[SOLVED] Improvement of function TCustomApplication.FindOptionIndex
lazarus / 64-FPC development

----------------------------------------------------------------------
lagprogramming · 2024-02-20 13:11

packages/fcl-base/src/custapp.pp has
Code: Pascal  [Select][+][-]window.onload = function(){var x1 =
document.getElementById("main_content_section"); if (x1) { var x =
document.getElementsByClassName("geshi");for (var i = 0; i function
TCustomApplication.FindOptionIndex(const S: String;  var Longopt:
Boolean; StartAt : Integer = -1): Integer; Var  SO,O : String;  I,P :
Integer; begin  If Not CaseSensitiveOptions then    SO:=UpperCase(S) 
else    SO:=S;  Result:=-1;  I:=StartAt;  if (I=-1) then   
I:=ParamCount;  While (Result=-1) and (I>0) do    begin   
O:=Params[i];    // - must be seen as an option value    If
(Length(O)>1) and (O[1]=FOptionChar) then      begin     
Delete(O,1,1);      LongOpt:=(Length(O)>0) and (O[1]=FOptionChar);   
 If LongOpt then        begin        Delete(O,1,1);       
P:=Pos('=',O);        If (P<>0) then          O:=Copy(O,1,P-1);      
end;      If Not CaseSensitiveOptions then        O:=UpperCase(O);   
 If (O=SO) then        Result:=i;      end;    Dec(i);    end;end;
1/2 The following lines contain a useless condition
Code: Pascal  [Select][+][-]window.onload = function(){var x1 =
document.getElementById("main_content_section"); if (x1) { var x =
document.getElementsByClassName("geshi");for (var i = 0; i     If
(Length(O)>1) and (O[1]=FOptionChar) then      begin     
Delete(O,1,1);      LongOpt:=(Length(O)>0) and (O[1]=FOptionChar);The
condition (Length(O)>0) is always true because before deleting the
first character using "Delete(O,1,1);" the length of variable O was at
least 2(due to the "If (Length(O)>1)" condition).
So, "LongOpt:=(Length(O)>0) and (O[1]=FOptionChar);" has been replaced
with "LongOpt:=(O[1]=FOptionChar);".

2/2 After that, "If (P<>0) then O:=Copy(O,1,P-1);" has been replaced
with "If (P<>0) then SetLength(O,P-1);".

Here is a patch

----------------------------------------------------------------------
AlexTP · 2024-02-23 12:17

posted to https://gitlab.com/freepascal.org/fpc/source/-/issues/40663
.

----------------------------------------------------------------------
Thaddy · 2024-02-23 15:06

You know I like micro optimizations, but in this case (Assuming {$B-}
state) it does not seem to matter in the generated assembler. Although
your logic is correct.