Spaces:
Running
Running
pull0322
Browse files- __pycache__/analyze.cpython-310.pyc +0 -0
- __pycache__/transcription.cpython-310.pyc +0 -0
- app.py +223 -222
__pycache__/analyze.cpython-310.pyc
CHANGED
Binary files a/__pycache__/analyze.cpython-310.pyc and b/__pycache__/analyze.cpython-310.pyc differ
|
|
__pycache__/transcription.cpython-310.pyc
CHANGED
Binary files a/__pycache__/transcription.cpython-310.pyc and b/__pycache__/transcription.cpython-310.pyc differ
|
|
app.py
CHANGED
@@ -1,223 +1,224 @@
|
|
1 |
-
from flask import Flask, request, jsonify, render_template, send_from_directory
|
2 |
-
import base64
|
3 |
-
from pydub import AudioSegment # 変換用にpydubをインポート
|
4 |
-
import os
|
5 |
-
import shutil
|
6 |
-
from process import AudioProcessor
|
7 |
-
from transcription import TranscriptionMaker
|
8 |
-
from analyze import TextAnalyzer
|
9 |
-
import json
|
10 |
-
|
11 |
-
process=AudioProcessor()
|
12 |
-
transcripter = TranscriptionMaker()
|
13 |
-
app = Flask(__name__)
|
14 |
-
|
15 |
-
users = []
|
16 |
-
transcription_text=""
|
17 |
-
harassment_keywords = [
|
18 |
-
"バカ", "馬鹿", "
|
19 |
-
"きもい", "キモい", "ブス", "デブ", "ハゲ",
|
20 |
-
"セクハラ", "パワハラ", "モラハラ"
|
21 |
-
]
|
22 |
-
total_audio = ""
|
23 |
-
|
24 |
-
# トップページ(テンプレート: index.html)
|
25 |
-
@app.route('/')
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
@app.route('/userregister', methods=['GET', 'POST'])
|
42 |
-
def userregister():
|
43 |
-
return render_template('userRegister.html')
|
44 |
-
|
45 |
-
#人数確認
|
46 |
-
@app.route('/confirm', methods=['GET']) # 基本的にGETで取得する想定なので、GETのみに変更
|
47 |
-
def confirm():
|
48 |
-
return jsonify({'members': users}), 200
|
49 |
-
|
50 |
-
#リセット画面(テンプレート: reset.html)
|
51 |
-
@app.route('/reset_html', methods=['GET', 'POST'])
|
52 |
-
def reset_html():
|
53 |
-
return render_template('reset.html')
|
54 |
-
|
55 |
-
#メンバー削除&累積音声削除
|
56 |
-
@app.route('/reset_member', methods=['GET', 'POST'])
|
57 |
-
def reset_member():
|
58 |
-
global users
|
59 |
-
global total_audio
|
60 |
-
print(total_audio)
|
61 |
-
process.delete_files_in_directory(total_audio)
|
62 |
-
try:
|
63 |
-
data = request.get_json()
|
64 |
-
if not data or "names" not in data:
|
65 |
-
return jsonify({"status": "error", "message": "Invalid request body"}), 400 # 400 Bad Request
|
66 |
-
|
67 |
-
names = data.get("names", [])
|
68 |
-
base_audio_dir = "/tmp/data/base_audio"
|
69 |
-
|
70 |
-
for name in names:
|
71 |
-
file_path = os.path.join(base_audio_dir, f"{name}.wav")
|
72 |
-
if os.path.exists(file_path):
|
73 |
-
try:
|
74 |
-
os.remove(file_path)
|
75 |
-
print(f"{file_path} を削除しました。")
|
76 |
-
except Exception as e:
|
77 |
-
print(f"削除中にエラーが発生しました: {e}")
|
78 |
-
# ファイル削除に失敗した場合も、エラーを返す
|
79 |
-
return jsonify({"status": "error", "message": f"Failed to delete {name}: {e}"}), 500
|
80 |
-
|
81 |
-
else:
|
82 |
-
print(f"ファイルが存在しません: {file_path}")
|
