adding expression_primary type + new tokens

lexer/lexer.mll

@@ -18,6 +18,7 @@ rule read_token = parse
   | "SUM"                         { SUM }
   | "COUNT"                       { COUNT }
   | "DISTINCT"                    { DISTINCT }
+  | "SUBSTRING"                   { SUBSTRING }
   | "FROM"                        { FROM }
   | "INNER"                       { INNER }
   | "LEFT"                        { LEFT }
@@ -31,13 +32,13 @@ rule read_token = parse
   | "AND"                         { AND }
   | "WHERE"                       { WHERE }
   | "HAVING"                      { HAVING }
+  | "NOT"                         { NOT }
   | "BETWEEN"                     { BETWEEN }
   | "IN"                          { IN }
   | "LIKE"                        { LIKE }
   | "GROUP"                       { GROUP }
   | "BY"                          { BY }
   | "*"                           { ASTERISK }
-  | "."                           { DOT }
   | "="                           { EQUALS_OPERATOR }
   | "("                           { LEFT_PAREN }
   | ")"                           { RIGHT_PAREN }
@@ -49,7 +50,7 @@ rule read_token = parse
   | ':'                           { COLON }
   | whitespace                    { read_token lexbuf }
   | digit+ as integer             { INTEGER (int_of_string integer) }
-  | digit+"."digit+               { FLOAT }
-  | '.'digit+                     { FLOAT }
+  | digit+"."digit+ as number     { FLOAT (float_of_string number) }
+  | '.'digit+ as number           { FLOAT (float_of_string number) }
   | alpha alphanumeric* as ident  { IDENT ident }
   | eof                           { EOF }

lib/ast.ml

@@ -2,13 +2,16 @@ type query =
   (*| Select of column list * table list option * filter option*)
   | Select of column list * table_expression
   | CreateSchema of string
-  | CreateTable of table
+  | CreateTable of table_scope option * table
   | DropSchema of string
   | DropTable of table
+  | DropColumn of string
+and table_scope = 
+  | Global
+  | Local
 and column = 
   | Asterisk
-(*  | Column of string *)
-  | Column of expression * as_clause option
+  | Column of expression_primary * as_clause option
 and as_clause =
   | As of string
 and table_expression =
@@ -25,19 +28,18 @@ and join_type =
   | Union
   | Natural
 and condition = 
-  | Condition of predicand * predicate
+  | Condition of expression_primary * predicate
   | And of condition * condition
   | Or of condition * condition
   | Not of condition
-and predicand = expression
 and predicate = 
-  | Comparison of operator * predicand
-  | Between of predicand * predicand
-  | NotBetween of predicand * predicand
-  | In of predicand list
-  | NotIn of predicand list
-  | Like of predicand
-  | NotLike of predicand
+  | Comparison of operator * expression_primary
+  | Between of expression_primary * expression_primary
+  | NotBetween of expression_primary * expression_primary
+  | In of expression_primary list
+  | NotIn of expression_primary list
+  | Like of expression_primary  
+  | NotLike of expression_primary
 and operator =
   | Equals
   | NotEquals
@@ -47,11 +49,11 @@ and operator =
   | GreaterEquals
 and filter = condition
 and group =
-  | Group of quantifier option * expression list option
+  | Group of quantifier option * expression_primary list option
 and aggregate = 
   | Aggregate of func * filter option
 and func =
-  | Function of function_type * quantifier option * expression
+  | Function of function_type * quantifier option * expression_primary
 and function_type = 
   | Avg
   | Max
@@ -61,11 +63,22 @@ and function_type =
 and quantifier = 
   | All
   | Distinct
-and expression = 
+and expression_primary = 
   | Ref of string
   | StringLiteral of string
   | DateLiteral of string
   | TimeLiteral of string
+  | TimestampLiteral of string
+  | Concatenation of expression_primary * expression_primary
+  | Numeric of expression_primary * sign * expression_primary
+  | Signed of sign * expression_primary
+  | Substring of expression_primary * expression_primary
+and sign =
+  | Plus 
+  | Minus
+  | Times
+  | Divide
+
 
