Class: Tabula::TextStripper

Inherits:
Object
  • Object
show all
Defined in:
lib/tabula/pdf/text_stripper.rb

Overview

Extracts text elements from PDF pages using pdf-reader. Uses pdf-reader's PageTextReceiver for proper font encoding and CMap handling.

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(page) ⇒ TextStripper

Returns a new instance of TextStripper.

Parameters:

  • page (PDF::Reader::Page) —

    pdf-reader page object



10
11
12
13
14
15
 
# File 'lib/tabula/pdf/text_stripper.rb', line 10

def initialize(page)
  @page = page
  @text_elements = []
  @min_char_width = Float::INFINITY
  @min_char_height = Float::INFINITY
end

Instance Attribute Details

#min_char_height ⇒ Object (readonly)

Returns the value of attribute min_char_height.



80
81
82
 
# File 'lib/tabula/pdf/text_stripper.rb', line 80

def min_char_height
  @min_char_height
end

#min_char_width ⇒ Object (readonly)

Returns the value of attribute min_char_width.



80
81
82
 
# File 'lib/tabula/pdf/text_stripper.rb', line 80

def min_char_width
  @min_char_width
end

Instance Method Details

#extract ⇒ Array<TextElement>

Extract text elements from the page

Returns:



19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
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
 
# File 'lib/tabula/pdf/text_stripper.rb', line 19

def extract
  # Use pdf-reader's PageTextReceiver for proper font encoding
  receiver = PDF::Reader::PageTextReceiver.new
  receiver.page = @page
  @page.walk(receiver)

  # Get merged text runs for readable output
  # merge: true combines adjacent characters into words/phrases
  runs = receiver.runs(
    merge: true,
    skip_zero_width: true,
    skip_overlapping: true
  )

  # Get page dimensions and rotation
  rotation = @page.attributes[:Rotate] || 0

  runs.each do |run|
    next if run.text.nil? || run.text.empty?
    next unless printable?(run.text)

    # pdf-reader already applies rotation transformation
    # For rotated pages, y coordinates are negative
    # For non-rotated pages, we need to flip from bottom-origin to top-origin
    if [90, 270].include?(rotation)
      # Rotated pages: y is negative, convert to positive
      top = run.y.abs
    else
      # Non-rotated pages: convert from bottom-origin to top-origin
      page_height = calculate_page_height
      top = page_height - run.y
    end

    left = run.x
    width = run.width
    height = run.font_size

    # Detect text direction from Unicode character properties
    direction = rtl_text?(run.text) ? TextElement::DIRECTION_RTL : TextElement::DIRECTION_LTR

    element = TextElement.new(
      top: top,
      left: left,
      width: width,
      height: height,
      text: run.text,
      font_name: nil, # pdf-reader doesn't expose font name in runs
      font_size: run.font_size,
      width_of_space: estimate_space_width(run),
      direction: direction
    )

    @text_elements << element

    @min_char_width = [@min_char_width, width].min if width.positive?
    @min_char_height = [@min_char_height, height].min if height.positive?
  end

  @text_elements
end