excavate 1.0.1 → 1.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: d579e821dbcdc3891ec86ccc385b815735aec447f7aaccf6980430002b77e009
-  data.tar.gz: 409103ecdac9190c965a2e9414dc507b02e9a0654dcc7c1eeab01eec018ce916
+  metadata.gz: ed5cec17e4c8c985ae34d044fa170ccbecdc3bd783a0cd0916454a8c567149d7
+  data.tar.gz: 750dcde7f88158d2e107763a5fc733c4a72ac1030b53e91c6fd7e2697b167686
 SHA512:
-  metadata.gz: 2431a0910c6d163b790dddcd5745129017848aca78093b935156ef8ef0ff4d0c4cf8ecbc8c6fb1b8eafcc29078556236632b9e0e9884c5c5794b24073829bbe0
-  data.tar.gz: 4450fa82f329350881d0ea3f25f2396ec40c3c34af5523753d59f7a3540a3ea5dde9bcdb77bd8873f6ab5064f4d8beeb1b81bb05db41368a16a598a970669de8
+  metadata.gz: b62877409f75d68041b7b973ee364356e6e47b09bd714ebd4430f0df627643213bd8ff2dc3ca40d4b8d141c2fda3ff6b9b686f4720fc8997424b169effff169e
+  data.tar.gz: e3b0dee5b195329ecbc6cc85ffc5fc64c7b4c736481ac3c964b0353828c59e0176b71ad3c70ba53b7503a01419f5e6f205e28fd3531fe3584449ad8b39a66082

data/.rubocop_todo.yml

@@ -1,6 +1,6 @@
 # This configuration was generated by
 # `rubocop --auto-gen-config`
-# on 2026-02-25 10:37:11 UTC using RuboCop version 1.84.2.
+# on 2026-03-17 11:57:02 UTC using RuboCop version 1.85.1.
 # The point is for the user to remove these configuration records
 # one by one as the offenses are removed from the code base.
 # Note that changes in the inspected code, or installation of new
@@ -37,16 +37,10 @@ Lint/DuplicateBranch:
   Exclude:
     - 'lib/excavate/utils.rb'
 
-# Offense count: 1
-# This cop supports safe autocorrection (--autocorrect).
-Lint/ScriptPermission:
-  Exclude:
-    - 'scripts/create_multi_nested_fixture.rb'
-
-# Offense count: 5
+# Offense count: 4
 # Configuration parameters: CountComments, CountAsOne, AllowedMethods, AllowedPatterns.
 Metrics/MethodLength:
-  Max: 15
+  Max: 12
 
 # Offense count: 46
 # Configuration parameters: Prefixes, AllowedPatterns.
@@ -71,6 +65,16 @@ RSpec/MultipleExpectations:
 RSpec/NestedGroups:
   Max: 4
 
+# Offense count: 2
+# This cop supports safe autocorrection (--autocorrect).
+# Configuration parameters: AllowOnlyRestArgument, UseAnonymousForwarding, RedundantRestArgumentNames, RedundantKeywordRestArgumentNames, RedundantBlockArgumentNames.
+# RedundantRestArgumentNames: args, arguments
+# RedundantKeywordRestArgumentNames: kwargs, options, opts
+# RedundantBlockArgumentNames: blk, block, proc
+Style/ArgumentsForwarding:
+  Exclude:
+    - 'lib/excavate/archive.rb'
+
 # Offense count: 2
 # This cop supports unsafe autocorrection (--autocorrect-all).
 Style/IdenticalConditionalBranches:

data/lib/excavate/archive.rb

@@ -18,7 +18,7 @@ module Excavate
       @archive = archive
     end
 
-    def files(recursive_packages: false, files: [], filter: nil, &block)
+    def files(recursive_packages: false, files: [], filter: nil, &)
       # Auto-enable recursive_packages when extracting specific files
       recursive_packages = true if files.any?
 
