Beispiel #1
0
  /**
   * Get a Part (except a relationships part), but not its relationships part or related parts.
   * Useful if you need quick access to just this part. This can be called directly from outside the
   * library, in which case the Part will not be owned by a Package until the calling code makes it
   * so.
   *
   * @see To get a Part and all its related parts, and add all to a package, use getPart.
   * @param zf
   * @param resolvedPartUri
   * @return
   * @throws URISyntaxException
   * @throws InvalidFormatException
   */
  public static Part getRawPart(
      ZipFile zf, ContentTypeManager ctm, String resolvedPartUri, Relationship rel)
      throws Docx4JException {
    Part part = null;

    InputStream is = null;
    try {
      try {
        log.debug("resolved uri: " + resolvedPartUri);
        is = getInputStreamFromZippedPart(zf, resolvedPartUri);

        // Get a subclass of Part appropriate for this content type
        // This will throw UnrecognisedPartException in the absence of
        // specific knowledge. Hence it is important to get the is
        // first, as we do above.
        part = ctm.getPart("/" + resolvedPartUri, rel);

        if (part instanceof org.docx4j.openpackaging.parts.ThemePart) {

          ((org.docx4j.openpackaging.parts.JaxbXmlPart) part).setJAXBContext(Context.jcThemePart);
          ((org.docx4j.openpackaging.parts.JaxbXmlPart) part).unmarshal(is);

        } else if (part instanceof org.docx4j.openpackaging.parts.DocPropsCorePart) {

          ((org.docx4j.openpackaging.parts.JaxbXmlPart) part)
              .setJAXBContext(Context.jcDocPropsCore);
          ((org.docx4j.openpackaging.parts.JaxbXmlPart) part).unmarshal(is);

        } else if (part instanceof org.docx4j.openpackaging.parts.DocPropsCustomPart) {

          ((org.docx4j.openpackaging.parts.JaxbXmlPart) part)
              .setJAXBContext(Context.jcDocPropsCustom);
          ((org.docx4j.openpackaging.parts.JaxbXmlPart) part).unmarshal(is);

        } else if (part instanceof org.docx4j.openpackaging.parts.DocPropsExtendedPart) {

          ((org.docx4j.openpackaging.parts.JaxbXmlPart) part)
              .setJAXBContext(Context.jcDocPropsExtended);
          ((org.docx4j.openpackaging.parts.JaxbXmlPart) part).unmarshal(is);

        } else if (part
            instanceof org.docx4j.openpackaging.parts.CustomXmlDataStoragePropertiesPart) {

          ((org.docx4j.openpackaging.parts.JaxbXmlPart) part)
              .setJAXBContext(Context.jcCustomXmlProperties);
          ((org.docx4j.openpackaging.parts.JaxbXmlPart) part).unmarshal(is);

        } else if (part
            instanceof org.docx4j.openpackaging.parts.digitalsignature.XmlSignaturePart) {

          ((org.docx4j.openpackaging.parts.JaxbXmlPart) part).setJAXBContext(Context.jcXmlDSig);
          ((org.docx4j.openpackaging.parts.JaxbXmlPart) part).unmarshal(is);

        } else if (part instanceof org.docx4j.openpackaging.parts.JaxbXmlPart) {

          // MainDocument part, Styles part, Font part etc

          ((org.docx4j.openpackaging.parts.JaxbXmlPart) part).setJAXBContext(Context.jc);
          ((org.docx4j.openpackaging.parts.JaxbXmlPart) part).unmarshal(is);

        } else if (part instanceof org.docx4j.openpackaging.parts.WordprocessingML.BinaryPart) {

          log.debug("Detected BinaryPart " + part.getClass().getName());
          if (conserveMemory) {
            ((BinaryPart) part).setBinaryDataRef(zf.getName(), resolvedPartUri);
          } else {
            ((BinaryPart) part).setBinaryData(is);
          }

        } else if (part instanceof org.docx4j.openpackaging.parts.CustomXmlDataStoragePart) {

          // Is it a part we know?
          try {
            Unmarshaller u = Context.jc.createUnmarshaller();
            Object o = u.unmarshal(is);
            log.debug(o.getClass().getName());

            PartName name = part.getPartName();

            if (o instanceof org.opendope.conditions.Conditions) {

              part = new ConditionsPart(name);
              ((ConditionsPart) part).setJaxbElement((org.opendope.conditions.Conditions) o);