83 |
-
|
84 |
-
# usersリストを更新
|
85 |
-
users = [u for u in users if u not in names]
|
86 |
-
|
87 |
-
# 成功した場合のレスポンス
|
88 |
-
return jsonify({"status": "success", "message": "Members deleted successfully", "users": users}), 200 # 200 OK
|
89 |
-
|
90 |
-
except Exception as e:
|
91 |
-
print(f"An unexpected error occurred: {e}")
|
92 |
-
return jsonify({"status": "error", "message": f"Internal server error: {e}"}), 500 # 500 Internal Server Error
|
93 |
-
|
94 |
-
# 書き起こし作成エンドポイント
|
95 |
-
@app.route('/transcription',methods =['GET','POST'])
|
96 |
-
def transcription():
|
97 |
-
global transcription_text
|
98 |
-
global total_audio
|
99 |
-
try:
|
100 |
-
audio_directory = transcripter.merge_segments(total_audio)
|
101 |
-
transcription_text = transcripter.create_transcription(audio_directory)
|
102 |
-
with open(transcription_text,'r',encoding='utf-8') as file:
|
103 |
-
file_content = file.read()
|
104 |
-
print(file_content)
|
105 |
-
return jsonify({'transcription': file_content}),200
|
106 |
-
except Exception as e:
|
107 |
-
return jsonify({"error": str(e)}),500
|
108 |
-
|
109 |
-
# AI分析エンドポイント
|
110 |
-
@app.route('/analyze',methods =['GET','POST'])
|
111 |
-
def analyze():
|
112 |
-
global transcription_text
|
113 |
-
analyzer = TextAnalyzer(transcription_text, harassment_keywords)
|
114 |
-
api_key = os.environ.get("DEEPSEEK")
|
115 |
-
if api_key is None:
|
116 |
-
raise ValueError("DEEPSEEK_API_KEY が設定されていません。")
|
117 |
-
|
118 |
-
results = analyzer.analyze(api_key=api_key)
|
119 |
-
|
120 |
-
print(json.dumps(results, ensure_ascii=False, indent=2))
|
121 |
-
|
122 |
-
if "deepseek_analysis" in results and results["deepseek_analysis"]:
|
123 |
-
deepseek_data = results["deepseek_analysis"]
|
124 |
-
conversation_level = deepseek_data.get("conversationLevel")
|
125 |
-
harassment_present = deepseek_data.get("harassmentPresent")
|
126 |
-
harassment_type = deepseek_data.get("harassmentType")
|
127 |
-
repetition = deepseek_data.get("repetition")
|
128 |
-
pleasantConversation = deepseek_data.get("pleasantConversation")
|
129 |
-
blameOrHarassment = deepseek_data.get("blameOrHarassment")
|
130 |
-
|
131 |
-
print("\n--- DeepSeek 分析結果 ---")
|
132 |
-
print(f"会話レベル: {conversation_level}")
|
133 |
-
print(f"ハラスメントの有無: {harassment_present}")
|
134 |
-
print(f"ハラスメントの種類: {harassment_type}")
|
135 |
-
print(f"繰り返しの程度: {repetition}")
|
136 |
-
print(f"会話の心地よさ: {pleasantConversation}")
|
137 |
-
print(f"非難またはハラスメントの程度: {blameOrHarassment}")
|
138 |
-
return jsonify({"results": results}),200
|
139 |
-
|
140 |
-
|
141 |
-
#
|
142 |
-
@app.route('/upload_audio', methods=['POST'])
|
143 |
-
def upload_audio():
|
144 |
-
global total_audio
|
145 |
-
try:
|
146 |
-
data = request.get_json()
|
147 |
-
# name か users のいずれかが必須。どちらも無い場合はエラー
|
148 |
-
if not data or 'audio_data' not in data or ('name' not in data and 'users' not in data):
