All text from camelCase to SNAKE_CASE

regex notepad++

12,857

I suggest the following regex approach:

Find What: (\b[a-z]+|\G(?!^))((?:[A-Z]|\d+)[a-z]*)
Replace With: \U\1_\2
Match Case: ON.

This will turn camelCase87LikeThis words to CAMEL_CASE_87_LIKE_THIS. If you need to add support for those camel words that start with an uppercase letter, use the following regex modification:

(\G(?!^)|\b[a-zA-Z][a-z]*)([A-Z][a-z]*|\d+)

See the regex demo (also tested in Notepad++). Note the placement of the \G inside the regex and added A-Z.

Details:

(\b[a-z]+|\G(?!^)) - Group 1 capturing either of the two alternatives:
- \b[a-z]+ - start of a word (\b is the initial word boundary here) followed with 1+ lowercase ASCII letters
- |- or
- \G(?!^) - the end position of the previous successful match
((?:[A-Z]|\d+)[a-z]*) - Group 2 capturing:
- (?:[A-Z]|\d+) - either an uppercase ASCII letter ([A-Z]) or (|) 1+ digits (\d+)
- [a-z]* - 0+ lowercase ASCII letters.

The \U\1_\2 replacement pattern turns all the chars to uppercase with \U and inserts a _ between the two groups (inserted with \1 and \2 backreferences).

12,857

Dancia

Updated on September 16, 2022

Comments

Dancia over 1 year
I am trying to do some text manipulations using Notepad++ macros. My last step is converting camelCase strings to SNAKE_CASE. So far no luck. I'm not very familiar with regex so can't write my own solution.

Example text file input:
```
firstLine(874),
secondLine(15),
thirdLineOfText87(0x0001);
```
Desired output:
```
FIRST_LINE(874),
SECOND_LINE(15),
THIRD_LINE_OF_TEXT_87(0x0001);
```
Regex or any plugin is an acceptable answer.
- Yorga Babuscan about 4 years
  
  FYI, according to a post at the Notepad++ forum, there are conventions about the case types and its names. So, the correct names for the cases asked are lowerCamelCase and SCREAMING_SNAKE_CASE.
Dancia almost 7 years

How should find what look like if camelCases started with upper letter, example: FirstLine(874)?
Dancia almost 7 years

First approach misses all _ except last one, and second approach generates from ThirdLineOfText87(0x0001); -> THIRD_LINEOF_TEXT_87(0x0001); missed second _
Wiktor Stribiżew almost 7 years

You may use (\G(?!^)|\b[a-zA-Z][a-z]*)([A-Z][a-z]*|\d+), the trick was to put \G branch as the first alternative in the alternation group and adding support for the uppercase letter at the beginning of the word.
41686d6564 stands w. Palestine over 5 years

This replaces FfdffDF with FFDFF_D_F which, I believe, isn't the desired behavior. Here's a tweaked version that takes into account cases like SomethingIO and IOSomething: (\G(?!^)|\b(?:[A-Z]{2}|[a-zA-Z][a-z]*))(?=[a-zA-Z]{2,}|\d)([‌A-Z](?:[A-Z]|[a-z]*)‌|\d+).
ChaseMoskal over 4 years

hmm.. MySuperCoolTest.js becomes MY_SUPERCoolTest.js -- i can't get the regex to affect the entire string?
Wiktor Stribiżew over 4 years

@ChaseMoskal The regex in the answer is for strings like caMelCase, not CaMelCase. If you use the regex from my comment you will get MY_SUPER_COOL_TEST.js (just make sure Match case is on)
Wiktor Stribiżew about 3 years

@MoacirSchmidt Then you probably want (\G(?!^)|\b[[:alpha:]][[:lower:]]*)([[:upper:]][[:lower:]]*|‌\d+)|\b([[:upper:]][‌[:lower:]]*)\b and replace with \U(?{3}$3:$1_$2)
Wiktor Stribiżew about 3 years

@MoacirSchmidt Not the one in the previous comment.
Moacir Schmidt about 3 years

On the link you referenced Pascal is replaced by a single _ (underscore)
Wiktor Stribiżew about 3 years

@MoacirSchmidt Because regex101.com is not Notepad++. \U(?{3}$3:$1_$2) replacement will work in Notepad++.