[FEATURE] Get all texts/images/shapes/SVGs from PDF page

[Akxe]

First of all, this library is awesome. But I am missing one key aspect from this library to be able to use it. Also, I am not the only one. I will try to create some a draft of what I need, along while taking into consideration what others need too.

---

Get all of a type from a page:

To be able to modify a PDF also means to be able to change existing parts of it.

A PDFNode abstract class will be created. It would be the parent to PDFText, PDFImage, PDFSVG and `PDFShape.

class PDFNode {
  blendMode: BlendMode;
  opacity: number;
  rotate: Rotation;
  x: number;
  y: number;

  // Getter for its local width/height.
  get width(): number;
  get height(): number;
}

PDFPage.getAllTexts(): PDFTextNode[]

This is needed to be able to position one text base on another, Place signature (PNG) into the correct spot etc.

PDFText class is almost a copy of PDFPageDrawTextOptions interface.

class PDFTextNode extends PDFNode {
  color: Color;
  font: PDFFont;
  lineHeight: number;
  maxWidth?: number;
  size: number;
  wordBreaks: string[];
  xSkew: Rotation
  ySkew: Rotation;

  content: string;
}

The PDFPage.getAllTexts(): PDFText[] function will enable consumer, to find and replace text #564, Aligning new text to ones that are already in place. Or just find if the texts are in vector format, or in native text format.

PDFPage.getAllImages(): PDFImageNode[]

class PDFImageNode extends PDFNode {
  height: number;
  width: number;
  xSkew: Rotation
  ySkew: Rotation;
}

PDFPage.getAllShapes(): PDFShapeNode[]

The PDFShapeNode is an abstract parent class to PDFShapeLineNode, PDFShapeElipseNode, PDFShapeRectangleNode (and PDFShapePathNode?).

class PDFShapeNode extends PDFNode {
  color: Color;
}

TODO: Copy from PDFPageDrawXxxOptions.

Square is a special case of PDFShapeRectangle, same goes for circle.

PDFPage.getAllSVGs(): PDFSVGNode[]

class PDFSVGNode extends PDFNode {
  borderColor: Color
  borderDashArray: number[];
  borderDashPhase: number;
  borderLineCap: LineCapStyle
  borderOpacity: number;
  borderWidth: number;
  path: string;
}

Edits to PDFPage

All of the PDFPageDrawXxx functions will now return the appropriate PDFNode.

class PDFPage {
  // ... previously implemented

  // Added Return types
  drawCircle(options: PDFPageDrawCircleOptions): PDFShapeElipseNode;
  drawEllipse(options: PDFPageDrawEllipseOptions): PDFShapeElipseNode;
  drawImage(image: PDFImage, options: PDFPageDrawImageOptions): PDFImageNode;
  drawLine(options: PDFPageDrawLineOptions): PDFShapeLineNode;
  drawRectangle(options: PDFPageDrawRectangleOptions): PDFShapeRectangleNode;
  drawSquare(options: PDFPageDrawSquareOptions): PDFShapeSquareNode;
  drawSVGPath(image: string, options: PDFPageDrawSVGOptions): PDFSVGNode;
  drawText(text: string, options: PDFPageDrawTextOptions): PDFTextNode;

  // New function to appned created Node to document (optional)
  drawNode<N extends PDFNode>(node: N) => N;
}

[Byteschmiede]

I would love to see this. Is there any method to actual edit existing text elements on a pdf page? Like replace words or something like that?

[se181018]

I would also love to see this feature.
I found an abandoned project which, at least I think, capable of also modifying pdfs maybe it coud be of some help.
https://github.com/galkahana/HummusJS

And I think this also correlates to #177

[Hopding]

Added this to the roadmap for tracking: #998.