Changes the lexical class of a given character. This is especially useful for implementing compatibility packages.
   Char must be an integer character code, which would typically be
   written in character-quote syntax, eg 0'a instead of 98.
Class is the name of a character class. The following table lists the existing character classes and the default set of characters in that class:
 Class          Default member characters
---------------------------------------------------------
 upper_case     all upper case letters
 underline      _
 lower_case     all lower case letters
 digit          digits
 blank_space    space, tab and nonprintable characters
 end_of_line    newline (NL)
 atom_quote     '
 string_quote   "
 list_quote	`
 chars_quote
 radix
 ascii
 solo           ! ;
 special        ( [ { ) ] } , |
 line_comment   %
 escape         \
 first_comment  /
 second_comment *
 symbol         # + - . : < = > ? @ ^ ~ & $
 terminator
The table corresponds to the default eclipse_language settings. Using other language dialects typically changes some of the assignments on a module-local basis (e.g. for the iso dialect, the meaning of single and double quotes is reversed).
It is not recommended to change the class of the special characters, since in some cases it might make it impossible to correctly parse Prolog terms.
Note: the default class of the back-quote character ` was changed from symbol to list_quote in ECLiPSe 7.0.
Success:
   % The following example illustrates the use
   % of set_chtab/2 to redefine the class of the
   % dollar symbol.
   %
   [eclipse]: X = $a.
                 ^ (here?)
   syntax error: postfix/infix operator expected
   [eclipse]: set_chtab(0'$, lower_case).
   yes.
   [eclipse]: X = $a.
   X = $a
   yes.
   [eclipse]:
Error:
   set_chtab("a",symbol).       (Error 5)
   set_chtab(97,fred).          (Error 6)