feat(v2): add LBA snapshot extractor and command
This commit is contained in:
137
apps/ingestion/extractors/lba.py
Normal file
137
apps/ingestion/extractors/lba.py
Normal file
@ -0,0 +1,137 @@
|
||||
from __future__ import annotations
|
||||
|
||||
from typing import Any
|
||||
|
||||
from django.conf import settings
|
||||
|
||||
from .base import BaseSnapshotExtractor, ExtractorConfigError, ExtractorNormalizationError, ExtractorParseError
|
||||
from .http import ResponsibleHttpClient
|
||||
|
||||
|
||||
def _first_non_empty(record: dict[str, Any], *keys: str) -> Any:
|
||||
for key in keys:
|
||||
value = record.get(key)
|
||||
if value not in (None, ""):
|
||||
return value
|
||||
return None
|
||||
|
||||
|
||||
class LBASnapshotExtractor(BaseSnapshotExtractor):
|
||||
"""
|
||||
LBA (Lega Basket Serie A) MVP extractor.
|
||||
|
||||
Scope is intentionally conservative:
|
||||
- one configured public stats endpoint
|
||||
- one configured season label
|
||||
- normalized player-season rows only
|
||||
"""
|
||||
|
||||
extractor_name = "lba"
|
||||
source_name = "lba"
|
||||
|
||||
def __init__(self, *, http_client: ResponsibleHttpClient | None = None):
|
||||
self.url = settings.EXTRACTOR_LBA_STATS_URL.strip()
|
||||
self.season_label = settings.EXTRACTOR_LBA_SEASON_LABEL.strip()
|
||||
self.competition_external_id = settings.EXTRACTOR_LBA_COMPETITION_EXTERNAL_ID.strip()
|
||||
self.competition_name = settings.EXTRACTOR_LBA_COMPETITION_NAME.strip()
|
||||
self.include_raw_payload = settings.EXTRACTOR_INCLUDE_RAW_PAYLOAD
|
||||
self.http_client = http_client or ResponsibleHttpClient(
|
||||
user_agent=settings.EXTRACTOR_USER_AGENT,
|
||||
timeout_seconds=settings.EXTRACTOR_HTTP_TIMEOUT_SECONDS,
|
||||
retries=settings.EXTRACTOR_HTTP_RETRIES,
|
||||
retry_sleep_seconds=settings.EXTRACTOR_RETRY_SLEEP_SECONDS,
|
||||
request_delay_seconds=settings.EXTRACTOR_REQUEST_DELAY_SECONDS,
|
||||
)
|
||||
|
||||
if not self.url:
|
||||
raise ExtractorConfigError("EXTRACTOR_LBA_STATS_URL is required for lba extractor.")
|
||||
if not self.season_label:
|
||||
raise ExtractorConfigError("EXTRACTOR_LBA_SEASON_LABEL is required for lba extractor.")
|
||||
if not self.competition_external_id:
|
||||
raise ExtractorConfigError("EXTRACTOR_LBA_COMPETITION_EXTERNAL_ID is required.")
|
||||
if not self.competition_name:
|
||||
raise ExtractorConfigError("EXTRACTOR_LBA_COMPETITION_NAME is required.")
|
||||
|
||||
def fetch(self) -> Any:
|
||||
return self.http_client.get_json(self.url)
|
||||
|
||||
def parse(self, payload: Any) -> list[dict[str, Any]]:
|
||||
if isinstance(payload, list):
|
||||
return payload
|
||||
if not isinstance(payload, dict):
|
||||
raise ExtractorParseError("LBA payload must be a JSON object or array.")
|
||||
|
||||
for key in ("records", "data", "players", "items"):
|
||||
rows = payload.get(key)
|
||||
if isinstance(rows, list):
|
||||
return rows
|
||||
|
||||
raise ExtractorParseError("LBA payload must contain one of: records, data, players, items.")
