Validate generated conference fields

agenda/add_new_conference.py

@@ -80,6 +80,7 @@ Common optional fields:
   or loosely translate an address. If no reliable rendering is available, omit
   `address`; coordinates are preferable to an unreadable or inaccurate address.
 - `free`, `price`, `currency`, `hackathon`, `online`, `attendees`.
+- When `free: true`, omit `price` and `currency` because they are redundant.
 - Do not include `going`, `registered`, `accommodation_booked`,
   `transport_booked`, or `trip` unless the source explicitly says they apply to
   my attendance.
@@ -528,6 +529,27 @@ def validate_country(conf: dict[str, typing.Any]) -> None:
     conf["country"] = match.alpha_2.lower()
 
 
+def validate_series(
+    conf: dict[str, typing.Any], series: dict[str, ConferenceSeries]
+) -> None:
+    """Ensure a generated series ID exists in conference_series.yaml."""
+    series_id = conf.get("series")
+    if series_id is None:
+        return
+    if not isinstance(series_id, str) or series_id not in series:
+        raise ValueError(
+            f"Generated conference uses unknown series {series_id!r}. "
+            "Add it to conference_series.yaml first or remove the series field."
+        )
+
+
+def normalize_free_event_fields(conf: dict[str, typing.Any]) -> None:
+    """Remove redundant pricing fields from free conferences."""
+    if conf.get("free") is True:
+        conf.pop("price", None)
+        conf.pop("currency", None)
+
+
 def parse_yaml_datetime(value: typing.Any) -> datetime | None:
     """Convert YAML date/datetime values to a datetime."""
     if isinstance(value, datetime):
@@ -749,6 +771,8 @@ def add_new_conference(url: str, yaml_path: str) -> bool:
     assert isinstance(new_conf, dict)
 
     validate_country(new_conf)
+    validate_series(new_conf, series)
+    normalize_free_event_fields(new_conf)
     apply_event_metadata_dates(new_conf, soup)
     normalize_dates_field(new_conf)
     normalise_end_field(new_conf, source_text)

tests/test_add_new_conference.py

@@ -3,11 +3,12 @@
 from datetime import date, datetime
 import typing
 
-import lxml.html  # type: ignore[import-untyped]
+import lxml.html
 import pytest
 import yaml
 
 from agenda import add_new_conference
+from agenda.conference import ConferenceSeries
 
 
 def test_parse_osm_url_mlat_mlon() -> None:
@@ -231,6 +232,7 @@ def test_build_prompt_includes_nested_dates_and_series() -> None:
     assert "March 2027" in prompt
     assert "For an address written in a non-Latin script" in prompt
     assert "If no reliable rendering is available, omit" in prompt
+    assert "When `free: true`, omit `price` and `currency`" in prompt
 
 
 def test_validate_country_normalises_name() -> None:
@@ -242,6 +244,53 @@ def test_validate_country_normalises_name() -> None:
     assert conf["country"] == "gb"
 
 
+def test_validate_series_rejects_unknown_generated_id() -> None:
+    """Generated series IDs must exist in conference_series.yaml."""
+    conf: dict[str, typing.Any] = {"series": "state-of-the-map-asia"}
+
+    with pytest.raises(ValueError, match="unknown series 'state-of-the-map-asia'"):
+        add_new_conference.validate_series(
+            conf,
+            {"state-of-the-map": {"name": "State of the Map"}},
+        )
+
+
+def test_validate_series_accepts_known_id_or_missing_field() -> None:
+    """Known series IDs and conferences without a series should pass."""
+    series: dict[str, ConferenceSeries] = {
+        "state-of-the-map": {"name": "State of the Map"}
+    }
+
+    add_new_conference.validate_series({"series": "state-of-the-map"}, series)
+    add_new_conference.validate_series({}, series)
+
+
+def test_normalize_free_event_fields_removes_price_and_currency() -> None:
+    """Free events should not retain redundant pricing fields."""
+    conf: dict[str, typing.Any] = {
+        "free": True,
+        "price": 0,
+        "currency": "JPY",
+    }
+
+    add_new_conference.normalize_free_event_fields(conf)
+
+    assert conf == {"free": True}
+
+
+def test_normalize_free_event_fields_keeps_paid_event_pricing() -> None:
+    """Pricing fields should remain for events that are not marked free."""
+    conf: dict[str, typing.Any] = {
+        "free": False,
+        "price": 25,
+        "currency": "GBP",
+    }
+
+    add_new_conference.normalize_free_event_fields(conf)
+
+    assert conf == {"free": False, "price": 25, "currency": "GBP"}
+
+
 def test_normalise_end_field_defaults_single_day_date() -> None:
     """Non-Geomob conferences should default end to the start date."""
     conf: dict[str, typing.Any] = {
@@ -370,6 +419,12 @@ def test_add_new_conference_reuses_generic_url_for_new_year(
 ) -> None:
     """Generic URLs with digits in the domain should not be skipped early."""
     yaml_path = tmp_path / "conferences.yaml"
+    (tmp_path / "conference_series.yaml").write_text(
+        yaml.dump(
+            {"foss4g-north-america": {"name": "FOSS4G North America"}},
+            sort_keys=False,
+        )
+    )
     yaml_path.write_text(
         yaml.dump(
             [