近日在和朋友讨论 MaskMatch 时偶得2个不错的算法。
函数1 只支持'*','?'模糊匹配。速度比采用递归算法的快近2倍,比TMask方法快很多。
函数2 完全支持正规表达式。速度于之前的相同。(不会正规表达式的朋友慎用)
// ===========================
// Function 1
// ===========================
// Check if the string can match the wildcard. It can be used for unicode strings as well!
// C: 2004-07-24 | M: 2004-07-24
function MaskMatch(const aPattern, aSource: string): Boolean;
var
StringPtr, PatternPtr: PChar;
StringRes, PatternRes: PChar;
begin
Result := False;
StringPtr := PChar(UpperCase(aSource));
PatternPtr := PChar(UpperCase(aPattern));
StringRes := nil;
PatternRes := nil;
repeat
repeat // ohne vorangegangenes "*"
case PatternPtr^ of
#0 : begin
Result := StringPtr^ = #0;
if Result or (StringRes = nil) or (PatternRes = nil) then Exit;
StringPtr := StringRes;
PatternPtr := PatternRes;
Break;
end;
'*': begin
Inc(PatternPtr);
PatternRes := PatternPtr;
Break;
end;
'?': begin
if StringPtr^ = #0 then Exit;
Inc(StringPtr);
Inc(PatternPtr);
end;
else begin
if StringPtr^ = #0 then Exit;
if StringPtr^ <> PatternPtr^ then
begin
if (StringRes = nil) or (PatternRes = nil) then Exit;
StringPtr := StringRes;
PatternPtr := PatternRes;
Break;
end else
begin
Inc(StringPtr);
Inc(PatternPtr);
end;
end;
end;
until False;
repeat // mit vorangegangenem "*"
case PatternPtr^ of
#0 : begin
Result := True;
Exit;
end;
'*': begin
Inc(PatternPtr);
PatternRes := PatternPtr;
end;
'?': begin
if StringPtr^ = #0 then Exit;
Inc(StringPtr);
Inc(PatternPtr);
end;
else begin
repeat
if StringPtr^ = #0 then Exit;
if StringPtr^ = PatternPtr^ then Break;
Inc(StringPtr);
until False;
Inc(StringPtr);
StringRes := StringPtr;
Inc(PatternPtr);
Break;
end;
end;
until False;
until False;
end;
// ===========================
// Function 2
// ===========================
function _MatchPattern(aPattern, aSource: PChar): Boolean;
begin
Result := True;
while (True) do
begin
case aPattern[0] of
#0 : begin
//End of pattern reached.
Result := (aSource[0] = #0); //TRUE if end of aSource.
Exit;
end;
'*': begin //Match zero or more occurances of any char.
if (aPattern[1] = #0) then
begin
//Match any number of trailing chars.
Result := True;
Exit;
end else
Inc(aPattern);
while (aSource[0] <> #0) do
begin
//Try to match any substring of aSource.
if (_MatchPattern(aSource, aPattern)) then
begin
Result := True;
Exit;
end;
//Continue testing next char...
Inc(aSource);
end;
end;
'?': begin //Match any one char.
if (aSource[0] = #0) then
begin
Result := False;
Exit;
end;
[page]
//Continue testing next char...
Inc(aSource);
Inc(aPattern);
end;
'[': begin //Match given set of chars.
if (aPattern[1] in [#0,'[',']']) then
begin
//Invalid Set - So no match.
Result := False;
Exit;
end;
if (aPattern[1] = '^') then
begin
//Match for exclusion of given set...
Inc(aPattern, 2);
Result := True;
while (aPattern[0] <> ']') do
begin
if (aPattern[1] = '-') then
begin
//Match char exclusion range.
if (aSource[0] >= aPattern[0]) and (aSource[0] <= aPattern[2]) then
begin
//Given char failed set exclusion range.
Result := False;
Break;
end else
Inc(aPattern, 3);
end else
begin
//Match individual char exclusion.
if (aSource[0] = aPattern[0]) then
begin
//Given char failed set element exclusion.
Result := False;
Break;
end else
Inc(aPattern);
end;
end;
end else
begin
//Match for inclusion of given set...
Inc(aPattern);
Result := False;
while (aPattern[0] <> ']') do
begin
if (aPattern[1] = '-') then
begin
//Match char inclusion range.
if (aSource[0] >= aPattern[0]) and (aSource[0] <= aPattern[2]) then
begin
//Given char matched set range inclusion.
// Continue testing...
Result := True;
Break;
end else
Inc(aPattern, 3);
end else
begin
//Match individual char inclusion.
if (aSource[0] = aPattern[0]) then
begin
//Given char matched set element inclusion.
// Continue testing...
Result := True;
Break;
end else
Inc(aPattern);
end;
end;
end;
if (Result) then
begin
//Match was found. Continue further.
Inc(aSource);
//Position Pattern to char after "]"
while (aPattern[0] <> ']') and (aPattern[0] <> #0) do Inc(aPattern);
if (aPattern[0] = #0) then
begin
//Invalid Pattern - missing "]"
Result := False;
Exit;
end else
Inc(aPattern);
end else
Exit;
end;
else begin //Match given single char.
if (aSource[0] <> aPattern[0]) then
begin
Result := False;
Break;
end;
//Continue testing next char...
Inc(aSource);
Inc(aPattern);
end;
end;
end;
end;
function MatchPattern(const aPattern, aSource: string): Boolean;
begin
Result := _MatchPattern(PChar(aPattern), PChar(aSource));
end;