better swipe to delete word

[pxeemo]

this pattern doesn't separate words only by spaces for example links that were known as one word, now will be deleted like this:
https :// github. com/ => 4 words

[dessalines]

This is just for links specifically? I'm not sure why that would be preferred over simple word deletes.

Where did this complicated regex come from? Regex("^\\s?(\\w+\\W?|[^\\s\\w]+|\\s+)")

[pxeemo]

^ for starting the line (to just match the first after cursor)
$ for end of the line (to match the last one before cursor)
\\s matches whitespaces (spaces, tabs, break lines, ...)
? means one or none
\\w is for characters used in words (A to Z, a to z, 0 to 9 and particularly _, for all languages)
\\W is for non-word characters (.,!?-*;:=+({[]}), ... and whitespaces)
+ means one or more
* means none or many
[^\\s\\w]: not whitespaces (\\s) and not word characters (\\w)
(one|two): matches "one" or "two"

so simply it's not used only for links it matches:

• a whitespace (if exist) and one word and a non-word character after that (if exist) (matches end of the line signs)
• non-word and non-whitespace characters together (..., == like in codes or link :// or any other)
• all the whitespaces together

and delete last word regex matches:

• one word and a non-word character after that (if exist) and any whitespaces after that (if exist)
• non-word and non-whitespace characters together and any whitespaces after that (if exits)
• delete all whitespaces if there isn't any words

then it can split things like math (123+635=??) and codes in any language print("hello_world") and any others those not separated with spaces (like thumb-key) (it also has the same functionality as before)

[dessalines]

Did you come up with this regex on your own? Is there an SO link or something for this desired pattern?

[pxeemo]

I actually wrote it myself but it can be tested in sites like regex101.com it also has explaination in right panel

I was thinking it was better to remove \W? or change it to [^\s\w]? to not match space after word in this regex. I also tested it on a debug build and worked fine.

[dessalines]

K I tested and it seemed to work well. In the future add some code comments as to what this matches for.