feat: import fruit DB from XML — scripts/import_fruits.py (story #03)

Adds a repeatable Python import script that populates fruits, fruit_synonyms,
and fruit_images from 03-data/obstsorten.xml and the three image directories.
Idempotent: deletes synonyms and images per fruit before re-inserting; upserts
fruit row by osdb_number. Reads DATABASE_URL from env.

- map_typ: 23 XML typ codes → 17-value fruit_type enum; aggregate types coerced
  with logged warning; typ='p' (2 entries) mapped to Pfirsiche
- parse_synonyms: comma-split with whitespace/newline strip, empty filter
- parse_date: YYYYMMDD → date, fallback None → DB default NOW()
- find_images: globs {id}_*_s0.* across einzelfruechte/blueten/baume dirs
- 42 unit tests (no DB required) cover all mapping, parsing, and glob logic
- Makefile test target extended to include Python test suite
- 03-data/ and obstsorten.zip added to .gitignore (large binary data)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
2026-06-17 13:24:29 +02:00
co-authored by Claude Sonnet 4.6
parent 841f53fac8
commit 4d4c5f399f
4 changed files with 421 additions and 1 deletions
+237
View File
@@ -0,0 +1,237 @@
#!/usr/bin/env python3
"""
Import fruits from 03-data/obstsorten.xml into the OSDB database.
Usage:
DATABASE_URL=postgres://... python3 scripts/import_fruits.py
Idempotent: deletes synonyms and images for each fruit before re-inserting.
Upserts the fruit row by osdb_number.
"""
import glob
import os
import sys
import xml.etree.ElementTree as ET
from datetime import date, datetime
import psycopg2
DATA_DIR = os.path.join(os.path.dirname(__file__), "..", "03-data")
XML_PATH = os.path.join(DATA_DIR, "obstsorten.xml")
# typ values that are aggregate categories — coerced with a logged warning
AGGREGATE_TYPS = {"x", "bq", "aprpfi", "mirren", "pflzwe", "bo", "k"}
TYP_MAP = {
"a": "Apfelsorten",
"x": "Apfelsorten",
"b": "Birnensorten",
"bq": "Birnensorten",
"q": "Quittensorten",
"apr": "Aprikosen",
"aprpfi": "Aprikosen",
"pfi": "Pfirsiche",
"p": "Pfirsiche",
"mir": "Mirabellen",
"mirren": "Mirabellen",
"ren": "Renekloden",
"pfl": "Pflaumen",
"pflzwe": "Pflaumen",
"zwe": "Zwetschen",
"k": "Sauerkirschen",
"bo": "Brombeeren",
"bob": "Brombeeren",
"boe": "Erdbeeren",
"boh": "Himbeeren",
"boj": "Johannisbeeren",
"bos": "Stachelbeeren",
"wei": "Wein",
"w": "Wein",
}
IMAGE_DIRS = [
("einzelfruechte", "fruit"),
("blueten", "flower"),
("baume", "tree"),
]
def map_typ(typ: str):
"""Map XML typ code to fruit_type enum value. Returns None for unknowns."""
return TYP_MAP.get(typ)
def parse_synonyms(text) -> list:
"""Split comma-separated synonym text into a clean list."""
if not text:
return []
return [s.strip() for s in text.split(",") if s.strip()]
def parse_date(text) -> date | None:
"""Parse YYYYMMDD string to date. Returns None on failure or empty input."""
if not text:
return None
try:
return datetime.strptime(text.strip(), "%Y%m%d").date()
except ValueError:
return None
def find_images(fruit_id: str, data_dir: str) -> list:
"""
Find all _s0 images for a fruit across the three image directories.
Returns list of dicts with keys: filename, data, image_type.
"""
results = []
for subdir, image_type in IMAGE_DIRS:
pattern = os.path.join(data_dir, subdir, f"{fruit_id}_*_s0.*")
for path in sorted(glob.glob(pattern)):
with open(path, "rb") as f:
data = f.read()
results.append(
{
"filename": os.path.basename(path),
"data": data,
"image_type": image_type,
}
)
return results
def import_fruits(conn, data_dir: str = None) -> dict:
"""
Import all fruits from obstsorten.xml.
Returns counts dict: fruits, synonyms, images, warnings, skipped.
