excavate 0.2.4 → 0.3.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 4dedf2dff68fab0c162340e9dc591f1675612d2b6f46335c6fe599ecead65434
-  data.tar.gz: 8d8dd45fbe6ba733d4aa759a3f9f1bb5fc457be3fa0ae176b48bc0d1a417c8ec
+  metadata.gz: b66a3e8ff5083bd78711cff8289e18f958578d84fea44be53dcd4f5349508755
+  data.tar.gz: 1c65ff305ed20893b27470063280f60698d9920db597d42af8e75127c947784a
 SHA512:
-  metadata.gz: b3f6c9a0f90d398afe6ee88c64004aaa5aad6ef50382d095e9df67a2841b3ae7c1af46a2f760718837d10350d2de4b90f727b495c2d8fdd6f32ae9dd290b7de1
-  data.tar.gz: c33b1e35ec3c7478b9236ee705f683b937394f495bb60b57e7602841136e951cbd8eee71ceaabf586548218302e6c495010b7a30020726056d6dd3f47e2bd0ff
+  metadata.gz: feac4007753c19cb52d9f32432f96ab2a2cf6553556cb950ad15f210a0e8f3d0209622b81702e17d4a9fef3e975b42f7f97178d7d55dfc914731245093a159f0
+  data.tar.gz: 93bed15a89d5f620e45232f9de42afde7efaabc7e79734617c8443f2324858a6a97e6b9e40995ee18ad5bce4d36ea1ac584b238d257accf567e88472ea0fd991

data/.github/workflows/metanorma.yml

@@ -0,0 +1,76 @@
+name: metanorma
+
+on:
+  push:
+    branches: [ main ]
+  pull_request:
+
+concurrency:
+  group: '${{ github.workflow }}-${{ github.job }}-${{ github.head_ref || github.ref_name }}'
+  cancel-in-progress: true
+
+jobs:
+  prepare:
+    runs-on: ubuntu-latest
+    outputs:
+      head_tag: ${{ steps.check.outputs.head_tag }}
+      foreign_pr: ${{ steps.check.outputs.foreign_pr }}
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v2.3.4
+
+      - name: Retrieve tags
+        run: git fetch --depth=1 origin +refs/tags/*:refs/tags/* || true
+      - name: Set output variables
+        id: check
+        run: |
+          fpr="no"
+          tag=""
+          if [[ "${{ github.ref }}" == refs/heads/* ]]; then
+            tag="$(git tag --points-at HEAD)"
+          elif [[ "${{ github.ref }}" == refs/pull/* ]] && [ "${{ github.event.pull_request.head.repo.full_name }}" != "${{ github.event.pull_request.base.repo.full_name }}" ]; then
+            fpr="yes"
+          fi
+          echo "::set-output name=foreign_pr::${fpr}"
+          echo "::set-output name=head_tag::${tag}"
+
+  test:
+    name: Test with Metanorma on Ruby ${{ matrix.ruby }} ${{ matrix.os }}
+    needs: prepare
+    runs-on: ${{ matrix.os }}
+    continue-on-error: false
+    strategy:
+      fail-fast: false
+      matrix:
+        ruby: [ '2.7', '3.0', '3.1' ]
+        os: [ ubuntu-latest, windows-latest, macos-latest ]
+    steps:
+      - uses: actions/checkout@v3
+        with:
+          repository: metanorma/metanorma
+
+      - uses: metanorma/metanorma-build-scripts/inkscape-setup-action@main
+
+      - uses: actions/checkout@v3
+        with:
+          path: "excavate"
+
+      - run: 'echo ''gem "excavate", path: "./excavate"'' > Gemfile.devel'
+
+      - if: matrix.ruby == '3.0'
+        uses: ruby/setup-ruby@v1
+        with:
+          ruby-version: ${{ matrix.ruby }}
+          bundler-cache: false
+          rubygems: latest
+
+      - if: matrix.ruby == '3.0'
+        run: bundle
+
+      - if: matrix.ruby != '3.0'
+        uses: ruby/setup-ruby@v1
+        with:
+          ruby-version: ${{ matrix.ruby }}
+          bundler-cache: true
+
+      - run: bundle exec rake

