v0.0.5

Added:

• ocrd-nmalign-merge: allow 1:n plaintext (per-textline) files on right side
• tests and CI for the OCR-D processor

Changed:

• ocrd-nmalign-merge: skip empty lines on right side
• move setup.py → pyproject.toml
• adapt to OCR-D v3 API
• Docker image spec (labels, envvars, ocrd-all-tool.json) conforming to OCR-D