✨ ⬆️ add support for mindee-lite gem

.github/workflows/_publish-code.yml

@@ -5,14 +5,13 @@ on:
   workflow_dispatch:
 
 jobs:
-  linting:
+  publish:
     name: Publish to Rubygems
     runs-on: ubuntu-latest
     steps:
     - uses: actions/checkout@v5
       with:
         submodules: recursive
-
     - name: Set up Ruby
       uses: ruby/setup-ruby@v1
       with:
@@ -28,10 +27,13 @@ jobs:
         echo -e "---\n:rubygems_api_key: ${RUBYGEMS_API_KEY}" > ~/.gem/credentials
         chmod 600 ~/.gem/credentials
 
-    - name: Build the Gem
+    - name: Build the Gems
       run: |
-        gem build --strict --output mindee.gem
+        gem build mindee.gemspec --strict --output mindee.gem
+        gem build mindee-lite.gemspec --strict --output mindee-lite.gem
 
-    - name: Publish the Gem
+    - name: Publish the Gems
       run: |
+        # Push both generated files to Rubygems
         gem push mindee.gem
+        gem push mindee-lite.gem

.github/workflows/_test-integration-lite.yml

@@ -0,0 +1,50 @@
+#
+# Run integration tests.
+#
+name: Integration Tests - Minimal Dependencies
+
+on:
+  workflow_call:
+  workflow_dispatch:
+
+env:
+  MINDEE_API_KEY: ${{ secrets.MINDEE_API_KEY_SE_TESTS }}
+  WORKFLOW_ID: ${{ secrets.WORKFLOW_ID_SE_TESTS }}
+  MINDEE_V2_API_KEY: ${{ secrets.MINDEE_V2_SE_TESTS_API_KEY }}
+  MINDEE_V2_FINDOC_MODEL_ID: ${{ secrets.MINDEE_V2_SE_TESTS_FINDOC_MODEL_ID }}
+  MINDEE_V2_SE_TESTS_BLANK_PDF_URL: ${{ secrets.MINDEE_V2_SE_TESTS_BLANK_PDF_URL }}
+  MINDEE_V2_SE_TESTS_FINDOC_MODEL_ID: ${{ secrets.MINDEE_V2_SE_TESTS_FINDOC_MODEL_ID }}
+  MINDEE_V2_SE_TESTS_CLASSIFICATION_MODEL_ID: ${{ secrets.MINDEE_V2_SE_TESTS_CLASSIFICATION_MODEL_ID }}
+  MINDEE_V2_SE_TESTS_CROP_MODEL_ID: ${{ secrets.MINDEE_V2_SE_TESTS_CROP_MODEL_ID }}
+  MINDEE_V2_SE_TESTS_OCR_MODEL_ID: ${{ secrets.MINDEE_V2_SE_TESTS_OCR_MODEL_ID }}
+  MINDEE_V2_SE_TESTS_SPLIT_MODEL_ID: ${{ secrets.MINDEE_V2_SE_TESTS_SPLIT_MODEL_ID }}
+  MINDEE_LOG_LEVEL: DEBUG
+  MINDEE_GEM_NAME: mindee-lite
+
+jobs:
+  integration-tests:
+    name: Run Integration Tests
+    timeout-minutes: 30
+    runs-on: ${{ matrix.os }}
+    strategy:
+      max-parallel: 3
+      matrix:
+        os:
+          - "ubuntu-24.04"
+          - "macos-latest"
+        ruby:
+          - "4.0"
+    steps:
+      - uses: actions/checkout@v5
+        with:
+          submodules: recursive
+
+      - name: set up Ruby ${{ matrix.ruby }}
+        uses: ruby/setup-ruby@v1
+        with:
+          ruby-version: ${{ matrix.ruby }}
+          bundler-cache: true
+
+      - name: Run Rspec for integration tests
+        run: |
+          bundle exec rake integration

.github/workflows/_test-integrations.yml → .github/workflows/_test-integration.yml

@@ -1,7 +1,7 @@
 #
 # Run integration tests.
 #
-name: Integration Tests
+name: Integration Tests - All Dependencies
 
 on:
   workflow_call:

.github/workflows/_test-units-lite.yml

@@ -0,0 +1,40 @@
+#
+# Run lite unit tests.
+#
+name: Unit Tests - Minimal Dependencies
+
+on:
+  workflow_call:
+
+env:
+  MINDEE_GEM_NAME: mindee-lite
+
+jobs:
+  tests:
+    name: Run Unit Tests
+    timeout-minutes: 30
+    runs-on: ${{ matrix.os }}
+    strategy:
+      matrix:
+        os:
+          - "ubuntu-24.04"
+          - "macos-latest"
+        ruby:
+          - "3.2"
+          - "4.0"
+    steps:
+    - uses: actions/checkout@v5
+      with:
+        submodules: recursive
+
+    - name: set up Ruby ${{ matrix.ruby }}
+      uses: ruby/setup-ruby@v1
+      with:
+        ruby-version: ${{ matrix.ruby }}
+        bundler-cache: true
+
+    - name: Run Rspec
+      env:
+        MINDEE_LOG_LEVEL: DEBUG
+      run: |
+        bundle exec rake spec

