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

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357
  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.extensions.InternalElementMapping;
  29. import org.apache.fop.fo.pagination.Flow;
  30. import org.apache.fop.pdf.PDFFactory;
  31. import org.apache.fop.pdf.PDFParentTree;
  32. import org.apache.fop.pdf.PDFStructElem;
  33. import org.apache.fop.pdf.PDFStructTreeRoot;
  34. import org.apache.fop.pdf.StandardStructureAttributes.Table.Scope;
  35. import org.apache.fop.pdf.StandardStructureTypes;
  36. import org.apache.fop.pdf.StandardStructureTypes.Grouping;
  37. import org.apache.fop.pdf.StandardStructureTypes.Table;
  38. import org.apache.fop.pdf.StructureHierarchyMember;
  39. import org.apache.fop.pdf.StructureType;
  40. import org.apache.fop.util.XMLUtil;
  41. class PDFStructureTreeBuilder implements StructureTreeEventHandler {
  42. private static final String ROLE = "role";
  43. private static final Map<String, StructureElementBuilder> BUILDERS
  44. = new java.util.HashMap<String, StructureElementBuilder>();
  45. private static final StructureElementBuilder DEFAULT_BUILDER
  46. = new DefaultStructureElementBuilder(Grouping.NON_STRUCT);
  47. static {
  48. // Declarations and Pagination and Layout Formatting Objects
  49. StructureElementBuilder regionBuilder = new RegionBuilder();
  50. addBuilder("root", StandardStructureTypes.Grouping.DOCUMENT);
  51. addBuilder("page-sequence", new PageSequenceBuilder());
  52. addBuilder("static-content", regionBuilder);
  53. addBuilder("flow", regionBuilder);
  54. // Block-level Formatting Objects
  55. addBuilder("block", StandardStructureTypes.Paragraphlike.P);
  56. addBuilder("block-container", StandardStructureTypes.Grouping.DIV);
  57. // Inline-level Formatting Objects
  58. addBuilder("character", StandardStructureTypes.InlineLevelStructure.SPAN);
  59. addBuilder("external-graphic", new ImageBuilder());
  60. addBuilder("instream-foreign-object", new ImageBuilder());
  61. addBuilder("inline", StandardStructureTypes.InlineLevelStructure.SPAN);
  62. addBuilder("inline-container", StandardStructureTypes.Grouping.DIV);
  63. addBuilder("page-number", StandardStructureTypes.InlineLevelStructure.QUOTE);
  64. addBuilder("page-number-citation", StandardStructureTypes.InlineLevelStructure.QUOTE);
  65. addBuilder("page-number-citation-last", StandardStructureTypes.InlineLevelStructure.QUOTE);
  66. // Formatting Objects for Tables
  67. addBuilder("table-and-caption", StandardStructureTypes.Grouping.DIV);
  68. addBuilder("table", new TableBuilder());
  69. addBuilder("table-caption", StandardStructureTypes.Grouping.CAPTION);
  70. addBuilder("table-header", StandardStructureTypes.Table.THEAD);
  71. addBuilder("table-footer", new TableFooterBuilder());
  72. addBuilder("table-body", StandardStructureTypes.Table.TBODY);
  73. addBuilder("table-row", StandardStructureTypes.Table.TR);
  74. addBuilder("table-cell", new TableCellBuilder());
  75. // Formatting Objects for Lists
  76. addBuilder("list-block", StandardStructureTypes.List.L);
  77. addBuilder("list-item", StandardStructureTypes.List.LI);
  78. addBuilder("list-item-body", StandardStructureTypes.List.LBODY);
  79. addBuilder("list-item-label", StandardStructureTypes.List.LBL);
  80. // Dynamic Effects: Link and Multi Formatting Objects
  81. addBuilder("basic-link", StandardStructureTypes.InlineLevelStructure.LINK);
  82. // Out-of-Line Formatting Objects
  83. addBuilder("float", StandardStructureTypes.Grouping.DIV);
  84. addBuilder("footnote", StandardStructureTypes.InlineLevelStructure.NOTE);
  85. addBuilder("footnote-body", StandardStructureTypes.Grouping.SECT);
  86. addBuilder("wrapper", StandardStructureTypes.InlineLevelStructure.SPAN);
  87. addBuilder("marker", StandardStructureTypes.Grouping.PRIVATE);
  88. addBuilder("#PCDATA", new PlaceholderBuilder());
  89. }
  90. private static void addBuilder(String fo, StructureType structureType) {
  91. addBuilder(fo, new DefaultStructureElementBuilder(structureType));
  92. }
  93. private static void addBuilder(String fo, StructureElementBuilder mapper) {
  94. BUILDERS.put(fo, mapper);
  95. }
  96. private interface StructureElementBuilder {
