Spaces:
Sleeping
Sleeping
Commit
·
f4c39f1
1
Parent(s):
e6306f4
Take screenshots
Browse files- app.py +12 -5
- processing.py +53 -12
- requirements.txt +4 -1
- screenshot.py +15 -0
app.py
CHANGED
@@ -8,18 +8,25 @@ import datetime
|
|
8 |
|
9 |
app = FastAPI()
|
10 |
|
|
|
11 |
@app.get("/")
|
12 |
def index():
|
13 |
-
return HTMLResponse(
|
|
|
14 |
<p>Backend for gradio theme gallery.
|
15 |
<a href="https://huggingface.co/spaces/freddyaboulton/theme-gallery-static">https://huggingface.co/spaces/freddyaboulton/theme-gallery-stati</a>
|
16 |
-
</p>"""
|
|
|
|
|
17 |
|
18 |
scheduler = BackgroundScheduler()
|
19 |
-
scheduler.add_job(
|
20 |
-
|
|
|
|
|
|
|
|
|
21 |
scheduler.start()
|
22 |
|
23 |
if __name__ == "__main__":
|
24 |
uvicorn.run(app, port=7860, host="0.0.0.0")
|
25 |
-
|
|
|
8 |
|
9 |
app = FastAPI()
|
10 |
|
11 |
+
|
12 |
@app.get("/")
|
13 |
def index():
|
14 |
+
return HTMLResponse(
|
15 |
+
"""
|
16 |
<p>Backend for gradio theme gallery.
|
17 |
<a href="https://huggingface.co/spaces/freddyaboulton/theme-gallery-static">https://huggingface.co/spaces/freddyaboulton/theme-gallery-stati</a>
|
18 |
+
</p>"""
|
19 |
+
)
|
20 |
+
|
21 |
|
22 |
scheduler = BackgroundScheduler()
|
23 |
+
scheduler.add_job(
|
24 |
+
func=process_spaces,
|
25 |
+
trigger="interval",
|
26 |
+
seconds=1200,
|
27 |
+
next_run_time=datetime.datetime.now(),
|
28 |
+
)
|
29 |
scheduler.start()
|
30 |
|
31 |
if __name__ == "__main__":
|
32 |
uvicorn.run(app, port=7860, host="0.0.0.0")
|
|
processing.py
CHANGED
@@ -10,6 +10,10 @@ import json
|
|
10 |
import datetime
|
11 |
import tqdm
|
12 |
import requests
|
|
|
|
|
|
|
|
|
13 |
|
14 |
class SpaceData(TypedDict):
|
15 |
id: str
|
@@ -18,20 +22,32 @@ class SpaceData(TypedDict):
|
|
18 |
lastModified: str
|
19 |
status: str
|
20 |
|
|
|
21 |
repo = huggingface_hub.Repository(
|
22 |
local_dir="data",
|
23 |
repo_type="dataset",
|
24 |
clone_from="freddyaboulton/gradio-theme-subdomains",
|
25 |
-
token=os.getenv("HF_TOKEN")
|
26 |
)
|
27 |
repo.git_pull()
|
28 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
29 |
|
30 |
api = huggingface_hub.HfApi(token=os.getenv("HF_TOKEN"))
|
31 |
|
|
|
32 |
def get_theme_preview_spaces() -> List[SpaceInfo]:
|
33 |
return list(iter(api.list_spaces(filter="gradio-theme")))
|
34 |
|
|
|
35 |
def get_info(space_name: SpaceInfo) -> SpaceData | None:
|
36 |
if not space_name.id:
|
37 |
print(f"no space_name for {space_name}")
|
@@ -44,17 +60,42 @@ def get_info(space_name: SpaceInfo) -> SpaceData | None:
|
|
44 |
if subdomain is None:
|
45 |
print(f"no subdomain for {space_info.id}")
|
46 |
return None
|
47 |
-
|
48 |
-
status = space_info.runtime[
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
49 |
if status not in ["SLEEPING", "RUNNING", "RUNNING_BUILDING", "BUILDING"]:
|
50 |
print(f"Space not running, building, or sleeping {space_info.id}")
|
51 |
elif status == "SLEEPING":
|
52 |
requests.get(f"https://huggingface.co/spaces/{space_info.id}")
|
53 |
-
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
|
|
|
|
|
|
58 |
|
59 |
|
60 |
def get_all_info(spaces: List[SpaceInfo]) -> List[SpaceData]:
|
@@ -68,7 +109,7 @@ def process_spaces():
|
|
68 |
|
69 |
all_info = get_all_info(theme_spaces)
|
70 |
|
71 |
-
json.dump(all_info, open("data/
|
72 |
-
repo.push_to_hub(
|
73 |
-
|
74 |
-
|
|
|
10 |
import datetime
|
11 |
import tqdm
|
12 |
import requests
|
13 |
+
from pathlib import Path
|
14 |
+
from screenshot import get_screen_shot
|
15 |
+
import boto3
|
16 |
+
|
17 |
|
18 |
class SpaceData(TypedDict):
|
19 |
id: str
|
|
|
22 |
lastModified: str
