Use / to indicate the start and end of a search
Expressions that match patterns
- can have multiple matches
Match on the entire line
\d+ find a digit of any length
. match any character
\. match special character (in this case .)
^ = starts with
$ = ends with
r = raw string
\d = digit
\w = alphanumeric
\s = space, tab, new line
Capitalize to inverse:
\D = not a digit
\W = not alphanumeric
Match one of several characters
gr[ae]y -> gray, grey
Can match a range
[a-e]
[0-9]
Can negate
[^0-5]
Can repeat to match multiple using
?(zero or one)*(zero or more)+(one or more)
| for multiple patterns
( ) for groups
{ } for a range
For all vowels [aeioiAEIOU]
For all non-vowels [^aeioiAEIOU]
Any alphanumeric [0-9a-zA-Z] (same as \w)
Two vowels [aeiou][aeiou]
Need to escape heaps of the characters [()] = matches ( or )
/hij/ will find hij (case sensitive)
Put on end
/hij/g global (return all matches)
/hij/i case insensitive
/hij/m multiline
Specifying quantity
pattern-quantity
\d{3}-\d{3}-\d{4}
* 0 or more
+ 1 or more
? 0 or 1
{2} exactly 2
{2, 4} 2 to 4
\d+ = one or more digits \d{3,5} = between 3 and 5 \d{3,} = 3 or more
[aeiou]+ = one or more vowels
If we want constant-vowel, constanant-vowel n times
Use () to group
To match sayonara
([^aeiou][aeiou])+
To extract the contents of the HTML tag in Kindle exports
cat personal/notes/books/Hamming_art_science_engineering.html | sed "s/.*<div class='noteText'>//" | sed "s/<\/h3>//" > personal/notes/books/Hamming_art_science_engineering.md
Al Sweigart Yes, It's Time to Learn Regular Expressions PyCon 2017 (youtube)
Read This If You Don’t Know Enough About Regex