def process_local_dir(input_path, output_dir=None, max_books=None, heap_profile=False, overwrite=False):
if not os.path.exists(input_path):
print "Specified path does not exist: ", input_path
exit()
if os.path.isdir(input_path):
files = os.listdir(input_path)
basedir = input_path
else:
basedir = os.path.split(input_path)[0]
files = [os.path.basename(input_path)]
count = 0
for filename in files:
if filename.endswith(".csv") or filename.endswith(".csv.gz"):
count += 1
input_filename = os.path.join(basedir, filename)
print "----"
print "Processing #", count, " : ", input_filename
dest_filename_1ms, dest_filename_100ms = output_filenames(input_filename, output_dir)
feature_names = extractor.feature_names()
if (
not overwrite
and hdf.complete_hdf_exists(dest_filename_1ms, feature_names)
and hdf.complete_hdf_exists(dest_filename_100ms, feature_names)
):
print "Skipping %s found data files %s" % (input_filename, [dest_filename_1ms, dest_filename_100ms])
else:
process_local_file(input_filename, dest_filename_1ms, dest_filename_100ms, max_books, heap_profile)
else:
print "Unknown suffix for", filename
def process_s3_file(
input_bucket_name, input_key_name, output_bucket_name_1ms=None, output_bucket_name_100ms=None, overwrite=False
):
input_filename = cloud_helpers.download_file_from_s3(input_bucket_name, input_key_name)
dest_1ms, dest_100ms = output_filenames(input_filename, cloud_helpers.scratch_dir())
out_key_name_1ms = os.path.split(dest_1ms)[1]
out_key_name_100ms = os.path.split(dest_100ms)[1]
if output_bucket_name_1ms is None:
output_bucket_name_1ms = input_bucket_name + "-hdf-1ms"
if output_bucket_name_100ms is None:
output_bucket_name_100ms = input_bucket_name + "-hdf"
feature_names = extractor.feature_names()
if (
not overwrite
and cloud_helpers.hdf_already_on_s3(output_bucket_name_1ms, out_key_name_1ms, feature_names)
and cloud_helpers.hdf_already_on_s3(output_bucket_name_100ms, out_key_name_100ms, feature_names)
):
print "HDFs on S3 have same features, so skipping this input..."
return
else:
print "HDFs either not on S3 or have different features..."
# In some weird situation we might have a local copy of the HDF already
# finished but it just might not have been uploaded yet
if (
not overwrite
and hdf.complete_hdf_exists(dest_1ms, feature_names)
and hdf.complete_hdf_exists(dest_100ms, feature_names)
):
print "Found finished HDFs on local storage..."
header = hdf.header_from_hdf_filename(dest_1ms)
else:
print "Running feature generator..."
header = process_local_file(input_filename, dest_1ms, dest_100ms)
print "Header:", header
print "Uploading 1ms feature file", dest_1ms, "to", output_bucket_name_1ms, "/", out_key_name_1ms
cloud_helpers.upload_file_to_s3(dest_1ms, output_bucket_name_1ms, out_key_name_1ms, header)
print "Uploading 100ms feature file", dest_100ms, "to", output_bucket_name_100ms, "/", out_key_name_100ms
cloud_helpers.upload_file_to_s3(dest_100ms, output_bucket_name_100ms, out_key_name_100ms, header)
def process_s3_file(input_bucket_name, input_key_name,
output_bucket_name_1ms = None,
output_bucket_name_100ms = None,
overwrite = False):
input_filename = cloud_helpers.download_file_from_s3(input_bucket_name, input_key_name)
dest_1ms, dest_100ms = output_filenames(input_filename, cloud_helpers.scratch_dir())
out_key_name_1ms = os.path.split(dest_1ms)[1]
out_key_name_100ms = os.path.split(dest_100ms)[1]
if output_bucket_name_1ms is None:
output_bucket_name_1ms = input_bucket_name + "-hdf-1ms"
if output_bucket_name_100ms is None:
output_bucket_name_100ms = input_bucket_name + "-hdf"
feature_names = extractor.feature_names()
if not overwrite and \
cloud_helpers.hdf_already_on_s3(output_bucket_name_1ms, out_key_name_1ms, feature_names) and \
cloud_helpers.hdf_already_on_s3(output_bucket_name_100ms, out_key_name_100ms, feature_names):
print "HDFs on S3 have same features, so skipping this input..."
return
else:
print "HDFs either not on S3 or have different features..."
# In some weird situation we might have a local copy of the HDF already
# finished but it just might not have been uploaded yet
if not overwrite and hdf.complete_hdf_exists(dest_1ms, feature_names) and \
hdf.complete_hdf_exists(dest_100ms, feature_names):
print "Found finished HDFs on local storage..."
header = hdf.header_from_hdf_filename(dest_1ms)
else:
print "Running feature generator..."
header = process_local_file(input_filename, dest_1ms, dest_100ms)
print "Header:", header
print "Uploading 1ms feature file", dest_1ms, "to", output_bucket_name_1ms, "/", out_key_name_1ms
cloud_helpers.upload_file_to_s3(\
dest_1ms, output_bucket_name_1ms, out_key_name_1ms, header)
print "Uploading 100ms feature file", dest_100ms, "to", output_bucket_name_100ms, "/", out_key_name_100ms
cloud_helpers.upload_file_to_s3(\
dest_100ms, output_bucket_name_100ms, out_key_name_100ms, header)
def process_local_dir(input_path, output_dir = None, max_books = None,
heap_profile=False, overwrite = False):
if not os.path.exists(input_path):
print "Specified path does not exist: ", input_path
exit()
if os.path.isdir(input_path):
files = os.listdir(input_path)
basedir = input_path
else:
basedir = os.path.split(input_path)[0]
files = [os.path.basename(input_path)]
count = 0
for filename in files:
if filename.endswith('.csv') or filename.endswith('.csv.gz'):
count += 1
input_filename = os.path.join(basedir, filename)
print "----"
print "Processing #", count, " : ", input_filename
dest_filename_1ms, dest_filename_100ms = \
output_filenames (input_filename, output_dir)
feature_names = extractor.feature_names()
if not overwrite and \
hdf.complete_hdf_exists(dest_filename_1ms, feature_names) and \
hdf.complete_hdf_exists(dest_filename_100ms, feature_names):
print "Skipping %s found data files %s" \
% (input_filename, [dest_filename_1ms, dest_filename_100ms])
else:
process_local_file(input_filename,
dest_filename_1ms, dest_filename_100ms,
max_books, heap_profile)
else:
print "Unknown suffix for", filename
