Update steganography.py

steganography.py

@@ -13,21 +13,27 @@ DEFAULT_SAMPLE_RATE = 22050
 
 logging.basicConfig(level=logging.INFO)
 
-def text_to_spectrogram_image(text, base_width=512, height=256, max_font_size=80, margin=10, letter_spacing=5):
+def load_font(font_path, max_font_size):
     try:
-        font = ImageFont.truetype(DEFAULT_FONT_PATH, max_font_size)
+        return ImageFont.truetype(font_path, max_font_size)
     except IOError:
-        logging.warning(f"Font not found at {DEFAULT_FONT_PATH}. Using default font.")
-        font = ImageFont.load_default()
+        logging.warning(f"Font not found at {font_path}. Using default font.")
+        return ImageFont.load_default()
     except Exception as e:
         logging.error(f"An error occurred while loading the font: {e}")
         raise
 
+def create_text_image(text, font, base_width=512, height=256, margin=10, letter_spacing=5):
     draw = ImageDraw.Draw(Image.new("L", (1, 1)))
-
-    text_widths = [draw.textsize(char, font=font)[0] for char in text]
+    text_widths = [
+        draw.textbbox((0, 0), char, font=font)[2] - draw.textbbox((0, 0), char, font=font)[0]
+        for char in text
+    ]
     text_width = sum(text_widths) + letter_spacing * (len(text) - 1)
-    text_height = draw.textsize(text[0], font=font)[1]
+    text_height = (
+        draw.textbbox((0, 0), text[0], font=font)[3]
+        - draw.textbbox((0, 0), text[0], font=font)[1]
+    )
 
     width = max(base_width, text_width + margin * 2)
     height = max(height, text_height + margin * 2)
@@ -43,9 +49,7 @@ def text_to_spectrogram_image(text, base_width=512, height=256, max_font_size=80
         draw.text((current_x, text_start_y), char, font=font, fill="white")
         current_x += char_width + letter_spacing
 
-    image = np.array(image)
-    image = np.where(image > 0, 255, image)
-    return image
+    return np.array(image)
 
 def spectrogram_image_to_audio(image, sr=DEFAULT_SAMPLE_RATE):
     flipped_image = np.flipud(image)
@@ -54,7 +58,8 @@ def spectrogram_image_to_audio(image, sr=DEFAULT_SAMPLE_RATE):
     return y
 
 def create_audio_with_spectrogram(text, base_width, height, max_font_size, margin, letter_spacing):
-    spec_image = text_to_spectrogram_image(text, base_width, height, max_font_size, margin, letter_spacing)
+    font = load_font(DEFAULT_FONT_PATH, max_font_size)
+    spec_image = create_text_image(text, font, base_width, height, margin, letter_spacing)
     y = spectrogram_image_to_audio(spec_image)
 
     with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as temp_audio:
@@ -111,45 +116,50 @@ def gradio_image_to_audio_fn(upload_image):
 def gradio_decode_fn(upload_audio):
     return display_audio_spectrogram(upload_audio)
 