  97. PDFStructElem build(StructureHierarchyMember parent, Attributes attributes, PDFFactory pdfFactory,
  98. EventBroadcaster eventBroadcaster);
  99. }
  100. private static class DefaultStructureElementBuilder implements StructureElementBuilder {
  101. private final StructureType defaultStructureType;
  102. DefaultStructureElementBuilder(StructureType structureType) {
  103. this.defaultStructureType = structureType;
  104. }
  105. public final PDFStructElem build(StructureHierarchyMember parent, Attributes attributes,
  106. PDFFactory pdfFactory, EventBroadcaster eventBroadcaster) {
  107. String role = attributes.getValue(ROLE);
  108. StructureType structureType;
  109. if (role == null) {
  110. structureType = defaultStructureType;
  111. } else {
  112. structureType = StandardStructureTypes.get(role);
  113. if (structureType == null) {
  114. structureType = defaultStructureType;
  115. PDFEventProducer.Provider.get(eventBroadcaster).nonStandardStructureType(role, role,
  116. structureType.toString());
  117. }
  118. }
  119. PDFStructElem structElem = createStructureElement(parent, structureType);
  120. setAttributes(structElem, attributes);
  121. addKidToParent(structElem, parent, attributes);
  122. registerStructureElement(structElem, pdfFactory, attributes);
  123. return structElem;
  124. }
  125. protected PDFStructElem createStructureElement(StructureHierarchyMember parent,
  126. StructureType structureType) {
  127. return new PDFStructElem(parent, structureType);
  128. }
  129. protected void setAttributes(PDFStructElem structElem, Attributes attributes) {
  130. }
  131. protected void addKidToParent(PDFStructElem kid, StructureHierarchyMember parent,
  132. Attributes attributes) {
  133. parent.addKid(kid);
  134. }
  135. protected void registerStructureElement(PDFStructElem structureElement, PDFFactory pdfFactory,
  136. Attributes attributes) {
  137. pdfFactory.getDocument().registerStructureElement(structureElement);
  138. }
  139. }
  140. private static class PageSequenceBuilder extends DefaultStructureElementBuilder {
  141. PageSequenceBuilder() {
  142. super(StandardStructureTypes.Grouping.PART);
  143. }
  144. @Override
  145. protected PDFStructElem createStructureElement(StructureHierarchyMember parent,
  146. StructureType structureType) {
  147. return new PageSequenceStructElem(parent, structureType);
  148. }
  149. }
  150. private static class RegionBuilder extends DefaultStructureElementBuilder {
  151. RegionBuilder() {
  152. super(StandardStructureTypes.Grouping.SECT);
  153. }
  154. @Override
  155. protected void addKidToParent(PDFStructElem kid, StructureHierarchyMember parent,
  156. Attributes attributes) {
  157. String flowName = attributes.getValue(Flow.FLOW_NAME);
  158. ((PageSequenceStructElem) parent).addContent(flowName, kid);
  159. }
  160. }
  161. private static class ImageBuilder extends DefaultStructureElementBuilder {
  162. ImageBuilder() {
  163. super(StandardStructureTypes.Illustration.FIGURE);
  164. }
  165. @Override
  166. protected void setAttributes(PDFStructElem structElem, Attributes attributes) {
  167. String altTextNode = attributes.getValue(ExtensionElementMapping.URI, "alt-text");
  168. if (altTextNode == null) {
  169. altTextNode = "No alternate text specified";
  170. }
  171. structElem.put("Alt", altTextNode);
  172. }
  173. }
  174. private static class TableBuilder extends DefaultStructureElementBuilder {
  175. TableBuilder() {
  176. super(StandardStructureTypes.Table.TABLE);
  177. }
  178. @Override
  179. protected PDFStructElem createStructureElement(StructureHierarchyMember parent,
  180. StructureType structureType) {
  181. return new TableStructElem(parent, structureType);
  182. }
  183. }
  184. private static class TableFooterBuilder extends DefaultStructureElementBuilder {
  185. public TableFooterBuilder() {
  186. super(StandardStructureTypes.Table.TFOOT);
  187. }
  188. @Override
  189. protected void addKidToParent(PDFStructElem kid, StructureHierarchyMember parent,
  190. Attributes attributes) {
  191. ((TableStructElem) parent).addTableFooter(kid);
  192. }
  193. }
  194. private static class TableCellBuilder extends DefaultStructureElementBuilder {
  195. TableCellBuilder() {
  196. super(StandardStructureTypes.Table.TD);
  197. }
  198. @Override
