Merge branch 'develop'

Author: tsolomko

.gitignore

@@ -78,3 +78,6 @@ docs/
 
 # Vscode launch.json generated by Swift extension
 .vscode/launch.json
+
+# API baselines generate by swift package diagnose-api-breaking-changes
+api_baseline/

.jazzy.yaml

@@ -3,11 +3,11 @@ sourcekitten_sourcefile: docs.json
 clean: false
 author: Timofey Solomko
 module: SWCompression
-module_version: 4.8.2
+module_version: 4.8.3
 copyright: '© 2022 Timofey Solomko'
 readme: README.md
 github_url: https://github.com/tsolomko/SWCompression
-github_file_prefix: https://github.com/tsolomko/SWCompression/tree/4.8.2
+github_file_prefix: https://github.com/tsolomko/SWCompression/tree/4.8.3
 theme: fullwidth
 
 custom_categories:

CHANGELOG.md

@@ -1,5 +1,28 @@
 # Changelog
 
+## 4.8.3
+
+- There are now minimum deployment targets specified in Swift Package Manager manifest.
+- The properties of `GzipHeader` are now `var`-properties (instead of `let`).
+- GZip extra fields are now supported.
+    - Added `GzipHeader.ExtraField` struct.
+    - Added `GzipHeader.extraFields` property.
+    - Added a new `extraFields` argument to `GzipArchive.archive` function (with a default array empty value).
+- Fixed potential crashes that could occur when processing GZip archives truncated in a header or a "footer".
+- Some non-well-formed values of PAX extended header records no longer cause `TarError.wrongPaxHeaderEntry` to be thrown.
+    - The record values with newline characters are now fully processed.
+    - The record values that do not contain UTF-8 strings are now ignored.
+- swcomp changes:
+    - The symbolic links are now extracted with the values recorded in the containers.
+    - The hard links are now extracted from TAR containers instead of being ignored.
+    - Fixed build issues on Linux and Windows.
+    - `benchmark` is now a command group with two commands, `run` and `show`.
+    - Added `-a`, `--append` option to the `benchmark run` command.
+    - Added `-d`, `--description` option to the `benchmark run` command.
+    - Added `-t`, `--preserve-timestamp` option to the `benchmark run` command.
+    - The file format of saved results is now more flexible and allows multi-way comparisons.
+    - Improved precision of time measurements in benchmarks.
+
 ## 4.8.2
 
 - Swift 5.1 is no longer supported.

Package.swift

@@ -3,6 +3,12 @@ import PackageDescription
 
 let package = Package(
     name: "SWCompression",
+    platforms: [
+        .macOS(.v10_13),
+        .iOS(.v11),
+        .tvOS(.v11),
+        .watchOS(.v4)
+    ],
     products: [
         .library(
             name: "SWCompression",

README.md

@@ -171,6 +171,8 @@ Every function or type of SWCompression's public API is documented. This documen
 There is a small command-line program, "swcomp", which is included in this repository in "Sources/swcomp". It can be
 built using Swift Package Manager.
 
+__IMPORTANT:__ The "swcomp" command-line tool is NOT intended for general use.
+
 ## Contributing
 
 Whether you find a bug, have a suggestion, idea, feedback or something else, please

SWCompression.podspec

@@ -1,7 +1,7 @@
 Pod::Spec.new do |s|
 
   s.name         = "SWCompression"
-  s.version      = "4.8.2"
+  s.version      = "4.8.3"
   s.summary      = "A framework with functions for working with compression, archives and containers."
 
   s.description  = "A framework with (de)compression algorithms and functions for processing various archives and containers."

