thermograph/tests/test_content.py

"""Tests for the SSR content pages, ported from backend/tests/web/test_content.py
(repo-split Stage 3). The `client` fixture (conftest.py) fakes api_client by
calling the real backend payload builders against the same synthetic history
fixture backend/tests uses, so assertions here and there agree on the same
numbers. Tests that exercise backend-only concerns (indexnow submission, the
/api/v2/content/* JSON endpoints themselves, data-layer helpers like
cities.title_name) stay in backend/tests -- this file covers only what the SSR
rendering layer (content.py + format.py) does with that data.
"""
import math
import os
import re

import httpx
import pytest

import api_client
import content
import format as fmt

from conftest import B, SLUG


def test_robots_txt(client):
    r = client.get(f"{B}/robots.txt")
    assert r.status_code == 200
    assert "Sitemap:" in r.text and "/sitemap.xml" in r.text
    assert "Disallow: /thermograph/api/" in r.text


def test_sitemap_lists_city_urls(client):
    r = client.get(f"{B}/sitemap.xml")
    assert r.status_code == 200
    assert "<urlset" in r.text
    assert f"/climate/{SLUG}</loc>" in r.text
    assert f"/climate/{SLUG}/july</loc>" in r.text
    assert f"/climate/{SLUG}/records</loc>" in r.text


def test_city_page_renders_stats_in_html(client):
    r = client.get(f"{B}/climate/{SLUG}")
    assert r.status_code == 200
    b = r.text
    assert "London" in b and "climate" in b.lower()
    assert 'rel="canonical"' in b and f"/climate/{SLUG}" in b
    assert '"@type":"Dataset"' in b
    assert "average temperatures by month" in b.lower()
    assert re.search(r"-?\d+°C</span>", b)


def _jsonld_breadcrumbs(html: str) -> list[dict]:
    import json
    m = re.search(r'<script type="application/ld\+json">(.*?)</script>', html, re.S)
    assert m, "no JSON-LD block on page"
    graph = json.loads(m.group(1))["@graph"]
    return [n for n in graph if n["@type"] == "BreadcrumbList"]


@pytest.mark.parametrize("path", [f"/climate/{SLUG}", f"/climate/{SLUG}/records"])
def test_breadcrumb_jsonld_items_valid_for_google(client, path):
    (bc,) = _jsonld_breadcrumbs(client.get(B + path).text)
    els = bc["itemListElement"]
    assert [e["position"] for e in els] == list(range(1, len(els) + 1))
    for e in els[:-1]:
        assert e["item"].startswith("http"), f"crumb {e['name']!r} missing item"


def test_jsonld_url_uses_the_ssr_requests_own_origin(client):
    # A regression guard for a real bug caught during the port: the jsonld
    # "url" field must reflect the browser-facing request the SSR app
    # received, not the internal backend-call's own origin.
    for path in (f"/climate/{SLUG}", f"/climate/{SLUG}/records"):
        html = client.get(B + path).text
        m = re.search(r'"url":"(http://testserver[^"]*)"', html)
        assert m, f"{path}: jsonld url missing or not on the request's own origin"


def test_city_404(client):
    assert client.get(f"{B}/climate/nope-not-a-city").status_code == 404


def _unreachable(*_a, **_k):
    req = httpx.Request("GET", "http://backend.invalid/x")
    raise httpx.ConnectError("connection refused", request=req)


@pytest.mark.parametrize("path, attr", [
    (f"/climate/{SLUG}", "city"),
    (f"/climate/{SLUG}/july", "city_month"),
    (f"/climate/{SLUG}/records", "city_records"),
    ("/climate", "hub"),
    ("/", "home"),
    ("/sitemap.xml", "sitemap"),
])
def test_backend_unreachable_maps_to_503(client, monkeypatch, path, attr):
    """A backend that's down/timing out (httpx.ConnectError, TimeoutException,
    ...) never raises httpx.HTTPStatusError -- it doesn't even get a response
    -- so it must be caught separately from the 404/503 status-code passthrough
    or it escapes as a raw framework 500 instead of a clean 503."""
    monkeypatch.setattr(api_client, attr, _unreachable)
    assert client.get(f"{B}{path}").status_code == 503


