[release/7.0] Tar: Remove invalidation of whitespace in PAX extended …

…attributes (#78785)

* Remove invalidation of whitespace in TryGetNextExtendedAttribute (#78465)

* Remove invalidation of whitespace in TryGetNextExtendedAttribute

* Add requested test

* Remove TrimStart in PAX extended attributes (#78707)

* Remove TrimStart in PAX extended attributes

* Adjust existing tests with two extra inline datas.

Co-authored-by: carlossanlop <carlossanlop@users.noreply.github.com>

* Tar: Extra tests to confirm extra long paths are not treated as duplicate entries when the full path is in the extended attributes. (#78744)

Co-authored-by: carlossanlop <carlossanlop@users.noreply.github.com>

Co-authored-by: Stephen Toub <stoub@microsoft.com>
Co-authored-by: carlossanlop <carlossanlop@users.noreply.github.com>

src/libraries/System.Formats.Tar/src/System/Formats/Tar/TarHeader.Read.cs

@@ -723,13 +723,7 @@ private static bool TryGetNextExtendedAttribute(
             {
                 return false;
             }
-            line = line.Slice(spacePos + 1).TrimStart((byte)' ');
-
-            // If there are any more spaces, it's malformed.
-            if (line.IndexOf((byte)' ') >= 0)
-            {
-                return false;
-            }
+            line = line.Slice(spacePos + 1);
 
             // Find the equal separator.
             int equalPos = line.IndexOf((byte)'=');

src/libraries/System.Formats.Tar/tests/TarFile/TarFile.ExtractToDirectory.Stream.Tests.cs

@@ -177,5 +177,32 @@ public void Extract_UnseekableStream_BlockAlignmentPadding_DoesNotAffectNextEntr
 
             Assert.Equal(2, Directory.GetFileSystemEntries(destination.Path, "*", SearchOption.AllDirectories).Count());
         }
+
+        [Fact]
+        public void PaxNameCollision_DedupInExtendedAttributes()
+        {
+            using TempDirectory root = new();
+
+            string sharedRootFolders = Path.Join(root.Path, "folder with spaces", new string('a', 100));
+            string path1 = Path.Join(sharedRootFolders, "entry 1 with spaces.txt");
+            string path2 = Path.Join(sharedRootFolders, "entry 2 with spaces.txt");
+
+            using MemoryStream stream = new();
+            using (TarWriter writer = new(stream, TarEntryFormat.Pax, leaveOpen: true))
+            {
+                // Paths don't fit in the standard 'name' field, but they differ in the filename,
+                // which is fully stored as an extended attribute
+                PaxTarEntry entry1 = new(TarEntryType.RegularFile, path1);
+                writer.WriteEntry(entry1);
+                PaxTarEntry entry2 = new(TarEntryType.RegularFile, path2);
+                writer.WriteEntry(entry2);
+            }
+            stream.Position = 0;
+
+            TarFile.ExtractToDirectory(stream, root.Path, overwriteFiles: true);
+
+            Assert.True(File.Exists(path1));
+            Assert.True(Path.Exists(path2));
+        }
     }
 }

src/libraries/System.Formats.Tar/tests/TarFile/TarFile.ExtractToDirectoryAsync.Stream.Tests.cs

@@ -242,5 +242,32 @@ public async Task Extract_UnseekableStream_BlockAlignmentPadding_DoesNotAffectNe
 
