How to delete specific lines only with macro

Dominic Leblanc

Hi,

I have a text file that lists georeferenced line in UTM format and should go like this: line number, coordinate of the first point and coordinate of the last point.

Example:
L106900
359088.16 4927710.64
372696.93 4924424.02

But, when the file was made, it had generated a bunch of coordinates in between the first and last coordinates that I would like to get rid of. I started doing it manually but it’s thousands of lines I need to remove.

Any idea how to automate that?

Thanks,
D

guy038

Hello, @dominic-leblanc and All,

Well, here is a possible solution , using a regular expression

Given this example of INPUT text :

L106900
359088.16 4927710.64
368425.29 4925689.13
359928.56 4926782.72
372696.93 4924424.02


L106901
359261.78 4932588.32
375624.69 4926854.61
359923.84 4926958.17
378952.70 4937532.29
396284.33 4925818.24
382621.65 4920105.59

And the following regex S/R :

SEARCH/MARK (?-is)^L\d+\R.+\R\K(?s:.*?)(?=.+(?:\R+L\d+|\R*\z))

REPLACE Leave EMPTY

You would be left, after execution, with this OUTPUT text :

L106900
359088.16 4927710.64
372696.93 4924424.02


L106901
359261.78 4932588.32
382621.65 4920105.59

As you can see, this regex :

Keep the first line after a line beginning the uppercase letter L
Keep the last line before a line beginning the uppercase letter L or before the end of text / file
Delete any coordinates line located in between

IMPORTANT :

You must select the Regular expression search mode
You may check the Wrap around option, if you need to modify the whole text of current file
You can match the different occurrences, using the Find dialog and the Find Next button
You can see the bunch of all the lines to delete , using the Mark dialog and the Mark All button
When using the Replace dialog, you must use, exclusively the Replace All button ( NOT the Replace button ) because of the \K syntax part of the regex search expression !
Each line beginning with an L letter and digits may be preceded or not with some empty lines
The last bunch or coordinates lines may be followed or not with some empty lines

Best Regards,

guy038

P.S. :

If, by mistake, you re-run this regex search/replacement, you may be surprised of the message Replace All: n occurrences were replaced .... That is quite normal, because it replaces, this time, the zero-length bunch of in-between lines with nothing again !

Dominic Leblanc

@guy038

Wow! I’m impressed thank you! It didn’t work for some lines for which the number has a decimal though like L102300.1

guy038

Hi, @dominic-leblanc and All,

Ah…, OK ! As this fact wasn’t mentionned in your initial example, this case was not supposed to happen :-)

So, my new version of the search regex is simply :

SEARCH/MARK (?-is)^L[\d.]+\R.+\R\K(?s:.*?)(?=.+(?:\R+L[\d.]+|\R*\z))

Note that this version would also match if the initial georeferenced line was, for instance :

L.12.34.56.78.9

I suppose that it’s not a problem if your text does not contain such lines ;-))

BR

guy038

Dominic Leblanc

Thanks Guy, I really appreciate your help!