Add HPBF hyperlinks support to the extractor

git-svn-id: https://svn.apache.org/repos/asf/poi/trunk@690729 13f79535-47bb-0310-9956-ffa450edef68

src/documentation/content/xdocs/changes.xml

@@ -37,6 +37,7 @@
 
 		<!-- Don't forget to update status.xml too! -->
         <release version="3.1.1-alpha1" date="2008-??-??">
+           <action dev="POI-DEVELOPERS" type="add">Support for HPBF Publisher hyperlinks, including during text extraction</action>
            <action dev="POI-DEVELOPERS" type="fix">26321 and 44958 - preserve position of ArrayRecords and TableRecords among cell value records</action>
            <action dev="POI-DEVELOPERS" type="fix">Impove empty header or footer handling in HWPF HeaderStories</action>
            <action dev="POI-DEVELOPERS" type="fix">Avoid NPE in hssf.usermodel.HeaderFooter when stripping fields out</action>

src/documentation/content/xdocs/status.xml

@@ -34,6 +34,7 @@
 	<!-- Don't forget to update changes.xml too! -->
     <changes>
         <release version="3.1.1-alpha1" date="2008-??-??">
+           <action dev="POI-DEVELOPERS" type="add">Support for HPBF Publisher hyperlinks, including during text extraction</action>
            <action dev="POI-DEVELOPERS" type="fix">26321 and 44958 - preserve position of ArrayRecords and TableRecords among cell value records</action>
            <action dev="POI-DEVELOPERS" type="fix">Impove empty header or footer handling in HWPF HeaderStories</action>
            <action dev="POI-DEVELOPERS" type="fix">Avoid NPE in hssf.usermodel.HeaderFooter when stripping fields out</action>

src/scratchpad/src/org/apache/poi/hpbf/extractor/PublisherTextExtractor.java

@@ -24,6 +24,7 @@ import org.apache.poi.POIOLE2TextExtractor;
 import org.apache.poi.hpbf.HPBFDocument;
 import org.apache.poi.hpbf.model.qcbits.QCBit;
 import org.apache.poi.hpbf.model.qcbits.QCTextBit;
+import org.apache.poi.hpbf.model.qcbits.QCPLCBit.Type12;
 import org.apache.poi.poifs.filesystem.POIFSFileSystem;
 
 /**
@@ -31,6 +32,7 @@ import org.apache.poi.poifs.filesystem.POIFSFileSystem;
  */
 public class PublisherTextExtractor extends POIOLE2TextExtractor {
 	private HPBFDocument doc;
+	private boolean hyperlinksByDefault = false;
 	
 	public PublisherTextExtractor(HPBFDocument doc) {
 		super(doc);
@@ -43,6 +45,16 @@ public class PublisherTextExtractor extends POIOLE2TextExtractor {
 		this(new POIFSFileSystem(is));
 	}
 	
+	/**
+	 * Should a call to getText() return hyperlinks inline
+	 *  with the text?
+	 * Default is no
+	 */
+	public void setHyperlinksByDefault(boolean hyperlinksByDefault) {
+		this.hyperlinksByDefault = hyperlinksByDefault;
+	}
+
+	
 	public String getText() {
 		StringBuffer text = new StringBuffer();
 		
@@ -55,6 +67,24 @@ public class PublisherTextExtractor extends POIOLE2TextExtractor {
 			}
 		}
 		
+		// If requested, add in the hyperlinks
+		// Ideally, we'd do these inline, but the hyperlink
+		//  positions are relative to the text area the
+		//  hyperlink is in, and we have yet to figure out
+		//  how to tie that together.
+		if(hyperlinksByDefault) {
+			for(int i=0; i<bits.length; i++) {
+				if(bits[i] != null && bits[i] instanceof Type12) {
+					Type12 hyperlinks = (Type12)bits[i];
+					for(int j=0; j<hyperlinks.getNumberOfHyperlinks(); j++) {
+						text.append("<");
+						text.append(hyperlinks.getHyperlink(j));
+						text.append(">\n");
+					}
+				}
+			}
+		}
+		
 		// Get more text
 		// TODO
 		

src/scratchpad/src/org/apache/poi/hpbf/model/qcbits/QCPLCBit.java

@@ -167,6 +167,10 @@ public class QCPLCBit extends QCBit {
 	
 	/**
 	 * Type 12 holds hyperlinks, and is very complex.
+	 * There is normally one of these for each text
+	 *  area that contains at least one hyperlinks.
+	 * The character offsets are relative to the start
+	 *  of the text area that this applies to.
 	 */
 	public static class Type12 extends QCPLCBit {
 		private String[] hyperlinks;
@@ -249,6 +253,8 @@ public class QCPLCBit extends QCBit {
 		 * Returns where in the text (in characters) the
 		 *  hyperlink at the given index starts 
 		 *  applying to.
+		 * This position is relative to the text area that this
+		 *  PLCBit applies to.
 		 * @param number The hyperlink number, zero based
 		 */
 		public int getTextStartAt(int number) {
@@ -258,6 +264,8 @@ public class QCPLCBit extends QCBit {
 		 * Returns where in the text that this block
 		 *  of hyperlinks stops applying to. Normally,
 		 *  but not always the end of the text.
+		 * This position is relative to the text area that this
+		 *  PLCBit applies to.
 		 */
 		public int getAllTextEndAt() {
 			return preData[numberOfPLCs+1];

src/scratchpad/testcases/org/apache/poi/hpbf/extractor/TextPublisherTextExtractor.java

@@ -134,4 +134,41 @@ public class TextPublisherTextExtractor extends TestCase {
 		assertEquals(s2007, s2000);
 		assertEquals(s2007, s98);
 	}
+	
+	/**
+	 * Test that the hyperlink extraction stuff works as well
+	 *  as we can hope it to.
+	 */
+	public void testWithHyperlinks() throws Exception {
+		File f = new File(dir, "LinkAt10.pub");
+		HPBFDocument doc = new HPBFDocument(
+				new FileInputStream(f)
+		);
+
+		PublisherTextExtractor ext = 
+			new PublisherTextExtractor(doc);
+		ext.getText();
+		
+		// Default is no hyperlinks
+		assertEquals("1234567890LINK\n", ext.getText());
+		
+		// Turn on
+		ext.setHyperlinksByDefault(true);
+		assertEquals("1234567890LINK\n<http://poi.apache.org/>\n", ext.getText());
+		
+		
+		// Now a much more complex document
+		f = new File(dir, "Sample.pub");
+		ext = new PublisherTextExtractor(new FileInputStream(f));
+		ext.setHyperlinksByDefault(true);
+		String text = ext.getText();
+		
+		assertTrue(text.endsWith(
+				"<http://poi.apache.org/>\n" +
+				"<C:\\Documents and Settings\\Nick\\My Documents\\Booleans.xlsx>\n" +
+				"<>\n" +
+				"<mailto:dev@poi.apache.org?subject=HPBF>\n" +
+				"<mailto:dev@poi.apache.org?subject=HPBF>\n"
+		));
+	}
 }