Several basic concepts of regularity

I often see articles that talk about regularity, but only talk about methods, but rarely talk about the following basic concepts:

1. Greed: +,*,?,{m,n}, etc. are greedy matches by default, that is, as many matches as possible, also called maximum matches.
If you add ? afterwards, it will be converted into non-greedy matching and requires higher version support

2. Getting: The default is to use (x|y) to get matches. In many cases, it is just a test and does not necessarily require the matching data to be obtained. Especially in nested matching or big data, non-get matching (?:x|y) should be used, which improves efficiency and optimizes the program.

3. Consumption: The default is consumption matching, and it is generally non-consumption matching in pre-check.
For example, 2003-2-8 will become 2003-02-08
If the second match is used with /-(\d)-/ will start from 8, thus only replacing the first 2, error
If /-(\d)(?=-)/ is used, the second match starts from the second - that is, no characters are consumed

4. Pre-check: js is divided into positive pre-check and negative pre-check
As above (?=pattern) is a forward pre-check, match the search string at the beginning of any string matching pattern. Also (?!pattern) is a negative pre-check, matching the search string at the beginning of any string that does not match pattern. Negative pre-checks are sometimes used to expand [^], [^] are just some characters, and ?! can make the entire string.

5. Callback: It is generally used in substitution, that is, it returns unused substitution values ​​based on unused matching content, thereby simplifying the program and requiring higher version support.

6. Quote: \num Reference to the num-th match obtained.
For example, '(.)\1\1' matches AAA type. '(.)(.)\2\1' Match ABBA type.

Of course there are many more, these are just basic things that need to be mastered.

When encountering regular problems, the following two methods can be generally solved:

1. Classification, that is, list various situations based on all possible situations, such as numbers within 2003
0 0
1-999 [1-9]\d{0,2}
1000-1999 1\d{3}
2000-2003 200[0-3]

So the final match is (0|[1-9]\d{0,2}|1\d{3}|200[0-3])

2. Grouping, that is, dividing the entire sentence into different smallest units, such as ', %, _ appears in double
If it can be grouped, that is, if it is allowed to exist, there is
''
%% If it is an even number greater than 2, it can be subdivided into multiple smaller 2-character units
__
[^'%_] Non-the above characters

So the final match is ^(''|%%|__|[^'%_])*$


--------------------------------------------------------------------------------

replace + function 5.5+ support