router: add MatchText to NormalizedInput and NormalizeMatchText function (slice 9a)

Add MatchText field to NormalizedInput — a lossy lexical matching view
derived from ingress text: TrimSpace → NFKC → lowercase → collapse
Unicode whitespace. Does NOT fold ё→е, strip punctuation, strip wake
words, rewrite numbers, or invoke morphology.

Both ingress sites (voice + text) construct MatchText at entry. No
existing consumer reads MatchText yet — it is dark data for future
opt-in migration.

Vikunja: #725

cmd/mavend/voice.go

@@ -211,7 +211,7 @@ func (h *reactiveHandler) HandlePushToTalk(ctx context.Context, req voice.PushTo
 
 	// 2-5. the shared turn pipeline (confirm → clarify → route → dialogue →
 	// action → replier), identical to the text path.
-	replyText := h.runTurn(ctx, router.NormalizedInput{Text: text, Source: sourceVoice})
+	replyText := h.runTurn(ctx, router.NormalizedInput{Text: text, MatchText: router.NormalizeMatchText(text), Source: sourceVoice})
 
 	// 6. tts — synthesise the reply text; return to the voice server which
 	// ships it back on the conn.
@@ -244,7 +244,7 @@ func (h *reactiveHandler) upgradeAPI(api ipc.CoreAPI) {
 // HandlePushToTalk so text channels share the same routing logic.
 func (h *reactiveHandler) handleText(ctx context.Context, conversation, text string) string {
 	log.Printf("voice: handleText: %q", text)
-	return h.runTurn(withDialogueID(ctx, dialogueIDFor(sourceText, conversation)), router.NormalizedInput{Text: text, Source: sourceText})
+	return h.runTurn(withDialogueID(ctx, dialogueIDFor(sourceText, conversation)), router.NormalizedInput{Text: text, MatchText: router.NormalizeMatchText(text), Source: sourceText})
 }
 
 // turnSource is a local alias for router.InputSource, kept so the daemon code

internal/router/matchtext.go

@@ -0,0 +1,32 @@
+package router
+
+import (
+	"strings"
+	"unicode"
+
+	"golang.org/x/text/unicode/norm"
+)
+
+// NormalizeMatchText derives a lossy lexical matching view from raw ingress text.
+// The result is intended ONLY for consumers that explicitly opt into
+// case-insensitive, whitespace-collapsed matching. It must not be used for
+// user-visible text, entity identity, quoted content, free-form arguments,
+// or persisted utterances without proving the transformation is safe for
+// that consumer.
+//
+// Pipeline: TrimSpace → Unicode NFKC → lowercase → collapse Unicode whitespace.
+// Does NOT fold ё→е, strip punctuation, strip wake words, rewrite numbers,
+// or invoke morphology.
+func NormalizeMatchText(in string) string {
+	out := strings.TrimSpace(in)
+	out = norm.NFKC.String(out)
+	out = strings.ToLower(out)
+	// Collapse runs of Unicode whitespace (including non-breaking space,
+	// thin space, ideographic space, etc.) to a single ASCII space.
+	out = strings.Join(strings.FieldsFunc(out, func(r rune) bool {
+		return unicode.IsSpace(r)
+	}), " ")
+	return out
+}
+
+

internal/router/source.go

@@ -92,10 +92,12 @@ const (
 )
 
 // NormalizedInput — the typed ingress boundary for a turn. Text is the raw
-// utterance after STT (voice) or as typed (text). Source identifies the
-// channel. This slice performs no new linguistic normalization: text and voice
-// paths continue to converge onto the same turn path as they did before.
+// utterance after STT (voice) or as typed (text). MatchText is a lossy
+// lexical matching view derived from Text: whitespace-collapsed, NFKC-
+// normalized, lowercased. Consumers must opt into MatchText individually;
+// nothing reads it by default in this slice. Source identifies the channel.
 type NormalizedInput struct {
-	Text   string
-	Source InputSource
+	Text      string // original ingress text — byte-for-byte the value current consumers receive
+	MatchText string // derived lossy representation for case-insensitive lexical matching
+	Source    InputSource
 }