Upload video_to_audio.ipynb

docs/video_to_audio.ipynb

@@ -0,0 +1,139 @@
+{
+  "nbformat": 4,
+  "nbformat_minor": 0,
+  "metadata": {
+    "colab": {
+      "provenance": []
+    },
+    "kernelspec": {
+      "name": "python3",
+      "display_name": "Python 3"
+    },
+    "language_info": {
+      "name": "python"
+    }
+  },
+  "cells": [
+    {
+      "cell_type": "markdown",
+      "source": [
+        "# (YouTube) video to audio"
+      ],
+      "metadata": {
+        "id": "kNt1V_xZCYzb"
+      }
+    },
+    {
+      "cell_type": "markdown",
+      "source": [
+        "- Author: [Pierre Guillou](https://www.linkedin.com/in/pierreguillou/)\n",
+        "- Date: 08/10/2023"
+      ],
+      "metadata": {
+        "id": "Fa6V8oEynFe-"
+      }
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "%%capture\n",
+        "#!apt-get install -y ffmpeg\n",
+        "!python3 -m pip install -U yt-dlp\n",
+        "!pip install unidecode\n",
+        "!pip install gradio\n",
+        "!pip install pydub"
+      ],
+      "metadata": {
+        "id": "S4yB5r9RCdkH"
+      },
+      "execution_count": 1,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "import gradio as gr\n",
+        "import re, unidecode\n",
+        "from unidecode import unidecode\n",
+        "import yt_dlp\n",
+        "import os\n",
+        "import pydub\n",
+        "import numpy as np\n",
+        "\n",
+        "# no space, punctuation, accent in lower string\n",
+        "def cleanString(string):\n",
+        "    cleanString = unidecode(string)\n",
+        "    cleanString = re.sub('\\W+','_', cleanString)\n",
+        "    return cleanString.lower()\n",
+        "\n",
+        "# from audio file path to sample rate and numpy array\n",
+        "def read_audio(f, normalized=False):\n",
+        "    \"\"\"MP3 to numpy array\"\"\"\n",
+        "    a = pydub.AudioSegment.from_mp3(f)\n",
+        "    y = np.array(a.get_array_of_samples())\n",
+        "    if a.channels == 2:\n",
+        "        y = y.reshape((-1, 2))\n",
+        "    if normalized:\n",
+        "        return a.frame_rate, np.float32(y) / 2**15\n",
+        "    else:\n",
+        "        return a.frame_rate, y\n",
+        "\n",
+        "# from YouTube url to audio file path and sample rate + numpy array\n",
+        "def download_audio(url):\n",
+        "\n",
+        "    path_to_folder_audio_mp3 = \"./audio/\"\n",
+        "    ydl_opts = {\n",
+        "        'format': 'm4a/bestaudio/best',\n",
+        "        'outtmpl': f'{path_to_folder_audio_mp3}%(title)s',\n",
+        "        'postprocessors': [{\n",
+        "            'key': 'FFmpegExtractAudio',\n",
+        "            'preferredcodec': 'mp3',\n",
+        "        }]\n",
+        "    }\n",
+        "\n",
+        "    with yt_dlp.YoutubeDL(ydl_opts) as ydl:\n",
+        "        info_dict = ydl.extract_info(url, download=True)\n",
+        "        video_title = info_dict['title']\n",
+        "\n",
+        "        # Rename the audio file\n",
+        "        local_link = video_title + \".mp3\"\n",
+        "        new_local_link = cleanString(video_title) + \".mp3\"\n",
+        "        for filename in os.listdir(path_to_folder_audio_mp3):\n",
+        "            if cleanString(local_link) == cleanString(filename):\n",
+        "                os.rename(os.path.join(path_to_folder_audio_mp3, filename),os.path.join(path_to_folder_audio_mp3, new_local_link))\n",
+        "\n",
+        "        # get audio file path\n",
+        "        file_path = path_to_folder_audio_mp3 + new_local_link\n",
+        "\n",
+        "    return file_path, read_audio(file_path)\n",
+        "\n",
+        "# Gradio interface\n",
+        "iface = gr.Interface(fn=download_audio,\n",
+        "                     inputs=gr.Textbox(label=\"YouTube Video URL\"),\n",
+        "                     outputs=[\n",
+        "                         gr.File(label=\"Output Audio File\"),\n",
+        "                         gr.Audio(label=\"Play Audio\", show_download_button=False, format=\"mp3\"),\n",
+        "                     ],\n",
+        "                     allow_flagging=\"never\",\n",
+        "                     title=\"YouTube Video to Audio (mp3)\",\n",
+        "                     description=\"Just paste any YouTube video url and get its corresponding audio file in mp3.\",\n",
+        "                     )\n",
+        "iface.launch()"
+      ],
+      "metadata": {
+        "id": "9YvB5hBloP1f"
+      },
+      "execution_count": null,
+      "outputs": []
+    },
+    {
+      "cell_type": "markdown",
+      "source": [
+        "# END"
+      ],
+      "metadata": {
+        "id": "u9QYxqjtnzCD"
+      }
+    }
+  ]
+}