action cloudmersive_pdf_ocr_pdf_to_words_with_location {
  label: "Convert a PDF into words with location"
  description: "Converts a PDF into words/text with location information and other metdata via Optical Character Recognition.  This API is intended to be run on scanned documents.  If you want to OCR photos (e.g. taken with a smart phone camera), be sure to use the photo/toText API instead, as it is designed to unskew the image first."
  provider: cloudmersive
  method: POST
  path: "/ocr/pdf/to/words-with-location"
  encoding: json
  input: {
    type: "object"
    properties: {
      language: {
        type: "string"
      }
      preprocessing: {
        type: "string"
      }
    }
    additionalProperties: false
  }
  output: {
    type: "object"
    description: "Response from an OCR to words with location operation.  Includes the confience rating and converted text result, along with the locations of the words in the pages."
    properties: {
      OcrPages: {
        type: "array"
        description: "OCR page results"
        items: {
          type: "object"
          description: "OCR results of a page, including words of text and their location"
          properties: {
            PageNumber: {
              type: "integer"
              format: "int32"
              description: "Page number of the page that was OCR-ed, starting with 1 for the first page in the PDF file"
            }
            Successful: {
              type: "boolean"
              description: "True if successful, false otherwise"
            }
            Words: {
              type: "array"
              description: "Word elements in the image"
              items: {
                type: "object"
                description: "A single word in an OCR document"
                properties: {
                  BlockNumber: {
                    type: "integer"
                    format: "int32"
                    description: "Index of the containing block"
                  }
                  ConfidenceLevel: {
                    type: "number"
                    format: "double"
                    description: "Confidence level of the machine learning result; possible values are 0.0 (lowest accuracy) - 1.0 (highest accuracy)"
                  }
                  Height: {
                    type: "integer"
                    format: "int32"
                    description: "Height of the word in pixels"
                  }
                  LineNumber: {
                    type: "integer"
                    format: "int32"
                    description: "Line number of the word"
                  }
                  PageNumber: {
                    type: "integer"
                    format: "int32"
                    description: "Index of the containing page"
                  }
                  ParagraphNumber: {
                    type: "integer"
                    format: "int32"
                    description: "Index of the containing paragraph"
                  }
                  Width: {
                    type: "integer"
                    format: "int32"
                    description: "Width of the word in pixels"
                  }
                  WordNumber: {
                    type: "integer"
                    format: "int32"
                    description: "Index of the word in the line"
                  }
                  WordText: {
                    type: "string"
                    description: "Text of the word"
                  }
                  XLeft: {
                    type: "integer"
                    format: "int32"
                    description: "X location of the left edge of the word in pixels"
                  }
                  YTop: {
                    type: "integer"
                    format: "int32"
                    description: "Y location of the top edge of the word in pixels"
                  }
                }
              }
            }
          }
        }
      }
      Successful: {
        type: "boolean"
        description: "True if successful, false otherwise"
      }
    }
  }
}