|
23 |
status: str
|
24 |
|
25 |
+
|
26 |
repo = huggingface_hub.Repository(
|
27 |
local_dir="data",
|
28 |
repo_type="dataset",
|
29 |
clone_from="freddyaboulton/gradio-theme-subdomains",
|
30 |
+
token=os.getenv("HF_TOKEN"),
|
31 |
)
|
32 |
repo.git_pull()
|
33 |
|
34 |
+
screen_shot_dir = Path("data") / "images"
|
35 |
+
screen_shot_dir.mkdir(exist_ok=True, parents=True)
|
36 |
+
|
37 |
+
s3_client = boto3.client(
|
38 |
+
"s3",
|
39 |
+
aws_access_key_id=os.getenv("AWS_ACCESS_KEY"),
|
40 |
+
aws_secret_access_key=os.getenv("AWS_SECRET_ACCESS_KEY"),
|
41 |
+
)
|
42 |
+
|
43 |
|
44 |
api = huggingface_hub.HfApi(token=os.getenv("HF_TOKEN"))
|
45 |
|
46 |
+
|
47 |
def get_theme_preview_spaces() -> List[SpaceInfo]:
|
48 |
return list(iter(api.list_spaces(filter="gradio-theme")))
|
49 |
|
50 |
+
|
51 |
def get_info(space_name: SpaceInfo) -> SpaceData | None:
|
52 |
if not space_name.id:
|
53 |
print(f"no space_name for {space_name}")
|
|
|
60 |
if subdomain is None:
|
61 |
print(f"no subdomain for {space_info.id}")
|
62 |
return None
|
63 |
+
|
64 |
+
status = space_info.runtime["stage"]
|
65 |
+
img_id = space_info.id.replace("/", "_")
|
66 |
+
light_file = str(screen_shot_dir / Path(img_id + "_light.jpg"))
|
67 |
+
dark_file = str(screen_shot_dir / Path(img_id + "_dark.jpg"))
|
68 |
+
if status == "RUNNING":
|
69 |
+
get_screen_shot(
|
70 |
+
f"https://{space_info.subdomain}.hf.space?__theme=light", 10, light_file
|
71 |
+
)
|
72 |
+
get_screen_shot(
|
73 |
+
f"https://{space_info.subdomain}.hf.space?__theme=dark", 10, dark_file
|
74 |
+
)
|
75 |
+
s3_client.upload_file(
|
76 |
+
light_file,
|
77 |
+
"gradio-theme-screenshots",
|
78 |
+
img_id + "_light.jpg",
|
79 |
+
ExtraArgs={"ContentType": "image/jpg"},
|
80 |
+
)
|
81 |
+
s3_client.upload_file(
|
82 |
+
dark_file,
|
83 |
+
"gradio-theme-screenshots",
|
84 |
+
img_id + "_dark.jpg",
|
85 |
+
ExtraArgs={"ContentType": "image/jpg"},
|
86 |
+
)
|
87 |
if status not in ["SLEEPING", "RUNNING", "RUNNING_BUILDING", "BUILDING"]:
|
88 |
print(f"Space not running, building, or sleeping {space_info.id}")
|
89 |
elif status == "SLEEPING":
|
90 |
requests.get(f"https://huggingface.co/spaces/{space_info.id}")
|
91 |
+
return {
|
92 |
+
"id": space_info.id,
|
93 |
+
"likes": space_info.likes,
|
94 |
+
"subdomain": f"https://{space_info.subdomain}.hf.space",
|
95 |
+
"lastModified": space_info.lastModified,
|
96 |
+
"screenshot_id": img_id,
|
97 |
+
"status": status,
|
98 |
+
} # type: ignore
|
99 |
|
100 |
|
101 |
def get_all_info(spaces: List[SpaceInfo]) -> List[SpaceData]:
|
|
|
109 |
|
110 |
all_info = get_all_info(theme_spaces)
|
111 |
|
112 |
+
json.dump(all_info, open("data/val_subdomains.json", "w"))
|
113 |
+
repo.push_to_hub(
|
114 |
+
blocking=False, commit_message=f"Updating data at {datetime.datetime.now()}"
|
115 |
+
)
|
requirements.txt
CHANGED
@@ -1 +1,4 @@
|
|
1 |
-
apscheduler
|
|
|
|
|
|
|
|
1 |
+
apscheduler
|
2 |
+
selenium
|
3 |
+
webdriver-manager
|
4 |
+
boto3
|
screenshot.py
ADDED
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from selenium import webdriver
|
2 |
+
from selenium.webdriver import ChromeOptions
|
3 |
+
from webdriver_manager.chrome import ChromeDriverManager
|
4 |
+
import time
|
5 |
+
|
6 |
+
options = ChromeOptions()
|
7 |
+
options.add_argument("--headless=new")
|
8 |
+
ChromeDriverManager().install()
|
9 |
+
|
10 |
+
|
11 |
+
def get_screen_shot(url: str, load_time: int, path: str):
|
12 |
+
driver = webdriver.Chrome(options=options)
|
13 |
+
driver.get(url)
|
14 |
+
time.sleep(load_time)
|
15 |
+
driver.save_screenshot(path)
|