def setup(logname):
from src.modules.utils.config_loader import get_config
conf = get_config()
# associate command line arguments with logging levels
log_dict = {
'debug': logging.DEBUG,
'info': logging.INFO,
'warning': logging.WARNING,
'error': logging.ERROR,
'critical': logging.CRITICAL
}
# currently parses only the logging level, but you can add more arguments as well (see python docs on logging)
parser = argparse.ArgumentParser(usage='%(prog)s [-l]')
parser.add_argument('-l',
'--log',
nargs='?',
choices=log_dict.keys(),
default='info',
const='info',
help="logging level for console")
args, extra_args = parser.parse_known_args()
# create logging file if doesn't exist
log_file_path = os.path.join(conf.dirs.logs,
os.getlogin() + "_" + logname + ".log")
if not os.path.exists(conf.dirs.logs):
os.makedirs(conf.dirs.logs)
if not os.path.exists(log_file_path):
open(log_file_path, "w").close()
# set up logging to file (file always does debug level)
logging.basicConfig(level=logging.DEBUG,
format='%(asctime)s %(levelname)-4s|%(message)s',
datefmt='%Y/%m/%d %H:%M:%S',
filename=os.path.join(
conf.dirs.logs,
os.getlogin() + "_" + logname + ".log"),
filemode='a')
# set up logging to console (only do this if there are no handlers)
if len(logging.getLogger('').handlers) < 2:
console = logging.StreamHandler()
console.setLevel(log_dict[args.log])
formatter = logging.Formatter(
'%(asctime)s %(levelname)-4s|%(message)s', '%Y/%m/%d %H:%M:%S')
console.formatter = formatter
logging.getLogger('').addHandler(console)
logging.info(
"============================================================")
return conf
# ====== External package imports ================
import pandas as pd
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
import numpy as np
# ====== Internal package imports ================
from src.modules.data.database_io import DatabaseIO
from src.modules.utils.misc import run_and_catch_exceptions
# ============== Logging ========================
import logging
from src.modules.utils.setup import setup, IndentLogger
logger = IndentLogger(logging.getLogger(''), {})
# =========== Config File Loading ================
from src.modules.utils.config_loader import get_config
conf = get_config()
# ================================================
def interval_overlap(interval_a, interval_b):
"""Given two intervals on the real line, determine their overlap
:param interval_a: tuple of (left, right) for first point
:param interval_b: tuple of (left, right) for second point
:return: float which is overlap
"""
x1, x2 = interval_a
x3, x4 = interval_b
if x3 < x1:
if x4 < x1:
import glob
import os
# ####### Logging and config file loading ########
import logging
from src.modules.utils.setup import setup, IndentLogger
logger = IndentLogger(logging.getLogger(''), {})
# ########### Config File Loading ################
from src.modules.utils.config_loader import get_config
conf, confp = get_config()
##################################################
conf, confp = setup("process_subtitles")
# import / output files
subtitle_dir = confp.dirs.subtitles
output_file_name = "video_metadata_full.csv"
# create metadata file
f_out = open(subtitle_dir + output_file_name, 'w')
f_out.write("file,time_stamp,date_and_time\n")
for i, f_path in enumerate(glob.glob(subtitle_dir + "*.srt")):
with open(f_path) as f:
f_name = os.path.basename(f_path).split(".")[0]
print("Collecting metadata from file {}:{}".format(i, f_name))
for line in f.readlines():
if "AM" in line or "PM" in line:
f_out.write(line.replace(",", "").replace("\n", "") + '\n')
elif '-->' in line:
f_out.write(f_name + ',' + line[:8] + ',')
f_out.close()