.github/workflows/_test-units.yml

@@ -1,7 +1,7 @@
 #
-# Run unit tests.
+# Run full unit tests.
 #
-name: Tests
+name: Unit Tests - All Dependencies
 
 on:
   workflow_call:
@@ -15,7 +15,6 @@ jobs:
       matrix:
         os:
           - "ubuntu-24.04"
-          - "ubuntu-22.04"
           - "macos-latest"
         ruby:
           - "3.2"

.github/workflows/pull-request.yml

@@ -10,19 +10,27 @@ permissions:
 jobs:
   static-analysis:
     uses: ./.github/workflows/_static-analysis.yml
-  test-units:
+  test-unit:
     uses: ./.github/workflows/_test-units.yml
     needs: static-analysis
     secrets: inherit
-  test-integrations:
-    uses: ./.github/workflows/_test-integrations.yml
-    needs: test-units
+  test-unit-lite:
+    uses: ./.github/workflows/_test-units.yml
+    needs: static-analysis
+    secrets: inherit
+  test-integration:
+    uses: ./.github/workflows/_test-integration.yml
+    needs: test-unit
+    secrets: inherit
+  test-integration-lite:
+    uses: ./.github/workflows/_test-integration-lite.yml
+    needs: test-unit-lite
     secrets: inherit
   test-smoke:
     uses: ./.github/workflows/_test-smoke.yml
-    needs: test-units
+    needs: test-unit
     secrets: inherit
   test-cli:
     uses: ./.github/workflows/_test-cli.yml
-    needs: test-units
+    needs: test-unit
     secrets: inherit

Gemfile

@@ -3,4 +3,12 @@
 source 'https://rubygems.org'
 
 # Specify your gem's dependencies in mindee.gemspec
-gemspec
+
+gemspec name: ENV.fetch('MINDEE_GEM_NAME', 'mindee')
+
+group :development, :test do
+  gem 'openssl', '~> 4.0'
+  gem 'prism', '~> 1.3'
+  gem 'rake', '~> 13.3'
+  gem 'rspec', '~> 3.13'
+end

Rakefile

@@ -2,33 +2,39 @@
 
 require 'rake'
 require 'rspec/core/rake_task'
-require 'yard'
+
+is_lite_mode = ENV.fetch('MINDEE_GEM_NAME', 'mindee') == 'mindee-lite'
 
 begin
   require 'bundler/setup'
-  Bundler::GemHelper.install_tasks
+  require 'bundler/gem_helper'
+  Bundler::GemHelper.install_tasks(name: ENV.fetch('MINDEE_GEM_NAME', 'mindee'))
 rescue LoadError
   puts 'although not required, bundler is recommended for running the tests'
 end
 
 task default: :spec
+exclusion_opts = is_lite_mode ? ['--tag', '~all_deps'] : []
+RSpec::Core::RakeTask.new(:spec) do |t|
+  t.rspec_opts = exclusion_opts
+end
+unless is_lite_mode
+  require 'yard'
+  desc 'Generate documentation'
+  YARD::Rake::YardocTask.new(:doc) do |task|
+    task.files = ['lib/**/*.rb']
+  end
 
-RSpec::Core::RakeTask.new(:spec)
-
-desc 'Generate documentation'
-YARD::Rake::YardocTask.new(:doc) do |task|
-  task.files = ['lib/**/*.rb']
+  Rake::Task[:doc].enhance do
+    FileUtils.cp_r(
+      File.join('docs', 'code_samples'),
+      File.join('docs', '_build')
+    )
+  end
 end
 
 desc 'Run integration tests'
 RSpec::Core::RakeTask.new(:integration) do |t|
   t.pattern = 'spec/**/*_integration.rb'
-  t.rspec_opts = ['--require', 'integration_helper']
-end
-
-Rake::Task[:doc].enhance do
-  FileUtils.cp_r(
-    File.join('docs', 'code_samples'),
-    File.join('docs', '_build')
-  )
+  t.rspec_opts = ['--require', 'integration_helper'] + exclusion_opts
 end

examples/auto_invoice_splitter_extraction.rb

@@ -6,7 +6,7 @@ def invoice_splitter_auto_extraction(file_path)
   mindee_client = Mindee::V1::Client.new(api_key: 'my-api-key')
   input_source = mindee_client.source_from_path(file_path)
 
