fbreader/src/formats/fb2/FB2BookReader.cpp

   1 /*
   2  * Copyright (C) 2004-2008 Geometer Plus <contact@geometerplus.com>
   3  *
   4  * This program is free software; you can redistribute it and/or modify
   5  * it under the terms of the GNU General Public License as published by
   6  * the Free Software Foundation; either version 2 of the License, or
   7  * (at your option) any later version.
   8  *
   9  * This program is distributed in the hope that it will be useful,
  10  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  11  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12  * GNU General Public License for more details.
  13  *
  14  * You should have received a copy of the GNU General Public License
  15  * along with this program; if not, write to the Free Software
  16  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
  17  * 02110-1301, USA.
  18  */
  19
  20 #include <stdlib.h>
  21 #include <string.h>
  22
  23 #include <ZLInputStream.h>
  24 #include <ZLStringUtil.h>
  25
  26 #include <ZLTextParagraph.h>
  27
  28 #include "FB2BookReader.h"
  29 #include "Base64EncodedImage.h"
  30 #include "../../bookmodel/BookModel.h"
  31
  32 FB2BookReader::FB2BookReader(BookModel &model) : myModelReader(model) {
  33         myInsideCoverpage = false;
  34         myParagraphsBeforeBodyNumber = (size_t)-1;
  35         myInsidePoem = false;
  36         mySectionDepth = 0;
  37         myBodyCounter = 0;
  38         myReadMainText = false;
  39         myCurrentImage = 0;
  40         myProcessingImage = false;
  41         mySectionStarted = false;
  42         myInsideTitle = false;
  43 }
  44
  45 void FB2BookReader::characterDataHandler(const char *text, int len) {
  46         if ((len > 0) && (myProcessingImage || myModelReader.paragraphIsOpen())) {
  47                 std::string str(text, len);
  48                 if (myProcessingImage) {
  49                         myImageBuffer.push_back(str);
  50                 } else {
  51                         myModelReader.addData(str);
  52                         if (myInsideTitle) {
  53                                 myModelReader.addContentsData(str);
  54                         }
  55                 }
  56         }
  57 }
  58
  59 void FB2BookReader::startElementHandler(int tag, const char **xmlattributes) {
  60         const char *id = attributeValue(xmlattributes, "id");
  61         if (id != 0) {
  62                 if (!myReadMainText) {
  63                         myModelReader.setFootnoteTextModel(id);
  64                 }
  65                 myModelReader.addHyperlinkLabel(id);
  66         }
  67         switch (tag) {
  68                 case _P:
  69                         if (mySectionStarted) {
  70                                 mySectionStarted = false;
  71                         } else if (myInsideTitle) {
  72                                 static const std::string SPACE = " ";
  73                                 myModelReader.addContentsData(SPACE);
  74                         }
  75                         myModelReader.beginParagraph();
  76                         break;
  77                 case _V:
  78                         myModelReader.pushKind(VERSE);
  79                         myModelReader.beginParagraph();
  80                         break;
  81                 case _SUBTITLE:
  82                         myModelReader.pushKind(SUBTITLE);
  83                         myModelReader.beginParagraph();
  84                         break;
  85                 case _TEXT_AUTHOR:
  86                         myModelReader.pushKind(AUTHOR);
  87                         myModelReader.beginParagraph();
  88                         break;
  89                 case _DATE:
  90                         myModelReader.pushKind(DATE);
  91                         myModelReader.beginParagraph();
  92                         break;
  93                 case _CITE:
  94                         myModelReader.pushKind(CITE);
  95                         break;
  96                 case _SECTION:
  97                         if (myReadMainText) {
  98                                 myModelReader.insertEndOfSectionParagraph();
  99                                 ++mySectionDepth;
 100                                 myModelReader.beginContentsParagraph();
 101                                 mySectionStarted = true;
 102                         }
 103                         break;
 104                 case _TITLE:
 105                         if (myInsidePoem) {
 106                                 myModelReader.pushKind(POEM_TITLE);
 107                         } else if (mySectionDepth == 0) {
 108                                 myModelReader.insertEndOfSectionParagraph();
 109                                 myModelReader.pushKind(TITLE);
 110                         } else {
 111                                 myModelReader.pushKind(SECTION_TITLE);
 112                                 myModelReader.enterTitle();
 113                                 myInsideTitle = true;
 114                         }
 115                         break;
 116                 case _POEM:
 117                         myInsidePoem = true;
 118                         break;
 119                 case _STANZA:
 120                         myModelReader.pushKind(STANZA);
 121                         myModelReader.beginParagraph(ZLTextParagraph::BEFORE_SKIP_PARAGRAPH);
