extract constants, make urls shorts and clickable

constants.py

@@ -0,0 +1,400 @@
+"""
+Shared constants for the NLP resources application.
+Contains lists of countries, languages, domains, tasks, and utility functions.
+"""
+
+import re
+from urllib.parse import urlparse
+
+# Countries where Spanish is spoken
+COUNTRIES = [
+    "Spain",
+    "Mexico",
+    "Argentina",
+    "Colombia",
+    "Peru",
+    "Venezuela",
+    "Chile",
+    "Ecuador",
+    "Guatemala",
+    "Cuba",
+    "Bolivia",
+    "Dominican Republic",
+    "Honduras",
+    "Paraguay",
+    "El Salvador",
+    "Nicaragua",
+    "Costa Rica",
+    "Panama",
+    "Uruguay",
+    "Puerto Rico",
+]
+
+# Languages relevant to Spanish NLP
+LANGUAGES = [
+    "spanish",
+    "catalan",
+    "basque",
+    "galician",
+    "guarani",
+    "quechua",
+    "aymara",
+    "nauhatl",
+    "mapudungun",
+]
+
+# NLP tasks
+TASKS = [
+    "text classification",
+    "sentiment analysis",
+    "named entity recognition",
+    "part-of-speech tagging",
+    "question answering",
+    "text summarization",
+    "machine translation",
+    "language modeling",
+    "text generation",
+    "information extraction",
+    "semantic similarity",
+    "natural language inference",
+]
+
+# Domains for datasets
+DOMAINS = [
+    "clinical",
+    "legal",
+    "financial",
+    "scientific",
+    "news",
+    "social media",
+    "literature",
+    "general",
+]
+
+# Dataset types
+DATASET_TYPES = ["pretraining", "benchmark", "supervised fine-tuning", "alignment"]
+
+# Event types
+EVENT_TYPES = ["workshop", "talk", "AMA", "round table"]
+
+# Initiative types
+INITIATIVE_TYPES = [
+    "project",
+    "event",
+    "research group",
+    "community",
+    "research institute",
+    "non-profit",
+    "OS company",
+]
+
+# Technical levels (for events)
+TECHNICAL_LEVELS = ["1", "2", "3", "4", "5"]
+
+
+def format_url_for_display(url: str) -> str:
+    """
+    Format URL for display in tables - show only the meaningful part.
+
+    Args:
+        url: Full URL string
+
+    Returns:
+        Shortened, readable version of the URL
+    """
+    if not url or not url.strip():
+        return ""
+
+    url = url.strip()
+
+    # Remove protocol
+    if url.startswith(("http://", "https://")):
+        url = url.split("://", 1)[1]
+
+    # Special handling for common domains
+    if "huggingface.co" in url:
+        # Extract the meaningful part after huggingface.co
+        if "/datasets/" in url:
+            return url.split("/datasets/")[-1]
+        elif "/models/" in url:
+            return url.split("/models/")[-1]
+        elif "/collections/" in url:
+            return url.split("/collections/")[-1]
+        else:
+            # Return everything after huggingface.co/
+            parts = url.split("huggingface.co/")
+            return parts[-1] if len(parts) > 1 else url
+
+    elif "github.com" in url:
+        # Extract repo name (owner/repo)
+        parts = url.split("github.com/")
+        if len(parts) > 1:
+            repo_path = parts[-1].split("/")
+            if len(repo_path) >= 2:
+                return f"{repo_path[0]}/{repo_path[1]}"
+        return url
+
+    elif "zenodo.org" in url:
+        # Extract record ID
+        if "/record/" in url:
+            return f"zenodo:{url.split('/record/')[-1].split('/')[0]}"
+        return url
+
+    elif "arxiv.org" in url:
+        # Extract arXiv ID
+        if "/abs/" in url:
+            return f"arXiv:{url.split('/abs/')[-1]}"
+        elif "/pdf/" in url:
+            return f"arXiv:{url.split('/pdf/')[-1].replace('.pdf', '')}"
+        return url
+
+    elif "youtube.com" in url or "youtu.be" in url:
+        # Extract video ID or title if available
+        if "watch?v=" in url:
+            video_id = url.split("watch?v=")[-1].split("&")[0]
+            return f"YouTube:{video_id[:8]}..."
+        elif "youtu.be/" in url:
+            video_id = url.split("youtu.be/")[-1].split("?")[0]
+            return f"YouTube:{video_id[:8]}..."
+        return url
+
+    else:
+        # For other URLs, try to extract domain and path
+        try:
+            parsed = urlparse(
+                f"https://{url}" if not url.startswith(("http://", "https://")) else url
+            )
+            domain = parsed.netloc
+            path = parsed.path.strip("/")
+
+            if path:
+                # Show domain + first part of path
+                path_parts = path.split("/")
+                if len(path_parts) > 0 and path_parts[0]:
+                    return f"{domain}/{path_parts[0]}"
+
+            return domain
+        except:
+            # Fallback: limit length
+            return url[:30] + "..." if len(url) > 30 else url
+
+
+def make_url_clickable(url: str, display_text: str = None) -> str:
+    """
+    Convert URL to clickable HTML link.
+
+    Args:
+        url: Full URL
+        display_text: Text to display for the link (optional)
+
+    Returns:
+        HTML link string
+    """
+    if not url or not url.strip():
+        return ""
+
+    url = url.strip()
+
+    # Ensure URL has protocol
+    if not url.startswith(("http://", "https://")):
+        url = f"https://{url}"
+
+    # Use provided display text or format the URL
+    text = display_text if display_text else format_url_for_display(url)
+
+    return f'<a target="_blank" href="{url}" style="color: var(--link-text-color); text-decoration: underline; text-decoration-style: dotted;">{text}</a>'
+
+
+def get_column_display_names():
+    """
+    Return mapping of column names to pretty display names.
+
+    Returns:
+        Dictionary mapping column names to display names
+    """
+    return {
+        # Common fields
+        "name": "Name",
+        "submitted_by": "Submitted By",
+        "date_submitted": "Date Submitted",