def test_curated_event_renders(client):
    b = client.get(f"{B}/climate/new-orleans-louisiana-us").text
    assert "Notable weather in New Orleans" in b
    assert "Hurricane Katrina" in b


def test_uncurated_city_has_no_event_section(client):
    b = client.get(f"{B}/climate/shanghai-cn").text
    assert "city-event" not in b


def test_city_travel_cta_prefills_compare(client):
    b = client.get(f"{B}/climate/{SLUG}").text
    assert "Thinking of visiting" in b
    assert "/compare#loc=" in b


def test_month_and_records(client):
    assert client.get(f"{B}/climate/{SLUG}/july").status_code == 200
    assert client.get(f"{B}/climate/{SLUG}/records").status_code == 200
    assert client.get(f"{B}/climate/{SLUG}/notamonth").status_code == 404


def test_month_records_cover_every_metric_high_and_low(client):
    b = client.get(f"{B}/climate/{SLUG}/july").text
    assert "July records" in b
    for label in ("High", "Low", "Precip"):
        assert f'class="rc-metric">{label}<' in b
    n = b.count('rc-row rc-high')
    assert n >= 3 and b.count('rc-row rc-low') == n
    assert "Highest" in b and "Lowest" in b
    assert "Wettest" in b and "Driest" in b


def test_records_page_has_monthly_and_seasonal(client):
    b = client.get(f"{B}/climate/{SLUG}/records").text
    assert "Records by month" in b
    for month in ("January", "July", "December"):
        assert month in b
    assert f"/climate/{SLUG}/january" in b
    assert "Records by season" in b
    assert "Northern Hemisphere" in b
    assert 'Winter <span class="season-span">(Dec–Feb)</span>' in b
    assert 'Summer <span class="season-span">(Jun–Aug)</span>' in b
    assert "All-time records" in b
    assert '"@type":"Dataset"' in b
    assert re.search(r"-?\d+°C</span>", b)
    assert "monthly" in b.lower() and "seasonal" in b.lower()


def test_records_seasons_flip_in_southern_hemisphere(client):
    b = client.get(f"{B}/climate/sao-paulo-br/records").text
    assert "Southern Hemisphere" in b
    assert 'Summer <span class="season-span">(Dec–Feb)</span>' in b
    assert 'Winter <span class="season-span">(Jun–Aug)</span>' in b


def test_climate_pages_are_colour_coded(client):
    city = client.get(f"{B}/climate/{SLUG}").text
    assert "t-cell t-" in city
    assert "range-fill" in city and "--c1:var(--" in city
    recs = client.get(f"{B}/climate/{SLUG}/records").text
    assert "t-inline t-" in recs
    month = client.get(f"{B}/climate/{SLUG}/july").text
    assert "t-inline t-" in month


def test_records_show_both_extremes_per_metric(client):
    b = client.get(f"{B}/climate/{SLUG}/records").text
    assert "Daytime high" in b and "Overnight low" in b
    assert "▲" in b and "▼" in b
    assert b.count('class="rec-ext"') >= 48


def test_hub_glossary_about(client):
    assert client.get(f"{B}/climate").status_code == 200
    assert client.get(f"{B}/glossary").status_code == 200
    assert client.get(f"{B}/glossary/percentile").status_code == 200
    assert client.get(f"{B}/glossary/not-a-term").status_code == 404
    assert client.get(f"{B}/about").status_code == 200


def test_static_page_titles_come_from_content_yaml(client):
    from markupsafe import escape
    for path, key in ((f"{B}/climate", "hub"), (f"{B}/glossary", "glossary_index"),
                      (f"{B}/about", "about"), (f"{B}/privacy", "privacy")):
        html = client.get(path).text
        assert f"<title>{escape(content.PAGES[key]['title'])}</title>" in html, path


