onadit 0.0.1 → 1.0.0

package/dist/src/retrieval/passages.js

@@ -0,0 +1,160 @@
+/**
+ * Query-anchored passages: the few spans of a long row that answer the task.
+ *
+ * Deterministic, no model call. Finds query-term matches (lowercase, accents
+ * folded), scores fixed windows around them by how many distinct query terms
+ * they contain (rarer-in-this-row terms count more), and returns the best
+ * non-overlapping windows with exact character offsets into the stored
+ * content.
+ *
+ * No speaker attribution. Captures are stored as one flat string
+ * ("User: …\n\nAssistant: …"), and a message can itself contain "\n\nUser: "
+ * (quoted or invented dialogue), so the stored text has no verified message
+ * boundaries. A passage is quoted evidence; who said it is not claimed.
+ */
+import { tokenize } from "./lexical.js";
+const DEFAULT_MAX = 2;
+const DEFAULT_WINDOW = 360;
+/** Bound work on multi-megabyte rows: positions kept per term. */
+const MAX_HITS_PER_TERM = 400;
+// Function words that match everywhere and carry no evidence. Terms shorter
+// than 3 characters are dropped unless they contain a digit (IDs like "16").
+const STOPWORDS = new Set([
+    "the", "and", "for", "with", "that", "this", "what", "which", "why", "how", "does", "did",
+    "was", "were", "are", "from", "into", "about", "should", "would", "could", "have", "has",
+    "our", "your", "you", "not", "but", "can", "its", "use", "using", "any", "all", "when", "where",
+    "les", "des", "une", "est", "pour", "dans", "par", "sur", "avec", "que", "qui", "quoi", "pas",
+    "aux", "ces", "ses", "son", "sont", "mais", "ou", "plus", "comme", "entre", "tout", "tous",
+]);
+/** Lowercase + strip accents, keeping a 1:1 index mapping to the input. */
+export function foldForMatch(text) {
+    const lower = text.toLowerCase();
+    const base = lower.length === text.length ? lower : [...text].map((c) => oneChar(c.toLowerCase(), c)).join("");
+    // Fast path: most content is ASCII.
+    if (!/[^\x00-\x7f]/.test(base))
+        return base;
+    return base.replace(/[^\x00-\x7f]/g, (c) => oneChar(c.normalize("NFD").replace(/\p{M}/gu, ""), c));
+}
+function oneChar(candidate, original) {
+    return candidate.length === original.length ? candidate : original;
+}
+export function passageQueryTerms(query) {
+    const out = new Set();
+    // Dotted versions are literal. Ordinary separators keep the previous
+    // tokenization: "design-critic" still finds "design critic".
+    const tokens = (query.normalize("NFC").match(/[\p{L}\p{N}_]+(?:\.[\p{L}\p{N}_]+)*/gu) ?? [])
+        .flatMap((token) => /\.\d/u.test(token) ? [token] : tokenize(token));
+    for (const tok of tokens) {
+        const t = foldForMatch(tok.toLowerCase());
+        if (STOPWORDS.has(t))
+            continue;
+        if (t.length < 3 && !/\d/.test(t))
+            continue;
+        out.add(t);
+    }
+    return [...out];
+}
+function isWordChar(c) {
+    return c !== undefined && /[\p{L}\p{N}_]/u.test(c);
+}
+/** Index of the first element >= target in an ascending array. */
+function lowerBound(sorted, target) {
+    let lo = 0;
+    let hi = sorted.length;
+    while (lo < hi) {
+        const mid = (lo + hi) >> 1;
+        if (sorted[mid] < target)
+            lo = mid + 1;
+        else
+            hi = mid;
+    }
+    return lo;
+}
+/**
+ * Best query-matching passages in `content`, ordered by position. Empty when
+ * no query term occurs (callers fall back to their summary/snippet).
