Upload 4 files

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+data/samaritanus.db filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,14 +1,74 @@
 ---
-title: Samaritanus
-emoji: 🐨
+title: Samaritan Torah Search
+emoji: 📖
 colorFrom: blue
-colorTo: gray
+colorTo: indigo
 sdk: gradio
-sdk_version: 5.33.1
+sdk_version: 4.19.2
 app_file: app.py
 pinned: false
-license: mit
-short_description: Samaritan Torah Search
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# Samaritan Torah Search
+
+A Gradio application for searching through the Samaritan Torah text with both exact and fuzzy matching capabilities.
+
+## Features
+
+- **Search Types**:
+  - Exact Match: Find precise matches of your search terms
+  - Fuzzy Match: Find similar matches with more lenient matching rules
+
+- **Search Capabilities**:
+  - Full text search across all verses
+  - Highlighted search terms in results
+  - RTL (Right-to-Left) text support for Hebrew
+
+- **User Interface**:
+  - Clean, modern design
+  - Responsive layout
+  - Custom Hebrew font (NarkisClassic)
+  - Search button and Enter key support
+
+## Usage
+
+1. Enter your search term in the search box
+2. Choose between "Exact Match" or "Fuzzy Match"
+3. Click the Search button or press Enter
+4. View the results with highlighted matches
+
+## Technical Details
+
+- Built with Gradio
+- Uses SQLite with FTS5 for full-text search
+- Custom CSS for RTL support and styling
+- Responsive design for all screen sizes
+
+## Development
+
+### Local Setup
+
+1. Install dependencies:
+   ```bash
+   pip install -r requirements.txt
+   ```
+
+2. Run the app:
+   ```bash
+   python app.py
+   ```
+
+### Hugging Face Spaces
+
+This app is deployed on Hugging Face Spaces. The deployment is automatic when changes are pushed to the repository.
+
+## Data
+
+The app uses a SQLite database with FTS5 enabled for efficient full-text search. The database contains the complete text of the Samaritan Torah with the following structure:
+
+- `verses` table: Contains the main verse data
+- `verses_fts` virtual table: FTS5 table for full-text search
+
+## License
+
+MIT License 

