Spaces:

freddyaboulton
/

theme-gallery

Sleeping

App Files Files Community

freddyaboulton HF staff commited on Mar 14, 2023

Commit

69f3076

1 Parent(s): 0ba68c7

Modify backend

Browse files

Files changed (2) hide show

app.py +5 -41
processing.py +21 -37

app.py CHANGED Viewed

@@ -1,54 +1,18 @@
 from processing import process_spaces
 from apscheduler.schedulers.background import BackgroundScheduler
-from huggingface_hub import hf_hub_url
 from fastapi import FastAPI
 from fastapi.responses import HTMLResponse
 import uvicorn
-import requests
-def batch(iterable, n=1):
-    l = len(iterable)
-    for ndx in range(0, l, n):
-        yield iterable[ndx:min(ndx + n, l)]
-SPACE = """
-<a href="https://huggingface.co/spaces/{space_id}/" target="_blank" class="linkwrap">
-    <div class="blocker"></div>
-    <iframe width="300" height="300" src="{subdomain}" frameborder="0"></iframe>
-</a>
-"""
-def embed_theme_spaces():
-    subdomains = requests.get(hf_hub_url("freddyaboulton/gradio-theme-subdomains",
-                                         filename="subdomains.json",
-                                         repo_type="dataset")).json()
-    html = """
-    <style>
-        .linkwrap { position:relative; display:inline-block; }
-        .blocker { position:absolute; height:100%; width:100%; z-index:1; background:rgba(116, 113, 113, 0.102);  }
-        .linkwrap iframe { z-index: 2; }
-    </style>
-    <h1 style="font-family: sans-serif; text-align: center;">Gradio Theme Gallery</h1>
-    <h3 style="font-family: sans-serif; text-align: center;">Click on a theme to be taken to its preview!</h3>
-    <body style="background: rgb(255 165 0 / 38%)">
-    """
-    for row in batch(list(subdomains.keys()), n=4):
-        html += ("\n".join(["<div>"] +
-                           [SPACE.format(subdomain=subdomains[s], space_id=s) for s in row] +
-                           ["</div>"])
-        )
-    return html + "</body>"
 app = FastAPI()
-@app.get("/", response_class=HTMLResponse)
 def index():
-    return embed_theme_spaces()
 scheduler = BackgroundScheduler()
 scheduler.add_job(func=process_spaces, trigger="interval", seconds=360)

 from processing import process_spaces
 from apscheduler.schedulers.background import BackgroundScheduler
 from fastapi import FastAPI
 from fastapi.responses import HTMLResponse
 import uvicorn
 app = FastAPI()
+@app.get("/")
 def index():
+    return HTMLResponse("""
+    <p>Backend for gradio theme gallery.
+    <a href="https://huggingface.co/spaces/freddyaboulton/theme-gallery-static">https://huggingface.co/spaces/freddyaboulton/theme-gallery-stati</a>
+    </p>""")
 scheduler = BackgroundScheduler()
 scheduler.add_job(func=process_spaces, trigger="interval", seconds=360)

processing.py CHANGED Viewed

@@ -1,21 +1,20 @@
 from __future__ import annotations
-from typing import List, Dict, TypedDict
 import huggingface_hub
 from huggingface_hub.hf_api import SpaceInfo
-import requests
 from concurrent.futures import ThreadPoolExecutor
 import os
-import time
-import random
 import json
 import datetime
-class SpaceSubdomain(TypedDict):
-    name: str
     subdomain: str
 repo = huggingface_hub.Repository(
     local_dir="data",
@@ -29,49 +28,34 @@ repo.git_pull()
 api = huggingface_hub.HfApi(token=os.getenv("HF_TOKEN"))
 def get_theme_preview_spaces() -> List[SpaceInfo]:
-    return api.list_spaces(filter="gradio-theme")
-def get_subdomain(space_name: SpaceInfo) -> SpaceSubdomain | None:
-    time.sleep(random.random())
-    print(f"processing {space_name.id}")
     if not space_name.id:
         print(f"no space_name for {space_name}")
         return None
-    subdomain: str | None = (
-        requests.get(
-            f"https://huggingface.co/api/spaces/{space_name.id}/host",
-            headers={"Authorization": f"Bearer {os.getenv('HF_TOKEN')}"}
-            )
-            .json()
-            .get("host")
-    )
-    if not subdomain:
-        print(f"no subdomain for {space_name.id}")
         return None
-    return {'name': space_name.id, 'subdomain': subdomain}
-def get_all_subdomains(spaces: List[SpaceInfo], subdomains: Dict[str, str]) -> Dict[str, str]:
-    not_parsed_yet = [space for space in spaces if space.id not in subdomains]
     with ThreadPoolExecutor(max_workers=10) as executor:
-        new_subdomains = executor.map(get_subdomain, not_parsed_yet)
-    new_subdomains = {s["name"]: s["subdomain"] for s in new_subdomains if s}
-    new_subdomains.update(subdomains)
-    return new_subdomains
 def process_spaces():
-    theme_spaces = get_theme_preview_spaces()
-    subdomains: Dict[str, str] = json.load(open("data/subdomains.json"))
-    all_subdomains = get_all_subdomains(theme_spaces, subdomains)
-    json.dump(all_subdomains, open("data/subdomains.json", "w"))
     repo.push_to_hub(blocking=False, commit_message=f"Updating data at {datetime.datetime.now()}")

 from __future__ import annotations
+from typing import List, TypedDict
 import huggingface_hub
 from huggingface_hub.hf_api import SpaceInfo
 from concurrent.futures import ThreadPoolExecutor
 import os
 import json
 import datetime
+import tqdm
+class SpaceData(TypedDict):
+    id: str
+    likes: int
     subdomain: str
+    lastModified: str
 repo = huggingface_hub.Repository(
     local_dir="data",
 api = huggingface_hub.HfApi(token=os.getenv("HF_TOKEN"))
 def get_theme_preview_spaces() -> List[SpaceInfo]:
+    return list(iter(api.list_spaces(filter="gradio-theme")))
+def get_info(space_name: SpaceInfo) -> SpaceData | None:
     if not space_name.id:
         print(f"no space_name for {space_name}")
         return None
+    space_info = api.space_info(space_name.id, token=os.getenv("HF_TOKEN"))
+    subdomain: str | None = getattr(space_info, "subdomain", None)
+    if subdomain is None:
+        print(f"no subdomain for {space_info.id}")
         return None
+    return {"id": space_info.id, "likes": space_info.likes,
+            "subdomain": f"https://{space_info.subdomain}.hf.space",
+            "lastModified": space_info.lastModified} # type: ignore
+def get_all_info(spaces: List[SpaceInfo]) -> List[SpaceData]:
     with ThreadPoolExecutor(max_workers=10) as executor:
+        all_info = list(tqdm.tqdm(executor.map(get_info, spaces), total=len(spaces)))
+    return [info for info in all_info if info]
 def process_spaces():
+    theme_spaces = list(iter(get_theme_preview_spaces()))
+    all_info = get_all_info(theme_spaces)
+    json.dump(all_info, open("data/subdomains.json", "w"))
     repo.push_to_hub(blocking=False, commit_message=f"Updating data at {datetime.datetime.now()}")