saltzm / yadi Goto Github PK

yadi> m(M) :- movie(M,'Chabrol',,D), D=Y,Y=Z,Z=2000.
Info:
Query transformed into: m(M):-movie(M,'Chabrol',,2000)
2013-11-18 02:40:39,141 INFO sqlalchemy.engine.base.Engine SELECT movie._0 as M FROM movie WHERE movie._3 = 2000 AND movie._1 = 'Chabrol';
2013-11-18 02:40:39,141 INFO sqlalchemy.engine.base.Engine {}
++
++

Hard to intepret

decide whether or not to allow duplicate results

yadi> users(Name):-tuser(Name,,,),not(tuser(Name,,,'Lyon')).
Original query before processing: users():-tuser(,,),!tuser(,,'Lyon')
Info:
Query transformed into: users():-tuser(,,),!tuser(,_,'Lyon')
Traceback (most recent call last):
File "./yadi/TranslationWrapper/translateDatalogToSQL.py", line 28, in translateDatalogToSql
sql_query = QueryExecutor().execute_query(ast_query)
File "./yadi/queryExecutor/queryExecutor.py", line 14, in execute_query
sql = SQLGenerator().get_SQL_code(new_query,query)
File "./yadi/queryExecutor/sqlFactory.py", line 9, in get_SQL_code
return sql_gen.get_SQL_code(query,old_query)
File "./yadi/queryExecutor/sqlFactory.py", line 28, in get_SQL_code
negated_queries = self.get_negated_queries(query.get_relations(),var_dict)
File "./yadi/queryExecutor/sqlFactory.py", line 181, in get_negated_queries
if element.is_variable():
AttributeError: 'str' object has no attribute 'is_variable'

r(M,A):- movie(M, 'Ethan Coen', ,), featuring (M,A,,). is functioning

r(M,Ac):- movie(M,'Ethan Coen', ,), featuring (M,Ac,,). is not functioning

The only difference is the name of the actor variable A, Ac. Shouldn't Ac be accepted?

yadi> tweet(X, Y, Z).
Original query before processing: None:-tweet(X,Y,Z)
Query transformed into: answer(Y,X,Z):-tweet(X,Y,Z)

yadi> moviesDirectorNotActor(Title):- movie(Title,Director,,),not(featuring(Title,Director,)).
Original query before processing: moviesDirectorNotActor():-movie(,),!featuring()
Info:
Query transformed into: moviesDirectorNotActor():-movie(,),!featuring(_)
Traceback (most recent call last):
File "./yadi/TranslationWrapper/translateDatalogToSQL.py", line 28, in translateDatalogToSql
sql_query = QueryExecutor().execute_query(ast_query)
File "./yadi/queryExecutor/queryExecutor.py", line 14, in execute_query
sql = SQLGenerator().get_SQL_code(new_query,query)
File "./yadi/queryExecutor/sqlFactory.py", line 9, in get_SQL_code
return sql_gen.get_SQL_code(query,old_query)
File "./yadi/queryExecutor/sqlFactory.py", line 28, in get_SQL_code
negated_queries = self.get_negated_queries(query.get_relations(),var_dict)
File "./yadi/queryExecutor/sqlFactory.py", line 181, in get_negated_queries
if element.is_variable():
AttributeError: 'str' object has no attribute 'is_variable'

Lexical analysis/parsing to extract tokens from charstream (Lex-like)

• E.g. from "Movie(x, y)" to ("Movie", '(', "x", ',', "y", ')')
• Lex: http://dinosaur.compilertools.net/lex/index.html

yadi> asdf.
Traceback (most recent call last):
File "./yadi/TranslationWrapper/translateDatalogToSQL.py", line 24, in translateDatalogToSql
ast_query_list = ast_builder.buildAST(parsed_statement.asList())
File "./yadi/ASTFactory/ASTBuilder.py", line 35, in buildAST
x = self.handleStatement(code_line)
File "./yadi/ASTFactory/ASTBuilder.py", line 63, in handleStatement
statement_ast = self.handleFact(statement)
File "./yadi/ASTFactory/ASTBuilder.py", line 81, in handleFact
return self.rule_handler.handleFact(statement)
File "./yadi/ASTFactory/RuleHandler.py", line 26, in handleFact
relation_object_list, constraint_object_list = self.handleBody(fact)
File "./yadi/ASTFactory/RuleHandler.py", line 61, in handleBody
relation_list.append(self.handleRelation(body_part))
File "./yadi/ASTFactory/RuleHandler.py", line 77, in handleRelation
if not isinstance(relation[1], list):
IndexError: list index out of range

yadi> r(Title1,Title2,ReleaseDate):-movie(Title1,,,ReleaseDate),movie(Title2,,,ReleaseDate).
Original query before processing: r():-movie(,),movie(,)
Info:
Query transformed into: r():-movie(,),movie(,)
2013-11-18 02:51:50,747 INFO sqlalchemy.engine.base.Engine SELECT 'TRUE' FROM movie as movie1, movie as movie0;
2013-11-18 02:51:50,747 INFO sqlalchemy.engine.base.Engine {}
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+

