[question | feature] Diffs in content order, not filter order

[jamesdeluk]

Say there is a website with titles and text:

<p class='title'>1</p>
<p class='text'>a</p>
<p class='text'>b</p>
<p class='text'>c</p>

<p class='title'>2</p>
<p class='text'>d</p>
<p class='text'>e</p>
<p class='text'>f</p>

<...

I have two filters:

.title
.list

This will show in the diff as:

1
2
a
b
c
d
e
f

That is, all .title then all .list

What I'd like is to see it in the order of the page, i.e.:

1
a
b
c
2
d
e
f

Is this possible?

If not, feature request :)

[dgtlmoon]

try .title, .text as the filter?

[jamesdeluk]

Doesn't seem to work.

Specific example:

https://openrouter.ai/rankings

.text-lg, .text-foreground

The ideal output is:

Leaderboard
Grok Code Fast 1
Gemini 2.5 Flash
Claude Sonnet 4.5
gpt-oss-120b
Claude Opus 4.5
Gemini 2.0 Flash
Gemini 2.5 Flash Lite
Grok 4 Fast
Gemini 3 Pro Preview
Gemini 2.5 Pro

Market Share
google
...

[dgtlmoon]

@jamesdeluk i dont want to sound horrible, but when you report an issue - in the issue template it specifically says to include a link to the URL (in capitals!) and real data, it saves people a lot of time if you can do this, i can see that the data you put in your initial report is not in the same structure as the data in the realworld, so its rather difficult to give you an accurate report when the report itself is missing data

[dgtlmoon]

.text-foreground is used all over that page... not just for those headings

[jamesdeluk]

Yeah, I get that - it was originally a question/feature, not a bug report, hence not including "bug report" issues. But you make a valid point.

[jamesdeluk]

.text-lg is the header, .text-foreground is the content (and some other stuff, which I'm happy to deal with). I also have .text-xs as removed.

[dgtlmoon]

Can you clip some screenshot and show exactly what you mean? otherwise i'm basically trying to reverse engineer what you wrote in my brain and then extrapolate it out to a page with completely different structure

[jamesdeluk]

Website:

Config:

Output:

(ignore some of the "false positives" for now - the main point is that all headers/titles are at the top, whereas on the website they're throughout the content)