pdf_ocr_vlm 1.2.3

HTTP OCR engine for dart_pdf_editor that adds a selectable, searchable text layer to scanned PDFs using a self-hosted VLM or any JSON OCR service.

Changelog

1.2.3

• Version bump to align with dart_pdf_editor 1.2.3. No API changes since 1.2.2.

1.2.2

• Version bump to align with dart_pdf_editor 1.2.2. No API changes since 1.2.1.

1.2.1

• Add a package example and shorten the pubspec description for pub.dev scoring.

1.2.0

• Version bump to align with dart_pdf_editor 1.2.0.

0.1.0

• Initial release. VlmOcrEngine implements dart_pdf_editor's PdfOcrEngine by POSTing each page raster to an HTTP OCR service and mapping the recognized boxes back into PDF user space.
• VlmOcrEngine.dotsOcr preset targets a vLLM server hosting rednote-hilab/dots.ocr over its OpenAI-compatible chat API, with no adapter required.
• A small default JSON contract ({image, width, height} → {spans: [...]}) plus requestBody/responseParser hooks for custom or cloud backends.