Spaces:
Runtime error
Runtime error
Commit
•
f583bdc
1
Parent(s):
4f4d3a0
sejm
Browse files- app.py +106 -4
- requirements.txt +3 -0
app.py
CHANGED
@@ -1,19 +1,121 @@
|
|
1 |
import gradio as gr
|
2 |
from transformers import pipeline
|
3 |
import numpy as np
|
|
|
|
|
|
|
|
|
|
|
|
|
4 |
|
5 |
transcriber = pipeline("automatic-speech-recognition", model="filipzawadka/whisper-small-pl-2")
|
6 |
#transcriber = pipeline("automatic-speech-recognition", model="openai/whisper-small.en")
|
7 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
8 |
|
9 |
-
|
10 |
-
|
11 |
-
|
|
|
12 |
|
13 |
|
14 |
demo = gr.Interface(
|
15 |
fn=transcribe,
|
16 |
-
inputs=gr.Audio(type="filepath"),
|
|
|
17 |
outputs="text",
|
18 |
)
|
19 |
|
|
|
1 |
import gradio as gr
|
2 |
from transformers import pipeline
|
3 |
import numpy as np
|
4 |
+
import requests
|
5 |
+
import subprocess
|
6 |
+
import os
|
7 |
+
import urllib.parse
|
8 |
+
|
9 |
+
term = 9
|
10 |
|
11 |
transcriber = pipeline("automatic-speech-recognition", model="filipzawadka/whisper-small-pl-2")
|
12 |
#transcriber = pipeline("automatic-speech-recognition", model="openai/whisper-small.en")
|
13 |
|
14 |
+
def offset_time(link, start_offset, clip_length):
|
15 |
+
|
16 |
+
# Parse the URL
|
17 |
+
parsed_url = urllib.parse.urlparse(link)
|
18 |
+
query_params = urllib.parse.parse_qs(parsed_url.query)
|
19 |
+
|
20 |
+
# Change the start and stop time by a set number
|
21 |
+
# For example, adding 1000 to both start and stop times
|
22 |
+
start_time = int(query_params['startTime'][0]) + start_offset
|
23 |
+
stop_time = start_time + clip_length
|
24 |
+
|
25 |
+
# Rebuild the query with the new times
|
26 |
+
new_query_params = {'startTime': [str(start_time)], 'stopTime': [str(stop_time)]}
|
27 |
+
new_query = urllib.parse.urlencode(new_query_params, doseq=True)
|
28 |
+
|
29 |
+
# Rebuild the entire URL
|
30 |
+
return urllib.parse.urlunparse((
|
31 |
+
parsed_url.scheme,
|
32 |
+
parsed_url.netloc,
|
33 |
+
parsed_url.path,
|
34 |
+
parsed_url.params,
|
35 |
+
new_query,
|
36 |
+
parsed_url.fragment
|
37 |
+
))
|
38 |
+
|
39 |
+
def get_sejm_videos(term):
|
40 |
+
# Replace 'term9' with the desired term
|
41 |
+
url = f"https://api.sejm.gov.pl/sejm/term{term}/videos"
|
42 |
+
|
43 |
+
# Send a GET request to the API
|
44 |
+
response = requests.get(url)
|
45 |
+
|
46 |
+
# Check if the request was successful
|
47 |
+
if response.status_code == 200:
|
48 |
+
return response.json()
|
49 |
+
else:
|
50 |
+
return f"Error: {response.status_code}"
|
51 |
+
|
52 |
+
def get_today_sejm_videos(term):
|
53 |
+
# Replace 'term9' with the desired term
|
54 |
+
url = f"https://api.sejm.gov.pl/sejm/term{term}/videos/today"
|
55 |
+
|
56 |
+
# Send a GET request to the API
|
57 |
+
response = requests.get(url)
|
58 |
+
|
59 |
+
# Check if the request was successful
|
60 |
+
if response.status_code == 200:
|
61 |
+
return response.json()
|
62 |
+
else:
|
63 |
+
return f"Error: {response.status_code}"
|
64 |
+
def get_sejm_videos(term, since=None, till=None, title=None, video_type=None, comm=None):
|
65 |
+
base_url = f"https://api.sejm.gov.pl/sejm/term{term}/videos"
|
66 |
+
params = {}
|
67 |
+
|
68 |
+
if since:
|
69 |
+
params['since'] = since
|
70 |
+
if till:
|
71 |
+
params['till'] = till
|
72 |
+
if title:
|
73 |
+
params['title'] = title
|
74 |
+
if video_type:
|
75 |
+
params['type'] = video_type
|
76 |
+
if comm:
|
77 |
+
params['comm'] = comm
|
78 |
+
|
79 |
+
response = requests.get(base_url, params=params)
|
80 |
+
if response.status_code == 200:
|
81 |
+
return response.json()
|
82 |
+
else:
|
83 |
+
return f"Error: {response.status_code}"
|
84 |
+
|
85 |
+
def download_video(video_url, video_path):
|
86 |
+
response = requests.get(video_url)
|
87 |
+
if response.status_code == 200:
|
88 |
+
with open(video_path, 'wb') as file:
|
89 |
+
file.write(response.content)
|
90 |
+
return True
|
91 |
+
else:
|
92 |
+
print(f"Error downloading video: {response.status_code}")
|
93 |
+
return False
|
94 |
+
|
95 |
+
def extract_audio(video_path, audio_path):
|
96 |
+
command = ['ffmpeg', '-i', video_path, '-q:a', '0', '-map', 'a', audio_path, '-y']
|
97 |
+
subprocess.run(command)
|
98 |
+
if os.path.exists(audio_path):
|
99 |
+
print("Audio extracted successfully.")
|
100 |
+
else:
|
101 |
+
print("Error extracting audio.")
|
102 |
+
|
103 |
+
# 600000,10000
|
104 |
+
|
105 |
+
def transcribe(num1,num2):
|
106 |
+
|
107 |
+
videos = get_sejm_videos(term)
|
108 |
|
109 |
+
if download_video(offset_time(videos[0]['videoLink'],num1,num2), "./video.mp4"):
|
110 |
+
extract_audio("./video.mp4", "./audio.mp3")
|
111 |
+
print("./audio.mp3")
|
112 |
+
return transcriber("./audio.mp3")["text"]
|
113 |
|
114 |
|
115 |
demo = gr.Interface(
|
116 |
fn=transcribe,
|
117 |
+
#inputs=gr.Audio(type="filepath"),
|
118 |
+
inputs=[gr.Number(label="Number 1"), gr.Number(label="Number 2")],
|
119 |
outputs="text",
|
120 |
)
|
121 |
|
requirements.txt
CHANGED
@@ -91,3 +91,6 @@ widgetsnbextension==4.0.9
|
|
91 |
xxhash==3.4.1
|
92 |
yarl==1.9.4
|
93 |
librosa
|
|
|
|
|
|
|
|
91 |
xxhash==3.4.1
|
92 |
yarl==1.9.4
|
93 |
librosa
|
94 |
+
requests
|
95 |
+
os
|
96 |
+
subprocess
|