What regex to create a search limit boundary?

Aporosa · Oct 4, 2021, 1:05 AM

I need some kind of ‘stop search here / start again’ expression, as in the problem here:

The examples here are from downloaded metadata in html, from the naturalis.nl biodiversity database. Species names are split up by markup code, which I need to strip out.

Here’s a simple example:
Piper L.

So, stripping the markup out is a simple search and replace, resulting in the scientific genus Piper L.
Find: (.*)
Replace: $1
Result: Piper L.

But that strategy does not work on this more common binomial name example:
Desired result: Ficus septica Burm. fil.
Original html: Ficus septica Burm. fil.
Find: (.*)
Replace: $1
Result: Ficus septica Burm. fil.

I seem to need a way to stop the regex after the first couple, and start afresh for the next one. Instead, it is skipping to the last occurrence.

How can I improve my regex code?

Terry R · Oct 4, 2021, 1:18 AM

@Aporosa said in What regex to create a search limit boundary?:

How can I improve my regex code?

My small test shows this seems to work:
Find What<([^>]+)?>
Replace With: nothing in this field

Terry

PS this will also find any sequences of <> and since they make no sense being there this removes them as well.

Terry R · Oct 4, 2021, 1:30 AM

@Aporosa said in What regex to create a search limit boundary?:

How can I improve my regex code?

I looked more into your regex and the reason it didn’t work as planned was the (.*) is greedy. If your were to change your regex to (.*?), note the inclusion of a ? changes the regex from being greedy to lazy (not-greedy).

The difference is that your’s will find the longest sequence that the regex can be true with, the adjusted regex finds the shortest length the regex is true with.

Terry

Aporosa · Oct 4, 2021, 5:59 AM

@Terry-R said in What regex to create a search limit boundary?:

note the inclusion of a ? changes the regex from being greedy to lazy (not-greedy)

Thank you Terry R, not only was this a very clearly explained and effective solution, but now I understand the idea of a ‘greedy’ regex expression, something I didn’t get from textbook examples that I’d read.