fix: support new "most popular 100" & "bottom 100" HTML

The previous version had all 100 movies rendered into the HTML.  The new
version has only the top 25 rendered into HTML, but the whole list has
been made available as LD+JSON data.
Since we can easily support both, we don't (yet) remove the old parser.

tests/test_imdb.py

@@ -30,29 +30,43 @@ def test_score_conversion(score: int):
     assert score == score_from_imdb_rating(imdb_rating_from_score(score))
 
 
+@pytest.mark.parametrize(
+    "fixture",
+    (
+        ("most_popular_100.html.bz2"),
+        ("most_popular_100-20240714.html.bz2"),
+    ),
+)
 @pytest.mark.asyncio
-async def test_load_most_popular_100(monkeypatch):
+async def test_load_most_popular_100(monkeypatch, fixture: str):
-    with bz2.open(fixturesdir / "most_popular_100.html.bz2", "rb") as f:
+    with bz2.open(fixturesdir / fixture, "rb") as f:
         html = f.read()
     soup = bs4.BeautifulSoup(html, "html5lib")
 
     monkeypatch.setattr(imdb, "asoup_from_url", AsyncMock(return_value=soup))
 
     movie_ids = await imdb.load_most_popular_100()
-    assert len(movie_ids) == 100
+    assert len(set(movie_ids)) == 100
     assert all(id_.startswith("tt") for id_ in movie_ids)
 
 
+@pytest.mark.parametrize(
+    "fixture",
+    (
+        ("bottom_100.html.bz2"),
+        ("bottom_100-20240714.html.bz2"),
+    ),
+)
 @pytest.mark.asyncio
-async def test_load_bottom_100(monkeypatch):
+async def test_load_bottom_100(monkeypatch, fixture: str):
-    with bz2.open(fixturesdir / "bottom_100.html.bz2", "rb") as f:
+    with bz2.open(fixturesdir / fixture, "rb") as f:
         html = f.read()
     soup = bs4.BeautifulSoup(html, "html5lib")
 
     monkeypatch.setattr(imdb, "asoup_from_url", AsyncMock(return_value=soup))
 
     movie_ids = await imdb.load_bottom_100()
-    assert len(movie_ids) == 100
+    assert len(set(movie_ids)) == 100
     assert all(id_.startswith("tt") for id_ in movie_ids)
 
 

unwind/imdb.py

@@ -4,7 +4,7 @@ import re
 from collections import namedtuple
 from dataclasses import dataclass, field
 from datetime import datetime
-from typing import AsyncIterable
+from typing import AsyncIterable, Iterable
 from urllib.parse import urljoin
 
 import bs4
@@ -106,7 +106,9 @@ find_year = re.compile(
     r"(\([IVX]+\) )?\((?P<year>\d{4})(–( |\d{4})| (?P<type>[^)]+))?\)"  # noqa: RUF001
 ).fullmatch
 # find_year_2: e.g. "2024", "1971–2003", "2024–"  # noqa: RUF003
-find_year_2 = re.compile(r"(?P<year>\d{4})(–(?P<end_year>\d{4})?)?").fullmatch  # noqa: RUF001
+find_year_2 = re.compile(
+    r"(?P<year>\d{4})(–(?P<end_year>\d{4})?)?"  # noqa: RUF001
+).fullmatch
 find_movie_id = re.compile(r"/title/(?P<id>tt\d+)/").search
 find_movie_name = re.compile(r"\d+\. (?P<name>.+)").fullmatch
 # find_vote_count: e.g. "(5.9K)", "(1K)", "(8)"
@@ -398,17 +400,33 @@ async def load_ratings(user_id: ImdbUserId) -> AsyncIterable[Rating]:
             yield rating
 
 
-async def _ids_from_list_html(url: str) -> AsyncIterable[ImdbMovieId]:
+def _ids_from_list_html(soup: bs4.BeautifulSoup) -> Iterable[ImdbMovieId]:
-    """Return all IMDb movie IDs (`tt*`) from the given URL."""
+    """Return all IMDb movie IDs (`tt*`) from the given soup."""
     # document.querySelectorAll('li.ipc-metadata-list-summary-item a.ipc-title-link-wrapper')
     # .href: '/title/tt1213644/?ref_=chtbtm_t_1'
     # .text(): '1. Disaster Movie'
-    soup = await asoup_from_url(url)
     for item in soup.find_all("li", "ipc-metadata-list-summary-item"):
         if (link := item.find("a", "ipc-title-link-wrapper")) is not None:
             if (href := link.get("href")) is not None:
                 if match_ := find_movie_id(href):
-                    yield match_["id"]
+                    yield ImdbMovieId(match_["id"])
+
+
+def _items_from_ldjson(soup: bs4.BeautifulSoup) -> Iterable[dict]:
+    """Return all items from the LD+JSON block in the given soup."""
+    if (item := soup.find("script", type="application/ld+json")) is None:
+        raise RuntimeError("Could not find LD+JSON data.")
+    data = json.loads(item.string.strip())
+    if data["@type"] != "ItemList":
+        raise RuntimeError(f"Expected ItemList, got {data['@type']!a}.")
+    for item in data["itemListElement"]:
+        yield item["item"]
+
+
+def _ids_from_ldjson(soup: bs4.BeautifulSoup) -> Iterable[ImdbMovieId]:
+    for item in _items_from_ldjson(soup):
+        if match_ := find_movie_id(item["url"]):
+            yield ImdbMovieId(match_["id"])
 
 
 async def load_most_popular_100() -> list[ImdbMovieId]:
@@ -418,7 +436,11 @@ async def load_most_popular_100() -> list[ImdbMovieId]:
     As determined by IMDb users
     """
     url = "https://www.imdb.com/chart/moviemeter/"
-    ids = [tid async for tid in _ids_from_list_html(url)]
+    soup = await asoup_from_url(url)
+    try:
+        ids = list(_ids_from_ldjson(soup))
+    except RuntimeError:
+        ids = list(_ids_from_list_html(soup))
     if len(ids) != 100:
         raise RuntimeError(f"Expected exactly 100 items, got {len(ids)}")
     return ids
@@ -431,7 +453,11 @@ async def load_bottom_100() -> list[ImdbMovieId]:
     Bottom 100 as voted by IMDb users
     """
     url = "https://www.imdb.com/chart/bottom/"
-    ids = [tid async for tid in _ids_from_list_html(url)]
+    soup = await asoup_from_url(url)
+    try:
+        ids = list(_ids_from_ldjson(soup))
+    except RuntimeError:
+        ids = list(_ids_from_list_html(soup))
     if len(ids) != 100:
         raise RuntimeError(f"Expected exactly 100 items, got {len(ids)}")
     return ids