upstream/borgbackup
1
0
Fork 0
mirror of https://github.com/borgbackup/borg.git synced 2026-06-11 09:59:19 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions 6

unify scanning and listing of segment dirs / segment files and apply good practices

Browse source 
+ os.scandir instead of os.listdir
  Improved speed and added flexibility with attributes (name,path,is_dir(),is_file())
+ use is_dir / is_file to make sure  we're reading only dirs / files respectively
+ Filtering to particular start, end index range built in
+ Have seperate generators to avoid unnecessary less_than / greater_than checking for every dir/file when not required

Resolves #7597
This commit is contained in:
nain 2023-05-25 11:25:10 -04:00
parent 53bedfb63b
commit c9f35a16e9
1 changed files with 35 additions and 11 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

46
src/borg/repository.py
View file

@ -1319,28 +1319,52 @@ class LoggedIO:
safe_fadvise(fd.fileno(), 0, 0, 'DONTNEED')
fd.close()
def get_segment_dirs(self, start_index=None, end_index=None):
"""Returns generator yeilding required segment dirs as `os.DirEntry` objects. Start and end are inclusive.
"""
data_dir = os.path.join(self.path, 'data')
if start_index is None and end_index is None:
segment_dirs = (f for f in os.scandir(data_dir) if f.is_dir() and f.name.isdigit())
elif start_index is not None and end_index is None:
segment_dirs = (f for f in os.scandir(data_dir) if f.is_dir() and f.name.isdigit() and start_index <= int(f.name))
elif start_index is None and end_index is not None:
segment_dirs = (f for f in os.scandir(data_dir) if f.is_dir() and f.name.isdigit() and int(f.name) <= end_index)
elif start_index is not None and end_index is not None:
segment_dirs = (f for f in os.scandir(data_dir) if f.is_dir() and f.name.isdigit() and start_index <= int(f.name) <= end_index)
return segment_dirs
def get_segment_files(self, segment_dir, start_index=None, end_index=None):
"""Returns generator yeilding required segment files in segment_dir as `os.DirEntry` objects. Start and end are inclusive.
"""
if start_index is None and end_index is None:
segment_files = (f for f in os.scandir(segment_dir) if f.is_file() and f.name.isdigit())
elif start_index is not None and end_index is None:
segment_files = (f for f in os.scandir(segment_dir) if f.is_file() and f.name.isdigit() and start_index <= int(f.name))
elif start_index is None and end_index is not None:
segment_files = (f for f in os.scandir(segment_dir) if f.is_file() and f.name.isdigit() and int(f.name) <= end_index)
elif start_index is not None and end_index is not None:
segment_files = (f for f in os.scandir(segment_dir) if f.is_file() and f.name.isdigit() and start_index <= int(f.name) <= end_index)
return segment_files
def segment_iterator(self, segment=None, reverse=False):
if segment is None:
segment = 0 if not reverse else 2 ** 32 - 1
data_path = os.path.join(self.path, 'data')
start_segment_dir = segment // self.segments_per_dir
dirs = os.listdir(data_path)
if not reverse:
dirs = [dir for dir in dirs if dir.isdigit() and int(dir) >= start_segment_dir]
dirs = self.get_segment_dirs(start_index=start_segment_dir)
else:
dirs = [dir for dir in dirs if dir.isdigit() and int(dir) <= start_segment_dir]
dirs = sorted(dirs, key=int, reverse=reverse)
dirs = self.get_segment_dirs(end_index=start_segment_dir)
dirs = sorted(dirs, key=lambda dir: int(dir.name), reverse=reverse)
for dir in dirs:
filenames = os.listdir(os.path.join(data_path, dir))
if not reverse:
filenames = [filename for filename in filenames if filename.isdigit() and int(filename) >= segment]
files = self.get_segment_files(dir, start_index=segment)
else:
filenames = [filename for filename in filenames if filename.isdigit() and int(filename) <= segment]
filenames = sorted(filenames, key=int, reverse=reverse)
for filename in filenames:
files = self.get_segment_files(dir, end_index=segment)
files = sorted(files, key=lambda file: int(file.name), reverse=reverse)
for file in files:
# Note: Do not filter out logically deleted segments (see "File system interaction" above),
# since this is used by cleanup and txn state detection as well.
yield int(filename), os.path.join(data_path, dir, filename)
yield int(file.name), file.path
def get_latest_segment(self):
for segment, filename in self.segment_iterator(reverse=True):