You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Many words rely on the find term, for example: ' (Tick), postpone, [defined], synonym.
The term find is defined in Forth-2012 as: find: To search the dictionary for a definition name matching a given string.
Later it was updated to: find: To search the search order or a specified word list for a definition name matching a given string.
When the Recognizer API is provided, the standard search order can be ignored doing text translation, and another user-defined search order can be technically used. In such a case it would be unexpected if Tick cannot find an ordinary word that can be successfully translated.
Then, the mentioned words shall use the system's recognizer (the perceptor) to find a word. Of course, they shall analyze the recognized token type, accept appropriate tokens, and discard unknown or inappropriate tokens.
See also the discussed word apply-recognizer-filter ( sd.lexeme xt.recognizer xt.filter -- i*x tt | 0 ) xt.filter ( i*x tt -- i*x tt flag ).
To formally implement this requirement, we will need either update the term find, or update all the words that rely on this term.
The text was updated successfully, but these errors were encountered:
It's one more argument against allowing side effects for recognizers (see #7).
E.g., [defined] (that uses the perceptor) must not have any side effects regardless to what lexeme it's applied.
Many words rely on the find term, for example:
'
(Tick),postpone
,[defined]
,synonym
.The term find is defined in Forth-2012 as:
find: To search the dictionary for a definition name matching a given string.
Later it was updated to:
find: To search the search order or a specified word list for a definition name matching a given string.
When the Recognizer API is provided, the standard search order can be ignored doing text translation, and another user-defined search order can be technically used. In such a case it would be unexpected if Tick cannot find an ordinary word that can be successfully translated.
Then, the mentioned words shall use the system's recognizer (the perceptor) to find a word. Of course, they shall analyze the recognized token type, accept appropriate tokens, and discard unknown or inappropriate tokens.
apply-recognizer-filter ( sd.lexeme xt.recognizer xt.filter -- i*x tt | 0 ) xt.filter ( i*x tt -- i*x tt flag )
.To formally implement this requirement, we will need either update the term find, or update all the words that rely on this term.
The text was updated successfully, but these errors were encountered: