Cameron CordesandClaude Opus 5.5
bc41c737e5
`e2e.sh check` compares every indexed fixture's manifest wall clock with the `wall_clock` reported by /image/metadata, /memories and /photos/gps-summary; any disagreement fails. - Fixture videos are stamped like phone containers: the America/New_York wall as a UTC creation_time (the label previously showed the UTC time). - mtimes are pinned in America/New_York instead of the host's zone. - The e2e env sets TZ=America/New_York (the #120 preflight): the Windows exe has no /etc/timezone, so server_zone() was otherwise None. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
346 lines
14 KiB
Python
346 lines
14 KiB
Python
#!/usr/bin/env python3
|
|
"""Generate the synthetic e2e media library.
|
|
|
|
Every image is rendered with its own relative path, date and scenario
|
|
burned into the pixels, so an agent looking at a screenshot can tell
|
|
exactly which fixture is on screen. EXIF (dates, GPS, camera) is stamped
|
|
in a single exiftool batch. A `manifest.json` next to the media records
|
|
what was generated and what the server should derive from it, so tests
|
|
can assert against it instead of hardcoding counts.
|
|
|
|
The harness runs in WSL but the tools (ffmpeg.exe / exiftool.exe) are
|
|
Windows binaries, so every path handed to them goes through `wslpath -w`.
|
|
|
|
Usage: gen_fixtures.py <out_dir> [--import DIR]
|
|
<out_dir>/media/ the pristine library (never served directly)
|
|
<out_dir>/manifest.json what's in it
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import argparse
|
|
import datetime as dt
|
|
import json
|
|
import os
|
|
import shutil
|
|
import subprocess
|
|
import sys
|
|
from concurrent.futures import ThreadPoolExecutor
|
|
from dataclasses import dataclass, field
|
|
from pathlib import Path
|
|
from zoneinfo import ZoneInfo
|
|
|
|
FFMPEG = shutil.which("ffmpeg.exe") or shutil.which("ffmpeg")
|
|
EXIFTOOL = shutil.which("exiftool.exe") or shutil.which("exiftool")
|
|
# drawtext needs an explicit font on Windows ffmpeg builds. The colon after
|
|
# the drive letter must be escaped inside a filter argument.
|
|
FONT = "C\\:/Windows/Fonts/consola.ttf"
|
|
|
|
# The owner's display zone and the server zone the e2e env pins (TZ=...).
|
|
# Naive fixture datetimes without GPS are walls in this zone.
|
|
HOME_TZ = ZoneInfo("America/New_York")
|
|
|
|
PLACES = {
|
|
"santa_monica": (34.0100, -118.4960),
|
|
"paris": (48.8584, 2.2945),
|
|
"yosemite": (37.7456, -119.5936),
|
|
"tokyo": (35.6595, 139.7005),
|
|
}
|
|
|
|
PALETTE = [
|
|
"0x1f4e79", "0x7a3e9d", "0x2e7d32", "0xb71c1c", "0xe65100",
|
|
"0x006064", "0x4e342e", "0x283593", "0x827717", "0xad1457",
|
|
]
|
|
|
|
|
|
@dataclass
|
|
class Item:
|
|
rel: str # path relative to the library root, forward slashes
|
|
kind: str # "image" | "video"
|
|
size: tuple[int, int] = (1600, 1200)
|
|
# Capture wall clock: EXIF DateTimeOriginal as-is; for videos, the wall in
|
|
# HOME_TZ, stamped as UTC creation_time the way phones write containers.
|
|
taken: dt.datetime | None = None
|
|
gps: str | None = None # key into PLACES
|
|
camera: tuple[str, str] | None = None # (Make, Model)
|
|
mtime: dt.datetime | None = None # pinned file mtime (undated fallback)
|
|
duration: int = 0 # video seconds
|
|
tags: list[str] = field(default_factory=list)
|
|
scenario: str = ""
|
|
copy_of: str | None = None # byte-identical duplicate of another item
|
|
|
|
def label_date(self) -> str:
|
|
if self.taken:
|
|
return self.taken.strftime("%Y-%m-%d %H:%M")
|
|
return "no EXIF date"
|
|
|
|
|
|
def win(p: Path) -> str:
|
|
return subprocess.check_output(["wslpath", "-w", str(p)], text=True).strip()
|
|
|
|
|
|
def build_items(today: dt.date) -> list[Item]:
|
|
items: list[Item] = []
|
|
canon = ("Canon", "Canon EOS R5")
|
|
iphone = ("Apple", "iPhone 12")
|
|
pixel = ("Google", "Pixel 7")