 let rec pp_query fmt ast = 
   match ast with
@@ -86,9 +99,11 @@ and pp_column col =
 and pp_expression exp = 
   match exp with 
   | Ref(name) -> name
-  | StringLiteral(name) -> "'"^name^"'"
-  | DateLiteral(name) -> "'"^name^"'"
-  | TimeLiteral(name) -> "'"^name^"'"
+  | StringLiteral(s) -> "'" ^ s ^ "'"
+  | DateLiteral(d) -> "'" ^ d ^ "'"
+  | TimeLiteral(t) -> "'" ^ t ^ "'"
+  | TimestampLiteral(ts) -> "'" ^ ts ^ "'" 
+  | _ -> "Expression not yet supported"
 
 and pp_table_expression table_exp =
   match table_exp with
@@ -137,8 +152,8 @@ and pp_predicate pred =
   | Comparison(op, exp) -> pp_operator op ^ pp_expression exp 
   | Between(exp1, exp2) -> "BETWEEN " ^ pp_expression exp1 ^ " AND " ^pp_expression exp2
   | NotBetween(exp1, exp2) -> "NOT BETWEEN " ^ pp_expression exp1 ^ " AND " ^pp_expression exp2
-  | Like(exp) -> "LIKE " ^ pp_expression exp
-  | NotLike(exp) -> " NOT LIKE " ^ pp_expression exp
+  (*| Like(exp) -> "LIKE " ^ pp_expression exp
+  | NotLike(exp) -> " NOT LIKE " ^ pp_expression exp*)
   | _ -> failwith "Predicate not supported"
 
 and pp_operator op = 

parser/parser.mly

@@ -11,18 +11,19 @@ open Ast
 %token CROSS NATURAL UNION JOIN
 %token GREATER_THAN_OPERATOR LESS_THAN_OPERATOR EQUALS_OPERATOR
 %token MAX MIN SUM COUNT AVG
+%token SUBSTRING
 %token ASYMMETRIC SYMMETRIC
 %token <string> IDENT
-%token COMMA DOT PIPE
+%token COMMA PIPE
 %token LEFT_PAREN RIGHT_PAREN
-%token ASTERISK
+%token ASTERISK SOLIDUS
 %token AS ON GROUP BY FILTER
 %token OR AND NOT 
 %token TRUE FALSE UNKNOWN
 %token PLUS_SIGN MINUS_SIGN 
 %token DATE TIME TIMESTAMP
 %token <int> INTEGER 
-%token FLOAT
+%token <float> FLOAT
 %token UNDERSCORE QUOTE COLON
 %token EOF
 %start main
@@ -36,6 +37,7 @@ main:
   | table_definition EOF { $1 }
   | drop_schema_statement EOF { $1 }
   | drop_table_statement EOF { $1 }
+  | drop_column_definition EOF { $1 }
 
 (* 5.2 TOKEN / SEPARATOR *)
 
@@ -59,7 +61,7 @@ literal :
   | general_literal { $1 }
 
 unsigned_literal:
-(*  | unsigned_numeric_literal {}*)
+(*  | unsigned_numeric_literal { $1 }*)
   | general_literal { $1 }
 
 general_literal:
@@ -77,14 +79,15 @@ signed_numeric_literal:
   | unsigned_numeric_literal {}
 
 unsigned_numeric_literal :
-  | exact_numeric_literal {}
+  | exact_numeric_literal { $1 }
 
 exact_numeric_literal:
-  | FLOAT {}
+  | FLOAT { }
+  | INTEGER {}
 
 sign:
-  | PLUS_SIGN {}
-  | MINUS_SIGN {}
+  | PLUS_SIGN { Plus }
+  | MINUS_SIGN { Minus }
 
 unsigned_integer:
   | INTEGER { $1 }
@@ -92,7 +95,7 @@ unsigned_integer:
 datetime_literal:
   | date_literal { $1 }
   | time_literal { $1 }
-  (*| timestamp_literal {}*)
+  | timestamp_literal { $1 }
 
 date_literal:
   | DATE date_string { DateLiteral($2) }
@@ -100,12 +103,18 @@ date_literal:
 time_literal:
   | TIME time_string { TimeLiteral($2) }
 
