Get insert based HWPF tests working fine, delete ones still problematic

git-svn-id: https://svn.apache.org/repos/asf/poi/trunk@684959 13f79535-47bb-0310-9956-ffa450edef68

src/scratchpad/src/org/apache/poi/hwpf/HWPFDocument.java

@@ -317,6 +317,11 @@ public class HWPFDocument extends POIDocument
    *  document, but excludes any headers and footers.
    */
   public Range getRange() {
+	  // First up, trigger a full-recalculate
+	  // Needed in case of deletes etc
+	  getOverallRange();
+	  
+	  // Now, return the real one
 	  return new Range(
 			  _cpSplit.getMainDocumentStart(), 
 			  _cpSplit.getMainDocumentEnd(), 

src/scratchpad/src/org/apache/poi/hwpf/model/PAPFormattedDiskPage.java

@@ -64,10 +64,12 @@ public class PAPFormattedDiskPage extends FormattedDiskPage
     {
       super(documentStream, offset);
 
-      for (int x = 0; x < _crun; x++)
+      for (int x = 0; x < _crun; x++) {
-      {
+         int startAt = getStart(x) - fcMin;
-    	boolean isUnicode = tpt.isUnicodeAt( getStart(x) );
+         int endAt = getEnd(x) - fcMin;
-        _papxList.add(new PAPX(getStart(x) - fcMin, getEnd(x) - fcMin, getGrpprl(x), getParagraphHeight(x), dataStream, isUnicode));
+    	 boolean isUnicode = tpt.isUnicodeAt(startAt);
+    	 
+         _papxList.add(new PAPX(startAt, endAt, getGrpprl(x), getParagraphHeight(x), dataStream, isUnicode));
       }
       _fkp = null;
       _dataStream = dataStream;

src/scratchpad/src/org/apache/poi/hwpf/model/PropertyNode.java

@@ -85,18 +85,18 @@ public abstract class PropertyNode implements Comparable, Cloneable
   {
     int end = start + length;
 
-    if (_cpEnd > start)
+    if (_cpEnd > start) {
-    {
+        // The start of the change is before we end
-      if (_cpStart < end)
+    	
-      {
+        if (_cpStart < end) {
-        _cpEnd = end >= _cpEnd ? start : _cpEnd - length;
+            // The delete was somewhere in the middle of us
-        _cpStart = Math.min(start, _cpStart);
+            _cpEnd = end >= _cpEnd ? start : _cpEnd - length;
-      }
+            _cpStart = Math.min(start, _cpStart);
-      else
+        } else {
-      {
+            // The delete was before us
-        _cpEnd -= length;
+            _cpEnd -= length;
-        _cpStart -= length;
+            _cpStart -= length;
-      }
+        }
     }
   }
 

src/scratchpad/src/org/apache/poi/hwpf/model/SEPX.java

@@ -20,11 +20,14 @@
 
 package org.apache.poi.hwpf.model;
 
-import org.apache.poi.hwpf.sprm.SprmBuffer;
-import org.apache.poi.hwpf.sprm.SectionSprmUncompressor;
 import org.apache.poi.hwpf.sprm.SectionSprmCompressor;
+import org.apache.poi.hwpf.sprm.SectionSprmUncompressor;
 import org.apache.poi.hwpf.usermodel.SectionProperties;
 
+/**
+ * TODO - figure out if this works in characters, like most
+ *  things do, or in bytes as PAPX / CHPX does.
+ */
 public class SEPX extends PropertyNode
 {
 

src/scratchpad/src/org/apache/poi/hwpf/model/TextPiece.java

@@ -118,6 +118,9 @@ public class TextPiece extends PropertyNode implements Comparable
 	   if(end > buf.length()) {
 		   throw new StringIndexOutOfBoundsException("Index " + end + " out of range 0 -> " + buf.length());
 	   }
+	   if(end < start) {
+		   throw new StringIndexOutOfBoundsException("Asked for text from " + start + " to " + end + ", which has an end before the start!");
+	   }
 	   return buf.substring(start, end);
    }
 

src/scratchpad/src/org/apache/poi/hwpf/usermodel/Range.java

@@ -650,14 +650,6 @@ public class Range
                 absPlaceHolderIndex, 
 				(absPlaceHolderIndex + pPlaceHolder.length()), getDocument()
     );
-    if (subRange.usesUnicode()) {
-			absPlaceHolderIndex = getStartOffset() + (pOffset * 2);
-            subRange = new Range(
-                      absPlaceHolderIndex, 
-                      (absPlaceHolderIndex + (pPlaceHolder.length() * 2)), 
-					  getDocument()
-            );
-    }
 