data/.github/workflows/test-and-release.yml

@@ -0,0 +1,69 @@
+name: test-and-release
+
+on:
+  push:
+    branches: [ main ]
+    tags: [ 'v*' ]
+  pull_request:
+
+concurrency:
+  group: '${{ github.workflow }}-${{ github.job }}-${{ github.head_ref || github.ref_name }}'
+  cancel-in-progress: true
+
+jobs:
+  prepare:
+    runs-on: ubuntu-latest
+    outputs:
+      head_tag: ${{ steps.check.outputs.head_tag }}
+      foreign_pr: ${{ steps.check.outputs.foreign_pr }}
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v2.3.4
+
+      - name: Retrieve tags
+        run: git fetch --depth=1 origin +refs/tags/*:refs/tags/* || true
+      - name: Set output variables
+        id: check
+        run: |
+          fpr="no"
+          tag=""
+          if [[ "${{ github.ref }}" == refs/heads/* ]]; then
+            tag="$(git tag --points-at HEAD)"
+          elif [[ "${{ github.ref }}" == refs/pull/* ]] && [ "${{ github.event.pull_request.head.repo.full_name }}" != "${{ github.event.pull_request.base.repo.full_name }}" ]; then
+            fpr="yes"
+          fi
+          echo "::set-output name=foreign_pr::${fpr}"
+          echo "::set-output name=head_tag::${tag}"
+
+  test:
+    name: Test on Ruby ${{ matrix.ruby }} ${{ matrix.os }}
+    needs: prepare
+    runs-on: ${{ matrix.os }}
+    continue-on-error: false
+    strategy:
+      fail-fast: false
+      matrix:
+        ruby: [ '2.6', '2.7', '3.0', '3.1' ]
+        os: [ ubuntu-latest, windows-latest, macos-latest ]
+
+    steps:
+      - uses: actions/checkout@v3
+
+      - uses: ruby/setup-ruby@v1
+        with:
+          ruby-version: ${{ matrix.ruby }}
+          bundler-cache: true
+
+      - run: bundle exec rspec
+
+  release:
+    name: Release gem
+    needs: test
+    runs-on: ubuntu-latest
+    if: contains(github.ref, 'refs/tags/v')
+    steps:
+      - uses: actions/checkout@v3
+
+      - uses: cadwallion/publish-rubygems-action@master
+        env:
+          RUBYGEMS_API_KEY: ${{secrets.FONTIST_CI_RUBYGEMS_API_KEY}}

data/Gemfile

@@ -7,6 +7,6 @@ gemspec
 gem "gem-release"
 gem "rake", "~> 13.0"
 gem "rspec", "~> 3.2"
-gem "rubocop", "0.75.0"
+gem "rubocop"
 gem "rubocop-performance"
 gem "rubocop-rails"

data/README.adoc

@@ -61,6 +61,13 @@ end
 $ excavate --recursive path/to/archive.cab
 ----
 
+It supports recursive extraction of a directory containing archives:
+
+[source,sh]
+----
+$ excavate --recursive path/to/dir_with_archives
+----
+
 If you'd like to skip extraction of nested archives, just use:
 
 [source,sh]
@@ -68,6 +75,26 @@ If you'd like to skip extraction of nested archives, just use:
 $ excavate path/to/archive.cab
 ----
 
+To extract a particular file or files specify them as last arguments:
+
+[source,sh]
+----
+$ excavate --recursive archive.cab file1 dir/file2
+----
+
+Also `excavate` supports extraction from nested archives:
+
+[source,sh]
+----
+$ excavate --recursive archive.cab dir/nested.zip/file
+----
+
+And filtering:
+
+[source,sh]
+----
+$ excavate archive.cab --filter "**/specialfile*.txt"
+----
 
 == Dependencies
 

data/excavate.gemspec

@@ -1,5 +1,4 @@
 # frozen_string_literal: true
-
 require_relative "lib/excavate/version"
 
 Gem::Specification.new do |spec|
@@ -31,6 +30,6 @@ Gem::Specification.new do |spec|
   spec.add_runtime_dependency "libmspack", "~> 0.1"
   spec.add_runtime_dependency "ruby-ole", "~> 1.0"
   spec.add_runtime_dependency "rubyzip", "~> 2.3"
-  spec.add_runtime_dependency "seven_zip_ruby", "~> 1.0"
+  spec.add_runtime_dependency "seven-zip", "~> 1.4"
   spec.add_runtime_dependency "thor", "~> 1.0"
 end

data/lib/excavate/archive.rb

@@ -1,5 +1,8 @@
 module Excavate
   class Archive
+    INVALID_MEMORY_MESSAGE =
+      "invalid memory read at address=0x0000000000000000".freeze
+
     TYPES = { "cab" => Extractors::CabExtractor,
               "cpio" => Extractors::CpioExtractor,
               "exe" => Extractors::SevenZipExtractor,
@@ -14,9 +17,10 @@ module Excavate
       @archive = archive
     end
 
