.fop.events.EventBroadcaster;
  27. import org.apache.fop.fo.extensions.ExtensionElementMapping;
  28. import org.apache.fop.fo.pagination.Flow;
  29. import org.apache.fop.pdf.PDFFactory;
  30. import org.apache.fop.pdf.PDFParentTree;
  31. import org.apache.fop.pdf.PDFStructElem;
  32. import org.apache.fop.pdf.PDFStructTreeRoot;
  33. import org.apache.fop.pdf.StandardStructureAttributes.Table.Scope;
  34. import org.apache.fop.pdf.StandardStructureTypes;
  35. import org.apache.fop.pdf.StandardStructureTypes.Grouping;
  36. import org.apache.fop.pdf.StandardStructureTypes.Table;
  37. import org.apache.fop.pdf.StructureHierarchyMember;
  38. import org.apache.fop.pdf.StructureType;
  39. import org.apache.fop.util.XMLUtil;
  40. class PDFStructureTreeBuilder implements StructureTreeEventHandler {
  41. private static final String ROLE = "role";
  42. private static final Map<String, StructureElementBuilder> BUILDERS
  43. = new java.util.HashMap<String, StructureElementBuilder>();
  44. private static final StructureElementBuilder DEFAULT_BUILDER
  45. = new DefaultStructureElementBuilder(Grouping.NON_STRUCT);
  46. static {
  47. // Declarations and Pagination and Layout Formatting Objects
  48. StructureElementBuilder regionBuilder = new RegionBuilder();
  49. addBuilder("root", StandardStructureTypes.Grouping.DOCUMENT);
  50. addBuilder("page-sequence", new PageSequenceBuilder());
  51. addBuilder("static-content", regionBuilder);
  52. addBuilder("flow", regionBuilder);
  53. // Block-level Formatting Objects
  54. addBuilder("block", StandardStructureTypes.Paragraphlike.P);
  55. addBuilder("block-container", StandardStructureTypes.Grouping.DIV);
  56. // Inline-level Formatting Objects
  57. addBuilder("character", StandardStructureTypes.InlineLevelStructure.SPAN);
  58. addBuilder("external-graphic", new ImageBuilder());
  59. addBuilder("instream-foreign-object", new ImageBuilder());
  60. addBuilder("inline", StandardStructureTypes.InlineLevelStructure.SPAN);
  61. addBuilder("inline-container", StandardStructureTypes.Grouping.DIV);
  62. addBuilder("page-number", StandardStructureTypes.InlineLevelStructure.QUOTE);
  63. addBuilder("page-number-citation", StandardStructureTypes.InlineLevelStructure.QUOTE);
  64. addBuilder("page-number-citation-last", StandardStructureTypes.InlineLevelStructure.QUOTE);
  65. // Formatting Objects for Tables
  66. addBuilder("table-and-caption", StandardStructureTypes.Grouping.DIV);
  67. addBuilder("table", new TableBuilder());
  68. addBuilder("table-caption", StandardStructureTypes.Grouping.CAPTION);
  69. addBuilder("table-header", StandardStructureTypes.Table.THEAD);
  70. addBuilder("table-footer", new TableFooterBuilder());
  71. addBuilder("table-body", StandardStructureTypes.Table.TBODY);
  72. addBuilder("table-row", StandardStructureTypes.Table.TR);
  73. addBuilder("table-cell", new TableCellBuilder());
  74. // Formatting Objects for Lists
  75. addBuilder("list-block", StandardStructureTypes.List.L);
  76. addBuilder("list-item", StandardStructureTypes.List.LI);
  77. addBuilder("list-item-body", StandardStructureTypes.List.LBODY);
  78. addBuilder("list-item-label", StandardStructureTypes.List.LBL);
  79. // Dynamic Effects: Link and Multi Formatting Objects
  80. addBuilder("basic-link", StandardStructureTypes.InlineLevelStructure.LINK);
  81. // Out-of-Line Formatting Objects
  82. addBuilder("float", StandardStructureTypes.Grouping.DIV);
  83. addBuilder("footnote", StandardStructureTypes.InlineLevelStructure.NOTE);
  84. addBuilder("footnote-body", StandardStructureTypes.Grouping.SECT);
  85. addBuilder("wrapper", StandardStructureTypes.InlineLevelStructure.SPAN);
  86. addBuilder("marker", StandardStructureTypes.Grouping.PRIVATE);
  87. addBuilder("#PCDATA", new PlaceholderBuilder());
  88. }
  89. private static void addBuilder(String fo, StructureType structureType) {
  90. addBuilder(fo, new DefaultStructureElementBuilder(structureType));
  91. }
  92. private static void addBuilder(String fo, StructureElementBuilder mapper) {
  93. BUILDERS.put(fo, mapper);
  94. }
  95. private interface StructureElementBuilder {
  96. PDFStructElem build(StructureHierarchyMember parent, Attributes attributes, PDFFactory pdfFactory,
  97. EventBroadcaster eventBroadcaster);
  98. }
  99. private static class DefaultStructureElementBuilder implements StructureElementBuilder {