app.py

@@ -0,0 +1,225 @@
+import gradio as gr
+import sqlite3
+import pandas as pd
+from pathlib import Path
+import re
+
+def get_db_connection():
+    db_path = Path(__file__).parent / 'data' / 'samaritanus.db'
+    return sqlite3.connect(db_path)
+
+def highlight_text(text, search_term):
+    if not search_term:
+        return text
+    pattern = re.escape(search_term)
+    return re.sub(f'({pattern})', r'<mark>\1</mark>', text, flags=re.IGNORECASE)
+
+def search_verses(search_term, exact_match, page=1, per_page=10):
+    conn = get_db_connection()
+    
+    if exact_match:
+        # Exact match
+        query = """
+        SELECT v.*, 
+               COUNT(*) OVER() as total_count
+        FROM verses v
+        JOIN verses_fts fts ON v.id = fts.rowid
+        WHERE verses_fts MATCH ?
+        ORDER BY v.book, v.chapter, v.verse
+        LIMIT ? OFFSET ?
+        """
+        params = [search_term, per_page, (page - 1) * per_page]
+    else:
+        # Fuzzy match using SQLite's built-in fuzzy search
+        words = search_term.split()
+        # Permissive fuzzy match - use more lenient matching
+        fuzzy_terms = []
+        for word in words:
+            # Add the word itself
+            fuzzy_terms.append(word)
+            # Add prefix match
+            fuzzy_terms.append(f"{word}*")
+        match_expr = " OR ".join(fuzzy_terms)
+        
+        query = """
+        SELECT v.*, 
+               COUNT(*) OVER() as total_count
+        FROM verses v
+        JOIN verses_fts fts ON v.id = fts.rowid
+        WHERE verses_fts MATCH ?
+        ORDER BY v.book, v.chapter, v.verse
+        LIMIT ? OFFSET ?
+        """
+        params = [match_expr, per_page, (page - 1) * per_page]
+    
+    df = pd.read_sql_query(query, conn, params=params)
+    conn.close()
+    return df
+
+def format_results(results):
+    if results.empty:
+        return "No results found"
+    
+    formatted_results = []
+    for _, row in results.iterrows():
+        verse_ref = f"{row['book_name']} {row['chapter']}:{row['verse']}"
+        verse_text = row['text']
+        formatted_results.append(f"<div style='text-align: right; direction: rtl;'><b>{verse_ref}</b><br>{verse_text}</div><hr>")
+    
+    return "\n".join(formatted_results)
+
+def search(search_term, exact_match, page=1):
+    if not search_term:
+        return "Please enter a search term", 1, 1
+    
+    results = search_verses(search_term, exact_match, page)
+    
+    if results.empty:
+        return "No results found", 1, 1
+    
+    total_results = results['total_count'].iloc[0]
+    total_pages = (total_results + 9) // 10  # Round up division
+    
+    formatted_results = format_results(results)
+    result_count = f"Found {total_results} {'result' if total_results == 1 else 'results'}"
+    
+    return f"{result_count}\n\n{formatted_results}", page, total_pages
+
+# Custom CSS
+css = """
+@font-face {
+    font-family: 'NarkisClassic';
+    src: url('https://raw.githubusercontent.com/johnlockejrr/samaritanus_search/main/public/fonts/NarkisClassic-Regular.woff') format('woff');
+    font-weight: normal;
+    font-style: normal;
+}
+@font-face {
+    font-family: 'NarkisClassic';
+    src: url('https://raw.githubusercontent.com/johnlockejrr/samaritanus_search/main/public/fonts/NarkisClassic-Bold.woff') format('woff');
+    font-weight: bold;
+    font-style: normal;
+}
+.gradio-container {
+    font-family: 'NarkisClassic', serif;
+}
+mark {
+    background-color: yellow;
+    font-weight: bold;
+}
+"""
+
+# Create the Gradio interface
+with gr.Blocks(css=css) as demo:
+    gr.Markdown("# Samaritan Torah Search")
+    
+    # Store current page in session state
+    current_page = gr.State(1)
+    total_pages = gr.State(1)
+    
+    with gr.Row():
+        search_input = gr.Textbox(
+            label="Search",
+            placeholder="Enter search term...",
+            scale=4
+        )
+        search_type = gr.Radio(
+            choices=["Exact Match", "Fuzzy Match"],
+            value="Exact Match",
+            label="Search Type",
+            scale=1
+        )
+    
+    search_button = gr.Button("Search")
+    
+    with gr.Row():
+        output = gr.Markdown()
+    
+    with gr.Row(visible=False) as pagination_row:
+        with gr.Column(scale=1):
+            prev_button = gr.Button("←", size="sm")
+        with gr.Column(scale=3):
+            page_info = gr.Markdown()
+        with gr.Column(scale=1):
+            next_button = gr.Button("→", size="sm")
+    
+    def search_wrapper(search_term, search_type, page):
+        exact_match = search_type == "Exact Match"
+        return search(search_term, exact_match, page)
+    
+    def reset_page():
+        return 1, 1
+    
+    def update_pagination(page, total):
+        if total > 1:
+            return gr.Row.update(visible=True), f"Page {page} of {total}"
+        return gr.Row.update(visible=False), ""
+    
+    def prev_page(page, total):
+        if page > 1:
+            return page - 1
+        return page
+    
+    def next_page(page, total):
+        if page < total:
+            return page + 1
+        return page
+    
+    # Search button click
+    search_button.click(
+        fn=reset_page,
+        outputs=[current_page, total_pages]
+    ).then(
+        fn=search_wrapper,
+        inputs=[search_input, search_type, current_page],
+        outputs=[output, current_page, total_pages]
+    ).then(
+        fn=update_pagination,
+        inputs=[current_page, total_pages],
+        outputs=[pagination_row, page_info]
+    )
+    
+    # Search input submit
+    search_input.submit(
+        fn=reset_page,
+        outputs=[current_page, total_pages]
+    ).then(
+        fn=search_wrapper,
+        inputs=[search_input, search_type, current_page],
+        outputs=[output, current_page, total_pages]
+    ).then(
+        fn=update_pagination,
+        inputs=[current_page, total_pages],
+        outputs=[pagination_row, page_info]
+    )
+    
+    # Pagination buttons
+    prev_button.click(
+        fn=prev_page,
+        inputs=[current_page, total_pages],
+        outputs=current_page
+    ).then(
+        fn=search_wrapper,
+        inputs=[search_input, search_type, current_page],
+        outputs=[output, current_page, total_pages]
+    ).then(
+        fn=update_pagination,
+        inputs=[current_page, total_pages],
+        outputs=[pagination_row, page_info]
+    )
+    
+    next_button.click(
+        fn=next_page,
+        inputs=[current_page, total_pages],
+        outputs=current_page
+    ).then(
+        fn=search_wrapper,
+        inputs=[search_input, search_type, current_page],
+        outputs=[output, current_page, total_pages]
+    ).then(
+        fn=update_pagination,
+        inputs=[current_page, total_pages],
+        outputs=[pagination_row, page_info]
+    )
+
+if __name__ == "__main__":
+    demo.launch() 

data/samaritanus.db

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d4083e6bfb2a3bb13cd06170f3aa46d4b5bb75f494aeaa7f1f831f3575d1380
+size 1327104

requirements.txt

@@ -0,0 +1,3 @@
+gradio>=4.19.2
+pandas>=2.2.0
+sqlite3