SWCompression.xcodeproj/SWCompression.plist

@@ -15,9 +15,9 @@
 	<key>CFBundlePackageType</key>
 	<string>FMWK</string>
 	<key>CFBundleShortVersionString</key>
-	<string>4.8.2</string>
+	<string>4.8.3</string>
 	<key>CFBundleVersion</key>
-	<string>87</string>
+	<string>88</string>
 	<key>NSHumanReadableCopyright</key>
 	<string>Copyright © 2022 Timofey Solomko</string>
 </dict>

SWCompression.xcodeproj/TestSWCompression.plist

@@ -15,8 +15,8 @@
 	<key>CFBundlePackageType</key>
 	<string>BNDL</string>
 	<key>CFBundleShortVersionString</key>
-	<string>4.8.2</string>
+	<string>4.8.3</string>
 	<key>CFBundleVersion</key>
-	<string>87</string>
+	<string>88</string>
 </dict>
 </plist>

SWCompression.xcodeproj/project.pbxproj

@@ -213,6 +213,8 @@
 		06F066771FFB763400312A82 /* test8.bz2 in Resources */ = {isa = PBXBuildFile; fileRef = 06F066111FFB763300312A82 /* test8.bz2 */; };
 		06F276DF1F2BAB4A00E67335 /* 7zEntry.swift in Sources */ = {isa = PBXBuildFile; fileRef = 06F276DE1F2BAB4900E67335 /* 7zEntry.swift */; };
 		06FEAD921F54B9CD00AD016E /* EncodingTree.swift in Sources */ = {isa = PBXBuildFile; fileRef = 06FEAD911F54B9CD00AD016E /* EncodingTree.swift */; };
+		E6023B1B28F9B60000D6F3DC /* test4_extra_field.gz in Resources */ = {isa = PBXBuildFile; fileRef = E6023B1A28F9B60000D6F3DC /* test4_extra_field.gz */; };
+		E6023B1D28F9C92200D6F3DC /* GzipHeader+ExtraField.swift in Sources */ = {isa = PBXBuildFile; fileRef = E6023B1C28F9C92200D6F3DC /* GzipHeader+ExtraField.swift */; };
 		E604F3C22700C75F004BD38A /* test_dict_B5.lz4 in Resources */ = {isa = PBXBuildFile; fileRef = E604F3BF2700C75F004BD38A /* test_dict_B5.lz4 */; };
 		E604F3C32700C75F004BD38A /* test_dict_B5_BD.lz4 in Resources */ = {isa = PBXBuildFile; fileRef = E604F3C02700C75F004BD38A /* test_dict_B5_BD.lz4 */; };
 		E604F3C42700C75F004BD38A /* lz4_dict in Resources */ = {isa = PBXBuildFile; fileRef = E604F3C12700C75F004BD38A /* lz4_dict */; };
@@ -266,6 +268,7 @@
 		E6791E1826FD001A003852A9 /* DataError.swift in Sources */ = {isa = PBXBuildFile; fileRef = E6791E1726FD001A003852A9 /* DataError.swift */; };
 		E6791E1B26FD0094003852A9 /* LZ4.swift in Sources */ = {isa = PBXBuildFile; fileRef = E6791E1A26FD0094003852A9 /* LZ4.swift */; };
 		E6791E3326FD05EC003852A9 /* LZ4Tests.swift in Sources */ = {isa = PBXBuildFile; fileRef = E6791E3226FD05EC003852A9 /* LZ4Tests.swift */; };
+		E68DA766290420BC00259CB4 /* test_pax_record_newline.tar in Resources */ = {isa = PBXBuildFile; fileRef = E68DA765290420BC00259CB4 /* test_pax_record_newline.tar */; };
 		E694694327480EA6009C897A /* TarReader.swift in Sources */ = {isa = PBXBuildFile; fileRef = E694694227480EA6009C897A /* TarReader.swift */; };
 		E6974C5B2701AC2600E06C60 /* test_dict_B5_dictID.lz4 in Resources */ = {isa = PBXBuildFile; fileRef = E6974C5A2701AC2600E06C60 /* test_dict_B5_dictID.lz4 */; };
 		E69FAC922729ACD900D3C406 /* test_dos_latin_us.zip in Resources */ = {isa = PBXBuildFile; fileRef = E69FAC912729ACD900D3C406 /* test_dos_latin_us.zip */; };
@@ -497,6 +500,8 @@
 		06F276DE1F2BAB4900E67335 /* 7zEntry.swift */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.swift; path = 7zEntry.swift; sourceTree = "<group>"; };
