[ 
https://issues.apache.org/jira/browse/PDFBOX-6314?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18147701#comment-18147701
 ] 

ASF subversion and git services commented on PDFBOX-6314:
---------------------------------------------------------

Commit 1940504 from Tilman Hausherr in branch 'pdfbox/branches/2.0'
[ https://svn.apache.org/r1940504 ]

PDFBOX-6314: avoid cast for values above 0xFFFF, as suggested by Martin Mirchev

> SaslPrep rejects valid supplementary characters (e.g. U+20000) in AES-256 
> passwords
> -----------------------------------------------------------------------------------
>
>                 Key: PDFBOX-6314
>                 URL: https://issues.apache.org/jira/browse/PDFBOX-6314
>             Project: PDFBox
>          Issue Type: Bug
>          Components: PDModel
>    Affects Versions: 3.0.8 PDFBox, 4.0.0
>            Reporter: Martin Mirchev
>            Priority: Major
>
> {{SaslPrep.prohibited(int codepoint)}} casts the code point to {{char}} 
> before two of its checks:
> {code:java}
> return nonAsciiSpace((char)codepoint)
>         || asciiControl((char)codepoint)
>         || ...
> {code}
> For a code point above U+FFFF the cast keeps only the low 16 bits, so U+20000 
> (CJK Unified Ideographs Extension B) becomes U+0000 and is rejected as an 
> ASCII control character. Any supplementary character whose low 16 bits fall 
> in 0x00-0x1F, 0x7F or one of the non-ASCII space code points is rejected the 
> same way, although none of them is prohibited by RFC 4013.
> Through the public API (trunk bd0c9c0):
> {code:java}
> try (PDDocument doc = new PDDocument()) {
>     doc.addPage(new PDPage());
>     StandardProtectionPolicy pol = new 
> StandardProtectionPolicy("own\uD840\uDC00er", "", new AccessPermission());
>     pol.setEncryptionKeyLength(256);
>     doc.protect(pol);
>     doc.save(out);
> }
> // java.lang.IllegalArgumentException: Prohibited character 'CJK UNIFIED 
> IDEOGRAPHS EXTENSION B 20000' at position 3
> {code}
> {{pass\u00e9}} is accepted, and a real control character such as U+0001 is 
> still rejected. The same code is on the 3.0 and 2.0 branches.
> A possible fix is to apply the two {{char}} checks only to BMP code points:
> {code:java}
> return (codepoint <= Character.MAX_VALUE
>             && (nonAsciiSpace((char) codepoint) || asciiControl((char) 
> codepoint)))
>         || nonAsciiControl(codepoint)
>         || ...
> {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