feat: replace numeric subtyping with explicit modes 🔢 - #210

Merged

timfennis merged 31 commits into

feat/explicit-numeric-modes

Sep 7, 2026

Commits on Sep 3, 2026

feat: replace numeric subtyping with explicit modes 🔢
timfennis
committed
6cc3b23
fix: highlight advanced number literals everywhere 🎨
timfennis
committed
100a554
refactor(lexer): reject bare bigint literals 🚫
timfennis
committed
2126915
fix(stdlib): restore transcendental function docs 📚
timfennis
committed
15122be
refactor(core): store Number integers as BigInt 🧹
timfennis
committed
33a4f21
perf(vm): compare primitive numerics directly ⚡
timfennis
committed
662385d
perf(stdlib): prioritize homogeneous numeric overloads 🏎️
timfennis
committed
62d982f
perf(stdlib): fold borrowed numeric collections 🧮
timfennis
committed
c99f512
fix(bench): use Number for bigint benchmark 🔢
timfennis
committed
fc59d75
fix(stdlib): add explicit numeric overloads for randf and randi 🎲

timfennis
and
claude
committed
43a40c0
test(lsp): adapt completion grouping to sibling numeric types 🧪

timfennis
and
claude
committed
d7e4713
test: cast into typed collections for numeric dispatch 🎯

timfennis
and
claude
committed
322b50e
docs(manual): explain typed-collection dispatch and its cast 📚

timfennis
and
claude
committed
d553248
refactor: centralize numeric runtime handling 🔢
timfennis
committed
924b3f5
refactor(lexer): unify numeric literal handling 🔢
timfennis
committed
29e5650
Fix some performance degradation
timfennis
committed
b190841
perf(stdlib): read numeric operands straight from their slots ⚡

timfennis
and
claude
committed
847d1bd
chore: cleanup num code
timfennis
committed
123232b
chore: move number from core to vm
timfennis
committed
c4fc393
chore: improve documentation
timfennis
committed
888b9a0

Commits on Sep 5, 2026

fix(stdlib): keep i64::MIN remainders in range 🔢

timfennis
and
claude
committed
16ad243
fix(vm): accept Number floats from comparators ⚖️

timfennis
and
claude
committed
e1486ce
fix(macros): match the Number these params declare 🎭

timfennis
and
claude
committed
b563cf4
refactor(stdlib): give randi Int bounds only 🎲

timfennis
and
claude
committed
1d29e7f
fix(vm): repair exact-number results and hashing 🔢

timfennis
and
claude
committed
9b2e87a
fix(stdlib): report left shifts that overflow their Int result ⬅️

timfennis
and
claude
committed
fa503ca
fix(vm): say what is actually wrong with a rejected value 📏

timfennis
and
claude
committed
9578fd0
refactor(vm): drop a redundant range check 🧹

timfennis
and
claude
committed
e801fc1
docs(manual): spell out Number's zero-divisor result 📝

timfennis
and
claude
committed
2a6a1af

Commits on Sep 7, 2026

fix(vm): normalize every exact fraction, not just the ones from division 🔢

timfennis
and
claude
committed
46b2e3e
refactor(vm): make a whole fraction unrepresentable 🔒

timfennis
and
claude
committed
10751c1