kreuzberg 4.0.0.rc2 → 4.0.1

data/spec/binding/images_spec.rb

@@ -0,0 +1,738 @@
+# frozen_string_literal: true
+
+require 'spec_helper'
+require 'tempfile'
+require 'fileutils'
+
+RSpec.describe 'Image Extraction' do
+  describe 'PDF image extraction with metadata' do
+    it 'extracts images with format and dimensions' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true,
+          target_dpi: 150
+        )
+      )
+
+      pdf_path = test_document_path('pdf/with_images.pdf')
+      begin
+        result = Kreuzberg.extract_file_sync(path: pdf_path, config: config)
+
+        expect(result).not_to be_nil
+        expect(result.images).not_to be_nil
+        if result.images && !result.images.empty?
+          image = result.images.first
+          expect(image).to be_a(Kreuzberg::Result::Image)
+          expect(image.format).not_to be_nil
+          expect(image.width).to be > 0
+          expect(image.height).to be > 0
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'includes page numbers in extracted images' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true,
+          target_dpi: 150
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        expect(result.images).not_to be_nil
+        if result.images && !result.images.empty?
+          result.images.each do |image|
+            expect(image.page_number).to be > 0
+          end
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'respects target_dpi configuration parameter' do
+      dpi_values = [150, 300, 600]
+
+      dpi_values.each do |dpi|
+        config = Kreuzberg::Config::Extraction.new(
+          image_extraction: Kreuzberg::Config::ImageExtraction.new(
+            extract_images: true,
+            target_dpi: dpi
+          )
+        )
+
+        begin
+          result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+          expect(result).not_to be_nil
+          expect(result.images).not_to be_nil
+        rescue Kreuzberg::Errors::ValidationError
+          skip 'Test file not available'
+        end
+      end
+    end
+
+    it 'includes colorspace information in image metadata' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true,
+          target_dpi: 150
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        if result.images && !result.images.empty?
+          image = result.images.first
+          expect(image).to respond_to(:colorspace)
+          # Verify colorspace has meaningful value if present
+          if image.colorspace
+            expect(image.colorspace).not_to be_empty
+            expect(image.colorspace).to be_a(String)
+          end
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+  end
+
+  describe 'Image handling in composite documents' do
+    it 'extracts images from DOCX files' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+
+      begin
+        docx_path = test_document_path('office/document.docx')
+        result = Kreuzberg.extract_file_sync(path: docx_path, config: config)
+
+        expect(result).not_to be_nil
+        expect(result.content).not_to be_nil
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'extracts images from PPTX files' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+
+      begin
+        pptx_path = test_document_path('presentations/simple.pptx')
+        result = Kreuzberg.extract_file_sync(path: pptx_path, config: config)
+
+        expect(result).not_to be_nil
+        expect(result.content).not_to be_nil
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'handles documents with multiple images across pages' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true,
+          target_dpi: 150
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        expect(result.images).not_to be_nil
+        if result.images && result.images.length > 1
+          page_numbers = result.images.map(&:page_number).uniq
+          expect(page_numbers.length).to be > 1
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'preserves image index for sequential extraction' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        if result.images && result.images.length > 1
+          result.images.each_with_index do |image, _index|
+            expect(image.image_index).to be_a(Integer)
+          end
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+  end
+
+  describe 'Image format detection' do
+    it 'detects PNG format in extracted images' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true,
+          target_dpi: 150
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        if result.images && !result.images.empty?
+          formats = result.images.filter_map(&:format)
+          expect(formats).to be_an(Array)
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'detects JPEG format in extracted images' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true,
+          target_dpi: 150
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        if result.images && !result.images.empty?
+          result.images.each do |image|
+            expect(image.format).not_to be_nil
+            expect(image.format).to be_a(String)
+          end
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'handles WebP format detection if present' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        expect(result.images).not_to be_nil
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'provides consistent format strings across extractions' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+
+      begin
+        result1 = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+        result2 = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        if result1.images && result2.images && !result1.images.empty? && !result2.images.empty?
+          expect(result1.images.first.format).to eq(result2.images.first.format)
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+  end
+
+  describe 'Embedded vs referenced images' do
+    it 'extracts embedded images from documents' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        expect(result.images).not_to be_nil
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'handles image data field in extracted images' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true,
+          target_dpi: 150
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        if result.images && !result.images.empty?
