Multiple Or Conditions in Regex-CodePudding

Lets say I have messy data pattern:

TRND-0LL2134.SV
TRN_RTXDFT.SV
TRND_ZRSFTFF.SV
DR3345.SV

I am trying to use regex to filter on prefixes with "-" or "_", and those without.

Desired output

column1     column2     column3
=========== =========== ===========
TRND        -           0LL2134.SV
TRN         _           RTXDFT.SV
TRND        _           ZRSFTFF.SV
<blank>     <blank>     DR3345.SV

So far I used

\-|\_

to filter on prefixes with "-" or "" but have a challenge filtering on strings that do not have a prefix with either "-" or ""

CodePudding user response：

Use [-_]? to match the optional separator.

Use a non-greedy quantifier for the first word so that if there's no separator, it will be blank and the word will be associated with column 3.

^([A-Z0-9]*?)([-_]?)([A-Z0-9] \.SV)$

DEMO

CodePudding user response：

Use a ? to make a group optional

^(([A-Z] )[-_])?(\w .SV)$

Demo

If you want to match exactly the characters sequence TRN(D) than:

^((TRND?)[-_])?(\w .SV)$

CodePudding user response：

Your matching expression should contain 3 groups for data to be fed into each of the columns of your output and be anchored at the beginning of your test strings.

Specifically, try ...

/^((.*)([_-])(.*))|^([^_-] )$/

Capture groups #2,3,4 will contain the data to go into your columns 1,2,3, resp.; in case there is none of the separating characters in the test string, the content of column 3 will be in capture group #5

The way you've described your problem, this pattern will always match - the only difference is in how the test string is distributed into the column slots. If valid test strings need to end on .SV, use

/^((.*)([_-])(.*[.]SV))$|^([^_-] [.]SV)$/

As you haven't indicated in which programming environment you are operating this answer can not specifiy how to employ the regex.

Online examples are available on Regex101.com for Variant 1 and Variant 2