+ */
+export function extractPassages(content, query, opts = {}) {
+    const max = opts.max ?? DEFAULT_MAX;
+    const window = opts.windowChars ?? DEFAULT_WINDOW;
+    const terms = passageQueryTerms(query);
+    if (!content || terms.length === 0 || max <= 0)
+        return [];
+    const folded = foldForMatch(content);
+    const hitsByTerm = new Map();
+    for (const term of terms) {
+        const hits = [];
+        let from = 0;
+        while (hits.length < MAX_HITS_PER_TERM) {
+            const at = folded.indexOf(term, from);
+            if (at < 0)
+                break;
+            from = at + term.length;
+            if (isWordChar(folded[at - 1]) || isWordChar(folded[at + term.length]))
+                continue;
+            hits.push(at);
+        }
+        if (hits.length)
+            hitsByTerm.set(term, hits);
+    }
+    if (hitsByTerm.size === 0)
+        return [];
+    // Rare-in-this-row terms are more telling than ones repeated everywhere.
+    const weight = new Map();
+    for (const [term, hits] of hitsByTerm)
+        weight.set(term, 1 / Math.log2(2 + hits.length));
+    const candidates = [];
+    for (const hits of hitsByTerm.values()) {
+        for (const at of hits) {
+            let start = Math.max(0, at - Math.floor(window / 2));
+            const end = Math.min(content.length, start + window);
+            start = Math.max(0, end - window);
+            let score = 0;
+            let matched = 0;
+            for (const [term, termHits] of hitsByTerm) {
+                const first = lowerBound(termHits, start);
+                if (first < termHits.length && termHits[first] + term.length <= end) {
+                    score += weight.get(term);
+                    matched++;
+                }
+            }
+            candidates.push({ start, end, score, matched });
+        }
+    }
+    candidates.sort((a, b) => b.score - a.score || a.start - b.start);
+    const chosen = [];
+    for (const c of candidates) {
+        if (chosen.length >= max)
+            break;
+        if (chosen.some((p) => c.start < p.end && p.start < c.end))
+            continue;
+        chosen.push(c);
+    }
+    chosen.sort((a, b) => a.start - b.start);
+    return chosen.map((c) => {
+        const { start, end } = snapToWords(content, c.start, c.end);
+        let text = content.slice(start, end).replace(/\s+/g, " ").trim();
+        if (start > 0)
+            text = `…${text}`;
+        if (end < content.length)
+            text = `${text}…`;
+        return { text, start, end, terms_matched: c.matched, terms_total: terms.length };
+    });
+}
+/** Nudge window edges inward to whitespace so words are not cut in half. */
+function snapToWords(content, start, end) {
+    let s = start;
+    if (s > 0 && isWordChar(content[s - 1])) {
+        const next = content.slice(s, Math.min(end, s + 40)).search(/\s/);
+        if (next >= 0)
+            s += next + 1;
+    }
+    let e = end;
+    if (e < content.length && isWordChar(content[e])) {
+        const back = content.slice(Math.max(s, e - 40), e).search(/\s\S*$/);
