sij/khoj
1
0
Fork 0
mirror of https://github.com/khoj-ai/khoj.git synced 2024-11-27 17:35:07 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions 5

Fix date filter usage with multi word queries. Simplify date regex

Browse source
This commit is contained in:
Debanjum Singh Solanky 2022-07-14 21:32:58 +04:00
parent 67e9366c0f
commit 9de2097182
2 changed files with 23 additions and 24 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

24
src/search_filter/date_filter.py
View file

@ -11,11 +11,10 @@ import dateparser as dtparse
# Date Range Filter Regexes
# Example filter queries:
# - dt>=yesterday dt<"tomorrow"
# - dt>="yesterday" dt<"tomorrow"
# - dt>="last week"
# - dt:"next year"
date_regex = r'(?:dt([:><=]{1,2})\"?([\w\-\/]+))?\"?'
date_range_regex=f'.*?\s+{date_regex}\s*{date_regex}.*?'
# - dt:"2 years ago"
date_regex = r"dt([:><=]{1,2})\"(.*?)\""
def date_filter(query, entries, embeddings):
@ -54,20 +53,17 @@ def date_filter(query, entries, embeddings):
def extract_date_range(query):
# find date range filter in query
date_range_match = re.search(date_range_regex, query)
if not date_range_match or date_range_match.groups() == (None, None, None, None):
return None
date_range_matches = re.findall(date_regex, query)
# extract comparators (e.g >,<,=) applied on dates in date filter
date_comparators = [date_cmp for date_cmp in date_range_match.groups()[0::2] if date_cmp]
if len(date_range_matches) == 0:
return None
# extract, parse natural dates ranges from date range filter passed in query
# e.g today maps to (start_of_day, start_of_tomorrow)
query_dtranges = []
for date_str in date_range_match.groups()[1::2]:
if date_str and parse(date_str):
for index, (cmp, date_str) in enumerate(date_range_matches):
if parse(date_str):
dt_start, dt_end = parse(date_str)
query_dtranges.append((dt_start.timestamp(), dt_end.timestamp()))
date_range_matches[index] = [cmp, (dt_start.timestamp(), dt_end.timestamp())]
# Combine dates with their comparators to form date range intervals
# For e.g
@ -76,7 +72,7 @@ def extract_date_range(query):
# ---
effective_date_range = [0, inf]
date_range_considering_comparator = []
for ((dtrange_start, dtrange_end), cmp) in zip(query_dtranges, date_comparators):
for cmp, (dtrange_start, dtrange_end) in date_range_matches:
if cmp == '>':
date_range_considering_comparator += [[dtrange_end, inf]]
elif cmp == '>=':

23
tests/test_date_filter.py
View file

@ -94,17 +94,20 @@ def test_parse():
def test_date_filter_regex():
dtrange_match = re.search(date_filter.date_range_regex, 'head dt>"today" dt:"1984-01-01" tail')
assert dtrange_match.groups() == ('>', 'today', ':', '1984-01-01')
dtrange_match = re.findall(date_filter.date_regex, 'multi word head dt>"today" dt:"1984-01-01"')
assert dtrange_match == [('>', 'today'), (':', '1984-01-01')]
dtrange_match = re.search(date_filter.date_range_regex, 'head dt>="today" dt="1984-01-01"')
assert dtrange_match.groups() == ('>=', 'today', '=', '1984-01-01')
dtrange_match = re.findall(date_filter.date_regex, 'head dt>"today" dt:"1984-01-01" multi word tail')
assert dtrange_match == [('>', 'today'), (':', '1984-01-01')]
dtrange_match = re.search(date_filter.date_range_regex, 'head dt<"today" tail')
assert dtrange_match.groups() == ('<', 'today', None, None)
dtrange_match = re.findall(date_filter.date_regex, 'multi word head dt>="today" dt="1984-01-01"')
assert dtrange_match == [('>=', 'today'), ('=', '1984-01-01')]
dtrange_match = re.search(date_filter.date_range_regex, 'head dt<="today"')
assert dtrange_match.groups() == ('<=', 'today', None, None)
dtrange_match = re.findall(date_filter.date_regex, 'dt<"multi word date" multi word tail')
assert dtrange_match == [('<', 'multi word date')]
dtrange_match = re.search(date_filter.date_range_regex, 'head tail')
assert dtrange_match.groups() == (None, None, None, None)
dtrange_match = re.findall(date_filter.date_regex, 'head dt<="multi word date"')
assert dtrange_match == [('<=', 'multi word date')]
dtrange_match = re.findall(date_filter.date_regex, 'head tail')
assert dtrange_match == []