davanstrien HF staff commited on
Commit
26205f8
1 Parent(s): 80d8645

Add commit scheduler and usage stats tracking to

Browse files
Files changed (1) hide show
  1. app.py +59 -7
app.py CHANGED
@@ -1,6 +1,47 @@
1
- from huggingface_hub import HfApi
2
- import gradio as gr
3
  import re
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4
 
5
 
6
  def extract_slug(url):
@@ -33,9 +74,16 @@ def clone_collection(
33
  token=token,
34
  )
35
  for item in collection.items:
36
- api.add_collection_item(
37
- new_collection.slug, item.item_id, item_type=item.item_type
38
- )
 
 
 
 
 
 
 
39
  return f"[Collection]({collection.url}) has been cloned into [{new_collection.slug}]({new_collection.url})"
40
 
41
 
@@ -48,8 +96,12 @@ with gr.Blocks(css="style.css") as demo:
48
  gr.HTML(title)
49
  gr.HTML(
50
  """<p style='text-align: center;'>
51
- This space allows you to clone a <a href="https://huggingface.co/docs/hub/collections">Collection</a> from the Hugging Face Hub into your own namespace.
52
- You can then edit the collection to your liking!</p>"""
 
 
 
 
53
  )
54
  gr.Markdown("## Authentication")
55
  gr.Markdown(
 
1
+ import json
2
+ import os
3
  import re
4
+ from datetime import datetime
5
+ from pathlib import Path
6
+
7
+ import gradio as gr
8
+ from dotenv import load_dotenv
9
+ from huggingface_hub import CommitScheduler, HfApi
10
+ from huggingface_hub.utils import HfHubHTTPError
11
+
12
+ load_dotenv()
13
+
14
+ HF_TOKEN = os.getenv("HF_TOKEN")
15
+
16
+ JSON_DATASET_DIR = Path("dataset")
17
+ JSON_DATASET_DIR.mkdir(parents=True, exist_ok=True)
18
+ JSON_DATASET_PATH = JSON_DATASET_DIR / "dataset.jsonl"
19
+
20
+ scheduler = CommitScheduler(
21
+ repo_id="librarian-bots/collection_cloner-usage-stats",
22
+ repo_type="dataset",
23
+ folder_path=JSON_DATASET_DIR,
24
+ path_in_repo=str(JSON_DATASET_PATH),
25
+ token=HF_TOKEN,
26
+ )
27
+
28
+
29
+ def save_json(source_slug: str, destination_slug: str) -> None:
30
+ with scheduler.lock:
31
+ with JSON_DATASET_PATH.open("a") as f:
32
+ if source_slug.startswith("hf_"): # catch people accidentally adding tokens
33
+ return None
34
+ if destination_slug.startswith("hf_"):
35
+ return None
36
+ json.dump(
37
+ {
38
+ "source_collection": source_slug,
39
+ "destination_collection": destination_slug,
40
+ "datetime": datetime.now().isoformat(),
41
+ },
42
+ f,
43
+ )
44
+ f.write("\n")
45
 
46
 
47
  def extract_slug(url):
 
74
  token=token,
75
  )
76
  for item in collection.items:
77
+ try:
78
+ api.add_collection_item(
79
+ new_collection.slug, item.item_id, item_type=item.item_type
80
+ )
81
+ except HfHubHTTPError as e:
82
+ gr.Info(
83
+ f"Failed to add item {item.item_id} to collection {new_collection.slug} because it already exists in this collection."
84
+ )
85
+ if not private:
86
+ save_json(collection.slug, new_collection.slug)
87
  return f"[Collection]({collection.url}) has been cloned into [{new_collection.slug}]({new_collection.url})"
88
 
89
 
 
96
  gr.HTML(title)
97
  gr.HTML(
98
  """<p style='text-align: center;'>
99
+ This space allows you to clone a <a href="https://huggingface.co/docs/hub/collections">Collection</a> from the Hugging Face Hub into your own namespace.<p>
100
+ <p style='text-align: center;'> You can edit this cloned Collection to your liking!</p>"""
101
+ )
102
+ gr.Markdown(
103
+ """
104
+ **Note**: To track interest in this feature this Space keeps a record of clones which are cloned into public collection. Clones into private Collections are not tracked."""
105
  )
106
  gr.Markdown("## Authentication")
107
  gr.Markdown(