Text & characters

  1. Plain text
  2. Character-level metadata
  3. Words
  4. Search

Sample PDF for this guide: text.pdf — two pages of structured text with mixed fonts (Helvetica and Helvetica-Bold).

Plain text 

require "rpdfium"

Rpdfium.open("text.pdf") do |doc|
  page = doc.page(0)
  puts page.text.lines.first(3).join
end

Output:

Relazione di esempio
Azienda S.R.L. — P.IVA 01234567890
Sezione 1

Restrict to a region of the page:

page.text_in_bbox(left: 50, top: 100, right: 300, bottom: 150)

Character-level metadata

Per-char data is essential for layout-aware processing — bounding box, font, weight, origin, rotation angle, plus PDFium’s “character provenance” flags:

Rpdfium.open("text.pdf") do |doc|
  pp doc.page(0).chars.first
end

Output:

{char: "R",
 codepoint: 82,
 x0: 50.0, x1: 62.99, top: 50.0, bottom: 66.65,
 origin_x: 50.0, origin_y: 62.92,
 angle: 0.0,                  # radians (rotated text)
 fontsize: 18.0,
 font: "Helvetica-Bold",
 weight: 700,
 render_mode: 0,              # 0=fill 1=stroke 2=both 3=invisible
 generated: false,            # true → inserted by PDFium (e.g. spaces)
 hyphen: false,               # true → soft-hyphen for line break
 unicode_error: false,        # true → couldn't map glyph to unicode
 advance: 12.99,
 text_obj_id: 44737295712,    # groups chars of the same text object (varies per run)
 text_obj_ends_with_space: false}

generated / hyphen / unicode_error are the artefact recognition flags. Distinguishing real characters from PDFium-synthesized ones is crucial when you don’t want fake whitespace to widen a column.

Loose char boxes (proportional to font size, more stable for layout algorithms):

page.chars(loose: true)

Words

Cluster chars into words automatically:

Rpdfium.open("text.pdf") do |doc|
  words = doc.page(0).words(x_tolerance: 3.0, y_tolerance: 3.0)
  pp words.first(3).map { |w| w[:text] }
end

Output:

["Relazione", "di", "esempio"]

Each word carries its own bbox, font, fontsize, and the underlying chars.

page.search("esempio", match_case: false).each_match do |m|
  puts "found '#{m[:text]}' at #{m[:rects].first.inspect}"
end