|
||||
|
||||
def normalize_record(self, source_record: dict[str, Any]) -> dict[str, Any]:
|
||||
player_obj = source_record.get("player") if isinstance(source_record.get("player"), dict) else {}
|
||||
team_obj = source_record.get("team") if isinstance(source_record.get("team"), dict) else {}
|
||||
|
||||
full_name = _first_non_empty(
|
||||
source_record,
|
||||
"full_name",
|
||||
"player_name",
|
||||
"name",
|
||||
) or _first_non_empty(player_obj, "full_name", "name")
|
||||
first_name = _first_non_empty(source_record, "first_name") or _first_non_empty(player_obj, "first_name")
|
||||
last_name = _first_non_empty(source_record, "last_name") or _first_non_empty(player_obj, "last_name")
|
||||
player_external_id = _first_non_empty(
|
||||
source_record, "player_external_id", "player_id", "athlete_id"
|
||||
) or _first_non_empty(player_obj, "id", "player_id")
|
||||
team_external_id = _first_non_empty(source_record, "team_external_id", "team_id") or _first_non_empty(
|
||||
team_obj, "id", "team_id"
|
||||
)
|
||||
team_name = _first_non_empty(source_record, "team_name", "team") or _first_non_empty(team_obj, "name")
|
||||
|
||||
normalized = {
|
||||
"competition_external_id": self.competition_external_id,
|
||||
"competition_name": self.competition_name,
|
||||
"season": self.season_label,
|
||||
"team_external_id": team_external_id,
|
||||
"team_name": team_name,
|
||||
"player_external_id": player_external_id,
|
||||
"full_name": full_name,
|
||||
"first_name": first_name,
|
||||
"last_name": last_name,
|
||||
"birth_date": _first_non_empty(source_record, "birth_date") or _first_non_empty(
|
||||
player_obj, "birth_date", "dob"
|
||||
),
|
||||
"nationality": _first_non_empty(source_record, "nationality")
|
||||
or _first_non_empty(player_obj, "nationality", "country"),
|
||||
"height_cm": _first_non_empty(source_record, "height_cm") or _first_non_empty(player_obj, "height_cm"),
|
||||
"weight_kg": _first_non_empty(source_record, "weight_kg") or _first_non_empty(player_obj, "weight_kg"),
|
||||
"position": _first_non_empty(source_record, "position") or _first_non_empty(player_obj, "position"),
|
||||
"role": _first_non_empty(source_record, "role"),
|
||||
"games_played": _first_non_empty(source_record, "games_played", "gp"),
|
||||
"minutes_per_game": _first_non_empty(source_record, "minutes_per_game", "mpg"),
|
||||
"points_per_game": _first_non_empty(source_record, "points_per_game", "ppg"),
|
||||
"rebounds_per_game": _first_non_empty(source_record, "rebounds_per_game", "rpg"),
|
||||
"assists_per_game": _first_non_empty(source_record, "assists_per_game", "apg"),
|
||||
"steals_per_game": _first_non_empty(source_record, "steals_per_game", "spg"),
|
||||
"blocks_per_game": _first_non_empty(source_record, "blocks_per_game", "bpg"),
|
||||
"turnovers_per_game": _first_non_empty(source_record, "turnovers_per_game", "tov"),
|
||||
"fg_pct": _first_non_empty(source_record, "fg_pct", "fg_percentage"),
|
||||
"three_pt_pct": _first_non_empty(
|
||||
source_record, "three_pt_pct", "three_point_pct", "3p_pct", "three_pct"
|
||||
),
|
||||
"ft_pct": _first_non_empty(source_record, "ft_pct", "ft_percentage"),
|
||||
}
|
||||
|
||||
missing = [key for key, value in normalized.items() if key != "role" and value in (None, "")]
|
||||
if missing:
|
||||
raise ExtractorNormalizationError(f"lba row missing required fields: {', '.join(sorted(missing))}")
|
||||
|
||||
normalized["team_external_id"] = str(normalized["team_external_id"]).strip()
|
||||
normalized["player_external_id"] = str(normalized["player_external_id"]).strip()
|
||||
normalized["competition_external_id"] = str(normalized["competition_external_id"]).strip()
|
||||
normalized["season"] = str(normalized["season"]).strip()
|
||||
|
||||
if self.include_raw_payload:
|
||||
normalized["raw_payload"] = source_record
|
||||
|
||||
return normalized
|
||||
Reference in New Issue
Block a user