Author: tilman
Date: Thu Nov 23 04:32:48 2023
New Revision: 1914058
URL: http://svn.apache.org/viewvc?rev=1914058&view=rev
Log:
PDFBOX-5713: rewritten parser to accept more than 3 segments and rearrange them
as ASCII-BINARY-ASCII, as described in "3.3 IBM PC Format" of Adobe Technical
Note #5040
Modified:
pdfbox/branches/3.0/fontbox/pom.xml
pdfbox/branches/3.0/fontbox/src/main/java/org/apache/fontbox/pfb/PfbParser.java
pdfbox/branches/3.0/fontbox/src/test/java/org/apache/fontbox/pfb/PfbParserTest.java
Modified: pdfbox/branches/3.0/fontbox/pom.xml
URL:
http://svn.apache.org/viewvc/pdfbox/branches/3.0/fontbox/pom.xml?rev=1914058&r1=1914057&r2=1914058&view=diff
==============================================================================
--- pdfbox/branches/3.0/fontbox/pom.xml (original)
+++ pdfbox/branches/3.0/fontbox/pom.xml Thu Nov 23 04:32:48 2023
@@ -143,6 +143,18 @@
<sha512>2787fcecc0feb1c9e6ff0d8de6193658413863e44eaab572751ca7e6c3b369c0a9731f4952cb0821f307760f0422f77c5f0d3fe7df6b054643fb39423e8d70ee</sha512>
</configuration>
</execution>
+ <execution>
+ <id>PDFBOX-5713</id>
+ <phase>generate-test-resources</phase>
+ <goals>
+ <goal>wget</goal>
+ </goals>
+ <configuration>
+
<url>https://issues.apache.org/jira/secure/attachment/13064282/DejaVuSerifCondensed.pfb</url>
+
<outputDirectory>${project.build.directory}/fonts</outputDirectory>
+
<sha512>6ef13c3497862dc8e4c2a4261bc3a7ef3e2dd75e00ae2af4912b236b387225541db76c72854fbb2323d1064311ffdda9e64ed7065afc3a7d13f5b71b7df2f2ef</sha512>
+ </configuration>
+ </execution>
</executions>
</plugin>
</plugins>
Modified:
pdfbox/branches/3.0/fontbox/src/main/java/org/apache/fontbox/pfb/PfbParser.java
URL:
http://svn.apache.org/viewvc/pdfbox/branches/3.0/fontbox/src/main/java/org/apache/fontbox/pfb/PfbParser.java?rev=1914058&r1=1914057&r2=1914058&view=diff
==============================================================================
---
pdfbox/branches/3.0/fontbox/src/main/java/org/apache/fontbox/pfb/PfbParser.java
(original)
+++
pdfbox/branches/3.0/fontbox/src/main/java/org/apache/fontbox/pfb/PfbParser.java
Thu Nov 23 04:32:48 2023
@@ -23,7 +23,11 @@ import java.io.IOException;
import java.io.InputStream;
import java.nio.file.Files;
import java.nio.file.Paths;
+import java.util.ArrayList;
import java.util.Arrays;
+import java.util.List;
+import org.apache.commons.logging.Log;
+import org.apache.commons.logging.LogFactory;
/**
* Parser for a pfb-file.
@@ -33,6 +37,8 @@ import java.util.Arrays;
*/
public class PfbParser
{
+ private static final Log LOG = LogFactory.getLog(PfbParser.class);
+
/**
* the pfb header length.
* (start-marker (1 byte), ascii-/binary-marker (1 byte), size (4 byte))
@@ -56,10 +62,9 @@ public class PfbParser
private static final int BINARY_MARKER = 0x02;
/**
- * The record types in the pfb-file.
+ * the EOF marker.
*/
- private static final int[] PFB_RECORDS = {ASCII_MARKER, BINARY_MARKER,
- ASCII_MARKER};
+ private static final int EOF_MARKER = 0x03;
/**
* buffersize.
@@ -72,9 +77,9 @@ public class PfbParser
private byte[] pfbdata;
/**
- * the lengths of the records.