            } else if (o instanceof org.opendope.xpaths.Xpaths) {

              part = new XPathsPart(name);
              ((XPathsPart) part).setJaxbElement((org.opendope.xpaths.Xpaths) o);

            } else if (o instanceof org.opendope.questions.Questionnaire) {

              part = new QuestionsPart(name);
              ((QuestionsPart) part).setJaxbElement((org.opendope.questions.Questionnaire) o);

            } else if (o instanceof org.opendope.answers.Answers) {

              part = new StandardisedAnswersPart(name);
              ((StandardisedAnswersPart) part).setJaxbElement((org.opendope.answers.Answers) o);

            } else if (o instanceof org.opendope.components.Components) {

              part = new ComponentsPart(name);
              ((ComponentsPart) part).setJaxbElement((org.opendope.components.Components) o);

            } else if (o instanceof JAXBElement<?>
                && XmlUtils.unwrap(o) instanceof org.docx4j.bibliography.CTSources) {
              part = new BibliographyPart(name);
              ((BibliographyPart) part)
                  .setJaxbElement((JAXBElement<org.docx4j.bibliography.CTSources>) o);

            } else {

              log.warn("No known part after all for CustomXmlPart " + o.getClass().getName());

              CustomXmlDataStorage data = getCustomXmlDataStorageClass().factory();
              is.reset();
              data.setDocument(is); // Not necessarily JAXB, that's just our method name
              ((org.docx4j.openpackaging.parts.CustomXmlDataStoragePart) part).setData(data);
            }

          } catch (javax.xml.bind.UnmarshalException ue) {

            // No ...
            CustomXmlDataStorage data = getCustomXmlDataStorageClass().factory();
            is.reset();
            data.setDocument(is); // Not necessarily JAXB, that's just our method name
            ((org.docx4j.openpackaging.parts.CustomXmlDataStoragePart) part).setData(data);
          }

        } else if (part instanceof org.docx4j.openpackaging.parts.XmlPart) {

          try {
            ((XmlPart) part).setDocument(is);
          } catch (Docx4JException d) {
            // This isn't an XML part after all,
            // even though ContentTypeManager detected it as such
            // So get it as a binary part
            part = getBinaryPart(zf, ctm, resolvedPartUri);
            if (conserveMemory) {
              ((BinaryPart) part).setBinaryDataRef(zf.getName(), resolvedPartUri);
            } else {
              ((BinaryPart) part).setBinaryData(is);
            }
          }

        } else {
          // Shouldn't happen, since ContentTypeManagerImpl should
          // return an instance of one of the above, or throw an
          // Exception.

          log.error("No suitable part found for: " + resolvedPartUri);
          part = null;
        }

      } catch (PartUnrecognisedException e) {
        log.warn("PartUnrecognisedException shouldn't happen anymore!");
        // Try to get it as a binary part
        part = getBinaryPart(zf, ctm, resolvedPartUri);
        if (conserveMemory) {
          ((BinaryPart) part).setBinaryDataRef(zf.getName(), resolvedPartUri);
        } else {
          ((BinaryPart) part).setBinaryData(is);
        }
      }
    } catch (Exception ex) {
      // IOException, URISyntaxException
      ex.printStackTrace();
      throw new Docx4JException("Failed to getPart", ex);

    } finally {
      if (is != null) {
        try {
          is.close();
        } catch (IOException exc) {
          exc.printStackTrace();
        }
      }
    }
    return part;
  }
Beispiel #2
0
  /**
   * Get a Part (except a relationships part), but not its relationships part or related parts.
   * Useful if you need quick access to just this part. This can be called directly from outside the
   * library, in which case the Part will not be owned by a Package until the calling code makes it
   * so.
   *
   * @see To get a Part and all its related parts, and add all to a package, use getPart.
   * @param partByteArrays
   * @param ctm
   * @param resolvedPartUri
   * @param rel
   * @return
   * @throws Docx4JException including if result is null
   */
  public Part getRawPart(ContentTypeManager ctm, String resolvedPartUri, Relationship rel)
      throws Docx4JException {

    Part part = null;

    InputStream is = null;
    try {
      try {
        log.debug("resolved uri: " + resolvedPartUri);

