Spaces:
Build error
Build error
Update app.py
Browse files
app.py
CHANGED
|
@@ -2,46 +2,48 @@ import gradio as gr
|
|
| 2 |
from datetime import datetime, timedelta
|
| 3 |
import requests
|
| 4 |
import json
|
|
|
|
| 5 |
|
| 6 |
def wayback(website):
|
| 7 |
if not website:
|
| 8 |
-
return
|
| 9 |
|
| 10 |
-
# Rango de tiempo fijo de
|
| 11 |
end_date = datetime.now()
|
| 12 |
-
start_date = end_date - timedelta(days=
|
| 13 |
datefrom = start_date.strftime('%Y%m%d')
|
| 14 |
dateto = end_date.strftime('%Y%m%d')
|
| 15 |
|
| 16 |
-
# Construir la consulta a la Wayback Machine con un límite fijo de
|
| 17 |
-
query = f"?url={website}&output=json&from={datefrom}&to={dateto}&limit=
|
| 18 |
|
| 19 |
try:
|
| 20 |
response = requests.get(f"http://web.archive.org/cdx/search/cdx{query}")
|
| 21 |
if response.status_code != 200:
|
| 22 |
-
return
|
| 23 |
content = json.loads(response.text)
|
| 24 |
if len(content) <= 1:
|
| 25 |
-
return '
|
| 26 |
except json.JSONDecodeError as e:
|
| 27 |
-
return
|
| 28 |
except Exception as e:
|
| 29 |
-
return
|
| 30 |
|
| 31 |
-
# Procesar los resultados y devolverlos
|
| 32 |
results = []
|
| 33 |
for row in content[1:]:
|
| 34 |
date, page, status = [row[i] for i in [1, 2, 4]]
|
| 35 |
formatted_date = datetime.strptime(date, '%Y%m%d%H%M%S').strftime('%d/%m/%Y')
|
| 36 |
formatted_wayback_url = f"https://web.archive.org/web/{date}/{page}"
|
| 37 |
-
results.append(
|
| 38 |
-
|
|
|
|
| 39 |
|
| 40 |
# Crear la interfaz de Gradio
|
| 41 |
iface = gr.Interface(
|
| 42 |
fn=wayback,
|
| 43 |
inputs="text",
|
| 44 |
-
outputs="
|
| 45 |
title="Wayback Machine Lookup",
|
| 46 |
description="Busca instantáneas de una página web en la Wayback Machine. Introduce solo la URL."
|
| 47 |
)
|
|
|
|
| 2 |
from datetime import datetime, timedelta
|
| 3 |
import requests
|
| 4 |
import json
|
| 5 |
+
import pandas as pd
|
| 6 |
|
| 7 |
def wayback(website):
|
| 8 |
if not website:
|
| 9 |
+
return pd.DataFrame(columns=["Fecha", "URL"])
|
| 10 |
|
| 11 |
+
# Rango de tiempo fijo de 365 días hacia atrás
|
| 12 |
end_date = datetime.now()
|
| 13 |
+
start_date = end_date - timedelta(days=365)
|
| 14 |
datefrom = start_date.strftime('%Y%m%d')
|
| 15 |
dateto = end_date.strftime('%Y%m%d')
|
| 16 |
|
| 17 |
+
# Construir la consulta a la Wayback Machine con un límite fijo de 3000 resultados
|
| 18 |
+
query = f"?url={website}&output=json&from={datefrom}&to={dateto}&limit=3000"
|
| 19 |
|
| 20 |
try:
|
| 21 |
response = requests.get(f"http://web.archive.org/cdx/search/cdx{query}")
|
| 22 |
if response.status_code != 200:
|
| 23 |
+
return pd.DataFrame([{'Fecha': 'Error', 'URL': f'Respuesta no exitosa con código de estado {response.status_code}'}])
|
| 24 |
content = json.loads(response.text)
|
| 25 |
if len(content) <= 1:
|
| 26 |
+
return pd.DataFrame([{'Fecha': 'Error', 'URL': 'No hay datos para esta página web'}])
|
| 27 |
except json.JSONDecodeError as e:
|
| 28 |
+
return pd.DataFrame([{'Fecha': 'Error', 'URL': f'Error al analizar JSON: {e}'}])
|
| 29 |
except Exception as e:
|
| 30 |
+
return pd.DataFrame([{'Fecha': 'Error', 'URL': f'Error: {e}'}])
|
| 31 |
|
| 32 |
+
# Procesar los resultados y devolverlos en forma de DataFrame
|
| 33 |
results = []
|
| 34 |
for row in content[1:]:
|
| 35 |
date, page, status = [row[i] for i in [1, 2, 4]]
|
| 36 |
formatted_date = datetime.strptime(date, '%Y%m%d%H%M%S').strftime('%d/%m/%Y')
|
| 37 |
formatted_wayback_url = f"https://web.archive.org/web/{date}/{page}"
|
| 38 |
+
results.append({'Fecha': formatted_date, 'URL': formatted_wayback_url})
|
| 39 |
+
|
| 40 |
+
return pd.DataFrame(results)
|
| 41 |
|
| 42 |
# Crear la interfaz de Gradio
|
| 43 |
iface = gr.Interface(
|
| 44 |
fn=wayback,
|
| 45 |
inputs="text",
|
| 46 |
+
outputs="dataframe",
|
| 47 |
title="Wayback Machine Lookup",
|
| 48 |
description="Busca instantáneas de una página web en la Wayback Machine. Introduce solo la URL."
|
| 49 |
)
|