Merge 'integration' to 'rc/1.26'

.github/workflows/test-pr.yml

@@ -7,16 +7,19 @@ on:
 jobs:
   build:
     name: Checkout and Build
-    runs-on: ubuntu-20.04
+    runs-on: ubuntu-latest
+
+    continue-on-error: true
 
     strategy:
+      fail-fast: false
       matrix:
-        java-version: [8, 11, 16, 17]
+        java-version: [8, 11, 16, 17, 21]
 
     steps:
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@v4
       - name: JDK setup
-        uses: actions/setup-java@v2
+        uses: actions/setup-java@v4
         with:
           java-version: ${{ matrix.java-version }}
           distribution: 'temurin'
@@ -31,7 +34,7 @@ jobs:
           path: target/site/jacoco/
   coverage:
     name: Quality Assurance
-    runs-on: ubuntu-20.04
+    runs-on: ubuntu-latest
     needs: [ build ]
 
     steps:

README.md

@@ -29,7 +29,7 @@ Quick Start
 
 In order to build the parser you'll need:
 
- * Java 8 - 17, which can be downloaded [from Oracle](https://www.oracle.com/technetwork/java/javase/downloads/index.html), or for Linux users [OpenJDK](http://openjdk.java.net/install/index.html).
+ * Java 8 - 21, which can be downloaded [from Oracle](https://www.oracle.com/technetwork/java/javase/downloads/index.html), or for Linux users [OpenJDK](http://openjdk.java.net/install/index.html).
  * [Maven v3+](https://maven.apache.org/)
 
 ### Building the veraPDF Parser

src/main/java/org/verapdf/parser/COSParser.java

@@ -3,10 +3,8 @@
 import org.verapdf.as.exceptions.StringExceptions;
 import org.verapdf.cos.*;
 import org.verapdf.parser.postscript.PSObject;
-import org.verapdf.pd.encryption.StandardSecurityHandler;
 
 import java.io.IOException;
-import java.security.GeneralSecurityException;
 import java.util.LinkedList;
 import java.util.Queue;
 import java.util.logging.Level;

src/main/java/org/verapdf/pd/PDNameTreeNode.java

@@ -26,7 +26,6 @@
 import org.verapdf.cos.COSObjType;
 import org.verapdf.cos.COSObject;
 import org.verapdf.exceptions.LoopedException;
-import org.verapdf.pd.structure.NameTreeIterator;
 
 import java.util.*;
 
@@ -165,14 +164,22 @@ private COSObject getObject(String key, Set<COSKey> visitedKeys) {
 		return null;
 	}
 
+	private List<COSObject> getObjects() {
+		List<COSObject> result = new LinkedList<>(getNames().values());
+		for (PDNameTreeNode kid : getKids()) {
+			result.addAll(kid.getObjects());
+		}
+		return result;
+	}
+
 	@Override
-	public NameTreeIterator iterator() {
-		return new NameTreeIterator(this);
+	public Iterator<COSObject> iterator() {
+		return getObjects().iterator();
 	}
 
 	public Long size() {
 		long i = 0;
-		NameTreeIterator iterator = iterator();
+		Iterator<COSObject> iterator = iterator();
 		for (; iterator.hasNext(); i++) {
 			iterator.next();
 		}

src/main/java/org/verapdf/pd/font/PDCIDFont.java

@@ -162,16 +162,15 @@ public FontProgram getFontProgram() {
             this.isFontParsed = true;
 
