Features: 1) Add suggest, price, and quantity fields to ProductDocument; 2) Introduce prepare_suggest method for enhancing autocomplete functionality; 3) Enhance search with filters, function scoring, and result aggregation; 4) Add new analyzers and token filters for synonyms and stopwords.
Fixes: 1) Remove unused imports in `core/elasticsearch/__init__.py`. Extra: Refactor `process_query` for improved readability and functionality; update aggregation and result processing logic; reformat and clean up code.
This commit is contained in:
parent
efd927f4d1
commit
97829d23a6
2 changed files with 129 additions and 81 deletions
|
|
@ -1,12 +1,9 @@
|
|||
from django.conf import settings
|
||||
from django.http import Http404
|
||||
from django.shortcuts import get_object_or_404
|
||||
from django.utils.text import slugify
|
||||
from django.utils.translation import gettext_lazy as _
|
||||
from django_elasticsearch_dsl import fields
|
||||
from django_elasticsearch_dsl.registries import registry
|
||||
from elasticsearch import NotFoundError
|
||||
from elasticsearch.dsl import Q, Search
|
||||
from elasticsearch.dsl import SF, Q, Search
|
||||
from rest_framework.request import Request
|
||||
|
||||
from core.models import Brand, Category, Product
|
||||
|
|
@ -33,88 +30,120 @@ SMART_FIELDS = [
|
|||
]
|
||||
|
||||
|
||||
def process_query(query: str = "", request: Request | None = None):
|
||||
"""
|
||||
Perform a lenient, typo‑tolerant, multi‑index search.
|
||||
|
||||
* Full‑text with fuzziness for spelling mistakes
|
||||
* `bool_prefix` for edge‑ngram autocomplete / “icontains”
|
||||
"""
|
||||
def process_query(
|
||||
query: str = "", request: Request | None = None, filters: dict | None = None
|
||||
):
|
||||
if not query:
|
||||
raise ValueError(_("no search term provided."))
|
||||
raise ValueError("no search term provided.")
|
||||
filters = filters or {}
|
||||
base_q = Q(
|
||||
"bool",
|
||||
should=[
|
||||
Q(
|
||||
"multi_match",
|
||||
query=query,
|
||||
fields=SMART_FIELDS,
|
||||
fuzziness="AUTO",
|
||||
operator="and",
|
||||
),
|
||||
Q(
|
||||
"multi_match",
|
||||
query=query,
|
||||
fields=[f for f in SMART_FIELDS if f.endswith(".auto")],
|
||||
type="bool_prefix",
|
||||
),
|
||||
],
|
||||
minimum_should_match=1,
|
||||
)
|
||||
functions = [
|
||||
SF("field_value_factor", field="quantity", modifier="log1p", missing=0),
|
||||
SF("field_value_factor", field="rating", modifier="sqrt", missing=0),
|
||||
SF("exp", field="created_at", origin="now", scale="30d"), # newness boost
|
||||
]
|
||||
fq = [
|
||||
Q("function_score", query=base_q, functions=functions, boost_mode="sum"),
|
||||
]
|
||||
if "category" in filters:
|
||||
fq.append(Q("term", category__slug=filters["category"]))
|
||||
if "brand" in filters:
|
||||
fq.append(Q("term", brand__slug=filters["brand"]))
|
||||
if "price_min" in filters or "price_max" in filters:
|
||||
range_q = {}
|
||||
if "price_min" in filters:
|
||||
range_q["gte"] = filters["price_min"]
|
||||
if "price_max" in filters:
|
||||
range_q["lte"] = filters["price_max"]
|
||||
fq.append(Q("range", price=range_q))
|
||||
|
||||
query = query.strip()
|
||||
try:
|
||||
q = Q(
|
||||
"bool",
|
||||
should=[
|
||||
Q(
|
||||
"multi_match",
|
||||
query=query,
|
||||
fields=SMART_FIELDS,
|
||||
fuzziness="AUTO",
|
||||
operator="and",
|
||||
),
|
||||
Q(
|
||||
"multi_match",
|
||||
query=query,
|
||||
fields=[f for f in SMART_FIELDS if f.endswith(".auto")],
|
||||
type="bool_prefix",
|
||||
),
|
||||
],
|
||||
minimum_should_match=1,
|
||||
search = (
|
||||
Search(index=["products", "categories", "brands", "posts"])
|
||||
.query(base_q)
|
||||
.filter(*fq)
|
||||
.extra(size=20)
|
||||
.highlight("description", fragment_size=150)
|
||||
.extra(
|
||||
aggs={
|
||||
"by_category": {"terms": {"field": "category.keyword"}},
|
||||
"by_brand": {"terms": {"field": "brand.keyword"}},
|
||||
"price_stats": {"stats": {"field": "price"}},
|
||||
}
|
||||
)
|
||||
)
|
||||
response = search.execute()
|
||||
|
||||
results: dict = {"products": [], "categories": [], "brands": [], "posts": []}
|
||||
for hit in response.hits:
|
||||
obj_uuid = getattr(hit, "uuid", None) or hit.meta.id
|
||||
obj_name = getattr(hit, "name", None) or getattr(hit, "title", None) or "N/A"
|
||||
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
|
||||
idx = hit.meta.index
|
||||
if idx == "products" and request:
|
||||
