My passion for improving developer experience led me to show you how to use an AI assistant for generating a Bash script that can download protected video streams.

I like to keep things simple, and I especially don’t want to install a dedicated app for every single task. As a developer, I like to have control over my tools and automate repetitive tasks. Especially I want to reduce my dependencies and keep my system clean and secure.

Here, I want to show you how to use AI (in this case, ClaudeAI) to generate a Bash script that can download videos that are protected by obfuscating the video stream using a m3u81 playlist. Such a stream can’t be downloaded directly, but you can use a tool like ffmpeg2.

It took me approximately 10 minutes to generate and test the video downloader script using AI assistants and apply it to a sample video stream. I opened ClaudeAI and asked the following question:

Please write a Bash script that downloads a video given a link that points to a m3u8 file. The m3u8 file may contain links to video files or parts or even to other m3u8 files that contain similar information. The output should be one file that is the result of a video stream. Please make use of the ffmpeg library if possible.

The generated script worked out of the box for a m3u8 file that contained links to video stream parts. The script was a bit verbose by printing each stream part to the console, so I asked ClaudeAI to show the progress instead. I switched to ChatGPT because my free ClaudeAI credits were used up.

I want to change the console/terminal output of the following video stream downloader the way that the progress is shown with a overall percentage number. Ideally, the overall content length is known apriori, otherwise a progress indicator should be shown by using a spinning ASCII character in a fixed position.

(script omitted)

The modified script worked as expected, until the second m3u8 file was encountered. A proper error message was shown, and I asked ChatGPT to handle this case as well.

I think getting the overall content length works, the percentage is shown. However, I encountered the following error when the next part was loaded, and the script ended:

line 95: (00 * 3600) + (02 * 60) + 08: value too great for base (error token is “08”)

ChatGPT fixed the issue by prefixing each time component inside the arithmetic evaluation with 10# to ensure that Bash treats them as base-10 numbers, effectively removing any leading zero interpretation as octal numbers.

Finally, the script worked as expected and had a nice progress indicator. I was able to download video streams and watch them offline without any additional software (except ffmpeg). I took the time to refine the script by using Shellcheck and manually adding some enhancements. A further improvement could be to obtain the m3u8 file automatically by using another script that extracts the m3u8 file URL from the DOM of a media library URL.

The complete script m3u8-downloader.sh is shown below3.

Disclaimer: The script is generated by AI and may contain errors. Use it at your own risk. The script is intended solely for downloading video streams that are freely available and not protected by copyright. Users are responsible for ensuring compliance with applicable copyright laws and must obtain proper authorization or rights before using this code to download any copyrighted material.

#!/bin/bash

# -----------------------------------------------------------------------------
# The MIT License (MIT)
#
# Copyright (c) 2024 Daniel Dietrich
#
# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
# copies of the Software, and to permit persons to whom the Software is
# furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice shall be included in
# all copies or substantial portions of the Software.
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
# THE SOFTWARE.
# -----------------------------------------------------------------------------

# Function to display usage instructions
usage() {
    echo "Usage: $0 [-y] <m3u8_url> <output_filename>"
    echo "  -y   Overwrite existing output files without prompt."
    echo "Example: $0 \"https://example.com/stream.m3u8\" \"output.mp4\""
    exit 1
}

# Function to check if required commands are available
check_requirements() {
    local missing_requirements=0
    if ! command -v ffmpeg >/dev/null 2>&1; then
        echo "Error: ffmpeg is not installed. Please install it first."
        missing_requirements=1
    fi
    if ! command -v ffprobe >/dev/null 2>&1; then
        echo "Error: ffprobe is not installed. Please install it first."
        missing_requirements=1
    fi
    if ! command -v curl >/dev/null 2>&1; then
        echo "Error: curl is not installed. Please install it first."
        missing_requirements=1
    fi
    if [ $missing_requirements -eq 1 ]; then
        exit 1
    fi
}

# Function to validate URL
validate_url() {
    local url=$1
    if ! curl --output /dev/null --silent --head --fail "$url"; then
        echo "Error: Invalid URL or resource not accessible: $url"
        exit 1
    fi
}

# Function to determine total content length using ffprobe
get_content_length() {
    local url=$1
    local duration
    duration=$(ffprobe -i "$url" -show_entries format=duration -v quiet -of csv="p=0")
    echo "$duration"
}

# Function to get format name and long name using ffprobe
get_format() {
    local url=$1
    local format_info
    format_info=$(ffprobe -i "$url" -show_entries format=format_name,format_long_name -v quiet -of csv="p=0")