             if (fontDescriptor.canParseFontFile(ASAtom.FONT_FILE2) &&
-                    this.getSubtype() == ASAtom.CID_FONT_TYPE2) {
+                    this.getSubtypeEntryValue() == ASAtom.CID_FONT_TYPE2) {
                 COSStream trueTypeFontFile = fontDescriptor.getFontFile2();
                 COSKey key = trueTypeFontFile.getObjectKey();
                 COSObject cidToGIDMap = this.getCIDToGIDMap();
                 String fontProgramID = FontProgramIDGenerator.getCIDFontType2ProgramID(key, this.cMap, cidToGIDMap);
                 this.fontProgram = StaticResources.getCachedFont(fontProgramID);
                 if (fontProgram == null) {
                     try (ASInputStream fontData = trueTypeFontFile.getData(COSStream.FilterFlags.DECODE)) {
-                        this.fontProgram = new CIDFontType2Program(
-                                fontData, this.cMap, cidToGIDMap);
+                        this.fontProgram = new CIDFontType2Program(fontData, this.cMap, cidToGIDMap);
                         StaticResources.cacheFontProgram(fontProgramID, this.fontProgram);
                     } catch (IOException e) {
                         LOGGER.log(Level.FINE, "Can't read TrueType font program.", e);
@@ -193,17 +192,18 @@ public FontProgram getFontProgram() {
                             }
                         }
                     } else if (ASAtom.OPEN_TYPE == subtype.getName()) {
-                        ASAtom fontName = ASAtom.getASAtom(this.getName());
-                        boolean isCFF = fontName != ASAtom.TRUE_TYPE && fontName != ASAtom.CID_FONT_TYPE2;
+                        boolean isCFF = this.getSubtypeEntryValue() != ASAtom.TRUE_TYPE && this.getSubtypeEntryValue() != ASAtom.CID_FONT_TYPE2;
+                        boolean isCIDFontType2 = this.getSubtypeEntryValue() == ASAtom.CID_FONT_TYPE2;
                         boolean isSymbolic = this.isSymbolic();
                         COSObject encoding = this.getEncoding();
                         String fontProgramID = FontProgramIDGenerator.getOpenTypeFontProgramID(key, isCFF, isSymbolic, encoding, this.cMap, isSubset);
                         this.fontProgram = StaticResources.getCachedFont(fontProgramID);
                         if (fontProgram == null) {
+                            COSObject cidToGIDMap = this.getCIDToGIDMap();
                             try (ASInputStream fontData = fontFile.getData(COSStream.FilterFlags.DECODE)) {
                                 this.fontProgram = new OpenTypeFontProgram(
-                                        fontData, isCFF, isSymbolic, encoding,
-                                        this.cMap, isSubset);
+                                        fontData, isCFF, isCIDFontType2, isSymbolic, encoding,
+                                        this.cMap, isSubset, cidToGIDMap);
                                 StaticResources.cacheFontProgram(fontProgramID, this.fontProgram);
                             }
                         }

src/main/java/org/verapdf/pd/font/PDFont.java

@@ -116,6 +116,10 @@ public String getType() {
     public ASAtom getSubtype() {
         return this.subtype;
     }
+
+    public ASAtom getSubtypeEntryValue() {
+        return this.subtype;
+    }
 
     /**
      * @return true if the font flags in the font descriptor dictionary mark

src/main/java/org/verapdf/pd/font/cff/CFFCIDFontProgram.java

@@ -57,13 +57,8 @@ public class CFFCIDFontProgram extends CFFFontBaseParser implements FontProgram
     CFFCIDFontProgram(SeekableInputStream stream, CFFIndex definedNames, CFFIndex globalSubrs,
                       long topDictBeginOffset, long topDictEndOffset, CMap externalCMap,
                       boolean isSubset) {
-        super(stream);
-        this.definedNames = definedNames;
-        this.globalSubrs = globalSubrs;
-        this.topDictBeginOffset = topDictBeginOffset;
-        this.topDictEndOffset = topDictEndOffset;
+        super(stream, definedNames, globalSubrs, topDictBeginOffset, topDictEndOffset, isSubset);
         this.externalCMap = externalCMap;
-        this.isSubset = isSubset;
     }
 
     /**

src/main/java/org/verapdf/pd/font/cff/CFFFontBaseParser.java

@@ -36,7 +36,7 @@
  *
  * @author Sergey Shemyakov
  */
-abstract class CFFFontBaseParser extends CFFFileBaseParser {
+class CFFFontBaseParser extends CFFFileBaseParser {
 
     private static final Logger LOGGER = Logger.getLogger(CFFFontBaseParser.class.getCanonicalName());
 
@@ -72,13 +72,39 @@ abstract class CFFFontBaseParser extends CFFFileBaseParser {
     //Subrs
     protected long subrsOffset = -1;
 
+    private boolean containsROS = false;
+
     public CFFFontBaseParser(SeekableInputStream source) {
         super(source);
         stack = new ArrayList<>(48);
         this.charStringType = 2;
         this.charSetOffset = 0; // default
     }
 
