WildDex/dev_scripts/database_setup.py

import csv
import sqlite3
from pathlib import Path

DB_PATH = Path(__file__).parent.parent / "wilddex.db"
WILDLIFE_CSV = Path(__file__).parent.parent / "wildlife" / "portland_wildlife.csv"
DEX_ENTRIES_CSV = Path(__file__).parent.parent / "wildlife" / "portland_wildlife_dex_entries.csv"


def create_tables(conn: sqlite3.Connection) -> None:
    conn.executescript("""
        CREATE TABLE IF NOT EXISTS species (
            species_id          INTEGER PRIMARY KEY AUTOINCREMENT,
            inat_taxon_id       INTEGER UNIQUE NOT NULL,
            common_name         TEXT NOT NULL,
            scientific_name     TEXT NOT NULL,
            kingdom             TEXT,
            phylum              TEXT,
            class               TEXT,
            "order"             TEXT,
            family              TEXT,
            genus               TEXT,
            iconic_taxon        TEXT,
            conservation_status TEXT,
            pokedex_entry       TEXT,
            photo_url_thumb     TEXT,
            photo_url_med       TEXT,
            lat_min             REAL,
            lat_max             REAL,
            lng_min             REAL,
            lng_max             REAL,
            is_unlocked         INTEGER NOT NULL DEFAULT 0
        );

        CREATE TABLE IF NOT EXISTS species_detail (
            detail_id           INTEGER PRIMARY KEY AUTOINCREMENT,
            species_id          INTEGER NOT NULL REFERENCES species(species_id),
            size_range          TEXT,
            weight_range        TEXT,
            lifespan            TEXT,
            diet                TEXT,
            activity_period     TEXT,
            habitat_type        TEXT,
            behavior_notes      TEXT
        );

        CREATE TABLE IF NOT EXISTS sighting (
            sighting_id         INTEGER PRIMARY KEY AUTOINCREMENT,
            species_id          INTEGER NOT NULL REFERENCES species(species_id),
            observed_at         TEXT,
            latitude            REAL,
            longitude           REAL,
            location_label      TEXT,
            user_photo_path     TEXT,
            notes               TEXT
        );

        CREATE TABLE IF NOT EXISTS lookalike (
            id                      INTEGER PRIMARY KEY AUTOINCREMENT,
            species_id              INTEGER NOT NULL REFERENCES species(species_id),
            lookalike_species_id    INTEGER NOT NULL REFERENCES species(species_id),
            distinguishing_tip      TEXT
        );
    """)
    conn.commit()


def load_dex_entries(csv_path: Path) -> dict[int, str]:
    """Return a mapping of taxon_id -> pokedex description."""
    entries = {}
    with open(csv_path, newline="", encoding="utf-8") as f:
        for row in csv.DictReader(f):
            entries[int(row["taxon_id"])] = row["description"]
    return entries


def import_species(conn: sqlite3.Connection) -> None:
    dex_entries = load_dex_entries(DEX_ENTRIES_CSV)

    with open(WILDLIFE_CSV, newline="", encoding="utf-8") as f:
        rows = list(csv.DictReader(f))

    inserted = skipped = 0
    for row in rows:
        taxon_id = int(row["taxon_id"])
        try:
            conn.execute(
                """
                INSERT INTO species (
                    inat_taxon_id, common_name, scientific_name,
                    kingdom, phylum, class, "order", family, genus,
                    iconic_taxon, photo_url_thumb, photo_url_med,
                    pokedex_entry
                ) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
                """,
                (
                    taxon_id,
                    row["common_name"],
                    row["scientific_name"],
                    row.get("kingdom"),
                    row.get("phylum"),
                    row.get("class"),
                    row.get("order"),
                    row.get("family"),
                    row.get("genus"),
                    row.get("iconic_taxon"),
                    row.get("photo_url_thumb"),
                    row.get("photo_url_med"),
                    dex_entries.get(taxon_id),
                ),
            )
            inserted += 1
        except sqlite3.IntegrityError:
            skipped += 1

    conn.commit()
    print(f"Species import complete: {inserted} inserted, {skipped} skipped (already exist).")


def verify_db(conn: sqlite3.Connection) -> None:
    tables = ["species", "species_detail", "sighting", "lookalike"]
    for table in tables:
        (count,) = conn.execute(f"SELECT COUNT(*) FROM {table}").fetchone()
        print(f"  {table}: {count} rows")

    print("\nSample species rows:")
    rows = conn.execute(
        "SELECT inat_taxon_id, common_name, scientific_name, iconic_taxon, pokedex_entry IS NOT NULL FROM species LIMIT 5"
    ).fetchall()
    for taxon_id, common, sci, iconic, has_dex in rows:
        dex_flag = "has dex entry" if has_dex else "no dex entry"
        print(f"  [{taxon_id}] {common} ({sci}) — {iconic} — {dex_flag}")


if __name__ == "__main__":
    with sqlite3.connect(DB_PATH) as conn:
        create_tables(conn)
        import_species(conn)
        print("\nVerification:")
        verify_db(conn)
    print(f"\nDatabase ready at {DB_PATH}")