Add method to extract filter terms from query to all filters

- Test the get_filter_term method in all 3 word, file, date filters
- Make the existing can_filter method by default in base filter abstract class
This commit is contained in:
Debanjum Singh Solanky
2023-08-27 19:14:50 -07:00
parent 150105505b
commit d5a2de6222
7 changed files with 62 additions and 12 deletions

View File

@@ -12,9 +12,12 @@ class BaseFilter(ABC):
...
@abstractmethod
def can_filter(self, raw_query: str) -> bool:
def get_filter_terms(self, query: str) -> List[str]:
...
def can_filter(self, raw_query: str) -> bool:
return len(self.get_filter_terms(raw_query)) > 0
@abstractmethod
def apply(self, query: str, entries: List[Entry]) -> Tuple[str, Set[int]]:
...

View File

@@ -3,6 +3,7 @@ import re
import logging
from collections import defaultdict
from datetime import timedelta, datetime
from typing import List
from dateutil.relativedelta import relativedelta
from math import inf
@@ -45,9 +46,9 @@ class DateFilter(BaseFilter):
continue
self.date_to_entry_ids[date_in_entry].add(id)
def can_filter(self, raw_query):
"Check if query contains date filters"
return self.extract_date_range(raw_query) is not None
def get_filter_terms(self, query: str) -> List[str]:
"Get all filter terms in query"
return [f"dt{item[0]}'{item[1]}'" for item in re.findall(self.date_regex, query)]
def defilter(self, query):
# remove date range filter from query

View File

@@ -3,6 +3,7 @@ import re
import fnmatch
import logging
from collections import defaultdict
from typing import List
# Internal Packages
from khoj.search_filter.base_filter import BaseFilter
@@ -25,8 +26,9 @@ class FileFilter(BaseFilter):
for id, entry in enumerate(entries):
self.file_to_entry_map[getattr(entry, self.entry_key)].add(id)
def can_filter(self, raw_query):
return re.search(self.file_filter_regex, raw_query) is not None
def get_filter_terms(self, query: str) -> List[str]:
"Get all filter terms in query"
return [f'file:"{term}"' for term in re.findall(self.file_filter_regex, query)]
def defilter(self, query: str) -> str:
return re.sub(self.file_filter_regex, "", query).strip()

View File

@@ -2,6 +2,7 @@
import re
import logging
from collections import defaultdict
from typing import List
# Internal Packages
from khoj.search_filter.base_filter import BaseFilter
@@ -36,12 +37,11 @@ class WordFilter(BaseFilter):
return self.word_to_entry_index
def can_filter(self, raw_query):
"Check if query contains word filters"
required_words = re.findall(self.required_regex, raw_query)
blocked_words = re.findall(self.blocked_regex, raw_query)
return len(required_words) != 0 or len(blocked_words) != 0
def get_filter_terms(self, query: str) -> List[str]:
"Get all filter terms in query"
required_terms = [f"+{required_term}" for required_term in re.findall(self.required_regex, query)]
blocked_terms = [f"-{blocked_term}" for blocked_term in re.findall(self.blocked_regex, query)]
return required_terms + blocked_terms
def defilter(self, query: str) -> str:
return re.sub(self.blocked_regex, "", re.sub(self.required_regex, "", query)).strip()