@@ -26,7 +26,7 @@ module Excavate
       extract(target, recursive_packages: recursive_packages,
                       files: files, filter: filter)
 
-      all_files_in(target).map(&block)
+      all_files_in(target).map(&)
     ensure
       windows_safe_rm_rf(target)
     end
@@ -218,7 +218,11 @@ module Excavate
       replace_archive_with_contents(archive, target)
     rescue StandardError
       FileUtils.rm_rf(target)
-      raise unless normalized_extension(archive) == "exe"
+      # During recursive extraction of nested archives, silently skip
+      # any that fail (e.g. .msi files that aren't real OLE, .cab files
+      # with incompatible format, .exe files with unsupported compression).
+      # Only re-raise for file types we don't recognize as archives.
+      raise unless TYPES.key?(normalized_extension(archive))
     end
 
     def replace_archive_with_contents(archive, target)

data/lib/excavate/extractors/ole_extractor.rb

@@ -22,6 +22,8 @@ module Excavate
 
         Omnizip::Formats::Ole.open(@archive) do |ole|
           children(ole).each do |entry|
+            next if ole.directory?(entry)
+
             path = File.join(target, prepare_filename(entry))
             FileUtils.mkdir_p(File.dirname(path))
             content = ole.read(entry)

data/lib/excavate/extractors/xz_extractor.rb

@@ -28,20 +28,37 @@ module Excavate
       end
 
       def extract_tar_xz(target)
-        # Decompress XZ to get gzip data
-        gzip_data = Omnizip::Formats::Xz.decompress(@archive)
-        # Decompress gzip to get tar data
-        tar_data = Zlib::GzipReader.new(StringIO.new(gzip_data)).read
+        data = Omnizip::Formats::Xz.decompress(@archive)
+        data = strip_compression(data)
+        validate_tar!(data)
 
         # Write tar file and extract
         temp_tar = File.join(target, ".temp_#{Time.now.to_i}_#{rand(1000)}.tar")
-        File.binwrite(temp_tar, tar_data)
+        File.binwrite(temp_tar, data)
 
         TarExtractor.new(temp_tar).extract(target)
       ensure
         File.delete(temp_tar) if temp_tar && File.exist?(temp_tar)
       end
 
+      def strip_compression(data)
+        if FileMagic.detect_bytes(data) == :gzip
+          return Zlib::GzipReader.new(StringIO.new(data)).read
+        end
+
+        data
+      end
+
+      def validate_tar!(data)
+        inner_type = FileMagic.detect_bytes(data)
+        return if inner_type == :tar
+
+        inner_type ||= "unknown format"
+
+        raise UnknownArchiveError,
+              "Expected tar inside #{@archive}, got #{inner_type}"
+      end
+
       def extract_pure_xz(target)
         # Decompress XZ
         data = Omnizip::Formats::Xz.decompress(@archive)

data/lib/excavate/file_magic.rb

@@ -1,29 +1,32 @@
+# frozen_string_literal: true
+
 module Excavate
   class FileMagic
+    # [offset, magic_bytes, type]
+    SIGNATURES = [
+      [0, "MSCF\x00\x00\x00\x00".b, :cab],
+      [0, "\xFD7zXZ\x00".b, :xz],
+      [0, "\x1F\x8B".b, :gzip],
+      [257, "ustar".b, :tar],
+    ].freeze
+
+    MAX_READ = SIGNATURES.map { |o, m, _| o + m.bytesize }.max
+
     def self.detect(path)
-      new(path).detect
+      beginning = File.read(path, MAX_READ, mode: "rb")
+      detect_bytes(beginning)
     end
 
-    def initialize(path)
-      @path = path
-    end
+    def self.detect_bytes(data)
+      return nil if data.nil? || data.empty?
+
+      SIGNATURES.each do |offset, magic, type|
+        next if data.bytesize < offset + magic.bytesize
 
-    def detect
-      beginning = File.read(@path, 8, mode: "rb")
-      case beginning
-      when "MSCF\x00\x00\x00\x00".force_encoding("BINARY")
-        :cab
-      else
-        case beginning.byteslice(0, 6)
-        when "\xFD7zXZ\x00".force_encoding("BINARY")
-          :xz
-        else
-          case beginning.byteslice(0, 2)
-          when "\x1F\x8B".force_encoding("BINARY")
-            :gzip
-          end
-        end
+        return type if data.byteslice(offset, magic.bytesize) == magic
       end
+
+      nil
     end
   end
 end

data/lib/excavate/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Excavate
-  VERSION = "1.0.1"
+  VERSION = "1.0.2"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: excavate
 version: !ruby/object:Gem::Version
-  version: 1.0.1
+  version: 1.0.2
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2026-02-25 00:00:00.000000000 Z
+date: 2026-03-17 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: cabriolet