-    def files(recursive_packages: false)
+    def files(recursive_packages: false, files: [], filter: nil)
       target = Dir.mktmpdir
-      extract(target, recursive_packages: recursive_packages)
+      extract(target, recursive_packages: recursive_packages,
+                      files: files, filter: filter)
 
       all_files_in(target).map do |file|
         yield file
@@ -25,10 +29,105 @@ module Excavate
       FileUtils.rm_rf(target)
     end
 
-    def extract(target = nil, recursive_packages: false)
+    def extract(target = nil,
+                recursive_packages: false,
+                files: [],
+                filter: nil)
+      if files.size.positive?
+        extract_particular_files(target, files,
+                                 recursive_packages: recursive_packages)
+      elsif filter
+        extract_by_filter(target, filter,
+                          recursive_packages: recursive_packages)
+      else
+        extract_all(target, recursive_packages: recursive_packages)
+      end
+    end
+
+    private
+
+    def extract_particular_files(target, files, recursive_packages: false)
+      tmp = Dir.mktmpdir
+      extract_all(tmp, recursive_packages: recursive_packages)
+      found_files = find_files(tmp, files)
+      copy_files(found_files, target || Dir.pwd)
+    ensure
+      FileUtils.rm_rf(tmp)
+    end
+
+    def copy_files(files, target)
+      files.map do |file|
+        FileUtils.mkdir_p(target)
+        target_path = File.join(target, File.basename(file))
+        ensure_not_exist(target_path)
+
+        FileUtils.cp(file, target_path)
+
+        target_path
+      end
+    end
+
+    def ensure_not_exist(path)
+      if File.exist?(path)
+        type = File.directory?(path) ? "directory" : "file"
+        raise(TargetExistsError,
+              "Target #{type} `#{File.basename(path)}` already exists.")
+      end
+    end
+
+    def find_files(source, files)
+      all_files = all_files_in(source)
+
+      files.map do |target_file|
+        found_file = all_files.find do |source_file|
+          file_matches?(source_file, target_file, source)
+        end
+
+        unless found_file
+          raise(TargetNotFoundError, "File `#{target_file}` not found.")
+        end
+
+        found_file
+      end
+    end
+
+    def file_matches?(source_file, target_file, source_dir)
+      base_path(source_file, source_dir) == target_file
+    end
+
+    def base_path(path, prefix)
+      path.sub(prefix, "").sub(/^\//, "").sub(/^\\/, "")
+    end
+
+    def extract_by_filter(target, filter, recursive_packages: false)
+      tmp = Dir.mktmpdir
+      extract_all(tmp, recursive_packages: recursive_packages)
+      found_files = find_by_filter(tmp, filter)
+      copy_files(found_files, target || Dir.pwd)
+    end
+
+    def find_by_filter(source, filter)
+      all_files = all_files_in(source)
+
+      found_files = all_files.select do |source_file|
+        file_matches_filter?(source_file, filter, source)
+      end
+
+      if found_files.empty?
+        raise(TargetNotFoundError, "Filter `#{filter}` matched no file.")
+      end
+
+      found_files
+    end
+
+    def file_matches_filter?(source_file, filter, source_dir)
+      File.fnmatch?(filter, base_path(source_file, source_dir))
+    end
+
+    def extract_all(target, recursive_packages: false)
       source = File.expand_path(@archive)
       target ||= default_target(source)
-      raise(TargetNotEmptyError, "Target directory `#{File.basename(target)}` is not empty.") unless Dir.empty?(target)
+      ensure_empty(target)
 
       if recursive_packages
         extract_recursively(source, target)
@@ -39,11 +138,16 @@ module Excavate
       target
     end
 
-    private
+    def ensure_empty(path)
+      unless Dir.empty?(path)
+        raise(TargetNotEmptyError,
+              "Target directory `#{File.basename(path)}` is not empty.")
+      end
+    end
 
     def default_target(source)
       target = File.expand_path(File.basename(source, ".*"))
-      raise(TargetExistsError, "Target directory `#{File.basename(target)}` already exists.") if File.exist?(target)
+      ensure_not_exist(target)
 
       FileUtils.mkdir(target)
 
