Compi is born with certain perception which is "GENERIC".
Programmer can change Compi's reaction to incoming text using 
action "percept" as follows:

..percept.<PERCEPTION>


<PERCEPTION> can have following values:



Case <PERCEPTION>="GENERIC"      -  Compi starts with this perception by default.
--------------------------
each_char_is_separator.False        'when true, each letter or character 
                                    'is a single token  of type separator
                                    'text cannot be divides more than to letters;
                                    'please note that characters in this case are 
                                    'supplied from tokenizer to Compy according
                                    'setting of "tz_separator_case";

native_separators.True
separators_are_global.True
recognize_numbers.False             'when true, then Tokenizer assumes that every word starting
                                    'with digit is a number and assigns type "n" to it.
word_case_sensetive.True            'true means to make difference between low and upper case
                                    'when matching event of type word;
keyword_case_sensetive.True         'event-pattern of type keyword is case sensetive;
separator_case_sensetive.True

output_case."N"                 affects tokens which to be output: affects all types of tokens except
                                strings and numbers; description:
                                   N - non-specified; 
                                   U causes rasing to upper case;
                                   L causes lowering case;

output_text_to_console.True

recongnize_strings.False            'when true, then Tokenizer recognizes double quoted strings
                                    'as separate tokens and assigns type "str" to them.  
recognize_vb_strings.False          'true corresponds to "" VB stile of double quotation mark
                                    'coding.
      
tz_strip_qmarks.false             'first and last quotation marks in string  "XXX" are preserved;
tz_conv_str.""                    "P" causes   \ --> \\ and "" --> \" as for PHP

tz_word_case.""                   'please see below;
tz_separator_case.""
tz_id_case.""
tz_keyword_case.""
tz_string_case.""




Case "VBSCRIPT"
---------------
each_char_is_separator.False

native_separators.True
separators_are_global.True
recognize_numbers.True
word_case_sensetive.False
keyword_case_sensetive.False
separator_case_sensetive.False

output_case.""

recongnize_strings.True
recognize_vb_strings.True
tz_strip_qmarks.False

tz_word_case."L"                     'Tokenizer changes word case to low case when submits
                                     'event of type word to Compi.
tz_separator_case."L"
tz_id_case."L"
tz_keyword_case."L"
tz_string_case.""

tz_conv_str."P"




Case "HTML"                          'in the current Compi revision 1.0.14 used to preserve HTML text.
----------

each_char_is_separator.False

native_separators.False
separators_are_global.False
recognize_numbers.False
word_case_sensetive.True
keyword_case_sensetive.True
separator_case_sensetive.True

output_case."N"

recongnize_strings.False
recognize_vb_strings.False
tz_strip_qmarks.False

tz_word_case.""
tz_separator_case.""
tz_id_case.""
tz_keyword_case.""
tz_string_case.""

tz_conv_str.""