 	// this Range isn't a proper parent of the subRange() so we'll have to keep
 	// track of an updated endOffset on our own
@@ -674,12 +666,6 @@ public class Range
             (absPlaceHolderIndex + pPlaceHolder.length() + pValue.length()), 
 			getDocument()
     );
-    if (subRange.usesUnicode())
-            subRange = new Range(
-                      (absPlaceHolderIndex + (pValue.length() * 2)),
-                      (absPlaceHolderIndex + (pPlaceHolder.length() * 2) + 
-					  (pValue.length() * 2)), getDocument()
-            );
 
 	// deletes are automagically propagated
     subRange.delete();

src/scratchpad/testcases/org/apache/poi/hwpf/usermodel/TestRangeDelete.java

@@ -27,12 +27,14 @@ import org.apache.poi.hwpf.HWPFDocument;
 /**
  *	Test to see if Range.delete() works even if the Range contains a
  *	CharacterRun that uses Unicode characters.
- *
- * TODO - re-enable me when unicode paragraph stuff is fixed!
  */
 public class TestRangeDelete extends TestCase {
 
 	// u201c and u201d are "smart-quotes"
+	private String introText = 
+		"Introduction\r";
+	private String fillerText = 
+		"${delete} This is an MS-Word 97 formatted document created using NeoOffice v. 2.2.4 Patch 0 (OpenOffice.org v. 2.2.1).\r";
 	private String originalText =
 		"It is used to confirm that text delete works even if Unicode characters (such as \u201c\u2014\u201d (U+2014), \u201c\u2e8e\u201d (U+2E8E), or \u201c\u2714\u201d (U+2714)) are present.  Everybody should be thankful to the ${organization} ${delete} and all the POI contributors for their assistance in this matter.\r";
 	private String searchText = "${delete}";
@@ -64,31 +66,34 @@ public class TestRangeDelete extends TestCase {
 	public void testDocStructure() throws Exception {
 
 		HWPFDocument daDoc = new HWPFDocument(new FileInputStream(illustrativeDocFile));
+		Range range;
+		Section section;
+		Paragraph para;
 
-		Range range = daDoc.getOverallRange();
+		// First, check overall
-
+		range = daDoc.getOverallRange();
 		assertEquals(1, range.numSections());
-		Section section = range.getSection(0);
+		assertEquals(4, range.numParagraphs());
-
-		assertEquals(5, section.numParagraphs());
-		Paragraph para = section.getParagraph(2);
-
-		assertEquals(5, para.numCharacterRuns());
-
-		assertEquals(originalText, para.text());
 		
 		
-		// Now check on just the main text
+		// Now, onto just the doc bit
 		range = daDoc.getRange();
-		
+
 		assertEquals(1, range.numSections());
 		section = range.getSection(0);
 
-		assertEquals(5, section.numParagraphs());
+		assertEquals(4, section.numParagraphs());
+		
+		para = section.getParagraph(0);
+		assertEquals(1, para.numCharacterRuns());
+		assertEquals(introText, para.text());
+		
+		para = section.getParagraph(1);
+		assertEquals(2, para.numCharacterRuns());
+		assertEquals(fillerText, para.text());
+		
 		para = section.getParagraph(2);
-
+		assertEquals(6, para.numCharacterRuns());
-		assertEquals(5, para.numCharacterRuns());
-
 		assertEquals(originalText, para.text());
 	}
 
@@ -103,7 +108,7 @@ public class TestRangeDelete extends TestCase {
 		assertEquals(1, range.numSections());
 
 		Section section = range.getSection(0);
-		assertEquals(5, section.numParagraphs());
+		assertEquals(4, section.numParagraphs());
 
 		Paragraph para = section.getParagraph(2);
 
@@ -114,12 +119,7 @@ public class TestRangeDelete extends TestCase {
 		assertEquals(192, offset);
 
 		int absOffset = para.getStartOffset() + offset;
-		if (para.usesUnicode())
-			absOffset = para.getStartOffset() + (offset * 2);
-
 		Range subRange = new Range(absOffset, (absOffset + searchText.length()), para.getDocument());
-		if (subRange.usesUnicode())
-			subRange = new Range(absOffset, (absOffset + (searchText.length() * 2)), para.getDocument());
 
 		assertEquals(searchText, subRange.text());
 
@@ -131,7 +131,7 @@ public class TestRangeDelete extends TestCase {
 		assertEquals(1, range.numSections());
 		section = range.getSection(0);
 
-		assertEquals(5, section.numParagraphs());
+		assertEquals(4, section.numParagraphs());
 		para = section.getParagraph(2);
 
 		text = para.text();
@@ -154,7 +154,7 @@ public class TestRangeDelete extends TestCase {
 		assertEquals(1, range.numSections());
 
 		Section section = range.getSection(0);
-		assertEquals(5, section.numParagraphs());
+		assertEquals(4, section.numParagraphs());
 