|
149 |
-
return jsonify({"error": "音声データまたは名前がありません"}), 400
|
150 |
-
|
151 |
-
# Base64デコードして音声バイナリを取得
|
152 |
-
audio_binary = base64.b64decode(data['audio_data'])
|
153 |
-
|
154 |
-
upload_name = 'tmp'
|
155 |
-
audio_dir = "/tmp/data"
|
156 |
-
os.makedirs(audio_dir, exist_ok=True)
|
157 |
-
audio_path = os.path.join(audio_dir, f"{upload_name}.wav")
|
158 |
-
with open(audio_path, 'wb') as f:
|
159 |
-
f.write(audio_binary)
|
160 |
-
print(users)
|
161 |
-
# 各ユーザーの参照音声ファイルのパスをリストに格納
|
162 |
-
reference_paths = []
|
163 |
-
base_audio_dir = "/tmp/data/base_audio"
|
164 |
-
for user in users:
|
165 |
-
ref_path = os.path.abspath(os.path.join(base_audio_dir, f"{user}.wav"))
|
166 |
-
if not os.path.exists(ref_path):
|
167 |
-
return jsonify({"error": "参照音声ファイルが見つかりません", "details": ref_path}), 500
|
168 |
-
reference_paths.append(ref_path)
|
169 |
-
|
170 |
-
# 複数人の場合は参照パスのリストを、1人の場合は単一のパスを渡す
|
171 |
-
if len(users) > 1:
|
172 |
-
print("複数人の場合の処理")
|
173 |
-
matched_times, segments_dir = process.process_multi_audio(reference_paths, audio_path, threshold=0.05)
|
174 |
-
total_audio = transcripter.merge_segments(segments_dir)
|
175 |
-
# 各メンバーのrateを計算
|
176 |
-
total_time = sum(matched_times)
|
177 |
-
rates = [(time / total_time) * 100 if total_time > 0 else 0 for time in matched_times]
|
178 |
-
return jsonify({"rates": rates}), 200
|
179 |
-
else:
|
180 |
-
matched_time, unmatched_time, segments_dir = process.process_audio(reference_paths[0], audio_path, threshold=0.05)
|
181 |
-
total_audio = transcripter.merge_segments(segments_dir)
|
182 |
-
total_time = matched_time + unmatched_time
|
183 |
-
rate = (matched_time / total_time) * 100 if total_time > 0 else 0
|
184 |
-
return jsonify({"rate": rate}), 200
|
185 |
-
except Exception as e:
|
186 |
-
print("Error in /upload_audio:", str(e))
|
187 |
-
return jsonify({"error": "サーバーエラー", "details": str(e)}), 500
|
188 |
-
|
189 |
-
@app.route('/reset', methods=['GET'])
|
190 |
-
def reset():
|
191 |
-
global users
|
192 |
-
users=[]
|
193 |
-
return 200
|
194 |
-
|
195 |
-
|
196 |
-
|
197 |
-
|
198 |
-
|
199 |
-
|
200 |
-
|
201 |
-
|
202 |
-
|
203 |
-
|
204 |
-
|
205 |
-
|
206 |
-
|
207 |
-
users
|
208 |
-
|
209 |
-
|
210 |
-
|
211 |
-
|
212 |
-
|
213 |
-
|
214 |
-
|
215 |
-
|
216 |
-
|
217 |
-
|
218 |
-
|
219 |
-
|
220 |
-
|
221 |
-
|
222 |
-
|
|
|
223 |
app.run(debug=True, host="0.0.0.0", port=port)
|
|
|
1 |
+
from flask import Flask, request, jsonify, render_template, send_from_directory
|
2 |
+
import base64
|
3 |
+
from pydub import AudioSegment # 変換用にpydubをインポート
|
4 |
+
import os
|
5 |
+
import shutil
|
6 |
+
from process import AudioProcessor
|
7 |
+
from transcription import TranscriptionMaker
|
8 |
+
from analyze import TextAnalyzer
|
9 |
+
import json
|
10 |
+
|
11 |
+
process=AudioProcessor()
|
12 |
+
transcripter = TranscriptionMaker()
|
13 |
+
app = Flask(__name__)