+        if (back >= 0)
+            e = Math.max(s, e - 40) + back;
+    }
+    return { start: s, end: Math.max(e, s) };
+}
+//# sourceMappingURL=passages.js.map

package/dist/src/retrieval/passages.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"passages.js","sourceRoot":"","sources":["../../../src/retrieval/passages.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;GAaG;AACH,OAAO,EAAE,QAAQ,EAAE,MAAM,cAAc,CAAC;AAmBxC,MAAM,WAAW,GAAG,CAAC,CAAC;AACtB,MAAM,cAAc,GAAG,GAAG,CAAC;AAC3B,kEAAkE;AAClE,MAAM,iBAAiB,GAAG,GAAG,CAAC;AAE9B,4EAA4E;AAC5E,6EAA6E;AAC7E,MAAM,SAAS,GAAG,IAAI,GAAG,CAAC;IACxB,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,OAAO,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK;IACzF,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,OAAO,EAAE,QAAQ,EAAE,OAAO,EAAE,OAAO,EAAE,MAAM,EAAE,KAAK;IACxF,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,OAAO,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,OAAO;IAC/F,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK;IAC7F,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,OAAO,EAAE,MAAM,EAAE,MAAM;CAC3F,CAAC,CAAC;AAEH,2EAA2E;AAC3E,MAAM,UAAU,YAAY,CAAC,IAAY;IACvC,MAAM,KAAK,GAAG,IAAI,CAAC,WAAW,EAAE,CAAC;IACjC,MAAM,IAAI,GAAG,KAAK,CAAC,MAAM,KAAK,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,OAAO,CAAC,CAAC,CAAC,WAAW,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAC/G,oCAAoC;IACpC,IAAI,CAAC,cAAc,CAAC,IAAI,CAAC,IAAI,CAAC;QAAE,OAAO,IAAI,CAAC;IAC5C,OAAO,IAAI,CAAC,OAAO,CAAC,eAAe,EAAE,CAAC,CAAC,EAAE,EAAE,CAAC,OAAO,CAAC,CAAC,CAAC,SAAS,CAAC,KAAK,CAAC,CAAC,OAAO,CAAC,SAAS,EAAE,EAAE,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;AACrG,CAAC;AAED,SAAS,OAAO,CAAC,SAAiB,EAAE,QAAgB;IAClD,OAAO,SAAS,CAAC,MAAM,KAAK,QAAQ,CAAC,MAAM,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,QAAQ,CAAC;AACrE,CAAC;AAED,MAAM,UAAU,iBAAiB,CAAC,KAAa;IAC7C,MAAM,GAAG,GAAG,IAAI,GAAG,EAAU,CAAC;IAC9B,qEAAqE;IACrE,6DAA6D;IAC7D,MAAM,MAAM,GAAG,CAAC,KAAK,CAAC,SAAS,CAAC,KAAK,CAAC,CAAC,KAAK,CAAC,uCAAuC,CAAC,IAAI,EAAE,CAAC;SACzF,OAAO,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,OAAO,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,CAAC;IACvE,KAAK,MAAM,GAAG,IAAI,MAAM,EAAE,CAAC;QACzB,MAAM,CAAC,GAAG,YAAY,CAAC,GAAG,CAAC,WAAW,EAAE,CAAC,CAAC;QAC1C,IAAI,SAAS,CAAC,GAAG,CAAC,CAAC,CAAC;YAAE,SAAS;QAC/B,IAAI,CAAC,CAAC,MAAM,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC;YAAE,SAAS;QAC5C,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC;IACb,CAAC;IACD,OAAO,CAAC,GAAG,GAAG,CAAC,CAAC;AAClB,CAAC;AAED,SAAS,UAAU,CAAC,CAAqB;IACvC,OAAO,CAAC,KAAK,SAAS,IAAI,gBAAgB,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;AACrD,CAAC;AAED,kEAAkE;AAClE,SAAS,UAAU,CAAC,MAAgB,EAAE,MAAc;IAClD,IAAI,EAAE,GAAG,CAAC,CAAC;IACX,IAAI,EAAE,GAAG,MAAM,CAAC,MAAM,CAAC;IACvB,OAAO,EAAE,GAAG,EAAE,EAAE,CAAC;QACf,MAAM,GAAG,GAAG,CAAC,EAAE,GAAG,EAAE,CAAC,IAAI,CAAC,CAAC;QAC3B,IAAI,MAAM,CAAC,GAAG,CAAE,GAAG,MAAM;YAAE,EAAE,GAAG,GAAG,GAAG,CAAC,CAAC;;YACnC,EAAE,GAAG,GAAG,CAAC;IAChB,CAAC;IACD,OAAO,EAAE,CAAC;AACZ,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,eAAe,CAAC,OAAe,EAAE,KAAa,EAAE,OAAuB,