document.pdf

type Chart {
    type: String,           // Type of plot, e.g. 'line', 'bar', 'scatter'
    title: String,          // Title of the plot
    description: String,    // Description of the plot
}

type Table {
    description: String,    // Description of the table
    title: String,          // Title of the table
    caption: String,        // Caption of the table
    markdown: String,       // Markdown representation of the table
}

type Markdown {
    caption: String,        // Caption of the markdown content
    title: String,          // Title of the markdown content
    markdown: String,       // Additional markdown content
}

class Paragraph(BaseModel):
    content: str                 // Text content of the paragraph
    bbox: BoxCoords | None       // Bounding box coordinates of the paragraph
    type:

class DocumentPage(BaseModel):
    description: str | None    // Description of the page
    title: str | None          // Title of the page
    page_number: int | None    // Page number
    lines: list[Chart],       // Plots on the page
    tables: [Table],           // Tables on the page
    others: [Markdown],        // Other markdown content on the page
}

type DocumentPDF {
    title: String,          // Title of the document
    description: String,    // Description of the document
    pages: [DocumentPage],  // Pages in the document
}

{
    "title": "Sample PDF Document",                 # Title of the document
    "description": "...",                           # Description of the document
    "pages": [                                      # List of pages in the document
        {
            "description": "Page 1",                # Description of the first page
            "title": "Introduction",                # Title of the first page
            "page_number": 1,                       # Page number extracted from the first page
            "lines": [                              # Lines of text extracted from the first page
                {
                    "content": "Sales over time",
                    "bbox": {
                        "xywh": [
                            0., 0., 1., 1.0          # Normalized bounding box coordinates of the line {x, y, width, height}
                        ]
                    },
                    "type": "line"
                }
                ...                                # Other lines of text
            ],
            "paragraphs": [
                {
                    "content": "This is a paragraph of text.",
                    "bbox": {
                        "xywh": [
                            0., 0., 1., 1.          # Normalized bounding box coordinates of the line {x, y, width, height}
                        ]
                    },
                    "type": "paragraph"
                }
                ...                                 # Other paragraphs of text
            ]
            "tables": [                             # Tables extracted from the first page
                {
                    "data": [

                    ]
                }
                ...
            ],
            "others": [
                {
                    "caption": "Note",
                    "title": "Important Note",
                    "markdown": "This is an important note about the document."
                }
                ...
            ]
        },
        ... # Second page, third page, etc.
    ]

}