+          image = result.images.first
+          expect(image).to respond_to(:data)
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'preserves image metadata when extraction enabled' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        if result.images && !result.images.empty?
+          image = result.images.first
+          expect(image.width).to be_a(Integer)
+          expect(image.height).to be_a(Integer)
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'returns nil for images when extraction disabled' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: false
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        expect(result.images).to be_nil
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+  end
+
+  describe 'Error handling for corrupted images' do
+    it 'gracefully handles documents with malformed images' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+        expect(result).not_to be_nil
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'continues extraction when encountering problematic images' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        expect(result).not_to be_nil
+        expect(result.content).not_to be_nil
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'handles extraction with max_image_dimension constraint' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true,
+          max_image_dimension: 1000
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        if result.images && !result.images.empty?
+          result.images.each do |image|
+            expect(image.width).to be_a(Integer)
+            expect(image.height).to be_a(Integer)
+          end
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'respects auto_adjust_dpi configuration' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true,
+          auto_adjust_dpi: true,
+          min_dpi: 150,
+          max_dpi: 600
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        expect(result).not_to be_nil
+        expect(result.images).not_to be_nil
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+  end
+
+  describe 'Batch image extraction from multi-page documents' do
+    it 'extracts images from multi-page PDF in single operation' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true,
+          target_dpi: 150
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        expect(result).not_to be_nil
+        expect(result.images).not_to be_nil
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'maintains correct page associations for extracted images' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        if result.images && result.images.length > 1
+          result.images.each do |image|
+            expect(image.page_number).to be >= 1
+          end
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'preserves image order within document' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        if result.images && result.images.length > 1
+          (0...(result.images.length - 1)).each do |i|
+            expect(result.images[i].image_index).to be <= result.images[i + 1].image_index
+          end
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'handles multiple file batch extraction with images' do
+      paths = []
+      2.times do |i|
+        file = Tempfile.new("batch_image_test_#{i}.txt")
+        file.write("Image extraction test #{i}")
+        file.close
+        paths << file.path
+      end
+
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+
+      results = Kreuzberg.batch_extract_files_sync(paths: paths, config: config)
+
+      expect(results).to be_a(Array)
+      expect(results.length).to eq(2)
+      expect(results).to all(be_a(Kreuzberg::Result))
+    ensure
+      paths.each { |p| FileUtils.rm_f(p) }
+    end
+
+    it 'maintains correct image count across batch operations' do
+      paths = []
+      2.times do |i|
+        file = Tempfile.new("batch_count_#{i}.txt")
+        file.write("Content #{i}")
+        file.close
+        paths << file.path
+      end
+
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+
+      results = Kreuzberg.batch_extract_files_sync(paths: paths, config: config)
+
+      expect(results.length).to eq(paths.length)
+      expect(results).to all(be_a(Kreuzberg::Result))
+    ensure
+      paths.each { |p| FileUtils.rm_f(p) }
+    end
+  end
+
+  describe 'ImageExtraction configuration integration' do
+    it 'applies different DPI settings to affect extraction behavior' do
+      config_low = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true,
+          target_dpi: 72
+        )
+      )
+      config_high = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true,
+          target_dpi: 300
+        )
+      )
+
+      begin
+        result_low = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config_low)
+        result_high = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config_high)
+
+        # Both configurations should produce valid extraction
+        expect(result_low).not_to be_nil
+        expect(result_high).not_to be_nil
+        # Different DPI settings should be accepted
+        expect([result_low, result_high]).to all(be_a(Kreuzberg::Result))
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'respects extract_images false disables extraction' do
+      config_enabled = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+      config_disabled = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: false
+        )
+      )
+
+      begin
+        result_enabled = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config_enabled)
+        result_disabled = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config_disabled)
+
+        # Enabled should extract if images present
+        expect(result_enabled).not_to be_nil
+        # Disabled should return nil or empty images
+        expect(result_disabled.images).to be_empty if result_disabled.images
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'handles dimension constraints realistically' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true,
+          max_image_dimension: 1024
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        expect(result).not_to be_nil
+        # Dimension constraint should be applied
+        if result.images && !result.images.empty?