EAAE;IACvF,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,IAAI,WAAW,CAAC;IACpC,MAAM,MAAM,GAAG,IAAI,CAAC,WAAW,IAAI,cAAc,CAAC;IAClD,MAAM,KAAK,GAAG,iBAAiB,CAAC,KAAK,CAAC,CAAC;IACvC,IAAI,CAAC,OAAO,IAAI,KAAK,CAAC,MAAM,KAAK,CAAC,IAAI,GAAG,IAAI,CAAC;QAAE,OAAO,EAAE,CAAC;IAE1D,MAAM,MAAM,GAAG,YAAY,CAAC,OAAO,CAAC,CAAC;IACrC,MAAM,UAAU,GAAG,IAAI,GAAG,EAAoB,CAAC;IAC/C,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;QACzB,MAAM,IAAI,GAAa,EAAE,CAAC;QAC1B,IAAI,IAAI,GAAG,CAAC,CAAC;QACb,OAAO,IAAI,CAAC,MAAM,GAAG,iBAAiB,EAAE,CAAC;YACvC,MAAM,EAAE,GAAG,MAAM,CAAC,OAAO,CAAC,IAAI,EAAE,IAAI,CAAC,CAAC;YACtC,IAAI,EAAE,GAAG,CAAC;gBAAE,MAAM;YAClB,IAAI,GAAG,EAAE,GAAG,IAAI,CAAC,MAAM,CAAC;YACxB,IAAI,UAAU,CAAC,MAAM,CAAC,EAAE,GAAG,CAAC,CAAC,CAAC,IAAI,UAAU,CAAC,MAAM,CAAC,EAAE,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC;gBAAE,SAAS;YACjF,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QAChB,CAAC;QACD,IAAI,IAAI,CAAC,MAAM;YAAE,UAAU,CAAC,GAAG,CAAC,IAAI,EAAE,IAAI,CAAC,CAAC;IAC9C,CAAC;IACD,IAAI,UAAU,CAAC,IAAI,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IAErC,yEAAyE;IACzE,MAAM,MAAM,GAAG,IAAI,GAAG,EAAkB,CAAC;IACzC,KAAK,MAAM,CAAC,IAAI,EAAE,IAAI,CAAC,IAAI,UAAU;QAAE,MAAM,CAAC,GAAG,CAAC,IAAI,EAAE,CAAC,GAAG,IAAI,CAAC,IAAI,CAAC,CAAC,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC;IAExF,MAAM,UAAU,GAAqE,EAAE,CAAC;IACxF,KAAK,MAAM,IAAI,IAAI,UAAU,CAAC,MAAM,EAAE,EAAE,CAAC;QACvC,KAAK,MAAM,EAAE,IAAI,IAAI,EAAE,CAAC;YACtB,IAAI,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,EAAE,GAAG,IAAI,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC;YACrD,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,OAAO,CAAC,MAAM,EAAE,KAAK,GAAG,MAAM,CAAC,CAAC;YACrD,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,GAAG,GAAG,MAAM,CAAC,CAAC;YAClC,IAAI,KAAK,GAAG,CAAC,CAAC;YACd,IAAI,OAAO,GAAG,CAAC,CAAC;YAChB,KAAK,MAAM,CAAC,IAAI,EAAE,QAAQ,CAAC,IAAI,UAAU,EAAE,CAAC;gBAC1C,MAAM,KAAK,GAAG,UAAU,CAAC,QAAQ,EAAE,KAAK,CAAC,CAAC;gBAC1C,IAAI,KAAK,GAAG,QAAQ,CAAC,MAAM,IAAI,QAAQ,CAAC,KAAK,CAAE,GAAG,IAAI,CAAC,MAAM,IAAI,GAAG,EAAE,CAAC;oBACrE,KAAK,IAAI,MAAM,CAAC,GAAG,CAAC,IAAI,CAAE,CAAC;oBAC3B,OAAO,EAAE,CAAC;gBACZ,CAAC;YACH,CAAC;YACD,UAAU,CAAC,IAAI,CAAC,EAAE,KAAK,EAAE,GAAG,EAAE,KAAK,EAAE,OAAO,EAAE,CAAC,CAAC;QAClD,CAAC;IACH,CAAC;IACD,UAAU,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,IAAI,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC;IAElE,MAAM,MAAM,GAAsB,EAAE,CAAC;IACrC,KAAK,MAAM,CAAC,IAAI,UAAU,EAAE,CAAC;QAC3B,IAAI,MAAM,CAAC,MAAM,IAAI,GAAG;YAAE,MAAM;QAChC,IAAI,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,GAAG,CAAC;YAAE,SAAS;QACrE,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IACjB,CAAC;IACD,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC;IAEzC,OAAO,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE;QACtB,MAAM,EAAE,KAAK,EAAE,GAAG,EAAE,GAAG,WAAW,CAAC,OAAO,EAAE,CAAC,CAAC,KAAK,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC;QAC5D,IAAI,IAAI,GAAG,OAAO,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC;QACjE,IAAI,KAAK,GAAG,CAAC;YAAE,IAAI,GAAG,IAAI,IAAI,EAAE,CAAC;QACjC,IAAI,GAAG,GAAG,OAAO,CAAC,MAAM;YAAE,IAAI,GAAG,GAAG,IAAI,GAAG,CAAC;QAC