PDF Object Inspector (Developer Tool)

About the PDF Object Inspector (Developer Tool)

Open a PDF and browse its internals as defined in ISO 32000-1 §7: the header and version, the trailer, every cross-reference section (tables, cross-reference streams and hybrid files), each incremental update with the objects it changed, the document catalog, the page tree and every object. Dictionaries are pretty-printed with clickable references, so you can follow /Parent, /Resources or /Contents from object to object.

Streams are decoded (Flate, LZW with PNG and TIFF predictors, ASCIIHex, ASCII85, RunLength), shown as text or hex, and content streams are split into operators with their meaning — re Rectangle, Tf Set text font and size. Raw and decoded streams can be downloaded, JPEG images previewed, and scripts, actions and attachments are listed so you can spot active content. The file is read in your browser and never uploaded.

Common uses

• Debugging a PDF generator: checking the objects, fonts and content streams it writes.
• Finding out why a PDF renders differently in two readers, or why a form field misbehaves.
• Seeing what changed in each incremental update — including after a digital signature.
• Checking a PDF from an unknown sender for JavaScript, launch actions or attachments before opening it in a full reader.
• Learning the PDF format with real files next to the specification.

How a PDF is put together

A PDF file is a header (%PDF-1.7), a body of numbered objects (12 0 obj … endobj), a cross-reference section that lists the byte offset of every object, and a trailer that names the document catalog (/Root). A reader starts at the end of the file: startxref gives the offset of the last cross-reference section, whose trailer may point (/Prev) to an earlier one.

Since PDF 1.5 the cross-reference can itself be a compressed stream, and small objects can be packed into object streams — this tool shows such objects as “in object stream 12”. A hybrid file has both a classic table and a cross-reference stream (/XRefStm) so that old readers still open it.

Incremental updates and signatures

Programs that save “in place” append the changed objects and a new cross-reference section to the end of the file instead of rewriting it. Each such update is a revision here, with the objects it added or changed and those it deleted. The older revisions stay inside the file and can still be read.

A digital signature covers the file from the first byte to the end of the revision it was added in (its /ByteRange). When a later revision exists, the document was changed after signing — that may be fine (for example filling in a later signature field) or not. The overview lists every signature field and which revision it covers.

Active content

The overview lists objects with JavaScript, launch actions, an open action, additional actions, embedded files, submit-form actions, links to other files, web links, XFA forms and rich media. Finding them doesn’t mean a file is harmful — forms often use scripts — but a document you didn’t expect to be interactive shouldn’t have any.

Limitations

• Encrypted PDFs show their structure, but strings and streams are encrypted and objects inside object streams can’t be read. Save an unprotected copy with Unlock PDF and inspect that.
• Image formats inside PDFs (JPEG 2000, JBIG2, CCITT fax) are not decoded; download them as stored. JPEG images and 8-bit grey, RGB and CMYK pixels are previewed.
• The operator view lists the first 2,000 operators of a stream (the first 50,000 are analysed); download the decoded stream for all of it.
• Damaged files are repaired the way readers do it — by scanning for objects — and the overview says so.
• This is a viewer: nothing in the file is changed.

Privacy

Everything happens in your browser. What you enter or open here is not uploaded or stored by MySmartCoPilot.