+          result.images.each do |image|
+            # Image should respect dimension constraints
+            expect(image).not_to be_nil
+          end
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+  end
+
+  describe 'Integration with Extraction config' do
+    it 'accepts ImageExtraction config in Extraction' do
+      image_config = Kreuzberg::Config::ImageExtraction.new(
+        extract_images: true,
+        target_dpi: 600
+      )
+      config = Kreuzberg::Config::Extraction.new(image_extraction: image_config)
+
+      expect(config.image_extraction).to be_a(Kreuzberg::Config::ImageExtraction)
+      expect(config.image_extraction.target_dpi).to eq(600)
+    end
+
+    it 'accepts image extraction config as hash in Extraction' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: {
+          extract_images: true,
+          target_dpi: 600,
+          max_image_dimension: 3000
+        }
+      )
+
+      expect(config.image_extraction).to be_a(Kreuzberg::Config::ImageExtraction)
+      expect(config.image_extraction.extract_images).to be true
+      expect(config.image_extraction.target_dpi).to eq(600)
+      expect(config.image_extraction.max_image_dimension).to eq(3000)
+    end
+
+    it 'includes image extraction config in to_h' do
+      image_config = Kreuzberg::Config::ImageExtraction.new(
+        extract_images: true,
+        target_dpi: 600
+      )
+      config = Kreuzberg::Config::Extraction.new(image_extraction: image_config)
+
+      hash = config.to_h
+
+      expect(hash).to include(:image_extraction)
+      expect(hash[:image_extraction]).to be_a(Hash)
+      expect(hash[:image_extraction][:extract_images]).to be true
+      expect(hash[:image_extraction][:target_dpi]).to eq(600)
+    end
+
+    it 'combines image extraction with other configurations' do
+      config = Kreuzberg::Config::Extraction.new(
+        use_cache: true,
+        force_ocr: true,
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true,
+          target_dpi: 600
+        ),
+        ocr: Kreuzberg::Config::OCR.new(
+          backend: 'tesseract',
+          language: 'eng'
+        )
+      )
+
+      expect(config.use_cache).to be true
+      expect(config.force_ocr).to be true
+      expect(config.image_extraction.target_dpi).to eq(600)
+      expect(config.ocr.backend).to eq('tesseract')
+    end
+
+    it 'handles nil image extraction config' do
+      config = Kreuzberg::Config::Extraction.new(image_extraction: nil)
+
+      expect(config.image_extraction).to be_nil
+    end
+  end
+
+  describe 'Image metadata validation in real extractions' do
+    it 'validates extracted images have complete required metadata' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        if result.images && !result.images.empty?
+          result.images.each do |image|
+            # All extracted images must have these fields populated
+            expect(image).not_to be_nil
+            expect(image.format).not_to be_nil, 'Format is required'
+            expect(image.format).not_to be_empty
+            expect(image.image_index).to be >= 0, 'Image index must be non-negative'
+            expect(image.data).not_to be_nil, 'Image data is required'
+          end
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'includes optional metadata fields appropriately' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true,
+          target_dpi: 150
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        if result.images && !result.images.empty?
+          result.images.each do |image|
+            # Optional fields should be valid when present
+            expect(image.width).to be > 0, 'Width should be positive when present' if image.width
+            expect(image.height).to be > 0, 'Height should be positive when present' if image.height
+            expect(image.page_number).to be > 0, 'Page number should be positive' if image.page_number
+          end
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+
+    it 'ensures multiple images have different indices' do
+      config = Kreuzberg::Config::Extraction.new(
+        image_extraction: Kreuzberg::Config::ImageExtraction.new(
+          extract_images: true
+        )
+      )
+
+      begin
+        result = Kreuzberg.extract_file_sync(path: test_document_path('pdf/with_images.pdf'), config: config)
+
+        if result.images && result.images.length > 1
+          indices = result.images.map(&:image_index)
+          unique_indices = indices.uniq
+          expect(unique_indices.length).to eq(indices.length), 'Each image should have unique index'
+        end
+      rescue Kreuzberg::Errors::ValidationError
+        skip 'Test file not available'
+      end
+    end
+  end
+end