  199. protected void registerStructureElement(PDFStructElem structureElement, PDFFactory pdfFactory,
  200. Attributes attributes) {
  201. if (structureElement.getStructureType() == Table.TH) {
  202. String scopeAttribute = attributes.getValue(InternalElementMapping.URI,
  203. InternalElementMapping.SCOPE);
  204. Scope scope = (scopeAttribute == null)
  205. ? Scope.COLUMN
  206. : Scope.valueOf(scopeAttribute.toUpperCase(Locale.ENGLISH));
  207. pdfFactory.getDocument().registerStructureElement(structureElement, scope);
  208. } else {
  209. pdfFactory.getDocument().registerStructureElement(structureElement);
  210. }
  211. }
  212. @Override
  213. protected void setAttributes(PDFStructElem structElem, Attributes attributes) {
  214. String columnSpan = attributes.getValue("number-columns-spanned");
  215. if (columnSpan != null) {
  216. structElem.setTableAttributeColSpan(Integer.parseInt(columnSpan));
  217. }
  218. String rowSpan = attributes.getValue("number-rows-spanned");
  219. if (rowSpan != null) {
  220. structElem.setTableAttributeRowSpan(Integer.parseInt(rowSpan));
  221. }
  222. }
  223. }
  224. private static class PlaceholderBuilder implements StructureElementBuilder {
  225. public PDFStructElem build(StructureHierarchyMember parent, Attributes attributes,
  226. PDFFactory pdfFactory, EventBroadcaster eventBroadcaster) {
  227. PDFStructElem elem = new PDFStructElem.Placeholder(parent);
  228. parent.addKid(elem);
  229. return elem;
  230. }
  231. }
  232. private PDFFactory pdfFactory;
  233. private EventBroadcaster eventBroadcaster;
  234. private LinkedList<PDFStructElem> ancestors = new LinkedList<PDFStructElem>();
  235. private PDFStructElem rootStructureElement;
  236. void setPdfFactory(PDFFactory pdfFactory) {
  237. this.pdfFactory = pdfFactory;
  238. }
  239. void setEventBroadcaster(EventBroadcaster eventBroadcaster) {
  240. this.eventBroadcaster = eventBroadcaster;
  241. }
  242. void setLogicalStructureHandler(PDFLogicalStructureHandler logicalStructureHandler) {
  243. createRootStructureElement(logicalStructureHandler);
  244. }
  245. private void createRootStructureElement(PDFLogicalStructureHandler logicalStructureHandler) {
  246. assert rootStructureElement == null;
  247. PDFParentTree parentTree = logicalStructureHandler.getParentTree();
  248. PDFStructTreeRoot structTreeRoot = pdfFactory.getDocument().makeStructTreeRoot(parentTree);
  249. rootStructureElement = createStructureElement("root", structTreeRoot,
  250. new AttributesImpl(), pdfFactory, eventBroadcaster);
  251. }
  252. private static PDFStructElem createStructureElement(String name, StructureHierarchyMember parent,
  253. Attributes attributes, PDFFactory pdfFactory, EventBroadcaster eventBroadcaster) {
  254. StructureElementBuilder builder = BUILDERS.get(name);
  255. if (builder == null) {
  256. // TODO is a fallback really necessary?
  257. builder = DEFAULT_BUILDER;
  258. }
  259. return builder.build(parent, attributes, pdfFactory, eventBroadcaster);
  260. }
  261. public void startPageSequence(Locale language, String role) {
  262. ancestors = new LinkedList<PDFStructElem>();
  263. AttributesImpl attributes = new AttributesImpl();
  264. attributes.addAttribute("", ROLE, ROLE, XMLUtil.CDATA, role);
  265. PDFStructElem structElem = createStructureElement("page-sequence",
  266. rootStructureElement, attributes, pdfFactory, eventBroadcaster);
  267. if (language != null) {
  268. structElem.setLanguage(language);
  269. }
  270. ancestors.add(structElem);
  271. }
  272. public void endPageSequence() {
  273. }
  274. public StructureTreeElement startNode(String name, Attributes attributes) {
  275. PDFStructElem parent = ancestors.getFirst();
  276. PDFStructElem structElem = createStructureElement(name, parent, attributes,
  277. pdfFactory, eventBroadcaster);
  278. ancestors.addFirst(structElem);
  279. return structElem;
  280. }
  281. public void endNode(String name) {
  282. ancestors.removeFirst();
  283. }
  284. public StructureTreeElement startImageNode(String name, Attributes attributes) {
  285. return startNode(name, attributes);
  286. }
  287. public StructureTreeElement startReferencedNode(String name, Attributes attributes) {
  288. return startNode(name, attributes);
  289. }
  290. }