+        # Dataset fields
+        "github_url": "GitHub",
+        "huggingface_url": "HF Dataset",
+        "zenodo_url": "Zenodo",
+        "paper_url": "Paper",
+        "website_url": "Website",
+        "dataset_type": "Type",
+        "task": "Tasks",
+        "domain": "Domain",
+        "countries": "Countries",
+        "languages": "Languages",
+        # Model fields
+        "familia": "Family",
+        "available_sizes": "Sizes (B)",
+        "hf_collection_url": "HF Collection",
+        # Event fields
+        "titulo": "Title",
+        "ponente": "Speaker",
+        "bio": "Bio",
+        "tipo": "Type",
+        "etiquetas": "Tags",
+        "tema": "Topic",
+        "nivel_tecnico": "Tech Level",
+        "fecha": "Date",
+        "youtube": "YouTube",
+        # Shared task fields
+        "conference_name": "Conference",
+        "workshop_date": "Workshop Date",
+        "registration_deadline": "Registration",
+        "data_available_date": "Data Available",
+        "submission_deadline": "Submission",
+        "more_info_url": "More Info",
+        # Initiative fields
+        "type": "Type",
+    }
+
+
+def format_dataframe_for_display(df, url_columns=None, hide_columns=None):
+    """
+    Format a DataFrame for better display in Gradio tables with clickable URLs.
+
+    Args:
+        df: Pandas DataFrame
+        url_columns: List of column names that contain URLs
+        hide_columns: List of column names to hide
+
+    Returns:
+        Formatted DataFrame
+    """
+    if df.empty:
+        return df
+
+    # Make a copy to avoid modifying original
+    display_df = df.copy()
+
+    # Hide specified columns
+    if hide_columns:
+        display_df = display_df.drop(
+            columns=[col for col in hide_columns if col in display_df.columns]
+        )
+
+    # Format URL columns with clickable links
+    if url_columns:
+        for col in url_columns:
+            if col in display_df.columns:
+                display_df[col] = display_df[col].apply(
+                    lambda x: make_url_clickable(x) if x and str(x).strip() else ""
+                )
+
+    # Ensure first column content doesn't wrap (for name/title columns)
+    first_col = display_df.columns[0] if len(display_df.columns) > 0 else None
+    if first_col:
+        # Keep full text but ensure it displays in a single line (no wrapping)
+        # Replace line breaks and excessive whitespace to ensure single line display
+        display_df[first_col] = display_df[first_col].apply(
+            lambda x: str(x).replace("\n", " ").replace("\r", " ").strip() if x else ""
+        )
+
+    # Rename columns to pretty names
+    column_names = get_column_display_names()
+    display_df = display_df.rename(columns=column_names)
+
+    return display_df
+
+
+def format_dataframe_for_html_display(df, url_columns=None, hide_columns=None):
+    """
+    Format a DataFrame for HTML display with clickable links.
+
+    Args:
+        df: Pandas DataFrame
+        url_columns: List of column names that contain URLs
+        hide_columns: List of column names to hide
+
+    Returns:
+        HTML string representation of the DataFrame
+    """
+    if df.empty:
+        return "<p>No data available</p>"
+
+    # Make a copy to avoid modifying original
+    display_df = df.copy()
+
+    # Hide specified columns
+    if hide_columns:
+        display_df = display_df.drop(
+            columns=[col for col in hide_columns if col in display_df.columns]
+        )
+
+    # Format URL columns with clickable links
+    if url_columns:
+        for col in url_columns:
+            if col in display_df.columns:
+                display_df[col] = display_df[col].apply(
+                    lambda x: make_url_clickable(x) if x and str(x).strip() else ""
+                )
+
+    # Ensure first column content doesn't wrap (for name/title columns)
+    first_col = display_df.columns[0] if len(display_df.columns) > 0 else None
+    if first_col:
+        # Keep full text but ensure it displays in a single line (no wrapping)
+        # Replace line breaks and excessive whitespace to ensure single line display
+        display_df[first_col] = display_df[first_col].apply(
+            lambda x: str(x).replace("\n", " ").replace("\r", " ").strip() if x else ""
+        )
+
+    # Rename columns to pretty names
+    column_names = get_column_display_names()
+    display_df = display_df.rename(columns=column_names)
+
+    # Convert to HTML with custom styling
+    html = display_df.to_html(
+        escape=False,  # Allow HTML in cells
+        index=False,  # Don't show row indices
+        classes="dataframe-table",
+        table_id="resources-table",
+    )
+
+    # Add custom CSS styling
+    styled_html = f"""
+    <style>
+    .dataframe-table {{
+        border-collapse: collapse;
+        margin: 25px 0;
+        font-size: 0.9em;
+        font-family: sans-serif;
+        min-width: 400px;
+        box-shadow: 0 0 20px rgba(0, 0, 0, 0.15);
+        width: 100%;
+    }}
+    .dataframe-table thead tr {{
+        background-color: #009879;
+        color: #ffffff;
+        text-align: left;
+    }}
+    .dataframe-table th,
+    .dataframe-table td {{
+        padding: 12px 15px;
+        border: 1px solid #dddddd;
+    }}
+    .dataframe-table tbody tr {{
+        border-bottom: 1px solid #dddddd;
+    }}
+    .dataframe-table tbody tr:nth-of-type(even) {{
+        background-color: #f3f3f3;
+    }}
+    .dataframe-table tbody tr:hover {{
+        background-color: #f5f5f5;
+    }}
+    .dataframe-table a {{
+        color: #009879;
+        text-decoration: none;
+    }}
+    .dataframe-table a:hover {{
+        text-decoration: underline;
+    }}
+    </style>
+    {html}
+    """
+
+    return styled_html