 122                         myModelReader.endParagraph();
 123                         break;
 124                 case _EPIGRAPH:
 125                         myModelReader.pushKind(EPIGRAPH);
 126                         break;
 127                 case _ANNOTATION:
 128                         if (myBodyCounter == 0) {
 129                                 myModelReader.setMainTextModel();
 130                         }
 131                         myModelReader.pushKind(ANNOTATION);
 132                         break;
 133                 case _COVERPAGE:
 134                         if (myBodyCounter == 0) {
 135                                 myInsideCoverpage = true;
 136                                 myModelReader.setMainTextModel();
 137                         }
 138                         break;
 139                 case _SUB:
 140                         myModelReader.addControl(SUB, true);
 141                         break;
 142                 case _SUP:
 143                         myModelReader.addControl(SUP, true);
 144                         break;
 145                 case _CODE:
 146                         myModelReader.addControl(CODE, true);
 147                         break;
 148                 case _STRIKETHROUGH:
 149                         myModelReader.addControl(STRIKETHROUGH, true);
 150                         break;
 151                 case _STRONG:
 152                         myModelReader.addControl(STRONG, true);
 153                         break;
 154                 case _EMPHASIS:
 155                         myModelReader.addControl(EMPHASIS, true);
 156                         break;
 157                 case _A:
 158                 {
 159                         const char *ref = attributeValue(xmlattributes, myHrefAttributeName.c_str());
 160                         if (ref != 0) {
 161                                 const char *type = attributeValue(xmlattributes, "type");
 162                                 if((type != 0 && !strncmp(type, "note", 4)) || strstr(ref, "note"))
 163                                         myHyperlinkType = FOOTNOTE;
 164                                 else if(ref[0] == '#')
 165                                         myHyperlinkType = INTERNAL_HYPERLINK;
 166                                 else
 167                                         myHyperlinkType = EXTERNAL_HYPERLINK;
 168
 169                                 if(ref[0] == '#')
 170                                         ref++;
 171 /*      wrong
 172  *                              if (ref[0] == '#') {
 173                                         myHyperlinkType = FOOTNOTE;
 174                                         ++ref;
 175                                 } else {
 176                                         myHyperlinkType = EXTERNAL_HYPERLINK;
 177                                 }
 178 */
 179                                 myModelReader.addHyperlinkControl(myHyperlinkType, ref);
 180                         } else {
 181                                 myHyperlinkType = FOOTNOTE;
 182                                 myModelReader.addControl(myHyperlinkType, true);
 183                         }
 184                         break;
 185                 }
 186                 case _IMAGE:
 187                 {
 188                         const char *ref = attributeValue(xmlattributes, myHrefAttributeName.c_str());
 189                         const char *vOffset = attributeValue(xmlattributes, "voffset");
 190                         char offset = (vOffset != 0) ? atoi(vOffset) : 0;
 191                         if ((ref != 0) && (*ref == '#')) {
 192                                 ++ref;
 193                                 if ((myCoverImageReference != ref) ||
 194                                                 (myParagraphsBeforeBodyNumber != myModelReader.model().bookTextModel()->paragraphsNumber())) {
 195                                         myModelReader.addImageReference(ref);
 196                                 }
 197                                 if (myInsideCoverpage) {
 198                                         myCoverImageReference = ref;
 199                                 }
 200                         }
 201                         break;
 202                 }
 203                 case _BINARY:
 204                 {
 205                         const char *contentType = attributeValue(xmlattributes, "content-type");
 206                         if ((contentType != 0) && (id != 0)) {
 207                                 myCurrentImage = new Base64EncodedImage(contentType);
 208                                 myModelReader.addImage(id, myCurrentImage);
 209                                 myProcessingImage = true;
 210                         }
 211                         break;
 212                 }
 213                 case _EMPTY_LINE:
 214                         myModelReader.beginParagraph(ZLTextParagraph::EMPTY_LINE_PARAGRAPH);
 215                         myModelReader.endParagraph();
 216                         break;
 217                 case _BODY:
 218                         ++myBodyCounter;
 219                         myParagraphsBeforeBodyNumber = myModelReader.model().bookTextModel()->paragraphsNumber();
 220                         if ((myBodyCounter == 1) || (attributeValue(xmlattributes, "name") == 0)) {
 221                                 myModelReader.setMainTextModel();
