Read Aloud --suppress line numbers ?

The Read Aloud is boon for us terminally dyslexic fans. But we may also like to read those ms with nerdy line numbers. It turns out that the speaking of line numbers is more often a distraction when one is not in reviewer/editor mode. This fine zotero pdf reader has me using it to read prepublication ms's that I usually don't add to my library. If this suppress of speaking line numbers is already a feature I missed it, my apologies. If it's in the plan great. Otherwise this is a humble feature request.
  • I agree that this is an amazing addition to Zotero. To add to the comment on line numbers, I think more generally, all repeating texts in PDFs should be excluded – or at least it would be great to have the option.

    I am currently listening to a rather long technical document, which is already tedious. This is made worse by listening on every page to the header, footer, and a download link (as in a URL) that spans the left flank of the document.

    To the humble feature request to suppress line numbers, I therefore add headers, footers, sidebars and maybe URLs more generally.
  • For anyone else reviewing line-numbered manuscripts: I put together a small plugin that removes margin line numbers from the text Read Aloud speaks.

    https://github.com/zfergus/zotero-better-read-aloud

    It finds line numbers by looking for long, dense runs of consecutive numbers (dozens per page), so footnote markers and numbered reference lists are left alone. It also fixes two side effects of line numbers: lines misread as numbered list items, which make Read Aloud pause after every line, and words hyphenated across a line break being read in two halves. It only edits the copy of the text Read Aloud receives; the PDF and Zotero’s cache are untouched. Install the .xpi from the Releases page via Tools → Plugins → Install Plugin From File… (Zotero 10+).

    I’ve tested it on a handful of EUROGRAPHICS review copies. If it misbehaves on another template, please open an issue on GitHub.

    To the Zotero devs: the cleaner fix is probably in pdf-worker, dropping line-number glyphs before block segmentation, next to the existing header/footer detection in page-furniture.js. I’m happy to turn this into a PR if that’s useful.
Sign In or Register to comment.