paradizelost/paperless-ngx
1
0
Fork 0
mirror of https://github.com/paperless-ngx/paperless-ngx.git synced 2025-07-22 17:54:40 -05:00
Code Issues Packages Projects Releases Wiki Activity

Handle quotes

Browse Source
This commit is contained in:
shamoon 2025-07-20 06:47:57 -07:00
parent 9f55626ba6
commit c8ed0453a6
1 changed files with 21 additions and 31 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

50
src/documents/index.py
View File

@ -459,43 +459,33 @@ def get_permissions_criterias(user: User | None = None) -> list:
def rewrite_natural_date_keywords(query_string: str) -> str: def rewrite_natural_date_keywords(query_string: str) -> str:
""" """
Rewrites `added:today`, `created:yesterday` into whoosh datetime ranges. Rewrites natural date keywords (e.g. added:today or added:"yesterday") to UTC range syntax for Whoosh.
This prevents UTC confusion when searching with natural language date keywords.
""" """
replacements = {}
patterns = [
("added:today", "added"),
("added:yesterday", "added"),
("created:today", "created"),
("created:yesterday", "created"),
]
tz = get_current_timezone() tz = get_current_timezone()
local_now = now().astimezone(tz) local_now = now().astimezone(tz)
today_start_local = datetime.combine(local_now.date(), time.min).replace(tzinfo=tz) today = local_now.date()
today_end_local = datetime.combine(local_now.date(), time.max).replace(tzinfo=tz) yesterday = today - timedelta(days=1)
yesterday_start_local = today_start_local - timedelta(days=1)
yesterday_end_local = today_end_local - timedelta(days=1)
for pattern, field in patterns: ranges = {
if pattern in query_string: "today": (
if pattern.endswith("today"): datetime.combine(today, time.min, tzinfo=tz),
start = today_start_local datetime.combine(today, time.max, tzinfo=tz),
end = today_end_local ),
else: "yesterday": (
start = yesterday_start_local datetime.combine(yesterday, time.min, tzinfo=tz),
end = yesterday_end_local datetime.combine(yesterday, time.max, tzinfo=tz),
),
}
pattern = r"(\b(?:added|created))\s*:\s*[\"']?(today|yesterday)[\"']?"
def repl(m):
field, keyword = m.group(1), m.group(2)
start, end = ranges[keyword]
start_str = start.astimezone(timezone.utc).strftime("%Y%m%d%H%M%S") start_str = start.astimezone(timezone.utc).strftime("%Y%m%d%H%M%S")
end_str = end.astimezone(timezone.utc).strftime("%Y%m%d%H%M%S") end_str = end.astimezone(timezone.utc).strftime("%Y%m%d%H%M%S")
return f"{field}:[{start_str} TO {end_str}]"
range_expr = f"{field}:[{start_str} TO {end_str}]" return re.sub(pattern, repl, query_string)
logger.warning(f"RANGE: {range_expr}")
replacements[pattern] = range_expr
for match, replacement in replacements.items():
query_string = re.sub(rf"\b{re.escape(match)}\b", replacement, query_string)
return query_string