Adding parameter to more information for 'context' option.

[monba-gea]

Hello, recently I discovered urlscan and was testing with my mutt setup. It's great in what it does but I see limitations in some of my own mail,

$ cat /tmp/urlscan_sample.txt 
This is a sample test [1] to show 
the folding in urlscan [2], could be
bit nicer to show [3] one more 
character [4] near the linked
urls [5], rather than just <URL> [6].


[1]: https://12345.pl
[2]: https://12345.nl
[3]: https://12345.com
[4]: https://12345.xyz
[5]: https://12345.de
[6]: https://12345.net

$ cat /tmp/urlscan_sample.txt  | urlscan
[1]: <URL> [1]
[2]: <URL> [2]
[3]: <URL> [3]
[4]: <URL> [4]
[5]: <URL> [5]
[6]: <URL> [6]


[1]   < https://12345.pl
[2]   < https://12345.nl
[3]   < https://12345.com
[4]   < https://12345.xyz
[5]   < https://12345.de
[6]   < https://12345.net

I glanced the man page but couldn't find any parameter to give more context, having atleast the previous character is really helpful for example,

[1]: sample test [1]
..
..
[5]:  linked urls [5], 

[firecat53]

Yes, what you are seeing is a known limitation of the original implementation that's documented in the README:

Extraction of context from HTML messages leaves something to be desired.
Probably the ideal solution would be to extract context on a word basis rather
than on a paragraph basis.

Another smaller change could be adding a configurable context window (change the number of context lines) but that wouldn't fix the issue you are describing.

I'll see how extensive the changes would need to be to switch to a word based versus paragraph based context.