[Bug]: Indirect object reference in LastChar

[szepeviktor]

Is there an existing issue for this?

• I have searched the existing issues and found no similar reports

Are you using the latest version of this package?

• The issue I'm reporting exists in the latest release

Can other PDF readers read the file?

• The PDF I'm trying to read opens correctly in at least one other PDF reader

When running this snippet

$document = (new PdfParser())->parseFile('/path/to/file.pdf');

I run into the following issue/exception (Please attach the pdf)

Value "91 0 R" for dictionary key LastChar could not be parsed to a valid value type

What it is: The error points to a Font dictionary entry. /LastChar must be an integer. Here it is 91 0 R, an indirect object reference, which the parser can’t turn into a valid integer.

Role: Together with /FirstChar, /LastChar defines the character-code range covered by /Widths for simple fonts (Type1/TrueType). Not used for CID/Type 0 fonts.

Where to find it: In a font object under a page’s /Resources → /Font.

Why it fails: The library expects a number, not a reference - or the referenced object isn't a number.

Do you allow attachment files to be used in tests to prevent regressions?

• Yes, I give permission to use this file as a test file to prevent future regressions (And am authorized to give this permission)

[szepeviktor]

InvalidArgumentException: "R" is not a valid width when font /Widths contains indirect entries

The parser throws when a font’s /Widths array contains indirect references (e.g. 92 0 R). PDF allows array elements to be indirect; these should be dereferenced before validating as numbers.

Repro (evidence from mutool)
/Widths contains an indirect entry:

mutool show sample.pdf 'trailer/Root/Pages/Kids/0/Resources/Font/f-3-0/Widths'
… 294 294 0 92 0 R 258 321 262 0 92 0 R 92 0 R …

Dereferencing yields a number:

mutool show sample.pdf 92
92 0 obj
571
endobj

Exception observed:

PrinsFrank\PdfParser\Exception\InvalidArgumentException: ""R"" is not a valid width

Expected
Indirect width entries are dereferenced and returned as numeric values.

Actual
The parser treats the trailing R as invalid input and throws.

Proposal
In the font width resolution path, accept IndirectReference entries inside /Widths (and, for CID fonts, /W), dereference first, then validate as int|float.

[PrinsFrank]

@szepeviktor Thanks for the bug report! Do you by any chance have a PDF sample that this occurs in so I can verify the fix for this?

[szepeviktor]

Sorry, all of them are CV-s uploaded by users.

[PrinsFrank]

@szepeviktor Is it correct that these are two bugs? One for the "lastChar" record, and one for the widths array? Can you share a sample for both of these privately? I need to see the underlying structure of those referenced objects to implement this properly. ;)