Update app.py
Browse files
app.py
CHANGED
|
@@ -22,43 +22,43 @@ def link_find(url):
|
|
| 22 |
return out
|
| 23 |
#https://huggingface.co/spaces/Omnibus/crawl
|
| 24 |
|
| 25 |
-
def sitemap(url):
|
| 26 |
uri=""
|
| 27 |
if url != "" and url != None:
|
| 28 |
link1=link_find(url)
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
|
| 34 |
-
|
| 35 |
-
|
| 36 |
-
|
| 37 |
-
|
| 38 |
-
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
|
| 46 |
-
|
| 47 |
-
|
| 48 |
-
|
| 49 |
-
|
| 50 |
-
|
| 51 |
-
|
| 52 |
-
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
print (e)
|
| 58 |
return link1
|
| 59 |
with gr.Blocks() as app:
|
| 60 |
-
inp=gr.Textbox()
|
|
|
|
| 61 |
btn=gr.Button()
|
| 62 |
outp=gr.JSON()
|
| 63 |
-
btn.click(sitemap,inp,outp)
|
| 64 |
app.launch()
|
|
|
|
| 22 |
return out
|
| 23 |
#https://huggingface.co/spaces/Omnibus/crawl
|
| 24 |
|
| 25 |
+
def sitemap(url,level):
|
| 26 |
uri=""
|
| 27 |
if url != "" and url != None:
|
| 28 |
link1=link_find(url)
|
| 29 |
+
if level >=2:
|
| 30 |
+
for i,ea in enumerate(link1):
|
| 31 |
+
print(ea)
|
| 32 |
+
try:
|
| 33 |
+
if not ea['URL'].startswith("http"):
|
| 34 |
+
uri1=url.split("//")[0]
|
| 35 |
+
uri2=url.split("//")[1]
|
| 36 |
+
uri3=uri2.split("/")[0]
|
| 37 |
+
uri=f'{uri1}//{uri3}'
|
| 38 |
+
print(uri)
|
| 39 |
+
out_list=link_find(f"{uri}{ea['URL']}")
|
| 40 |
+
link1[i]['TREE']=out_list
|
| 41 |
+
if level>=3:
|
| 42 |
+
for n,na in enumerate(link1[i]['TREE']):
|
| 43 |
+
print(na)
|
| 44 |
+
try:
|
| 45 |
+
if not na['URL'].startswith("http"):
|
| 46 |
+
uri11=url.split("//")[0]
|
| 47 |
+
uri22=url.split("//")[1]
|
| 48 |
+
uri33=uri22.split("/")[0]
|
| 49 |
+
uri0=f'{uri11}//{uri33}'
|
| 50 |
+
print(uri0)
|
| 51 |
+
out_list1=link_find(f"{uri0}{na['URL']}")
|
| 52 |
+
link1[i]['TREE'][n]['TREE']=out_list1
|
| 53 |
+
except Exception as e:
|
| 54 |
+
print (e)
|
| 55 |
+
except Exception as e:
|
| 56 |
+
print (e)
|
|
|
|
| 57 |
return link1
|
| 58 |
with gr.Blocks() as app:
|
| 59 |
+
inp=gr.Textbox(label="URL")
|
| 60 |
+
level=gr.Slider(minimum=1,maximum=3,step=1,value=2)
|
| 61 |
btn=gr.Button()
|
| 62 |
outp=gr.JSON()
|
| 63 |
+
btn.click(sitemap,[inp,level],outp)
|
| 64 |
app.launch()
|