5C,OAAO,EAAE,IAAI,EAAE,KAAK,EAAE,GAAG,EAAE,aAAa,EAAE,CAAC,CAAC,OAAO,EAAE,WAAW,EAAE,KAAK,CAAC,MAAM,EAAE,CAAC;IACnF,CAAC,CAAC,CAAC;AACL,CAAC;AAED,4EAA4E;AAC5E,SAAS,WAAW,CAAC,OAAe,EAAE,KAAa,EAAE,GAAW;IAC9D,IAAI,CAAC,GAAG,KAAK,CAAC;IACd,IAAI,CAAC,GAAG,CAAC,IAAI,UAAU,CAAC,OAAO,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC;QACxC,MAAM,IAAI,GAAG,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,CAAC,GAAG,EAAE,CAAC,GAAG,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC;QAClE,IAAI,IAAI,IAAI,CAAC;YAAE,CAAC,IAAI,IAAI,GAAG,CAAC,CAAC;IAC/B,CAAC;IACD,IAAI,CAAC,GAAG,GAAG,CAAC;IACZ,IAAI,CAAC,GAAG,OAAO,CAAC,MAAM,IAAI,UAAU,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;QACjD,MAAM,IAAI,GAAG,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,GAAG,EAAE,CAAC,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC,QAAQ,CAAC,CAAC;QACpE,IAAI,IAAI,IAAI,CAAC;YAAE,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,GAAG,EAAE,CAAC,GAAG,IAAI,CAAC;IAChD,CAAC;IACD,OAAO,EAAE,KAAK,EAAE,CAAC,EAAE,GAAG,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,EAAE,CAAC;AAC3C,CAAC"}

package/dist/src/retrieval/post-filter.d.ts

@@ -0,0 +1,32 @@
+/**
+ * Post-filter utilities for topic/decision filtering on conversation search.
+ *
+ * Applied to the merged candidate pool BEFORE reranking and BEFORE top_k
+ * truncation. Parses JSON-encoded string arrays and performs case-insensitive
+ * exact-element matching.
+ */
+/**
+ * Safely parse a JSON-encoded string array.
+ * Returns [] on any failure (malformed JSON, non-array, non-string elements).
+ */
+export declare function parseJsonStringArray(raw: string | undefined | null): string[];
+/**
+ * Case-insensitive exact-element match.
+ * "agent" matches "Agent" but NOT "agent-access".
+ */
+export declare function matchesElement(elements: string[], target: string): boolean;
+/**
+ * Filter candidates by topic and/or decision.
+ * AND semantics when both filters are provided.
+ *
+ * Operates on any type with optional `topics`/`decisions` JSON string fields,
+ * including RerankCandidate.
+ */
+export declare function applyPostFilters<T extends {
+    topics?: string;
+    decisions?: string;
+}>(candidates: T[], filters: {
+    topic?: string;
+    decision?: string;
+}): T[];
+//# sourceMappingURL=post-filter.d.ts.map

package/dist/src/retrieval/post-filter.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"post-filter.d.ts","sourceRoot":"","sources":["../../../src/retrieval/post-filter.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AAEH;;;GAGG;AACH,wBAAgB,oBAAoB,CAClC,GAAG,EAAE,MAAM,GAAG,SAAS,GAAG,IAAI,GAC7B,MAAM,EAAE,CASV;AAED;;;GAGG;AACH,wBAAgB,cAAc,CAAC,QAAQ,EAAE,MAAM,EAAE,EAAE,MAAM,EAAE,MAAM,GAAG,OAAO,CAG1E;AAED;;;;;;GAMG;AACH,wBAAgB,gBAAgB,CAC9B,CAAC,SAAS;IAAE,MAAM,CAAC,EAAE,MAAM,CAAC;IAAC,SAAS,CAAC,EAAE,MAAM,CAAA;CAAE,EACjD,UAAU,EAAE,CAAC,EAAE,EAAE,OAAO,EAAE;IAAE,KAAK,CAAC,EAAE,MAAM,CAAC;IAAC,QAAQ,CAAC,EAAE,MAAM,CAAA;CAAE,GAAG,CAAC,EAAE,CActE"}

package/dist/src/retrieval/post-filter.js

@@ -0,0 +1,57 @@
+/**
+ * Post-filter utilities for topic/decision filtering on conversation search.