datasets_resource.py

@@ -4,6 +4,16 @@ import gradio as gr
 import pandas as pd
 from datasets import Dataset, load_dataset
 
+from constants import (
+    COUNTRIES,
+    DATASET_TYPES,
+    DOMAINS,
+    LANGUAGES,
+    TASKS,
+    format_dataframe_for_display,
+    format_dataframe_for_html_display,
+)
+
 # Dataset configuration
 DATASET_NAME = "somosnlp/recursos-pln-es"
 CONFIG_NAME = "datasets"
@@ -169,27 +179,56 @@ def create_all_tab():
             show_label=False,
         )
 
-        # Load initial data
-        initial_df = load_data()
+        # Load and format initial data with clickable links
+        def get_formatted_data():
+            df = load_data()
+            return format_dataframe_for_display(
+                df,
+                url_columns=[
+                    "github_url",
+                    "huggingface_url",
+                    "zenodo_url",
+                    "paper_url",
+                    "website_url",
+                ],
+                hide_columns=["date_submitted"],
+            )
 
+        # Use Dataframe component with HTML rendering enabled
         table = gr.Dataframe(
-            value=initial_df,
+            value=get_formatted_data(),
             label=RESOURCE_TITLE,
             show_label=False,
             interactive=False,
-            wrap=True,
+            wrap=False,  # Disable wrapping to show full text in single lines
+            datatype="markdown",  # Enable HTML rendering
         )
 
         # Connect search functionality
+        def search_and_format(query):
+            initial_df = load_data()
+            filtered_df = search_and_filter_data(initial_df, query)
+            return format_dataframe_for_display(
+                filtered_df,
+                url_columns=[
+                    "github_url",
+                    "huggingface_url",
+                    "zenodo_url",
+                    "paper_url",
+                    "website_url",
+                ],
+                hide_columns=["date_submitted"],
+            )
+
         search_box.change(
-            fn=lambda query: search_and_filter_data(initial_df, query),
+            fn=search_and_format,
             inputs=search_box,
             outputs=table,
         )
 
         # Refresh button to reload data
         refresh_btn = gr.Button("🔄 Refresh Data", variant="secondary")
-        refresh_btn.click(fn=lambda: load_data(), outputs=table)
+        refresh_btn.click(fn=get_formatted_data, outputs=table)
 
         return table
 
@@ -215,12 +254,7 @@ def create_contribute_tab():
 
             dataset_type_input = gr.Dropdown(
                 label="Dataset Type *",
-                choices=[
-                    "pretraining",
-                    "benchmark",
-                    "supervised fine-tuning",
-                    "alignment",
-                ],
+                choices=DATASET_TYPES,
                 info="Type of dataset (required)",
                 multiselect=False,
             )
@@ -262,18 +296,13 @@ def create_contribute_tab():
             # Multi-select fields
             task_input = gr.CheckboxGroup(
                 label="Tasks",
-                choices=[
-                    "language modeling",
-                    "text classification",
-                    "text generation",
-                    "summarization",
-                ],
+                choices=TASKS,
                 info="What tasks is this dataset suitable for?",
             )
 
             domain_input = gr.CheckboxGroup(
                 label="Domain",
-                choices=["legal", "clinical"],
+                choices=DOMAINS,
                 info="Specific domains covered by the dataset",
             )
 
@@ -531,12 +560,7 @@ def create_edit_tab():
             name_input = gr.Textbox(label="Name *", placeholder="Dataset name")
             dataset_type_input = gr.Dropdown(
                 label="Dataset Type *",
-                choices=[
-                    "pretraining",
-                    "benchmark",
-                    "supervised fine-tuning",
-                    "alignment",
-                ],
+                choices=DATASET_TYPES,
                 value="benchmark",
             )
 
@@ -563,73 +587,22 @@ def create_edit_tab():
 
             task_input = gr.CheckboxGroup(
                 label="Tasks",
-                choices=[
-                    "text classification",
-                    "sentiment analysis",
-                    "named entity recognition",
-                    "part-of-speech tagging",
-                    "question answering",
-                    "text summarization",
-                    "machine translation",
-                    "language modeling",
-                    "text generation",
-                    "information extraction",
-                    "semantic similarity",
-                    "natural language inference",
-                ],
+                choices=TASKS,
             )
 
             domain_input = gr.CheckboxGroup(
                 label="Domain",
-                choices=[
-                    "clinical",
-                    "legal",
-                    "financial",
-                    "scientific",
-                    "news",
-                    "social media",
-                    "literature",
-                    "general",
-                ],
+                choices=DOMAINS,
             )
 
             countries_input = gr.CheckboxGroup(
                 label="Countries",
-                choices=[
-                    "Spain",
-                    "Mexico",
-                    "Argentina",
-                    "Colombia",
-                    "Peru",
-                    "Venezuela",
-                    "Chile",
-                    "Ecuador",
-                    "Guatemala",
-                    "Cuba",
-                    "Bolivia",
-                    "Dominican Republic",
-                    "Honduras",
-                    "Paraguay",
-                    "El Salvador",
-                    "Nicaragua",
-                    "Costa Rica",
-                    "Panama",
-                    "Uruguay",
-                    "Puerto Rico",
-                ],
+                choices=COUNTRIES,
             )
 
             languages_input = gr.CheckboxGroup(
                 label="Languages",
-                choices=[
-                    "spanish",
-                    "catalan",
-                    "basque",
-                    "galician",
-                    "guarani",
-                    "quechua",
-                    "aymara",
-                ],
+                choices=LANGUAGES,
             )
 
             update_btn = gr.Button("💾 Update Entry", variant="primary")

events_resource.py

@@ -4,6 +4,13 @@ import gradio as gr
 import pandas as pd
 from datasets import Dataset, load_dataset
 
+from constants import (
+    EVENT_TYPES,
+    TECHNICAL_LEVELS,
+    format_dataframe_for_display,
+    format_dataframe_for_html_display,
+)
+
 # Dataset configuration
 DATASET_NAME = "somosnlp/recursos-pln-es"
 CONFIG_NAME = "events"
@@ -176,27 +183,44 @@ def create_all_tab():
             show_label=False,
         )
 
-        # Load initial data
-        initial_df = load_data()
+        # Load and format initial data with clickable links
+        def get_formatted_data():
+            df = load_data()
+            return format_dataframe_for_display(
+                df,
+                url_columns=["youtube"],
+                hide_columns=["date_submitted"],
+            )
 
+        # Use Dataframe component with HTML rendering enabled
         table = gr.Dataframe(
-            value=initial_df,
+            value=get_formatted_data(),
             label=RESOURCE_TITLE,
             show_label=False,
             interactive=False,
-            wrap=True,
+            wrap=False,  # Disable wrapping to show full text in single lines
+            datatype="markdown",  # Enable HTML rendering
         )
 
