CustomMatchers should override default matchers

[Thomas-Bel]

Hi,

For our app we would like to separately process certain url patterns, while allowing all other urls to be handled normally. I tried to implement this using both url=true and a custom matcher for the special case urls, but this approach fails. The default matching is performed first, and only afterwards the custom matchers are exercised.

Is this desired behaviour?

[joshswan]

That's the current expected behavior, though I do agree that it may not be optimal. As I just mentioned in #59, it might be better for you to just disable the default matcher and implement a custom URL matcher that you can control better (for example using the regex I posted there).

Working on solutions to solve both this and #59 in a new version - leaning towards removing Autolinker.js and implementing everything as matchers that can be more easily controlled/customized.

[Thomas-Bel]

Thanks for the quick response. For now I do appreciate the option to identify urls without the www or http prefix, so I have gotten it to work with patch-package by switching around the order of the two replacement operations in the Autolink file.

[khunshan85]

I am also having this requirement where I want only those urls to be matched that contain my specified domains. For example, if the text contains .google.com/ only then it should be considered a url, otherwise every other url should be considered as a text. How can I achieve this? I tried overriding the default matcher, but didn't work

[joshswan]

@khunshan85 Disable the default URL linking prop and implement a custom matcher with your own regex that targets Google URLs.

Here's a generic URL regex from the other issue to get you started: /(https?:\/\/|www\.)[-a-zA-Z0-9@:%._\+~#=]{1,256}\.(xn--)?[a-z0-9-]{2,20}\b([-a-zA-Z0-9@:%_\+\[\],.~#?&\/=]*[-a-zA-Z0-9@:%_\+\]~#?&\/=])*/i

[khunshan85]

Thanks for the quick response @joshswan . Let me try this