Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-09 03:04:10 -04:00
Code Issues Packages Projects Releases Wiki Activity

Initial stab at supporting smart quotes as delimiters in the search query parser

Browse Source
This commit is contained in:
Kovid Goyal 2018-01-18 12:33:07 +05:30
parent 13d0a7f353
commit ac3132e541
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
1 changed files with 24 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

41
src/calibre/utils/search_query_parser.py
View File

@ -142,14 +142,16 @@ class Parser(object):
WORD = 2 WORD = 2
QUOTED_WORD = 3 QUOTED_WORD = 3
EOF = 4 EOF = 4
REPLACEMENTS = tuple((u'\\' + x, unichr(i + 1)) for i, x in enumerate(ur'\"()“”'))
# Had to translate named constants to numeric values # Had to translate named constants to numeric values
lex_scanner = re.Scanner([ lex_scanner = re.Scanner([
(r'[()]', lambda x,t: (Parser.OPCODE, t)), (ur'[()]', lambda x,t: (Parser.OPCODE, t)),
(r'@.+?:[^")\s]+', lambda x,t: (Parser.WORD, unicode(t))), (ur'@.+?:[^"“)\s]+', lambda x,t: (Parser.WORD, unicode(t))),
(r'[^"()\s]+', lambda x,t: (Parser.WORD, unicode(t))), (ur'[^"“()\s]+', lambda x,t: (Parser.WORD, unicode(t))),
(r'".*?((?<!\\)")', lambda x,t: (Parser.QUOTED_WORD, t[1:-1])), (ur'".*?((?<!\\)")', lambda x,t: (Parser.QUOTED_WORD, t[1:-1])),
(r'\s+', None) (ur'“.*?((?<!\\)”)', lambda x,t: (Parser.QUOTED_WORD, t[1:-1])),
(ur'\s+', None)
], flags=re.DOTALL) ], flags=re.DOTALL)
def token(self, advance=False): def token(self, advance=False):
@ -179,21 +181,26 @@ class Parser(object):
def advance(self): def advance(self):
self.current_token += 1 self.current_token += 1
def parse(self, expr, locations): def tokenize(self, expr):
self.locations = locations
# Strip out escaped backslashes, quotes and parens so that the # Strip out escaped backslashes, quotes and parens so that the
# lex scanner doesn't get confused. We put them back later. # lex scanner doesn't get confused. We put them back later.
expr = expr.replace(u'\\\\', u'\x01').replace(u'\\"', u'\x02') for k, v in self.REPLACEMENTS:
expr = expr.replace(u'\\(', u'\x03').replace(u'\\)', u'\x04') expr = expr.replace(k, v)
self.tokens = self.lex_scanner.scan(expr)[0] tokens = self.lex_scanner.scan(expr)[0]
for (i,tok) in enumerate(self.tokens):
tt, tv = tok
if tt == self.WORD or tt == self.QUOTED_WORD:
self.tokens[i] = (tt,
tv.replace(u'\x01', u'\\').replace(u'\x02', u'"').
replace(u'\x03', u'(').replace(u'\x04', u')'))
def unescape(x):
for k, v in self.REPLACEMENTS:
x = x.replace(v, k[1:])
return x
return [
(tt, unescape(tv) if tt in (self.WORD, self.QUOTED_WORD) else tv)
for tt, tv in tokens
]
def parse(self, expr, locations):
self.locations = locations
self.tokens = self.tokenize(expr)
self.current_token = 0 self.current_token = 0
prog = self.or_expression() prog = self.or_expression()
if not self.is_eof(): if not self.is_eof():