1. 32

This release includes some polish and features that were requested by some users.

Notably:

  • The config file location now follows the cleaner XDG directory spec.
  • You can open a file while you are in the middle of renaming it. This makes it easier to remind yourself what the document was in the first place. Just press Ctrl-o.
  • You can optionally specify multiple folders to import pdfs from.
  • You can rename a file that is already imported, just press ‘r’ when in the Library list.
  • There is a screen on first startup, showing you where the config file will be initialized.
  • There is now a help screen (press h).
  • Everything is slightly prettier.
  1. 6

    For large files, pdftotext can take quite a long time to parse the entire document, which is stupid because we’re only using the first couple of lines for file name suggestions.

    Have you considered using some version of MuPDF? By using mutool I can run

    mutool draw -F txt some.pdf 1
    

    and I only get the first page, in text form. I’m not sure if it internally only parses as far as it has to, but from my experience MuPDF is a bit faster than Poppler, but lacks features, which as far as I see, your project doesn’t need.

    Also, I have to say that it’s a great idea to develop this. Running find . -type f | wc -l in my unsorted documents directory gives me 2446 files, with the number rising all the time ^^. I hope you’ll get this published in the standard repositories of some larger distros.

    1. 2

      Thanks so much for the mupdf pointer, will try it out!

      re: publishing in standard repos: Do you have a good link that describes how to bundle an executable up in a .deb, .rpm and get it into Arch without too much manual hassle? Would love for it to just package everything right in CI.

      1. 2

        For Arch, you can consider publishing it on the AUR. It may be moved into the repositories if a Trusted User or a developer adopts it.

        EDIT: I made one at https://aur.archlinux.org/packages/paperboy and also created two Haskell packages along the way, pdfinfo and titlecase. I can add you as a co-maintainer if you wish.

        1. 1

          Thanks so much for the mupdf pointer, will try it out!

          I’ve been helping in developing MuPDF for some time. If you have any questions about it or stumble upon an issue, feel free to ask on #mupdf on freenode, or report bugs at our bug tracker.

          1. 1

            Super cool, thanks! I managed to solve that specific problem by being told that pdftotext accepts “from” and “to” flags for which pages to convert (facepalm for not noticing myself).

            But I am very happy to have discovered that the tools part of mupdf exists, I can imagine doing interesting stuff with it in the future.

      2. 3

        Readme mentions David_Lynch_Teaches_Typing.zip. Turns out that actually exists. https://youtu.be/xxrCEV4P0OY

        1. 2

          :)

          a) That thing was too hilarious not to mention and b) I literally wrote the readme by looking at my download folder in annoyance at all the random crap that accumulated in it over the years.

        2. 3

          Installed, looks pretty neat. Be interesting if there’s a way to grab dates so renamed titles could be compositions (e.g. suggested date + suggested title + original filename + ‘.pdf’)

          Stories with similar links:

          1. Paperboy - a utility to rename a messy collection of pdfs authored by instant 4 years ago | 38 points | 10 comments