Martin Mirchev created PDFBOX-6328:
--------------------------------------

             Summary: ArrayIndexOutOfBoundsException when a CFF font has more 
fonts in its Name INDEX than entries in its Top DICT INDEX
                 Key: PDFBOX-6328
                 URL: https://issues.apache.org/jira/browse/PDFBOX-6328
             Project: PDFBox
          Issue Type: Bug
          Components: FontBox
    Affects Versions: 4.0.0, 3.0.8 PDFBox
            Reporter: Martin Mirchev
         Attachments: cff-ok.pdf, cff-two-names.pdf

{{CFFParser.parse()}} reads {{topDictIndex[i]}} for every entry of the Name 
INDEX without checking that the Top DICT INDEX has as many entries. A CFF whose 
Name INDEX lists two fonts but whose Top DICT INDEX has one throws:

{code}
java.lang.ArrayIndexOutOfBoundsException: Index 1 out of bounds for length 1
    at org.apache.fontbox.cff.CFFParser.parse(CFFParser.java:214)
{code}

{{PDType1CFont}} catches only {{IOException}} around the parse, so for a PDF 
with such a font embedded ({{FontFile3}}, {{/Subtype /Type1C}}) the exception 
escapes {{PDFRenderer.renderImage()}} and {{PDFTextStripper.getText()}}. 
Attached {{cff-two-names.pdf}} embeds {{FoglihtenNo07.otf}}'s CFF (from the 
fontbox test resources) with its Name INDEX rewritten in place to two names 
(same length, offsets unchanged); {{cff-ok.pdf}} embeds the unmodified CFF and 
renders and extracts normally (3/3 runs each).

A possible change: reject the font when the Top DICT INDEX is shorter than the 
Name INDEX. With it, {{cff-two-names.pdf}} logs "Can't read the embedded Type1C 
font", renders and extracts text with the fallback font:

{code}
--- a/fontbox/src/main/java/org/apache/fontbox/cff/CFFParser.java
+++ b/fontbox/src/main/java/org/apache/fontbox/cff/CFFParser.java
@@ -205,6 +205,12 @@
             throw new IOException("Top DICT INDEX missing in CFF font");
         }
         
+        if (topDictIndex.length < nameIndex.length)
+        {
+            throw new IOException("Top DICT INDEX has " + topDictIndex.length
+                    + " entries for " + nameIndex.length + " fonts in the Name 
INDEX");
+        }
+
         stringIndex = readStringIndexData(input);
         byte[][] globalSubrIndex = readIndexData(input);
 
{code}

This issue was found by an LLM-assisted tool and verified manually.




--
This message was sent by Atlassian Jira
(v10.3.26#10030026)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to