import shutil

import os

import argparse

import pandas as pd

import data_common

description = '''

parser = argparse.ArgumentParser(description=description,

formatter_class=argparse.RawTextHelpFormatter)

parser.add_argument('type_name', action="store", help=u'android或者server')

parser.add_argument('filename', action="store", help=u'测试结果文件')

parser.add_argument('--version', action='version',

version='%(prog)s 1.0 Rongzhong xu 2018 06 15')

options = parser.parse_args()

i = 1

file_list = []

results = []

detection_results = []

if options.type_name == 'server':

df = pd.read_table(options.filename, sep='\s',

names=['name','left', 'top', 'width', 'height', 'sep', 'value'])

rename = lambda x: os.path.basename(x)

df['name'] = df['name'].apply(rename)

for num in range(len(df)):

row = df.loc[num]

out = "# {}\n{}\n3 640 480 1\n0\n1\n".format(i, row['name'])

out = out + "1 {} {} {} {}\n".format(row['left'], row['top'], row['left']

+ row['width'], row['top'] + row['height'])

i = i + 1

#print(out)

file_list.append(row['name'])

results.append(out.rstrip('\n'))

if options.type_name == 'android':

print('In android')

df = pd.read_excel(options.filename, usecols=[0,4],names=['name','result'])

rename = lambda x: os.path.basename(x)

df['name'] = df['name'].apply(rename)

print(df.head())

for num in range(len(df)):

row = df.iloc[num]

result = row['result']

if result != '未检测到人脸':

temps = result.split('[')

left, top, = temps[1].strip(']').split(',')

right, bottom = temps[2].strip(']').split(',')

out = "# {}\n{}\n3 640 480 1\n0\n1\n".format(i, row['name'])

out = out + "1 {} {} {} {}\n".format(left, top, right, bottom)

detection_result = "{0} {1} {2} {3} {4} 1 {5}".format(

row['name'],left, top, int(right) - int(left),

int(bottom) - int(top), 0.99)

i = i + 1

#print(out)

file_list.append(row['name'])

results.append(out.rstrip('\n'))

detection_results.append(detection_result)

data_common.output_file("files.txt",file_list)

data_common.output_file("gt.txt",results)

data_common.output_file("dt.txt",detection_results)