def load_lasagne_nn(json_name, weights_name):
info = read_json(json_name)
nn = LasagneNN(architecture=info['arch'],
dim=info['dim'],
params=info['params'])
nn.load_weights(pickle.load(open(weights_name, 'rb'))['weights'])
return info, nn
def register():
teacher_name = input("请输入你的帐号姓名(3~6位):")
teacher_pwd = input("请输入你的帐号密码(6~12位):")
teacher_dict = file_utils.read_json("teacher_data.json", {})
t = modul.Teacher(teacher_name, teacher_pwd)
if teacher_name not in teacher_dict:
teacher_dict[t.name] = t.pwd
file_utils.write_json(teacher_dict, "teacher_data.json")
else:
print("已存在的用户，请重新输入:")
def login():
if file_utils.read_file("teacher_data.json"):
teacher_name = input("请输入你的帐号姓名:")
teacher_pwd = input("请输入你的帐号密码:")
teacher_dict = file_utils.read_json("teacher_data.json", {})
if teacher_dict.get(teacher_name) == teacher_pwd:
student_utils.manager_stu()
else:
print("账户名字或者密码不正确!")
else:
print("当前teacher账户为空，请先注册")
def manager_stu():
while True:
print(file_utils.read_file("student_ui.txt"))
num = input("请选择(1-5):")
student_list = file_utils.read_json("student_data.json", [])
if num == "1":
add_stu(student_list, "student_data.json")
elif num == "2":
search_stu(student_list)
elif num == "3":
mod_stu(student_list)
elif num == "4":
del_stu(student_list)
elif num == "5":
break
else:
print("输入错误")
if i in summary_sentence_indexes:
markdown += ' ***' + raw_sentences[i] + '***'
else:
markdown += raw_sentences[i]
elif mode is TRIM:
if i in summary_sentence_indexes:
markdown += ' ' + raw_sentences[i]
return markdown
def transform_slider_input(x):
return str(int(x/10))
# Read processed summary file
manifest = file_utils.read_json('summary_output.json')
# Get specific pieces of data
audio_url = manifest['audio_url']
sentences = manifest['sentences']
summary_indices = manifest['summary_indices']
# ========== UI ==================================================
streamlit.title('Audio Summarizer')
# Audio Player
streamlit.markdown('## Original Audio')
streamlit.audio(audio_url, format='audio/mp3', start_time=0)
def match_sites_dataframe(dataframe, matches_json="", top_n=5):
'''
Generates a dataframe of matched sites.
matches_json is an optional parameter for saving and loading slow to generate
description based matches.
INPUTS:
- dataframe
- matches_json -- A string representing the filename of a json file containing old matches to speed up processing
- top_n (int) -- Maximum amount of matches to return for each item
OUTPUTS:
- matches_df
'''
#Missing values should be represented by empty strings
dataframe = dataframe.fillna(value="")
#Ensure we have the correct columns
dataframe = pandas.DataFrame(dataframe.to_dict("records"),
columns=ALL_FIELDNAMES)
#Fill any columns we just added with "-1" to mark it wasn't originally there
dataframe = dataframe.fillna(value="-1")
#Make sure everything in that dataframe is a string
dataframe = dataframe.applymap(lambda x: str(x))
#Remove extra whitespace
dataframe = dataframe.applymap(lambda x: x.strip()
if type(x) == str else x)
if "Match Site" in dataframe.columns:
ndf = dataframe[dataframe["Match Site"] == "-1"]
if ndf.empty:
#No new rows.
return pandas.DataFrame()
odf = dataframe[dataframe["Match Site"] != "-1"]
if odf.empty:
old_rows = []
else:
old_rows = odf.to_dict("records")
new_rows = ndf.to_dict("records")
else:
new_rows = dataframe.to_dict("records")
old_rows = []
# Add a 'Description' field to new_rows
site_rows = [{
**row, "Description": row["Stock Description"]
} for row in new_rows]
old_site_rows = remove_duplicate_rows(old_rows)
old_item_ids_to_rows = generate_item_ids_to_rows(old_rows)
# Generate desc_matches based on matches_json
desc_matches = {}
if matches_json:
if file_utils.file_exists(matches_json):
desc_matches = file_utils.read_json(matches_json)
else:
desc_matches = match_by_description(site_rows, old_site_rows)
file_utils.save_json(matches_json, desc_matches)
matches_rows = match_sites(site_rows,
old_site_rows,
old_item_ids_to_rows,
desc_matches,
top_n=top_n)
matches_df = pandas.DataFrame(matches_rows, columns=OUTPUT_FIELDNAMES)
matches_df = matches_df.fillna(value="")
matches_df = matches_df[OUTPUT_FIELDNAMES]
return matches_df