Add boss seed data pipeline for export and import

Add seeder support for boss battles so new database instances come
pre-populated. Adds --export-bosses CLI flag to dump boss data from the
database to JSON seed files, and loads those files during normal seeding.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
2026-02-08 12:36:08 +01:00
parent 76855f4f56
commit 053dece33e
7 changed files with 250 additions and 27 deletions

View File

@@ -1,20 +1,26 @@
"""Entry point for running seeds.
Usage:
python -m app.seeds # Run seed
python -m app.seeds --verify # Run seed + verification
python -m app.seeds # Run seed
python -m app.seeds --verify # Run seed + verification
python -m app.seeds --export-bosses # Export boss data to seed JSON files
"""
import asyncio
import sys
from app.core.database import engine
from app.seeds.run import seed, verify
from app.seeds.run import export_bosses, seed, verify
async def main():
verbose = "--verbose" in sys.argv or "-v" in sys.argv
engine.echo = verbose
if "--export-bosses" in sys.argv:
await export_bosses()
return
await seed()
if "--verify" in sys.argv:
await verify()

View File

@@ -1,9 +1,11 @@
"""Database upsert helpers for seed data."""
from sqlalchemy import select
from sqlalchemy import delete, select
from sqlalchemy.dialects.postgresql import insert
from sqlalchemy.ext.asyncio import AsyncSession
from app.models.boss_battle import BossBattle
from app.models.boss_pokemon import BossPokemon
from app.models.evolution import Evolution
from app.models.game import Game
from app.models.pokemon import Pokemon
@@ -204,14 +206,69 @@ async def upsert_route_encounters(
return count
async def upsert_bosses(
session: AsyncSession,
version_group_id: int,
bosses: list[dict],
dex_to_id: dict[int, int],
) -> int:
"""Upsert boss battles for a version group, return count of bosses upserted."""
count = 0
for boss in bosses:
# Upsert the boss battle on (version_group_id, order) conflict
stmt = insert(BossBattle).values(
version_group_id=version_group_id,
name=boss["name"],
boss_type=boss["boss_type"],
badge_name=boss.get("badge_name"),
badge_image_url=boss.get("badge_image_url"),
level_cap=boss["level_cap"],
order=boss["order"],
location=boss["location"],
sprite_url=boss.get("sprite_url"),
).on_conflict_do_update(
constraint="uq_boss_battles_version_group_order",
set_={
"name": boss["name"],
"boss_type": boss["boss_type"],
"badge_name": boss.get("badge_name"),
"badge_image_url": boss.get("badge_image_url"),
"level_cap": boss["level_cap"],
"location": boss["location"],
"sprite_url": boss.get("sprite_url"),
},
).returning(BossBattle.id)
result = await session.execute(stmt)
boss_id = result.scalar_one()
# Delete existing boss_pokemon for this boss, then re-insert
await session.execute(
delete(BossPokemon).where(BossPokemon.boss_battle_id == boss_id)
)
for bp in boss.get("pokemon", []):
pokemon_id = dex_to_id.get(bp["pokeapi_id"])
if pokemon_id is None:
print(f" Warning: no pokemon_id for pokeapi_id {bp['pokeapi_id']}")
continue
session.add(BossPokemon(
boss_battle_id=boss_id,
pokemon_id=pokemon_id,
level=bp["level"],
order=bp["order"],
))
count += 1
await session.flush()
return count
async def upsert_evolutions(
session: AsyncSession,
evolutions: list[dict],
dex_to_id: dict[int, int],
) -> int:
"""Upsert evolution pairs, return count of upserted rows."""
# Clear existing evolutions and re-insert (simpler than complex upsert)
from sqlalchemy import delete
await session.execute(delete(Evolution))
count = 0