def test_glossary_terms_come_from_content_yaml(client):
    html = client.get(f"{B}/glossary").text
    for entry in content.GLOSSARY.values():
        assert entry["term"] in html
    term_html = client.get(f"{B}/glossary/percentile").text
    assert content.GLOSSARY["percentile"]["short"] in term_html


def test_footer_has_no_leaked_jinja_comment(client):
    for path in (f"{B}/about", f"{B}/climate/{SLUG}", f"{B}/climate/{SLUG}/records"):
        b = client.get(path).text
        footer = re.search(r"<footer.*?</footer>", b, re.S)
        assert footer, f"no footer on {path}"
        assert "#}" not in footer.group(0) and "{#" not in footer.group(0), path


def test_page_titles_lead_with_city_and_payload(client):
    for path, must_start, payload in (
        (f"{B}/climate/{SLUG}", "London, GB climate", "how unusual it is now"),
        (f"{B}/climate/{SLUG}/july", "London, GB in July", "normal weather"),
        (f"{B}/climate/{SLUG}/records", "London, GB weather records", "hottest & coldest"),
    ):
        b = client.get(path).text
        title = re.search(r"<title>(.*?)</title>", b, re.S).group(1)
        title = title.replace("&amp;", "&")
        assert title.startswith(must_start), title
        assert payload in title, title
        assert "England, United Kingdom" not in title
        assert len(must_start) <= 40
        og = re.search(r'<meta property="og:title" content="(.*?)"', b, re.S).group(1)
        assert og == re.search(r"<title>(.*?)</title>", b, re.S).group(1)


def test_meta_descriptions_lead_with_real_numbers(client):
    for path in (f"{B}/climate/{SLUG}", f"{B}/climate/{SLUG}/july",
                 f"{B}/climate/{SLUG}/records"):
        b = client.get(path).text
        desc = re.search(r'<meta name="description" content="(.*?)"', b, re.S).group(1)
        assert len(desc) <= 155, f"{len(desc)}: {desc}"
        assert re.search(r"-?\d+°C", desc), desc
        assert desc.endswith("years of local history."), desc
    rec = client.get(f"{B}/climate/{SLUG}/records").text
    desc = re.search(r'<meta name="description" content="(.*?)"', rec, re.S).group(1)
    assert re.search(r"\([A-Z][a-z]{2} \d{4}\)", desc), desc


_TEMP_SPAN = re.compile(r'<span class="temp" data-temp-f="(-?[\d.]+)"([^>]*)>(-?\d+)°([CF]?)</span>')


def _temp_spans(html):
    return [(float(f), int(n), letter) for f, _attrs, n, letter in _TEMP_SPAN.findall(html)]


def test_climate_pages_carry_convertible_temps(client):
    for path in (f"{B}/climate/{SLUG}", f"{B}/climate/{SLUG}/july", f"{B}/climate/{SLUG}/records"):
        b = client.get(path).text
        spans = _temp_spans(b)
        assert spans, f"no temperature spans on {path}"
        assert re.search(r"\d+°F \(-?\d+°C\)", b) is None


def test_units_default_to_the_citys_country(client):
    gb = client.get(f"{B}/climate/{SLUG}").text
    us = client.get(f"{B}/climate/seattle-washington-us").text

    gb_spans, us_spans = _temp_spans(gb), _temp_spans(us)
    assert gb_spans and us_spans
    assert {letter for _f, _n, letter in gb_spans if letter} == {"C"}
    assert {letter for _f, _n, letter in us_spans if letter} == {"F"}

    assert any(math.floor(f + 0.5) != n for f, n, _l in gb_spans), "data-temp-f was converted"
    assert all(math.floor(f + 0.5) == n for f, n, _l in us_spans)

    assert 'data-unit-default="C"' in gb
    assert 'data-unit-default="F"' in us


_PRECIP_SPAN = re.compile(r'<span class="precip" data-precip-in="([\d.]+)">([\d.]+) (in|mm)</span>')


def test_precip_and_wind_follow_the_citys_unit(client):
    gb = client.get(f"{B}/climate/{SLUG}/records").text
    us = client.get(f"{B}/climate/seattle-washington-us/records").text