        // Get a subclass of Part appropriate for this content type
        // This will throw UnrecognisedPartException in the absence of
        // specific knowledge. Hence it is important to get the is
        // first, as we do above.
        part = ctm.getPart("/" + resolvedPartUri, rel);

        log.info("ctm returned " + part.getClass().getName());

        if (part instanceof org.docx4j.openpackaging.parts.ThemePart
            || part instanceof org.docx4j.openpackaging.parts.DocPropsCorePart
            || part instanceof org.docx4j.openpackaging.parts.DocPropsCustomPart
            || part instanceof org.docx4j.openpackaging.parts.DocPropsExtendedPart
            || part instanceof org.docx4j.openpackaging.parts.CustomXmlDataStoragePropertiesPart
            || part instanceof org.docx4j.openpackaging.parts.digitalsignature.XmlSignaturePart
            || part instanceof org.docx4j.openpackaging.parts.JaxbXmlPart) {

          // Nothing to do here

        } else if (part instanceof org.docx4j.openpackaging.parts.WordprocessingML.BinaryPart) {

          log.debug("Detected BinaryPart " + part.getClass().getName());
          //					is = partStore.loadPart( resolvedPartUri);
          //					((BinaryPart)part).setBinaryData(is);

        } else if (part instanceof org.docx4j.openpackaging.parts.CustomXmlDataStoragePart) {
          // ContentTypeManager initially detects them as CustomXmlDataStoragePart;
          // the below changes as necessary

          // Is it a part we know?
          is = partStore.loadPart(resolvedPartUri);
          try {
            Unmarshaller u = Context.jc.createUnmarshaller();
            Object o = u.unmarshal(is);
            log.debug(o.getClass().getName());

            PartName name = part.getPartName();
            if (o instanceof CoverPageProperties) {

              part = new DocPropsCoverPagePart(name);
              ((DocPropsCoverPagePart) part).setJaxbElement((CoverPageProperties) o);

            } else if (o instanceof org.opendope.conditions.Conditions) {

              part = new ConditionsPart(name);
              ((ConditionsPart) part).setJaxbElement((org.opendope.conditions.Conditions) o);

            } else if (o instanceof org.opendope.xpaths.Xpaths) {

              part = new XPathsPart(name);
              ((XPathsPart) part).setJaxbElement((org.opendope.xpaths.Xpaths) o);

            } else if (o instanceof org.opendope.questions.Questionnaire) {

              part = new QuestionsPart(name);
              ((QuestionsPart) part).setJaxbElement((org.opendope.questions.Questionnaire) o);

            } else if (o instanceof org.opendope.answers.Answers) {

              part = new StandardisedAnswersPart(name);
              ((StandardisedAnswersPart) part).setJaxbElement((org.opendope.answers.Answers) o);

            } else if (o instanceof org.opendope.components.Components) {

              part = new ComponentsPart(name);
              ((ComponentsPart) part).setJaxbElement((org.opendope.components.Components) o);

            } else if (o instanceof JAXBElement<?>
                && XmlUtils.unwrap(o) instanceof org.docx4j.bibliography.CTSources) {
              part = new BibliographyPart(name);
              ((BibliographyPart) part)
                  .setJaxbElement((JAXBElement<org.docx4j.bibliography.CTSources>) o);

            } else {

              log.error("TODO: handle known CustomXmlPart part  " + o.getClass().getName());

              CustomXmlDataStorage data = getCustomXmlDataStorageClass().factory();
              is.reset();
              data.setDocument(is); // Not necessarily JAXB, that's just our method name
              ((org.docx4j.openpackaging.parts.CustomXmlDataStoragePart) part).setData(data);
            }

          } catch (javax.xml.bind.UnmarshalException ue) {

            log.warn("No JAXB model for this CustomXmlDataStorage part; " + ue.getMessage());

            CustomXmlDataStorage data = getCustomXmlDataStorageClass().factory();
            is.reset();
            data.setDocument(is); // Not necessarily JAXB, that's just our method name
            ((org.docx4j.openpackaging.parts.CustomXmlDataStoragePart) part).setData(data);
          }

        } else if (part instanceof org.docx4j.openpackaging.parts.XmlPart) {

          is = partStore.loadPart(resolvedPartUri);

          //					try {
          ((XmlPart) part).setDocument(is);