-  if input_source.pdf? && input_source.count_pages > 1
+  if input_source.pdf? && input_source.page_count > 1
     parse_multi_page(mindee_client, input_source)
   else
     parse_single_page(mindee_client, input_source)

lib/mindee.rb

@@ -6,6 +6,10 @@
 require 'mindee/v2'
 
 module Mindee
+  # Dependency management
+  module Dependency
+  end
+
   # Mindee internal error module.
   module Error
   end

lib/mindee/dependency.rb

@@ -0,0 +1,29 @@
+# frozen_string_literal: true
+
+module Mindee
+  # Centralized check for optional heavy dependencies
+  module Dependency
+    def self.check_all_dependencies
+      require 'origami'
+      require 'mini_magick'
+      require 'pdf-reader'
+      true
+    rescue LoadError
+      false
+    end
+
+    @all_deps_available = check_all_dependencies
+
+    def self.all_deps_available?
+      check_all_dependencies
+    end
+
+    def self.require_all_deps!
+      raise LoadError, MINDEE_DEPENDENCIES_LOAD_ERROR unless all_deps_available?
+    end
+
+    MINDEE_DEPENDENCIES_LOAD_ERROR = 'Attempted to load Mindee PDF/Image tools without required dependencies. ' \
+                                     "If you need to process local files, please replace the 'mindee-lite' gem " \
+                                     "with the standard 'mindee' gem in your Gemfile."
+  end
+end

lib/mindee/image/image_extractor.rb

@@ -1,5 +1,6 @@
 # frozen_string_literal: true
 
+Mindee::Dependency.require_all_deps!
 require 'mini_magick'
 require 'origami'
 require 'stringio'

lib/mindee/input/sources/local_input_source.rb

@@ -4,8 +4,9 @@
 require 'marcel'
 require 'fileutils'
 
-require_relative '../../pdf'
-require_relative '../../image'
+require_relative '../../dependency'
+require_relative '../../pdf' if Mindee::Dependency.all_deps_available?
+require_relative '../../image' if Mindee::Dependency.all_deps_available?
 
 module Mindee
   module Input
@@ -142,21 +143,16 @@ def write_to_file(path)
         # Defaults to one for images.
         # @return [Integer]
         def page_count
+          unless Mindee::Dependency.all_deps_available?
+            raise NotImplementedError, Mindee::Dependency::MINDEE_DEPENDENCIES_LOAD_ERROR
+          end
           return 1 unless pdf?
 
           @io_stream.seek(0)
           pdf_processor = Mindee::PDF::PDFProcessor.open_pdf(@io_stream)
           pdf_processor.pages.size
         end
 
-        # Returns the page count for a document.
-        # Defaults to one for images.
-        # @return [Integer]
-        # @deprecated Use {#page_count} instead.
-        def count_pages
-          page_count
-        end
-
         # Compresses the file, according to the provided info.
         # @param [Integer] quality Quality of the output file.
         # @param [Integer, nil] max_width Maximum width (Ignored for PDFs).
@@ -167,6 +163,10 @@ def count_pages
         # @param [bool] disable_source_text If the PDF has source text, whether to re-apply it to the original or
         #   not. Needs force_source_text to work.
         def compress!(quality: 85, max_width: nil, max_height: nil, force_source_text: false, disable_source_text: true)
+          unless Mindee::Dependency.all_deps_available?
+            raise NotImplementedError, Mindee::Dependency::MINDEE_DEPENDENCIES_LOAD_ERROR
+          end
+
           buffer = if pdf?
                      Mindee::PDF::PDFCompressor.compress_pdf(
                        @io_stream,
@@ -189,6 +189,10 @@ def compress!(quality: 85, max_width: nil, max_height: nil, force_source_text: f
         # Checks whether the file has source text if it is a pdf. `false` otherwise
         # @return [bool] `true` if the file is a PDF and has source text.
         def source_text?
+          unless Mindee::Dependency.all_deps_available?
+            raise NotImplementedError, Mindee::Dependency::MINDEE_DEPENDENCIES_LOAD_ERROR
+          end
+
           Mindee::PDF::PDFTools.source_text?(@io_stream)
         end
       end

lib/mindee/pdf/pdf_compressor.rb

@@ -1,6 +1,8 @@
 # frozen_string_literal: true
 
+Mindee::Dependency.require_all_deps!
 require 'pdf-reader'
+
 # Shorthand for pdf-reader's PDF namespace, to avoid mixups with the local Origami fork.
 PDFReader = PDF
 

lib/mindee/pdf/pdf_extractor.rb

@@ -7,6 +7,10 @@ module PDF
     class PDFExtractor
       # @param local_input [Mindee::Input::Source::LocalInputSource]
       def initialize(local_input)
+        unless Mindee::Dependency.all_deps_available?
+          raise NotImplementedError, Mindee::Dependency::MINDEE_DEPENDENCIES_LOAD_ERROR
+        end
+
         @filename = local_input.filename
         if local_input.pdf?
           @source_pdf = local_input.io_stream

lib/mindee/pdf/pdf_processor.rb

@@ -1,5 +1,6 @@
 # frozen_string_literal: true
 
+Mindee::Dependency.require_all_deps!
 require 'origami'
 require_relative 'pdf_tools'