+timestamp_literal:
+  | TIMESTAMP timestamp_string { TimestampLiteral($2) }
+
 date_string:
   | QUOTE unquoted_date_string QUOTE { $2 }
 
 time_string:
   | QUOTE unquoted_time_string QUOTE { $2 }
 
+timestamp_string:
+  | QUOTE unquoted_timestamp_string QUOTE { $2 }
+
 date_value:
   | years_value MINUS_SIGN months_value MINUS_SIGN days_value { String.concat "-" [$1;$3;$5] }
 
@@ -118,6 +127,9 @@ unquoted_date_string:
 unquoted_time_string:
   | time_value { $1 }
 
+unquoted_timestamp_string:
+  | unquoted_date_string unquoted_time_string { $1 ^ " " ^$2 }
+
 years_value :
   | datetime_value { $1 }
 
@@ -154,10 +166,6 @@ schema_name:
 
 (****************************)
 
-character_value_expression:
-  (* | concatenation {} *)
-  | character_factor { $1 }
-
 
 (* 6. SCALAR EXPRESSION *)
 
@@ -171,7 +179,8 @@ parenthesized_value_expression:
   | LEFT_PAREN value_expression RIGHT_PAREN { $2 }
 
 nonparenthesized_value_expression_primary:
-  | unsigned_value_specification { $1 }
+  (* Return StringLiteral, TimeLiteral, DateLiteral, TimestampLiteral *)
+  | unsigned_value_specification { $1 } 
   | column_reference { Ref($1) }
   | set_function_specification { Ref("function") } 
 
@@ -218,29 +227,70 @@ set_function_specification:
 
 value_expression:
   | common_value_expression { $1 }
+  (*| boolean_value_expression { $1 }*)
 
+(* To avoid cycles and therefore reduce/reduce conflicts
+instead of having all *_value_expression below,
+they are cascaded : 
+numeric_value_expression -> string_value_expression 
+  -> datetime_value_expression -> reference_value_expression
+in the last step of the *_value_expression preceding
+instead of value_expression_primary
+
+This way the parser covers all of them every time, and 
+prevent reduce/reduce conflicts *)
 common_value_expression:
-  (*| numeric_value_expression {}*)
-  (*| string_value_expression {} *)
-  | reference_value_expression { $1 }
+  | numeric_value_expression { $1 }
+(*  | string_value_expression { $1 } *)
+(*  | datetime_value_expression { $1 } *)
+(*  | reference_value_expression { $1 }*)
 
 reference_value_expression:
   | value_expression_primary { $1 }
 
 (*************************)
 
+(* 6.26 NUMERIC VALUE EXPRESSION *)
+
+numeric_value_expression:
+  | term { $1 }
+  | numeric_value_expression PLUS_SIGN term { Numeric($1, Plus, $3) }
+  | numeric_value_expression MINUS_SIGN term { Numeric($1, Minus, $3) }
+
+term:
+  | factor { $1 }
+  | term ASTERISK factor { Numeric($1, Times, $3) }
+  | term SOLIDUS factor { Numeric($1, Divide, $3) }
+
+factor:
+  | numeric_primary { $1 }
+  | sign numeric_primary { Signed($1, $2) }
+
+numeric_primary:
+  | string_value_expression { $1 }
+(*  | value_expression_primary { $1 }*)
+
+(*********************************)
+
 (* 6.28 STRING VALUE EXPRESSION *)
 
 string_value_expression :
   | character_value_expression { $1 } 
   (* | blob_value_expression {} *)
 
