Initials of [Torsten Berg] (called Torsten Reincke earlier) here on the Wiki.

----
[LV] ''tr'' is also a Unix filter for translating characters .

There are several options for tr.  Perhaps a Tcl version could be
written and added to [fileutil].

It appears that the tr specs are:

   * read stdin and write stdout
   * In function 1, take 2 strings. string1 contains single characters to be sought within the stdin.  string2 contains a map of single characters corresponding to each character in string1.  The resulting output replaces each occurance that matches an entry in string1 with the corresponding string2 character.  This sounds a lot like [string map].
   * note that one can use character class expressions or equivalence class expressions in string1 - unless one is doing case conversion
   * In function 2 (-c) , complement the set of characters specified by string1, then replace them with the characters in string2.
   * Function 3 (-d) indicates that all occurances of characters from string1 are deleted. 
   * Function 4 (-s) indicates that, after functions 1-3 are complete, sequentially repeated instances of the same character will be replaced by one instance of the character, if the character is found in string2. 
   * if string1 contains a character class and Function 4 is being performed, then string1 will be treated as if all characters of that class were specified.
   * if string2 contains a character class such as :upper: or :lower:, then it will be treated as containing only those characters matching the string1 characters.

  * A character in string1 or string2 can be either the actual character, a member of a character class, an octal sequence, or a special ''meta'' escape sequence such as \a, \b, \f, \r, \t, and \v.






----
[Category Person]
[Category Command]