podcast-dlgen.py - Easily download podcasts with names from a RSS feed and jdownloader

podcast-dlgen.py

#!/usr/bin/env python3
# /// script
# dependencies = [
#     "requests>=2.31.0",
#     "lxml>=4.9.4",
#     "pyperclip>=1.8.2",
# ]
# ///
# Inspired by https://maxbarrass.com/blog/2020/04/10/downloading-files-from-rss-with-jdownloader-packegizer%F0%9F%92%BB%F0%9F%92%A5%F0%9F%98%8D
# archive link: https://archive.is/rP7Qg

#! jDownloader2 Packagizer Rule:
# Sourceulr(s): contains | #name=(.+)(?=\....)
# Package Name: <jd:orgfiletype>
# Filename: <jd:source:1>.<jd:orgfiletype>
#!

#* $ uv run python-dlgen.py -h                                                                                            
#* usage: python-dlgen.py [-h] url
#*                                                                                                                      
#* Process podcast RSS feed and generate download list.
#*
#* positional arguments:
#*   url         URL of the RSS feed
#*
#* options:
#*   -h, --help  show this help message and exit
  
import argparse
import sys

try:
    import pyperclip
except ImportError:
    pyperclip = None

import requests
from lxml import etree

XSLT_STR = '''\
<?xml version="1.0" encoding="utf-8"?>
<xsl:stylesheet version="1.0" 
    xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
    xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd">
  
  <xsl:output method="text" encoding="utf-8"/>
  
  <xsl:template match="/">
    <xsl:apply-templates select="//item"/>
  </xsl:template>

  <xsl:template match="item">
    <xsl:variable name="url" select="enclosure/@url"/>
    <xsl:variable name="episode" select="itunes:episode"/>
    <xsl:variable name="title" select="normalize-space(title)"/>
    <xsl:variable name="mime-type" select="enclosure/@type"/>

    <!-- Build filename -->
    <xsl:variable name="base-name">
      <xsl:choose>
        <xsl:when test="$episode != ''">
          <xsl:value-of select="concat($episode, ' - ', $title)"/>
        </xsl:when>
        <xsl:otherwise>
          <xsl:value-of select="$title"/>
        </xsl:otherwise>
      </xsl:choose>
    </xsl:variable>

    <!-- Determine extension -->
    <xsl:variable name="extension">
      <xsl:choose>
        <xsl:when test="contains($mime-type, 'mpeg')">mp3</xsl:when>
        <xsl:when test="contains($mime-type, 'm4a')">m4a</xsl:when>
        <xsl:when test="contains($mime-type, 'mp4')">mp4</xsl:when>
        <xsl:otherwise>
          <!-- Fallback: Check URL extension -->
          <xsl:variable name="url-lower" select="translate($url, 'ABCDEFGHIJKLMNOPQRSTUVWXYZ', 'abcdefghijklmnopqrstuvwxyz')"/>
          <xsl:choose>
            <xsl:when test="contains($url-lower, '.mp3')">mp3</xsl:when>
            <xsl:when test="contains($url-lower, '.m4a')">m4a</xsl:when>
            <xsl:when test="contains($url-lower, '.mp4')">mp4</xsl:when>
            <xsl:otherwise>mp3</xsl:otherwise>
          </xsl:choose>
        </xsl:otherwise>
      </xsl:choose>
    </xsl:variable>

    <xsl:value-of select="concat(normalize-space($url), '#name=', $base-name, '.', $extension)"/>
    <xsl:text>&#10;</xsl:text>
  </xsl:template>
</xsl:stylesheet>
'''

def main():
    parser = argparse.ArgumentParser(description='Process podcast RSS feed and generate download list.')
    parser.add_argument('url', help='URL of the RSS feed')
    args = parser.parse_args()

    try:
        response = requests.get(args.url)
        response.raise_for_status()
    except requests.exceptions.RequestException as e:
        print(f"Error downloading RSS feed: {e}", file=sys.stderr)
        sys.exit(1)

    try:
        # Parse XML and XSLT with consistent encoding
        xml_doc = etree.fromstring(response.content)
        xslt_doc = etree.fromstring(XSLT_STR.encode('utf-8'))
        transformer = etree.XSLT(xslt_doc)
        result = transformer(xml_doc)
        output = str(result).strip()
    except Exception as e:
        print(f"Processing Error: {e}", file=sys.stderr)
        sys.exit(1)

    print(output)
    
    if pyperclip is not None:
        try:
            pyperclip.copy(output)
            print("Output copied to clipboard.", file=sys.stderr)
        except pyperclip.PyperclipException as e:
            print(f"Clipboard Error: {e}", file=sys.stderr)

if __name__ == '__main__':
    main()