Manage your scientific article library using PDF metadata, Unix-style.

Most PDF articles found online have useless metadata. To manage a library you will have to rely on

• the filesystem, which is limited
• or an external database = additional state to maintain + can lock you into an "ecosystem" :‑###..

However, PDF metadata is stored alongside the content and can contain a list of authors and a journal for easier searching.

• interoperability: files stay where they were and their metadata is open to other programs (pro-tip: look into fzf),
• freedom: no proprietary databases with special tools enforcing their workflow,
• hackability: the scripts are short, POSIX-compliant, pipe-able and accept wildcards, where applicable.

git clone https://github.com/ivan-boikov/fam && cd fam && make install

By default it will install symlinks in $HOME/.local/bin. Change to whatever inside the Makefile, but be sure it's in $PATH. Run make uninstall to uninstall.

1. Interactively find a correct DOI of a PDF with doi-repair. No automatic internet searches and fancy algorithms -- you are way more reliable in finding a correct DOI anyway. Then, data from CrossRef is used to populate standard PDF metadata with a list of authors, a title, and an ISO-690-like reference with a DOI.
2. Enjoy the profits: use doi-rename to unify your filenames or chain doi-infer with doi-to-bibtex to generate a BibTeX file

doi-infer *.pdf | doi-to-bibtex > literature.bib

(don't ask too much too quickly, don't anger CrossRef) or be extra fancy and search recursively with a bit cleaner output

find <library path> -name '*.pdf' | doi-infer | doi-to-bibtex | sed 's/\}, /\},\n\t/g' > literature.bib

or whatever else you have, the data is open, so use it!

The things you already have most likely:

• pdftk, an excellent PDF editing utility, highly useful on its own
• pdfinfo from the poppler package (a dependency of pdftk)
• a PDF viewer callable by xdg-open
• standard Linux utilities: grep, curl and others

No guarantees, things might crash and burn, so be careful. Works on my machine™.

• embedding keywords
• optional removal of accented characters
• support conference proceedings