m3u8

Warnings

• breaking In version 6.0.0, the 'BITRATE' attribute within `StreamInfo` (e.g., `playlist.playlists[0].stream_info.bandwidth['BITRATE']`) is now parsed as an integer type instead of a string. Code expecting a string for bitrate values will need to be updated.
  Fix: Ensure your code handles `BITRATE` values as integers. If string representation is required, cast it explicitly: `str(playlist.playlists[0].stream_info.bandwidth['BITRATE'])`.
• breaking Starting from version 5.2.0, `target_duration` is consistently parsed as an integer. Previous versions might have returned it as a float or string, which could break code expecting those types.
  Fix: Adjust code to expect and handle `playlist.target_duration` as an integer. Cast if other types are explicitly needed: `float(playlist.target_duration)`.
• breaking Version 4.0.0 dropped compatibility with Python 2.x and updated to require Python 3.7 or newer. Additionally, it now leverages built-in ISO 8601 parsing for Python 3.11+.
  Fix: Upgrade your Python environment to 3.7 or higher. Review and update any Python 2 specific code if migrating an older project.
• gotcha When parsing playlists, especially those loaded from a URI, segment URIs can be relative. Accessing `segment.uri` directly might yield a relative path. Use `segment.absolute_uri` to get the full, resolved URL if absolute paths are required for downloading or further processing.
  Fix: Always prefer `segment.absolute_uri` when constructing URLs for segments or keys to ensure correct resolution against the playlist's `base_uri`.
• gotcha The library, by default, ignores non-standard M3U8 tags (i.e., tags not defined in the HLS specification). If you need to parse or extract information from custom tags (e.g., `#MOVIE-NAME:`), you must provide a custom parser function.
  Fix: Pass a `custom_tags_parser` function to `m3u8.load()` or `m3u8.loads()`. This function will be called for each line of the playlist, allowing you to process and store custom tag data.

Install

• pip install m3u8 Install latest version

Imports

• load

  import m3u8
  playlist = m3u8.load('http://example.com/playlist.m3u8')

  The primary functions `load` and `loads` are typically accessed directly from the imported `m3u8` module.
• loads

  import m3u8
  playlist = m3u8.loads('#EXTM3U8\n...')

  Used for parsing M3U8 content directly from a string.

Quickstart

This quickstart demonstrates how to load an M3U8 playlist from a string using `m3u8.loads()` and access its basic properties like target duration and individual segments. An commented-out section shows how to load directly from a URL using `m3u8.load()` for live HLS streams.

import m3u8
import requests
import os

# Example: Load a remote M3U8 playlist
# In a real scenario, use a specific, stable M3U8 URL.
# For a runnable example, we'll use a placeholder or local file.
# For demonstration, we'll simulate a playlist string.

# --- Option 1: Load from a URL (requires internet access) ---
# remote_m3u8_url = 'https://example.com/path/to/playlist.m3u8' # Replace with a real M3U8 URL if testing live
# try:
#     playlist = m3u8.load(remote_m3u8_url)
#     print(f"Loaded playlist from URL: {remote_m3u8_url}")
#     if playlist.segments:
#         print(f"First segment URI: {playlist.segments[0].uri}")
#     else:
#         print("No segments found in remote playlist.")
# except requests.exceptions.RequestException as e:
#     print(f"Error loading remote playlist: {e}")

# --- Option 2: Load from a string (always runnable) ---
local_m3u8_content = """
#EXTM3U
#EXT-X-VERSION:3
#EXT-X-TARGETDURATION:10
#EXT-X-MEDIA-SEQUENCE:0
#EXTINF:9.009,
http://example.com/segment0.ts
#EXTINF:10.000,
http://example.com/segment1.ts
#EXT-X-ENDLIST
"""

playlist_from_string = m3u8.loads(local_m3u8_content)
print("\n--- Playlist from String ---")
print(f"Target Duration: {playlist_from_string.target_duration}")
print(f"Number of segments: {len(playlist_from_string.segments)}")
if playlist_from_string.segments:
    print(f"First segment URI: {playlist_from_string.segments[0].uri}")

# Iterate through segments
print("\nSegments:")
for i, segment in enumerate(playlist_from_string.segments):
    print(f"  Segment {i}: Duration={segment.duration}, URI={segment.uri}")