+ * the lengths of the records (ASCII, BINARY, ASCII)
*/
- private int[] lengths;
+ private final int[] lengths = new int[3];
// sample (pfb-file)
// 00000000 80 01 8b 15 00 00 25 21 50 53 2d 41 64 6f 62 65
@@ -123,43 +128,92 @@ public class PfbParser
{
throw new IOException("PFB header missing");
}
+ // read into segments and keep them
+ List<Integer> typeList = new ArrayList<>(3);
+ List<byte[]> barrList = new ArrayList<>();
ByteArrayInputStream in = new ByteArrayInputStream(pfb);
- pfbdata = new byte[pfb.length - PFB_HEADER_LENGTH];
- lengths = new int[PFB_RECORDS.length];
- int pointer = 0;
- for (int records = 0; records < PFB_RECORDS.length; records++)
+ int total = 0;
+ do
{
- if (in.read() != START_MARKER)
+ int r = in.read();
+ if (r == -1 && total > 0)
+ {
+ break; // EOF
+ }
+ if (r != START_MARKER)
{
throw new IOException("Start marker missing");
}
-
- if (in.read() != PFB_RECORDS[records])
+ int recordType = in.read();
+ if (recordType == EOF_MARKER)
{
- throw new IOException("Incorrect record type");
+ break;
+ }
+ if (recordType != ASCII_MARKER && recordType != BINARY_MARKER)
+ {
+ throw new IOException("Incorrect record type: " + recordType);
}
int size = in.read();
size += in.read() << 8;
size += in.read() << 16;
size += in.read() << 24;
- lengths[records] = size;
- if (pointer >= pfbdata.length)
+ LOG.debug("record type: " + recordType + ", segment size: " +
size);
+ byte ar[] = new byte[size];
+ int got = in.read(ar);
+ if (got != size)
{
- throw new EOFException("attempted to read past EOF");
+ throw new EOFException("EOF while reading PFB font");
}
- if (size > pfbdata.length - pointer)
+ total += size;
+ typeList.add(recordType);
+ barrList.add(ar);
+ }
+ while (true);
+
+ // We now have ASCII and binary segments. Lets arrange these so that
the ASCII segments
+ // come first, then the binary segments, then the last ASCII segment
if it is
+ // 0000... cleartomark
+
+ pfbdata = new byte[total];
+ byte[] cleartomarkSegment = null;
+ int dstPos = 0;
+
+ // copy the ASCII segments
+ for (int i = 0; i < typeList.size(); ++i)
+ {
+ if (typeList.get(i) != ASCII_MARKER)
{
- throw new EOFException("attempted to read " + size + " bytes
at position " + pointer +
- " into array of size " + pfbdata.length + ", but only
space for " +
- (pfbdata.length - pointer) + " bytes left");
+ continue;
}
- int got = in.read(pfbdata, pointer, size);
- if (got < 0)
+ byte[] ar = barrList.get(i);
+ if (i == typeList.size() - 1 && ar.length < 600 && new
String(ar).contains("cleartomark"))
{
- throw new EOFException();
+ cleartomarkSegment = ar;
+ continue;
}
- pointer += got;
+ System.arraycopy(ar, 0, pfbdata, dstPos, ar.length);
+ dstPos += ar.length;
+ }
+ lengths[0] = dstPos;
+
+ // copy the binary segments
+ for (int i = 0; i < typeList.size(); ++i)
+ {
+ if (typeList.get(i) != BINARY_MARKER)
+ {
+ continue;
+ }
+ byte[] ar = barrList.get(i);
+ System.arraycopy(ar, 0, pfbdata, dstPos, ar.length);
+ dstPos += ar.length;
+ }
+ lengths[1] = dstPos - lengths[0];
+
+ if (cleartomarkSegment != null)
+ {
+ System.arraycopy(cleartomarkSegment, 0, pfbdata, dstPos,
cleartomarkSegment.length);
+ lengths[2] = cleartomarkSegment.length;
}
}
Modified:
pdfbox/branches/3.0/fontbox/src/test/java/org/apache/fontbox/pfb/PfbParserTest.java
URL:
http://svn.apache.org/viewvc/pdfbox/branches/3.0/fontbox/src/test/java/org/apache/fontbox/pfb/PfbParserTest.java?rev=1914058&r1=1914057&r2=1914058&view=diff
==============================================================================
---
pdfbox/branches/3.0/fontbox/src/test/java/org/apache/fontbox/pfb/PfbParserTest.java
(original)
+++
pdfbox/branches/3.0/fontbox/src/test/java/org/apache/fontbox/pfb/PfbParserTest.java
Thu Nov 23 04:32:48 2023
@@ -66,6 +66,34 @@ class PfbParserTest
}
/**
+ * PDFBOX-5713: font with several binary segments.
+ *
+ * @throws IOException
+ */
+ @Test
+ void testPfbPDFBox5713() throws IOException
+ {
+ Type1Font font;
+ try (InputStream is = new
FileInputStream("target/fonts/DejaVuSerifCondensed.pfb"))
+ {
+ font = Type1Font.createWithPFB(is);
+ }
+ Assertions.assertEquals("Version 2.33", font.getVersion());
+ Assertions.assertEquals("DejaVuSerifCondensed", font.getFontName());
+ Assertions.assertEquals("DejaVu Serif Condensed", font.getFullName());
+ Assertions.assertEquals("DejaVu Serif Condensed",
font.getFamilyName());
+ Assertions.assertEquals("Copyright [c] 2003 by Bitstream, Inc. All
Rights Reserved.", font.getNotice());
+ Assertions.assertEquals(false, font.isFixedPitch());
+ Assertions.assertEquals(false, font.isForceBold());
+ Assertions.assertEquals(0, font.getItalicAngle());
+ Assertions.assertEquals("Book", font.getWeight());
+ Assertions.assertTrue(font.getEncoding() instanceof BuiltInEncoding);
+ Assertions.assertEquals(5959, font.getASCIISegment().length);
+ Assertions.assertEquals(1056090, font.getBinarySegment().length);
+ Assertions.assertEquals(3399, font.getCharStringsDict().size());
+ }
+
+ /**
* Test 0 length font.
*/
@Test