-with gr.Blocks(title="Audio Steganography", css="footer{display:none !important}", theme=gr.themes.Soft(primary_hue="green", secondary_hue="green", spacing_size="sm", radius_size="lg")) as txt2spec:
-    with gr.Tab("Text to Spectrogram"):
-        with gr.Group():
-            text = gr.Textbox(lines=2, placeholder="Enter your text:", label="Text", info="Enter the text you want to convert to audio.")
-            with gr.Row(variant="panel"):
-                base_width = gr.Slider(value=512, label="Image Width", visible=False)
-                height = gr.Slider(value=256, label="Image Height", visible=False)
-                max_font_size = gr.Slider(minimum=10, maximum=130, step=5, value=80, label="Font size")
-                margin = gr.Slider(minimum=0, maximum=50, step=1, value=10, label="Indent")
-                letter_spacing = gr.Slider(minimum=0, maximum=50, step=1, value=5, label="Letter spacing")
-            generate_button = gr.Button("Generate", variant="primary", size="lg")
-
-        with gr.Column(variant="panel"):
+def create_gradio_interface():
+    with gr.Blocks(title="Audio Steganography", css="footer{display:none !important}", theme=gr.themes.Soft(primary_hue="green", secondary_hue="green", spacing_size="sm", radius_size="lg")) as txt2spec:
+        with gr.Tab("Text to Spectrogram"):
             with gr.Group():
-                output_audio = gr.Audio(type="filepath", label="Generated audio")
-                output_spectrogram = gr.Image(type="filepath", label="Spectrogram")
-
-        generate_button.click(gradio_interface_fn, inputs=[text, base_width, height, max_font_size, margin, letter_spacing], outputs=[output_audio, output_spectrogram])
+                text = gr.Textbox(lines=2, placeholder="Enter your text:", label="Text", info="Enter the text you want to convert to audio.")
+                with gr.Row(variant="panel"):
+                    base_width = gr.Slider(value=512, label="Image Width", visible=False)
+                    height = gr.Slider(value=256, label="Image Height", visible=False)
+                    max_font_size = gr.Slider(minimum=10, maximum=130, step=5, value=80, label="Font size")
+                    margin = gr.Slider(minimum=0, maximum=50, step=1, value=10, label="Indent")
+                    letter_spacing = gr.Slider(minimum=0, maximum=50, step=1, value=5, label="Letter spacing")
+                generate_button = gr.Button("Generate", variant="primary", size="lg")
+
+            with gr.Column(variant="panel"):
+                with gr.Group():
+                    output_audio = gr.Audio(type="filepath", label="Generated audio")
+                    output_spectrogram = gr.Image(type="filepath", label="Spectrogram")
+
+            generate_button.click(gradio_interface_fn, inputs=[text, base_width, height, max_font_size, margin, letter_spacing], outputs=[output_audio, output_spectrogram])
+
+        with gr.Tab("Image to Spectrogram"):
+            with gr.Group():
+                with gr.Column():
+                    upload_image = gr.Image(type="filepath", label="Upload image")
+                    convert_button = gr.Button("Convert to audio", variant="primary", size="lg")
 
-    with gr.Tab("Image to Spectrogram"):
-        with gr.Group():
-            with gr.Column():
-                upload_image = gr.Image(type="filepath", label="Upload image")
-                convert_button = gr.Button("Convert to audio", variant="primary", size="lg")
+            with gr.Column(variant="panel"):
+                output_audio_from_image = gr.Audio(type="filepath", label="Generated audio")
 
-        with gr.Column(variant="panel"):
-            output_audio_from_image = gr.Audio(type="filepath", label="Generated audio")
+            convert_button.click(gradio_image_to_audio_fn, inputs=[upload_image], outputs=[output_audio_from_image])
 
-        convert_button.click(gradio_image_to_audio_fn, inputs=[upload_image], outputs=[output_audio_from_image])
+        with gr.Tab("Audio Spectrogram"):
+            with gr.Group():
+                with gr.Column():
+                    upload_audio = gr.Audio(type="filepath", label="Upload audio", scale=3)
+                    decode_button = gr.Button("Show spectrogram", variant="primary", size="lg")
 
-    with gr.Tab("Audio Spectrogram"):
-        with gr.Group():
-            with gr.Column():
-                upload_audio = gr.Audio(type="filepath", label="Upload audio", scale=3)
-                decode_button = gr.Button("Show spectrogram", variant="primary", size="lg")
+            with gr.Column(variant="panel"):
+                decoded_image = gr.Image(type="filepath", label="Audio Spectrogram")
 
-        with gr.Column(variant="panel"):
-            decoded_image = gr.Image(type="filepath", label="Audio Spectrogram")
+            decode_button.click(gradio_decode_fn, inputs=[upload_audio], outputs=[decoded_image])
 
-        decode_button.click(gradio_decode_fn, inputs=[upload_audio], outputs=[decoded_image])
+    return txt2spec
 
-txt2spec.launch(share=True)
+if __name__ == "__main__":
+    txt2spec = create_gradio_interface()
+    txt2spec.launch(share=True)