Heuristically check submissions to the AES Student Recording Competition comply with the rules.

check-submissions.py

# -*- coding: UTF-8 -*-

# Copyright 2015 Brecht De Man

# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version. 
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# See <http://www.gnu.org/licenses/> for a copy of the GNU General Public License. 


# Made for automatic, heuristic checking of rule compliance of submissions to 
# the AES Student Recording Competitions. 

# INSTRUCTIONS
# Put in a folder (e.g. 'main') which contain folders (e.g. 'main/category 1') which 
# contain the submission folders (e.g. 'main/category 1/87654_traditionalacoustic') 
# or zipped submissions (e.g. 'main/category 1/87654_traditionalacoustic.zip').
# Then run script from Terminal or similar. 

#TODO: more rigorous file name checking (regex)

import os # for getting folder contents
import zipfile # for unzipping folders
import wave # for reading .wav files
import re # for counting pages in PDF document

# This function is where you set the audio constraints.
def check_wav_file(wav_file_path, n_channels): 
    wav_file = wave.open(wav_file_path, 'r')
    alertstr = ''
    # check sample rate
    if wav_file.getframerate() != 48000: 
        alertstr += '\n\t\t\t! Sample rate = '+str(wav_file.getframerate())
    # check bit depth
    if wav_file.getsampwidth()*8 != 24: # note: sample width is in bytes, not bits
        alertstr += '\n\t\t\t! Bit depth = '+str(wav_file.getsampwidth())
    # check number of channels equal to n_channels
    if wav_file.getnchannels() != n_channels: 
        alertstr += '\n\t\t\t! Number of channels = '+str(wav_file.getnchannels())
    # check length of file
    file_length_in_seconds = wav_file.getnframes()/wav_file.getframerate()
    if file_length_in_seconds > 300 or file_length_in_seconds < 180: 
        alertstr += '\n\t\t\t! File length = '+seconds2timestr(file_length_in_seconds)
    return alertstr
    
# Turn number of seconds (int) to '[minutes] min [seconds] s' (string)
def seconds2timestr(time_in_seconds):
    if time_in_seconds is not None: 
        time_in_minutes = int(time_in_seconds/60)
        remaining_seconds = int(time_in_seconds%60)
        return str(time_in_minutes) + " min " + str(remaining_seconds) + " s"
    else:
        return 'N/A'

# Count pages in a PDF document  
rxcountpages = re.compile(b'/Type\s*/Page([^s]|$)', re.MULTILINE|re.DOTALL)  
#rxcountpages = re.compile(r"$/Type\s*/Page([/\s]|$)", re.MULTILINE|re.DOTALL)
def count_pages(filename): 
    f = open(filename, 'rb')
    data = f.read()
    return len(rxcountpages.findall(data))  

category_list = os.listdir('.')
for category in category_list: 
    # check if folder
    if os.path.isdir(category):
        # print category to console
        print(category+': ')
        
        # unzip each zipped file
        for file in os.listdir(category):
            if file.endswith(".zip"):
                zip_ref = zipfile.ZipFile(category+'/'+file, 'r')
                zip_ref.extractall(category+'/'+file[:-4]) # extract to eponymous folder
                zip_ref.close()
                os.remove(category+'/'+file) # remove zip file
        
        # go over each folder in here
        submission_list = os.listdir(category)
        count_submissions = 0
        for submission in submission_list: 
            if os.path.isdir(category+'/'+submission): # check if submission folder is really a folder
                # count total number of submissions for this category
                count_submissions += 1
            
                # print name to console
                print('\t'+submission)
            
                # name: warning if deviates from specified format?
                #TODO 
                
                # get all files in submission folder
                file_list = os.listdir(category+'/'+submission)
                
                # get wav files
                wav_files = [s for s in file_list if s.endswith('.wav') and not s.startswith('.')]
                if len(wav_files) == 1: # stereo mix
                    # check file name convention
                    if not wav_files[0].endswith(('_traditionalacoustic.wav','_traditionalstudio.wav', '_modernstudioelectro.wav', '_soundforvisual.wav')): 
                        print('\t\t! WAV file name: '+wav_files[0])
                    # print any deviations from rules to console
                    alertstr = check_wav_file(category+'/'+submission+'/'+wav_files[0], 2) 
                    if len(alertstr)>0: 
                        print('\t\t'+wav_files[0]+': '+alertstr)
                elif len(wav_files) == 6: # 5.1 mix
                    for wav_file in wav_files: 
                        # check file name convention
                        if not wav_file.endswith(('_(C).wav','_(R).wav', '_(L).wav', '_(Rs).wav', '_(Ls).wav', '_(LFE).wav', )) or\
                           not any(s in wav_file for s in ['_traditionalacoustic_','_traditionalstudio_', 
                               '_modernstudioelectro_', '_soundforvisual_']): 
                            print('\t\t! WAV file name: '+wav_file)
                        # print any deviations from rules to console
                        alertstr =  check_wav_file(category+'/'+submission+'/'+wav_file, 1) 
                        if len(alertstr)>0: 
                            print('\t\t'+wav_file+': '+alertstr)
                else: # anything other than 1 or 6 files
                    print('\t\t! Weird number of .WAV files: '+str(len(wav_files))+' found: '+', '.join(wav_files))
                
                # get PDF files
                pdf_files = [s for s in file_list if s.endswith('.pdf') and not s.startswith('.')]
                if len(pdf_files) == 0: 
                    print('\t\t! No PDF files found.')
                elif len(pdf_files)>1:
                    print('\t\t! More than one PDF file found: '+', '.join(pdf_files))
                else: # read PDF and count pages; alert if over 6
                    # check file name convention
                    if not pdf_files[0].endswith(('_traditionalacoustic.pdf','_traditionalstudio.pdf', '_modernstudioelectro.pdf', '_soundforvisual.pdf')): 
                        print('\t\t! PDF file name: '+pdf_files[0])
                    n_pages = count_pages(category+'/'+submission+'/'+pdf_files[0])
                    if n_pages>6 or n_pages<1: # too long or error
                        print('\t\t! Number of documentation pages: '+str(n_pages))
                
                # show any other file (except hidden files)
                other_files = [s for s in file_list if not (s.endswith('.wav') or s.endswith('.pdf') or s.startswith('.'))]
                for file in other_files: # print to console
                    print('\t\t\t! Extra file: '+file)
                
            
        # show number of submissions for this category
        print('Number of submissions: '+str(count_submissions))

# info on what is checked
print('''Checked: all .wav and .pdf files.
Checked: exactly 1 (stereo) or 6 (5.1) .wav files. 
Checked: all .wav files end in _[categoryname] (plus _(L), _(Rs), _(LFE)... for 5.1).
Checked: all audio 48 kHz / 24 bit.
Checked: exactly 1 PDF file ending in _[categoryname].
Checked: between 1 and 6 documentation pages.''')