Bugfix: whitespace in object name

[HJLebbink]

The url_decode function now correctly decodes "a+b%2Bc" into "a b+c" instead of the incorrect "a+b+c".
Every function that uses either url_encode or url_decode now has dedicated tests to ensure correct handling of whitespace.

[Copilot]

Pull Request Overview

This PR fixes handling of object names containing whitespace or UTF-8 characters by switching to rand_object_name_utf8, refactoring common test logic into helper functions, and updating URL encoding/decoding utilities and their usage throughout the S3 client.

• Replaced rand_object_name with rand_object_name_utf8 in tests and added tests for names with spaces.
• Refactored repetitive test cases into generic test_* helper functions.
• Introduced url_decode, url_decode_old, and url_encode in src/s3/utils.rs and updated all urlencode/urldecode calls in the client.
• Added form-based decoding logic (url_decode_w_enc) in list_objects response parsing.

Reviewed Changes

Copilot reviewed 13 out of 13 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
tests/test_upload_download_object.rs	Use UTF-8–safe random names and whitespace test
tests/test_object_delete.rs	Refactor delete tests and add whitespace case
tests/test_object_copy.rs	Refactor copy tests and add whitespace case
tests/test_list_objects.rs	Refactor list tests and add whitespace case
tests/test_get_object.rs	Refactor get tests and add whitespace case
tests/test_bucket_create_delete.rs	Refactor bucket tests and add whitespace case
src/s3/utils.rs	Add url_decode_old, url_decode, url_encode
src/s3/response/list_objects.rs	Swap to url_decode_w_enc and remove old decode
src/s3/multimap.rs	Replace urlencode with url_encode
src/s3/client/delete_bucket.rs	Early return when bucket doesn't exist
src/s3/builders/put_object.rs	Replace urlencode with url_encode
src/s3/builders/copy_object.rs	Replace urlencode with url_encode
common/src/cleanup_guard.rs	Comment out debug output

Comments suppressed due to low confidence (5)

tests/test_list_objects.rs:162

• [nitpick] The comment contains a typo: it should say yielding "a+b%2Bc" instead of "a+b2Bc", and clarify the clause about percent-decoding.

/// In percent-encoding, "a b+c" becomes "a%20b%2Bc", but some S3 implementations may do

src/s3/utils.rs:52

• [nitpick] Replace the TODO placeholder with a concise description of what this function does and when to use it.

/// Decodes a URL-encoded string, TODO

src/s3/utils.rs:57

• [nitpick] Provide details on whether this function performs percent-decoding or form-decoding and how it differs from url_decode_old.

/// Decodes a URL-encoded string, TODO

src/s3/utils.rs:65

• [nitpick] Clarify in the doc comment whether this uses percent-encoding and reference relevant RFC or behavior.

/// Encodes a string using URL encoding, TODO

src/s3/utils.rs:53

• Add unit tests for url_decode_old, url_decode, and url_encode to verify correct handling of spaces, plus signs, and percent-encoded sequences.

pub fn url_decode_old(s: &str) -> Result<Cow<str>, FromUtf8Error> {