Features: 1) Add support for .sat suffix fields in search query with phrase_prefix match; 2) Incorporate lenient fuzzy match and prefix match into a combined query logic; 3) Enhance function_score with additional filtering for better query relevance.

Fixes: 1) Handle missing or empty query input with a more explicit check; 2) Resolve potential response errors by catching `NotFoundError` exceptions in search execution.

Extra: Refactor code structure to improve readability and modularity, including better slug generation logic and streamlined image URL handling.
This commit is contained in:
Egor Pavlovich Gorbunov 2025-06-19 16:31:54 +03:00
parent 570990fd87
commit 5efac0d5ff

View file

@ -30,42 +30,48 @@ SMART_FIELDS = [
"title.ngram^3", "title.ngram^3",
"title.phonetic", "title.phonetic",
"title.auto^4", "title.auto^4",
"name.sat^6",
"title.sat^4",
] ]
def process_query(query: str = "", request: Request | None = None): def process_query(query: str = "", request: Request | None = None):
""" if not (query := query.strip()):
Perform a lenient, typotolerant, multiindex search.
* Fulltext with fuzziness for spelling mistakes
* `bool_prefix` for edgengram autocomplete / icontains
"""
if not query:
raise ValueError(_("no search term provided.")) raise ValueError(_("no search term provided."))
query = query.strip() sat_match = Q(
try: "multi_match",
q = Q( query=query,
"bool", type="phrase_prefix",
should=[ fields=[f for f in SMART_FIELDS if ".sat" in f],
Q( )
fuzzy_match = Q(
"multi_match", "multi_match",
query=query, query=query,
fields=SMART_FIELDS, fields=SMART_FIELDS,
fuzziness="AUTO", fuzziness="AUTO",
operator="and", operator="and",
), )
Q(
prefix_match = Q(
"multi_match", "multi_match",
query=query, query=query,
fields=[f for f in SMART_FIELDS if f.endswith(".auto")], fields=[f for f in SMART_FIELDS if f.endswith(".auto")],
type="bool_prefix", type="bool_prefix",
), )
],
combined = Q(
"bool",
should=[sat_match, fuzzy_match, prefix_match],
minimum_should_match=1, minimum_should_match=1,
) )
functions = [ functions = [
{
"filter": Q("prefix", **{"name.raw": query.lower()}),
"weight": 5,
},
{ {
"gauss": { "gauss": {
"sales_rank": { "sales_rank": {
@ -81,7 +87,7 @@ def process_query(query: str = "", request: Request | None = None):
boosted = Q( boosted = Q(
"function_score", "function_score",
query=q, query=combined,
boost_mode="sum", boost_mode="sum",
score_mode="sum", score_mode="sum",
functions=functions, functions=functions,
@ -92,38 +98,46 @@ def process_query(query: str = "", request: Request | None = None):
.query(boosted) .query(boosted)
.extra(size=100) .extra(size=100)
) )
response = search.execute()
results: dict = {"products": [], "categories": [], "brands": [], "posts": []} try:
response = search.execute()
except NotFoundError:
raise Http404
results = {"products": [], "categories": [], "brands": [], "posts": []}
for hit in response.hits: for hit in response.hits:
obj_uuid = getattr(hit, "uuid", None) or hit.meta.id obj_uuid = getattr(hit, "uuid", None) or hit.meta.id
obj_name = ( obj_name = getattr(hit, "name", None) or getattr(hit, "title", None) or "N/A"
getattr(hit, "name", None) or getattr(hit, "title", None) or "N/A" raw_slug = getattr(hit, "slug", None) or ""
obj_slug = (
raw_slug or slugify(obj_name)
if hit.meta.index in {"brands", "categories"}
else raw_slug
) )
obj_slug = ""
raw_slug = getattr(hit, "slug", None)
if raw_slug:
obj_slug = raw_slug
elif hit.meta.index == "brands":
obj_slug = slugify(obj_name)
elif hit.meta.index == "categories":
obj_slug = slugify(f"{obj_name}")
image_url = None image_url = None
idx = hit.meta.index idx = hit.meta.index
if idx == "products" and request:
if request:
if idx == "products":
prod = get_object_or_404(Product, uuid=obj_uuid) prod = get_object_or_404(Product, uuid=obj_uuid)
first = prod.images.order_by("priority").first() first = prod.images.order_by("priority").first()
if first and first.image: image_url = (
image_url = request.build_absolute_uri(first.image.url) request.build_absolute_uri(first.image.url)
elif idx == "brands" and request: if first and first.image
else None
)
if idx == "brands":
brand = get_object_or_404(Brand, uuid=obj_uuid) brand = get_object_or_404(Brand, uuid=obj_uuid)
if brand.small_logo: image_url = (
image_url = request.build_absolute_uri(brand.small_logo.url) request.build_absolute_uri(brand.small_logo.url)
elif idx == "categories" and request: if brand.small_logo
else None
)
if idx == "categories":
cat = get_object_or_404(Category, uuid=obj_uuid) cat = get_object_or_404(Category, uuid=obj_uuid)
if cat.image: image_url = (
image_url = request.build_absolute_uri(cat.image.url) request.build_absolute_uri(cat.image.url) if cat.image else None
)
results[idx].append( results[idx].append(
{ {
@ -135,8 +149,6 @@ def process_query(query: str = "", request: Request | None = None):
) )
return results return results
except NotFoundError:
raise Http404
LANGUAGE_ANALYZER_MAP = { LANGUAGE_ANALYZER_MAP = {