PDF::Inspector::Text.analyze does not return strings within repeat block

[ruinunes]

When using PDF::Inspector::Text.analyze(some_pdf).strings, it does not return strings that were added via

repeat(1..page_count) do
  text_box "Hello", at: [0, bounds.top]
end

Is there a way to get a hold of these strings?

[pointlessone]

Hi Rui,

Could you please provide a minimal example that demonstrates the issue?

[ruinunes]

The minimal example is pretty much as above.
In a prawn document, if we write text_box "Hello", at: [0, bounds.top], the Hello string can be displayed when outputting with for example puts PDF::Inspector::Text.analyze(some_pdf).strings. However, if the text_box or text are sitting inside a repeat block, then the string will not be available in the strings array, when inspecting via PDF::Inspector::Text.analyze(some_pdf).strings.

text_box "Hello A", at: [0, bounds.top]
repeat(1..page_count) do
  text_box "Hello B", at: [100, bounds.top]
end

Hello B will not be available in the strings collection. Only Hello A.

[ruinunes]

@pointlessone what are the requirements for a minimal example?

[pointlessone]

A good example should be:

• Complete — someone could copy provided code into a file, run it and it would demonstrate your issue. It also means that if your issue requires any other resources (e.g. fonts, images, etc.) they should be provided as well. Also make sure you povide relevant information about your environment: Prawn version (latest release is implied, but not always the case), OS (including version), Ruby engine and version, etc.
• Minimal — the code should include the minimal number of instructions required to demonstrate the issue. Please don't post 4k lines of your production code for someone else to sift through. That is of course unless that's what required to demonstrate the issue. Time is the most constrained resource on any OSS projects. The more work you do, the less there's left for someone else. It's more likely someone would be willing to invest smaller amount of time. Higher chance your issue would be resolved sooner.

Don't get me wrong. I (and everyone else) appreciate that you're reporting issues. That is a valuable contribution. Sometimes that is all people can contribute. But is they can do more it helps a lot.

[JoelWAnna]

failing test example

    it 'includes footer in analysed text' do
        pdf = Prawn::Document.new
        page_1 = 'First Page'
        page_2 = 'Second Page'
        pdf.text page_1, align: :center
        pdf.start_new_page
        pdf.text page_2, align: :center

        pdf.repeat(:all) do
          pdf.text_box 'test'
        end

        rendered_pdf = pdf.render
        page_analysis = PDF::Inspector::Page.analyze(rendered_pdf)
        expect(page_analysis.pages.size).to eq 2
        strings = PDF::Inspector::Text.analyze(rendered_pdf).strings
        expect(strings).to eq [page_1, 'test', page_2, 'test']

    end

[cupofjoakim]

Having the same issue here.

  it "contains client details" do
    text_analysis = PDF::Inspector::Text.analyze(@pdf.render)
    content_string = text_analysis.strings.join(" ")
    expect(content_string).to include("Client Name")
  end

And the expected string is in a header that is repeated on all pages:```

  def header(header_height)
    repeat :all do
      # Header
      bounding_box([0, bounds.top], width: 180, height: header_height) do
        move_down 8
        font_size 9
        text "Client name"
      end
    end
  end

[henrik]

I'm seeing the same with create_stamp/stamp. Worked fine with pdf-inspector 1.0.2 and pdf-reader 1.3.3. But fails (the text is not included) with pdf-inspector 1.3.0 and pdf-reader 2.0.0.

The underlying issue/change in behaviour seems to be with pdf-reader. Issue here: yob/pdf-reader#268