Read a search role's own page for its text in enrich and on the Fetch button when its reader has detail.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
This commit is contained in:
1 parent
07e3bb4756
commit
b148647ac3
5 files changed
+346
-40
No files matched your search
@@ -8,12 +8,15 @@ from pathlib import Path
|
||||
|
||||
import pytest
|
||||
|
||||
import httpx
|
||||
|
||||
from app.jobs import posting
|
||||
from app.storage import Store
|
||||
from desk.models import Item
|
||||
from desk.tools import scrape
|
||||
from desk.tools.article import ArticleContent
|
||||
from desk.tools.checks import Page
|
||||
from plugins import ats
|
||||
from prepped import upsert
|
||||
|
||||
ROOT = Path(__file__).resolve().parents[1]
|
||||
@@ -164,3 +167,82 @@ def test_a_kept_link_is_followed_without_fetching_the_ad_page_again(
|
||||
web = Web(monkeypatch, {}, {LINK: "The real posting."})
|
||||
assert posting.fetch_posting(store, uid).article_text == "The real posting."
|
||||
assert web.asked == [LINK]
|
||||
|
||||
|
||||
AD = (ROOT / "tests" / "fixtures" / "search" / "jobbank-detail.html").read_bytes()
|
||||
AD_URL = "https://jobbank.dk/job/1234/acme/role"
|
||||
|
||||
|
||||
def search_role(store: Store, provider: str = "jobbank", url: str = AD_URL) -> str:
|
||||
"""A role of a saved search as a pull stores it: `provider:id`, source
|
||||
`search`, the search's target in `raw`."""
|
||||
target = f"{provider}:cph"
|
||||
sub = store.add_subscription("search", target, label=target, params={"q": "x"})
|
||||
row = {"source_id": "1234", "title": "Researcher", "url": url, "description": "",
|
||||
"company": "Acme A/S"}
|
||||
item = ats.item_of(row, target, target)
|
||||
item.source = "search"
|
||||
upsert(store, item, sub)
|
||||
return item.uid
|
||||
|
||||
|
||||
class Pages:
|
||||
"""`ats._get` answering from `answers` (URL -> body or the exception to
|
||||
raise), remembering what it was asked."""
|
||||
|
||||
def __init__(self, monkeypatch: pytest.MonkeyPatch, answers: dict[str, object]) -> None:
|
||||
self.asked: list[str] = []
|
||||
self.answers = answers
|
||||
monkeypatch.setattr(ats, "_get", self.get)
|
||||
|
||||
def get(self, client: object, url: str, host: str, slug: str, **kw: object) -> bytes:
|
||||
self.asked.append(url)
|
||||
answer = self.answers[url]
|
||||
if isinstance(answer, Exception):
|
||||
raise answer
|
||||
return answer # type: ignore[return-value]
|
||||
|
||||
|
||||
def test_a_search_role_whose_reader_has_detail_is_read_by_its_detail(
|
||||
store: Store, monkeypatch: pytest.MonkeyPatch) -> None:
|
||||
uid = search_role(store)
|
||||
pages = Pages(monkeypatch, {AD_URL: AD})
|
||||
web = Web(monkeypatch, {}, {})
|
||||
|
||||
got = posting.fetch_posting(store, uid)
|
||||
|
||||
assert got.article_text and got.article_at and got.posting_url == ""
|
||||
assert pages.asked == [AD_URL] and web.asked == [] # no scrape of the link
|
||||
with pytest.raises(posting.AlreadyFetched):
|
||||
posting.fetch_posting(store, uid)
|
||||
|
||||
|
||||
@pytest.mark.parametrize(("answer", "message"), [
|
||||
(httpx.HTTPStatusError("404", request=httpx.Request("GET", AD_URL),
|
||||
response=httpx.Response(404, request=httpx.Request("GET", AD_URL))),
|
||||
"gone"),
|
||||
(httpx.HTTPStatusError("429", request=httpx.Request("GET", AD_URL),
|
||||
response=httpx.Response(429, request=httpx.Request("GET", AD_URL))),
|
||||
"refusing"),
|
||||
(httpx.ConnectError("down"), "could not be reached"),
|
||||
(b"<html>Just a moment...</html>", "Nothing readable"),
|
||||
(b'<script type="application/ld+json">{"@type": "JobPosting"}</script>', "no readable text"),
|
||||
])
|
||||
def test_a_failed_detail_read_says_why(store: Store, monkeypatch: pytest.MonkeyPatch,
|
||||
answer: object, message: str) -> None:
|
||||
uid = search_role(store)
|
||||
Pages(monkeypatch, {AD_URL: answer})
|
||||
with pytest.raises(posting.Unreadable, match=message) as failure:
|
||||
posting.fetch_posting(store, uid)
|
||||
assert failure.value.status == 422
|
||||
stored = store.role(uid)
|
||||
assert stored is not None and stored.article_text == ""
|
||||
|
||||
|
||||
def test_a_search_role_without_detail_is_scraped_from_its_link(
|
||||
store: Store, monkeypatch: pytest.MonkeyPatch) -> None:
|
||||
uid = search_role(store, "adzuna", "https://www.adzuna.dk/details/1")
|
||||
pages = Pages(monkeypatch, {})
|
||||
Web(monkeypatch, {}, {"https://www.adzuna.dk/details/1": "The posting."})
|
||||
assert posting.fetch_posting(store, uid).article_text == "The posting."
|
||||
assert pages.asked == []
|
||||
Reference in new issue
Block a user