Bash uses the Extended Regular Expression (ERE) dialect. Regular Expression patterns that use capturing groups (parentheses) will have their captured strings assigned to the BASH_REMATCH array variable for later retrieval.
| Symbol | Meaning |
|---|---|
. |
Any character except newline |
^ |
Start of line. When within brackets, negates the match |
$ |
End of line |
? |
Preceding item will be matched zero or one times |
* |
Preceding item will be matched zero or more times |
+ |
Preceding item will be matched one or more times |
{n} |
Preceding item is matched exactly n times |
{n,} |
Preceding item is matched n or more times |
{,m} |
Preceding item is matched at most m times |
{n,m} |
Preceding item is matched at least n times,but not more than m times |
\b |
Word boundary |
\B |
Not word boundary |
\< |
Start of word |
\> |
End of word |
\w |
Word character, same as [[:alnum:]] |
\W |
Not word character, same as [^[:alnum:]] |
\\ |
Literal \ |
[abq] |
Range (a or b or q) |
[^abq] |
Not a or b or q |
[a-q] |
Letters between a and q, inclusive |
[]ab] |
match literal ] or a or b. Place ] first |
[ab^] |
a or b or literal ^. Place ^ anywhere but first |
[aq-] |
a or q or literal -. Place - last. |
(a|b) |
a or b |
(...) |
Group |
\n |
Backreference, ie \1 matches first group |
Note: most meta-characters lose their special meaning inside brackets
| Symbol | Meaning |
|---|---|
[:alnum:] |
Digits and letters |
[:alpha:] |
All letters |
[:cntrl:] |
Control characters |
[:digit:] |
Digits |
[:graph:] |
Printed characters |
[:lower:] |
Lower case letters |
[:print:] |
Printed characters and spaces |
[:punct:] |
Punctuation |
[:space:] |
Blank characters |
[:upper:] |
Upper case letters |
[:xdigit:] |
Hexadecimal digits |
Note: In actual usage these character classes will be inside brackets,thus:
[[:alpha:]] but could also be [[:alpha:]7] match a letter or 7