# Copyright (C) 2026 The Software Heritage developers
# See the AUTHORS file at the top-level directory of this distribution
# License: GNU Affero General Public License version 3, or any later version
# See top-level LICENSE file for more information
import hashlib
from urllib.parse import urlparse, urlunparse
from django_ratelimit.decorators import ratelimit
from django.http import HttpRequest, HttpResponse, HttpResponseRedirect, QueryDict
from django.shortcuts import render
from swh.web.utils import (
archive,
cache_get,
cache_set,
origin_visit_types,
reverse,
strtobool,
)
from swh.web.utils.identifiers import resolve_swhid
[docs]
@ratelimit(key="user_or_ip", rate="60/m")
def browse_search_view(request: HttpRequest) -> HttpResponse:
origins: list = []
total_results: int | None = -1
previous_page_url = None
next_page_url = None
if "no_js" in request.GET:
# apply same client side processing when javascript is disabled
# get search parameter values
query = request.GET.get("q", "").strip()
with_visit = strtobool(request.GET.get("with_visit", "on"))
with_content = strtobool(request.GET.get("with_content", "on"))
visit_type = request.GET.get("visit_type", "any")
use_ql = strtobool(request.GET.get("use_ql", "off"))
search_in_metadata = strtobool(request.GET.get("search_in_metadata", "off"))
page_token = request.GET.get("page_token")
prev_page_token = None
next_page_token = None
limit = 100
# if a SWHID is provided, try to resolve it and redirect to object page
if query.startswith("swh:"):
swhid_resolved = resolve_swhid(query)
if swhid_resolved["browse_url"]:
return HttpResponseRedirect(swhid_resolved["browse_url"])
# if an origin URL is provided and is archived, redirect to its browsing page
try:
if origin_info := archive.lookup_origin_visit_latest(
query,
require_snapshot=True,
type=visit_type if visit_type != "any" else None,
):
return HttpResponseRedirect(
reverse(
"browse-origin",
query_params={"origin_url": origin_info["origin"]},
)
)
except Exception:
pass
if search_in_metadata:
origins = list(
archive.search_origin_metadata(
query,
limit,
return_metadata=False,
)
)
else:
origins, next_page_token, total_results = archive.search_origin(
url_pattern=query,
use_ql=use_ql,
limit=limit,
with_visit=with_visit,
with_content=with_content,
visit_types=[visit_type] if visit_type not in (None, "any") else None,
page_token=page_token,
)
total_results = total_results or 0
# cache page tokens for previous and next links
key = "origin_search" + "_".join(
map(str, [query, with_visit, with_content, visit_type, use_ql, limit])
)
cache_key = hashlib.md5(key.encode()).hexdigest()
page_tokens = cache_get(cache_key) or []
if next_page_token and (not page_tokens or page_tokens[-1] == page_token):
page_tokens.append(next_page_token)
cache_set(cache_key, page_tokens)
# compute previous and next links if any
current_url = request.build_absolute_uri()
parsed_url = urlparse(current_url)
query_dict = QueryDict(parsed_url.query, mutable=True)
if page_token:
pos = page_tokens.index(page_token)
if pos > 0:
prev_page_token = page_tokens[pos - 1]
query_dict["page_token"] = prev_page_token
previous_page_url = urlunparse(
parsed_url._replace(query=query_dict.urlencode(safe="/;:"))
)
elif pos == 0:
query_dict.pop("page_token", None)
previous_page_url = urlunparse(
parsed_url._replace(query=query_dict.urlencode(safe="/;:"))
)
if next_page_token:
query_dict["page_token"] = next_page_token
next_page_url = urlunparse(
parsed_url._replace(query=query_dict.urlencode(safe="/;:"))
)
return render(
request,
"browse-search.html",
{
"heading": "Search software origins to browse",
"visit_types": origin_visit_types(use_cache=True),
"total_results": total_results,
"origins": origins,
"previous_page_url": previous_page_url,
"next_page_url": next_page_url,
},
)
