[rekonq] Re: Review Request: Text highlighting in the url bar simplified

[Pierre Rossi]

> On 2010-11-19 22:58:49, Jon Ander Peñalba wrote:
> > I think that this solves the problem. I don't know if I've added all the symbols that cause trouble, but if not, it's really simple to update :)
> 
> Benjamin Poulain wrote:
>     In my opinion, it is just the wrong way to solve the problem. You still have not addressed the case where you match "XX" in a string of "XXXXXXXXXXXXX".
>     
>     In you last patch, you replace special symbols, but those can appear in an URL, you want to match them. If you want to escape all the control character in a meaningful way, your code will end up being bigger than the previous code.
>     
>     The objective of you patch is "Text highlighting in the url bar simplified", the new code is neither simplified, neither correct.
> 
> Jon Ander Peñalba wrote:
>     I think it's simpler and way more readable than our current solution.
>     I don't understand the problem of "XX" in "XXXXX". The result is "<b>XX</b><b>XX</b>X" and that's what's supposed to be.
>     
>     I know that hard-coding all the symbols and removing them from the string is normally a bad solution, but in this case I think we have more to gain, the previous code is quite hard to follow.

If I take a site like http://www.snowandrock.com/all/ski/fcp-category/home I might want to look it up in the history by typing "snow+rock" and this currently doesn't work with the regexp approach.
as for the </b><b> I don't really like it, if you type X then it'd result in <b>X</b><b>X</b><b>X</b><b>X</b><b>X</b>, even though it doesn't matter that much, that's a bit of overhead compared to <b>XXXXX</b>.
Weighing the pros and cons, I think the regexp-based solution is going to become as unreadable as the current implementation quite soon at this rate, and I fear it might never be as robust.

I agree that the current function used to do that can seem somewhat cryptic, and I take the blame for it, I should probably write a comment block there to explain what it does, since it's probably something that's bound to be changed and/or fine tuned in the future.

http://xkcd.com/208/


- Pierre


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://git.reviewboard.kde.org/r/100154/#review365
-----------------------------------------------------------


On 2010-11-19 22:56:28, Jon Ander Peñalba wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://git.reviewboard.kde.org/r/100154/
> -----------------------------------------------------------
> 
> (Updated 2010-11-19 22:56:28)
> 
> 
> Review request for rekonq.
> 
> 
> Summary
> -------
> 
> I've simplified text highlighting. I think the behaviour hasn't changed, but a second opinion is welcome :)
> 
> I've removed the Qt::escape (and updated the test accordingly) because I find it useless, but if it needs to be there for any reason there's no problem in putting it back.
> 
> 
> Diffs
> -----
> 
>   src/tests/listitem_test.cpp fc0b62e 
>   src/urlbar/listitem.cpp a0462e7 
> 
> Diff: http://git.reviewboard.kde.org/r/100154/diff
> 
> 
> Testing
> -------
> 
> The 'listitem_test' test passes.
> 
> 
> Thanks,
> 
> Jon Ander
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.kde.org/pipermail/rekonq/attachments/20101120/7c47d9cd/attachment-0001.htm