|
|
|
|
# Dated album with GPS — map pins, date grouping, tag search.
|
|
for i in range(6):
|
|
items.append(Item(
|
|
f"2019/2019-07-04 Beach Day/beach_{i + 1:02d}.jpg", "image",
|
|
taken=dt.datetime(2019, 7, 4, 10 + i, 15), gps="santa_monica",
|
|
camera=canon, tags=["beach", "summer"], scenario="dated album + GPS",
|
|
))
|
|
# Multi-day trip; mix of orientations.
|
|
for i in range(8):
|
|
items.append(Item(
|
|
f"2021/Paris Trip/IMG_{1000 + i}.jpg", "image",
|
|
size=(1200, 1600) if i % 3 == 0 else (1600, 1200),
|
|
taken=dt.datetime(2021, 5, 12 + i // 3, 9 + i, 30), gps="paris",
|
|
camera=iphone, tags=["travel", "paris"], scenario="trip, mixed orientation",
|
|
))
|
|
for i in range(5):
|
|
items.append(Item(
|
|
f"2022/Hiking/hike_{i + 1:02d}.jpg", "image",
|
|
taken=dt.datetime(2022, 8, 20, 7 + i, 0), gps="yosemite",
|
|
camera=pixel, tags=["outdoors"], scenario="album with a video",
|
|
))
|
|
items.append(Item(
|
|
"2022/Hiking/summit.mp4", "video", size=(1280, 720), duration=6,
|
|
taken=dt.datetime(2022, 8, 20, 12, 30), tags=["outdoors"],
|
|
scenario="landscape video (HLS)",
|
|
))
|
|
# No GPS, tagged people-ish album.
|
|
for i in range(6):
|
|
items.append(Item(
|
|
f"2023/Birthday/party_{i + 1:02d}.jpg", "image",
|
|
taken=dt.datetime(2023, 11, 3, 18, 5 * i), camera=iphone,
|
|
tags=["family"], scenario="dated, no GPS",
|
|
))
|
|
# Enough items to force scrolling / pagination in the grid.
|
|
start = dt.datetime(2020, 1, 3, 12, 0)
|
|
for i in range(60):
|
|
items.append(Item(
|
|
f"Grid Stress/photo_{i + 1:03d}.jpg", "image", size=(800, 600),
|
|
taken=start + dt.timedelta(days=6 * i), camera=pixel,
|
|
scenario="bulk grid / scroll",
|
|
))
|
|
# Dates that only exist in the filename (the memories.rs regexes).
|
|
items += [
|
|
Item("Screenshots/Screenshot_20220315-101500.png", "image", size=(1080, 2340),
|
|
scenario="date from filename (YYYYMMDD-HHMMSS), PNG, no EXIF"),
|
|
Item("Screenshots/IMG_20210704_123000.jpg", "image",
|
|
scenario="date from filename (YYYYMMDD_HHMMSS), no EXIF"),
|
|
Item("Screenshots/2020-02-29_08-00-00.jpg", "image",
|
|
scenario="date from filename (leap day), no EXIF"),
|
|
]
|
|
# No date anywhere: falls back to the pinned mtime.
|
|
for i in range(3):
|
|
items.append(Item(
|
|
f"Undated/scan_{i + 1:02d}.jpg", "image",
|
|
mtime=dt.datetime(2015, 3, 1 + i, 12, 0),
|
|
scenario="no EXIF, no filename date (mtime fallback)",
|
|
))
|
|
# "On this day" — same month/day as generation date in earlier years.
|
|
for year in (2018, 2020, 2024):
|
|
taken = _same_day(today, year, 14)
|
|
items.append(Item(
|
|
f"On This Day/otd_{year}.jpg", "image", taken=taken, gps="tokyo",
|
|
camera=canon, scenario=f"memories: on this day ({year})",
|
|
))
|
|
# Excluded via EXCLUDED_DIRS=/private — never indexed, must not appear anywhere.
|
|
items.append(Item(
|
|
"private/otd_hidden.jpg", "image", taken=_same_day(today, 2019, 9),
|
|
scenario="EXCLUDED_DIRS: never indexed or shown",
|
|
))
|
|
# Portrait video, filename-dated.
|
|
items.append(Item(
|
|
"Videos/VID_20230610_150000.mp4", "video", size=(720, 1280), duration=4,
|
|
taken=dt.datetime(2023, 6, 10, 15, 0), scenario="portrait video",
|
|
))
|
|
# Byte-identical duplicate for duplicate detection.
|
|
items.append(Item(
|
|
"Duplicates/IMG_1000 copy.jpg", "image", copy_of="2021/Paris Trip/IMG_1000.jpg",
|
|
scenario="exact duplicate of 2021/Paris Trip/IMG_1000.jpg",
|
|
))
|
|
return items
|
|
|
|
|
|
def expected_wall_clock(it: Item) -> str:
|
|
"""The wall clock the server should report (`wall_clock`): the filename
|
|
date for the filename class (it outranks EXIF), else the capture wall,
|
|
else the pinned mtime rendered in HOME_TZ."""