If someone turns on duplicates, for example, or turns off pretty print, we need a way to store and access that information. The creation of a container class will probably be necessary.

yadi> tuser(,,).
Info:
Query transformed into: answer():-tuser(,,)
2013-11-18 02:15:37,701 INFO sqlalchemy.engine.base.Engine SELECT 'TRUE' FROM tuser;
2013-11-18 02:15:37,701 INFO sqlalchemy.engine.base.Engine {}
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+
| TRUE |
+------+

yadi> r(Title1,Title2,Release_date):-movie(Title1,,,Release_date),movie(Title2,,,Release_date).
Original query before processing: r(date):-movie(,,date),movie(,,date)
Info:
Query transformed into: r(date):-movie(,,date),movie(,_,_date)
2013-11-18 02:50:13,618 INFO sqlalchemy.engine.base.Engine SELECT _date FROM movie as movie1, movie as movie0 WHERE movie1._2 = _date AND movie0._2 = _date;
2013-11-18 02:50:13,618 INFO sqlalchemy.engine.base.Engine {}
2013-11-18 02:50:13,619 INFO sqlalchemy.engine.base.Engine ROLLBACK
(ProgrammingError) column "_date" does not exist
LINE 1: SELECT _date FROM movie as movie1, movie as movie0 WHERE mov...
^
'SELECT _date FROM movie as movie1, movie as movie0 WHERE movie1._2 = _date AND movie0._2 = _date;' {}

yadi> asdf.
Traceback (most recent call last):
File "./yadi/TranslationWrapper/translateDatalogToSQL.py", line 24, in translateDatalogToSql
ast_query_list = ast_builder.buildAST(parsed_statement.asList())
File "./yadi/ASTFactory/ASTBuilder.py", line 35, in buildAST
x = self.handleStatement(code_line)
File "./yadi/ASTFactory/ASTBuilder.py", line 63, in handleStatement
statement_ast = self.handleFact(statement)
File "./yadi/ASTFactory/ASTBuilder.py", line 81, in handleFact
return self.rule_handler.handleFact(statement)
File "./yadi/ASTFactory/RuleHandler.py", line 26, in handleFact
relation_object_list, constraint_object_list = self.handleBody(fact)
File "./yadi/ASTFactory/RuleHandler.py", line 61, in handleBody
relation_list.append(self.handleRelation(body_part))
File "./yadi/ASTFactory/RuleHandler.py", line 77, in handleRelation
if not isinstance(relation[1], list):
IndexError: list index out of range

yadi> tuser(x,y).
Info:
Query transformed into: answer():-tuser(x,y)
2013-11-18 02:18:45,647 INFO sqlalchemy.engine.base.Engine SELECT 'TRUE' FROM tuser WHERE tuser._1 = y AND tuser._0 = x;
2013-11-18 02:18:45,647 INFO sqlalchemy.engine.base.Engine {}
2013-11-18 02:18:45,648 INFO sqlalchemy.engine.base.Engine ROLLBACK
(ProgrammingError) column "y" does not exist
LINE 1: SELECT 'TRUE' FROM tuser WHERE tuser._1 = y AND tuser._0 = x...
^
"SELECT 'TRUE' FROM tuser WHERE tuser._1 = y AND tuser._0 = x;" {}

Query transformed into: movie(A):-movie(A,B,C,D,E,F)
2013-11-20 14:02:06,699 INFO sqlalchemy.engine.base.Engine SELECT movie._0 as A FROM movie;
2013-11-20 14:02:06,699 INFO sqlalchemy.engine.base.Engine {}
+------------------------+
| Home Alone |
+------------------------+
| Life Is Beautiful |
+------------------------+
| Casino |
+------------------------+
| Toy Story |
+------------------------+
| No Country for Old Men |
+------------------------+
yadi> movie(E) :- movie(A, B, C, D, E, F).
Original query before processing: movie(E):-movie(A,B,C,D,E,F)
Info:
Query transformed into: movie(E):-movie(A,B,C,D,E,F)
2013-11-20 14:02:12,415 INFO sqlalchemy.engine.base.Engine SELECT movie._4 as E FROM movie;
2013-11-20 14:02:12,415 INFO sqlalchemy.engine.base.Engine {}
2013-11-20 14:02:12,419 INFO sqlalchemy.engine.base.Engine ROLLBACK
(ProgrammingError) column movie._4 does not exist
LINE 1: SELECT movie._4 as E FROM movie;
^
'SELECT movie._4 as E FROM movie;' {}

Since the SQL selects only based on head variables, if you do something like movie(A):-movie(A,B,C,D,E,F), even if there are only 3 columns in the movie relation, it will return successfully instead of raising an error. I think it should raise an error.