Apache
/
poi
mirror of https://github.com/apache/poi.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Hopefully fix bug 39228 - big overhaul of how slides and notes are build, based on Yegor's discoveries. Hopefully, we now return the right slide #, and in the right order 

git-svn-id: https://svn.apache.org/repos/asf/jakarta/poi/trunk@417541 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Nick Burch 2006-06-27 18:15:32 +00:00
parent 993972171b
commit 2db5e373a2
4 changed files with 218 additions and 156 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
src/scratchpad/src/org/apache/poi/hslf/model/Sheet.java
View File

@ -73,9 +73,18 @@ public abstract class Sheet
public SlideShow getSlideShow() { return _slideShow; } public SlideShow getSlideShow() { return _slideShow; }
/** /**
* Set the SlideShow we're attached to * Set the SlideShow we're attached to.
* Also passes it on to our child RichTextRuns
*/ */
public void setSlideShow(SlideShow ss) { _slideShow = ss; } public void setSlideShow(SlideShow ss) {
_slideShow = ss;
TextRun[] trs = getTextRuns();
if(trs != null) {
for(int i=0; i<trs.length; i++) {
trs[i].supplySlideShow(_slideShow);
}
}
}
/** /**

15
src/scratchpad/src/org/apache/poi/hslf/model/TextRun.java
View File

@ -25,6 +25,7 @@ import java.util.Vector;
import org.apache.poi.hslf.record.*; import org.apache.poi.hslf.record.*;
import org.apache.poi.hslf.record.StyleTextPropAtom.TextPropCollection; import org.apache.poi.hslf.record.StyleTextPropAtom.TextPropCollection;
import org.apache.poi.hslf.usermodel.RichTextRun; import org.apache.poi.hslf.usermodel.RichTextRun;
import org.apache.poi.hslf.usermodel.SlideShow;
import org.apache.poi.util.StringUtil; import org.apache.poi.util.StringUtil;
/** /**
@ -45,6 +46,7 @@ public class TextRun
protected StyleTextPropAtom _styleAtom; protected StyleTextPropAtom _styleAtom;
protected boolean _isUnicode; protected boolean _isUnicode;
protected RichTextRun[] _rtRuns; protected RichTextRun[] _rtRuns;
private SlideShow slideShow;
/** /**
* Constructs a Text Run from a Unicode text block * Constructs a Text Run from a Unicode text block
@ -500,4 +502,17 @@ public class TextRun
public void setRunType(int type) { public void setRunType(int type) {
_headerAtom.setTextType(type); _headerAtom.setTextType(type);
} }
/**
* Supply the SlideShow we belong to.
* Also passes it on to our child RichTextRuns
*/
public void supplySlideShow(SlideShow ss) {
slideShow = ss;
if(_rtRuns != null) {
for(int i=0; i<_rtRuns.length; i++) {
_rtRuns[i].supplySlideShow(slideShow);
}
}
}
} }

