Spinner while copying; handle dirt files better

This commit is contained in:
Reid 'arrdem' McKenzie 2021-12-05 11:35:19 -07:00
parent 6605d28377
commit 5089b8c94f
2 changed files with 128 additions and 73 deletions

View file

@ -14,5 +14,6 @@ zapp_binary(
], ],
deps = [ deps = [
py_requirement("ExifRead"), py_requirement("ExifRead"),
py_requirement("yaspin"),
] ]
) )

View file

@ -29,6 +29,14 @@ from .util import *
# FIXME: use piexif, which supports writeback not exifread. # FIXME: use piexif, which supports writeback not exifread.
import exifread import exifread
from yaspin import Spinner, yaspin
_print = print
def print(*strs, **kwargs):
_print("\r", *strs, **kwargs)
parser = argparse.ArgumentParser() parser = argparse.ArgumentParser()
@ -38,6 +46,17 @@ parser.add_argument("destructive", action="store_true", default=False)
MODIFIED_ISO_DATE = "%Y:%m:%dT%H:%M:%SF%f" MODIFIED_ISO_DATE = "%Y:%m:%dT%H:%M:%SF%f"
SPINNER = Spinner(["|", "/", "-", "\\"], 200)
KNOWN_IMG_TYPES = {
".jpg": ".jpeg",
".jpeg": ".jpeg",
".png": ".png",
".mov": ".mov",
".gif": ".gif",
".mp4": ".mp4",
".m4a": ".m4a",
".oga": ".oga", # How the hell do I have ogg files kicking around
}
def exif_tags(p: Path) -> object: def exif_tags(p: Path) -> object:
@ -60,14 +79,29 @@ def safe_strptime(date, format):
try: try:
return datetime.strptime(date, format) return datetime.strptime(date, format)
except ValueError: except ValueError:
return None pass
def date_from_name(p: Path): def safe_ymdhmms(date):
"""Try to munge a datestamp out of a path.""" fmt = (
r"(?P<year>\d{4})(?P<month>\d{2})(?P<day>\d{2})"
r" "
r"(?P<hour>\d{2})(?P<minute>\d{2})(?P<second>\d{2})(?P<millisecond>\d{3})"
)
m = re.match(fmt, date)
if m:
return datetime(
year=int(m.group("year")),
month=int(m.group("month")),
day=int(m.group("day")),
hour=int(m.group("hour")),
minute=int(m.group("minute")),
second=int(m.group("second")),
microsecond=int(m.group("millisecond")) * 1000,
)
fname = ".".join(p.name.split(".")[:-1])
def date_from_name(fname: str):
# Discard common prefixes # Discard common prefixes
fname = fname.replace("IMG_", "") fname = fname.replace("IMG_", "")
fname = fname.replace("PXL_", "") fname = fname.replace("PXL_", "")
@ -93,54 +127,59 @@ def date_from_name(p: Path):
# Try to guess the date # Try to guess the date
# File date formats: # File date formats:
for fmt in [ for unfmt in [
# Our date format # Our date format
MODIFIED_ISO_DATE, lambda d: safe_strptime(d, MODIFIED_ISO_DATE),
# A bug # A bug
# 2014:08:21T19:4640F1408672000 # 2014:08:21T19:4640F1408672000
# 2015:12:14T23:0933F1450159773 # 2015:12:14T23:0933F1450159773
"%Y:%m:%dT%H:%M%SF%f", lambda d: safe_strptime(d, "%Y:%m:%dT%H:%M%SF%f"),
# 2020-12-21 17.15.09.0 # 2020-12-21 17.15.09.0
"%Y-%m-%d %H.%M.%S.%f", lambda d: safe_strptime(d, "%Y-%m-%d %H.%M.%S.%f"),
# 2020-12-21 17.15.09 # 2020-12-21 17.15.09
"%Y-%m-%d %H.%M.%S", lambda d: safe_strptime(d, "%Y-%m-%d %H.%M.%S"),
# 2019-02-09 12.45.32-6 # 2019-02-09 12.45.32-6
# 2019-01-13 13.43.45-16 # 2019-01-13 13.43.45-16
"%Y-%m-%d %H.%M.%S-%f", lambda d: safe_strptime(d, "%Y-%m-%d %H.%M.%S-%f"),
# Note the _1 or such may not be millis, but we assume it is. # Note the _1 or such may not be millis, but we assume it is.
# 20171113_130826_1 # 20171113_130826_1
# 20171113 130826 1 # 20171113 130826 1
"%Y%m%d %H%M%S %f", lambda d: safe_strptime(d, "%Y%m%d %H%M%S %f"),
# 20180404_114639 # 20180404_114639
# 20180404 114639 # 20180404 114639
"%Y%m%d %H%M%S", lambda d: safe_strptime(d, "%Y%m%d %H%M%S"),
# 2017-11-05_15:15:55 # 2017-11-05_15:15:55
# 2017-11-05 15:15:55 # 2017-11-05 15:15:55
"%Y-%m-%d %H:%M:%S", lambda d: safe_strptime(d, "%Y-%m-%d %H:%M:%S"),
lambda d: safe_strptime(d, "%Y%m%d %h%m%s%f"),
# HACK:
# Python doesn't support %s as milliseconds; these don't quite work.
# So use a custom matcher.
# 20210526 002327780
# 20210417_220753284 # 20210417_220753284
# 20210417 220753284 # 20210417 220753284
# 20210304 204755545 # 20210304 204755545
"%Y%m%d %h%m%s%f", # 20211111 224304117
safe_ymdhmms,
]: ]:
try: val = unfmt(fname)
