refactor(builtin): default end to self.length() on view constructors

[Yu-zh]

What

The view/slice constructors in builtin all declared end? : Int with no default and then re-derived the missing case inside the body, each with a slightly different idiom:

let end = match end { Some(end) | (None with end = len) => end }
let end_offset = if end_offset is Some(o) { o } else { self.length() }
match end { None => fixed[start:], Some(e) => fixed[start:e] }

A default-argument expression can reference an earlier parameter, so end? : Int = self.length() states the same thing in the signature. The body then works with a plain Int and the unwrapping goes away — 26 functions, −62 lines net.

Covered types: ArrayView, MutArrayView, BytesView, StringView, ReadOnlyArray, UninitializedArray (plus the deprecated String::substring).

Why it is behavior-preserving

• len-based cases — the default is literally the len the body already computed.
• StringView::{get_view,sub} used self.end() in the None branch. self.start() + self.length() is the same value, since StringView::length() is defined as end() - start().
• {String,StringView}::clamped_view clamped the Some branch into [0, len]. self.length() is already in range, so it clamps to itself.
• ReadOnlyArray::view fell back to fixed[start:], whose own default is the fixed array's length — and ReadOnlyArray::length() is defined as exactly that.

moon info produces no .mbti diff: the generated interface renders both forms as end? : Int, so this is not a source- or interface-breaking change for downstream users.

Not converted

• Iter::view genuinely needs the Option: an iterator has no length, and None selects a drop-only fast path that Some cannot express.
• {Array,FixedArray}::fill take a range but are not view constructors, so they are left alone here.

Verification

moon check --deny-warn --target all          # clean
moon info --target wasm,wasm-gc,js,native    # no diff
moon fmt                                     # no diff
moon bundle --all                            # clean
moon test --target all                       # wasm 6966 / wasm-gc 6967 / js 6911 / native 6885, 0 failed
moon test --release --target js,wasm,wasm-gc # 0 failed
moon test --release --target native          # 6885 passed, 0 failed
MOONC_INTERNAL_PARAMS='use_js_builtin_string = 1 |' \
  moon test --target wasm-gc [--release]     # 6967 passed, 0 failed

[coveralls]

Coverage Report for CI Build 5891

Coverage decreased (-0.02%) to 90.447%

Details

• Coverage decreased (-0.02%) from the base build.
• Patch coverage: 6 of 6 lines across 2 files are fully covered (100%).
• No coverage regressions found.

Uncovered Changes

No uncovered changes found.

Coverage Regressions

No coverage regressions found.

---

Coverage Stats

Relevant Lines:	17440
Covered Lines:	15774
Line Coverage:	90.45%
Coverage Strength:	132021.41 hits per line

---

💛 - Coveralls

[bobzhang]

what's the motivation? for easy inlining?
Note this could introduce some duplicated computation

[Yu-zh]

duplicated computations are caused by the wrappers and I plan to inline them as well (if the size of default values are reasonably small)