 222                                 myReadMainText = true;
 223                         }
 224                         myModelReader.pushKind(REGULAR);
 225                         break;
 226                 default:
 227                         break;
 228         }
 229 }
 230
 231 void FB2BookReader::endElementHandler(int tag) {
 232         switch (tag) {
 233                 case _P:
 234                         myModelReader.endParagraph();
 235                         break;
 236                 case _V:
 237                 case _SUBTITLE:
 238                 case _TEXT_AUTHOR:
 239                 case _DATE:
 240                         myModelReader.popKind();
 241                         myModelReader.endParagraph();
 242                         break;
 243                 case _CITE:
 244                         myModelReader.popKind();
 245                         break;
 246                 case _SECTION:
 247                         if (myReadMainText) {
 248                                 myModelReader.endContentsParagraph();
 249                                 --mySectionDepth;
 250                                 mySectionStarted = false;
 251                         } else {
 252                                 myModelReader.unsetTextModel();
 253                         }
 254                         break;
 255                 case _TITLE:
 256                         myModelReader.exitTitle();
 257                         myModelReader.popKind();
 258                         myInsideTitle = false;
 259                         break;
 260                 case _POEM:
 261                         myInsidePoem = false;
 262                         break;
 263                 case _STANZA:
 264                         myModelReader.beginParagraph(ZLTextParagraph::AFTER_SKIP_PARAGRAPH);
 265                         myModelReader.endParagraph();
 266                         myModelReader.popKind();
 267                         break;
 268                 case _EPIGRAPH:
 269                         myModelReader.popKind();
 270                         break;
 271                 case _ANNOTATION:
 272                         myModelReader.popKind();
 273                         if (myBodyCounter == 0) {
 274                                 myModelReader.insertEndOfSectionParagraph();
 275                                 myModelReader.unsetTextModel();
 276                         }
 277                         break;
 278                 case _COVERPAGE:
 279                         if (myBodyCounter == 0) {
 280                                 myInsideCoverpage = false;
 281                                 myModelReader.insertEndOfSectionParagraph();
 282                                 myModelReader.unsetTextModel();
 283                         }
 284                         break;
 285                 case _SUB:
 286                         myModelReader.addControl(SUB, false);
 287                         break;
 288                 case _SUP:
 289                         myModelReader.addControl(SUP, false);
 290                         break;
 291                 case _CODE:
 292                         myModelReader.addControl(CODE, false);
 293                         break;
 294                 case _STRIKETHROUGH:
 295                         myModelReader.addControl(STRIKETHROUGH, false);
 296                         break;
 297                 case _STRONG:
 298                         myModelReader.addControl(STRONG, false);
 299                         break;
 300                 case _EMPHASIS:
 301                         myModelReader.addControl(EMPHASIS, false);
 302                         break;
 303                 case _A:
 304                         myModelReader.addControl(myHyperlinkType, false);
 305                         break;
 306                 case _BINARY:
 307                         if (!myImageBuffer.empty() && (myCurrentImage != 0)) {
 308                                 myCurrentImage->addData(myImageBuffer);
 309                                 myImageBuffer.clear();
 310                                 myCurrentImage = 0;
 311                         }
 312                         myProcessingImage = false;
 313                         break;
 314                 case _BODY:
 315                         myModelReader.popKind();
 316                         myModelReader.unsetTextModel();
 317                         myReadMainText = false;
 318                         break;
 319                 default:
 320                         break;
 321         }
 322 }
 323
 324 bool FB2BookReader::processNamespaces() const {
 325         return true;
 326 }
 327
 328 void FB2BookReader::namespaceListChangedHandler() {
 329         const std::string XLINK_REFERENCE = "http://www.w3.org/1999/xlink";
 330         const std::map<std::string,std::string> namespaceMap = namespaces();
 331         for (std::map<std::string,std::string>::const_iterator it = namespaceMap.begin(); it != namespaceMap.end(); ++it) {
 332                 if (ZLStringUtil::stringStartsWith(it->second, XLINK_REFERENCE)) {
 333                         myHrefAttributeName = it->first + ":href";
 334                         return;
 335                 }
 336         }
 337         myHrefAttributeName = "";
 338 }
 339
 340 bool FB2BookReader::readBook(const std::string &fileName) {
 341         myHrefAttributeName = "";
 342         return readDocument(fileName);
 343 }