#typescript/src/float

#typescript/src/float

Strictcli canonical float form (SCF), the shortest-round-trip float formatting shared byte-for-byte with the Python and Go implementations. It is pinned by the committed conformance/float_vectors.json and matches go/strictcli/float.go formatFloatCanonical and Python _format_float_canonical exactly.

Rules: 1. Digit source is the shortest decimal string that round-trips to the identical IEEE-754 double. ECMAScript Number::toString is specified to produce exactly that, so no digit post-processing is needed. 2. Integer-valued floats in fixed notation always carry a trailing ".0". 3. -0.0 is preserved as "-0.0". 4. Fixed notation for |x| in [1e-6, 1e21); scientific outside. Zero is a carve-out: 0.0 and -0.0 are always fixed. ECMAScript's own notation switch matches this band exactly: toString uses exponential notation precisely when the decimal exponent is >= 21 or <= -7, and a shortest round-trip digit string never crosses the band boundary (it would have to round-trip to a double on the other side of 1e21 / 1e-6, which are exactly representable comparison points in both siblings). 5. Scientific spelling: lowercase "e", explicit exponent sign, no exponent zero-padding ("1e+21", "1e-7", "1.5e+300") -- ECMAScript's native exponential spelling, verbatim. 6. The trailing ".0" is only ever added in the fixed branch, never in the scientific branch.

#formatFloatCanonical

export function formatFloatCanonical(v: number): string

Strictcli canonical float form (SCF), the shortest-round-trip float formatting shared byte-for-byte with the Python and Go implementations. It is pinned by the committed conformance/float_vectors.json and matches go/strictcli/float.go formatFloatCanonical and Python _format_float_canonical exactly.

Rules: 1. Digit source is the shortest decimal string that round-trips to the identical IEEE-754 double. ECMAScript Number::toString is specified to produce exactly that, so no digit post-processing is needed. 2. Integer-valued floats in fixed notation always carry a trailing ".0". 3. -0.0 is preserved as "-0.0". 4. Fixed notation for |x| in [1e-6, 1e21); scientific outside. Zero is a carve-out: 0.0 and -0.0 are always fixed. ECMAScript's own notation switch matches this band exactly: toString uses exponential notation precisely when the decimal exponent is >= 21 or <= -7, and a shortest round-trip digit string never crosses the band boundary (it would have to round-trip to a double on the other side of 1e21 / 1e-6, which are exactly representable comparison points in both siblings). 5. Scientific spelling: lowercase "e", explicit exponent sign, no exponent zero-padding ("1e+21", "1e-7", "1.5e+300") -- ECMAScript's native exponential spelling, verbatim. 6. The trailing ".0" is only ever added in the fixed branch, never in the scientific branch.