cmd/go: install "malformed file path" for source code filenames outside the Unicode BMP

[nigeltao]

As reported at google/wuffs#62 (comment)

$ go install github.com/google/wuffs@v0.3.0-beta.12
go: downloading github.com/google/wuffs v0.3.0-beta.12
go install: github.com/google/wuffs@v0.3.0-beta.12: create zip: test/data/non-ascii/😻.txt: malformed file path "test/data/non-ascii/😻.txt": invalid char '😻'

The github.com/google/wuffs repository indeed has a file named test/data/non-ascii/😻.txt and e.g. git clone seems OK with it.

😻 is U+1F63B SMILING CAT FACE WITH HEART-SHAPED EYES (in UTF-8: [F0 9F 98 BB]), outside of the Unicode Basic Multilingual Plane (BMP).

If I'm reading the ZIP specification correctly, section "4.6.9 -Info-ZIP Unicode Path Extra Field (0x7075)" suggests that it's possible for ZIP files to contain (UTF-8 encoded) Unicode file names.

[ALTree]

I think the names of files in modules have additional restrictions, not everything that is permitted by the zip specification is allowed.

https://go.dev/ref/mod#zip-path-size-constraints

File and directory names within a module may consist of Unicode letters, ASCII digits, the ASCII space character (U+0020), and the ASCII punctuation characters !#$%&()+,-.=@[]^_{}~.

[seankhliao]

Duplicate of #45549

[catundercar]

I think the names of files in modules have additional restrictions, not everything that is permitted by the zip specification is allowed.

https://go.dev/ref/mod#zip-path-size-constraints

File and directory names within a module may consist of Unicode letters, ASCII digits, the ASCII space character (U+0020), and the ASCII punctuation characters !#$%&()+,-.=@[]^_{}~.

Hello, I have a small question: If we do not check other files which do not affect compilation in the module, would it affect we get and use the module?