some notes regarding prolog and working with DCG grammars

Author: step21

notes.md

@@ -0,0 +1,89 @@
+# DCG / Parsing notes
+
+These are various notes and snippets from several sources. Such as clocksin/mellish, tokenize swi-prolog package, dcg_tut by anne hogdon and amzi prolog tutorial.
+
+##
+
+DCGs or what they are converted to are basically difference lists such as `lex([lex, fsdf, sd,"."]-X).` DCGs are a nicer/better notation for that.
+
+
+
+## Alternatives in DCGS
+
+* The `;` operator allows alternatives, kind of like a switch statement(?)
+* `{Var = "somevalue"}` to unify to a certain value if not unified by some other predicate
+* `optional(:Match, :Default)//`` https://www.swi-prolog.org/pldoc/man?predicate=optional//2
+
+## Conversion between a predicate with arguments to a list (or reverse)
+
+`=..` univ - converts predicate and args to list or reverse (list to predicate and args)
+
+## Lookahead
+
+`look_ahead(T), [T] --> [T].`
+`phrase(look_ahead(T), [a], Rest).`
+
+This is actually first removing it then putitng it back.
+
+
+## Trees
+
+* How to map to tree?
+* Recursion but then - probably an extra state or variable than carries tree state... or just look at examples from Anne.
+* How to build parse tree https://cs.union.edu/~striegnk/learn-prolog-now/html/node67.html
+
+### Sorted tree dictionary
+
+* chpt 7 clocksin/mellish
+* Examples of trees?
+
+## For formatting output
+
+`try_literals(X) :- phrase(cliche, X) ,format('~s~n', [X]).`
+
+
+
+## Helper
+
+* `atomic(X).` test of atom or number
+
+## For lists
+
+* `efface` - removes first occurence
+* `delete` - removes all occurences
+* `last()` - last list element
+* first?
+* `nextto(X, Y, L)` (X and Y are consecutive elements in L)
+* findall
+
+
+## Reading and writing files
+
+* `read_file_to_codes(File, Codes, [encoding(utf8)]),`
+* `open(Filename, read, Output).`
+* `close(Output).`
+
+## One or more of some
+
+names ---> [name, conj, names].
+names ---> [name].
+names ---> [].
+
+## 
+
+* Unify for one additional Var - `noun(thing,X) --> [X], {location(X,_)}.`
+
+N to count terms
+AST to add nodes ?
+
+wordlist([X|Y]) --> word(X), whitespace, wordlist(Y).
+wordlist([X]) --> whitespace, wordlist(X).
+wordlist([X]) --> word(X).
+wordlist([X]) --> word(X), whitespace.
+
+word(W) --> charlist(X), {name(W,X)}.
+
+charlist([X|Y]) --> chr(X), charlist(Y).
+charlist([X]) --> chr(X).
+
+chr(X) --> [X],{X>=48}.