Working on #48 : Simple version of ExpressionParser

src/parsers/ExpressionParser.py

@@ -1,22 +1,23 @@
 from core.builtin_concepts_ids import BuiltinConcepts
 from core.sheerka.services.SheerkaExecute import ParserInput
+from core.tokenizer import TokenKind
+from core.utils import get_text_from_tokens
 from parsers.BaseParser import BaseExprParser
 from parsers.FunctionParser import FunctionParser
 from parsers.LogicalOperatorParser import LogicalOperatorParser
 from parsers.RelationalOperatorParser import RelationalOperatorParser
-from parsers.expressions import ParenthesisNode, NameExprNode
+from parsers.expressions import ParenthesisNode, NameExprNode, VariableNode
 
 
 class ExpressionParser(BaseExprParser):
     """
-    Parses xxx (== | > | < | >= | <= | != | in | not in) yyy
-    Nothing else
+    Parses expressions
     """
 
     NAME = "Expression"
 
     def __init__(self, **kwargs):
-        super().__init__(ExpressionParser.NAME, 60, False, yield_eof=True)
+        super().__init__(ExpressionParser.NAME, 60, False, yield_eof=False)
         self.logical_parser = LogicalOperatorParser()
         self.relational_parser = RelationalOperatorParser()
         self.function_parser = FunctionParser()
@@ -62,11 +63,38 @@ class ExpressionParser(BaseExprParser):
 
     def parse_input(self):
         pos = self.parser_input.pos
-        for parser in []: # [self.logical_parser, self.relational_parser, self.function_parser]:
+        for parser in []:  # [self.logical_parser, self.relational_parser, self.function_parser]:
             self.parser_input.seek(pos)  # reset position
             if parser.reset_parser(self.context, self.parser_input):
                 res = parser.parse_input()
                 if res and not parser.has_error:
                     return res
 
-        return NameExprNode(self.parser_input.start, self.parser_input.end, self.parser_input.as_tokens())
+        # try to recognize a VariableNode
+        dots_found = []
+        for i, token in enumerate(self.parser_input.as_tokens()):
+            if token.type == TokenKind.DOT:
+                dots_found.append(i)
+                continue
+
+            if not (token.type == TokenKind.WHITESPACE or
+                    token.type == TokenKind.IDENTIFIER and token.value.isidentifier()):
+                return NameExprNode(self.parser_input.start, self.parser_input.end, self.parser_input.as_tokens())
+
+        if len(dots_found) == 0:
+            return VariableNode(pos, self.parser_input.end, self.parser_input.as_tokens(), self.parser_input.as_text())
+
+        parts = []
+        current_dot_pos = pos
+        for dot_found in dots_found:
+            parts.append(get_text_from_tokens(self.parser_input.tokens[current_dot_pos: dot_found]))
+            current_dot_pos = dot_found + 1
+
+        # do not forget the trailing part
+        parts.append(get_text_from_tokens(self.parser_input.tokens[current_dot_pos: self.parser_input.end + 1]))
+
+        return VariableNode(self.parser_input.start,
+                            self.parser_input.end,
+                            self.parser_input.as_tokens(),
+                            parts[0],
+                            *parts[1:])