Page History: Regular Expression Function
Compare Page Revisions
Page Revision: 2018/10/15 14:10
The following is the
Regular Expression Match function lifted from user Rave's
MMUNIX - UNIX-like commands GREP routine. That in turn used Rob Pike's algorithm. It is reasonably fast and brings a powerful feature to MMBasic. It is not a full-blooded regexp function but catches the following functionality:
. matches any single character
[abc] matches a or b or c
? matches zero or one of the previous character
* matches zero or more of the previous character e.g. A* is any number of A's
+ matches one or more of the previous character
^ matches begin of text e.g. ^A means the string must start with A
$ matches end of text e.g. A$ means the string must end with A
\ Following character to be matched literally. e.g. \? matches ?, \\ matches \ etc.
I extracted the RegExp function so that it could be used as a general function to compare if a string "looks like"...
Notes:- It is case sensitive.
- Not a complete set of expressions, but what there is will cover 90% of requirements
- Recursive matching is not supported, e.g.
[hc]?at matching "at", "hat", and "cat" will not work correctly. - RegExp returns a non-zero on a match (it might not be 1).
Syntax:=RegExp(RegularExpressionString$,MyString$)
Example: simple match for a string that "looks like" it
might be a time. RegExp is used here to determine the "look and feel" of the tested string - the actual values remain to be checked. See
IsTime for a complete solution.
Print RegExp("^[012][0123456789]:[012345][0123456789]$","10:55")
Dependencies:None
Function RegExp(regex$,text$) ' case sensitive
Local s,n,m,f
n=Len(regex$):m=Len(text$)
RegExp=0
If Mid$(regex$,1,1)="^" Then
MatchHere regex$,text$,2,1,n,m,f
RegExp=f
Else
For s=1 To m
MatchHere regex$,text$,1,s,n,m,f
If f Then RegExp=s: Exit For
Next
EndIf
End Function
Sub MatchHere(regex$,text$,r,s,n,m,f)
Local i
Do While r <= n
If Mid$(regex$,r)="$" Then f=s>m: Exit Sub
If Mid$(regex$,r,1)="[" Then
i=Instr(r+2,regex$,"]")
If i Then i=i+1 Else i=r+1
ElseIf Mid$(regex$,r,1)="\" Then
i=r+2
Else
i=r+1
EndIf
If Mid$(regex$,i,1)="*" Then
MatchStar regex$,text$,r,i,s,n,m,f
Exit Sub
EndIf
If Mid$(regex$,i,1)="?" Then
MatchOptional regex$,text$,r,i,s,n,m,f
Exit Sub
EndIf
If s>m Then f=0: Exit Sub
i=r
MatchChar regex$,text$,r,s,f
If Not f Then Exit Sub
If Mid$(regex$,r+1,1)="+" Then
MatchStar regex$,text$,i,r+1,s+1,n,m,f
Exit Sub
EndIf
r=r+1: s=s+1
Loop
f=1
End Sub
Sub MatchStar(regex$,text$,r,i,s,n,m,f)
Do
MatchHere regex$,text$,i+1,(s),n,m,f
If f or s>m Then Exit Sub
MatchChar regex$,text$,(r),s,f
If Not f Then Exit Sub
s=s+1
Loop
End Sub
Sub MatchOptional(regex$,text$,r,i,s,n,m,f)
MatchHere regex$,text$,i+1,(s),n,m,f
If f or s>m Then Exit Sub
MatchChar regex$,text$,(r),s,f
If Not f Then Exit Sub
MatchHere regex$,text$,i+1,s+1,n,m,f
End Sub
Sub MatchChar(regex$,text$,r,s,f)
Local i,j
f=Mid$(regex$,r,1)="."
If f Then Exit Sub
If Mid$(regex$,r,1)="[" Then
i=Instr(r+2,regex$,"]")
If i Then
j=Instr(r+1,regex$,Mid$(text$,s,1))
f=j>0 And j<i:r=i
Exit Sub
EndIf
ElseIf Mid$(regex$,r,1)=Chr$(92) Then
r=r+1
EndIf
f=Mid$(regex$,r,1)=Mid$(text$,s,1)
End Sub