View File

@@ -5,14 +5,18 @@ from pathlib import Path
from sqlalchemy import func, select
from sqlalchemy.ext.asyncio import AsyncSession
from sqlalchemy.orm import selectinload
from app.core.database import async_session
from app.models.boss_battle import BossBattle
from app.models.boss_pokemon import BossPokemon
from app.models.game import Game
from app.models.pokemon import Pokemon
from app.models.route import Route
from app.models.route_encounter import RouteEncounter
from app.models.version_group import VersionGroup
from app.seeds.loader import (
upsert_bosses,
upsert_evolutions,
upsert_games,
upsert_pokemon,
@@ -129,7 +133,26 @@ async def seed():
print(f"\nTotal routes: {total_routes}")
print(f"Total encounters: {total_encounters}")
# 5. Upsert evolutions
# 5. Per version group: upsert bosses
total_bosses = 0
for vg_slug, vg_info in vg_data.items():
vg_id = vg_slug_to_id[vg_slug]
first_game_slug = list(vg_info["games"].keys())[0]
bosses_file = DATA_DIR / f"{first_game_slug}-bosses.json"
if not bosses_file.exists():
continue
bosses_data = load_json(f"{first_game_slug}-bosses.json")
if not bosses_data:
continue
boss_count = await upsert_bosses(session, vg_id, bosses_data, dex_to_id)
total_bosses += boss_count
print(f" {vg_slug}: {boss_count} bosses")
print(f"Total bosses: {total_bosses}")
# 6. Upsert evolutions
evolutions_path = DATA_DIR / "evolutions.json"
if evolutions_path.exists():
evolutions_data = load_json("evolutions.json")
@@ -152,12 +175,14 @@ async def verify():
pokemon_count = (await session.execute(select(func.count(Pokemon.id)))).scalar()
routes_count = (await session.execute(select(func.count(Route.id)))).scalar()
enc_count = (await session.execute(select(func.count(RouteEncounter.id)))).scalar()
boss_count = (await session.execute(select(func.count(BossBattle.id)))).scalar()
print(f"Version Groups: {vg_count}")
print(f"Games: {games_count}")
print(f"Pokemon: {pokemon_count}")
print(f"Routes: {routes_count}")
print(f"Route Encounters: {enc_count}")
print(f"Boss Battles: {boss_count}")
# Per-version-group route counts
result = await session.execute(
@@ -181,4 +206,87 @@ async def verify():
for row in result:
print(f" {row[0]}: {row[1]}")
# Per-version-group boss counts
result = await session.execute(
select(VersionGroup.slug, func.count(BossBattle.id))
.join(BossBattle, BossBattle.version_group_id == VersionGroup.id)
.group_by(VersionGroup.slug)
.order_by(VersionGroup.slug)
)
print("\nBosses per version group:")
for row in result:
print(f" {row[0]}: {row[1]}")
print("\nVerification complete!")
async def export_bosses():
"""Export boss battles from the database to seed JSON files."""
print("Exporting boss battles...")
async with async_session() as session:
# Load version group data to get the first game slug for filenames
with open(VG_JSON) as f:
vg_data = json.load(f)
# Query all version groups with their games
vg_result = await session.execute(
select(VersionGroup).options(selectinload(VersionGroup.games))
)
version_groups = vg_result.scalars().all()
slug_to_vg = {vg.slug: vg for vg in version_groups}
exported = 0
for vg_slug, vg_info in vg_data.items():
vg = slug_to_vg.get(vg_slug)
if vg is None:
continue
# Query boss battles for this version group
result = await session.execute(
select(BossBattle)
.where(BossBattle.version_group_id == vg.id)
.options(
selectinload(BossBattle.pokemon).selectinload(BossPokemon.pokemon),
)
.order_by(BossBattle.order)
)
bosses = result.scalars().all()
if not bosses:
continue
first_game_slug = list(vg_info["games"].keys())[0]
data = [
{
"name": b.name,
"boss_type": b.boss_type,
"badge_name": b.badge_name,
"badge_image_url": b.badge_image_url,
"level_cap": b.level_cap,
"order": b.order,
"location": b.location,
"sprite_url": b.sprite_url,
"pokemon": [
{
"pokeapi_id": bp.pokemon.pokeapi_id,
"pokemon_name": bp.pokemon.name,
"level": bp.level,
"order": bp.order,
}
for bp in sorted(b.pokemon, key=lambda p: p.order)
],
}
for b in bosses
]
out_path = DATA_DIR / f"{first_game_slug}-bosses.json"
with open(out_path, "w") as f:
json.dump(data, f, indent=2)
f.write("\n")
print(f" {vg_slug}: {len(bosses)} bosses -> {out_path.name}")
exported += 1
print(f"Exported bosses for {exported} version groups.")