+character_value_expression:
+(* Return Concatenation of char_expression * expression_primary *)
+  | concatenation { $1 } 
+(* Return Char of string *)
+  | character_factor { $1 }
+
 character_primary:
-  | value_expression_primary { $1 }
-(*  | string_value_function {} *)
+  | datetime_value_expression { $1 }
+(*  | value_expression_primary { $1 }*)
+  | string_value_function { $1 } 
 
 concatenation:
-  | character_value_expression concatenation_operator character_factor {}
+  | character_value_expression concatenation_operator character_factor { Concatenation($1, $3) }
 
 character_factor :
   | character_primary { $1 }
@@ -250,13 +300,39 @@ character_factor :
 (* 6.29 STRING VALUE FUNCTION *)
 
 string_value_function:
-  | character_value_function {} 
+  | character_value_function { $1 } 
 
 character_value_function :
-  | {}
+  | character_substring_function { $1 }
 
+character_substring_function :
+  | SUBSTRING LEFT_PAREN character_value_expression FROM start_position RIGHT_PAREN { Substring($3, $5) }
+
+start_position:
+  | numeric_value_expression { $1 }
 (*************************)
 
+(* 6.30 DATETIME VALUE EXPRESSION *)
+
+datetime_value_expression:
+  | datetime_term { $1 }
+
+datetime_term :
+  | datetime_factor { $1 }
+
+datetime_factor:
+  | datetime_primary { $1 }
+
+datetime_primary:
+  | reference_value_expression { $1 }
+(*  | datetime_value_function { $1 }*)
+
+(**********************************)
+
+(* 6.31 DATETIME VALUE FUCTION *)
+
+(*******************************)
+
 (* 6.34 BOOLEAN VALUE EXPRESSION *)
 
 boolean_value_expression:
@@ -330,11 +406,11 @@ table_expression:
   | from_clause { TableExpression(Some($1), None, None) }
   | from_clause where_clause { TableExpression(Some($1), Some($2), None) }
   | from_clause where_clause group_by_clause { TableExpression(Some($1), Some($2), Some($3)) }
-(*  | from_clause where_clause group_by_clause having_clause { Some($1) }
-  | from_clause group_by_clause { Some($1) }
-  | from_clause group_by_clause having_clause { Some($1) }
-  | from_clause having_clause { Some($1) }
-*)
+  | from_clause where_clause group_by_clause having_clause { TableExpression(Some($1), Some($2), Some($3)) }
+  | from_clause where_clause having_clause { TableExpression(Some($1), Some($2), None) }
+  | from_clause group_by_clause { TableExpression(Some($1), None, Some($2)) }
+  | from_clause group_by_clause having_clause { TableExpression(Some($1), None, Some($2)) }
+  | from_clause having_clause { TableExpression(Some($1), None, None) }
 
 (************************)
 
@@ -484,15 +560,17 @@ select_sublist_element :
   | derived_column { $1 }
 (*  | qualified_asterisk {} *)
 
+(*
 qualified_asterisk:
-  | asterisked_identifier_chain {}
+  | asterisked_identifier_chain DOT ASTERISK {}
 
 asterisked_identifier_chain:
   | asterisked_identifier {}
   | asterisked_identifier_chain DOT asterisked_identifier {}
 
 asterisked_identifier :
-  | IDENT {}
+  | ASTERISK {}
+*)
 
 derived_column:
   | value_expression { Column($1, None) }
@@ -666,22 +744,23 @@ drop_schema_statement:
 (* 11.3 TABLE DEFINITION *)
 
 table_definition :
-  | CREATE TABLE table_name  { CreateTable($3) }
+  | CREATE TABLE table_name  { CreateTable(None, $3) }
+  | CREATE table_scope TABLE table_name  { CreateTable(Some($2), $4) }
 
 table_scope :
-  | global_or_local TEMPORARY {}
+  | global_or_local TEMPORARY { $1 }
 
 global_or_local :
-  | GLOBAL {}
-  | LOCAL {}
+  | GLOBAL { Global }
+  | LOCAL { Local }
 
 (*************************)
 
 (* 11.18 DROP COLUMN DEFINITION *)
 
 drop_column_definition:
-  | DROP column_name {}
-  | DROP COLUMN column_name {}
+  | DROP column_name { DropColumn($2) }
+  | DROP COLUMN column_name { DropColumn($3) }
 
 (********************************)