258
src/scratchpad/src/org/apache/poi/hslf/usermodel/SlideShow.java
View File

@ -247,19 +247,34 @@ public class SlideShow
} }
} }
} }
/**
* For a given SlideAtomsSet, return the core record, based on the refID from the
* SlidePersistAtom
*/
private Record getCoreRecordForSAS(SlideAtomsSet sas) {
SlidePersistAtom spa = sas.getSlidePersistAtom();
int refID = spa.getRefID();
return getCoreRecordForRefID(refID);
}
/**
* For a given refID (the internal, 0 based numbering scheme), return the
* core record
* @param refID the refID
*/
private Record getCoreRecordForRefID(int refID) {
Integer coreRecordId = (Integer)
_sheetIdToCoreRecordsLookup.get(new Integer(refID));
Record r = _mostRecentCoreRecords[coreRecordId.intValue()];
return r;
}
/** /**
* Build up model level Slide and Notes objects, from the underlying * Build up model level Slide and Notes objects, from the underlying
* records. * records.
*/ */
private void buildSlidesAndNotes() { private void buildSlidesAndNotes() {
// For holding the Slide Records
Vector slidesV = new Vector(10);
// For holding the Notes Records
Vector notesV = new Vector(10);
// For holding the Meta Sheet Records
Vector metaSheetsV = new Vector(10);
// Ensure we really found a Document record earlier // Ensure we really found a Document record earlier
// If we didn't, then the file is probably corrupt // If we didn't, then the file is probably corrupt
if(_documentRecord == null) { if(_documentRecord == null) {
@ -269,167 +284,102 @@ public class SlideShow
// Fetch the SlideListWithTexts in the most up-to-date Document Record // Fetch the SlideListWithTexts in the most up-to-date Document Record
// //
// Then, use this to find the Slide records, and also the Notes record // As far as we understand it:
// for each Slide (if it has one) // * The first SlideListWithText will contain a SlideAtomsSet
// for each of the master slides
// * The second SlideListWithText will contain a SlideAtomsSet
// for each of the slides, in their current order
// These SlideAtomsSets will normally contain text
// * The third SlideListWithText (if present), will contain a
// SlideAtomsSet for each Notes
// These SlideAtomsSets will not normally contain text
// //
// The following matching algorithm is based on looking at the output // Having indentified the masters, slides and notes + their orders,
// of org.apache.poi.hslf.dev.SlideIdListing on a number of files: // we have to go and find their matching records
// // We always use the latest versions of these records, and use the
// 1) Get the SlideAtomSets from the SlideListWithTexts of the most // SlideAtom/NotesAtom to match them with the StyleAtomSet
// up-to-date Document
// 2) Get the SlidePersistAtoms from all of these
// 3) Get the RefId, which corresponds to a "sheet ID" from the
// PersistPtr Stuff
// 4) Grab the record at that ID, and see if it's a slide or a notes
// 5) Build a mapping between the SlideIdentifier ID and the RefId
// for both slides and notes
// 6) Loop over all the slides
// 7) Look each slide's SlideAtom to see if it has associated Notes -
// if it does, the ID will be SlideIdentifier for those notes
// (Note: might not be the same as the SlideIdentifier of the Slide)
// 8) Generate the model representations, giving them the matching
// slide atom sets, IDs etc
SlideListWithText[] slwts = _documentRecord.getSlideListWithTexts(); SlideListWithText masterSLWT = _documentRecord.getMasterSlideListWithText();
SlideListWithText slidesSLWT = _documentRecord.getSlideSlideListWithText();
SlideListWithText notesSLWT = _documentRecord.getNotesSlideListWithText();
// To hold the lookup from SlideIdentifier IDs to RefIDs // Start by finding the notes records to go with the entries in
Hashtable slideSlideIdToRefid = new Hashtable(); // notesSLWT
Hashtable notesSlideIdToRefid = new Hashtable(); org.apache.poi.hslf.record.Notes[] notesRecords;
// To hold the lookup from SlideIdentifier IDs to SlideAtomsSets SlideAtomsSet[] notesSets = new SlideAtomsSet[0];
Hashtable slideSlideIdToSlideAtomsSet = new Hashtable(); Hashtable slideIdToNotes = new Hashtable();
Hashtable notesSlideIdToSlideAtomsSet = new Hashtable(); if(notesSLWT == null) {
// None
// Loop over all the SlideListWithTexts, getting their notesRecords = new org.apache.poi.hslf.record.Notes[0];
// SlideAtomSets } else {
for(int i=0; i<slwts.length; i++) { // Match up the records and the SlideAtomSets
SlideAtomsSet[] sas = slwts[i].getSlideAtomsSets(); notesSets = notesSLWT.getSlideAtomsSets();
for(int j=0; j<sas.length; j++) { notesRecords = new org.apache.poi.hslf.record.Notes[notesSets.length];
// What does this SlidePersistAtom point to? for(int i=0; i<notesSets.length; i++) {
SlidePersistAtom spa = sas[j].getSlidePersistAtom(); // Get the right core record
Integer slideIdentifier = new Integer( spa.getSlideIdentifier() ); Record r = getCoreRecordForSAS(notesSets[i]);
Integer slideRefId = new Integer( spa.getRefID() );
// Grab the record it points to // Ensure it really is a notes record
Integer coreRecordId = (Integer) if(r instanceof org.apache.poi.hslf.record.Notes) {
_sheetIdToCoreRecordsLookup.get(slideRefId); notesRecords[i] = (org.apache.poi.hslf.record.Notes)r;
Record r = _mostRecentCoreRecords[coreRecordId.intValue()];
// Add the IDs to the appropriate lookups
if(r instanceof org.apache.poi.hslf.record.Slide) {
slideSlideIdToRefid.put( slideIdentifier, slideRefId );
// Save the SlideAtomsSet
slideSlideIdToSlideAtomsSet.put( slideIdentifier, sas[j] );
} else if(r instanceof org.apache.poi.hslf.record.Notes) {
