kopia lustrzana https://dev.funkwhale.audio/funkwhale/funkwhale
248 wiersze
8.1 KiB
Python
248 wiersze
8.1 KiB
Python
import re
|
|
|
|
from django.contrib.postgres.search import SearchQuery
|
|
from django.db.models import Q
|
|
|
|
from . import utils
|
|
|
|
|
|
QUERY_REGEX = re.compile(r'(((?P<key>\w+):)?(?P<value>"[^"]+"|[\S]+))')
|
|
|
|
|
|
def parse_query(query):
|
|
"""
|
|
Given a search query such as "hello is:issue status:opened",
|
|
returns a list of dictionnaries discribing each query token
|
|
"""
|
|
matches = [m.groupdict() for m in QUERY_REGEX.finditer(query.lower())]
|
|
for m in matches:
|
|
if m["value"].startswith('"') and m["value"].endswith('"'):
|
|
m["value"] = m["value"][1:-1]
|
|
return matches
|
|
|
|
|
|
def normalize_query(
|
|
query_string,
|
|
findterms=re.compile(r'"([^"]+)"|(\S+)').findall,
|
|
normspace=re.compile(r"\s{2,}").sub,
|
|
):
|
|
""" Splits the query string in invidual keywords, getting rid of unecessary spaces
|
|
and grouping quoted words together.
|
|
Example:
|
|
|
|
>>> normalize_query(' some random words "with quotes " and spaces')
|
|
['some', 'random', 'words', 'with quotes', 'and', 'spaces']
|
|
|
|
"""
|
|
return [normspace(" ", (t[0] or t[1]).strip()) for t in findterms(query_string)]
|
|
|
|
|
|
def get_query(query_string, search_fields):
|
|
""" Returns a query, that is a combination of Q objects. That combination
|
|
aims to search keywords within a model by testing the given search fields.
|
|
|
|
"""
|
|
query = None # Query to search for every search term
|
|
terms = normalize_query(query_string)
|
|
for term in terms:
|
|
or_query = None # Query to search for a given term in each field
|
|
for field_name in search_fields:
|
|
q = Q(**{"%s__icontains" % field_name: term})
|
|
if or_query is None:
|
|
or_query = q
|
|
else:
|
|
or_query = or_query | q
|
|
if query is None:
|
|
query = or_query
|
|
else:
|
|
query = query & or_query
|
|
return query
|
|
|
|
|
|
def remove_chars(string, chars):
|
|
for char in chars:
|
|
string = string.replace(char, "")
|
|
return string
|
|
|
|
|
|
def get_fts_query(query_string, fts_fields=["body_text"], model=None):
|
|
search_type = "raw"
|
|
if query_string.startswith('"') and query_string.endswith('"'):
|
|
# we pass the query directly to the FTS engine
|
|
query_string = query_string[1:-1]
|
|
else:
|
|
query_string = remove_chars(query_string, ['"', "&", "(", ")", "!", "'"])
|
|
parts = query_string.replace(":", "").split(" ")
|
|
parts = ["{}:*".format(p) for p in parts if p]
|
|
if not parts:
|
|
return Q(pk=None)
|
|
|
|
query_string = "&".join(parts)
|
|
|
|
if not fts_fields or not query_string.strip():
|
|
return Q(pk=None)
|
|
query = None
|
|
for field in fts_fields:
|
|
if "__" in field and model:
|
|
# When we have a nested lookup, we switch to a subquery for enhanced performance
|
|
fk_field_name, lookup = (
|
|
field.split("__")[0],
|
|
"__".join(field.split("__")[1:]),
|
|
)
|
|
fk_field = model._meta.get_field(fk_field_name)
|
|
related_model = fk_field.related_model
|
|
subquery = related_model.objects.filter(
|
|
**{
|
|
lookup: SearchQuery(
|
|
query_string, search_type=search_type, config="english_nostop"
|
|
)
|
|
}
|
|
).values_list("pk", flat=True)
|
|
new_query = Q(**{"{}__in".format(fk_field_name): list(subquery)})
|
|
else:
|
|
new_query = Q(
|
|
**{
|
|
field: SearchQuery(
|
|
query_string, search_type=search_type, config="english_nostop"
|
|
)
|
|
}
|
|
)
|
|
query = utils.join_queries_or(query, new_query)
|
|
|
|
return query
|
|
|
|
|
|
def filter_tokens(tokens, valid):
|
|
return [t for t in tokens if t["key"] in valid]
|
|
|
|
|
|
def apply(qs, config_data):
|
|
for k in ["filter_query", "search_query"]:
|
|
q = config_data.get(k)
|
|