    # Transform the output into "<name> (<long_name>)"
    echo "$format_info" | awk -F, '{print $1 " (" $2 ")"}'
}

# Function to display spinner
spinner() {
    local pid=$1
    local delay=0.1
    local spinstr="|/-\\"
    while ps a | awk '{print $1}' | grep -q "$pid"; do
        local temp=${spinstr#?}
        printf " [%c] " "$spinstr"
        spinstr=$temp${spinstr%"$temp"}
        sleep $delay
        printf "\b\b\b\b\b\b"
    done
}

# Function to download and process the m3u8 stream
download_stream() {
    local url=$1
    local output=$2
    echo "Starting download of: $url"
    echo "Format: $(get_format "$url")"
    echo "Output will be saved as: $output"

    # Get the total duration of the stream
    local total_duration
    total_duration=$(get_content_length "$url")
    total_duration=${total_duration%.*}  # Remove any decimal part
    if [[ ! "$total_duration" =~ ^[0-9]+$ ]]; then
        echo "Content length not determinable."
        ffmpeg -i "$url" \
            -c copy \
            -bsf:a aac_adtstoasc \
            -movflags +faststart \
            -y \
            "$output" > /dev/null 2>&1 &
        spinner $!
    else
        # Convert total duration to HH:MM:SS format
        printf -v formatted_duration '%02d:%02d:%02d' "$(echo "$total_duration/3600" | bc)" "$(echo "$total_duration%3600/60" | bc)" "$(echo "$total_duration%60" | bc)"
        echo "Content length determined: $formatted_duration"
        ffmpeg -i "$url" \
            -c copy \
            -bsf:a aac_adtstoasc \
            -movflags +faststart \
            -y \
            "$output" 2>&1 | while read -r line; do
            if [[ "$line" =~ time=([0-9]+):([0-9]+):([0-9]+) ]]; then
                hours=${BASH_REMATCH[1]}
                minutes=${BASH_REMATCH[2]}
                seconds=${BASH_REMATCH[3]}
                # Remove leading zeros by using arithmetic evaluation
                current_duration=$((10#$hours * 3600 + 10#$minutes * 60 + 10#$seconds))
                if [ "$total_duration" -ne 0 ]; then
                    progress=$(awk "BEGIN {printf \"%.2f\", ($current_duration / $total_duration) * 100}")
                    echo -ne "Progress: $progress% \r"
                fi
            fi
        done
    fi

    if [ "${PIPESTATUS[0]}" -eq 0 ]; then
        echo "Download completed successfully!"
        echo "File saved as: $output"
    else
        echo "Error: Download failed!"
        rm -f "$output"
        exit 1
    fi
}

# Main script execution starts here

# Check if correct number of arguments provided
if [ $# -lt 2 ] || [ $# -gt 3 ]; then
    usage
fi

# Initialize variables
FORCE_OVERWRITE=0

# Parse command line options
while getopts ":y" opt; do
    case ${opt} in
        y )
            FORCE_OVERWRITE=1
            ;;
        \? )
            usage
            ;;
    esac
done
shift $((OPTIND -1))

# Get command line arguments
M3U8_URL=$1
OUTPUT_FILE=$2

# Check if output file already exists
if [ "$FORCE_OVERWRITE" -eq 0 ] && [ -f "$OUTPUT_FILE" ]; then
    read -p "File $OUTPUT_FILE already exists. Overwrite? (y/n) " -n 1 -r
    echo
    if [[ ! $REPLY =~ ^[Yy]$ ]]; then
        echo "Operation cancelled."
        exit 1
    fi
fi

# Check for required commands
check_requirements

# Validate URL
validate_url "$M3U8_URL"

# Start download
download_stream "$M3U8_URL" "$OUTPUT_FILE"

In fact, writing this script was just a kata exercise for me. By continuously using AI to generate and improve code, I learn a lot about my own development flow and how to improve it. Many ideas emerge from these exercises. Let’s see where this journey will take me next.

I hope this post was helpful and you enjoyed reading it. If you have any questions or feedback, please follow the links below to discuss them with me.

Thanks for reading and have a great time!
Daniel

P.S. My next journal post might be about the beautiful country of Portugal; I recently watched a German documentary4 about it on the ZDF media library.

  1. m3u8 is a multimedia playlist file format that is used to serve audio and video content over the internet. The m3u8 URL can be obtained by playing the video in the web browser, opening the developer tools by pressing ⌥ ⌘ I on macOS, selecting Network and filtering by m3u8. ↵

  2. ffmpeg is a free and open-source project consisting of a large software suite of libraries and programs for handling video, audio, and other multimedia files and streams. It can be installed on macOS by using Homebrew brew install ffmpeg or on Ubuntu by using sudo apt install ffmpeg. ↵

  3. The script might need to be made executable by running chmod +x m3u8-downloader.sh before it can be executed. ↵

  4. The video about Portugal will be available on the ZDF media library until 15.12.2025. ↵