Regularexpressions (3) working ideas of regularexpressions

Assume there is such a string of text: aaa1 bbb2 aa11 bb22 a111 b222 AAAA

You can use the expression [A-Za-Z] + \ D + to extract the first six strings. TestCode:

 
Uses regularexpressions; Procedure tform1.formcreate (Sender: tobject); var RegEx: iregex; {regularexpressions advocates iregex instead of tregex} match: imatch; {imatch indicates one of the matched strings} begin {build expression} RegEx: = tregex. create ('[A-Za-Z] + \ D +'); {the first string matching the expression is matched.} match: = RegEx. match ('aaa1 bbb2 aa11 bb22 a111 b222 aaa'); {match is available. value to obtain the matched text. here it will be aaa1} {match is available. length: Get the length of the text. here it will be: 4} {match is available. index to obtain the position of the string in the original string; here it will be: 1, and the second will be 6} {belowProgramAll matched substrings will be traversed} while match. success do {imatch. success: Successful match} begin showmessage (match. value); match: = match. nextmatch; {imatch. nextmatch: Continue to match the next one, and return the next imatch object} end;

  
 

Still face the text above: aaa1 bbb2 aa11 bb22 a111 b222 AAAA

If you want to extract "letters" or "Numbers" based on the previous match, you must use a subexpression;

For example, the expression can be written as follows: ([A-Za-Z] +) (\ D +)

The matched subexpressions are placed in imatch. Groups;
Groups is a collection of igroup objects. Its total number of elements = number of subexpressions + 1. Example:

 
Uses regularexpressions; Procedure tform1.formcreate (Sender: tobject); var RegEx: iregex; match: imatch; s, W, N: string; begin RegEx: = tregex. create ('([A-Za-Z] +) (\ D +)'); match: = RegEx. match ('aaa1 bbb2 aa11 bb22 a111 b222 aaa'); While match. success do begin S: = match. groups [0]. value; {groups [0] match result for placing the entire expression} W: = match. groups [1]. value; {groups [1] match result for placing the first subexpression} n: = match. groups [2]. value; {groups [2] match result for placing the second subexpression} showmessagefmt ('substring: % s; letter: % s; number: % s', [s, W, n]); match: = match. nextmatch; end; (* sub-string: aaa1; letter: AAA; number: 1 sub-string: bbb2; letter: BBB; number: 2 sub-string: aa11; letter: AA; number: 11 substring: bb22; letter: BB; number: 22 substring: a111; letter: A; number: 111 substring: b222; letter: B; Number: 222 *) end;

  
 

Through regular expressions, we mainly do two jobs: search and replace. The above is the basic idea of searching. Let's look at replacement again. The example is directly:

 
Uses regularexpressions; Procedure tform1.formcreate (Sender: tobject); var RegEx: iregex; STR: string; begin RegEx: = tregex. create ('([A-Za-Z] +) (\ D +)'); {Replace the text matched by the expression with ◆} STR: = RegEx. replace ('aaa1 bbb2 aa11 bb22 a111 b222 aaa', '◆ '); showmessage (STR ); {◆ AAAA} {Replace the text matched by the expression with "matching result of the first subexpression"} STR: = RegEx. replace ('aaa1 bbb2 aa11 bb22 a111 b222 aaa', '$ 1'); showmessage (STR ); {aaa bbb AA bb a B AAAA} {Replace the text matched by the expression with "matching result of the second subexpression", which is equivalent to deleting the number} STR: = RegEx. replace ('aaa1 bbb2 aa11 bb22 a111 b222 aaa', '$ 2'); showmessage (STR); {1 2 11 22 111 222 AAAA} end;

  
 

Use $1 $2 in replacement. The following lists all similar references:

$ N // the nth subexpression. $0 indicates the entire expression $ & // with $0 $ + // The Last subexpression; in the above example, it is equivalent to $2 $ _ // replacing the entire input string $ '// replacing all text of the matching input string $' // replacing all of the matched input strings text; this is written in Delphi: $ ''$ // Since $ is used for escape, if you need this symbol itself, You Can $ <Name> // specify the name of the subexpression; the current version is not supported.

  
 

Not all of the above, but it is already the main context; other content also focuses on this idea.