+    public CFFFontBaseParser(SeekableInputStream stream, CFFIndex definedNames, CFFIndex globalSubrs,
+                        long topDictBeginOffset, long topDictEndOffset, boolean isSubset) {
+        this(stream);
+        this.definedNames = definedNames;
+        this.globalSubrs = globalSubrs;
+        this.topDictBeginOffset = topDictBeginOffset;
+        this.topDictEndOffset = topDictEndOffset;
+        this.isSubset = isSubset;
+    }
+
+    protected boolean containsROS() {
+        try {
+            this.source.seek(topDictBeginOffset);
+            while (this.source.getOffset() < topDictEndOffset) {
+                readTopDictUnit();
+            }     
+            if (containsROS) {
+                return true;
+            }
+        } catch (IOException ignored) {
+        }
+        return false;
+    }
+
     protected void readTopDictUnit() throws IOException {
         try {
             int next = this.source.peek();
@@ -125,6 +151,9 @@ protected void readTopDictUnit() throws IOException {
                                             this.stack.get(this.stack.size() - 1).getInteger();
                                     this.stack.clear();
                                     break;
+                                case 30:
+                                    this.containsROS = true;
+                                    break;
                                 default:
                                     readTopDictTwoByteOps(next);
                             }

src/main/java/org/verapdf/pd/font/cff/CFFFontProgram.java

@@ -76,7 +76,7 @@ public void parseFont() throws IOException {
             }
             this.definedNames = this.readIndex();
             CFFIndex globalSubrs = this.readIndex();
-            if (isCIDFont(top.get(0))) {
+            if (isCIDFont(top, globalSubrs, topOffset)) {
                 font = new CFFCIDFontProgram(this.source, this.definedNames, globalSubrs,
                         topOffset + top.getOffset(0) - 1 + top.getOffsetShift(),
                         topOffset + top.getOffset(1) - 1 + top.getOffsetShift(),
@@ -91,9 +91,10 @@ public void parseFont() throws IOException {
         }
     }
 
-    private boolean isCIDFont(byte[] topDict) {
+    private boolean isCIDFont(CFFIndex top, CFFIndex globalSubrs, long topOffset) {
         try {
             byte rosOffset;
+            byte[] topDict = top.get(0);
             int supplementFirstByte = topDict[4] & 0xFF;    // checking if first operator is really ROS
             if (supplementFirstByte < 247 && supplementFirstByte > 31) {
                 rosOffset = 5;
@@ -104,25 +105,25 @@ private boolean isCIDFont(byte[] topDict) {
             } else if (supplementFirstByte == 29) {
                 rosOffset = 9;
             } else {
-                return isContainsROS(topDict);
+                return containsROS(top, globalSubrs, topOffset);
             }
             if (topDict[rosOffset] == 12 && topDict[rosOffset + 1] == 30) {
                 isCIDFont = true;
                 return true;
             }
-            return isContainsROS(topDict);
+            return containsROS(top, globalSubrs, topOffset);
         } catch (ArrayIndexOutOfBoundsException ex) {
-            return isContainsROS(topDict);
+            return containsROS(top, globalSubrs, topOffset);
         }
     }
 
-    private boolean isContainsROS(byte[] topDict) {
-        for (int rosOffset = 0; rosOffset < topDict.length - 2; rosOffset++) {
-            if (topDict[rosOffset] == 12 && topDict[rosOffset + 1] == 30) {
-                LOGGER.log(Level.WARNING, "The Top DICT does not begin with ROS operator");
-                isCIDFont = true;
-                return true;
-            }
+    private boolean containsROS(CFFIndex top, CFFIndex globalSubrs, long topOffset) {
+        if (new CFFFontBaseParser(this.source, this.definedNames, globalSubrs,
+                topOffset + top.getOffset(0) - 1 + top.getOffsetShift(),
+                topOffset + top.getOffset(1) - 1 + top.getOffsetShift(),
+                this.isSubset).containsROS()) {
+            LOGGER.log(Level.WARNING, "The Top DICT does not begin with ROS operator");
+            return true;
         }
         return false;
     }

src/main/java/org/verapdf/pd/font/cff/CFFType1FontProgram.java

@@ -58,15 +58,10 @@ public class CFFType1FontProgram extends CFFFontBaseParser implements FontProgra
     CFFType1FontProgram(SeekableInputStream stream, CFFIndex definedNames, CFFIndex globalSubrs,
                         long topDictBeginOffset, long topDictEndOffset,
                         CMap externalCMap, boolean isSubset) {
-        super(stream);
+        super(stream, definedNames, globalSubrs, topDictBeginOffset, topDictEndOffset, isSubset);
         encodingOffset = 0;
         encoding = new int[256];
-        this.definedNames = definedNames;
-        this.globalSubrs = globalSubrs;
-        this.topDictBeginOffset = topDictBeginOffset;
-        this.topDictEndOffset = topDictEndOffset;
         this.externalCMap = externalCMap;
-        this.isSubset = isSubset;
         fontMatrix = new float[6];
         System.arraycopy(DEFAULT_FONT_MATRIX, 0, this.fontMatrix, 0,
                 DEFAULT_FONT_MATRIX.length);

