Class: DynamicPDFApi::PdfText

Inherits:
Endpoint
  • Object
show all
Defined in:
lib/ruby_client/PdfText.rb

Overview

Represents the pdf text endpoint.

Instance Attribute Summary collapse

Attributes inherited from Endpoint

#_endpoint_name, #api_key, #base_url

Instance Method Summary collapse

Methods inherited from Endpoint

#init

Constructor Details

#initialize(resource, start_page = 1, page_count = 0, text_order = TextOrder::STREAM) ⇒ PdfText

Initializes a new instance of the PdfText class.



18
19
20
21
22
23
24
25
 
# File 'lib/ruby_client/PdfText.rb', line 18

def initialize(resource, start_page = 1, page_count = 0, text_order = TextOrder::STREAM)
  @_endpoint_name = 'pdf-text'
  super()
  @resource = resource
  @start_page = start_page
  @page_count = page_count
  @text_order = text_order
end

Instance Attribute Details

#page_count ⇒ Object

Gets or sets the page count.



35
36
37
 
# File 'lib/ruby_client/PdfText.rb', line 35

def page_count
  @page_count
end

#start_page ⇒ Object

Gets or sets the start page.



30
31
32
 
# File 'lib/ruby_client/PdfText.rb', line 30

def start_page
  @start_page
end

#text_order ⇒ Object

Gets or sets the text extraction order.



40
41
42
 
# File 'lib/ruby_client/PdfText.rb', line 40

def text_order
  @text_order
end

Instance Method Details

#process ⇒ Object

Process the pdf resource to get pdf’s text.



46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
 
# File 'lib/ruby_client/PdfText.rb', line 46

def process
  header = {
    'Authorization': "Bearer #{@api_key}",
    'Content-Length': @resource.data.length.to_s,
    'Expect': '100-continue',
    'Content-Type': 'application/pdf'
  }
  uri = URI.parse("#{@base_url}/v1.0/#{@_endpoint_name}")
  params = { 'startPage' => @start_page, 'pageCount' => @page_count, 'textOrder' => @text_order }
  uri.query = URI.encode_www_form(params)

  request = Net::HTTP::Post.new(uri.request_uri, header)
  request.set_form_data(params)
  req_options = {
    use_ssl: uri.scheme == 'https',
    verify_mode: OpenSSL::SSL::VERIFY_NONE

  }

  request.content_type = 'application/pdf'
  request.body = @resource.data

  response = Net::HTTP.start(uri.hostname, uri.port, req_options) do |http|
    http.request(request)
  end
  out_data = response.body
  ret_object = PdfTextResponse.new(out_data)
  ret_object.is_successful = false
  ret_object.status_code = response.code
  if ret_object.status_code == '200'
    ret_object.is_successful = true
  else
    if ret_object.status_code == '401'
      raise "Invalid api key specified."
    end
    out_data_json = JSON.parse(out_data)
    ret_object.error_json = out_data
    ret_object.error_message = if !out_data_json['message'].nil?
                                 out_data_json['message']
                               else
                                 "status_code : #{Net::HTTPResponse::CODE_TO_OBJ[ret_object.status_code]}"
                               end
    ret_object.error_id = out_data_json['id']
  end

  ret_object
end