@@ -51,7 +155,7 @@ module Excavate
     end
 
     def extract_recursively(archive, target)
-      extract_once(archive, target)
+      extract_to_directory(archive, target)
 
       all_files_in(target).each do |file|
         next unless archive?(file)
@@ -60,14 +164,39 @@ module Excavate
       end
     end
 
+    def extract_to_directory(archive, target)
+      if File.directory?(archive)
+        duplicate_dir(archive, target)
+      elsif !archive?(archive)
+        copy_file(archive, target)
+      else
+        extract_once(archive, target)
+      end
+    end
+
+    def duplicate_dir(source, target)
+      Dir.chdir(source) do
+        (Dir.entries(".") - [".", ".."]).each do |entry|
+          FileUtils.cp_r(entry, target)
+        end
+      end
+    end
+
+    def copy_file(archive, target)
+      FileUtils.cp(archive, target)
+    end
+
     def extract_once(archive, target)
       extension = normalized_extension(archive)
       extractor_class = TYPES[extension]
-      raise(UnknownArchiveError, "Could not unarchive `#{archive}`.") unless extractor_class
+      unless extractor_class
+        raise(UnknownArchiveError, "Could not unarchive `#{archive}`.")
+      end
 
       extractor_class.new(archive).extract(target)
     rescue StandardError => e
-      raise unless extension == "exe" && e.message.start_with?("Invalid file format")
+      raise unless extension == "exe" &&
+        e.message.start_with?("Invalid file format")
 
       Extractors::CabExtractor.new(archive).extract(target)
     end
@@ -81,7 +210,7 @@ module Excavate
     rescue FFI::NullPointerError => e
       FileUtils.rmdir(target)
       raise unless normalized_extension(archive) == "exe" &&
-        e.message.start_with?("invalid memory read at address=0x0000000000000000")
+        e.message.start_with?(INVALID_MEMORY_MESSAGE)
     end
 
     def normalized_extension(file)

data/lib/excavate/cli.rb

@@ -8,6 +8,13 @@ module Excavate
     STATUS_UNKNOWN_ERROR = 1
     STATUS_TARGET_EXISTS = 2
     STATUS_TARGET_NOT_EMPTY = 3
+    STATUS_TARGET_NOT_FOUND = 4
+
+    ERROR_TO_STATUS = {
+      TargetExistsError => STATUS_TARGET_EXISTS,
+      TargetNotEmptyError => STATUS_TARGET_NOT_EMPTY,
+      TargetNotFoundError => STATUS_TARGET_NOT_FOUND,
+    }.freeze
 
     def self.exit_on_failure?
       false
@@ -23,15 +30,22 @@ module Excavate
       super(args, config)
     end
 
-    desc "extract ARCHIVE", "Extract ARCHIVE to a new directory"
-    option :recursive, aliases: :r, type: :boolean, default: false, desc: "Also extract all nested archives."
-    def extract(archive)
-      target = Excavate::Archive.new(archive).extract(recursive_packages: options[:recursive])
-      success("Successfully extracted to #{File.basename(target)}/")
-    rescue TargetExistsError => e
-      error(e.message, STATUS_TARGET_EXISTS)
-    rescue TargetNotEmptyError => e
-      error(e.message, STATUS_TARGET_NOT_EMPTY)
+    desc "extract ARCHIVE [FILE...]",
+         "Extract FILE or all files from ARCHIVE to a new directory"
+    option :recursive, aliases: :r, type: :boolean, default: false,
+                       desc: "Also extract all nested archives."
+    option :filter, type: :string,
+                    desc: "Filter by pattern (supports **, *, ?, etc)"
+    def extract(archive, *files)
+      target = Excavate::Archive.new(archive).extract(
+        recursive_packages: options[:recursive],
+        files: files,
+        filter: options[:filter],
+      )
+
+      success("Successfully extracted to #{format_paths(target)}")
+    rescue Error => e
+      handle_error(e)
     end
     default_task :extract
 