|
|
import re
|
|
name = it.rel.rsplit("/", 1)[-1]
|
|
m = (re.search(r"(\d{4})(\d{2})(\d{2})[_-](\d{2})(\d{2})(\d{2})", name)
|
|
or re.search(r"(\d{4})-(\d{2})-(\d{2})_(\d{2})-(\d{2})-(\d{2})", name))
|
|
if m:
|
|
wall = dt.datetime(*map(int, m.groups()))
|
|
else:
|
|
wall = it.taken or it.mtime or dt.datetime(2024, 1, 1)
|
|
return wall.strftime("%Y-%m-%dT%H:%M:%S")
|
|
|
|
|
|
def _same_day(today: dt.date, year: int, hour: int) -> dt.datetime:
|
|
# Feb 29 in a non-leap year would be invalid; fall back to Feb 28.
|
|
try:
|
|
return dt.datetime(year, today.month, today.day, hour, 0)
|
|
except ValueError:
|
|
return dt.datetime(year, today.month, 28, hour, 0)
|
|
|
|
|
|
def render(item: Item, idx: int, media: Path) -> None:
|
|
out = media / item.rel
|
|
out.parent.mkdir(parents=True, exist_ok=True)
|
|
w, h = item.size
|
|
color = PALETTE[idx % len(PALETTE)]
|
|
lines = [item.rel, item.label_date(), item.scenario]
|
|
# Consolas glyphs are ~0.55em wide; shrink so the longest line fits.
|
|
longest = max(len(line) for line in lines)
|
|
fs = max(12, min(min(w, h) // 22, int(w * 0.92 / (longest * 0.55))))
|
|
text = ",".join(
|
|
f"drawtext=fontfile='{FONT}':text='{_esc(line)}':fontcolor=white:"
|
|
f"fontsize={fs}:x=(w-text_w)/2:y=(h/2)+({n - 1}*{int(fs * 1.6)})"
|
|
f":box=1:boxcolor=black@0.45:boxborderw=8"
|
|
for n, line in enumerate(lines)
|
|
)
|
|
if item.kind == "image":
|
|
cmd = [
|
|
FFMPEG, "-hide_banner", "-loglevel", "error", "-y",
|
|
"-f", "lavfi", "-i",
|
|
f"gradients=s={w}x{h}:c0={color}:c1=0x101010:seed={idx}:duration=1",
|
|
"-vf", text, "-frames:v", "1",
|
|
]
|
|
cmd += ["-q:v", "3"] if out.suffix.lower() == ".jpg" else []
|
|
else:
|
|
cmd = [
|
|
FFMPEG, "-hide_banner", "-loglevel", "error", "-y",
|
|
"-f", "lavfi", "-i", f"testsrc2=s={w}x{h}:r=30:d={item.duration}",
|
|
"-f", "lavfi", "-i", f"sine=frequency={300 + idx * 20}:d={item.duration}",
|
|
"-vf", text, "-c:v", "libx264", "-pix_fmt", "yuv420p",
|
|
"-c:a", "aac", "-shortest",
|
|
]
|
|
if item.taken:
|
|
utc = item.taken.replace(tzinfo=HOME_TZ).astimezone(dt.timezone.utc)
|
|
cmd += ["-metadata", f"creation_time={utc.strftime('%Y-%m-%dT%H:%M:%S')}Z"]
|
|
subprocess.run(cmd + [win(out)], check=True)
|
|
|
|
|
|
def _esc(s: str) -> str:
|
|
# drawtext text escaping: backslash, colon, quote, percent, comma.
|
|
for ch in ("\\", ":", "'", "%", ","):
|
|
s = s.replace(ch, "\\" + ch)
|
|
return s
|
|
|
|
|
|
def stamp_exif(items: list[Item], media: Path, work: Path) -> None:
|
|
args: list[str] = []
|
|
for it in items:
|
|
if it.kind != "image" or it.copy_of or not (it.taken or it.gps or it.camera):
|
|
continue
|
|
if not it.rel.lower().endswith(".jpg"):
|
|
continue
|
|
if it.taken:
|
|
stamp = it.taken.strftime("%Y:%m:%d %H:%M:%S")
|
|
args += [f"-DateTimeOriginal={stamp}", f"-CreateDate={stamp}"]
|
|
if it.camera:
|
|
args += [f"-Make={it.camera[0]}", f"-Model={it.camera[1]}"]
|
|
if it.gps:
|
|
lat, lon = PLACES[it.gps]
|
|
args += [
|
|
f"-GPSLatitude={abs(lat)}", f"-GPSLatitudeRef={'N' if lat >= 0 else 'S'}",
|
|
f"-GPSLongitude={abs(lon)}", f"-GPSLongitudeRef={'E' if lon >= 0 else 'W'}",
|
|
]
|
|
args += [win(media / it.rel), "-execute"]
|
|
argfile = work / "exif_args.txt"
|
|
argfile.write_text("\n".join(args) + "\n", encoding="utf-8")
|
|
subprocess.run(
|
|
[EXIFTOOL, "-@", win(argfile), "-common_args", "-q", "-overwrite_original",
|
|
"-charset", "filename=utf8"],
|
|
check=True,
|
|
)
|
|
argfile.unlink()
|
|
|
|
|
|
def main() -> int:
|
|
ap = argparse.ArgumentParser()
|
|
ap.add_argument("out_dir", type=Path)
|
|
ap.add_argument("--import", dest="import_dir", type=Path,
|
|
help="also copy this folder (read-only source) into media/imported/")
|
|
a = ap.parse_args()
|
|
if not FFMPEG or not EXIFTOOL:
|
|
print("ffmpeg(.exe) and exiftool(.exe) must be on PATH", file=sys.stderr)
|
|
return 1
|
|
|
|
out: Path = a.out_dir.resolve()
|
|
media = out / "media"
|
|
if media.exists():
|
|
shutil.rmtree(media)
|
|
media.mkdir(parents=True)
|
|
(media / "Empty Album").mkdir() # edge case: folder with no media
|
|
|
|
today = dt.date.today()
|
|
items = build_items(today)
|
|
originals = [(i, it) for i, it in enumerate(items) if not it.copy_of]
|
|
with ThreadPoolExecutor(max_workers=8) as pool:
|
|
list(pool.map(lambda p: render(p[1], p[0], media), originals))
|
|
stamp_exif(items, media, out)
|
|
|
|
for it in items:
|
|
if it.copy_of:
|
|
(media / it.rel).parent.mkdir(parents=True, exist_ok=True)
|
|
shutil.copyfile(media / it.copy_of, media / it.rel)
|
|
# Pin mtimes last: exiftool rewrites (and so touches) the files.
|
|
for it in items:
|
|
when = it.mtime or it.taken or dt.datetime(2024, 1, 1)
|
|
ts = when.replace(tzinfo=HOME_TZ).timestamp() # not WSL's local zone
|
|
os.utime(media / it.rel, (ts, ts))
|
|
|
|
imported = 0
|
|
if a.import_dir:
|
|
dest = media / "imported"
|
|
shutil.copytree(a.import_dir, dest) # copy: the source is never touched
|
|
imported = sum(1 for p in dest.rglob("*") if p.is_file())
|
|
|
|
by_path = {it.rel: it for it in items}
|
|
tags: dict[str, list[str]] = {}
|
|
for it in items:
|
|
for t in it.tags:
|
|
tags.setdefault(t, []).append(it.rel)
|
|
manifest = {
|
|
"generated_on": today.isoformat(),
|
|
"note": "'On This Day' fixtures match generated_on's month/day; regenerate "
|
|
"(e2e.sh fixtures --force) on another day to keep memories populated.",
|
|
"counts": {
|
|
"images": sum(1 for it in items if it.kind == "image"),
|
|
"videos": sum(1 for it in items if it.kind == "video"),
|
|
"total": len(items),
|
|
# What the server should index: EXCLUDED_DIRS is skipped at ingest.
|
|
"indexed": sum(1 for it in items if not it.rel.startswith("private/")) + imported,
|
|
"imported_files": imported,
|
|
},
|
|
"excluded_dirs": ["/private"],
|
|
"places": {k: {"lat": v[0], "lon": v[1]} for k, v in PLACES.items()},
|
|
"tags": tags,
|
|
"items": [
|
|
{
|
|
"path": it.rel, "kind": it.kind, "scenario": it.scenario,
|
|
"width": it.size[0], "height": it.size[1],
|
|
"taken": it.taken.isoformat() if it.taken else None,
|
|
"gps": it.gps, "camera": it.camera[1] if it.camera else None,
|
|
"tags": it.tags, "duplicate_of": it.copy_of,
|
|
# A duplicate carries its source's bytes, EXIF included.
|
|
"wall_clock": expected_wall_clock(by_path.get(it.copy_of, it)),
|
|
}
|
|
for it in items
|
|
],
|
|
}
|
|
(out / "manifest.json").write_text(json.dumps(manifest, indent=2) + "\n")
|
|
print(f"generated {len(items)} fixtures ({imported} imported) in {media}")
|
|
return 0
|
|
|
|
|
|
if __name__ == "__main__":
|
|
sys.exit(main())
|