notesSlideIdToRefid.put( slideIdentifier, slideRefId );
// Save the SlideAtomsSet
notesSlideIdToSlideAtomsSet.put( slideIdentifier, sas[j] );
} else if(r.getRecordType() == RecordTypes.MainMaster.typeID) {
// Skip for now, we don't do Master slides yet
} else { } else {
throw new IllegalStateException("SlidePersistAtom had a RefId that pointed to something other than a Slide or a Notes, was a " + r + " with type " + r.getRecordType()); System.err.println("A Notes SlideAtomSet at " + i + " said its record was at refID " + notesSets[i].getSlidePersistAtom().getRefID() + ", but that was actually a " + r);
}
// Record the match between slide id and these notes
SlidePersistAtom spa = notesSets[i].getSlidePersistAtom();
Integer slideId = new Integer(spa.getSlideIdentifier());
slideIdToNotes.put(slideId, new Integer(i));
}
}
// Now, do the same thing for our slides
org.apache.poi.hslf.record.Slide[] slidesRecords;
SlideAtomsSet[] slidesSets = new SlideAtomsSet[0];
if(slidesSLWT == null) {
// None
slidesRecords = new org.apache.poi.hslf.record.Slide[0];
} else {
// Match up the records and the SlideAtomSets
slidesSets = slidesSLWT.getSlideAtomsSets();
slidesRecords = new org.apache.poi.hslf.record.Slide[slidesSets.length];
for(int i=0; i<slidesSets.length; i++) {
// Get the right core record
Record r = getCoreRecordForSAS(slidesSets[i]);
// Ensure it really is a slide record
if(r instanceof org.apache.poi.hslf.record.Slide) {
slidesRecords[i] = (org.apache.poi.hslf.record.Slide)r;
} else {
System.err.println("A Slide SlideAtomSet at " + i + " said its record was at refID " + slidesSets[i].getSlidePersistAtom().getRefID() + ", but that was actually a " + r);
} }
} }
} }
// Now, create a model representation of a slide for each // Finally, generate model objects for everything
// slide + slideatomset we found // Notes first
// Do it in order of the SlideIdentifiers _notes = new Notes[notesRecords.length];
int[] slideIDs = new int[slideSlideIdToRefid.size()];
int pos = 0;
Enumeration e = slideSlideIdToRefid.keys();
while(e.hasMoreElements()) {
Integer id = (Integer)e.nextElement();
slideIDs[pos] = id.intValue();
pos++;
}
// Sort
Arrays.sort(slideIDs);
// Create
for(int i=0; i<slideIDs.length; i++) {
// Build up the list of all the IDs we might want to use
int slideIdentifier = slideIDs[i];
Integer slideIdentifierI = new Integer(slideIdentifier);
int slideNumber = (i+1);
Integer slideRefI = (Integer)slideSlideIdToRefid.get(slideIdentifierI);
Integer slideCoreRecNumI = (Integer)_sheetIdToCoreRecordsLookup.get(slideRefI);
int slideCoreRecNum = slideCoreRecNumI.intValue();
// Fetch the Slide record
org.apache.poi.hslf.record.Slide s = (org.apache.poi.hslf.record.Slide)
_mostRecentCoreRecords[slideCoreRecNum];
// Do we have a notes for this slide?
org.apache.poi.hslf.record.Notes n = null;
if(s.getSlideAtom().getNotesID() > 0) {
// Get the SlideIdentifier of the Notes
// (Note - might not be the same as the SlideIdentifier of the Slide)
int notesSlideIdentifier = s.getSlideAtom().getNotesID();
Integer notesSlideIdentifierI = new Integer(notesSlideIdentifier);
// Grab the notes record
Integer notesRefI = (Integer)notesSlideIdToRefid.get(notesSlideIdentifierI);
Integer notesCoreRecNum = (Integer)_sheetIdToCoreRecordsLookup.get(notesRefI);
n = (org.apache.poi.hslf.record.Notes)
_mostRecentCoreRecords[notesCoreRecNum.intValue()];
}
// Grab the matching SlideAtomSet
SlideAtomsSet sas = (SlideAtomsSet)
slideSlideIdToSlideAtomsSet.get(slideIdentifierI);
// Build the notes model, if there's notes
Notes notes = null;
if(n != null) {
// TODO: Use this
SlideAtomsSet nsas = (SlideAtomsSet)
notesSlideIdToSlideAtomsSet.get(slideIdentifierI);
// Create the model view of the notes
notes = new Notes(n);
notesV.add(notes);
}
// Build the slide model
Slide slide = new Slide(s, notes, sas, slideIdentifier, slideNumber);
slidesV.add(slide);
}
// ******************* Finish up ****************
// Finish setting up the notes
_notes = new Notes[notesV.size()];
for(int i=0; i<_notes.length; i++) { for(int i=0; i<_notes.length; i++) {
_notes[i] = (Notes)notesV.get(i); _notes[i] = new Notes(notesRecords[i]);
_notes[i].setSlideShow(this); _notes[i].setSlideShow(this);
// Now supply ourselves to all the rich text runs
// of this note's TextRuns
TextRun[] trs = _notes[i].getTextRuns();
for(int j=0; j<trs.length; j++) {
RichTextRun[] rtrs = trs[j].getRichTextRuns();
for(int k=0; k<rtrs.length; k++) {
rtrs[k].supplySlideShow(this);
}
}
} }
// Then slides
_slides = new Slide[slidesRecords.length];
// Create our Slides
_slides = new Slide[slidesV.size()];
for(int i=0; i<_slides.length; i++) { for(int i=0; i<_slides.length; i++) {
_slides[i] = (Slide)slidesV.get(i); SlideAtomsSet sas = slidesSets[i];
_slides[i].setSlideShow(this); int slideIdentifier = sas.getSlidePersistAtom().getSlideIdentifier();
Integer slideIdentifierI = new Integer(slideIdentifier);
// Now supply ourselves to all the rich text runs
// of this slide's TextRuns // Do we have a notes for this?
TextRun[] trs = _slides[i].getTextRuns(); Notes notes = null;
for(int j=0; j<trs.length; j++) { if(slideIdToNotes.containsKey(slideIdentifierI)) {
RichTextRun[] rtrs = trs[j].getRichTextRuns(); Integer notesPos = (Integer)slideIdToNotes.get(slideIdentifierI);
for(int k=0; k<rtrs.length; k++) { notes = _notes[notesPos.intValue()];
rtrs[k].supplySlideShow(this);
}
} }
// Now, build our slide
_slides[i] = new Slide(slidesRecords[i], notes, sas, slideIdentifier, (i+1));
_slides[i].setSlideShow(this);
} }
} }

