svn commit: r1923833 - /pdfbox/branches/2.0/pdfbox/src/main/java/org/apache/pdfbox/pdfparser/PDFStreamParser.java

[tilman]

Author: tilman
Date: Sat Feb 15 10:26:28 2025
New Revision: 1923833

URL: http://svn.apache.org/viewvc?rev=1923833&view=rev
Log:
PDFBOX-5957: refactor, add comments, improve logging

Modified:
    
pdfbox/branches/2.0/pdfbox/src/main/java/org/apache/pdfbox/pdfparser/PDFStreamParser.java

Modified: 
pdfbox/branches/2.0/pdfbox/src/main/java/org/apache/pdfbox/pdfparser/PDFStreamParser.java
URL: 
http://svn.apache.org/viewvc/pdfbox/branches/2.0/pdfbox/src/main/java/org/apache/pdfbox/pdfparser/PDFStreamParser.java?rev=1923833&r1=1923832&r2=1923833&view=diff
==============================================================================
--- 
pdfbox/branches/2.0/pdfbox/src/main/java/org/apache/pdfbox/pdfparser/PDFStreamParser.java
 (original)
+++ 
pdfbox/branches/2.0/pdfbox/src/main/java/org/apache/pdfbox/pdfparser/PDFStreamParser.java
 Sat Feb 15 10:26:28 2025
@@ -372,7 +372,8 @@ public class PDFStreamParser extends Bas
         boolean noBinData = true;
         int startOpIdx = -1;
         int endOpIdx = -1;
-        
+        String s = "";
+
         if (readBytes > 0)
         {
             for (int bIdx = 0; bIdx < readBytes; bIdx++)
@@ -399,25 +400,27 @@ public class PDFStreamParser extends Bas
             // PDFBOX-3742: just assuming that 1-3 non blanks is a PDF 
operator isn't enough
             if (endOpIdx != -1 && startOpIdx != -1)
             {
-                // usually, the operator here is Q, sometimes EMC 
(PDFBOX-2376), S (PDFBOX-3784).
-                String s = new String(binCharTestArr, startOpIdx, endOpIdx - 
startOpIdx);
+                // usually, the operator here is Q, sometimes EMC 
(PDFBOX-2376), S (PDFBOX-3784)
+                s = new String(binCharTestArr, startOpIdx, endOpIdx - 
startOpIdx);
                 if (!"Q".equals(s) && !"EMC".equals(s) && !"S".equals(s))
                 {
+                    // operator is not Q, not EMC, not S -> assume binary data
                     noBinData = false;
                 }
             }
 
-            // only if not close to eof
-            if (readBytes == MAX_BIN_CHAR_TEST_LENGTH) 
+            // only if not close to EOF
+            if (startOpIdx != -1 && readBytes == MAX_BIN_CHAR_TEST_LENGTH) 
             {
-                // a PDF operator is 1-3 bytes long
-                if (startOpIdx != -1 && endOpIdx == -1)
+                if (endOpIdx == -1)
                 {
                     endOpIdx = MAX_BIN_CHAR_TEST_LENGTH;
+                    s = new String(binCharTestArr, startOpIdx, endOpIdx - 
startOpIdx);
                 }
-                if (endOpIdx != -1 && startOpIdx != -1 && endOpIdx - 
startOpIdx > 3)
+                // a PDF operator is 1-3 bytes long
+                if (endOpIdx - startOpIdx > 3)
                 {
-                    noBinData = false;
+                    noBinData = false; // "operator" too long, assume binary 
data
                 }
             }
             pdfSource.unread(binCharTestArr, 0, readBytes);
@@ -425,7 +428,7 @@ public class PDFStreamParser extends Bas
         if (!noBinData)
         {
             LOG.warn("ignoring 'EI' assumed to be in the middle of inline 
image at stream offset " + 
-                    pdfSource.getPosition());
+                    pdfSource.getPosition() + ", s = '" + s + "'");
         }
         return noBinData;
     }