translation cache for app version

app.py

@@ -5,13 +5,12 @@ import time
 import requests
 import pytz
 import unittest
+import sqlite3  # Import sqlite3 for database handling
 
 import gradio as gr
-from utils import process_json_files, flatten_text_with_line_breaks, calculate_tanach_statistics, build_word_index
-
-import logging
 from deep_translator import GoogleTranslator
 from deep_translator.exceptions import NotValidLength, RequestError
+from utils import process_json_files, flatten_text_with_line_breaks, calculate_tanach_statistics, build_word_index
 
 # Set up logging
 logging.basicConfig(level=logging.DEBUG, format='%(asctime)s - %(levelname)s - %(message)s')
@@ -20,6 +19,26 @@ logging.basicConfig(level=logging.DEBUG, format='%(asctime)s - %(levelname)s - %
 TANACH_DATA = process_json_files(1, 39)
 WORD_INDEX = build_word_index(TANACH_DATA)
 
+# --- Database Setup ---
+# Use a connection function to ensure each thread gets its own connection
+def get_db_connection():
+    conn = sqlite3.connect('translation_cache.db')
+    conn.row_factory = sqlite3.Row  # This line allows accessing columns by name
+    return conn
+
+# Create the database table if it doesn't exist
+with get_db_connection() as conn:
+    cursor = conn.cursor()
+    cursor.execute('''
+        CREATE TABLE IF NOT EXISTS translations (
+        book_id INTEGER,
+        chapter_id INTEGER,
+        english_text TEXT,
+        PRIMARY KEY (book_id, chapter_id)
+        )
+    ''')
+    conn.commit()
+
 # --- Utility Functions ---
 
 def get_current_word_data(client_time_str):
@@ -43,46 +62,48 @@ def get_formatted_verse(book_id, chapter_id, verse_id, highlight_word=True):
     chapter_text = TANACH_DATA[book_id]["text"][chapter_id]
     flattened_chapter = flatten_text_with_line_breaks(chapter_text)
 
-    # Highlight the word *before* joining with <br>
-    if highlight_word and 0 <= verse_id - 1 < len(flattened_chapter):
-        flattened_chapter[verse_id - 1] = \
-          f"<span class='highlight'>{flattened_chapter[verse_id - 1]}</span>"
-
     return '<br>'.join(flattened_chapter)
 
+def translate_chapter(hebrew_chapter, book_id, chapter_id):
+    """Translates a Hebrew chapter to English, caching the result in the database."""
 
-def translate_verse(hebrew_verse, highlight_word=True):
-    """Translates a Hebrew verse to English, splitting into chunks if necessary."""
-    try:
-        translator = GoogleTranslator(source='iw', target='en')
-        max_length = 2000  # Slightly below the limit to be safe
-        translated_text = ""
+    # Get a new database connection for this thread
+    with get_db_connection() as conn:
+        cursor = conn.cursor()
 
-        # Split the verse into chunks smaller than the max length
-        chunks = [hebrew_verse[i:i + max_length] for i in range(0, len(hebrew_verse), max_length)]
+        # Check if translation exists in the database
+        cursor.execute(
+        "SELECT english_text FROM translations WHERE book_id=? AND chapter_id=?",
+        (book_id, chapter_id)
+        )
+        result = cursor.fetchone()
 
-        for chunk_index, chunk in enumerate(chunks):
-            # Translate the current chunk
-            translated_chunk = translator.translate(chunk)
+        if result:
+            return result['english_text'].split('\n') # Retrieve from database and split into lines
 
-            # If it's not the first chunk, find the last line break and start from there
-            if chunk_index > 0:
-                last_line_break = translated_chunk.rfind('<br>', 0, 100) # Find last <br> in first 100 chars
-                if last_line_break != -1:
-                    translated_text += translated_chunk[last_line_break + 4:] # Add from after <br>
-                else:
-                    translated_text += translated_chunk
-            else:
-                translated_text += translated_chunk
+        try:
+            translator = GoogleTranslator(source='iw', target='en')
+            max_length = 2000 # Slightly below the limit to be safe
+            translated_text = ""
 
-        return translated_text
+            # Split the chapter into chunks smaller than the max length
+            chunks = [hebrew_chapter[i:i + max_length] for i in range(0, len(hebrew_chapter), max_length)]
 
-    except RequestError as e:
-        logging.warning(f"Translation failed: Request Error - {e}")
-        return "Translation unavailable: Request Error"
+            for chunk in chunks:
+                translated_text += translator.translate(chunk)
 
+            # Store the translation in the database
+            cursor.execute(
+                "INSERT INTO translations (book_id, chapter_id, english_text) VALUES (?, ?, ?)",
+                (book_id, chapter_id, translated_text)
+            )
+            conn.commit()
 
