An open source, non-profit web search engine
Keep source on documents read off term pages; tighten the username reservation
Claude567fe8aa0df2
diff --git a/mwmbl/platform/api.py b/mwmbl/platform/api.pyindex 7f039f4..bee6d45 100644--- a/mwmbl/platform/api.py+++ b/mwmbl/platform/api.py@@ -128,10 +128,12 @@ def register(request, registration: Registration):raise InvalidRequest("Email already registered")username = registration.username or generate_username()- username_taken = MwmblUser.objects.filter(username=registration.username).exists()- # Reserved even before migration 0049 makes the user, so nobody can take it first and fail it.- if registration.username and (username_taken or registration.username == SEED_CRAWLER_USERNAME):- raise InvalidRequest("Username already exists")+ if registration.username:+ username_taken = MwmblUser.objects.filter(username=registration.username).exists()+ # Reserved even before migration 0049 makes the user, so nobody can take it first and+ # fail it - and in any case, as allauth's blacklist is, so nobody can pass for it.+ if username_taken or registration.username.lower() == SEED_CRAWLER_USERNAME:+ raise InvalidRequest("Username already exists")user = MwmblUser(username=username, email=registration.email)user.set_password(registration.password)diff --git a/mwmbl/tinysearchengine/rank.py b/mwmbl/tinysearchengine/rank.pyindex d1fd7cf..3aac63a 100644--- a/mwmbl/tinysearchengine/rank.py+++ b/mwmbl/tinysearchengine/rank.py@@ -411,6 +411,7 @@ class Ranker:result.score,result.term,remove_curate_state(result.state),+ source=result.source,)for result in items_wrong_state]diff --git a/test/test_rank.py b/test/test_rank.pyindex be14d81..4b3ed70 100644--- a/test/test_rank.py+++ b/test/test_rank.py@@ -8,7 +8,7 @@ from requests.exceptions import RetryErrorfrom urllib3.exceptions import MaxRetryError, ResponseErrorfrom mwmbl.tinysearchengine import rank-from mwmbl.tinysearchengine.indexer import Document+from mwmbl.tinysearchengine.indexer import Document, DocumentSourcefrom mwmbl.tinysearchengine.ltr_rank import LTRRankerfrom mwmbl.tinysearchengine.rank import (NUM_WIKI_RESULTS,@@ -84,6 +84,25 @@ def test_search_retrieved_ranks_like_search_without_external_search():assert ranker.external_search_calls == []+def test_a_staan_copy_from_the_index_keeps_its_source():+ """Read off a term's page rather than the curation term's, as it is for any query other+ than the one it was stored under, it still goes out labelled as Staan's."""+ staan_copy = Document(+ "Tokio async runtime",+ "https://tokio.rs/",+ "An async runtime for Rust.",+ 1.0,+ term="tokio",+ source=DocumentSource.STAAN_INDEXED,+ )+ ranker = _TrackingRanker()+ ranker.tiny_index.retrieve.side_effect = lambda term: [staan_copy] if term == "tokio" else []++ (result,) = ranker.search("async tokio ", [], use_external_search=False)++ assert result.source == DocumentSource.STAAN_INDEXED++def _make_retry_error(status_code: int) -> RetryError:reason = ResponseError(f"too many {status_code} error responses")max_retry_error = MaxRetryError(diff --git a/test/test_system_users.py b/test/test_system_users.pyindex 914414c..55f90c0 100644--- a/test/test_system_users.py+++ b/test/test_system_users.py@@ -35,6 +35,19 @@ def test_the_seed_crawlers_username_cannot_be_registered(client):assert not User.objects.filter(username=SEED_CRAWLER_USERNAME).exists()+def test_a_lookalike_of_the_seed_crawlers_username_cannot_be_registered(client):+ response = client.post(+ "/api/v1/platform/register",+ data=json.dumps(+ {"email": "[email protected]", "password": "correctpassword", "username": SEED_CRAWLER_USERNAME.title()}+ ),+ content_type="application/json",+ )++ assert response.status_code == 400++def test_allauth_refuses_the_seed_crawlers_username():"""Settings cannot import SEED_CRAWLER_USERNAME, so they repeat it."""assert settings.ACCOUNT_USERNAME_BLACKLIST == [SEED_CRAWLER_USERNAME]