    gb_p, us_p = _PRECIP_SPAN.findall(gb), _PRECIP_SPAN.findall(us)
    assert gb_p and us_p

    assert {u for _raw, _shown, u in gb_p} == {"mm"}
    assert {u for _raw, _shown, u in us_p} == {"in"}

    for raw, shown, _u in gb_p:
        assert abs(float(shown) - float(raw) * 25.4) < 0.51, (raw, shown)
    for raw, shown, _u in us_p:
        assert abs(float(shown) - float(raw)) < 0.005, (raw, shown)


def test_wind_and_humidity_render_per_unit_system():
    # The synthetic history carries only tmax/tmin/precip, so wind and
    # humidity never reach a rendered page in these tests -- exercise
    # format.py's renderers directly, same as the backend's equivalent test.
    with fmt.unit_scope("F"):
        assert fmt.wind_text(10) == "10 mph"
        assert '<span class="wind" data-wind-mph="10.0">10 mph</span>' == fmt.wind(10)
        assert fmt.fmt("humid", 7.75) == "7.8 g/m³"
    with fmt.unit_scope("C"):
        assert fmt.wind_text(10) == "16 km/h"
        assert 'data-wind-mph="10.0"' in fmt.wind(10)
        assert fmt.fmt("humid", 7.75) == "7.8 g/m³"


def test_precip_precision_differs_by_unit():
    with fmt.unit_scope("F"):
        assert fmt.precip_text(0.04) == "0.04 in"
    with fmt.unit_scope("C"):
        assert fmt.precip_text(0.04) == "1 mm"
        assert fmt.precip_text(1.81) == "46 mm"


def test_measure_conversion_constants_match_the_frontend():
    units_js = os.path.join(os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))),
                            "frontend", "units.js")
    with open(units_js, encoding="utf-8") as f:
        src = f.read()
    mm = float(re.search(r"MM_PER_IN\s*=\s*([\d.]+)", src).group(1))
    kmh = float(re.search(r"KMH_PER_MPH\s*=\s*([\d.]+)", src).group(1))
    assert mm == fmt.MM_PER_IN
    assert kmh == fmt.KMH_PER_MPH


def test_unit_default_absent_without_a_city(client):
    for path in (f"{B}/", f"{B}/climate", f"{B}/about", f"{B}/glossary"):
        assert "data-unit-default" not in client.get(path).text, path


def test_unit_does_not_leak_between_requests(client):
    assert 'data-unit-default="C"' in client.get(f"{B}/climate/{SLUG}").text
    about = client.get(f"{B}/about").text
    assert "data-unit-default" not in about
    assert {l for _f, _n, l in _temp_spans(about) if l} == {"F"}

    us = client.get(f"{B}/climate/seattle-washington-us").text
    assert {l for _f, _n, l in _temp_spans(us) if l} == {"F"}
    gb_again = client.get(f"{B}/climate/{SLUG}").text
    assert {l for _f, _n, l in _temp_spans(gb_again) if l} == {"C"}


def test_f_country_list_matches_the_frontend():
    units_js = os.path.join(os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))),
                            "frontend", "units.js")
    with open(units_js, encoding="utf-8") as f:
        src = f.read()
    literal = re.search(r"F_REGIONS = new Set\(\[(.*?)\]\)", src, re.S).group(1)
    assert set(re.findall(r'"([A-Z]{2})"', literal)) == set(fmt.F_COUNTRIES)


def test_celsius_rounding_matches_js():
    assert fmt._round_half_up(0.5) == 1
    assert fmt._round_half_up(1.5) == 2
    assert fmt._round_half_up(2.5) == 3
    assert fmt._round_half_up(-0.5) == 0
    assert fmt._c(31.1) == 0