|
14 |
+
|
15 |
+
users = []
|
16 |
+
transcription_text=""
|
17 |
+
harassment_keywords = [
|
18 |
+
"バカ", "馬鹿", "���ホ", "死ね", "クソ", "うざい",
|
19 |
+
"きもい", "キモい", "ブス", "デブ", "ハゲ",
|
20 |
+
"セクハラ", "パワハラ", "モラハラ"
|
21 |
+
]
|
22 |
+
total_audio = ""
|
23 |
+
|
24 |
+
# トップページ(テンプレート: index.html)
|
25 |
+
@app.route('/index', methods=['GET', 'POST'])
|
26 |
+
def index():
|
27 |
+
return render_template('index.html', users = users)
|
28 |
+
|
29 |
+
# フィードバック画面(テンプレート: feedback.html)
|
30 |
+
@app.route('/feedback', methods=['GET', 'POST'])
|
31 |
+
def feedback():
|
32 |
+
return render_template('feedback.html')
|
33 |
+
|
34 |
+
# 会話詳細画面(テンプレート: talkDetail.html)
|
35 |
+
@app.route('/talk_detail', methods=['GET', 'POST'])
|
36 |
+
def talk_detail():
|
37 |
+
return render_template('talkDetail.html')
|
38 |
+
|
39 |
+
# 音声登録画面(テンプレート: userRegister.html)
|
40 |
+
@app.route('/')
|
41 |
+
@app.route('/userregister', methods=['GET', 'POST'])
|
42 |
+
def userregister():
|
43 |
+
return render_template('userRegister.html')
|
44 |
+
|
45 |
+
#人数確認
|
46 |
+
@app.route('/confirm', methods=['GET']) # 基本的にGETで取得する想定なので、GETのみに変更
|
47 |
+
def confirm():
|
48 |
+
return jsonify({'members': users}), 200
|
49 |
+
|
50 |
+
#リセット画面(テンプレート: reset.html)
|
51 |
+
@app.route('/reset_html', methods=['GET', 'POST'])
|
52 |
+
def reset_html():
|
53 |
+
return render_template('reset.html')
|
54 |
+
|
55 |
+
#メンバー削除&累積音声削除
|
56 |
+
@app.route('/reset_member', methods=['GET', 'POST'])
|
57 |
+
def reset_member():
|
58 |
+
global users
|
59 |
+
global total_audio
|
60 |
+
print(total_audio)
|
61 |
+
process.delete_files_in_directory(total_audio)
|
62 |
+
try:
|
63 |
+
data = request.get_json()
|
64 |
+
if not data or "names" not in data:
|
65 |
+
return jsonify({"status": "error", "message": "Invalid request body"}), 400 # 400 Bad Request
|
66 |
+
|
67 |
+
names = data.get("names", [])
|
68 |
+
base_audio_dir = "/tmp/data/base_audio"
|
69 |
+
|
70 |
+
for name in names:
|
71 |
+
file_path = os.path.join(base_audio_dir, f"{name}.wav")
|
72 |
+
if os.path.exists(file_path):
|
73 |
+
try:
|
74 |
+
os.remove(file_path)
|
75 |
+
print(f"{file_path} を削除しました。")
|
76 |
+
except Exception as e:
|
77 |
+
print(f"削除中にエラーが発生しました: {e}")
|
78 |
+
# ファイル削除に失敗した場合も、エラーを返す
|
79 |
+
return jsonify({"status": "error", "message": f"Failed to delete {name}: {e}"}), 500
|
80 |
+
|
81 |
+
else:
|
82 |
+
print(f"ファイルが存在しません: {file_path}")
|
83 |
+
|
84 |
+
# usersリストを更新
|
85 |
+
users = [u for u in users if u not in names]
|
86 |
+
|
87 |
+
# 成功した場合のレスポンス
|
88 |
+
return jsonify({"status": "success", "message": "Members deleted successfully", "users": users}), 200 # 200 OK
|
89 |
+
|
90 |
+
except Exception as e:
|
91 |
+
print(f"An unexpected error occurred: {e}")
|
92 |
+
return jsonify({"status": "error", "message": f"Internal server error: {e}"}), 500 # 500 Internal Server Error