          // Experimental 22/6/2011; don't fall back to binary (which we used to)

          //					} catch (Docx4JException d) {
          //						// This isn't an XML part after all,
          //						// even though ContentTypeManager detected it as such
          //						// So get it as a binary part
          //						part = getBinaryPart(partByteArrays, ctm, resolvedPartUri);
          //						log.warn("Could not parse as XML, so using BinaryPart for "
          //								+ resolvedPartUri);
          //						((BinaryPart)part).setBinaryData(is);
          //					}

        } else {
          // Shouldn't happen, since ContentTypeManagerImpl should
          // return an instance of one of the above, or throw an
          // Exception.

          log.error("No suitable part found for: " + resolvedPartUri);
          part = null;
        }

      } catch (PartUnrecognisedException e) {
        log.error("PartUnrecognisedException shouldn't happen anymore!", e);
        // Try to get it as a binary part
        part = getBinaryPart(ctm, resolvedPartUri);
        log.warn("Using BinaryPart for " + resolvedPartUri);

        //				is = partStore.loadPart( resolvedPartUri);
        //				((BinaryPart)part).setBinaryData(is);
      }
    } catch (Exception ex) {
      // IOException, URISyntaxException
      ex.printStackTrace();
      throw new Docx4JException("Failed to getPart", ex);

    } finally {
      IOUtils.closeQuietly(is);
    }

    if (part == null) {
      throw new Docx4JException(
          "cannot find part "
              + resolvedPartUri
              + " from rel "
              + rel.getId()
              + "="
              + rel.getTarget());
    }

    return part;
  }
  private WordprocessingMLPackage fetchComponents(
      WordprocessingMLPackage srcPackage, ContentAccessor contentAccessor) throws Docx4JException {

    // convert components to altChunk
    Map<Integer, CTAltChunk> replacements = new HashMap<Integer, CTAltChunk>();
    Integer index = 0;
    justGotAComponent = false;

    LinkedList<Integer> continuousBeforeIndex = new LinkedList<Integer>();
    List<Boolean> continuousBefore = new ArrayList<Boolean>();

    List<Boolean> continuousAfter = new ArrayList<Boolean>();

    for (Object block : contentAccessor.getContent()) {

      // Object ublock = XmlUtils.unwrap(block);
      if (block instanceof org.docx4j.wml.SdtBlock) {

        org.docx4j.wml.SdtBlock sdt = (org.docx4j.wml.SdtBlock) block;

        Tag tag = getSdtPr(sdt).getTag();

        if (tag == null) {
          List<Object> newContent = new ArrayList<Object>();
          newContent.add(sdt);
          continue;
        }

        log.info(tag.getVal());

        HashMap<String, String> map = QueryString.parseQueryString(tag.getVal(), true);

        String componentId = map.get(BINDING_ROLE_COMPONENT);
        if (componentId == null) continue;

        // Convert the sdt to a w:altChunk
        // .. get the IRI
        String iri = ComponentsPart.getComponentById(components, componentId).getIri();
        log.debug("Fetching " + iri);

        if (docxFetcher == null) {
          log.error("You need a docxFetcher (and the MergeDocx extension) to fetch components");
          return srcPackage;
        }

        // .. create the part
        AlternativeFormatInputPart afiPart =
            new AlternativeFormatInputPart(
                getNewPartName(
                    "/chunk", ".docx", srcPackage.getMainDocumentPart().getRelationshipsPart()));
        afiPart.setBinaryData(docxFetcher.getDocxFromIRI(iri));

        afiPart.setContentType(
            new ContentType(
                "application/vnd.openxmlformats-officedocument.wordprocessingml.document.main+xml")); // docx

        Relationship altChunkRel = srcPackage.getMainDocumentPart().addTargetPart(afiPart);
        CTAltChunk ac = Context.getWmlObjectFactory().createCTAltChunk();
        ac.setId(altChunkRel.getId());

        replacements.put(index, ac);

        /*
         * 2011 12 11 TODO.  Rethink support for
         * od:continuousBefore and od:continuousAfter.
         */

