feat: upload certificate pdf to dataset (#4)

- feat: upload certificate pdf to dataset (c3e2b8d9f9647217794ff4debca4ce9f9f18ae48)

app.py

@@ -11,13 +11,14 @@ import os
 import sys
 import pandas as pd
 import json
+import shutil
 
 api = HfApi()
 HF_TOKEN = os.environ.get("HF_TOKEN")
 
 # Public dataset repo containing the pdfs of already certified users
-DATASET_REPO_URL = "https://huggingface.co/datasets/wseo/huggingface-krew-hackathon23"
-CERTIFIED_USERS_FILENAME = "usernames.csv"
+DATASET_REPO_URL = f"https://wseo:{HF_TOKEN}@huggingface.co/datasets/pseudolab/huggingface-krew-hackathon2023"
+CERTIFIED_USERS_FILENAME = "certified.csv"
 
 ORGANIZATION = "pseudolab"
 
@@ -101,10 +102,10 @@ def generate_certificate(certificate_template, first_name, last_name, hf_usernam
     d.text((538, 419), name, fill=(87,87,87), anchor="mm", font=name_font)
 
     # Debug line id
-    #d.line(((863, 0), (863, 1400)), "gray")
+    #d.line(((815, 0), (815, 1400)), "gray")
 
     # Date of certification
-    d.text((863, 336), f"HKH23-{username}", fill=(117,117,117), anchor="mm", font=username_font)
+    d.text((815, 327), f"HKH23-{hf_username}", fill=(117,117,117), font=username_font)
 
     pdf = im.convert('RGB')
     pdf.save('certificate.pdf')
@@ -112,27 +113,65 @@ def generate_certificate(certificate_template, first_name, last_name, hf_usernam
     return im, "./certificate.pdf"
 
 
+def create_initial_csv(path):
+    """Create an initial CSV file with headers if it doesn't exist."""
+    # Define the headers for our CSV file
+    headers = ['hf_username', 'first_name', 'last_name', 'certificate_type', 'datetime', 'pdf_path']
+    # Create a new DataFrame with no data and these headers
+    df = pd.DataFrame(columns=headers)
+    # Save the DataFrame to a CSV file
+    df.to_csv(path, index=False)
+
+
 def add_certified_user(hf_username, first_name, last_name, certificate_type):
-  """
-  Add the certified user to the database
-  """
-    
-  print("ADD CERTIFIED USER")
-  repo = Repository(local_dir="usernames", clone_from=DATASET_REPO_URL, use_auth_token=HF_TOKEN)
-  repo.git_pull()
+    """
+    Add the certified user to the dataset and include their certificate PDF.
+    """
+    print("ADD CERTIFIED USER")
+    repo = Repository(local_dir="data", clone_from=DATASET_REPO_URL)
+    repo.git_pull()
 
-  history = pd.read_csv(os.path.join("usernames", CERTIFIED_USERS_FILENAME))
+    csv_full_path = os.path.join("data", CERTIFIED_USERS_FILENAME)
 
-  # Check if this hf_username is already in our dataset:
-  check = history.loc[history['hf_username'] == hf_username]
-  if not check.empty:
-    history = history.drop(labels=check.index[0], axis=0)
-  
-  new_row = pd.DataFrame({'hf_username': hf_username, 'first_name': first_name, 'last_name': last_name, 'certificate_type': certificate_type, 'datetime': time.time()}, index=[0])
-  history = pd.concat([new_row, history[:]]).reset_index(drop=True)
+    if not os.path.isfile(csv_full_path):
+        create_initial_csv(csv_full_path)
+
+    history = pd.read_csv(csv_full_path)
+
+    # Check if this hf_username is already in our dataset:
+    check = history.loc[history['hf_username'] == hf_username]
+    if not check.empty:
+        history = history.drop(labels=check.index[0], axis=0)
+
+    pdfs_repo_path = os.path.join("data", "pdfs")
+
+    # Copy the PDF from its current location to the target directory in the repository
+    pdf_repo_filename = f"{hf_username}.pdf"  # Create a specific name for the PDF file
+    pdf_repo_path_full = os.path.join(pdfs_repo_path, pdf_repo_filename)
+    
+    # Create the pdfs directory if it doesn't exist
+    os.makedirs(pdfs_repo_path, exist_ok=True)
     
