HCha/Stockfish
1
0
Fork 0
mirror of https://github.com/HChaZZY/Stockfish.git synced 2025-12-23 02:27:00 +08:00
Code Issues Packages Projects Releases Wiki Activity

Add split_count argument to shuffle_binpack.py

Browse Source 
this optional argument allows for splitting the input binpack in multiple output binpacks while shuffling.
This commit is contained in:
Joost VandeVondele
2020-12-24 14:34:26 +01:00
committed by nodchip
parent 4f6fdca31f
commit bb6188430d
1 changed files with 31 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files

42
script/shuffle_binpack.py
View File

@@ -25,16 +25,18 @@ def index_binpack(file):
return index
def copy_binpack_indexed(in_file, index, out_file):
def copy_binpack_indexed(in_file, index, out_files):
print('Copying...')
total_size = 0
report_every = 100
prev_mib = -report_every
nextfile = 0
for offset, size in index:
in_file.seek(offset, os.SEEK_SET)
data = in_file.read(size)
assert len(data) == size
out_file.write(data)
out_files[nextfile].write(data)
nextfile = (nextfile + 1) % len(out_files)
total_size += size
mib = total_size // 1024 // 1024
@@ -44,26 +46,44 @@ def copy_binpack_indexed(in_file, index, out_file):
def main():
if len(sys.argv) < 3:
print('Usage: python shuffle_binpack.py infile outfile')
print('Usage: python shuffle_binpack.py infile outfile [split_count]')
return
in_filename = sys.argv[1]
out_filename = sys.argv[2]
if (Path(out_filename).exists()):
print('Output path already exists. Please specify a path to a file that does not exist.')
return
if len(sys.argv) > 3:
# split the infile in split_count pieces, creating new outfile names based on the provided name
basefile = sys.argv[2]
split_count = int(sys.argv[3])
base=os.path.splitext(basefile)[0]
ext=os.path.splitext(basefile)[1]
out_filenames = []
for i in range(split_count):
out_filenames.append(base+"_{}".format(i)+ext)
else:
out_filenames = [sys.argv[2]]
for out_filename in out_filenames:
if (Path(out_filename).exists()):
print('Output path {} already exists. Please specify a path to a file that does not exist.'.format(out_filename))
return
print(out_filenames)
in_file = open(in_filename, 'rb')
out_file = open(out_filename, 'wb')
index = index_binpack(in_file)
print('Shuffling...')
random.shuffle(index)
copy_binpack_indexed(in_file, index, out_file)
out_files = []
for out_filename in out_filenames:
out_files.append(open(out_filename, 'wb'))
copy_binpack_indexed(in_file, index, out_files)
in_file.close()
out_file.close()
for out_file in out_files:
out_file.close()
main()