Imported from vcaldo/continuo (skills_scripts/playlist2sqlite/SKILL.md). Install upstream with npx skills add vcaldo/continuo --skill playlist2sqlite. Copyright stays with the author.

Playlist2SQLite

Extract and parse M3U/M3U8 playlist files into portable SQL dump files for querying and analysis.

Overview

M3U playlists (commonly used for IPTV, streaming services, and media collections) contain channel/stream metadata in a semi-structured text format. This skill provides a bash script to parse M3U files (from local files or URLs) and export as SQL dump files that can be imported into any SQLite database.

Features:

• Download large playlists reliably with retry logic
• Handle files with 100k+ entries
• Proper SQL escaping for all edge cases
• Progress indicators for large files
• Automatic cleanup on errors

Quick Start

# Parse from URL
scripts/playlist2sqlite.sh --url "https://provider.com/iptv.m3u" --name my-iptv

# Parse from local file
scripts/playlist2sqlite.sh --input playlist.m3u --name channels

# Append to existing SQL file
scripts/playlist2sqlite.sh --input new-channels.m3u --name channels --append

# Verbose mode with progress
scripts/playlist2sqlite.sh --url "https://example.com/list.m3u" --name streams --verbose

# For very large playlists (increase timeout)
scripts/playlist2sqlite.sh --url "https://large-provider.com/huge.m3u" --name huge --timeout 3600 --progress

# Import into SQLite database
sqlite3 channels.db < channels.sql

Output Format

The script outputs a .sql file (SQL dump) containing:

• Transaction wrapping for performance
• Table creation statements
• Index creation statements
• INSERT statements for each channel

This portable format can be imported into any SQLite database.

Database Schema

The SQL creates a channels table with the following structure:

M3U Format Reference

M3U files follow this structure:

#EXTM3U
#EXTINF:-1 tvg-id="channel1" tvg-name="Channel 1" tvg-logo="http://logo.png" group-title="News",Channel 1
http://stream.url/channel1
#EXTINF:-1 tvg-id="channel2" tvg-name="Channel 2" tvg-logo="http://logo2.png" group-title="Sports",Channel 2
http://stream.url/channel2

The script parses:

• #EXTM3U header (validates format)
• #EXTINF metadata lines (extracts tvg-*, group-title, channel name)
• Stream URLs (line immediately following EXTINF)

Script Usage

Located at: scripts/playlist2sqlite.sh

Options

--name NAME        Output file name, creates NAME.sql (required)
--input FILE       M3U/M3U8 file to parse (mutually exclusive with --url)
--url URL          M3U/M3U8 URL to download and parse (mutually exclusive with --input)
--append           Append to existing SQL file (default: recreate)
--verbose          Show detailed progress
--progress         Show download/parsing progress (auto-enabled for large files)
--timeout SECS     Max download time in seconds (default: 1800 = 30 minutes)
--retries N        Number of download retries (default: 3)
--help             Display usage information

Examples

From URL (most common):

./scripts/playlist2sqlite.sh --url "https://provider.com/iptv.m3u" --name my-iptv
# Output: my-iptv.sql

From local file:

./scripts/playlist2sqlite.sh --input iptv.m3u --name channels
# Output: channels.sql

Append new channels:

./scripts/playlist2sqlite.sh --input additional.m3u --name channels --append

Large playlist with extended timeout:

./scripts/playlist2sqlite.sh --url "https://example.com/large.m3u" --name streams \
    --timeout 3600 --verbose

Importing the SQL

After generating the SQL file, import it into a SQLite database:

# Create new database and import
sqlite3 channels.db < channels.sql

# Or import into existing database
sqlite3 existing.db < channels.sql

Querying the Database

After importing, query with standard SQLite:

# Count channels by group
sqlite3 channels.db "SELECT group_title, COUNT(*) FROM channels GROUP BY group_title;"

# Find all sports channels
sqlite3 channels.db "SELECT tvg_name, url FROM channels WHERE group_title LIKE '%Sport%';"

# Export specific group to CSV
sqlite3 -header -csv channels.db "SELECT * FROM channels WHERE group_title='Movies';" > movies.csv

# Find channels without logos
sqlite3 channels.db "SELECT tvg_name FROM channels WHERE tvg_logo IS NULL OR tvg_logo='';"

Large Playlist Handling

The script is optimized for very large playlists:

Automatic Optimizations

• Progress indicators: Automatically enabled for files >50MB
• Transaction wrapping: All inserts wrapped in transaction for faster imports
• Retry logic: Downloads automatically retry up to 3 times on failure
• Memory efficiency: Streaming parser (doesn't load entire file into memory)

Recommended Settings for Large Playlists

Example: 100k+ Entry Playlist

./scripts/playlist2sqlite.sh \
    --url "https://large-provider.com/mega.m3u" \
    --name mega-list \
    --timeout 7200 \
    --retries 5 \
    --verbose

Troubleshooting

Download Issues

"Download failed after X attempts"

• Check if URL is accessible: curl -I "URL"
• Increase timeout: --timeout 3600
• Increase retries: --retries 5
• Some providers rate-limit; wait and retry

"Downloaded file appears to be HTML"

• URL returned error page instead of M3U
• Check if URL requires authentication
• Verify URL is correct and not expired
• Try accessing URL in browser to see the actual error