         # Connect search functionality
+        def search_and_format(query):
+            initial_df = load_data()
+            filtered_df = search_and_filter_data(initial_df, query)
+            return format_dataframe_for_display(
+                filtered_df,
+                url_columns=["youtube"],
+                hide_columns=["date_submitted"],
+            )
+
         search_box.change(
-            fn=lambda query: search_and_filter_data(initial_df, query),
+            fn=search_and_format,
             inputs=search_box,
             outputs=table,
         )
 
         # Refresh button to reload data
         refresh_btn = gr.Button("🔄 Refresh Data", variant="secondary")
-        refresh_btn.click(fn=lambda: load_data(), outputs=table)
+        refresh_btn.click(fn=get_formatted_data, outputs=table)
 
         return table
 
@@ -235,7 +259,7 @@ def create_contribute_tab():
 
             tipo_input = gr.Dropdown(
                 label="Tipo *",
-                choices=["workshop", "talk", "AMA", "round table"],
+                choices=EVENT_TYPES,
                 info="Type of event (required)",
                 multiselect=False,
             )
@@ -254,7 +278,7 @@ def create_contribute_tab():
 
             nivel_tecnico_input = gr.Dropdown(
                 label="Nivel Técnico *",
-                choices=["1", "2", "3", "4", "5"],
+                choices=TECHNICAL_LEVELS,
                 info="Technical level from 1 (beginner) to 5 (expert) (required)",
                 multiselect=False,
             )
@@ -467,13 +491,13 @@ def create_edit_tab():
             bio_input = gr.Textbox(label="Bio *", lines=2, placeholder="Speaker bio")
             tipo_input = gr.Dropdown(
                 label="Tipo *",
-                choices=["workshop", "talk", "AMA", "round table"],
+                choices=EVENT_TYPES,
                 value="talk",
             )
             etiquetas_input = gr.Textbox(label="Etiquetas *", placeholder="Tags")
             tema_input = gr.Textbox(label="Tema *", placeholder="Topic")
             nivel_tecnico_input = gr.Dropdown(
-                label="Nivel Técnico *", choices=["1", "2", "3", "4", "5"], value="3"
+                label="Nivel Técnico *", choices=TECHNICAL_LEVELS, value="3"
             )
             fecha_input = gr.Textbox(label="Fecha *", placeholder="DD/MM/YYYY")
             youtube_input = gr.Textbox(

initiatives_resource.py

@@ -4,6 +4,14 @@ import gradio as gr
 import pandas as pd
 from datasets import Dataset, load_dataset
 
+from constants import (
+    COUNTRIES,
+    INITIATIVE_TYPES,
+    LANGUAGES,
+    format_dataframe_for_display,
+    format_dataframe_for_html_display,
+)
+
 # Dataset configuration
 DATASET_NAME = "somosnlp/recursos-pln-es"
 CONFIG_NAME = "initiatives"
@@ -147,27 +155,44 @@ def create_all_tab():
             show_label=False,
         )
 
-        # Load initial data
-        initial_df = load_data()
+        # Load and format initial data with clickable links
+        def get_formatted_data():
+            df = load_data()
+            return format_dataframe_for_display(
+                df,
+                url_columns=["website_url"],
+                hide_columns=["date_submitted"],
+            )
 
+        # Use Dataframe component with HTML rendering enabled
         table = gr.Dataframe(
-            value=initial_df,
+            value=get_formatted_data(),
             label=RESOURCE_TITLE,
             show_label=False,
             interactive=False,
-            wrap=True,
+            wrap=False,  # Disable wrapping to show full text in single lines
+            datatype="markdown",  # Enable HTML rendering
         )
 
         # Connect search functionality
+        def search_and_format(query):
+            initial_df = load_data()
+            filtered_df = search_and_filter_data(initial_df, query)
+            return format_dataframe_for_display(
+                filtered_df,
+                url_columns=["website_url"],
+                hide_columns=["date_submitted"],
+            )
+
         search_box.change(
-            fn=lambda query: search_and_filter_data(initial_df, query),
+            fn=search_and_format,
             inputs=search_box,
             outputs=table,
         )
 
         # Refresh button to reload data
         refresh_btn = gr.Button("🔄 Refresh Data", variant="secondary")
-        refresh_btn.click(fn=lambda: load_data(), outputs=table)
+        refresh_btn.click(fn=get_formatted_data, outputs=table)
 
         return table
 
@@ -193,15 +218,7 @@ def create_contribute_tab():
 
             type_input = gr.Dropdown(
                 label="Type *",
-                choices=[
-                    "project",
-                    "event",
-                    "research group",
-                    "community",
-                    "research institute",
-                    "non-profit",
-                    "OS company",
-                ],
+                choices=INITIATIVE_TYPES,
                 info="Type of initiative (required)",
                 multiselect=False,
             )
@@ -430,53 +447,16 @@ def create_edit_tab():
             name_input = gr.Textbox(label="Name *", placeholder="Initiative name")
             type_input = gr.Dropdown(
                 label="Type *",
-                choices=[
-                    "project",
-                    "event",
-                    "research group",
-                    "community",
-                    "research institute",
-                    "non-profit",
-                    "OS company",
-                ],
+                choices=INITIATIVE_TYPES,
                 value="project",
             )
             countries_input = gr.CheckboxGroup(
                 label="Countries *",
-                choices=[
-                    "Spain",
-                    "Mexico",
-                    "Argentina",
-                    "Colombia",
-                    "Peru",
-                    "Venezuela",
-                    "Chile",
-                    "Ecuador",
-                    "Guatemala",
-                    "Cuba",
-                    "Bolivia",
-                    "Dominican Republic",
-                    "Honduras",
-                    "Paraguay",
-                    "El Salvador",
-                    "Nicaragua",
-                    "Costa Rica",
-                    "Panama",
-                    "Uruguay",
-                    "Puerto Rico",
-                ],
+                choices=COUNTRIES,
             )
             languages_input = gr.CheckboxGroup(
                 label="Languages *",
-                choices=[
-                    "spanish",
-                    "catalan",
-                    "basque",
-                    "galician",
-                    "guarani",
-                    "quechua",
-                    "aymara",
-                ],
+                choices=LANGUAGES,
             )
             website_url_input = gr.Textbox(
                 label="Website URL *", placeholder="https://..."