-# --- Gradio Interface ---
+            return translated_text.split('\n') # Return as list of lines
+
+        except RequestError as e:
+            logging.warning(f"Translation failed: Request Error - {e}")
+            return ["Translation unavailable: Request Error"]
 
 def update_tanach_display(client_time_str, timezone):
     """Updates the Gradio interface with client time, verse info, and translations."""
@@ -118,56 +139,66 @@ def update_tanach_display(client_time_str, timezone):
     """
 
     # Get and format Hebrew and English verses
-    hebrew_verse = get_formatted_verse(book_id, chapter_id, verse_id)
-    english_verse = translate_verse('\n'.join(hebrew_verse.split('<br>')), highlight_word=False)
+    hebrew_verse = get_formatted_verse(book_id, chapter_id, verse_id, highlight_word=False)
+    hebrew_verses = hebrew_verse.split("<br>")
 
-    return verse_info, hebrew_verse, english_verse
+    # Translate the entire chapter and get the correct verse
+    hebrew_chapter = flatten_text_with_line_breaks(TANACH_DATA[book_id]["text"][chapter_id])
+    english_chapter = translate_chapter('\n'.join(hebrew_chapter), book_id, chapter_id)
 
+    # Highlight the current verse in both Hebrew and English
+    hebrew_verses[verse_id - 1] = f"<span class='highlight'>{hebrew_verses[verse_id - 1]}</span>"
+    english_chapter[verse_id - 1] = f"<span class='highlight'>{english_chapter[verse_id - 1]}</span>"
 
-# --- Gradio Interface ---
+    # Join the verses back with <br> for display
+    hebrew_verse = "<br>".join(hebrew_verses)
+    english_verse = "<br>".join(english_chapter)
+
+    return verse_info, hebrew_verse, english_verse
 
+# --- Gradio Interface ---
 with gr.Blocks(css="""
- .container {
- display: flex;
- flex-direction: column;
- align-items: center;
- font-family: 'Times New Roman', serif;
- }
- /* Add this highlight class styling */
- .highlight {
- background-color: #FFFF00; /* Yellow highlight */
- padding: 2px 5px;
- border-radius: 5px;
- }
- #verse-info {
- margin-bottom: 20px;
- text-align: center;
- }
- #verses {
- display: flex;
- flex-direction: row;
- justify-content: center;
- align-items: flex-start;
- gap: 50px;
- }
- #hebrew-verse {
- font-size: 18px;
- line-height: 1.5;
- margin-bottom: 20px;
- text-align: right;
- direction: rtl;
- }
- #english-verse {
- font-size: 18px;
- line-height: 1.5;
- margin-bottom: 20px;
- }
+.container {
+display: flex;
+flex-direction: column;
+align-items: center;
+font-family: 'Times New Roman', serif;
+}
+/* Add this highlight class styling */
+.highlight {
+background-color: #FFFF00; /* Yellow highlight */
+padding: 2px 5px;
+border-radius: 5px;
+}
+#verse-info {
+margin-bottom: 20px;
+text-align: center;
+}
+#verses {
+display: flex;
+flex-direction: row;
+justify-content: center;
+align-items: flex-start;
+gap: 50px;
+}
+#hebrew-verse {
+font-size: 18px;
+line-height: 1.5;
+margin-bottom: 20px;
+text-align: right;
+direction: rtl;
+}
+#english-verse {
+font-size: 18px;
+line-height: 1.5;
+margin-bottom: 20px;
+}
 """) as iface:
     with gr.Row():
         timezone_input = gr.Dropdown(
-          choices=[tz for tz in pytz.common_timezones],
-          label="Select Your Timezone",
-          value="UTC"  # Set a default timezone
+            choices=[tz for tz in pytz.common_timezones],
+            label="Select Your Timezone",
+            value="UTC"  # Set a default timezone
         )
 
     with gr.Row():
@@ -183,9 +214,9 @@ with gr.Blocks(css="""
 
     # Update the display with verse information and translations when the button is clicked
     advance_button.click(
-        fn=lambda tz: update_tanach_display(datetime.datetime.now(pytz.timezone(tz)).strftime("%H:%M:%S"), tz),
-        inputs=[timezone_input],
-        outputs=[verse_info_output, hebrew_verse_output, english_verse_output],
+    fn=lambda tz: update_tanach_display(datetime.datetime.now(pytz.timezone(tz)).strftime("%H:%M:%S"), tz),
+    inputs=[timezone_input],
+    outputs=[verse_info_output, hebrew_verse_output, english_verse_output],
     )
 
 class TestWordIndex(unittest.TestCase):

translation_cache.db

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5785f0efbbb8d92d5fcd3f09ebebfaa18bf67d7cf76e6bf815a68960dca4d806
+size 1363968