 		06FEAD911F54B9CD00AD016E /* EncodingTree.swift */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.swift; name = EncodingTree.swift; path = Sources/Common/CodingTree/EncodingTree.swift; sourceTree = SOURCE_ROOT; };
 		06FED40B1DD7717E0013DFB2 /* BZip2.swift */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.swift; path = BZip2.swift; sourceTree = "<group>"; };
+		E6023B1A28F9B60000D6F3DC /* test4_extra_field.gz */ = {isa = PBXFileReference; lastKnownFileType = archive.gzip; path = test4_extra_field.gz; sourceTree = "<group>"; };
+		E6023B1C28F9C92200D6F3DC /* GzipHeader+ExtraField.swift */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.swift; path = "GzipHeader+ExtraField.swift"; sourceTree = "<group>"; };
 		E604F3BF2700C75F004BD38A /* test_dict_B5.lz4 */ = {isa = PBXFileReference; lastKnownFileType = file; path = test_dict_B5.lz4; sourceTree = "<group>"; };
 		E604F3C02700C75F004BD38A /* test_dict_B5_BD.lz4 */ = {isa = PBXFileReference; lastKnownFileType = file; path = test_dict_B5_BD.lz4; sourceTree = "<group>"; };
 		E604F3C12700C75F004BD38A /* lz4_dict */ = {isa = PBXFileReference; lastKnownFileType = file; path = lz4_dict; sourceTree = "<group>"; };
@@ -551,6 +556,7 @@
 		E6791E1726FD001A003852A9 /* DataError.swift */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.swift; path = DataError.swift; sourceTree = "<group>"; };
 		E6791E1A26FD0094003852A9 /* LZ4.swift */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.swift; path = LZ4.swift; sourceTree = "<group>"; };
 		E6791E3226FD05EC003852A9 /* LZ4Tests.swift */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.swift; path = LZ4Tests.swift; sourceTree = "<group>"; };
+		E68DA765290420BC00259CB4 /* test_pax_record_newline.tar */ = {isa = PBXFileReference; lastKnownFileType = archive.tar; path = test_pax_record_newline.tar; sourceTree = "<group>"; };
 		E694694227480EA6009C897A /* TarReader.swift */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.swift; path = TarReader.swift; sourceTree = "<group>"; };
 		E6974C5A2701AC2600E06C60 /* test_dict_B5_dictID.lz4 */ = {isa = PBXFileReference; lastKnownFileType = file; path = test_dict_B5_dictID.lz4; sourceTree = "<group>"; };
 		E69FAC912729ACD900D3C406 /* test_dos_latin_us.zip */ = {isa = PBXFileReference; lastKnownFileType = archive.zip; path = test_dos_latin_us.zip; sourceTree = "<group>"; };
@@ -714,6 +720,7 @@
 			children = (
 				063364E21DC52979007E313F /* GzipArchive.swift */,
 				061C062A1F0E8A1D00832F0C /* GzipHeader.swift */,
+				E6023B1C28F9C92200D6F3DC /* GzipHeader+ExtraField.swift */,
 				061C06251F0E8A0300832F0C /* GzipError.swift */,
 				0686A63F1FA4C18800E89C9E /* FileSystemType+Gzip.swift */,
 			);
@@ -875,6 +882,7 @@
 				06F065BF1FFB763300312A82 /* test2.gz */,
 				06F065BC1FFB763300312A82 /* test3.gz */,
 				06F065BB1FFB763300312A82 /* test4.gz */,
+				E6023B1A28F9B60000D6F3DC /* test4_extra_field.gz */,
 				06F065B91FFB763300312A82 /* test5.gz */,
 				06F065BE1FFB763300312A82 /* test6.gz */,
 				06F065BD1FFB763300312A82 /* test7.gz */,
@@ -953,6 +961,7 @@