return datetime.strptime(fname, fmt) if val is not None:
except ValueError: return val
continue
else:
def date_from_path(p: Path):
"""Try to munge a datestamp out of a path."""
fname = ".".join(p.name.split(".")[:-1])
date = date_from_name(fname)
if not date:
print(f"Warning: Unable to infer datetime from {fname!r}", file=sys.stderr) print(f"Warning: Unable to infer datetime from {fname!r}", file=sys.stderr)
return date
def normalize_ext(p: Path): def normalize_ext(p: Path):
renaming = { renaming = KNOWN_IMG_TYPES
".jpg": ".jpeg",
".jpeg": ".jpeg",
".png": ".png",
".mov": ".mov",
".gif": ".gif",
".mp4": ".mp4",
".m4a": ".m4a",
".oga": ".oga", # How the hell do I have ogg files kicking around
}
exts = [e.lower() for e in p.suffixes] exts = [e.lower() for e in p.suffixes]
# Guess an ext out of potentially many, allowing only for folding of effective dupes # Guess an ext out of potentially many, allowing only for folding of effective dupes
exts = set(renaming[e] for e in exts if e in renaming) exts = set(renaming[e] for e in exts if e in renaming)
@ -264,7 +303,7 @@ def img_info(p: Path) -> ImgInfo:
) )
if date and (date := safe_strptime(date, "%Y:%m:%d %H:%M:%S")): if date and (date := safe_strptime(date, "%Y:%m:%d %H:%M:%S")):
pass pass
elif date := date_from_name(p): elif date := date_from_path(p):
dirty |= True dirty |= True
else: else:
# The oldest of the mtime and the ctime # The oldest of the mtime and the ctime
@ -285,6 +324,9 @@ def img_info(p: Path) -> ImgInfo:
date = date.replace(microsecond=subsec) date = date.replace(microsecond=subsec)
if not (2015 <= date.year <= datetime.now().year):
raise ValueError(f"{p}'s inferred date ({date!r}) is beyond the sanity-check range!")
return ImgInfo( return ImgInfo(
p, p,
tags, tags,
@ -310,11 +352,12 @@ def main():
raise OSError() raise OSError()
src.rename(target) # Execute the rename src.rename(target) # Execute the rename
except OSError: # cross-device move except OSError: # cross-device move
with yaspin(SPINNER):
copyfile(src, target) copyfile(src, target)
if opts.destructive: if opts.destructive:
src.chmod(0o644)
src.unlink() src.unlink()
print(" unlink: ok") print(" unlink: ok")
@ -323,15 +366,20 @@ def main():
sequence_name = None sequence_name = None
sequence = 0 sequence = 0
for src in list(opts.src_dir.glob("**/*")): for src in opts.src_dir.glob("**/*"):
print(f"{src}:")
ext = "." + src.name.lower().split(".")[-1]
print(f" msg: ext inferred as {ext}")
if src.is_dir(): if src.is_dir():
continue continue
elif src.name.startswith("."): elif ext in ["thm", "lrv", "ico", "sav"] or src.name.startswith("._"):
if opts.destructive:
src.unlink()
continue continue
print(f"{src}:") elif ext in KNOWN_IMG_TYPES:
info = img_info(src) info = img_info(src)
year_dir = Path(opts.dest_dir / str(info.date.year)) year_dir = Path(opts.dest_dir / str(info.date.year))
year_dir.mkdir(exist_ok=True) # Ignore existing and continue year_dir.mkdir(exist_ok=True) # Ignore existing and continue
@ -358,18 +406,24 @@ def main():
if not target.exists(): if not target.exists():
# src & !target => copy # src & !target => copy
_copy(src, target) _copy(src, target)
elif src == target: elif src == target:
# src == target; skip DO NOT DELETE SRC # src == target; skip DO NOT DELETE SRC
pass pass
elif checksum_path_blocks(src) == checksum_path_blocks(target): elif checksum_path_blocks(src) == checksum_path_blocks(target):
print(f" ok: {target}") print(f" ok: {target}")
# src != target && id(src) == id(target); delete src # src != target && id(src) == id(target); delete src
if opts.destructive: if opts.destructive:
src.chmod(0o644)
src.unlink() src.unlink()
else: else:
# src != target && id(src) != id(target); replace target with src? # src != target && id(src) != id(target); replace target with src?
print(f" warning: {target} is a content-id collision with a different checksum; skipping") print(f" warning: {target} is a content-id collision with a different checksum; skipping")
else:
print(f" msg: unknown filetype {ext}")
if __name__ == "__main__": if __name__ == "__main__":
main() main()