88
src/scratchpad/testcases/org/apache/poi/hslf/usermodel/TestSlideOrdering.java Normal file
View File

@ -0,0 +1,88 @@
/* ====================================================================
Copyright 2002-2004 Apache Software Foundation
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
==================================================================== */
package org.apache.poi.hslf.usermodel;
import junit.framework.TestCase;
import org.apache.poi.hslf.*;
import org.apache.poi.hslf.model.*;
/**
* Tests that SlideShow returns Sheets in the right order
*
* @author Nick Burch (nick at torchbox dot com)
*/
public class TestSlideOrdering extends TestCase {
// Simple slideshow, record order matches slide order
private SlideShow ssA;
// Complex slideshow, record order doesn't match slide order
private SlideShow ssB;
public TestSlideOrdering() throws Exception {
String dirname = System.getProperty("HSLF.testdata.path");
String filenameA = dirname + "/basic_test_ppt_file.ppt";
HSLFSlideShow hssA = new HSLFSlideShow(filenameA);
ssA = new SlideShow(hssA);
String filenameB = dirname + "/incorrect_slide_order.ppt";
HSLFSlideShow hssB = new HSLFSlideShow(filenameB);
ssB = new SlideShow(hssB);
}
/**
* Test the simple case - record order matches slide order
*/
public void testSimpleCase() throws Exception {
assertEquals(2, ssA.getSlides().length);
Slide s1 = ssA.getSlides()[0];
Slide s2 = ssA.getSlides()[1];
String[] firstTRs = new String[] {
"This is a test title",
"This is the title on page 2"
};
assertEquals(firstTRs[0], s1.getTextRuns()[0].getText());
assertEquals(firstTRs[1], s2.getTextRuns()[0].getText());
}
/**
* Test the complex case - record order differs from slide order
*/
public void testComplexCase() throws Exception {
assertEquals(3, ssB.getSlides().length);
Slide s1 = ssB.getSlides()[0];
Slide s2 = ssB.getSlides()[1];
Slide s3 = ssB.getSlides()[2];
String[] firstTRs = new String[] {
"Slide 1",
"Slide 2",
"Slide 3"
};
assertEquals(firstTRs[0], s1.getTextRuns()[0].getText());
assertEquals(firstTRs[1], s2.getTextRuns()[0].getText());
assertEquals(firstTRs[2], s3.getTextRuns()[0].getText());
}
}