models_resource.py

@@ -4,6 +4,8 @@ import gradio as gr
 import pandas as pd
 from datasets import Dataset, load_dataset
 
+from constants import format_dataframe_for_display, format_dataframe_for_html_display
+
 # Dataset configuration
 DATASET_NAME = "somosnlp/recursos-pln-es"
 CONFIG_NAME = "models"
@@ -160,27 +162,43 @@ def create_all_tab():
             show_label=False,
         )
 
-        # Load initial data
-        initial_df = load_data()
+        # Load and format initial data with clickable links
+        def get_formatted_data():
+            df = load_data()
+            return format_dataframe_for_display(
+                df,
+                url_columns=["hf_collection_url", "website_url", "paper_url"],
+                hide_columns=["date_submitted"],
+            )
 
         table = gr.Dataframe(
-            value=initial_df,
+            value=get_formatted_data(),
             label=RESOURCE_TITLE,
             show_label=False,
             interactive=False,
             wrap=True,
+            datatype="markdown",
         )
 
         # Connect search functionality
+        def search_and_format(query):
+            initial_df = load_data()
+            filtered_df = search_and_filter_data(initial_df, query)
+            return format_dataframe_for_display(
+                filtered_df,
+                url_columns=["hf_collection_url", "website_url", "paper_url"],
+                hide_columns=["date_submitted"],
+            )
+
         search_box.change(
-            fn=lambda query: search_and_filter_data(initial_df, query),
+            fn=search_and_format,
             inputs=search_box,
             outputs=table,
         )
 
         # Refresh button to reload data
         refresh_btn = gr.Button("🔄 Refresh Data", variant="secondary")
-        refresh_btn.click(fn=lambda: load_data(), outputs=table)
+        refresh_btn.click(fn=get_formatted_data, outputs=table)
 
         return table
 
@@ -256,22 +274,20 @@ def create_contribute_tab():
         )
 
 
-
 def search_entries(query: str) -> pd.DataFrame:
     """Search for entries by familia or URL."""
     if not query.strip():
         return pd.DataFrame()
-    
+
     df = load_data()
     if df.empty:
         return df
-    
+
     # Search in familia and hf_collection_url columns
-    mask = (
-        df['familia'].str.contains(query, case=False, na=False) |
-        df['hf_collection_url'].str.contains(query, case=False, na=False)
-    )
-    
+    mask = df["familia"].str.contains(query, case=False, na=False) | df[
+        "hf_collection_url"
+    ].str.contains(query, case=False, na=False)
+
     return df[mask]
 
 
@@ -279,72 +295,85 @@ def load_entry_for_edit(selected_entry: str) -> tuple:
     """Load a specific entry for editing."""
     if not selected_entry:
         return ("",) * 5  # Return empty values for all fields
-    
+
     df = load_data()
     if df.empty:
         return ("",) * 5
-    
+
     # Find the entry by familia or hf_collection_url
-    entry = df[df['familia'] == selected_entry].iloc[0] if (df['familia'] == selected_entry).any() else df[df['hf_collection_url'] == selected_entry].iloc[0]
-    
+    entry = (
+        df[df["familia"] == selected_entry].iloc[0]
+        if (df["familia"] == selected_entry).any()
+        else df[df["hf_collection_url"] == selected_entry].iloc[0]
+    )
+
     return (
-        entry['familia'],
-        entry['available_sizes'],
-        entry['hf_collection_url'],
-        entry['website_url'],
-        entry['paper_url']
+        entry["familia"],
+        entry["available_sizes"],
+        entry["hf_collection_url"],
+        entry["website_url"],
+        entry["paper_url"],
     )
 
 
 def update_entry(
     original_identifier: str,
-    familia: str, available_sizes: str, hf_collection_url: str,
-    website_url: str, paper_url: str, profile: gr.OAuthProfile | None
+    familia: str,
+    available_sizes: str,
+    hf_collection_url: str,
+    website_url: str,
+    paper_url: str,
+    profile: gr.OAuthProfile | None,
 ):
     """Update an existing entry."""
     if not profile:
         return "❌ Please log in to edit entries."
-    
+
     username = profile.username
     if not username:
         return "❌ Could not get username from profile."
-    
+
     if not original_identifier:
         return "❌ No entry selected to edit."
-    
+
     if not hf_collection_url.strip():
         return "❌ Hugging Face Collection URL is required."
-    
+
     # Validate URLs
-    for url_field, url_value in [("Hugging Face Collection URL", hf_collection_url), 
-                                  ("Website URL", website_url), ("Paper URL", paper_url)]:
+    for url_field, url_value in [
+        ("Hugging Face Collection URL", hf_collection_url),
+        ("Website URL", website_url),
+        ("Paper URL", paper_url),
+    ]:
         if url_value.strip() and not validate_url(url_value):
             return f"❌ Invalid {url_field}. Please provide a valid URL."
-    
+
     # Validate available_sizes format
     if available_sizes.strip() and not validate_sizes(available_sizes):
         return "❌ Invalid available sizes format. Use comma-separated numbers (e.g., '0.1, 1.3, 7, 14')."
-    
+
     try:
         # Load existing dataset
         existing_dataset = load_dataset(DATASET_NAME, CONFIG_NAME, split="train")
         existing_df = existing_dataset.to_pandas()
-        
+
         # Find and update the entry
-        mask = (existing_df['familia'] == original_identifier) | (existing_df['hf_collection_url'] == original_identifier)
+        mask = (existing_df["familia"] == original_identifier) | (
+            existing_df["hf_collection_url"] == original_identifier
+        )
         if not mask.any():
             return f"❌ Entry '{original_identifier}' not found."
-        
+
         # Update the entry
         current_time = datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ")
-        
-        existing_df.loc[mask, 'familia'] = familia
-        existing_df.loc[mask, 'available_sizes'] = available_sizes
-        existing_df.loc[mask, 'hf_collection_url'] = hf_collection_url
-        existing_df.loc[mask, 'website_url'] = website_url
-        existing_df.loc[mask, 'paper_url'] = paper_url
-        existing_df.loc[mask, 'date_submitted'] = current_time
-        
+
+        existing_df.loc[mask, "familia"] = familia
+        existing_df.loc[mask, "available_sizes"] = available_sizes
+        existing_df.loc[mask, "hf_collection_url"] = hf_collection_url
+        existing_df.loc[mask, "website_url"] = website_url
+        existing_df.loc[mask, "paper_url"] = paper_url
+        existing_df.loc[mask, "date_submitted"] = current_time
+
         # Convert back to Dataset and push to hub
         updated_dataset = Dataset.from_pandas(existing_df)
         updated_dataset.push_to_hub(
@@ -352,104 +381,131 @@ def update_entry(
             config_name=CONFIG_NAME,
             commit_message=f"Update model entry: {familia or hf_collection_url} (edited by {username})",
         )
-        
+
         return f"✅ Successfully updated '{familia or hf_collection_url}'!"
-        
+
     except Exception as e:
         return f"❌ Error updating entry: {str(e)}"
 
