Gobbler Batch Directory Conversion

Convert entire directories of audio files or documents to markdown in a single command. Handles concurrent processing, progress tracking, and graceful error recovery.

Quick Start

Batch convert audio recordings:

bash

gobbler batch directory ./recordings -o ./output --pattern "*.mp3"

Batch convert PDF documents:

bash

gobbler batch directory ./docs -o ./output --pattern "*.pdf" --type document

Convert everything (auto-detect types):

bash

gobbler batch directory ./mixed -o ./output --pattern "*.*"

Pattern Matching Guide

Glob patterns control which files get processed:

Pattern	Matches
`*.mp3`	All MP3 files
`*.pdf`	All PDF files
`recording_*.wav`	Files starting with "recording_"
`*.{mp3,wav}`	MP3 and WAV files
`.`	All files (relies on auto-detection)

Patterns are case-sensitive. Use *.* with caution - it processes everything the tool recognizes.

File Type Detection

When --type is not specified, gobbler auto-detects based on extension:

Audio formats: mp3, wav, m4a, flac, ogg, wma, aac

Document formats: pdf, docx, doc, pptx, ppt, xlsx, xls

Unrecognized extensions are skipped with a warning.

Options Reference

Flag	Description	Default
`-o, --output`	Output directory for markdown files	Required
`--pattern, -p`	Glob pattern for file matching	`.`
`--type, -t`	Force file type: `audio` or `document`	Auto-detect
`--concurrency, -c`	Parallel processing limit	`3`

Batch Processing Mechanics

Concurrency: Files process in parallel up to the concurrency limit. Higher values speed up processing but use more memory.

Auto-queue: When batch contains >10 items, processing moves to background with Redis. You'll get a job ID to check status.

Progress tracking: Real-time progress bar shows completion percentage and current file.

Output structure:

code

output_dir/
  meeting_notes.md      # from meeting_notes.mp3
  quarterly_report.md   # from quarterly_report.pdf
  interview.md          # from interview.wav

Output filenames match input filenames with .md extension.

Saving Output

Tip: If the user has configured output.default_directory in ~/.config/gobbler/config.yml, save files there.

bash

gobbler batch directory ./recordings -o "$OUTPUT_DIR" --pattern "*.mp3"

Error Handling

Batch processing continues even when individual files fail:

•Failed files are logged but don't stop the batch
•Summary report shows: X successful, Y failed
•Failed file list included in output for retry

Example summary:

code

Processed 47 files successfully
3 files failed to process

Performance Tips

Concurrency tuning:

•Default (3): Safe for most systems
•Increase to 5-8: Faster on good hardware with available memory
•Reduce to 1-2: For large files or limited memory

bash

# Faster processing
gobbler batch directory ./files -o ./out --pattern "*.wav" --concurrency 5

# Conservative for large files
gobbler batch directory ./videos -o ./out --pattern "*.m4a" --concurrency 1

Memory considerations:

•Large audio files (>100MB) benefit from lower concurrency
•Document batches are generally lighter on memory
•Monitor system resources on first large batch

Troubleshooting

Pattern not matching any files

•Check case sensitivity (*.MP3 != *.mp3)
•Verify files exist in the source directory
•Test pattern with ls ./source/*.mp3 first

Mixed file types producing errors

•Use specific patterns instead of *.*
•Or explicitly set --type when directory contains one type

Memory issues with large batches

•Reduce concurrency: --concurrency 1
•Process in smaller batches by using more specific patterns
•Large audio files (podcasts, long recordings) need lower concurrency

Processing seems stuck

•Large files take time - check progress bar
•

10 files triggers background queue; check job status with gobbler jobs list
•Retry the batch if network issues caused API timeouts