 		Paragraph para = section.getParagraph(2);
 
@@ -163,26 +163,23 @@ public class TestRangeDelete extends TestCase {
 
 		boolean keepLooking = true;
 		while (keepLooking) {
-
+			// Reload the range every time
+			range = daDoc.getRange();
 			int offset = range.text().indexOf(searchText);
 			if (offset >= 0) {
 
 				int absOffset = range.getStartOffset() + offset;
-				if (range.usesUnicode())
-					absOffset = range.getStartOffset() + (offset * 2);
 
 				Range subRange = new Range(
 					absOffset, (absOffset + searchText.length()), range.getDocument());
-				if (subRange.usesUnicode())
-					subRange = new Range(
-						absOffset, (absOffset + (searchText.length() * 2)), range.getDocument());
 
 				assertEquals(searchText, subRange.text());
 
 				subRange.delete();
 
-			} else
+			} else {
 				keepLooking = false;
+			}
 		}
 
 		// we need to let the model re-calculate the Range before we use it
@@ -191,7 +188,11 @@ public class TestRangeDelete extends TestCase {
 		assertEquals(1, range.numSections());
 		section = range.getSection(0);
 
-		assertEquals(5, section.numParagraphs());
+		assertEquals(4, section.numParagraphs());
+
+		para = section.getParagraph(0);
+		text = para.text();
+		assertEquals(introText, text);
 
 		para = section.getParagraph(1);
 		text = para.text();

src/scratchpad/testcases/org/apache/poi/hwpf/usermodel/TestRangeInsertion.java

@@ -36,7 +36,7 @@ public class TestRangeInsertion extends TestCase {
 	private String originalText =
 		"It is used to confirm that text insertion works even if Unicode characters (such as \u201c\u2014\u201d (U+2014), \u201c\u2e8e\u201d (U+2E8E), or \u201c\u2714\u201d (U+2714)) are present.\r";
 	private String textToInsert = "Look at me!  I'm cool!  ";
-	private int insertionPoint = 244;
+	private int insertionPoint = 122;
 
 	private String illustrativeDocFile;
 
@@ -69,12 +69,21 @@ public class TestRangeInsertion extends TestCase {
 
 		assertEquals(3, section.numParagraphs());
 		Paragraph para = section.getParagraph(2);
+		assertEquals(originalText, para.text());
 
-		assertEquals(3, para.numCharacterRuns());
+		assertEquals(6, para.numCharacterRuns());
-		String text = para.getCharacterRun(0).text() + para.getCharacterRun(1).text() +
+		String text = 
-			para.getCharacterRun(2).text();
+			para.getCharacterRun(0).text() + 
+			para.getCharacterRun(1).text() +
+			para.getCharacterRun(2).text() +
+			para.getCharacterRun(3).text() +
+			para.getCharacterRun(4).text() +
+			para.getCharacterRun(5).text()
+		;
 
 		assertEquals(originalText, text);
+		
+		assertEquals(insertionPoint, para.getStartOffset());
 	}
 
 	/**
@@ -105,10 +114,17 @@ public class TestRangeInsertion extends TestCase {
 
 		assertEquals(3, section.numParagraphs());
 		Paragraph para = section.getParagraph(2);
+		assertEquals((textToInsert + originalText), para.text());
 
-		assertEquals(3, para.numCharacterRuns());
+		assertEquals(6, para.numCharacterRuns());
-		String text = para.getCharacterRun(0).text() + para.getCharacterRun(1).text() +
+		String text = 
-			para.getCharacterRun(2).text();
+			para.getCharacterRun(0).text() + 
+			para.getCharacterRun(1).text() +
+			para.getCharacterRun(2).text() +
+			para.getCharacterRun(3).text() +
+			para.getCharacterRun(4).text() +
+			para.getCharacterRun(5).text()
+		;
 
 		// System.out.println(text);
 

