Is there a function called MidStr??

[OC DelGuy]

I looked at the reference doc for FPC and see RightStr and LeftStr.  They both return characters from the end and the beginning , respectively.  Is there a MidStr or something that extracts characters from the string in between?

Code: Pascal  [Select][+][-]

1. Saying:='Impossible is for the unwilling';
2. WriteLn(LeftStr(Saying,10);
3. WriteLn(RightStr(Saying,9);
4.  

Output:

Code: Text  [Select][+][-]

1. Impossible
2. unwilling
3.  

I'm looking for something like this:

Code: Pascal  [Select][+][-]

1. WriteLn(MidStr(saying,15,3));

Output:

Code: Text  [Select][+][-]

1. for

[dseligo]

You have function Copy: https://www.freepascal.org/docs-html/rtl/system/copy.html

And UTF8Copy: https://lazarus-ccr.sourceforge.io/docs/lazutils/lazutf8/utf8copy.html

[Bogen85]

Yes.
It is copy
https://www.freepascal.org/docs-html/rtl/system/copy.html

Code: Pascal  [Select][+][-]

1. program midstring;
2. {$mode objfpc}
3. {$h+}
4.  
5. var
6.   s: string;
7.   midStr: string;
8. begin
9.   s := 'That is a copy!';
10.   midStr := Copy(s, 6, 4); // midStr will be 'is a'
11.   writeln(midStr);
12. end.

[OC DelGuy]

Yes.
It is copy
https://www.freepascal.org/docs-html/rtl/system/copy.html

Right-o Bogen!  Thanks!

[TRon]

or: https://www.freepascal.org/docs-html/rtl/sysutils/tstringhelper.substring.html

[Bogen85]

And UTF8Copy: https://lazarus-ccr.sourceforge.io/docs/lazutils/lazutf8/utf8copy.html

This! 

Is likely the preferred overall.

[Thaddy]

Other midstr option (complete program) using type helper for AnsiString:

Code: Pascal  [Select][+][-]

1. {$mode objfpc}{$H+}
2. uses sysutils;
3. begin
4.   writeln('Impossible is for the unwilling'.substring(14,3));  // prints for
5. end.

[Bogen85]

or: https://www.freepascal.org/docs-html/rtl/sysutils/tstringhelper.substring.html

This is not UTF-8 aware though.

Code: Pascal  [Select][+][-]

1. program midstringhelp;
2. {$mode objfpc}
3. {$h+}
4. {$codepage utf8}
5.  
6. uses SysUtils, StrUtils;
7.  
8. var
9.   s, midStr: string;
10. begin
11.   s := 'þháþ íß á ©óöü!';
12.   midStr := s.Substring(6, 4); // midStr should be 'íß á'
13.   writeln(midStr);
14. end.

The expected 4 "characters" (not bytes) at "character" index (not byte) index is not what is returned.

[Thaddy]

True, but not for long: there is much work being done for a unicode16 based rtl.

It is already in the unicodertl branch.
https://gitlab.com/freepascal.org/fpc/source/-/blob/unicodertl/rtl/objpas/sysutils/syshelph.inc

Either way, to OP it does not matter because in his question is not about UTF8.

[marcov]

Or simply use midstr ? 

https://www.freepascal.org/docs-html/rtl/strutils/midstr.html

afaik it was added as a wrapper around copy for VB migration purposes in D6 or so.

[Thaddy]

Oh, forgot all about that. It was added to Delphi even earlier than D6, I believe.

[marcov]

Possible, but only D6 added strutils, so then it must have moved from elsewhere.

[Thaddy]

Possible, but only D6 added strutils, so then it must have moved from elsewhere.

Or we wrote one ourselves, memory leak, but about strutils introduced in D6 you are right. It is not in D4.

[PascalDragon]

True, but not for long: there is much work being done for a unicode16 based rtl.

Even with a UTF-16-based RTL one needs to be careful not to cut a surrogate pair. Or characters that consist of multiple code points (e.g. a¨ for a normalized ä) - a problem which would also exist with UTF-32. Also having the UTF-16 RTL available will not mean that all targets will switch to that. Some might stay ANSI/UTF-8 (e.g. for *nix systems the ANSI/UTF-8 RTL would mean fewer conversions while for Windows the UTF-16 RTL means fewer conversions).

[OC DelGuy]

You have function Copy: **s://www.freepascal.org/docs-html/rtl/system/copy.html

And UTF8Copy: **s://lazarus-ccr.sourceforge.io/docs/lazutils/lazutf8/utf8copy.html

These two links don't seem to work...