{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "d62a02e5",
   "metadata": {},
   "outputs": [],
   "source": [
    "from bs4 import BeautifulSoup\n",
    "import feedparser\n",
    "import re\n",
    "import tqdm\n",
    "import os\n",
    "import pandas as pd\n",
    "import sqlite3\n",
    "import time\n",
    "import random\n",
    "import requests\n",
    "import multiprocessing\n",
    "import numpy as np\n",
    "import json\n",
    "import urllib\n",
    "import html\n",
    "from collections import defaultdict\n",
    "\n",
    "from suno_utils.audio import Audio\n",
    "from suno_utils.utils.text import normalize_whitespace\n",
    "from suno_utils.utils.podcasts import load_podcast_db\n",
    "\n",
    "\n",
    "RAW_RSS_DIR = \"/mnt/data-ssd-1/data/podcasts/bulk_rss_feeds/raw_feeds/\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9dc2c520",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a8cc4ac3",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "ec79adc1",
   "metadata": {},
   "source": [
    "## Basic set up"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "419f68e3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1203 french-canadian anchor podcasts\n"
     ]
    }
   ],
   "source": [
    "main_df = load_podcast_db(\n",
    "    \"/mnt/data-ssd-1/data/podcasts/meta/podcastindex_feeds.db\", english_only=False, anchor_only=True\n",
    ")\n",
    "main_df = main_df[main_df[\"language\"] == \"fr-ca\"].reset_index(drop=True)\n",
    "frecan_podcast_ids = set(main_df[\"uid\"])\n",
    "frecan_rss_filenames = [fn for fn in os.listdir(RAW_RSS_DIR) if fn.split(\".\")[0] in frecan_podcast_ids]\n",
    "print(len(frecan_rss_filenames), \"french-canadian anchor podcasts\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "76206802",
   "metadata": {},
   "outputs": [],
   "source": [
    "Audio.from_file(\"8714e1c1-9e94-a00c-7871-eecdfcd33156.mp3\").get_slice(to_s=10*60).to_mp3(\"sample_1.mp3\")\n",
    "Audio.from_file(\"c165de9fb1e96e296e33870d3ea9c3a9.mp4\").get_slice(to_s=10*60).to_mp3(\"sample_2.mp3\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ee700004",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fa622df4",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c0c041dd",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a2f51920",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8c4da379",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}