ref: 464b83a87fb9d1841de5f619db8a515161bbbbce
dir: /README.md/
# PDF as a file system
Yeah. WIP.
There is nothing to look at yet. You can build it and run things like
`./6.out my.pdf Root Pages Kids 0 Kids 0 Kids 0 Contents .` to dump a
specific object (text, images, fonts, etc).
## TODO
* ASCIIHexDecode filter
* CryptDecode filter
* does DCTDecode filter need to deal with ColorTransform?
* do we need predictor 5 ("optimum")?
* object stream dictionary doesn't support `Extends`