PDF feature

[AtomicNess123]

Does ripgrep search within PDFs?

[okdana]

You can use it to do a 'dumb' search of PDFs (the same way you'd search any other binary file) with the -a option. If you want to search in a way that semi-respects the actual PDF format, you can use the --pre option with something like pdftotext. There's an example of that in the user guide

[okdana]

If you mean you want it to find all of those words together (apple AND cake AND cream AND oven), no, it can't really do that right now, though in simpler cases you could use a pattern like apple.*cake|cake.*apple.

(Also, ripgrep is designed mainly to search lines, not entire documents. You can use the -U option to search across multiple lines, but if you're trying to do like full-text document searches it's not really the best tool. Maybe that'll change if/when #1497 is implemented, but idk.)

If you mean you want it to find any of those words (apple OR cake OR cream OR oven), it's just like grep:

rg 'apple|cake|cream|oven'
or
rg -e apple -e cake -e cream -e oven

[AtomicNess123]

Thanks, I understand. However, I can't understand the grammar

apple.*cake|cake.*apple.

Is this searching documents that contain the two words? What if I want to search for three or more words?

The example for any of those words is clear, thanks.

[AtomicNess123]

My last question pertains to the "rg --pre ./preprocess 'The Commentz-Walter algorithm' 1995-watson.pdf" syntax. Would it be possible to search in more than one PDF, e.g., in a folder of PDFs? This is not clear. Thanks!

[okdana]

Is this searching documents that contain the two words? What if I want to search for three or more words?

It matches either apple followed by cake or cake followed by apple (on the same line, normally). If you want to match more words you have to expand the pattern to include every possible combination, or you can try to use look-around with PCRE, but it obv becomes irritating/infeasible very quickly.

You can read about regular-expression syntax in general at e.g. https://www.regular-expressions.info/quickstart.html, and about the specific syntax that ripgrep uses at https://docs.rs/regex/*/regex/#syntax

Would it be possible to search in more than one PDF, e.g., in a folder of PDFs?

Just give it the path to a folder instead of a file.

You'll probably want to at least skim through the manual (rg --help) and the first few links here, especially the guide i linked earlier — they have answers to a lot of basic usage questions like that

[AtomicNess123]

Perfect. Thanks for all your help and time.

[AtomicNess123]

Sorry to re-open: I just discovered https://github.com/phiresky/ripgrep-all. This function would do what rg does without having the need to set up a preprocess file. With the same functionality?Could you confirm this?

[BurntSushi]

I haven't used it. Why not try it and find out?

[AtomicNess123]

Yes, it does!