[Copilot]

Under the Kitty keyboard protocol, key events report the unshifted base character as the primary key code, carrying any shifted character in the alternate-key sub-parameter. KittyKeyboard.GetKeyCode previously unwound KeyEvent.Code only for Digit* and Key* prefixes, causing main-keyboard punctuation positions under Shift to fall through to ev.Key[0] and emit the shifted symbol as the primary key code.

Changes

• Punctuation key code unwinding (src/XTerm.NET/Input/KittyKeyboard.cs):

  • When ev.ShiftKey is set, map the 11 main-keyboard punctuation Code values (Backquote, Minus, Equal, BracketLeft, BracketRight, Backslash, Semicolon, Quote, Comma, Period, Slash) to their US-positional base characters (\``, -, =, [, ], `, ;, ', ,, ., /).
  • Sequenced after Digit* and Key* handlers to preserve hot-path performance for letter and digit inputs.

• Regression and end-to-end tests (src/XTerm.NET.Tests/Input/):

  • Added Shifted_punctuation_appears_as_a_sub_parameter covering all 11 punctuation keys with ReportAlternateKeys (AllKeys | Alternates).
  • Added coverage for ReportAllKeysAsEscapeCodes and Ctrl+Shift disambiguated chords.
  • Added Terminal.GenerateKittyKeyInput test asserting mode-aware unwinding through the terminal entry point.

// Shift+= (key: "+", code: "Equal")
// Before: CSI 43;2u    ('+' reported as key code)
// After:  CSI 61;2u    ('=' base key code)

// Ctrl+Shift+= under ReportAlternateKeys
// Before: CSI 43:43;6u (duplicated shifted symbol)
// After:  CSI 61:43;6u (base key 61 with shifted alternate 43)

• Fixes Kitty keyboard: Shift is unwound only for Digit* and Key* codes, so shifted punctuation reports the shifted character as the key code #165

[github-actions]

Perf comparison — this change, against its base

3 run(s) of each side, alternating on one machine. Allocation is a count and is gated exactly. Time is a measurement, so its gate is derived from the spread this job just observed in itself rather than fixed in advance.

corpus	bytes/char	gen0/Mchar	ns/char	Δ time	noise	gate
scroll-ascii	0.00 → 0.00	0.00 → 0.00	3.36 → 3.36	+0.1%	±1%	4%
sgr-churn	0.00 → 0.00	0.00 → 0.00	9.02 → 8.99	-0.3%	±39%	117%
truecolor	0.00 → 0.00	0.00 → 0.00	10.00 → 9.96	-0.4%	±0%	4%
alt-redraw	0.00 → 0.00	0.00 → 0.00	14.40 → 14.36	-0.3%	±1%	4%
unicode	7.66 → 7.66	0.45 → 0.45	35.31 → 36.14	+2.3%	±6%	18%
flood	0.00 → 0.00	0.00 → 0.00	94.58 → 94.44	-0.1%	±1%	4%

Each corpus is gated at max(4%, 3 × its own noise). A wide noise column means this runner was busy and the timing half of the table should be read as advisory; the allocation half is exact either way.

assemblies measured

• base: XTerm.NET 2.0.5.0 mvid:4e8fedf9-7edf-4be2-a232-82a728abd815
• head: XTerm.NET 2.0.5.0 mvid:55751f9c-4ba5-449d-8e92-4047946a145e

---

Perf comparison — cumulative, everything since 2.0.3

3 run(s) of each side, alternating on one machine. Allocation is a count and is gated exactly. Time is a measurement, so its gate is derived from the spread this job just observed in itself rather than fixed in advance.

corpus	bytes/char	gen0/Mchar	ns/char	Δ time	noise	gate
scroll-ascii	0.00 → 0.00	0.00 → 0.00	3.38 → 3.36	-0.4%	±1%	5%
sgr-churn	0.00 → 0.00	0.00 → 0.00	9.07 → 8.99	-0.9%	±39%	117%
truecolor	0.00 → 0.00	0.00 → 0.00	10.11 → 9.96	-1.5%	±1%	5%
alt-redraw	0.00 → 0.00	0.00 → 0.00	14.28 → 14.36	+0.6%	±3%	10%
unicode	7.66 → 7.66	0.45 → 0.45	35.33 → 36.14	+2.3%	±5%	15%
flood	0.00 → 0.00	0.00 → 0.00	94.63 → 94.44	-0.2%	±1%	5%

Each corpus is gated at max(5%, 3 × its own noise). A wide noise column means this runner was busy and the timing half of the table should be read as advisory; the allocation half is exact either way.

assemblies measured

• base: XTerm.NET 2.0.3.0 mvid:9ba0ce98-0d1e-4424-8aa1-9ab012cdc07a
• head: XTerm.NET 2.0.5.0 mvid:55751f9c-4ba5-449d-8e92-4047946a145e