How can I highlight more than one occurrence of a match?

[icarojerry]

Supposing I have a text like:

• "Google Chrome is a cross-platform web browser developed by Google."

And use #:~:text=google, only the first occurrence is highlighted, while I would like to highlight all occurrences.

All the text is in the same context and fragment.

Is it possible to do that?

[bokand]

There's no way to specify "highlight all" but you can specify multiple instances. In your example:

#:~:text=Google-,Chrome&text=browser%20developed%20by-,Google

Would do the trick.

[icarojerry]

I know that, I am using this way for some cases.
But for example, if you access this link:
https://en.wikipedia.org/wiki/Google_Chrome#:~:text=google

Only the first word "Google" is selected, however if you search CTRL+F for the word "Google", all occurrences will be highlighted.

I would like to simulate this effect. Is there any intention in the project for this to be implemented?
Could have a prefix to indicate highlight all occurrences, like "text=~Google" or "text=*Google*"

Thanks for the answer,

[bokand]

There's no current intention to do that but it's an interesting feature request and could be considered in the future.

Do you have a use case that could be used to motivate it?

[icarojerry]

Yes I have. I work on a research site with long texts. When a user opens one of the results, I send the matches at the URL to keep the highlights in the text, without having to process the search again.

But, look these cases:

1. Using #:~:text=azimute
   

2. Using Ctrl + F -> azimute
   

[ratibus]

Hi

Indeed multiple matches support could be used to highlight some data in tables when we know a pattern but not the specific data, when we create the link.
For example, imagine a table with daily data and a column "date", formatted like yyyy-mm-dd. If I want to highlight the rows of every day of June, I could send the following fragment : #:~:text=-06-

Text being a regexp could also offer many more usage :)

[DavidGP]

I found a website where a syntax like [URL]?s[]=MySearchTerm&s[]=MyOtherSearchTerm works:

https://www.ipwiki.de/arbeitnehmererfinderrecht:beschraenkte_inanspruchnahme?s[]=inanspruchnahme&s[]=Arbeitgeber&s[]=Diensterfindung

However, this syntax doesn't seem to work anywhere else, so it's probably not a feature of Chrome but a function of that website.

[tomayac]

However, this syntax doesn't seem to work anywhere else, so it's probably not a feature of Chrome but a function of that website.

It's a website feature of the specific site in question, but you can emulate it using multiple Text Fragments. This is a manual process, though, where you need to hand-craft a link.

[RokeJulianLockhart]

#107 (comment)

@icarojerry, potentially, an array system would be the most straight forward method – #:~:text=hello would print all instances, whereas #:~:text=hello[0,2:7] would print position 0, +2, and +3 to +7.

[onurusluca]

Would love to have this. Any updates?

[ideosyncretic]

+1 for this

[purpleslurple]

So 22 year old Quip might work, in a pinch: https://purpleslurple.com/quip.php?url=https%3A%2F%2Fen.wikipedia.org%2Fwiki%2FGoogle_Chrome&text=Google&annotation=Look%20at%20that!%20All%20occurrences%20highlighted.%20%3A)

Another (clearer?) example: https://purpleslurple.com/quip.php?text=Granular%20Addressability%20in%20HTML%20Documents%20-%20ON%20THE%20FLY&url=https%3A%2F%2Fdougengelbart.org%2Fcolloquium%2Fforum%2Fba-unrev-talk%2F0209%2Fmsg00015.html

You could grab: https://github.com/purpleslurple/PurpleSlurple/blob/master/quip.php
and https://github.com/purpleslurple/PurpleSlurple/blob/master/quip-bookmarklet.js and run it from your server (off localhost is handy too), if you don't want it served off PurpleSlurple server.

When I last touched that code you could make multiple selections by simply wash repeat cycle on a given page. For some reason, that now breaks. (Maybe Github sessions have changed? My code hasn't. :)

PurpleSlurple is pretty cool too: PurpleSlurple.com