task5_2.py

'''
Python Training
'''
import os
import codecs
import argparse
# import xml.etree.ElementTree as ET
from bs4 import BeautifulSoup


INPUT_PATH = 'All_xml'
OUTPUT_PATH = 'task5_output_pcent.txt'

def run(input_path, out_path):
    '''
    start program
    '''
    utteranceDict = {}
    trans_utter = {}
    for filename in os.listdir(input_path):
        # if filename.endswith('.xml') and ('MPS' in filename or 'XboxOneServer' in filename):
        if filename.endswith('.xml') and 'MPS' in filename:
            print(filename)
            with codecs.open(os.path.join(input_path, filename), 'r', encoding='utf-16') as f:
                lines = f.readlines()
                for line in lines:
                    # if 'Recognition Text' in line:
                    if 'Word Text' in line \
                        and 'Pron' in line \
                        and 'Confidence' in line \
                        and 'BS' in line \
                        and 'DR' in line \
                        and 'Correct' in line:
                        temp = line.split('=')
                        trans = temp[1].split('\"')[1]
                        utter = temp[2].split('\"')[1]
                        
                        # utter = line.split('\"')[1]
                        if utter not in utteranceDict:
                            utteranceDict[utter] = 1
                        else:
                            utteranceDict[utter] += 1
                            
                        if 'bing' in trans or 'xbox' in trans:
                            trans_utter[trans] = utter


            # if 'MPS' in filename:
            #     pass

            # soup = BeautifulSoup(codecs.open(os.path.join(input_path, filename), 'r', encoding='utf-16'), "lxml")
            # print(soup.tag)
            # xmlp = ET.XMLParser(encoding='utf-16')
            # try:
            #     f = ET.parse(os.path.join(input_path, filename), parser=xmlp)
            #     for elem in f.iter('Recognition'):
            #         utter = elem.get("Text")
            #         if utter is not None and len(utter) > 0 and utter not in utteranceDict:
            #             utteranceDict[utter] = 0
            #         else:
            #             utteranceDict[utter] += 1
            # except (RuntimeError):
            #     print(filename)
    count = 0
    for utter in utteranceDict:
        count += utteranceDict[utter]

    utter_pcent = {}
    for trans in trans_utter:
        utter_pcent[trans_utter[trans]] = float(utteranceDict[trans_utter[trans]]) / count

    print(len(utteranceDict))
    with codecs.open(out_path, 'w', encoding='utf-16') as f:
        for utter in utter_pcent:
            # if utteranceDict[utter] >= 50:
                f.write('\t'.join((utter, "{:10.10f}".format(utter_pcent[utter]))) + '\n')
    

if __name__ == '__main__':
    parser = argparse.ArgumentParser(description='prodcess file')
    parser.add_argument('--input_path', type=str, default=INPUT_PATH, help='input xml file path')
    # parser.add_argument('--input_hyp', type=str, default=INPUT_HYP, help='input hyp file')
    parser.add_argument('--output_path', type=str, default=OUTPUT_PATH, help='output file path')
    args = parser.parse_args()
    run(args.input_path, args.output_path)