Regex is something I have always been slighly afraid of - long, intimidating and barely readable strings. Today was the day I learned regex, and started to use them to be more productive.

Most of the examples and resources I used come from RegexOne.

Basic Matching

The most basic regex is simply the text you want to match.

/abc/ -> matches anything starting with "abc"

Easy enough.

Digits and Characters with \d and \d

We can also match any non digit character using \D. Digits match using \d\. For example, we can match 12 using \d\. Or a1 with \D\d.

Wildcards with the .

. matches anything. Anything. Even whitespace. You can match a literal . by escaping with .

Groups with []

You can specify groups to match using []. For example we can match man and can but not dan with [mc].

Excluding matches with ^

You can exclude matches using ^. We can match hog and dog but ignore bog using [^b]. For completeness we can do [^b]og.

Ranges

You can define ranges, for example to match English words the follow is often used: [A-Za-z0-9_]. Writing that is tiresome, so you can use the metacharacter \w which accomplishes the same thing. An example of a range is [A-D], which matches anything containing A, B, C or D.

Matching a repetitions

You can match repetitions using {}. For example a{2} matches two or more a letters. It also works with ranges, for example [a-c]{2} matches a, b or c twice in a row. You can also match bear with b{1}..r{1}.

* and + for arbitrary matching

You can match any number of occurrences with * and at least one or more with +. For example a+ matches any string with at least one a. Or [abc]+ matches any string with either a, b or c.

Optional matches with ?

? indicates a character is optional. ab?c will match both abc and ac, since b? is optional.

Take the following:

1 file found?
2 files found?
24 files found?
0 files found.

We can match the first three, but not the last, \?. Or a complete match with \d?.*\?. An optional digit, any number of characters and a ?.

Whitespace with \s

You can match whitespace with \s. For example:

  a
    b
c

We can match the first two with \s+. This means “starting with whitespace”. We can improve it with \s+.*, to look for characters after the whitespace.

Matching start end ends with ^ and $

We can match the start of a line with ^ and the end with $. For example, we can match Australia with ^A.*a$. Starting with A, and arbitrary amount of characters, then ending with a.

Captures

We can specify which part of the regex we want to keep, or capture, using (). Say we have a bunch of files, and we want the filename part of all the .pdf files.

file_record.pdf
file_aaa.pdf
other.png

We can match and capture the pdf file names using (^file.*)\.pdf. Anything starting with file and ending with .pdf, but only keep the filename part.

Multiple capture groups

You can capture multiple sets of charaters in the same expression. Take this example:

Jan 1987
Mar 1983
Oct 2012

We can match the month and year, and then just the year part, with the following:

(\D{3}\s+(\d{4}))

Which reads:

any three non digits
at least one whitespace
any four digits

Another example:

1280x720 
1920x1600   
1024x768

We want to capture both the width and height. the height is three and an optional fourth digit. The regex looks like this:

(\d{4})x+(\d{3}\d?)

Or with |

You can match using logical OR using |. For example you can match “I love cats” and “I love dogs” with I love (cats|dogs).

Some other metacharacters

\S matches non whitespace \b matches the boundary between a word and non word character.

Conclusion

Learning just some basic regex make the whole system a lot less scary, and is very exciting. I will continue practising and learning regex.

20180728 Getting Started with Regex