Fix average glyph width calculation. Avoid divison by zero. Using average width if space is not found in font in case of text extraction. Unignore a parsing test

Author: Snipx

basics/src/main/java/com/itextpdf/basics/font/CidFont.java

@@ -23,7 +23,9 @@ public class CidFont extends FontProgram {
         for(int code : hMetrics.getKeys()) {
             avgWidth += hMetrics.get(code);
         }
-        avgWidth /= hMetrics.size();
+        if (hMetrics.size() != 0) {
+            avgWidth /= hMetrics.size();
+        }
         Map<String, Object> fontDesc = CidFontProperties.getAllFonts().get(fontNames.getFontName());
         if (fontDesc == null) {
             throw new PdfException("no.such.predefined.font.1").setMessageParams(fontName);

basics/src/main/java/com/itextpdf/basics/font/TrueTypeFont.java

@@ -302,7 +302,9 @@ private void initializeFontProperties() throws IOException {
             avgWidth += glyph.getWidth();
         }
 
-        avgWidth /= codeToGlyph.size();
+        if (codeToGlyph.size() != 0) {
+            avgWidth /= codeToGlyph.size();
+        }
 
         readGdefTable();
         readGsubTable();

basics/src/main/java/com/itextpdf/basics/font/Type1Font.java

@@ -340,7 +340,9 @@ protected void process() throws IOException {
             avgWidth += WX;
             widthCount++;
         }
-        avgWidth /= widthCount;
+        if (widthCount != 0) {
+            avgWidth /= widthCount;
+        }
         if (startKernPairs) {
             String metricsPath = fontParser.getAfmPath();
             if (metricsPath != null) {

core/src/main/java/com/itextpdf/core/font/DocTrueTypeFont.java

@@ -46,9 +46,12 @@ static TrueTypeFont createFontProgram(PdfDictionary fontDictionary, FontEncoding
             }
             if (widths[i] > 0) {
                 glyphsWithWidths++;
+                fontProgram.avgWidth += widths[i];
             }
         }
-        fontProgram.avgWidth /= glyphsWithWidths;
+        if (glyphsWithWidths != 0) {
+            fontProgram.avgWidth /= glyphsWithWidths;
+        }
         return fontProgram;
     }
 
@@ -67,8 +70,11 @@ static TrueTypeFont createFontProgram(PdfDictionary fontDictionary, CMapToUnicod
                 Glyph glyph = new Glyph(cid, width, cid2Uni.get(cid));
                 fontProgram.codeToGlyph.put(cid, glyph);
                 fontProgram.unicodeToGlyph.put(glyph.getUnicode(), glyph);
+                fontProgram.avgWidth += width;
+            }
+            if (fontProgram.codeToGlyph.size() != 0) {
+                fontProgram.avgWidth /= fontProgram.codeToGlyph.size();
             }
-            fontProgram.avgWidth /= fontProgram.codeToGlyph.size();
         }
 
         if (fontProgram.codeToGlyph.get(0) == null) {

core/src/main/java/com/itextpdf/core/font/DocType1Font.java

@@ -1,9 +1,7 @@
 package com.itextpdf.core.font;
 
-import com.itextpdf.basics.IntHashtable;
 import com.itextpdf.basics.font.FontEncoding;
 import com.itextpdf.basics.font.Type1Font;
-import com.itextpdf.basics.font.cmap.CMapToUnicode;
 import com.itextpdf.basics.font.otf.Glyph;
 import com.itextpdf.core.pdf.PdfArray;
 import com.itextpdf.core.pdf.PdfDictionary;
@@ -12,9 +10,6 @@
 import com.itextpdf.core.pdf.PdfStream;
 import com.itextpdf.core.pdf.PdfString;
 
-import java.io.IOException;
-import java.util.Map;
-
 class DocType1Font extends Type1Font implements DocFontProgram {
 
     private PdfStream fontFile;
@@ -64,9 +59,12 @@ static Type1Font createFontProgram(PdfDictionary fontDictionary, FontEncoding fo
             }
             if (widths[i] > 0) {
                 glyphsWithWidths++;
+                fontProgram.avgWidth += widths[i];
             }
         }
-        fontProgram.avgWidth /= glyphsWithWidths;
+        if (glyphsWithWidths != 0) {
+            fontProgram.avgWidth /= glyphsWithWidths;
+        }
         return fontProgram;
     }
 

core/src/main/java/com/itextpdf/core/parser/TextRenderInfo.java

@@ -309,9 +309,11 @@ private float convertHeightFromTextSpaceToUserSpace(float height){
      */
     private float getUnscaledFontSpaceWidth(){
         char charToUse = ' ';
-        if (gs.getFont().getWidth(charToUse) == 0)
-            charToUse = '\u00A0';
-        return getStringWidth(String.valueOf(charToUse));
+        if (gs.getFont().getWidth(charToUse) == 0) {
+            return gs.getFont().getFontProgram().getAvgWidth() / 1000f;
+        } else {
+            return getStringWidth(String.valueOf(charToUse));
+        }
     }
 
     /**

model/src/test/java/com/itextpdf/model/SimpleTextExtractionStrategyTest.java

@@ -7,9 +7,9 @@
 import com.itextpdf.core.font.PdfFont;
 import com.itextpdf.core.font.PdfFontFactory;
 import com.itextpdf.core.parser.GlyphTextEventListener;
+import com.itextpdf.core.parser.PdfTextExtractor;
 import com.itextpdf.core.parser.SimpleTextExtractionStrategy;
 import com.itextpdf.core.parser.TextExtractionStrategy;
-import com.itextpdf.core.parser.PdfTextExtractor;
 import com.itextpdf.core.pdf.PdfDocument;
 import com.itextpdf.core.pdf.PdfReader;
 import com.itextpdf.core.pdf.PdfString;
@@ -27,7 +27,6 @@
 import java.io.IOException;
 
 import org.junit.Assert;
-import org.junit.Ignore;
 import org.junit.Test;
 import org.junit.experimental.categories.Category;
 
@@ -190,7 +189,6 @@ public void extractFromPage229() throws IOException {
     }
 
     @Test
-    @Ignore("Seems to be a problem with decode")
     public void extractFromIsoTc171() throws IOException {
         if (this.getClass() != SimpleTextExtractionStrategyTest.class)
             return;