mirror of
https://github.com/Aidenwu0209/PaddleOCR-Skills.git
synced 2026-09-14 20:06:58 +08:00
Aidenwu0209
841df44a3f
Upstream PaddleOCR refactored the skills in PR #18090 (2026-06-03) to use the official `paddleocr api` CLI directly, removing scripts/ and references/. This mirror keeps the script-based version as the default path and adds the CLI as a documented alternative. Changes: - skills/paddleocr-text-recognition/SKILL.md: add "Alternative: paddleocr CLI" section with commands, output format, and Scripts-vs-CLI comparison table; update frontmatter compatibility note - skills/paddleocr-doc-parsing/SKILL.md: same, with --model_type doc_parsing, --page_ranges, --save_resources, --prettify_markdown - README.md / README_cn.md: add "Two Ways to Run the Skills" section, update synced-source line (checked against upstream main HEAD 211989f0, 2026-06-26), note #18090 refactor, and clarify CLI needs only PADDLEOCR_ACCESS_TOKEN - docs/QUICK_REFERENCE.md: add "CLI 方式（paddleocr api）" section + diff table - docs/LARGE_FILES.md: add banner noting CLI --page_ranges replaces split_pdf.py - docs/version3.x/deployment/skills.{en,}.md: add archived-mirror banner pointing to new integrations/ location upstream No scripts, references, or frontmatter metadata.openclaw fields changed. Scripts verified via `uv run scripts/smoke_test.py --skip-api-test`.