-  history.to_csv(os.path.join("usernames", CERTIFIED_USERS_FILENAME), index=False)
-  repo.push_to_hub(commit_message="Update certified users list")
+    shutil.copy('./certificate.pdf', pdf_repo_path_full)  # Copy the file
+
+    # Now, add a new entry to your CSV for this user and their PDF
+    new_row = pd.DataFrame({
+        'hf_username': hf_username,
+        'first_name': first_name,
+        'last_name': last_name,
+        'certificate_type': certificate_type,
+        'datetime': time.time(),  # This captures the current time
+        'pdf_path': pdf_repo_path_full[5:]  # This is the relative path to the PDF within the repo
+    }, index=[0])
+    
+    history = pd.concat([new_row, history[:]]).reset_index(drop=True)
+    
+    # Save the updated CSV
+    history.to_csv(os.path.join("data", CERTIFIED_USERS_FILENAME), index=False)
+
+    # Add the PDF and CSV changes to the repo and push
+    repo.git_add()
+    repo.push_to_hub(commit_message="Update certified users list and add PDF")
 
 
 def create_certificate(passed, certificate_type, hf_username, first_name, last_name):  
@@ -149,7 +188,7 @@ def create_certificate(passed, certificate_type, hf_username, first_name, last_n
         # Generate a certificate of 
         certificate, pdf = generate_certificate("./certificate-excellence.png", first_name, last_name, hf_username)
         # Add this user to our database
-        # add_certified_user(hf_username, first_name, last_name, certificate_type)    
+        add_certified_user(hf_username, first_name, last_name, certificate_type)    
         # Add a message
         message = """
         Congratulations, you successfully completed the 2023 Hackathon 🎉! \n 
@@ -161,7 +200,7 @@ def create_certificate(passed, certificate_type, hf_username, first_name, last_n
         # Generate a certificate of completion
         certificate, pdf = generate_certificate("./certificate-completion.png", first_name, last_name, hf_username)
         # Add this user to our database
-        # add_certified_user(hf_username, first_name, last_name, certificate_type)    
+        add_certified_user(hf_username, first_name, last_name, certificate_type)    
         # Add a message
         message = """
         Congratulations, you successfully completed the 2023 Hackathon 🎉! \n 
@@ -177,7 +216,7 @@ def create_certificate(passed, certificate_type, hf_username, first_name, last_n
         # Add a message
         message = """
           You didn't pass the minimum of one contribution to get a certificate of completion. 
-          For more information about the certification process, refer to the submit page.
+          For more information about the certification process, refer to the hackathon page.
           If the results here differ from your contributions, make sure you moved your space to the pseudolab organization.
           """
     return certificate, message, pdf
@@ -200,11 +239,11 @@ with gr.Blocks() as demo:
     # Get your 2023 Hackathon Certificate 🎓
     The certification process is completely free:
     - To get a *certificate of completion*: you need to **contribute to at least one model, dataset, or space**.
-    - To get a *certificate of excellence*: you need to **contribute to models, datasets, and spaces**.
+    - To get a *certificate of excellence*: you need to **contribute to models, datasets, and spaces**. *(Yes, all three!)*
     
     For more information about the certification process [check the hackathon page on certification](https://pseudo-lab.github.io/huggingface-hackathon23/submit.html#certification).
     
-    Don't hesitate to share your certificate on Twitter (tag me [@wonhseo](https://twitter.com/wonhseo) and [@huggingface](https://twitter.com/huggingface)) and on LinkedIn.
+    Don't hesitate to share your certificate on Twitter (tag me [@wonhseo](https://twitter.com/wonhseo), [@pseudolab](https://twitter.com/pseudolab), and [@huggingface](https://twitter.com/huggingface)) and on LinkedIn.
     """)
     
     hf_username = gr.Textbox(placeholder="wseo", label="Your Hugging Face Username (case sensitive)")