src/main/java/org/verapdf/pd/font/cmap/CMapParser.java

@@ -75,9 +75,10 @@ public void parse() throws IOException, PostScriptException {
             getBaseParser().initializeToken();
             //Skipping starting comments
             getBaseParser().skipSpaces(true);
+            COSObject nextObject = nextObject();
             while (getBaseParser().getToken().type != Token.Type.TT_EOF) {
-                COSObject nextObject = nextObject();
                 processObject(nextObject);
+                nextObject = nextObject();
             }
         } finally {
             this.getSource().close();    // We close stream after first reading attempt

src/main/java/org/verapdf/pd/font/opentype/OpenTypeFontProgram.java

@@ -27,6 +27,7 @@
 import org.verapdf.pd.font.FontProgram;
 import org.verapdf.pd.font.cff.CFFFontProgram;
 import org.verapdf.pd.font.cmap.CMap;
+import org.verapdf.pd.font.truetype.CIDFontType2Program;
 import org.verapdf.pd.font.truetype.TrueTypeFontProgram;
 import org.verapdf.tools.StaticResources;
 import org.verapdf.tools.resource.ASFileStreamCloser;
@@ -45,6 +46,7 @@ public class OpenTypeFontProgram implements FontProgram {
     // See TrueTypeFontParser table logic
 
     private final boolean isCFF;
+    private final boolean isCIDFontType2;
     private final boolean isSymbolic;
     private final boolean isSubset;
     private final COSObject encoding;
@@ -54,6 +56,7 @@ public class OpenTypeFontProgram implements FontProgram {
     private boolean attemptedParsing = false;
     private boolean successfullyParsed = false;
     private final CMap externalCMap;
+    private final COSObject cidToGIDMap;
 
     /**
      * Constructor from stream, containing font data, and encoding details.
@@ -62,14 +65,16 @@ public class OpenTypeFontProgram implements FontProgram {
      * @param isSymbolic is true if font is marked as symbolic.
      * @param encoding   is value of /Encoding in font dictionary.
      */
-    public OpenTypeFontProgram(ASInputStream source, boolean isCFF, boolean isSymbolic,
-                               COSObject encoding, CMap externalCMap, boolean isSubset) {
+    public OpenTypeFontProgram(ASInputStream source, boolean isCFF, boolean isCIDFontType2, boolean isSymbolic,
+                               COSObject encoding, CMap externalCMap, boolean isSubset, COSObject cidToGIDMap) {
         this.source = source;
         this.isCFF = isCFF;
+        this.isCIDFontType2 = isCIDFontType2;
         this.isSymbolic = isSymbolic;
         this.encoding = encoding;
         this.externalCMap = externalCMap;
         this.isSubset = isSubset;
+        this.cidToGIDMap = cidToGIDMap;
     }
 
     /**
@@ -134,14 +139,17 @@ public boolean isSuccessfulParsing() {
     public void parseFont() throws IOException {
         if (!attemptedParsing) {
             attemptedParsing = true;
-            if (!isCFF) {
-                this.font = new TrueTypeFontProgram(source, isSymbolic, encoding);
+            if (isCIDFontType2) {
+                this.font = new CIDFontType2Program(source, externalCMap, cidToGIDMap);
                 this.font.parseFont();
-            } else {
+            } else if (isCFF) {
                 try (ASInputStream cffTable = getCFFTable()) {
                     this.font = new CFFFontProgram(cffTable, externalCMap, isSubset);
                     this.font.parseFont();
                 }
+            } else {
+                this.font = new TrueTypeFontProgram(source, isSymbolic, encoding);
+                this.font.parseFont();
             }
             StaticResources.cacheFontProgram(null, this.font);
             this.successfullyParsed = true;

src/main/java/org/verapdf/pd/font/truetype/BaseTrueTypeProgram.java

@@ -84,7 +84,7 @@ public void parseFont() throws IOException {
      * this True Type font.
      */
     public int getNrOfCMaps() {
-        if(this.parser.getCmapParser() != null) {
+        if (this.parser.getCmapParser() != null) {
             return this.parser.getCmapParser().getCmapInfos().length;
         } else {
             return 0;