        // This is handled in this class
        if (map.get(BINDING_ROLE_COMPONENT_BEFORE) != null
            && map.get(BINDING_ROLE_COMPONENT_BEFORE).equals("true")) {
          continuousBefore.add(Boolean.TRUE);
          continuousBeforeIndex.addFirst(index);
          log.info("ctsBefore index: " + index);
        } else {
          continuousBefore.add(Boolean.FALSE);
          continuousBeforeIndex.addFirst(index);
        }

        // The following is handled in ProcessAltChunk
        if (map.get(BINDING_ROLE_COMPONENT_AFTER) != null
            && map.get(BINDING_ROLE_COMPONENT_AFTER).equals("true")) {
          continuousAfter.add(Boolean.TRUE);
        } else {
          continuousAfter.add(Boolean.TRUE);
        }

        justGotAComponent = true;
      }
      index++;
    }

    if (!justGotAComponent) {
      return srcPackage;
    }

    // Now replace in list
    for (Integer key : replacements.keySet()) {
      contentAccessor.getContent().set(key, replacements.get(key));
    }

    // Go through docx in reverse order
    List<Object> bodyChildren = contentAccessor.getContent();
    int i = 0;
    for (Integer indexIntoBody : continuousBeforeIndex) {

      if (continuousBefore.get(i)) {
        // Element before the w:altChunk
        if (indexIntoBody == 0) {
          // // Insert a sectPr right at the beginning of the docx?
          // // TODO check this isn't necessary
          // SectPr newSectPr =
          // Context.getWmlObjectFactory().createSectPr();
          // SectPr.Type type =
          // Context.getWmlObjectFactory().createSectPrType();
          // type.setVal("continuous");
          // newSectPr.setType( type );
          //
          // bodyChildren.add(0, newSectPr);

        } else {
          Object block = bodyChildren.get(indexIntoBody.intValue() - 1);
          if (block instanceof P
              && ((P) block).getPPr() != null
              && ((P) block).getPPr().getSectPr() != null) {
            makeContinuous(((P) block).getPPr().getSectPr());
          } else if (block instanceof P) {
            // More likely
            PPr ppr = ((P) block).getPPr();
            if (ppr == null) {
              ppr = Context.getWmlObjectFactory().createPPr();
              ((P) block).setPPr(ppr);
            }
            SectPr newSectPr = Context.getWmlObjectFactory().createSectPr();
            SectPr.Type type = Context.getWmlObjectFactory().createSectPrType();
            type.setVal("continuous");
            newSectPr.setType(type);

            ppr.setSectPr(newSectPr);
          } else {
            // Equally likely - its a table or something, so add a p
            P newP = Context.getWmlObjectFactory().createP();
            PPr ppr = Context.getWmlObjectFactory().createPPr();
            newP.setPPr(ppr);

            SectPr newSectPr = Context.getWmlObjectFactory().createSectPr();
            SectPr.Type type = Context.getWmlObjectFactory().createSectPrType();
            type.setVal("continuous");
            newSectPr.setType(type);
            ppr.setSectPr(newSectPr);

            bodyChildren.add(indexIntoBody.intValue(), newP); // add
            // before
            // altChunk
          }
        }
      }
      // else nothing specified, so go with normal MergeDocx behaviour

      i++;
    }

    // process altChunk
    try {
      // Use reflection, so docx4j can be built
      // by users who don't have the MergeDocx utility
      Class<?> documentBuilder = Class.forName("com.plutext.merge.ProcessAltChunk");
      // Method method = documentBuilder.getMethod("merge",
      // wmlPkgList.getClass());
      Method[] methods = documentBuilder.getMethods();
      Method processMethod = null;
      for (int j = 0; j < methods.length; j++) {
        log.debug(methods[j].getName());
        if (methods[j].getName().equals("process")) {
          processMethod = methods[j];
        }
      }
      if (processMethod == null) throw new NoSuchMethodException();
      return (WordprocessingMLPackage) processMethod.invoke(null, srcPackage);

    } catch (ClassNotFoundException e) {
      extensionMissing(e);
      justGotAComponent = false;
      return srcPackage;
      // throw new Docx4JException("Problem processing w:altChunk", e);
    } catch (NoSuchMethodException e) {
      // Degrade gracefully
      extensionMissing(e);
      justGotAComponent = false;
      return srcPackage;
      // throw new Docx4JException("Problem processing w:altChunk", e);
    } catch (Exception e) {
      throw new Docx4JException("Problem processing w:altChunk", e);
    }
  }