             Assert.Equal(2, Directory.GetFileSystemEntries(destination.Path, "*", SearchOption.AllDirectories).Count());
         }
+
+        [Fact]
+        public async Task PaxNameCollision_DedupInExtendedAttributesAsync()
+        {
+            using TempDirectory root = new();
+
+            string sharedRootFolders = Path.Join(root.Path, "folder with spaces", new string('a', 100));
+            string path1 = Path.Join(sharedRootFolders, "entry 1 with spaces.txt");
+            string path2 = Path.Join(sharedRootFolders, "entry 2 with spaces.txt");
+
+            await using MemoryStream stream = new();
+            await using (TarWriter writer = new(stream, TarEntryFormat.Pax, leaveOpen: true))
+            {
+                // Paths don't fit in the standard 'name' field, but they differ in the filename,
+                // which is fully stored as an extended attribute
+                PaxTarEntry entry1 = new(TarEntryType.RegularFile, path1);
+                await writer.WriteEntryAsync(entry1);
+                PaxTarEntry entry2 = new(TarEntryType.RegularFile, path2);
+                await writer.WriteEntryAsync(entry2);
+            }
+            stream.Position = 0;
+
+            await TarFile.ExtractToDirectoryAsync(stream, root.Path, overwriteFiles: true);
+
+            Assert.True(File.Exists(path1));
+            Assert.True(Path.Exists(path2));
+        }
     }
 }

src/libraries/System.Formats.Tar/tests/TarReader/TarReader.File.GlobalExtendedAttributes.Tests.cs

@@ -82,5 +82,29 @@ public void ExtractGlobalExtendedAttributesEntry_Throws()
                 Assert.Throws<InvalidOperationException>(() => entry.ExtractToFile(Path.Join(root.Path, "file"), overwrite: true));
             }
         }
+
+        [Theory]
+        [InlineData("key", "value")]
+        [InlineData("key    ", "value    ")]
+        [InlineData("    key", "    value")]
+        [InlineData("    key   ", "    value    ")]
+        [InlineData("    key spaced   ", "    value spaced    ")]
+        [InlineData("many sla/s\\hes", "/////////////\\\\\\///////////")]
+        public void GlobalExtendedAttribute_Roundtrips(string key, string value)
+        {
+            var stream = new MemoryStream();
+            using (var writer = new TarWriter(stream, leaveOpen: true))
+            {
+                writer.WriteEntry(new PaxGlobalExtendedAttributesTarEntry(new Dictionary<string, string>() { { key, value } }));
+            }
+
+            stream.Position = 0;
+            using (var reader = new TarReader(stream))
+            {
+                PaxGlobalExtendedAttributesTarEntry entry = Assert.IsType<PaxGlobalExtendedAttributesTarEntry>(reader.GetNextEntry());
+                Assert.Equal(1, entry.GlobalExtendedAttributes.Count);
+                Assert.Equal(KeyValuePair.Create(key, value), entry.GlobalExtendedAttributes.First());
+            }
+        }
     }
 }

src/libraries/System.Formats.Tar/tests/TarReader/TarReader.File.Tests.cs

@@ -325,6 +325,30 @@ public void PaxSizeLargerThanMaxAllowedByStream()
             Assert.Throws<ArgumentOutOfRangeException>(() => reader.GetNextEntry());
         }
 
+        [Theory]
+        [InlineData("key", "value")]
+        [InlineData("key    ", "value    ")]
+        [InlineData("    key", "    value")]
+        [InlineData("    key   ", "    value    ")]
+        [InlineData("    key spaced   ", "    value spaced    ")]
+        [InlineData("many sla/s\\hes", "/////////////\\\\\\///////////")]
+        public void PaxExtendedAttribute_Roundtrips(string key, string value)
+        {
+            var stream = new MemoryStream();
+            using (var writer = new TarWriter(stream, leaveOpen: true))
+            {
+                writer.WriteEntry(new PaxTarEntry(TarEntryType.Directory, "entryName", new Dictionary<string, string>() { { key, value } }));
+            }
+
+            stream.Position = 0;
+            using (var reader = new TarReader(stream))
+            {
+                PaxTarEntry entry = Assert.IsType<PaxTarEntry>(reader.GetNextEntry());
+                Assert.Equal(5, entry.ExtendedAttributes.Count);
+                Assert.Contains(KeyValuePair.Create(key, value), entry.ExtendedAttributes);
+            }
+        }
+
         private static void VerifyDataStreamOfTarUncompressedInternal(string testFolderName, string testCaseName, bool copyData)
         {
             using MemoryStream archiveStream = GetTarMemoryStream(CompressionMethod.Uncompressed, testFolderName, testCaseName);