Slow downloads

• Large playlists (>100MB) can take 10+ minutes on slow connections
• Use --progress to see download status
• Increase --timeout for slow connections

Parsing Issues

"File contains no #EXTINF entries"

• File is not a valid M3U playlist
• Check file content: head -20 file.m3u
• May be wrong format (e.g., plain text URL list)

"Entries skipped" in output

• Normal for M3U files with malformed entries
• Use --verbose to see which entries were skipped
• Usually caused by missing URL after EXTINF line

Special characters causing issues

• Script handles quotes, newlines, and unicode automatically
• If import fails, check for unusual binary content in M3U

Import Issues

SQLite "database is locked"

• Close other applications using the database
• For large imports, use: sqlite3 -cmd ".timeout 30000" db.db < file.sql

Import very slow

• The SQL file uses transactions, which should be fast
• Ensure database is on SSD, not spinning disk
• For very large files, import may take several minutes

Out of memory

• SQLite handles large imports well, but OS needs memory
• For 100k+ entries, ensure 1GB+ free RAM
• Consider splitting very large playlists

Common Use Cases

1. IPTV Management: Parse provider playlists from URLs, filter/search channels, export subsets
2. Playlist Analysis: Count channels per group, identify missing metadata, find duplicates
3. Playlist Merging: Combine multiple M3U files into single database, deduplicate
4. Channel Migration: Extract channels from old format, transform, export to new format
5. EPG Integration: Use tvg-id to link channels with Electronic Program Guide data
6. Backup: Store playlist data in portable SQL format
7. Intelligent Search: Find channels broadcasting specific programs/events

Channel Search

Find channels broadcasting specific programs or events by searching the web for broadcasting information and matching against your local channel database.

Quick Start

# Find channels showing Barcelona match
./scripts/search-channel.sh --db channels.db --query "jogo do Barça"

# Search for Champions League in Brazil
./scripts/search-channel.sh --db channels.db --query "Champions League" --country BR

# JSON output for integration
./scripts/search-channel.sh --db channels.db --query "NBA Finals" --json

How It Works

1. Web Search: Searches DuckDuckGo for "{query} onde assistir canais transmissão"
2. Channel Extraction: Extracts channel names from search results (ESPN, TNT, Globo, etc.)
3. Fuzzy Matching: Matches found channels against your SQLite database
4. Ranking: Scores results by match quality, sports relevance, and channel popularity
5. Output: Returns top matches with channel name, category, and stream URL

Script Usage

Located at: scripts/search-channel.sh

--db FILE          SQLite database file (required)
--query "TEXT"     Event/program to search for (required)
--country CODE     Country filter: BR, US, ES, PT, etc.
--lang LANG        Search language: pt, es, en (default: pt)
--max N            Maximum results to show (default: 10)
--json             Output results as JSON
--verbose          Show detailed progress
--no-cache         Skip cache for web search
--help             Display help

Examples

Sports Event (Portuguese):

./scripts/search-channel.sh --db channels.db --query "jogo do Flamengo"

# Output:
# Found 5 possible channels for "jogo do Flamengo":
#
# 1. [⭐⭐⭐] SporTV HD
#    Category: Sports BR
#    URL: http://stream.url/sportv
#
# 2. [⭐⭐⭐] Premiere FC
#    Category: Sports BR
#    URL: http://stream.url/premiere

Soccer Match (Spanish):

./scripts/search-channel.sh --db channels.db --query "Real Madrid vs Barcelona" --lang es

# Searches for: "Real Madrid vs Barcelona donde ver canal transmisión en vivo"

Movies:

./scripts/search-channel.sh --db channels.db --query "filme Gladiador" --lang pt

JSON Output (for integration):

./scripts/search-channel.sh --db channels.db --query "UFC 300" --json

# Output:
# {
#   "query": "UFC 300",
#   "count": 3,
#   "results": [
#     {"name": "ESPN Brasil HD", "category": "Sports", "url": "...", "score": 180},
#     {"name": "Combate", "category": "Sports", "url": "...", "score": 150}
#   ]
# }

Scoring Logic

Channels are ranked by score based on:

Event Type Detection:

• Sports keywords → prioritize sports channels
• Movie keywords → prioritize entertainment channels
• News keywords → prioritize news channels

Caching

Web search results are cached for 1 hour in /tmp/playlist2sqlite-cache/ to reduce API calls. Use --no-cache to force fresh search.

Supported Languages

Known Channel Patterns

The script recognizes popular broadcasting channels:

Sports: ESPN, Fox Sports, TNT Sports, SporTV, Premiere, BandSports, beIN Sports, Sky Sports, DAZN, Movistar, Star+

Entertainment: HBO, HBO Max, Netflix, Prime Video, Disney+, Paramount+

News: GloboNews, CNN, CNN Brasil, BBC

Notes

• Script requires bash, sqlite3, curl (for URL mode), and standard Unix tools (sed, awk, grep)
• Handles both M3U and M3U8 formats (M3U8 is UTF-8 encoded M3U)
• Non-standard EXTINF attributes are preserved in raw_extinf column
• Empty/invalid URLs are skipped with warning in verbose mode
• URL downloads use curl with automatic retry and temp file cleanup
• Output is .sql file (SQL dump), not .db (SQLite binary)
• SQL is wrapped in transaction for faster imports