Files
web-at-transit/script/upload_seed_images.py
T
2026-05-24 22:55:17 +07:00

219 lines
6.9 KiB
Python

#!/usr/bin/env python3
"""
Upload images from script/images/ to SeaweedFS filer and generate
seed_images.sql with INSERT statements for the images table.
Usage:
python script/upload_seed_images.py --filer-url http://VPS_IP:3103
python script/upload_seed_images.py --filer-url http://VPS_IP:3103 --dry-run
Output:
script/seed_images.sql (always written)
"""
import argparse
import hashlib
import mimetypes
import os
import sys
import uuid
from pathlib import Path
try:
import requests
except ImportError:
print("pip install requests", file=sys.stderr)
sys.exit(1)
SCRIPT_DIR = Path(__file__).parent
IMAGES_DIR = SCRIPT_DIR / "images"
OUTPUT_SQL = SCRIPT_DIR / "seed_images.sql"
# --------------------------------------------------------------------------- #
# folder → (entity_type, kind, display_order_hint)
# --------------------------------------------------------------------------- #
FOLDER_MAP = {
"stops": ("station", "photo", 0),
"qrcode": ("route", "map", 9),
"partners": ("sponsor", "logo", 0),
"landmarks": ("misc", "photo", 0),
"network": ("misc", "map", 0),
"extras": ("misc", "photo", 0),
"icons": ("misc", "icon", 0),
}
def classify(rel: Path):
"""Return (entity_type, entity_id, kind, display_order) or None to skip."""
parts = rel.parts # e.g. ('routes', '902', 'car.jpeg')
folder = parts[0]
if folder == "routes" and len(parts) == 3:
route_id = parts[1]
stem = Path(parts[2]).stem.lower()
kind = "hero" if stem == "hero" else "photo"
order = 1 if kind == "hero" else 0
return "route", route_id, kind, order
if folder in FOLDER_MAP and len(parts) == 2:
entity_type, kind, order = FOLDER_MAP[folder]
entity_id = Path(parts[1]).stem
return entity_type, entity_id, kind, order
return None
def get_mime(path: Path) -> str:
# mimetypes doesn't know .webp by default on some platforms
ext = path.suffix.lower()
if ext == ".webp":
return "image/webp"
mime, _ = mimetypes.guess_type(str(path))
return mime or "application/octet-stream"
def sha256_of(path: Path) -> str:
h = hashlib.sha256()
with open(path, "rb") as f:
for chunk in iter(lambda: f.read(65536), b""):
h.update(chunk)
return h.hexdigest()
def get_dimensions(path: Path):
try:
from PIL import Image
with Image.open(path) as img:
return img.size # (width, height)
except Exception:
return None, None
def upload(filer_url: str, entity_type: str, entity_id: str,
filename: str, mime: str, data: bytes) -> dict:
"""
PUT raw bytes to filer — mirrors seaweed.rs upload logic.
Returns {"fid": ..., "url_path": ...}
"""
key = (f"/buckets/transit/{sanitize(entity_type)}"
f"/{sanitize(entity_id)}/{sanitize_name(filename)}")
url = filer_url.rstrip("/") + key
resp = requests.put(url, data=data,
headers={"Content-Type": mime}, timeout=30)
resp.raise_for_status()
body = {}
try:
body = resp.json()
except Exception:
pass
fid = body.get("fid") or key # fallback: path itself (matches Rust code)
url_path = key.removeprefix("/buckets/transit") # stored in DB without prefix
return {"fid": fid, "url_path": url_path}
def sanitize(s: str) -> str:
return "".join(c if c.isalnum() or c in "-_" else "_" for c in s)
def sanitize_name(s: str) -> str:
return "".join(c if c.isalnum() or c in "-_." else "_" for c in s)
def q(s: str) -> str:
"""SQL single-quote escape."""
return s.replace("'", "''")
# --------------------------------------------------------------------------- #
def main():
parser = argparse.ArgumentParser()
parser.add_argument("--filer-url", default="http://localhost:3103",
help="SeaweedFS filer base URL (default: http://localhost:3103)")
parser.add_argument("--dry-run", action="store_true",
help="Skip actual upload; use path as fid placeholder")
args = parser.parse_args()
filer_url = args.filer_url.rstrip("/")
dry_run = args.dry_run
sql_lines = [
"-- Auto-generated by upload_seed_images.py",
"-- Run this AFTER 0003_seed.sql is applied.",
"-- ON CONFLICT (fid) DO NOTHING → safe to re-run.",
"",
]
image_files = sorted(p for p in IMAGES_DIR.rglob("*") if p.is_file())
ok = skip = 0
errors = []
for img_path in image_files:
rel = img_path.relative_to(IMAGES_DIR)
meta = classify(rel)
if meta is None:
print(f" SKIP {rel}")
skip += 1
continue
entity_type, entity_id, kind, display_order = meta
mime = get_mime(img_path)
file_size = img_path.stat().st_size
sha = sha256_of(img_path)
width, height = get_dimensions(img_path)
img_id = str(uuid.uuid4())
if dry_run:
safe_key = (f"/buckets/transit/{sanitize(entity_type)}"
f"/{sanitize(entity_id)}/{sanitize_name(img_path.name)}")
fid = safe_key # use path as pseudo-fid
url_path = safe_key.removeprefix("/buckets/transit")
print(f" DRY {rel}")
else:
try:
data = img_path.read_bytes()
result = upload(filer_url, entity_type, entity_id,
img_path.name, mime, data)
fid = result["fid"]
url_path = result["url_path"]
print(f" OK {rel} fid={fid}")
ok += 1
except Exception as exc:
print(f" ERR {rel} {exc}", file=sys.stderr)
errors.append((str(rel), str(exc)))
continue
# Build meta JSONB inline
meta_kv = [f'"mime":"{q(mime)}"', f'"size_bytes":{file_size}',
f'"sha256":"{sha}"']
if width is not None:
meta_kv.append(f'"width":{width}')
meta_kv.append(f'"height":{height}')
meta_json = "{" + ",".join(meta_kv) + "}"
sql_lines.append(f"-- {rel}")
sql_lines.append(
f"INSERT INTO images "
f"(id, entity_type, entity_id, kind, fid, url_path, display_order, meta) VALUES ("
f"'{img_id}', '{entity_type}', '{q(entity_id)}', '{kind}', "
f"'{q(fid)}', '{q(url_path)}', {display_order}, "
f"'{meta_json}'::jsonb"
f") ON CONFLICT (fid) DO NOTHING;"
)
sql_lines += [
"",
f"-- done: ok={ok if not dry_run else 'DRY'}, skip={skip}, err={len(errors)}",
]
OUTPUT_SQL.write_text("\n".join(sql_lines) + "\n", encoding="utf-8")
print(f"\nSQL written → {OUTPUT_SQL.relative_to(SCRIPT_DIR.parent)}")
if errors:
print(f"Errors ({len(errors)}):")
for r, e in errors:
print(f" {r}: {e}")
sys.exit(1)
if __name__ == "__main__":
main()