searchbox2.py

import ast
import re


class SearchBox:
    ast_table = {
        ast.Eq: ':',
        ast.Is: '=',
        ast.Or: 'or',
        ast.And: 'and',
        ast.Gt: '>',
        ast.GtE: '>=',
        ast.Lt: '<',
        ast.LtE: '<='
    }

    def __init__(self, q):
        self.q = q
        self._sub(' ', 'and')
        self._sub('=', 'is')
        self._sub(':', '==')

    def parse(self):
        """

        :return:
        >>> SearchBox('').parse()
        >>> SearchBox('a').parse()
        'a'
        >>> SearchBox('a:b').parse()
        (':', ['a', 'b'])
        >>> SearchBox('a:b c:d').parse()
        ('and', [(':', ['a', 'b']), (':', ['c', 'd'])])
        >>> SearchBox('a:b:c').parse()
        >>> SearchBox("a:'b c'").parse()
        (':', ['a', 'b c'])
        >>> SearchBox('a:"b c"').parse()
        (':', ['a', 'b c'])
        >>> SearchBox("'a b':c").parse()
        (':', ['a b', 'c'])
        >>> SearchBox("'a:b':c").parse()
        (':', ['a:b', 'c'])
        >>> SearchBox('tag:微信').parse()
        (':', ['tag', '微信'])
        """
        try:
            return self._eval_node(ast.parse(self.q, mode='eval').body)
        except SyntaxError:
            return None

    def _sub(self, src, dst):
        match_obj = re.fullmatch(r'(?:({expr})({src}))+({expr})'.format(
            expr='(?:\".*?\"|\'.*?\'|\S)+',
            src=src
        ), self.q)

        if match_obj:
            new_q = list(match_obj.groups())
            for i, fragment in enumerate(new_q):
                if fragment[0] == src:
                    if not any(f[0] not in '"\'' and src in f for f in {new_q[i-1], new_q[i+1]}):
                        new_q[i] = ' ' + dst + ' '

            self.q = ''.join(new_q)

    def _eval_node(self, node):
        if isinstance(node, ast.Name):
            return node.id
        elif isinstance(node, ast.Str):
            return node.s
        elif isinstance(node, ast.Num):
            return node.n
        elif isinstance(node, ast.Expr):
            return self._eval_node(node.value)
        elif isinstance(node, ast.BoolOp):
            return self.ast_table[type(node.op)], [self._eval_node(v) for v in node.values]
        elif isinstance(node, ast.Compare):
            return self.ast_table[type(node.ops[0])], \
                   [self._eval_node(node.left), self._eval_node(node.comparators[0])]
        else:
            raise SyntaxError


if __name__ == '__main__':
    for q in [
        "'a b':c",
        'a:"b c"',
        "a:'b c'",
        'a:b c:d'
    ]:
        print(SearchBox(q).q)