Strings and special characters removal

[JLWest]

I'm processing a string thru the following statement which removes special characters.

if Str in [#3,#4,#5,#6,#7,#8,#9] then Str:= ' '; end;

However I'm still getting some strange characters on the output. For example an 'RT' with a small 3.

I would like the output to be limited to 'A' .. 'Z' and '0' ..'9'. 
If different I will store in a listbox and manually examine for correction.

I don't know how to write the statements.
Can someone give me an idea where to look or search criteria.

Thanks.

[jamie]

First of all, please use a different word instead of "str" it is reserved for the STR procedure in the Run time.

Anyways.

   if not mystr in ['A'..'Z', '0'..'9'] then MyStr := '';

[JLWest]

First of all, please use a different word instead of "str" it is reserved for the STR procedure in the Run time.

Anyways.

   if not mystr in ['A'..'Z', '0'..'9'] then MyStr := '';

I'll try it . Made the change to tStr from Str.

Thanks

[Blaazen]

Code: Pascal  [Select][+][-]

1. if not mystr in ['A'..'Z', '0'..'9'] then MyStr := '';

This will not work for string, mystr must be type char here, and not has precedence to in operator.

Code: Pascal  [Select][+][-]

1. var myStr: Char;
2. ...
3. if not (mystr in ['A'..'Z', '0'..'9']) then MyStr := #0;

String sshould be precessed char by char + take care if it is UTF8.

[Kays]

[…] mystr must be type char here […]

Well, then you just iterate over it:

Code: Pascal  [Select][+][-]

1. {$mode objFPC}
2. {$longstrings on}
3. uses
4.         sysUtils;
5.  
6. function allCapsAndDigits(const s: string): string;
7. var
8.         c: char;
9. begin
10.         for c in s do
11.         begin
12.                 if c in ['0'..'9', 'A'..'Z'] then
13.                 begin
14.                         appendStr(allCapsAndDigits, c);
15.                 end;
16.         end;
17. end;

[…] If different I will store in a listbox and manually examine for correction. […]

Huh, if you're processing user input, you ideally check whether a valid character has been entered every time as a key has been pressed. This way you don't waste time storing and post-processing input you'll discard anyway.

[dbannon]

Kays, nice, I love recursion. (to understand recursion, first you need to understand recursion)

Anyway, perhaps another approach (that does not require a new function) might be -

Code: Pascal  [Select][+][-]

1.     Index := 1;
2.     while Index <= length(MyStr) do
3.         if not (MyStr[Index] in ['0'..'9', 'A'..'Z']) then
4.             delete(MyStr, Index, 1)
5.         else inc(Index);                

Will cleanly remove any UTF8 codes too.
Davo

[JLWest]

[…] mystr must be type char here […]

Well, then you just iterate over it:

Code: Pascal  [Select][+][-]

1. {$mode objFPC}
2. {$longstrings on}
3. uses
4.         sysUtils;
5.  
6. function allCapsAndDigits(const s: string): string;
7. var
8.         c: char;
9. begin
10.         for c in s do
11.         begin
12.                 if c in ['0'..'9', 'A'..'Z'] then
13.                 begin
14.                         appendStr(allCapsAndDigits, c);
15.                 end;
16.         end;
17. end;

[…] If different I will store in a listbox and manually examine for correction. […]

Huh, if you're processing user input, you ideally check whether a valid character has been entered every time as a key has been pressed. This way you don't waste time storing and post-processing input you'll discard anyway.

NO, Not processing input. Reading a file and saving the data record to a to  a list. Then going thru the listbox and correction the text.
  box

[PascalDragon]

First of all, please use a different word instead of "str" it is reserved for the STR procedure in the Run time.

It's not reserved. Otherwise the compiler would complain. It simply might be confusing if one looks at the code.

[Blaazen]

NO, Not processing input. Reading a file and saving the data record to a to  a list. Then going thru the listbox and correction the text.

And can there be UTF8 characters or is the plain file ASCII?

[furious programming]

Kays, nice, I love recursion. (to understand recursion, first you need to understand recursion)

But this code does not use recursion. @Kays used the function name instead of Result variable. The code of this function can be written differently (in a simpler way), which will give the same result:

Code: Pascal  [Select][+][-]

1. function ToAlphanumericString(const AString: String): String;
2. var
3.   Character: Char;
4. begin
5.   Result := '';
6.  
7.   for Character in AString do
8.     if Character in ['0'..'9', 'A'..'Z'] then
9.       Result += Character;
10. end;

Anyway, perhaps another approach (that does not require a new function) might be - […]

No, never use such a structure (Delete in a loop) because it is slow and unclear.

[lucamar]

NO, Not processing input. Reading a file and saving the data record to a to  a list. Then going thru the listbox and correction the text.

And can there be UTF8 characters or is the plain file ASCII?

In this particular case it doesn' really matter whether there are UTF 8 characters: they'll be discarded anyway. Remember that any byte in a UTF 8 char will have its last bit set and will not satisfy the condition: "char in ['A..Z', '0..9']".

What would matter is wheter the file can contain UTF-16 or 32-bit unicode chars. Or be in other not-so-respectful multibyte encoding.

[JLWest]

NO, Not processing input. Reading a file and saving the data record to a to  a list. Then going thru the listbox and correction the text.

And can there be UTF8 characters or is the plain file ASCII?

No this I think is a UTF-8 file.

[Zvoni]

Errrr.... maybe a stupid question:
What's wrong with defining an Array containing all illegal characters, then just loop through that array firing of a ReplaceStr or StringReplace against the Target-String?

[JLWest]

Errrr.... maybe a stupid question:
What's wrong with defining an Array containing all illegal characters, then just loop through that array firing of a ReplaceStr or StringReplace against the Target-String?

Probly would work, but I got it to work fine with an above solution.

[dbannon]

Errrr.... maybe a stupid question:
What's wrong with defining an Array containing all illegal characters, then just loop through that array firing of a ReplaceStr or StringReplace against the Target-String?

He has a small set of acceptable characters. If the string is just ascii then there is a another, not much bigger set of non acceptable ones. But it sounds like the string maybe UTF8 (or some other unicode ?) - that makes the unacceptable list quite big, er, huge !

Davo