 
-
 def create_edit_tab():
     """Create the edit tab for modifying existing entries."""
     with gr.TabItem("✏️ Edit", id=f"{RESOURCE_TYPE}_edit"):
         gr.Markdown(f"### Edit Existing {RESOURCE_TITLE}")
         gr.Markdown("Please log in to edit entries:")
         login_button = gr.LoginButton(elem_id=f"{RESOURCE_TYPE}-edit-oauth-button")
-        
+
         gr.Markdown("Search for an entry to edit:")
-        
+
         with gr.Row():
             search_input = gr.Textbox(
                 label="Search by familia or collection URL",
                 placeholder="Enter model familia or Hugging Face collection URL...",
-                scale=3
+                scale=3,
             )
             search_btn = gr.Button("🔍 Search", scale=1)
-        
+
         search_results = gr.Dropdown(
-            label="Select entry to edit",
-            choices=[],
-            interactive=True
+            label="Select entry to edit", choices=[], interactive=True
         )
-        
+
         gr.Markdown("---")
         gr.Markdown("**Edit the selected entry:**")
-        
+
         with gr.Column(visible=False) as edit_form:
-            hf_collection_url_input = gr.Textbox(label="Hugging Face Collection URL *", placeholder="https://huggingface.co/collections/...")
-            familia_input = gr.Textbox(label="Familia", placeholder="e.g., BERT, GPT, T5...")
+            hf_collection_url_input = gr.Textbox(
+                label="Hugging Face Collection URL *",
+                placeholder="https://huggingface.co/collections/...",
+            )
+            familia_input = gr.Textbox(
+                label="Familia", placeholder="e.g., BERT, GPT, T5..."
+            )
             available_sizes_input = gr.Textbox(
                 label="Available Sizes (in B parameters)",
                 placeholder="e.g., 0.1, 1.3, 7, 14",
-                info="Comma-separated list of model sizes in billions of parameters"
+                info="Comma-separated list of model sizes in billions of parameters",
+            )
+            website_url_input = gr.Textbox(
+                label="Website URL", placeholder="https://..."
+            )
+            paper_url_input = gr.Textbox(
+                label="Paper URL", placeholder="https://arxiv.org/..."
             )
-            website_url_input = gr.Textbox(label="Website URL", placeholder="https://...")
-            paper_url_input = gr.Textbox(label="Paper URL", placeholder="https://arxiv.org/...")
-            
+
             update_btn = gr.Button("💾 Update Entry", variant="primary")
             result_msg = gr.Markdown()
-        
+
         # Store the original identifier for updating
         original_identifier_state = gr.State("")
-        
+
         def search_and_update_dropdown(query):
             results_df = search_entries(query)
             if results_df.empty:
                 return gr.Dropdown(choices=[], value=None)
             else:
                 # Use familia if available, otherwise use hf_collection_url
-                choices = [entry if entry else url for entry, url in zip(results_df['familia'].fillna(''), results_df['hf_collection_url'])]
+                choices = [
+                    entry if entry else url
+                    for entry, url in zip(
+                        results_df["familia"].fillna(""),
+                        results_df["hf_collection_url"],
+                    )
+                ]
                 return gr.Dropdown(choices=choices, value=None)
-        
+
         def load_entry_and_show_form(selected_entry):
             if not selected_entry:
                 return (gr.Column(visible=False), "", *[("",) * 5])
-            
+
             entry_data = load_entry_for_edit(selected_entry)
             return (gr.Column(visible=True), selected_entry, *entry_data)
-        
+
         # Event handlers
         search_btn.click(
             fn=search_and_update_dropdown,
             inputs=[search_input],
-            outputs=[search_results]
+            outputs=[search_results],
         )
-        
+
         search_results.change(
             fn=load_entry_and_show_form,
             inputs=[search_results],
             outputs=[
-                edit_form, original_identifier_state,
-                familia_input, available_sizes_input, hf_collection_url_input,
-                website_url_input, paper_url_input
-            ]
+                edit_form,
+                original_identifier_state,
+                familia_input,
+                available_sizes_input,
+                hf_collection_url_input,
+                website_url_input,
+                paper_url_input,
+            ],
         )
-        
+
         update_btn.click(
             fn=update_entry,
             inputs=[
                 original_identifier_state,
-                familia_input, available_sizes_input, hf_collection_url_input,
-                website_url_input, paper_url_input
+                familia_input,
+                available_sizes_input,
+                hf_collection_url_input,
+                website_url_input,
+                paper_url_input,
             ],
-            outputs=[result_msg]
+            outputs=[result_msg],
         )
-        
+
         return (
-            search_input, search_btn, search_results, edit_form,
-            familia_input, available_sizes_input, hf_collection_url_input,
-            website_url_input, paper_url_input, update_btn, result_msg
+            search_input,
+            search_btn,
+            search_results,
+            edit_form,
+            familia_input,
+            available_sizes_input,
+            hf_collection_url_input,
+            website_url_input,
+            paper_url_input,
+            update_btn,
+            result_msg,
         )
 
 

shared_tasks_resource.py

@@ -4,6 +4,8 @@ import gradio as gr
 import pandas as pd
 from datasets import Dataset, load_dataset
 