 				0698B10A2104E11200A7C551 /* test_gnu_inc_format.tar */,
 				0698B10C2106136500A7C551 /* test_big_num_field.tar */,
 				0698B10E2106344200A7C551 /* test_negative_mtime.tar */,
+				E68DA765290420BC00259CB4 /* test_pax_record_newline.tar */,
 			);
 			path = TAR;
 			sourceTree = "<group>";
@@ -1213,6 +1222,7 @@
 				E652FEE027009BDD006BC312 /* test5_legacy.lz4 in Resources */,
 				06F066721FFB763400312A82 /* test7.bz2 in Resources */,
 				06F0665B1FFB763400312A82 /* test10.lzma in Resources */,
+				E68DA766290420BC00259CB4 /* test_pax_record_newline.tar in Resources */,
 				E652FEFE2700A028006BC312 /* test_B7_BD.lz4 in Resources */,
 				06F0663C1FFB763400312A82 /* test3.xz in Resources */,
 				06F066591FFB763400312A82 /* test_win.tar in Resources */,
@@ -1278,6 +1288,7 @@
 				E652FED6270091B6006BC312 /* test_skippable_frame.lz4 in Resources */,
 				06F0665A1FFB763400312A82 /* test11.lzma in Resources */,
 				E652FEE427009BDD006BC312 /* test3_legacy.lz4 in Resources */,
+				E6023B1B28F9B60000D6F3DC /* test4_extra_field.gz in Resources */,
 				06F066401FFB763400312A82 /* test_zip_bzip2.zip in Resources */,
 				06F066461FFB763400312A82 /* test_empty_file.zip in Resources */,
 				06F066331FFB763400312A82 /* random_file.zlib in Resources */,
@@ -1406,6 +1417,7 @@
 				06A9606A1F1E7E0D0078E6D1 /* 7zSubstreamInfo.swift in Sources */,
 				E6C4150726FE230A00F9D36F /* XxHash32.swift in Sources */,
 				0694A74D1F7C0DF00023BC99 /* BurrowsWheeler.swift in Sources */,
+				E6023B1D28F9C92200D6F3DC /* GzipHeader+ExtraField.swift in Sources */,
 				06CC3FDD1F8AAE8B00BD576D /* MsbBitReader+7z.swift in Sources */,
 				06A3933B1DE0709300182E12 /* Deflate.swift in Sources */,
 				06092A1A1FA4CB0300DE9FD5 /* CompressionMethod.swift in Sources */,
@@ -1516,7 +1528,7 @@
 				CLANG_WARN_SUSPICIOUS_MOVE = YES;
 				CLANG_WARN_UNREACHABLE_CODE = YES;
 				CLANG_WARN__DUPLICATE_METHOD_MATCH = YES;
-				CURRENT_PROJECT_VERSION = 87;
+				CURRENT_PROJECT_VERSION = 88;
 				DEAD_CODE_STRIPPING = YES;
 				DEBUG_INFORMATION_FORMAT = dwarf;
 				EAGER_LINKING = YES;
@@ -1601,7 +1613,7 @@
 				CLANG_WARN_SUSPICIOUS_MOVE = YES;
 				CLANG_WARN_UNREACHABLE_CODE = YES;
 				CLANG_WARN__DUPLICATE_METHOD_MATCH = YES;
-				CURRENT_PROJECT_VERSION = 87;
+				CURRENT_PROJECT_VERSION = 88;
 				DEAD_CODE_STRIPPING = YES;
 				DEBUG_INFORMATION_FORMAT = "dwarf-with-dsym";
 				EAGER_LINKING = YES;
@@ -1666,7 +1678,7 @@
 				APPLICATION_EXTENSION_API_ONLY = YES;
 				DEFINES_MODULE = YES;
 				DYLIB_COMPATIBILITY_VERSION = 1;
-				DYLIB_CURRENT_VERSION = 87;
+				DYLIB_CURRENT_VERSION = 88;
 				DYLIB_INSTALL_NAME_BASE = "@rpath";
 				INFOPLIST_FILE = SWCompression.xcodeproj/SWCompression.plist;
 				INSTALL_PATH = "$(LOCAL_LIBRARY_DIR)/Frameworks";
@@ -1693,7 +1705,7 @@
 				APPLICATION_EXTENSION_API_ONLY = YES;
 				DEFINES_MODULE = YES;
 				DYLIB_COMPATIBILITY_VERSION = 1;
-				DYLIB_CURRENT_VERSION = 87;
+				DYLIB_CURRENT_VERSION = 88;
 				DYLIB_INSTALL_NAME_BASE = "@rpath";
 				INFOPLIST_FILE = SWCompression.xcodeproj/SWCompression.plist;
 				INSTALL_PATH = "$(LOCAL_LIBRARY_DIR)/Frameworks";