|
93 |
+
|
94 |
+
# 書き起こし作成エンドポイント
|
95 |
+
@app.route('/transcription',methods =['GET','POST'])
|
96 |
+
def transcription():
|
97 |
+
global transcription_text
|
98 |
+
global total_audio
|
99 |
+
try:
|
100 |
+
audio_directory = transcripter.merge_segments(total_audio)
|
101 |
+
transcription_text = transcripter.create_transcription(audio_directory)
|
102 |
+
with open(transcription_text,'r',encoding='utf-8') as file:
|
103 |
+
file_content = file.read()
|
104 |
+
print(file_content)
|
105 |
+
return jsonify({'transcription': file_content}),200
|
106 |
+
except Exception as e:
|
107 |
+
return jsonify({"error": str(e)}),500
|
108 |
+
|
109 |
+
# AI分析エンドポイント
|
110 |
+
@app.route('/analyze',methods =['GET','POST'])
|
111 |
+
def analyze():
|
112 |
+
global transcription_text
|
113 |
+
analyzer = TextAnalyzer(transcription_text, harassment_keywords)
|
114 |
+
api_key = os.environ.get("DEEPSEEK")
|
115 |
+
if api_key is None:
|
116 |
+
raise ValueError("DEEPSEEK_API_KEY が設定されていません。")
|
117 |
+
|
118 |
+
results = analyzer.analyze(api_key=api_key)
|
119 |
+
|
120 |
+
print(json.dumps(results, ensure_ascii=False, indent=2))
|
121 |
+
|
122 |
+
if "deepseek_analysis" in results and results["deepseek_analysis"]:
|
123 |
+
deepseek_data = results["deepseek_analysis"]
|
124 |
+
conversation_level = deepseek_data.get("conversationLevel")
|
125 |
+
harassment_present = deepseek_data.get("harassmentPresent")
|
126 |
+
harassment_type = deepseek_data.get("harassmentType")
|
127 |
+
repetition = deepseek_data.get("repetition")
|
128 |
+
pleasantConversation = deepseek_data.get("pleasantConversation")
|
129 |
+
blameOrHarassment = deepseek_data.get("blameOrHarassment")
|
130 |
+
|
131 |
+
print("\n--- DeepSeek 分析結果 ---")
|
132 |
+
print(f"会話レベル: {conversation_level}")
|
133 |
+
print(f"ハラスメントの有無: {harassment_present}")
|
134 |
+
print(f"ハラスメントの種類: {harassment_type}")
|
135 |
+
print(f"繰り返しの程度: {repetition}")
|
136 |
+
print(f"会話の心地よさ: {pleasantConversation}")
|
137 |
+
print(f"非難またはハラスメントの程度: {blameOrHarassment}")
|
138 |
+
return jsonify({"results": results}),200
|
139 |
+
|
140 |
+
|
141 |
+
# 音声アップロード&解析エ��ドポイント
|
142 |
+
@app.route('/upload_audio', methods=['POST'])
|
143 |
+
def upload_audio():
|
144 |
+
global total_audio
|
145 |
+
try:
|
146 |
+
data = request.get_json()
|
147 |
+
# name か users のいずれかが必須。どちらも無い場合はエラー
|
148 |
+
if not data or 'audio_data' not in data or ('name' not in data and 'users' not in data):
|
149 |
+
return jsonify({"error": "音声データまたは名前がありません"}), 400
|
150 |
+
|
151 |
+
# Base64デコードして音声バイナリを取得
|
152 |
+
audio_binary = base64.b64decode(data['audio_data'])
|
153 |
+
|
154 |
+
upload_name = 'tmp'
|
155 |
+
audio_dir = "/tmp/data"
|
156 |
+
os.makedirs(audio_dir, exist_ok=True)
|
157 |
+
audio_path = os.path.join(audio_dir, f"{upload_name}.wav")
|
158 |
+
with open(audio_path, 'wb') as f:
|
159 |
+