+from constants import format_dataframe_for_display, format_dataframe_for_html_display
+
 # Dataset configuration
 DATASET_NAME = "somosnlp/recursos-pln-es"
 CONFIG_NAME = "shared_tasks"
@@ -176,27 +178,43 @@ def create_all_tab():
             show_label=False,
         )
 
-        # Load initial data
-        initial_df = load_data()
+        # Load and format initial data with clickable links
+        def get_formatted_data():
+            df = load_data()
+            return format_dataframe_for_display(
+                df,
+                url_columns=["more_info_url"],
+                hide_columns=["date_submitted"],
+            )
 
         table = gr.Dataframe(
-            value=initial_df,
+            value=get_formatted_data(),
             label=RESOURCE_TITLE,
             show_label=False,
             interactive=False,
             wrap=True,
+            datatype="markdown",
         )
 
         # Connect search functionality
+        def search_and_format(query):
+            initial_df = load_data()
+            filtered_df = search_and_filter_data(initial_df, query)
+            return format_dataframe_for_display(
+                filtered_df,
+                url_columns=["more_info_url"],
+                hide_columns=["date_submitted"],
+            )
+
         search_box.change(
-            fn=lambda query: search_and_filter_data(initial_df, query),
+            fn=search_and_format,
             inputs=search_box,
             outputs=table,
         )
 
         # Refresh button to reload data
         refresh_btn = gr.Button("🔄 Refresh Data", variant="secondary")
-        refresh_btn.click(fn=lambda: load_data(), outputs=table)
+        refresh_btn.click(fn=get_formatted_data, outputs=table)
 
         return table
 
@@ -289,22 +307,20 @@ def create_contribute_tab():
         )
 
 
-
 def search_entries(query: str) -> pd.DataFrame:
     """Search for entries by name or conference."""
     if not query.strip():
         return pd.DataFrame()
-    
+
     df = load_data()
     if df.empty:
         return df
-    
+
     # Search in name and conference_name columns
-    mask = (
-        df['name'].str.contains(query, case=False, na=False) |
-        df['conference_name'].str.contains(query, case=False, na=False)
-    )
-    
+    mask = df["name"].str.contains(query, case=False, na=False) | df[
+        "conference_name"
+    ].str.contains(query, case=False, na=False)
+
     return df[mask]
 
 
@@ -312,81 +328,97 @@ def load_entry_for_edit(selected_entry: str) -> tuple:
     """Load a specific entry for editing."""
     if not selected_entry:
         return ("",) * 7  # Return empty values for all fields
-    
+
     df = load_data()
     if df.empty:
         return ("",) * 7
-    
+
     # Find the entry by name
-    entry = df[df['name'] == selected_entry].iloc[0]
-    
+    entry = df[df["name"] == selected_entry].iloc[0]
+
     return (
-        entry['name'],
-        entry['conference_name'],
-        entry['workshop_date'],
-        entry['registration_deadline'],
-        entry['data_available_date'],
-        entry['submission_deadline'],
-        entry['more_info_url']
+        entry["name"],
+        entry["conference_name"],
+        entry["workshop_date"],
+        entry["registration_deadline"],
+        entry["data_available_date"],
+        entry["submission_deadline"],
+        entry["more_info_url"],
     )
 
 
 def update_entry(
     original_name: str,
-    name: str, conference_name: str, workshop_date: str, registration_deadline: str,
-    data_available_date: str, submission_deadline: str, more_info_url: str,
+    name: str,
+    conference_name: str,
+    workshop_date: str,
+    registration_deadline: str,
+    data_available_date: str,
+    submission_deadline: str,
+    more_info_url: str,
     profile: gr.OAuthProfile | None,
 ):
     """Update an existing entry."""
     if not profile:
         return "❌ Please log in to edit entries."
-    
+
     username = profile.username
     if not username:
         return "❌ Could not get username from profile."
-    
+
     if not original_name:
         return "❌ No entry selected to edit."
-    
+
     # Validate required fields
-    required_fields = [name, conference_name, workshop_date, registration_deadline, 
-                      data_available_date, submission_deadline, more_info_url]
+    required_fields = [
+        name,
+        conference_name,
+        workshop_date,
+        registration_deadline,
+        data_available_date,
+        submission_deadline,
+        more_info_url,
+    ]
     if not all(field.strip() for field in required_fields):
         return "❌ All fields are required."
-    
+
     # Validate URL
     if not validate_url(more_info_url):
         return "❌ Invalid URL. Please provide a valid URL."
-    
+
     # Validate dates
-    date_fields = [("Workshop Date", workshop_date), ("Registration Deadline", registration_deadline),
-                   ("Data Available Date", data_available_date), ("Submission Deadline", submission_deadline)]
+    date_fields = [
+        ("Workshop Date", workshop_date),
+        ("Registration Deadline", registration_deadline),
+        ("Data Available Date", data_available_date),
+        ("Submission Deadline", submission_deadline),
+    ]
     for field_name, date_value in date_fields:
         if not validate_date(date_value):
             return f"❌ Invalid {field_name}. Please use DD/MM/YYYY format."
-    
+
     try:
         # Load existing dataset
         existing_dataset = load_dataset(DATASET_NAME, CONFIG_NAME, split="train")
         existing_df = existing_dataset.to_pandas()
-        
+
         # Find and update the entry
-        mask = existing_df['name'] == original_name
+        mask = existing_df["name"] == original_name
         if not mask.any():
             return f"❌ Entry '{original_name}' not found."
-        
+
         # Update the entry
         current_time = datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ")
-        
-        existing_df.loc[mask, 'name'] = name
-        existing_df.loc[mask, 'conference_name'] = conference_name
-        existing_df.loc[mask, 'workshop_date'] = workshop_date
-        existing_df.loc[mask, 'registration_deadline'] = registration_deadline
-        existing_df.loc[mask, 'data_available_date'] = data_available_date
-        existing_df.loc[mask, 'submission_deadline'] = submission_deadline
-        existing_df.loc[mask, 'more_info_url'] = more_info_url
-        existing_df.loc[mask, 'date_submitted'] = current_time
-        
+
+        existing_df.loc[mask, "name"] = name
+        existing_df.loc[mask, "conference_name"] = conference_name
+        existing_df.loc[mask, "workshop_date"] = workshop_date
+        existing_df.loc[mask, "registration_deadline"] = registration_deadline
+        existing_df.loc[mask, "data_available_date"] = data_available_date
+        existing_df.loc[mask, "submission_deadline"] = submission_deadline
+        existing_df.loc[mask, "more_info_url"] = more_info_url
+        existing_df.loc[mask, "date_submitted"] = current_time
+
         # Convert back to Dataset and push to hub
         updated_dataset = Dataset.from_pandas(existing_df)
         updated_dataset.push_to_hub(
@@ -394,109 +426,136 @@ def update_entry(
             config_name=CONFIG_NAME,
             commit_message=f"Update shared task entry: {name} (edited by {username})",
         )
-        
+
         return f"✅ Successfully updated '{name}'!"
-        
+
     except Exception as e:
         return f"❌ Error updating entry: {str(e)}"
 
