Update modules/text.py

modules/text.py

@@ -0,0 +1,77 @@
+# modules/text.py
+
+# -*- coding: utf-8 -*-
+#
+# PROJECT:      CognitiveEDA v5.0 - The QuantumLeap Intelligence Platform
+#
+# DESCRIPTION:  Specialized module for basic text analysis, focused on generating
+#               a word cloud visualization from a text-heavy column.
+
+import base64
+import io
+import logging
+
+import pandas as pd
+from wordcloud import WordCloud
+
+def generate_word_cloud(df: pd.DataFrame, text_col: str) -> str:
+    """
+    Generates a word cloud from a text column and returns it as an HTML object.
+
+    The function processes the text, generates a word cloud image, encodes it
+    in Base64, and embeds it within an HTML string for display in Gradio.
+
+    Args:
+        df: The input DataFrame.
+        text_col: The name of the column containing text data.
+
+    Returns:
+        An HTML string containing the word cloud image or a status/error message.
+    """
+    # 1. Input Validation
+    if not text_col:
+        return "<p style='text-align:center; padding: 20px;'>Select a text column to generate a word cloud.</p>"
+    
+    if text_col not in df.columns:
+        return f"<p style='text-align:center; color:red; padding: 20px;'>Error: Column '{text_col}' not found in the dataset.</p>"
+
+    try:
+        logging.info(f"Generating word cloud for column '{text_col}'")
+        
+        # 2. Text Corpus Preparation
+        # Concatenate all non-null text entries into a single string
+        text_corpus = ' '.join(df[text_col].dropna().astype(str))
+
+        if not text_corpus.strip():
+            logging.warning(f"Column '{text_col}' contains no text data to generate a cloud.")
+            return "<p style='text-align:center; padding: 20px;'>No text data available in this column to generate a cloud.</p>"
+
+        # 3. Word Cloud Generation
+        wordcloud = WordCloud(
+            width=800,
+            height=400,
+            background_color='white',
+            colormap='viridis',
+            max_words=150,
+            collocations=False # Avoids generating two-word phrases
+        ).generate(text_corpus)
+
+        # 4. Image Encoding
+        buf = io.BytesIO()
+        wordcloud.to_image().save(buf, format='png')
+        img_str = base64.b64encode(buf.getvalue()).decode('utf-8')
+        
+        # 5. HTML Output
+        # The style attribute makes the image responsive to container width
+        html_content = (
+            f'<div style="text-align:center; padding: 10px;">'
+            f'<img src="data:image/png;base64,{img_str}" alt="Word Cloud for {text_col}" '
+            f'style="border-radius: 8px; max-width: 100%; height: auto;">'
+            f'</div>'
+        )
+        return html_content
+
+    except Exception as e:
+        logging.error(f"Word cloud generation failed for column '{text_col}': {e}", exc_info=True)
+        error_msg = f"Could not generate word cloud. An unexpected error occurred: {e}"
+        return f"<p style='text-align:center; color:red; padding: 20px;'>❌ **Error:** {error_msg}</p>"