Tar: Detect magic numbers in archives compressed with popular algorithms

[carlossanlop]

As suggested here: #74316 (comment)

Ideally, when reading a tar.xyz compressed file using TarFile or TarArchive APIs, we should detect the magic numbers at the beginning of the archive, determine the archive is compressed, and throw a friendlier error message. We currently throw Unable to parse number in ThrowInvalidNumber(). And we could offer two variations of the same exception:

• If we support the compression algorithm, suggest wrapping the archive with a stream class for that algorithm (DeflateStream, GZipStream, ZLibStream).
• If we don't support the compression algorithm, but it has magic numbers, detect them, mention the algorithm is currently unsupported.

Note: We have proposal #70517 to track the request to be able to open tar.gz files, since we support gzip. TarFile and TarArchive support the tar format only, no compression.

Tagging subscribers to this area: @dotnet/area-system-formats-tar
See info in area-owners.md if you want to be subscribed.

Issue Details

---

As suggested here: #74316 (comment)

Ideally, when reading a tar.xyz compressed file using TarFile or TarArchive APIs, we should detect the magic numbers at the beginning of the archive, determine the archive is compressed, and throw a friendlier error message. We currently throw Unable to parse number in ThrowInvalidNumber(). And we could offer two variations of the same exception:

• If we support the compression algorithm, suggest wrapping the archive with a stream class for that algorithm (DeflateStream, GZipStream, ZLibStream).
• If we don't support the compression algorithm, but it has magic numbers, detect them, mention the algorithm is currently unsupported.

Author:	carlossanlop
Assignees:	-
Labels:	area-System.Formats.Tar
Milestone:	Future