f.write(audio_binary)
|
160 |
+
print(users)
|
161 |
+
# 各ユーザーの参照音声ファイルのパスをリストに格納
|
162 |
+
reference_paths = []
|
163 |
+
base_audio_dir = "/tmp/data/base_audio"
|
164 |
+
for user in users:
|
165 |
+
ref_path = os.path.abspath(os.path.join(base_audio_dir, f"{user}.wav"))
|
166 |
+
if not os.path.exists(ref_path):
|
167 |
+
return jsonify({"error": "参照音声ファイルが見つかりません", "details": ref_path}), 500
|
168 |
+
reference_paths.append(ref_path)
|
169 |
+
|
170 |
+
# 複数人の場合は参照パスのリストを、1人の場合は単一のパスを渡す
|
171 |
+
if len(users) > 1:
|
172 |
+
print("複数人の場合の処理")
|
173 |
+
matched_times, segments_dir = process.process_multi_audio(reference_paths, audio_path, threshold=0.05)
|
174 |
+
total_audio = transcripter.merge_segments(segments_dir)
|
175 |
+
# 各メンバーのrateを計算
|
176 |
+
total_time = sum(matched_times)
|
177 |
+
rates = [(time / total_time) * 100 if total_time > 0 else 0 for time in matched_times]
|
178 |
+
return jsonify({"rates": rates}), 200
|
179 |
+
else:
|
180 |
+
matched_time, unmatched_time, segments_dir = process.process_audio(reference_paths[0], audio_path, threshold=0.05)
|
181 |
+
total_audio = transcripter.merge_segments(segments_dir)
|
182 |
+
total_time = matched_time + unmatched_time
|
183 |
+
rate = (matched_time / total_time) * 100 if total_time > 0 else 0
|
184 |
+
return jsonify({"rate": rate}), 200
|
185 |
+
except Exception as e:
|
186 |
+
print("Error in /upload_audio:", str(e))
|
187 |
+
return jsonify({"error": "サーバーエラー", "details": str(e)}), 500
|
188 |
+
|
189 |
+
@app.route('/reset', methods=['GET'])
|
190 |
+
def reset():
|
191 |
+
global users
|
192 |
+
users = []
|
193 |
+
return jsonify({"status": "success", "message": "Users reset"}), 200
|
194 |
+
|
195 |
+
|
196 |
+
@app.route('/upload_base_audio', methods=['POST'])
|
197 |
+
def upload_base_audio():
|
198 |
+
global users#グローバル変数を編集できるようにする
|
199 |
+
try:
|
200 |
+
data = request.get_json()
|
201 |
+
if not data or 'audio_data' not in data or 'name' not in data:
|
202 |
+
return jsonify({"error": "音声データまたは名前がありません"}), 400
|
203 |
+
name = data['name'] # 名前を取得
|
204 |
+
print(name)
|
205 |
+
|
206 |
+
|
207 |
+
users.append(name)
|
208 |
+
users=list(set(users))#重複排除
|
209 |
+
print(users)
|
210 |
+
|
211 |
+
|
212 |
+
audio_path=process.save_audio_from_base64(
|
213 |
+
base64_audio=data['audio_data'], # 音声データ
|
214 |
+
output_dir= "/tmp/data/base_audio", #保存先
|
215 |
+
output_filename=f"{name}.wav" # 固定ファイル名(必要に応じて generate_filename() で一意のファイル名に変更可能)
|
216 |
+
)
|
217 |
+
return jsonify({"state": "Registration Success!", "path": audio_path}), 200
|
218 |
+
except Exception as e:
|
219 |
+
print("Error in /upload_base_audio:", str(e))
|
220 |
+
return jsonify({"error": "サーバーエラー", "details": str(e)}), 500
|
221 |
+
|
222 |
+
if __name__ == '__main__':
|
223 |
+
port = int(os.environ.get("PORT", 7860))
|
224 |
app.run(debug=True, host="0.0.0.0", port=port)
|