  100. private final StructureType defaultStructureType;
  101. DefaultStructureElementBuilder(StructureType structureType) {
  102. this.defaultStructureType = structureType;
  103. }
  104. public final PDFStructElem build(StructureHierarchyMember parent, Attributes attributes,
  105. PDFFactory pdfFactory, EventBroadcaster eventBroadcaster) {
  106. String role = attributes.getValue(ROLE);
  107. StructureType structureType;
  108. if (role == null) {
  109. structureType = defaultStructureType;
  110. } else {
  111. structureType = StandardStructureTypes.get(role);
  112. if (structureType == null) {
  113. structureType = defaultStructureType;
  114. PDFEventProducer.Provider.get(eventBroadcaster).nonStandardStructureType(role, role,
  115. structureType.toString());
  116. }
  117. }
  118. PDFStructElem structElem = createStructureElement(parent, structureType);
  119. setAttributes(structElem, attributes);
  120. addKidToParent(structElem, parent, attributes);
  121. registerStructureElement(structElem, pdfFactory);
  122. return structElem;
  123. }
  124. protected PDFStructElem createStructureElement(StructureHierarchyMember parent,
  125. StructureType structureType) {
  126. return new PDFStructElem(parent, structureType);
  127. }
  128. protected void setAttributes(PDFStructElem structElem, Attributes attributes) {
  129. }
  130. protected void addKidToParent(PDFStructElem kid, StructureHierarchyMember parent,
  131. Attributes attributes) {
  132. parent.addKid(kid);
  133. }
  134. protected void registerStructureElement(PDFStructElem structureElement, PDFFactory pdfFactory) {
  135. pdfFactory.getDocument().registerStructureElement(structureElement);
  136. }
  137. }
  138. private static class PageSequenceBuilder extends DefaultStructureElementBuilder {
  139. PageSequenceBuilder() {
  140. super(StandardStructureTypes.Grouping.PART);
  141. }
  142. @Override
  143. protected PDFStructElem createStructureElement(StructureHierarchyMember parent,
  144. StructureType structureType) {
  145. return new PageSequenceStructElem(parent, structureType);
  146. }
  147. }
  148. private static class RegionBuilder extends DefaultStructureElementBuilder {
  149. RegionBuilder() {
  150. super(StandardStructureTypes.Grouping.SECT);
  151. }
  152. @Override
  153. protected void addKidToParent(PDFStructElem kid, StructureHierarchyMember parent,
  154. Attributes attributes) {
  155. String flowName = attributes.getValue(Flow.FLOW_NAME);
  156. ((PageSequenceStructElem) parent).addContent(flowName, kid);
  157. }
  158. }
  159. private static class ImageBuilder extends DefaultStructureElementBuilder {
  160. ImageBuilder() {
  161. super(StandardStructureTypes.Illustration.FIGURE);
  162. }
  163. @Override
  164. protected void setAttributes(PDFStructElem structElem, Attributes attributes) {
  165. String altTextNode = attributes.getValue(ExtensionElementMapping.URI, "alt-text");
  166. if (altTextNode == null) {
  167. altTextNode = "No alternate text specified";
  168. }
  169. structElem.put("Alt", altTextNode);
  170. }
  171. }
  172. private static class TableBuilder extends DefaultStructureElementBuilder {
  173. TableBuilder() {
  174. super(StandardStructureTypes.Table.TABLE);
  175. }
  176. @Override
  177. protected PDFStructElem createStructureElement(StructureHierarchyMember parent,
  178. StructureType structureType) {
  179. return new TableStructElem(parent, structureType);
  180. }
  181. }
  182. private static class TableFooterBuilder extends DefaultStructureElementBuilder {
  183. public TableFooterBuilder() {
  184. super(StandardStructureTypes.Table.TFOOT);
  185. }
  186. @Override
  187. protected void addKidToParent(PDFStructElem kid, StructureHierarchyMember parent,
  188. Attributes attributes) {
  189. ((TableStructElem) parent).addTableFooter(kid);
  190. }
  191. }
  192. private static class TableCellBuilder extends DefaultStructureElementBuilder {
  193. TableCellBuilder() {
  194. super(StandardStructureTypes.Table.TD);
  195. }
  196. @Override
  197. protected PDFStructElem createStructureElement(StructureHierarchyMember parent,
  198. StructureType structureType) {
  199. PDFStructElem grandParent = ((PDFStructElem) parent).getParentStructElem();
  200. //TODO What to do with cells from table-footer? Currently they are mapped on TD.
