[matt-edmondson]

Requested by Matt · project thread

Fixes #378, fixes #368, fixes #379, fixes #356, fixes #365, fixes #469, fixes #470, fixes #468, fixes #357, fixes #337, fixes #465, fixes #377.

Chord symbols

• Chord.Parse rejects "C+7" and "C+13" and reads "C+9"/"C+11" as a major triad with #9/#11 (regression from #362) #378: a + at the start of the body is the augmented mark before a number too. C+7 is Caug7, C+9 is Caug9, Cm+7 is Cm7#5. C+5 still reads as a raised fifth, and C7+9 is still C7#9.
• Chord.Parse rejects Cmaj7sus4 / Cmaj9sus4 / Cmaj7sus2, yet Chord.ToString() prints exactly those symbols, so a Progression holding one cannot be re-parsed #368: a suspension can follow a major seventh. Cmaj7sus4, Cmaj9sus4 and CM7sus2 parse, and the writer no longer treats them as inexpressible, so the exhaustive round-trip test now covers them.
• Chord.Parse rejects lead-sheet parenthesised spellings (Cm(maj7), C7(9), C7(13), Cm7(11)), so charts using them fail Progression.Parse #379: inside parentheses a bare 9/11/13 is an added tone (C7(13) is C7add13, C7(b9,13) is C7b9add13). A major-seventh mark supplies a missing seventh (Cm(maj7) is Cmmaj7). C7(9,9), C9(9) and C7(maj7) are still refused. The README grammar is updated.
• Chord.Voice(octave, inversion) puts the root back in the bass from the 4th inversion of extended chords: C9 inversion 4 gives C5 D5 E5 G5 A#5 #356: Voice(octave, inversion) raises each tone below the target bass by as many octaves as it takes, so the 4th to 6th inversions of C9/C13 put the ninth, eleventh or thirteenth lowest. The root comes from MIDI arithmetic. A voicing outside MIDI 0..127 throws one ArgumentOutOfRangeException(octave).

Roman numerals

• Key.RomanNumeralOf drops sixths and tensions: C6 → "I", G13 → "V7", C7b9 → "I7", even though ChordFromRomanNumeral parses "I6", "V13" and "V7b9" #469: RomanNumeralOf writes the chord's whole body, so C6 → I6, G13 → V13, C7b9 → I7b9, Cmaj7sus4 → Imaj7sus4. The slash bass is dropped (documented), because I6 already means an added sixth.
• Key.ChordFromRomanNumeral reads b/# after an upper-case numeral as the root's accidental ("Vb9" → F#9, "Ib5" → B5) #365: ChordFromRomanNumeral reads the suffix as a chord body against the root it has already resolved (a new internal ChordSymbolReader.TryReadBody), so Vb9 is G with a flat nine rather than F#9, and Ib5 is C major-flat-five.
• A new test checks that ChordFromRomanNumeral(RomanNumeralOf(c)) == c for 46 chord shapes on every degree of C major and C minor.

Analysis

• Roman numerals, functions and cadences use the mode's own degree index in pentatonic and hexatonic keys: in C major pentatonic, G is "IV" and G → C is reported as a Plagal cadence #470: in a mode with fewer than seven degrees, roman numerals, functions and cadences (and ChordFromRomanNumeral) are spelled against major or natural minor on the same tonic, chosen by the mode's third. G in C major pentatonic is V/Dominant and G → C is Authentic. C in A minor pentatonic is III. Modes with more than seven degrees are unchanged.
• Progression.Cadences reports a Half cadence for every V→V pair: "4/4 C | G | G | G" yields three half cadences #357: V → V is no longer a half cadence, so a held dominant gives one half cadence where it arrives.
• ChromaticChords reports a minor key's own dominant (E / E7 in A minor) as "borrowed from parallel major" #337: a minor key's own V/V7 (major triad or dominant seventh on degree 5, with the raised leading tone) is diatonic to ChromaticChords. A key counts as minor when its scale has a minor third and a minor seventh. Fm in C is still borrowed, and D7 is still V/V.
• ChromaticChords labels a dominant of a diminished triad as a secondary dominant: F#7 in C major is reported as "V/vii°", F# in A minor as "V/ii°" #465: only a major or minor triad can be a secondary-dominant target. F#7 in C and F# in A minor are no longer V/vii°/V/ii°. The existing row that expected V/III+ in A harmonic minor was replaced with V/iv, as the issue asks.
• Progression.InferKey ignores sevenths and extensions: "C7 | Fmaj7 | C7" is inferred as C major, though C7's B♭ is not in C major #468: InferKeys multiplies the root/triad score by the fraction of chord tones, slash bass included, that are in the scale. A minor key's dominant counts its leading tone as diatonic. C7 | Fmaj7 | C7 and Fmaj7 | C7 now infer F major, and Bbmaj7 | F7 infers Bb major. The comment now matches the code for sus, power and altered-fifth chords (they score 1.0).
• ChromaticChords never reports ChromaticKind.AugmentedSixth; Italian, French and German sixths fall through to Chromatic #377: ChromaticKind.AugmentedSixth is now produced. A chord holds pitch classes, not spellings, so the variants are read by shape on the lowered sixth degree. Ab7no5 is It+6 and Ab7b5 is Fr+6. Ab7 is Ger+6 only when it resolves to the dominant or to a cadential six-four (tonic over the dominant). Otherwise it stays chromatic. The rule is documented on the enum member and in the README.

Testing

dotnet test Semantics.Test -c Release: 1836 total, 0 failed, 12 skipped. The new cases are in ChordGrammarTests, ChordRoundTripTests, InversionTransposeTests, RomanNumeralParseTests and the new AnalysisFixesTests.

---

Generated by Claude Code

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
96.6% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud