Remove all text before or after a certain marker

[plutocrat]

I've been whittling away at a page and have substantially reduced the amount of text in the diff by using CSS selectors. However the information I'm looking for resides in a huge table and I think I've reached the limit of what I can do with CSS / Xpath.

Is there a way to ignore/remove all text after a certain marker. Or, in that case, all text on a page before a certain marker. I can do it in sed, awk etc, but can't figure out if there's an easy way in changedetect.

If that's not currently possible with the given tools, would that be interesting as a feature request?

[dgtlmoon]

Yeah this is a nice idea, to remove any text before an element, or after an element.

I thought about this functionality when I started the project, but actually I forgot :D or didnt have time yet, open to a PR, you could scrape the xpath extraction data JSON which actually already contains most of the information (offset from top, size etc)

[dgtlmoon]

Why arent you just using VisualSelector and selecting ONLY the element you want to follow? can you give some example URLs and more information about the problem you're trying to solve?

[plutocrat]

The use case is if I want to select a block of text that isn't defined by xpaths, or css. So take this page for example.
http://www.aastocks.com/en/mpf/compare.aspx?comp1=600269
I can select the fund price with the visual selector, or any single cell. Or I can select the entire table. But what I'm actually interested in is all the text between
Fund Performance.
blah
blah
blah
Fund Characteristics.
So if I could automatically discard any text before a certain point, that would make life a lot easier.

[dgtlmoon]

Just select the table, then under "match text" put in the following lines

1-month
3-month
6-month

etc

does that help?

[sirkelsalot]

I was trying to accomplish something similar on craigslist.com as well. For example:

https://washingtondc.craigslist.org/search/sss?postal=22042&query=kickr&search_distance=10#search=1~gallery~0~0

In this case, I want to ignore everything after the string "more from nearby areas (sorted by distance)" as shown below.

In my use case, these results can be hours away from where I live and it seems there is no way to turn it off within the craigslist website, it auto shows up if there are not enough local results. I only want to monitor for changes in the top section.

They are not separate "sections" that I can select only the top half with the visual file selector.

[Constantin1489]

@siopao try this xpath://ol/li[following-sibling::li[@class="nearby-separator"]] with xpath3.1 since CD 0.45.8.(I didn't try the previous version. I mentioned the version because I'm assure)

[Constantin1489]

@plutocrat you can use xpath3.1 since CD 0.45.8.
I hope this inspires you.