  201. if (grandParent.getStructureType() == StandardStructureTypes.Table.THEAD) {
  202. structureType = StandardStructureTypes.Table.TH;
  203. } else {
  204. structureType = StandardStructureTypes.Table.TD;
  205. }
  206. return super.createStructureElement(parent, structureType);
  207. }
  208. @Override
  209. protected void registerStructureElement(PDFStructElem structureElement, PDFFactory pdfFactory) {
  210. if (structureElement.getStructureType() == Table.TH) {
  211. pdfFactory.getDocument().registerStructureElement(structureElement, Scope.COLUMN);
  212. } else {
  213. pdfFactory.getDocument().registerStructureElement(structureElement);
  214. }
  215. }
  216. @Override
  217. protected void setAttributes(PDFStructElem structElem, Attributes attributes) {
  218. String columnSpan = attributes.getValue("number-columns-spanned");
  219. if (columnSpan != null) {
  220. structElem.setTableAttributeColSpan(Integer.parseInt(columnSpan));
  221. }
  222. String rowSpan = attributes.getValue("number-rows-spanned");
  223. if (rowSpan != null) {
  224. structElem.setTableAttributeRowSpan(Integer.parseInt(rowSpan));
  225. }
  226. }
  227. }
  228. private static class PlaceholderBuilder implements StructureElementBuilder {
  229. public PDFStructElem build(StructureHierarchyMember parent, Attributes attributes,
  230. PDFFactory pdfFactory, EventBroadcaster eventBroadcaster) {
  231. PDFStructElem elem = new PDFStructElem.Placeholder(parent);
  232. parent.addKid(elem);
  233. return elem;
  234. }
  235. }
  236. private PDFFactory pdfFactory;
  237. private EventBroadcaster eventBroadcaster;
  238. private LinkedList<PDFStructElem> ancestors = new LinkedList<PDFStructElem>();
  239. private PDFStructElem rootStructureElement;
  240. void setPdfFactory(PDFFactory pdfFactory) {
  241. this.pdfFactory = pdfFactory;
  242. }
  243. void setEventBroadcaster(EventBroadcaster eventBroadcaster) {
  244. this.eventBroadcaster = eventBroadcaster;
  245. }
  246. void setLogicalStructureHandler(PDFLogicalStructureHandler logicalStructureHandler) {
  247. createRootStructureElement(logicalStructureHandler);
  248. }
  249. private void createRootStructureElement(PDFLogicalStructureHandler logicalStructureHandler) {
  250. assert rootStructureElement == null;
  251. PDFParentTree parentTree = logicalStructureHandler.getParentTree();
  252. PDFStructTreeRoot structTreeRoot = pdfFactory.getDocument().makeStructTreeRoot(parentTree);
  253. rootStructureElement = createStructureElement("root", structTreeRoot,
  254. new AttributesImpl(), pdfFactory, eventBroadcaster);
  255. }
  256. private static PDFStructElem createStructureElement(String name, StructureHierarchyMember parent,
  257. Attributes attributes, PDFFactory pdfFactory, EventBroadcaster eventBroadcaster) {
  258. StructureElementBuilder builder = BUILDERS.get(name);
  259. if (builder == null) {
  260. // TODO is a fallback really necessary?
  261. builder = DEFAULT_BUILDER;
  262. }
  263. return builder.build(parent, attributes, pdfFactory, eventBroadcaster);
  264. }
  265. public void startPageSequence(Locale language, String role) {
  266. ancestors = new LinkedList<PDFStructElem>();
  267. AttributesImpl attributes = new AttributesImpl();
  268. attributes.addAttribute("", ROLE, ROLE, XMLUtil.CDATA, role);
  269. PDFStructElem structElem = createStructureElement("page-sequence",
  270. rootStructureElement, attributes, pdfFactory, eventBroadcaster);
  271. if (language != null) {
  272. structElem.setLanguage(language);
  273. }
  274. ancestors.add(structElem);
  275. }
  276. public void endPageSequence() {
  277. }
  278. public StructureTreeElement startNode(String name, Attributes attributes) {
  279. PDFStructElem parent = ancestors.getFirst();
  280. PDFStructElem structElem = createStructureElement(name, parent, attributes,
  281. pdfFactory, eventBroadcaster);
  282. ancestors.addFirst(structElem);
  283. return structElem;
  284. }
  285. public void endNode(String name) {
  286. ancestors.removeFirst();
  287. }
  288. public StructureTreeElement startImageNode(String name, Attributes attributes) {
  289. return startNode(name, attributes);
  290. }
  291. public StructureTreeElement startReferencedNode(String name, Attributes attributes) {
  292. return startNode(name, attributes);
  293. }
  294. }