Make opinionated judgements about string/symbol decoding

[Shaptic]

What

The conversion behavior of strings and symbols has changed:

• before: scvString -> Uint8Array OR string
• after: scvString -> decoded as UTF-8, returns string
• before: scvSymbol -> Uint8Array OR string
• after: scvString -> always decoded as an ascii string

Why

The following is true:

• symbols MUST be strings, as their input is enforced to be [a-zA-Z0-9_]+
• strings are "bytes with a hint of being text," so we should respect that

Previously, users were seeing byte arrays when they expected strings, so this rectifies that unintuitive behavior.

Known Limitations

UTF-8 encoding is assumed for scvStrings (because it's ascii-compatible), but there's not really a compelling reason to assume anything else (utf8 is assumed throughout most of the toolchain, and if you want something different you can always [a] request a feature or [b] use raw scvBytes)

[github-actions]

Size Change: +2.36 kB (0%)

Total Size: 2.98 MB

Filename	Size	Change
dist/stellar-base.js	2.18 MB	+1.87 kB (0%)
dist/stellar-base.min.js	793 kB	+489 B (0%)

_{compressed-size-action}