Sources/GZip/GzipArchive.swift

@@ -78,15 +78,18 @@ public class GzipArchive: Archive {
 
     private static func processMember(_ bitReader: LsbBitReader) throws -> Member {
         // Valid GZip archive must contain at least 20 bytes of data (10 for the header, 2 for an empty Deflate block,
-        // and 8 for checksums).
-        guard bitReader.bitsLeft >= 20 * 8
+        // and 8 for checksums). In addition, since GZip format is "byte-oriented" we should ensure that members are
+        // byte-aligned.
+        guard bitReader.isAligned && bitReader.bytesLeft >= 20
             else { throw GzipError.wrongMagic }
 
         let header = try GzipHeader(bitReader)
 
         let memberData = try Deflate.decompress(bitReader)
         bitReader.align()
 
+        guard bitReader.bytesLeft >= 8
+            else { throw GzipError.wrongMagic }
         let crc32 = bitReader.uint32()
         let isize = bitReader.uint64(fromBytes: 4)
         guard UInt64(truncatingIfNeeded: memberData.count) % (UInt64(truncatingIfNeeded: 1) << 32) == isize
@@ -110,14 +113,20 @@ public class GzipArchive: Archive {
      - Parameter isTextFile: Set to true, if the file which will be archived is text file or ASCII-file.
      - Parameter osType: Type of the system on which this archive will be created.
      - Parameter modificationTime: Last time the file was modified.
+     - Parameter extraFields: Any extra fields. Note that no extra field is allowed to have second byte of the extra
+     field (subfield) ID equal to zero. In addition, the length of a field's binary content must be less than
+     `UInt16.max`, while the total sum of the binary content length of all extra fields plus 4 for each field must also
+     not exceed `UInt16.max`. See GZip format specification for more details.
 
-     - Throws: `GzipError.cannotEncodeISOLatin1` if file name of comment cannot be encoded with ISO-Latin-1 encoding.
+     - Throws: `GzipError.cannotEncodeISOLatin1` if a file name or a comment cannot be encoded with ISO-Latin-1 encoding
+     or if the total sum of the binary content length of all extra fields plus 4 for each field exceeds `UInt16.max`.
 
      - Returns: Resulting archive's data.
      */
     public static func archive(data: Data, comment: String? = nil, fileName: String? = nil,
                                writeHeaderCRC: Bool = false, isTextFile: Bool = false,
-                               osType: FileSystemType? = nil, modificationTime: Date? = nil) throws -> Data {
+                               osType: FileSystemType? = nil, modificationTime: Date? = nil,
+                               extraFields: [GzipHeader.ExtraField] = []) throws -> Data {
         var flags: UInt8 = 0
 
         var commentData = Data()
@@ -146,6 +155,10 @@ public class GzipArchive: Archive {
             }
         }
 
+        if !extraFields.isEmpty {
+            flags |= 1 << 2
+        }
+
         if writeHeaderCRC {
             flags |= 1 << 1
         }
@@ -167,14 +180,35 @@ public class GzipArchive: Archive {
         var headerBytes: [UInt8] = [
             0x1f, 0x8b, // 'magic' bytes.
             8, // Compression method (DEFLATE).
-            flags // Flags; currently no flags are set.
+            flags
         ]
         for i in 0..<4 {
             headerBytes.append(mtimeBytes[i])
         }
         headerBytes.append(2) // Extra flags; 2 means that DEFLATE used slowest algorithm.
         headerBytes.append(os)
 
+        if !extraFields.isEmpty {
+            let xlen = extraFields.reduce(0) { $0 + 4 + $1.bytes.count }
+            guard xlen <= UInt16.max
+                else { throw GzipError.cannotEncodeISOLatin1 }
+            headerBytes.append((xlen & 0xFF).toUInt8())
+            headerBytes.append(((xlen >> 8) & 0xFF).toUInt8())
+
+            for extraField in extraFields {
+                headerBytes.append(extraField.si1)
+                headerBytes.append(extraField.si2)
+
+                let len = extraField.bytes.count
+                headerBytes.append((len & 0xFF).toUInt8())
+                headerBytes.append(((len >> 8) & 0xFF).toUInt8())
+
+                for byte in extraField.bytes {
+                    headerBytes.append(byte)
+                }
+            }
+        }
+
         var outData = Data(headerBytes)
 
         outData.append(fileNameData)