if q:
|
|
qs = qs.filter(q)
|
|
distinct = config_data.get("distinct", False)
|
|
if distinct:
|
|
qs = qs.distinct()
|
|
return qs
|
|
|
|
|
|
class SearchConfig:
|
|
def __init__(self, search_fields={}, filter_fields={}, types=[]):
|
|
self.filter_fields = filter_fields
|
|
self.search_fields = search_fields
|
|
self.types = types
|
|
|
|
def clean(self, query):
|
|
tokens = parse_query(query)
|
|
cleaned_data = {}
|
|
cleaned_data["types"] = self.clean_types(filter_tokens(tokens, ["is"]))
|
|
cleaned_data["search_query"] = self.clean_search_query(
|
|
filter_tokens(tokens, [None, "in"] + list(self.search_fields.keys()))
|
|
)
|
|
unhandled_tokens = [
|
|
t
|
|
for t in tokens
|
|
if t["key"] not in [None, "is", "in"] + list(self.search_fields.keys())
|
|
]
|
|
cleaned_data["filter_query"], matching_filters = self.clean_filter_query(
|
|
unhandled_tokens
|
|
)
|
|
if matching_filters:
|
|
cleaned_data["distinct"] = any(
|
|
[
|
|
self.filter_fields[k].get("distinct", False)
|
|
for k in matching_filters
|
|
if k in self.filter_fields
|
|
]
|
|
)
|
|
else:
|
|
cleaned_data["distinct"] = False
|
|
return cleaned_data
|
|
|
|
def clean_search_query(self, tokens):
|
|
if not self.search_fields or not tokens:
|
|
return
|
|
|
|
fields_subset = {
|
|
f for t in filter_tokens(tokens, ["in"]) for f in t["value"].split(",")
|
|
} or set(self.search_fields.keys())
|
|
fields_subset = set(self.search_fields.keys()) & fields_subset
|
|
to_fields = [self.search_fields[k]["to"] for k in fields_subset]
|
|
|
|
specific_field_query = None
|
|
for token in tokens:
|
|
if token["key"] not in self.search_fields:
|
|
continue
|
|
to = self.search_fields[token["key"]]["to"]
|
|
try:
|
|
field = token["field"]
|
|
value = field.clean(token["value"])
|
|
except KeyError:
|
|
# no cleaning to apply
|
|
value = token["value"]
|
|
q = Q(**{"{}__icontains".format(to): value})
|
|
if not specific_field_query:
|
|
specific_field_query = q
|
|
else:
|
|
specific_field_query &= q
|
|
query_string = " ".join([t["value"] for t in filter_tokens(tokens, [None])])
|
|
unhandled_tokens_query = get_query(query_string, sorted(to_fields))
|
|
|
|
if specific_field_query and unhandled_tokens_query:
|
|
return unhandled_tokens_query & specific_field_query
|
|
elif specific_field_query:
|
|
return specific_field_query
|
|
elif unhandled_tokens_query:
|
|
return unhandled_tokens_query
|
|
return None
|
|
|
|
def clean_filter_query(self, tokens):
|
|
if not self.filter_fields or not tokens:
|
|
return None, []
|
|
|
|
matching = [t for t in tokens if t["key"] in self.filter_fields]
|
|
queries = [self.get_filter_query(token) for token in matching]
|
|
query = None
|
|
for q in queries:
|
|
if not query:
|
|
query = q
|
|
else:
|
|
query = query & q
|
|
return query, [m["key"] for m in matching]
|
|
|
|
def get_filter_query(self, token):
|
|
raw_value = token["value"]
|
|
try:
|
|
field = self.filter_fields[token["key"]]["field"]
|
|
value = field.clean(raw_value)
|
|
except KeyError:
|
|
# no cleaning to apply
|
|
value = raw_value
|
|
try:
|
|
query_field = self.filter_fields[token["key"]]["to"]
|
|
return Q(**{query_field: value})
|
|
except KeyError:
|
|
pass
|
|
|
|
# we don't have a basic filter -> field mapping, this likely means we
|
|
# have a dynamic handler in the config
|
|
handler = self.filter_fields[token["key"]]["handler"]
|
|
value = handler(value)
|
|
return value
|
|
|
|
def clean_types(self, tokens):
|
|
if not self.types:
|
|
return []
|
|
|
|
if not tokens:
|
|
# no filtering on type, we return all types
|
|
return [t for key, t in self.types]
|
|
types = []
|
|
for token in tokens:
|
|
for key, t in self.types:
|
|
if key.lower() == token["value"]:
|
|
types.append(t)
|
|
|
|
return types
|