+ *
+ * Applied to the merged candidate pool BEFORE reranking and BEFORE top_k
+ * truncation. Parses JSON-encoded string arrays and performs case-insensitive
+ * exact-element matching.
+ */
+/**
+ * Safely parse a JSON-encoded string array.
+ * Returns [] on any failure (malformed JSON, non-array, non-string elements).
+ */
+export function parseJsonStringArray(raw) {
+    if (!raw)
+        return [];
+    try {
+        const parsed = JSON.parse(raw);
+        if (!Array.isArray(parsed))
+            return [];
+        return parsed.filter((v) => typeof v === "string");
+    }
+    catch {
+        return [];
+    }
+}
+/**
+ * Case-insensitive exact-element match.
+ * "agent" matches "Agent" but NOT "agent-access".
+ */
+export function matchesElement(elements, target) {
+    const lower = target.toLowerCase();
+    return elements.some((e) => e.toLowerCase() === lower);
+}
+/**
+ * Filter candidates by topic and/or decision.
+ * AND semantics when both filters are provided.
+ *
+ * Operates on any type with optional `topics`/`decisions` JSON string fields,
+ * including RerankCandidate.
+ */
+export function applyPostFilters(candidates, filters) {
+    if (!filters.topic && !filters.decision)
+        return candidates;
+    return candidates.filter((c) => {
+        if (filters.topic) {
+            const elements = parseJsonStringArray(c.topics);
+            if (!matchesElement(elements, filters.topic))
+                return false;
+        }
+        if (filters.decision) {
+            const elements = parseJsonStringArray(c.decisions);
+            if (!matchesElement(elements, filters.decision))
+                return false;
+        }
+        return true;
+    });
+}
+//# sourceMappingURL=post-filter.js.map

package/dist/src/retrieval/post-filter.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"post-filter.js","sourceRoot":"","sources":["../../../src/retrieval/post-filter.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AAEH;;;GAGG;AACH,MAAM,UAAU,oBAAoB,CAClC,GAA8B;IAE9B,IAAI,CAAC,GAAG;QAAE,OAAO,EAAE,CAAC;IACpB,IAAI,CAAC;QACH,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;QAC/B,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,MAAM,CAAC;YAAE,OAAO,EAAE,CAAC;QACtC,OAAO,MAAM,CAAC,MAAM,CAAC,CAAC,CAAU,EAAE,EAAE,CAAC,OAAO,CAAC,KAAK,QAAQ,CAAC,CAAC;IAC9D,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,EAAE,CAAC;IACZ,CAAC;AACH,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,cAAc,CAAC,QAAkB,EAAE,MAAc;IAC/D,MAAM,KAAK,GAAG,MAAM,CAAC,WAAW,EAAE,CAAC;IACnC,OAAO,QAAQ,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,EAAE,KAAK,KAAK,CAAC,CAAC;AACzD,CAAC;AAED;;;;;;GAMG;AACH,MAAM,UAAU,gBAAgB,CAE9B,UAAe,EAAE,OAA8C;IAC/D,IAAI,CAAC,OAAO,CAAC,KAAK,IAAI,CAAC,OAAO,CAAC,QAAQ;QAAE,OAAO,UAAU,CAAC;IAE3D,OAAO,UAAU,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE;QAC7B,IAAI,OAAO,CAAC,KAAK,EAAE,CAAC;YAClB,MAAM,QAAQ,GAAG,oBAAoB,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC;YAChD,IAAI,CAAC,cAAc,CAAC,QAAQ,EAAE,OAAO,CAAC,KAAK,CAAC;gBAAE,OAAO,KAAK,CAAC;QAC7D,CAAC;QACD,IAAI,OAAO,CAAC,QAAQ,EAAE,CAAC;YACrB,MAAM,QAAQ,GAAG,oBAAoB,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC;YACnD,IAAI,CAAC,cAAc,CAAC,QAAQ,EAAE,OAAO,CAAC,QAAQ,CAAC;gBAAE,OAAO,KAAK,CAAC;QAChE,CAAC;QACD,OAAO,IAAI,CAAC;IACd,CAAC,CAAC,CAAC;AACL,CAAC"}

package/dist/src/retrieval/project-index.d.ts

@@ -0,0 +1,71 @@
+export interface SessionProject {
+    cwd: string | null;
+    root: string | null;
+}
+export interface TranscriptHomes {
+    claudeProjects: string[];
+    codexRoots: string[];
+}
+/**
+ * Every Claude Code / Codex home on this machine: the defaults, per-account
+ * siblings (`~/.claude-<name>`, `~/.codex-<name>`), and CLAUDE_CONFIG_DIR /
+ * CODEX_HOME. Lookups are by session id of already-captured rows, so reading
+ * an extra home can link a session but never imports anything.