"""
if data_dir is None:
data_dir = DATA_DIR
xml_path = os.path.join(data_dir, "obstsorten.xml")
tree = ET.parse(xml_path)
root = tree.getroot()
counts = {"fruits": 0, "synonyms": 0, "images": {"fruit": 0, "flower": 0, "tree": 0}, "warnings": 0, "skipped": 0}
warnings = []
with conn.cursor() as cur:
for sorte in root.findall("sorte"):
osdb_number = (sorte.findtext("id") or "").strip()
name = (sorte.findtext("name") or "").strip()
typ = (sorte.findtext("typ") or "").strip()
synonym_text = sorte.findtext("synonym")
added_text = (sorte.findtext("added") or "").strip()
if not osdb_number or not name:
counts["skipped"] += 1
warnings.append(f"SKIP: missing id or name (id={osdb_number!r})")
continue
fruit_type = map_typ(typ)
if fruit_type is None:
counts["skipped"] += 1
warnings.append(f"SKIP: unknown typ={typ!r} for id={osdb_number!r}")
counts["warnings"] += 1
continue
if typ in AGGREGATE_TYPS:
warnings.append(f"COERCE: typ={typ!r}{fruit_type!r} for id={osdb_number!r}")
counts["warnings"] += 1
created_at = parse_date(added_text)
synonyms = parse_synonyms(synonym_text)
# Upsert fruit — get id back for child inserts
if created_at:
cur.execute(
"""
INSERT INTO fruits (name, osdb_number, fruit_type, created_at, updated_at)
VALUES (%s, %s, %s, %s, NOW())
ON CONFLICT (osdb_number) DO UPDATE
SET name = EXCLUDED.name,
fruit_type = EXCLUDED.fruit_type,
comment = NULL,
updated_at = NOW()
RETURNING id
""",
(name, osdb_number, fruit_type, created_at),
)
else:
cur.execute(
"""
INSERT INTO fruits (name, osdb_number, fruit_type, updated_at)
VALUES (%s, %s, %s, NOW())
ON CONFLICT (osdb_number) DO UPDATE
SET name = EXCLUDED.name,
fruit_type = EXCLUDED.fruit_type,
comment = NULL,
updated_at = NOW()
RETURNING id
""",
(name, osdb_number, fruit_type),
)
fruit_id = cur.fetchone()[0]
counts["fruits"] += 1
# Idempotent: delete existing synonyms and images
cur.execute("DELETE FROM fruit_synonyms WHERE fruit_id = %s", (fruit_id,))
cur.execute("DELETE FROM fruit_images WHERE fruit_id = %s", (fruit_id,))
# Insert synonyms
for syn in synonyms:
cur.execute(
"INSERT INTO fruit_synonyms (fruit_id, synonym) VALUES (%s, %s)",
(fruit_id, syn),
)
counts["synonyms"] += 1
# Insert images
for img in find_images(osdb_number, data_dir):
cur.execute(
"""
INSERT INTO fruit_images (fruit_id, filename, data, image_type, created_at)
VALUES (%s, %s, %s, %s, NOW())
""",
(fruit_id, img["filename"], psycopg2.Binary(img["data"]), img["image_type"]),
)
counts["images"][img["image_type"]] += 1
conn.commit()
for w in warnings[:20]:
print(f" WARNING: {w}", file=sys.stderr)
if len(warnings) > 20:
print(f" ... and {len(warnings) - 20} more warnings", file=sys.stderr)
return counts
def main():
database_url = os.environ.get("DATABASE_URL")
if not database_url:
print("Error: DATABASE_URL environment variable not set", file=sys.stderr)
sys.exit(1)
print("Connecting to database...")
conn = psycopg2.connect(database_url)
try:
print("Importing fruits...")
counts = import_fruits(conn)
finally:
conn.close()
total_images = sum(counts["images"].values())
print(f"\nDone.")
print(f" Fruits upserted: {counts['fruits']}")
print(f" Synonyms inserted: {counts['synonyms']}")
print(f" Images inserted: {total_images}")
print(f" fruit: {counts['images']['fruit']}")
print(f" flower: {counts['images']['flower']}")
print(f" tree: {counts['images']['tree']}")
print(f" Warnings: {counts['warnings']}")
print(f" Skipped: {counts['skipped']}")
if __name__ == "__main__":
main()