shreevatsa / pdf-glyph-mapping Goto Github PK

I am getting segfault at this line while dumping the Tjs. The page number on which this happens (sometimes on 1000+, sometimes on 10000+) is not deterministic, but out of ten times I have tried it, it has happened every single time.

I get (with mupdf-tools 1.18.0-2, Arch Linux) .jpg output by this command instead of .png. So we need to explicitly specify the format in the command (if it is possible) for doit.sh to work.

I had abandoned this project in favour of https://github.com/shreevatsa/pdf-explorer but that itself seems abandoned so it may be good to resurrect one or both of them.

At minimum, it seems the Cargo.toml here needs to be changed to:

clap = "=3.0.0-beta.2"
clap_derive = "=3.0.0-beta.2"

per here. On top of that, may be good to clarify the API a bit, e.g. remove dependency on mupdf and have clear boundaries / separate binaries: one for just dumping the text content stream, etc?

On running without making any changes to the code (other than .png being replaced with .jpg in doit.sh), no Tjs-*.map files were generated even after dump-tjs ran successfully.

Bold (can be got from font name) and Italic.

To save compute time spent in extracting operators, etc. from PDF, save the results in a convenient format (perhaps text, or use 'serde') for text-only analysis for quick improvements and testing of stuff like glyph maps and regex.