 
-
 def create_edit_tab():
     """Create the edit tab for modifying existing entries."""
     with gr.TabItem("✏️ Edit", id=f"{RESOURCE_TYPE}_edit"):
         gr.Markdown(f"### Edit Existing {RESOURCE_TITLE}")
         gr.Markdown("Please log in to edit entries:")
         login_button = gr.LoginButton(elem_id=f"{RESOURCE_TYPE}-edit-oauth-button")
-        
+
         gr.Markdown("Search for an entry to edit:")
-        
+
         with gr.Row():
             search_input = gr.Textbox(
                 label="Search by name or conference",
                 placeholder="Enter shared task name or conference name...",
-                scale=3
+                scale=3,
             )
             search_btn = gr.Button("🔍 Search", scale=1)
-        
+
         search_results = gr.Dropdown(
-            label="Select entry to edit",
-            choices=[],
-            interactive=True
+            label="Select entry to edit", choices=[], interactive=True
         )
-        
+
         gr.Markdown("---")
         gr.Markdown("**Edit the selected entry:**")
-        
+
         with gr.Column(visible=False) as edit_form:
             name_input = gr.Textbox(label="Name *", placeholder="Shared task name")
-            conference_name_input = gr.Textbox(label="Conference Name *", placeholder="Conference or workshop name")
-            
+            conference_name_input = gr.Textbox(
+                label="Conference Name *", placeholder="Conference or workshop name"
+            )
+
             gr.Markdown("**Important Dates** (all required, format: DD/MM/YYYY)")
             with gr.Row():
-                workshop_date_input = gr.Textbox(label="Workshop Date *", placeholder="DD/MM/YYYY")
-                registration_deadline_input = gr.Textbox(label="Registration Deadline *", placeholder="DD/MM/YYYY")
+                workshop_date_input = gr.Textbox(
+                    label="Workshop Date *", placeholder="DD/MM/YYYY"
+                )
+                registration_deadline_input = gr.Textbox(
+                    label="Registration Deadline *", placeholder="DD/MM/YYYY"
+                )
             with gr.Row():
-                data_available_date_input = gr.Textbox(label="Data Available Date *", placeholder="DD/MM/YYYY")
-                submission_deadline_input = gr.Textbox(label="Submission Deadline *", placeholder="DD/MM/YYYY")
-            
-            more_info_url_input = gr.Textbox(label="More Info URL *", placeholder="https://...")
-            
+                data_available_date_input = gr.Textbox(
+                    label="Data Available Date *", placeholder="DD/MM/YYYY"
+                )
+                submission_deadline_input = gr.Textbox(
+                    label="Submission Deadline *", placeholder="DD/MM/YYYY"
+                )
+
+            more_info_url_input = gr.Textbox(
+                label="More Info URL *", placeholder="https://..."
+            )
+
             update_btn = gr.Button("💾 Update Entry", variant="primary")
             result_msg = gr.Markdown()
-        
+
         # Store the original name for updating
         original_name_state = gr.State("")
-        
+
         def search_and_update_dropdown(query):
             results_df = search_entries(query)
             if results_df.empty:
                 return gr.Dropdown(choices=[], value=None)
             else:
-                choices = results_df['name'].tolist()
+                choices = results_df["name"].tolist()
                 return gr.Dropdown(choices=choices, value=None)
-        
+
         def load_entry_and_show_form(selected_entry):
             if not selected_entry:
                 return (gr.Column(visible=False), "", *[("",) * 7])
-            
+
             entry_data = load_entry_for_edit(selected_entry)
             return (gr.Column(visible=True), selected_entry, *entry_data)
-        
+
         # Event handlers
         search_btn.click(
             fn=search_and_update_dropdown,
             inputs=[search_input],
-            outputs=[search_results]
+            outputs=[search_results],
         )
-        
+
         search_results.change(
             fn=load_entry_and_show_form,
             inputs=[search_results],
             outputs=[
-                edit_form, original_name_state,
-                name_input, conference_name_input, workshop_date_input,
-                registration_deadline_input, data_available_date_input,
-                submission_deadline_input, more_info_url_input
-            ]
+                edit_form,
+                original_name_state,
+                name_input,
+                conference_name_input,
+                workshop_date_input,
+                registration_deadline_input,
+                data_available_date_input,
+                submission_deadline_input,
+                more_info_url_input,
+            ],
         )
-        
+
         update_btn.click(
             fn=update_entry,
             inputs=[
                 original_name_state,
-                name_input, conference_name_input, workshop_date_input,
-                registration_deadline_input, data_available_date_input,
-                submission_deadline_input, more_info_url_input,
+                name_input,
+                conference_name_input,
+                workshop_date_input,
+                registration_deadline_input,
+                data_available_date_input,
+                submission_deadline_input,
+                more_info_url_input,
             ],
-            outputs=[result_msg]
+            outputs=[result_msg],
         )
-        
+
         return (
-            search_input, search_btn, search_results, edit_form,
-            name_input, conference_name_input, workshop_date_input,
-            registration_deadline_input, data_available_date_input,
-            submission_deadline_input, more_info_url_input, update_btn, result_msg
+            search_input,
+            search_btn,
+            search_results,
+            edit_form,
+            name_input,
+            conference_name_input,
+            workshop_date_input,
+            registration_deadline_input,
+            data_available_date_input,
+            submission_deadline_input,
+            more_info_url_input,
+            update_btn,
+            result_msg,
         )