X-Git-Url: https://git.phdru.name/?a=blobdiff_plain;f=parser%2Fparser.py;h=dbd054964820852fd530b4cfab0f1da2c8e79e30;hb=240857e1b474ca2d8d3e97f0ac234ca3e4b3ad3a;hp=38c4cd236084d9bfb050214193503d64ced96c4c;hpb=2167fc0e09f4343c0fe67c8c754e29a41769e684;p=phdru.name%2Fcgi-bin%2Fblog-ru%2Fsearch-tags.git

diff --git a/parser/parser.py b/parser/parser.py
old mode 100644
new mode 100755
index 38c4cd2..dbd0549
--- a/parser/parser.py
+++ b/parser/parser.py
@@ -1,58 +1,66 @@
-# Parse query
-
-from ply import lex, yacc
-
-literals = '()'
-
-tokens = ('NAME', 'AND_OP', 'OR_OP', 'NOT_OP')
-
-t_NAME = '[a-z][a-z0-9_]*'
-
-t_AND_OP = '&'
-
-t_OR_OP = r'\|'
-
-t_NOT_OP = '!'
-
-t_ignore = '[ \t]+'
-
-def t_error(t):
-    """Avoid warnings on stderr"""
-
-lexer = lex.lex()
-
-def p_expression_name(p):
-    """expression : NAME"""
-    p[0] = ('NAME', p[1])
-
-def p_expression_and_and(p):
-    """expression : expression AND_OP AND_OP expression"""
-    p[0] = ('AND', p[1], p[4])
-
-def p_expression_and(p):
-    """expression : expression AND_OP expression"""
-    p[0] = ('AND', p[1], p[3])
-
-def p_expression_not(p):
-    """expression : NOT_OP expression"""
-    p[0] = ('NOT', p[2])
-
-def p_expression_or(p):
-    """expression : expression OR_OP expression"""
-    p[0] = ('OR', p[1], p[3])
-
-def p_expression_parens(p):
-    """expression : '(' expression ')'"""
-    p[0] = ('PARENS', p[2])
-
-def p_error(p):
-    """Avoid warnings on stderr"""
-    yacc.restart()
-
-precedence = (
-    ('left', 'OR_OP'),
-    ('left', 'AND_OP'),
-    ('right', 'NOT_OP'),
-)
-
-parser = yacc.yacc()
+#! /usr/bin/env python
+
+import os
+from lark import Lark, Transformer
+
+
+class TagsTransformer(Transformer):
+    def and_expression(self, items):
+        left = items[0]
+        right = items[2]
+        if len(items) > 3:
+            right = self.and_expression(items[2:])
+        return ('AND', left, right)
+
+    def or_expression(self, items):
+        left = items[0]
+        right = items[2]
+        if len(items) > 3:
+            right = self.or_expression(items[2:])
+        return ('OR', left, right)
+
+    def not_expression(self, items):
+        return ('NOT', items[1])
+
+    def expression_parens(self, items):
+        return ('PARENS', items[0])
+
+    def name(self, name):
+        return ('NAME', name[0].value)
+
+
+# cache
+_grammar = None
+
+
+def load_grammar():
+    global _grammar
+    parser_dir = os.path.dirname(__file__)
+    with open(os.path.join(parser_dir, 'grammar.ebnf'), 'rt') as grammar_file:
+        grammar_text = grammar_file.read()
+    grammar_lines = [line for line in grammar_text.splitlines()
+                     if not line.startswith('#')]
+    grammar_text = '\n'.join(grammar_lines)
+    _grammar = Lark(grammar_text)
+
+
+def parse(input):
+    if _grammar is None:
+        load_grammar()
+    tree = _grammar.parse(input)
+    return TagsTransformer().transform(tree)
+
+
+if __name__ == '__main__':
+    print '----------'
+    print parse('test')
+    print parse('!test')
+    print parse('not test')
+    print parse('foo or bar')
+    print parse('foo && bar')
+    print parse('foo && bar && baz')
+    print parse('!foo && bar && baz')
+    print parse('(test)')
+    print parse('(foo || bar)')
+    print parse('(foo and !bar)')
+    print '----------'