[bfrancojr]

Fixes #133.

A value that doesn't fit its (sbe.encoding) primitive is now refused instead of losing its high bits. Before, int32 v [(sbe.encoding) = "int8"] set to 300 was written as 44 and read back as 44. On read, uint32 2^31 in an int32 field wrapped to a negative number. The draft says "Emitters MUST reject values that fall outside the narrowed type's range". Ported from protowire-go #255 (issue #252).

• Marshal reads the value by the field's kind and checks it against the primitive's range: sbe: field M.v: value 300 is outside the range of int8, [-128, 127]. A double under "float" is checked against ±3.4028234663852886e+38 (NaN and ±Inf pass). An enum, laid out as uint8 by default, refuses a value past 255 or below 0.
• Unmarshal sign-extends per the primitive, then checks the result against the field's range: sbe: field M.v: wire value 2147483648 (uint32) does not fit a int32 field, [-2147483648, 2147483647]. A bool still reads any nonzero value as true.
• Codec::New refuses (sbe.encoding) on a string or bytes field, and an integer/float crossing, with Go's three texts.
• View is unchanged, as in Go's commit.
• The ranges are computed per field from the descriptor at encode and decode time. Go precomputes them at bind time. Doing it per call avoids adding fields to the public FieldTemplate struct. Go's %g in the float messages is reproduced by a small GoG helper in the new internal src/sbe/encoding_range.h.

Also fixed (same code path): a double field under "float" used to write the low 4 bytes of the double's bits, and a float field under "double" the float's bits zero-extended. Both read back as garbage. They now convert the value, as Go does. Go's 100-pairing matrix test pins it.

Stack and merge order: #123 → #124 → #137 (#134) → #140 (#132) → this PR. Base: sbe-header-bounds-132. This PR builds on #132's rewrite of FieldEncodingSize and its Go-style error wrapping.

Evidence, at 0529a25, on macOS:

• ctest: 100% passed, 689 of 689. That includes 7 SbeEncodingRange.* tests ported from Go's encoding_range_test.go: Marshal, MarshalFloat, MarshalNested, Unmarshal, KindBind, and Matrix (68 pairings bind and round-trip, 32 are refused, the counts Go asserts). It also includes UnmarshalText, which pins the full read-error text.
• Mutation check: with the write and read checks forced off, Marshal, MarshalFloat, MarshalNested, Unmarshal and UnmarshalText fail.
• scripts/go_differential.sh against protowire-go main 207fdfc: exit 1, from the same 450 PXF diff lines as at sbe: tell two schemas' template ids apart (#134) #137 (cmp). Every section ran (954 damaged streams decoded alike, 18 SBE schemas converted alike, 362 documents), and no differing line mentions sbe. The remaining PXF diffs are pxf: Marshal leaves out an unset (pxf.default) or (pxf.required) field, so it doesn't read back #129, pxf: FormatDocument drops every directive, so a @dataset document formats to nothing #130 and pxf: reject null on a (pxf.required) or @required field #131.
• clang-format 18 --dry-run --Werror: clean.
• clang-tidy 23 on template.cc, marshal.cc and unmarshal.cc: every remaining hit is on a line this PR does not touch. Those lines pass CI's tidy-18.

[bfrancojr]

Rebased onto main e1c1b72 (#136 merged); head is now 448b8af. The PR's own diff is unchanged, checked by comparing its added and removed lines before and after the rebase. Rebased as part of the SBE stack; evidence is on #123.

[bfrancojr]

Rebased with the stack onto main df6b102; head is now 456dd6e. Evidence is on #124.

[bfrancojr]

Rebased onto main 197e05b, after #124 and #137 merged as squashes: 456dd6e → b945fdf. The old #124 and #137 commits were dropped (git rebase --onto origin/main bcedc19), so this PR's own commit is the only one replayed. Its diff is unchanged, apart from where its CHANGELOG line lands. At the stack top (39e1261): build OK, ctest 100% tests passed out of 705. The remote head was read back with git ls-remote: b945fdf.

[bfrancojr]

Rebased onto main 9525e9b, after #140 merged as a squash: b945fdf → 72c54f1. I dropped #140's old commit with git rebase --onto origin/main 3486d29. The diff is unchanged apart from where the CHANGELOG line lands. At the stack top (2ee007c): build OK, ctest 100% tests passed out of 705. The remote head was read back with git ls-remote: 72c54f1.