prod = get_object_or_404(Product, uuid=obj_uuid)
|
||||
first = prod.images.order_by("priority").first()
|
||||
if first and first.image:
|
||||
image_url = request.build_absolute_uri(first.image.url)
|
||||
elif idx == "brands" and request:
|
||||
brand = get_object_or_404(Brand, uuid=obj_uuid)
|
||||
if brand.small_logo:
|
||||
image_url = request.build_absolute_uri(brand.small_logo.url)
|
||||
elif idx == "categories" and request:
|
||||
cat = get_object_or_404(Category, uuid=obj_uuid)
|
||||
if cat.image:
|
||||
image_url = request.build_absolute_uri(cat.image.url)
|
||||
|
||||
results[idx].append(
|
||||
{
|
||||
"uuid": str(obj_uuid),
|
||||
"name": obj_name,
|
||||
"slug": obj_slug,
|
||||
"image": image_url,
|
||||
}
|
||||
)
|
||||
|
||||
search = (
|
||||
Search(index=["products", "categories", "brands", "posts"])
|
||||
.query(q)
|
||||
.extra(size=100)
|
||||
)
|
||||
response = search.execute()
|
||||
facets = {
|
||||
"categories": [
|
||||
(b.key, b.doc_count) for b in response.aggregations.by_category.buckets
|
||||
],
|
||||
"brands": [
|
||||
(b.key, b.doc_count) for b in response.aggregations.by_brand.buckets
|
||||
],
|
||||
"price": {
|
||||
"min": response.aggregations.price_stats.min,
|
||||
"max": response.aggregations.price_stats.max,
|
||||
"avg": response.aggregations.price_stats.avg,
|
||||
},
|
||||
}
|
||||
|
||||
results: dict = {"products": [], "categories": [], "brands": [], "posts": []}
|
||||
for hit in response.hits:
|
||||
obj_uuid = getattr(hit, "uuid", None) or hit.meta.id
|
||||
obj_name = (
|
||||
getattr(hit, "name", None) or getattr(hit, "title", None) or "N/A"
|
||||
)
|
||||
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
|
||||
idx = hit.meta.index
|
||||
if idx == "products" and request:
|
||||
prod = get_object_or_404(Product, uuid=obj_uuid)
|
||||
first = prod.images.order_by("priority").first()
|
||||
if first and first.image:
|
||||
image_url = request.build_absolute_uri(first.image.url)
|
||||
elif idx == "brands" and request:
|
||||
brand = get_object_or_404(Brand, uuid=obj_uuid)
|
||||
if brand.small_logo:
|
||||
image_url = request.build_absolute_uri(brand.small_logo.url)
|
||||
elif idx == "categories" and request:
|
||||
cat = get_object_or_404(Category, uuid=obj_uuid)
|
||||
if cat.image:
|
||||
image_url = request.build_absolute_uri(cat.image.url)
|
||||
|
||||
results[idx].append(
|
||||
{
|
||||
"uuid": str(obj_uuid),
|
||||
"name": obj_name,
|
||||
"slug": obj_slug,
|
||||
"image": image_url,
|
||||
}
|
||||
)
|
||||
|
||||
return results
|
||||
except NotFoundError:
|
||||
raise Http404
|
||||
return {"results": results, "facets": facets}
|
||||
|
||||
|
||||
LANGUAGE_ANALYZER_MAP = {
|
||||
|
|
@ -163,6 +192,8 @@ COMMON_ANALYSIS = {
|
|||
"encoder": "double_metaphone",
|
||||
"replace": False,
|
||||
},
|
||||
"synonym_filter": {"type": "synonym", "synonyms_path": "analysis/synonyms.txt"},
|
||||
"english_stop": {"type": "stop", "stopwords": "_english_"},
|
||||
},
|
||||
"analyzer": {
|
||||
"autocomplete": {
|
||||
|
|
@ -177,6 +208,10 @@ COMMON_ANALYSIS = {
|
|||
"tokenizer": "standard",
|
||||
"filter": ["lowercase", "asciifolding", "ngram_filter"],
|
||||
},
|
||||
"synonym_analyzer": {
|
||||
"tokenizer": "standard",
|
||||
"filter": ["lowercase", "asciifolding", "synonym_filter"],
|
||||
},
|
||||
"name_phonetic": {
|
||||
"tokenizer": "standard",
|
||||
"filter": ["lowercase", "asciifolding", "double_metaphone"],
|
||||
|
|
|
|||
|
|
@ -52,6 +52,12 @@ class _BaseDoc(ActiveOnlyMixin, Document):
|
|||
|
||||
|
||||
class ProductDocument(_BaseDoc):
|
||||
suggest = fields.CompletionField(
|
||||
analyzer="autocomplete",
|
||||
search_analyzer="autocomplete_search",
|
||||
)
|
||||
price = fields.FloatField(attr="price")
|
||||
quantity = fields.IntegerField(attr="quantity")
|
||||
rating = fields.FloatField(attr="rating")
|
||||
|
||||
class Index(_BaseDoc.Index):
|
||||
|
|
@ -61,6 +67,13 @@ class ProductDocument(_BaseDoc):
|
|||
model = Product
|
||||
fields = ["uuid"]
|
||||
|
||||
def prepare_suggest(self, instance):
|
||||
terms = [instance.name]
|
||||
if instance.brand:
|
||||
terms.append(instance.brand.name)
|
||||
terms.append(instance.category.name)
|
||||
return {"input": terms, "weight": int(instance.quantity)}
|
||||
|
||||
|
||||
_add_multilang_fields(ProductDocument)
|
||||
registry.register_document(ProductDocument)
|
||||
|
|
|
|||
Loading…
Reference in a new issue