src/scratchpad/testcases/org/apache/poi/hwpf/usermodel/TestRangeProperties.java

@@ -257,6 +257,91 @@ public class TestRangeProperties extends TestCase {
 		assertEquals(p2_parts[0] + "\r", r.getParagraph(11).text());
 	}
 	public void testUnicodeStyling() throws Exception {
-		// TODO
+		Range r = u.getRange();
+		String[] p1_parts = u_page_1.split("\r");
+		
+		Paragraph p1 = r.getParagraph(0);
+		Paragraph p7 = r.getParagraph(6);
+		
+		// Line ending in its own run each time!
+		assertEquals(2, p1.numCharacterRuns());
+		assertEquals(2, p7.numCharacterRuns());
+		
+		CharacterRun c1a = p1.getCharacterRun(0);
+		CharacterRun c1b = p1.getCharacterRun(1);
+		CharacterRun c7a = p7.getCharacterRun(0);
+		CharacterRun c7b = p7.getCharacterRun(1);
+		
+		assertEquals("Times New Roman", c1a.getFontName()); // No Calibri
+		assertEquals(22, c1a.getFontSize());
+		
+		assertEquals("Times New Roman", c1b.getFontName()); // No Calibri
+		assertEquals(22, c1b.getFontSize());
+		
+		assertEquals("Times New Roman", c7a.getFontName());
+		assertEquals(48, c7a.getFontSize());
+		
+		assertEquals("Times New Roman", c7b.getFontName());
+		assertEquals(48, c7b.getFontSize());
+		
+		// Now check where they crop up
+		assertEquals(
+				0, 
+				c1a.getStartOffset()
+		);
+		assertEquals(
+				p1_parts[0].length(), 
+				c1a.getEndOffset()
+		);
+		
+		assertEquals(
+				p1_parts[0].length(), 
+				c1b.getStartOffset()
+		);
+		assertEquals(
+				p1_parts[0].length()+1, 
+				c1b.getEndOffset()
+		);
+		
+		assertEquals(
+				p1_parts[0].length() + 1 + 
+				p1_parts[1].length() + 1 + 
+				p1_parts[2].length() + 1 + 
+				p1_parts[3].length() + 1 + 
+				p1_parts[4].length() + 1 + 
+				p1_parts[5].length() + 1, 
+				c7a.getStartOffset()
+		);
+		assertEquals(
+				p1_parts[0].length() + 1 + 
+				p1_parts[1].length() + 1 + 
+				p1_parts[2].length() + 1 + 
+				p1_parts[3].length() + 1 + 
+				p1_parts[4].length() + 1 + 
+				p1_parts[5].length() + 1 +
+				1,
+				c7a.getEndOffset()
+		);
+		
+		assertEquals(
+				p1_parts[0].length() + 1 + 
+				p1_parts[1].length() + 1 + 
+				p1_parts[2].length() + 1 + 
+				p1_parts[3].length() + 1 + 
+				p1_parts[4].length() + 1 + 
+				p1_parts[5].length() + 1 +
+				1, 
+				c7b.getStartOffset()
+		);
+		assertEquals(
+				p1_parts[0].length() + 1 + 
+				p1_parts[1].length() + 1 + 
+				p1_parts[2].length() + 1 + 
+				p1_parts[3].length() + 1 + 
+				p1_parts[4].length() + 1 + 
+				p1_parts[5].length() + 1 +
+				p1_parts[6].length() + 1,
+				c7b.getEndOffset()
+		);
 	}
 }

src/scratchpad/testcases/org/apache/poi/hwpf/usermodel/TestRangeReplacement.java

@@ -70,12 +70,18 @@ public class TestRangeReplacement extends TestCase {
 		assertEquals(1, range.numSections());
 		Section section = range.getSection(0);
 
-		assertEquals(5, section.numParagraphs());
+		assertEquals(4, section.numParagraphs());
 		Paragraph para = section.getParagraph(2);
 
-		assertEquals(5, para.numCharacterRuns());
+		assertEquals(6, para.numCharacterRuns());
-		String text = para.getCharacterRun(0).text() + para.getCharacterRun(1).text() +
+		String text = 
-			para.getCharacterRun(2).text() + para.getCharacterRun(3).text() + para.getCharacterRun(4).text();
+			para.getCharacterRun(0).text() + 
+			para.getCharacterRun(1).text() +
+			para.getCharacterRun(2).text() +
+			para.getCharacterRun(3).text() +
+			para.getCharacterRun(4).text() +
+			para.getCharacterRun(5).text()
+		;
 
 		assertEquals(originalText, text);
 	}
@@ -91,7 +97,7 @@ public class TestRangeReplacement extends TestCase {
 		assertEquals(1, range.numSections());
 
 		Section section = range.getSection(0);
-		assertEquals(5, section.numParagraphs());
+		assertEquals(4, section.numParagraphs());
 
 		Paragraph para = section.getParagraph(2);
 
@@ -124,7 +130,7 @@ public class TestRangeReplacement extends TestCase {
 		assertEquals(1, range.numSections());
 
 		Section section = range.getSection(0);
-		assertEquals(5, section.numParagraphs());
+		assertEquals(4, section.numParagraphs());
 
 		Paragraph para = section.getParagraph(2);
 
@@ -135,7 +141,7 @@ public class TestRangeReplacement extends TestCase {
 
 		assertEquals(1, range.numSections());
 		section = range.getSection(0);
-		assertEquals(5, section.numParagraphs());
+		assertEquals(4, section.numParagraphs());
 
 		para = section.getParagraph(2);
 		text = para.text();