Tom Bertalan
Home > QRembedProject Files

QRembed

Usage

Embed a file or directory in QR/DataMatrix images optionally using 7z volumes.

python bin/qrembed.py [-h] [-o OUTPUT] [--method {qr,datamatrix}] [--chunked] [--chunk-size CHUNK_SIZE] [--retain-zips] input_path
  • input_path: Path to the file or directory to embed
  • -o OUTPUT, --output OUTPUT: Output image file (default: INPUT.qr.png)
  • --method {qr,datamatrix}: Encoding method: qr or datamatrix (default: qr)
  • --chunked: Enable chunked archive mode (split file/dir into 7z volumes for multiple codes)
  • --chunk-size CHUNK_SIZE: Chunk size in bytes (default: 2900 for QR, required for --chunked)
  • --retain-zips: Move 7z volume files to the target parent directory after encoding

Note:

  • Uses the largest QR code size and lowest error correction to maximize data capacity.
  • If the file is too large to fit in a QR code, an error will be raised unless chunked mode is used.

Be warned--this doen't scale super well: a 10kB file took 4 (dense) QR codes, and I expect this is linear. But if your only data storage medium is physical QR codes, I guess this could be useful?

Decoding

What seemed to work for me as to use Binary Eye to download the QR codes directly to files, saving as FILENAME.7z.001 etc. (note the two periods). Then I could use 7Zipper to extract by starting with the 001th.

Progress

  • CLI can encode any file or directory as a series of QR or DataMatrix codes, using 7z volumes for multipart support.
  • Default QR chunk size is set to 2900 bytes (empirically determined).
  • Output PNGs and (optionally) 7z volumes are moved to the target's parent directory.
  • Manual transfer and reassembly is possible, but tedious for large files.
  • No standard for including filenames in QR payloads; most phone QR readers do not extract or display custom headers.

Possible Next Steps

  • Write a Python script to decode a folder of QR images and reassemble the original file/volume.
  • Explore or build a mobile app for batch QR scanning and automatic reassembly.
  • Add support for other 2D codes (e.g., Aztec) or optimize for text files.
  • Improve error handling and user feedback for failed QR encodings.
  • Add optional metadata (e.g., chunk number, total chunks) in a way that is easy to parse by custom tools.
  • Document the decoding/reassembly process for users.
  • Consider designing a minimal mobile "app" (e.g., a Python script, PWA, or small APK) that can be distributed via QR codes (ideally <10 QRs for the installer itself).
  • This would enable fully offline, low-tech distribution of the decoding tool.