Add text, image and shape overlays to PDF pages

[McanKul]

Goal

Ship the first useful Edit PDF workflow: add new content without rasterizing the original document.

Depends on

• Reusable visual PDF editor canvas issue.

Scope

• Add text boxes with position, size, color and basic alignment.
• Add PNG/JPEG images with move, resize and aspect-ratio lock.
• Add rectangle, line and freehand drawing objects with color, width and opacity.
• Apply edits to one or more chosen pages.
• Generate vector/text/image overlay PDFs locally and merge them with qpdf.
• Embed a Unicode-capable bundled font and document its license.
• Show a faithful preview before saving.

Out of scope

• Reflowing or replacing existing PDF text.
• Editing embedded source images.
• Digital certificate signatures.

Acceptance criteria

• Existing text and vectors remain selectable and are not rasterized.
• Turkish characters render correctly in saved output.
• Placement matches preview on portrait, landscape, cropped and rotated pages.
• Transparency works for supported objects.
• Original files are never overwritten.
• Output opens without repair warnings in qpdf check and at least two common PDF viewers.
• Rust unit/integration tests cover escaping, page geometry and malformed input; frontend tests cover edit serialization.

Security/performance

Treat PDFs and images as untrusted input, enforce size limits, use bounded decoding, and avoid shell interpolation.