def identify_audio(filename):
"""
Identify a given song.
You can provide a filename of some audio.
You must provide start time to end time for
scanning.
"""
cmd = '{} "{}" > song.out'.format(EP_PATH, filename)
os.system(cmd)
with open('song.out', 'rb') as fp:
data = fp.read()
fp.close()
data = json.loads(data.strip())[0]
print json.dumps(data, indent=3)
fp_code = data["code"]
cmd = 'curl -S http://localhost:8080/query?fp_code={} > song.out'.format(fp_code)
os.system(cmd)
with open('song.out', 'rb') as fp:
data = fp.read()
fp.close()
data = json.loads(data.strip())
print json.dumps(data, indent=3)
if data["match"]:
print "SUCCESSFUL MATCH!"
title = songMap[str(data["track_id"])]
print "Matched to: {}".format(title)
return {
"success": True,
"match": title,
"score": data["score"]
};
else:
return {
"success": False
};
def read_id(): global lastId try: with open(ID_FNAME, 'rb') as fp: lastId = fp.read() fp.close() except IOError: lastId = "songid1" pass
def do_GET(self):
print self.path
if self.path == '/identify':
self.send_response(200)
self.send_header('Content-type', 'text/json')
self.end_headers()
self.wfile.write(json.dumps(identify_audio("file.mp3")))
elif self.path == '/learn':
self.send_response(200)
self.send_header('Content-type', 'text/html')
self.end_headers()
with open("learn.title", "rb") as fp:
title = fp.read()
fp.close()
learn_song("file.mp3", title)
self.wfile.write('OK')
def learn_song(filename, description):
global lastId
cmd = '{} "{}" > song.out'.format(EP_PATH, filename)
with open('song.out', 'rb') as fp:
subprocess.call(cmd, shell=True, stdout=fp)
os.system(cmd)
data = fp.read()
fp.close()
data = json.loads(data.strip())[0]
#print json.dumps(data, indent=3)
try:
fp_code = data["code"]
except KeyError:
print "Could not learn song with title: {}".format(description)
return
length = data["metadata"]["duration"]
cmd = 'curl http://localhost:8080/ingest -d "fp_code={}&track_id={}&length={}&codever=4.12" > /dev/null'.format(fp_code, lastId, length)
songMap[lastId] = description
lastId = 'songid' + str(int(lastId.split('songid')[1]) + 1)
save_id()
save_map()
os.system(cmd)
print "Learned song with title: {}".format(description)
def batch_test():
# 参数表： 测试图片集路径，结果存储路径，发票类型
curr = time.strftime('%Y.%m.%d', time.localtime(time.time()))
do_ocr = "y"
if len(sys.argv) == 1:
images_dir = input("Please enter pictures dir:")
result_save_path = input("Please enter result save path:")
invoicetype = input(
"Please enter the invoice type(special=01, normal=04):")
elif len(sys.argv) == 4:
images_dir = sys.argv[1]
result_save_path = sys.argv[2]
invoicetype = sys.argv[3]
elif len(sys.argv) == 5:
images_dir = sys.argv[1]
result_save_path = sys.argv[2]
invoicetype = sys.argv[3]
do_ocr = sys.argv[4]
else:
print("Wrong input args.\n")
return
root_path = result_save_path
result_save_path = os.path.join(result_save_path, curr)
if not os.path.exists(result_save_path):
os.makedirs(result_save_path)
image_list = glob.glob(os.path.join(images_dir, "*.jpg"))
if len(image_list) == 0:
image_list = glob.glob(os.path.join(images_dir, "*.bmp"))
if len(image_list) == 0:
print("No pics in " + images_dir + " !\n")
return
num_of_imges = len(image_list)
if do_ocr == "y" or do_ocr == "Y":
# 是否执行ocr操作
print("Now begin to ocr...")
ocr_result = []
ocr_erorr_list = []
for img in image_list:
try:
tuple_info = json.loads(init(img))
tuple_info['filename'] = img
ocr_result.append(tuple_info)
except Exception as _:
ocr_erorr_list.append(img)
print("ocr is complete, get %d / %d pics info." %
(len(ocr_result), num_of_imges))
file_ocr_result = os.path.join(result_save_path, "ocr_result.json")
file_ocr_error_list = os.path.join(result_save_path,
"ocr_error_files.txt")
print("Now saving the result...")
with open(file_ocr_result, 'w') as re, open(file_ocr_error_list,
'w') as er:
re.write(json.dumps(ocr_result, indent=4, ensure_ascii=False))
for i in ocr_erorr_list:
er.writelines(i + '\n')
else:
# 不执行ocr直接用已有结果对比chinese_info.json
file_ocr_result = os.path.join(result_save_path, "ocr_result.json")
with open(file_ocr_result, "r", encoding="utf-8") as fp:
content = fp.read()
content = content.replace('¥', '￥')
if content.startswith(u'\ufeff'):
content = content.encode('utf8')[3:].decode('utf8')
ocr_result = json.loads(content)
print("Chinese infor query begin...")
succeed, query_error = Get_Chinese_Info(ocr_result, result_save_path,
invoicetype, root_path)
file_result = str(succeed) + "(%d)_ocr_correct.txt" % num_of_imges
file_result = os.path.join(result_save_path, file_result)
try:
with open(file_result, "w") as wp:
wp.writelines(file_result + " %.2f " % (succeed / num_of_imges) +
'\n')
wp.writelines("%d not find in web：没有查询到中文数据（不一定ocr错误）" %
(num_of_imges - succeed) + '\n')
wp.writelines("%d query_error_pics：本地有中文数据，但五元组错误" %
(query_error) + '\n')
# wp.writelines("%d ocr_error_files：ocr没有正确生成五元组"%(len(ocr_erorr_list))+'\n')
except Exception as e:
print(e)
finally:
print("Test end.")