Add tests for reading zip archives using a different codepage encoding

rikyoz · Nov 2, 2024 · f349dc4 · f349dc4
1 parent 26bf10f
commit f349dc4
Show file tree

Hide file tree

Showing 2 changed files with 34 additions and 1 deletion.
diff --git a/tests/CMakeLists.txt b/tests/CMakeLists.txt
@@ -54,7 +54,7 @@ target_include_directories( ${TESTS_BASE_TARGET} INTERFACE
 if( BIT7Z_TESTS_FILESYSTEM )
     CPMAddPackage( NAME bit7z_test_data
                    GITHUB_REPOSITORY "rikyoz/bit7z-test-data"
-                   VERSION 1.11
+                   VERSION 1.12
                    DOWNLOAD_ONLY YES
                    GIT_PROGRESS ON )
     if( bit7z_test_data_ADDED )

diff --git a/tests/src/test_bitinputarchive.cpp b/tests/src/test_bitinputarchive.cpp
@@ -28,6 +28,7 @@
 #include <bit7z/bitformat.hpp>
 #include <bit7z/bittypes.hpp>
 #include <internal/fs.hpp>
+#include <internal/stringutil.hpp>
 
 #include <algorithm>
 #include <cstddef>
@@ -1537,4 +1538,36 @@ TEMPLATE_TEST_CASE( "BitInputArchive: Reading a nested zip archive",
         REQUIRE_NOTHROW( reader.test() );
         REQUIRE( reader.contains( italy.name ) );
     }
+}
+
+TEMPLATE_TEST_CASE( "BitInputArchive: Reading a zip archive using a different encoding",
+                    "[bitinputarchive]", tstring, buffer_t, stream_t ) {
+    const TestDirectory testDir{ fs::path{ test_archives_dir } / "metadata" / "unicode" };
+
+    const fs::path arcFileName = "codepage.zip";
+
+    TestType inputArchive{};
+    getInputArchive( arcFileName, inputArchive );
+    const Bit7zLibrary lib{ test::sevenzip_lib_path() };
+    const BitArchiveReader reader{ lib, inputArchive, BitFormat::Zip };
+    REQUIRE( reader.itemsCount() == 1 );
+
+    constexpr auto expectedItemName = BIT7Z_STRING( "ユニコード.pdf" );
+
+    // The archive uses the Shift-JS encoding (Codepage 932) for the file names.
+    // If we do not set the codepage to be used, 7-Zip will report a wrongly-encoded string for the name.
+    REQUIRE_FALSE( reader.itemAt( 0 ).name() == expectedItemName );
+
+    // Setting the correct codepage will make 7-Zip correctly encode the string.
+    reader.useFormatProperty( L"cp", 932u );
+    REQUIRE( reader.itemAt( 0 ).name() == expectedItemName );
+
+    TempTestDirectory testOutDir{ "test_bitinputarchive" };
+    INFO( "Output directory: " << testOutDir )
+
+    REQUIRE_NOTHROW( reader.extractTo( testOutDir ) );
+
+    const auto extractedFilePath = tstring_to_path( expectedItemName );
+    REQUIRE( fs::exists( extractedFilePath ) );
+    REQUIRE( fs::remove( extractedFilePath ) );
 }