@@ -42,9 +56,24 @@ module Excavate
       STATUS_SUCCESS
     end
 
+    def handle_error(exception)
+      status = ERROR_TO_STATUS[exception.class]
+      raise exception unless status
+
+      error(exception.message, status)
+    end
+
     def error(message, status)
       say(message, :red)
       status
     end
+
+    def format_paths(path_or_paths)
+      paths = Array(path_or_paths).map do |x|
+        File.directory?(x) ? "#{File.basename(x)}/" : File.basename(x)
+      end
+
+      paths.join(", ")
+    end
   end
 end

data/lib/excavate/extractors/ole_extractor.rb

@@ -15,6 +15,8 @@ module Excavate
 
         Ole::Storage.open(@archive) do |ole|
           children(ole).each do |file|
+            next if ole.file.directory?(file)
+
             filename = prepare_filename(file)
             path = File.join(target, filename)
             content = ole.file.read(file)

data/lib/excavate/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Excavate
-  VERSION = "0.2.4"
+  VERSION = "0.3.1"
 end

data/lib/excavate.rb

@@ -8,7 +8,12 @@ require_relative "excavate/utils"
 
 module Excavate
   class Error < StandardError; end
-  class UnknownArchiveError < Error; end
+
   class TargetExistsError < Error; end
+
   class TargetNotEmptyError < Error; end
+
+  class TargetNotFoundError < Error; end
+
+  class UnknownArchiveError < Error; end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: excavate
 version: !ruby/object:Gem::Version
-  version: 0.2.4
+  version: 0.3.1
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2021-04-02 00:00:00.000000000 Z
+date: 2022-08-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: arr-pm
@@ -81,19 +81,19 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '2.3'
 - !ruby/object:Gem::Dependency
-  name: seven_zip_ruby
+  name: seven-zip
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.0'
+        version: '1.4'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.0'
+        version: '1.4'
 - !ruby/object:Gem::Dependency
   name: thor
   requirement: !ruby/object:Gem::Requirement
@@ -116,8 +116,8 @@ executables:
 extensions: []
 extra_rdoc_files: []
 files:
-- ".github/workflows/release.yml"
-- ".github/workflows/rspec.yml"
+- ".github/workflows/metanorma.yml"
+- ".github/workflows/test-and-release.yml"
 - ".gitignore"
 - ".rspec"
 - ".rubocop.yml"
@@ -168,7 +168,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.0.3
+rubygems_version: 3.3.7
 signing_key: 
 specification_version: 4
 summary: Extract nested archives with a single command.

data/.github/workflows/release.yml

@@ -1,36 +0,0 @@
-name: release
-
-on:
-  push:
-    tags:
-      - 'v*'
-
-jobs:
-  release:
-    runs-on: ubuntu-18.04
-    steps:
-      - uses: actions/checkout@v1
-
-      - uses: ruby/setup-ruby@v1
-        with:
-          ruby-version: '2.6'
-
-      - run: bundle config set path 'vendor/bundle'
-
-      - run: bundle install --jobs 4 --retry 3
-
-      - run: bundle exec rspec
-
-      - name: Publish to rubygems.org
-        env:
-          RUBYGEMS_API_KEY: ${{secrets.FONTIST_CI_RUBYGEMS_API_KEY}}
-        run: |
-          gem install gem-release
-          touch ~/.gem/credentials
-          cat > ~/.gem/credentials << EOF
-          ---
-          :rubygems_api_key: ${RUBYGEMS_API_KEY}
-          EOF
-          chmod 0600 ~/.gem/credentials
-          git status
-          gem release

data/.github/workflows/rspec.yml

@@ -1,47 +0,0 @@
-name: rspec
-
-on:
-  push:
-    branches: [ master, main ]
-  pull_request:
-
-jobs:
-  build:
-    name: Test on Ruby ${{ matrix.ruby }} ${{ matrix.os }}
-    runs-on: ${{ matrix.os }}
-    continue-on-error: ${{ matrix.experimental }}
-    strategy:
-      fail-fast: false
-      matrix:
-        ruby: [ '2.4', '2.5', '2.6', '2.7' ]
-        os: [ ubuntu-latest, windows-latest, macos-latest ]
-        experimental: [ false ]
-        include:
-          - ruby: '3.0'
-            os: 'ubuntu-latest'
-            experimental: true
-          - ruby: '3.0'
-            os: 'windows-latest'
-            experimental: true
-          - ruby: '3.0'
-            os: 'macos-latest'
-            experimental: true
-
-    steps:
-      - uses: actions/checkout@master
-
-      - uses: ruby/setup-ruby@v1
-        with:
-          ruby-version: ${{ matrix.ruby }}
-
-      - uses: actions/cache@v1
-        with:
-          path: vendor/bundle
-          key: bundle-${{ matrix.os }}-${{ matrix.ruby }}-${{ hashFiles('**/*.gemspec') }}
-          restore-keys: bundle-${{ matrix.os }}-${{ matrix.ruby }}
-
-      - run: bundle config set path 'vendor/bundle'
-
-      - run: bundle install --jobs 4 --retry 3
-
-      - run: bundle exec rspec