def test_pct_ordinal_matches_backend_grading():
    # format.pct_ordinal is a ported copy of backend/data/grading.py's
    # pct_ordinal (the frontend can no longer import backend code) -- pin the
    # two against the same cases so a future edit to either can't drift silently.
    assert fmt.pct_ordinal(66.4) == "66th"
    assert fmt.pct_ordinal(1.2) == "1st"
    assert fmt.pct_ordinal(22) == "22nd"
    assert fmt.pct_ordinal(3) == "3rd"
    assert fmt.pct_ordinal(11) == "11th"
    assert fmt.pct_ordinal(99.6) == "99th"
    assert fmt.pct_ordinal(100) == "99th"
    assert fmt.pct_ordinal(0) == "1st"
    assert fmt.pct_ordinal(None) == "—"


def test_city_page_percentiles_are_real_ordinals(client):
    html = client.get(f"{B}/climate/{SLUG}").text
    assert "th pct" in html or "st pct" in html or "nd pct" in html or "rd pct" in html
    for bad in re.findall(r"\d+\.\d+(?:st|nd|rd|th)|100th|\b0th|\b\d*[123]th", html):
        raise AssertionError(f"malformed percentile ordinal: {bad!r}")


def test_city_page_etag_is_stable(client):
    r1 = client.get(f"{B}/climate/{SLUG}")
    r2 = client.get(f"{B}/climate/{SLUG}")
    assert r1.headers["etag"] == r2.headers["etag"]
    r3 = client.get(f"{B}/climate/{SLUG}", headers={"If-None-Match": r1.headers["etag"]})
    assert r3.status_code == 304


def test_seo_pages_load_unit_and_account_scripts(client):
    b = client.get(f"{B}/climate/{SLUG}").text
    for src in (f"{B}/units.js", f"{B}/account.js", f"{B}/climate.js"):
        assert f'src="{src}"' in b


def test_indexnow_key_file_served(client):
    from conftest import FAKE_INDEXNOW_KEY
    r = client.get(f"{B}/{FAKE_INDEXNOW_KEY}.txt")
    assert r.status_code == 200
    assert r.text.strip() == FAKE_INDEXNOW_KEY
    assert r.headers["content-type"].startswith("text/plain")


def test_sitemap_lastmod_is_stable(client):
    import datetime
    body = client.get(f"{B}/sitemap.xml").text
    mods = set(re.findall(r"<lastmod>([^<]+)</lastmod>", body))
    assert len(mods) == 1
    datetime.date.fromisoformat(next(iter(mods)))


def test_search_verification_meta(client, monkeypatch):
    monkeypatch.setenv("THERMOGRAPH_GOOGLE_VERIFY", "gtok123")
    monkeypatch.setenv("THERMOGRAPH_BING_VERIFY", "btok456")
    seo = client.get(f"{B}/climate/{SLUG}").text
    home = client.get(f"{B}/").text
    for b in (seo, home):
        assert '<meta name="google-site-verification" content="gtok123">' in b
        assert '<meta name="msvalidate.01" content="btok456">' in b


BRAND_PAGES = [B + p for p in ("/", "/climate", "/glossary", "/about", "/privacy")]


@pytest.mark.parametrize("path", BRAND_PAGES)
def test_brand_lockup_links_home(client, path):
    html = client.get(path).text
    brand = html.split('<div class="brand">', 1)[1].split("</header>", 1)[0]
    head = brand.split("</h1>")[0] if "<h1" in brand else brand.split("</p>")[0]
    assert "<a href=" in head, f"{path}: brand is not a link"
    href = head.split('<a href="', 1)[1].split('"', 1)[0]
    assert href in ("./", f"{B}/"), f"{path}: brand links to {href!r}, not home"
    assert '<span class="logo">' in head.split("<a href=", 1)[1], \
        f"{path}: the mark sits outside the home link"


def test_form_is_hidden_for_now(client):
    """The footer digest signup is temporarily hidden (commented out in the
    base template). The /digest endpoint (backend-owned) stays live; only the
    UI is pulled. Restore the form and flip this back to asserting presence.
    Ported from backend/tests/notifications/test_digest.py (repo-split Stage 4)
    -- every page it checks is frontend-owned now."""
    for path in ["/", "/about", "/climate", "/privacy", "/glossary"]:
        html = client.get(f"{B}{path}").text
        assert "data-digest" not in html, path