+ */
+export declare function discoverTranscriptHomes(home?: string, env?: NodeJS.ProcessEnv): TranscriptHomes;
+/** `auto-capture:<source>:<session>:<fp>` → `<source>:<session>`. */
+export declare function sessionKeyFromIdempotencyKey(key: string | undefined | null): string | null;
+/**
+ * Repository a directory belongs to. Worktrees resolve to the main repository
+ * (via the shared git dir), so a project keeps one identity across worktrees.
+ * The home directory and "/" are not projects. A directory that no longer
+ * exists keeps its own path as identity.
+ */
+export declare function resolveProjectRoot(cwd: string | null | undefined): string | null;
+export declare class ProjectIndex {
+    private readonly dataDir;
+    private readonly homes;
+    private data;
+    private rowsCache;
+    constructor(dataDir: string, homes?: TranscriptHomes);
+    private get file();
+    private load;
+    private save;
+    get(sessionKey: string): SessionProject | undefined;
+    /** Capture time: record the cwd the parser read from the transcript. */
+    record(sessionKey: string, cwd: string | null | undefined): void;
+    /**
+     * Resolve every auto-captured session in `conversations_hot` that the index
+     * has not seen yet, or saw without a cwd, from transcripts still on disk.
+     * New sessions with no transcript are recorded as unknown (counted once);
+     * known-unknown sessions are retried and only change when a cwd turns up.
+     */
+    backfillFromTable(): Promise<{
+        resolved: number;
+        unknown: number;
+    }>;
+    /** sessionId → transcript path, for both hosts across every home. */
+    private transcriptFiles;
+    /**
+     * Row ids per project root, from the conversations table. Recomputed when
+     * the table version moves (any add, delete or update — a replaced capture
+     * keeps the row count but changes the version) or the index gains entries.
+     */
+    rowsByProject(): Promise<{
+        byRoot: Map<string, string[]>;
+        projectById: Map<string, string>;
+    }>;
+}
+/**
+ * After a save job completes: link its session to the project it ran in.
+ * Every save path (daemon capture, `onadit capture`, POST /save) goes through
+ * the save job, so this is the one place capture-time identity is recorded.
+ * Best-effort — a failed index write never fails the save.
+ */
+export declare function recordSavedSessionProject(index: ProjectIndex, payload: unknown): void;
+/** First `cwd` recorded in a transcript's opening lines (both hosts write it early). */
+export declare function readTranscriptCwd(file: string): string | null;
+/** The daemon installs its index once; retrieval reads it (null in tests/CLI). */
+export declare function setProjectIndex(index: ProjectIndex | null): void;
+export declare function getProjectIndex(): ProjectIndex | null;
+//# sourceMappingURL=project-index.d.ts.map

package/dist/src/retrieval/project-index.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"project-index.d.ts","sourceRoot":"","sources":["../../../src/retrieval/project-index.ts"],"names":[],"mappings":"AA4BA,MAAM,WAAW,cAAc;IAC7B,GAAG,EAAE,MAAM,GAAG,IAAI,CAAC;IACnB,IAAI,EAAE,MAAM,GAAG,IAAI,CAAC;CACrB;AASD,MAAM,WAAW,eAAe;IAC9B,cAAc,EAAE,MAAM,EAAE,CAAC;IACzB,UAAU,EAAE,MAAM,EAAE,CAAC;CACtB;AAED;;;;;GAKG;AACH,wBAAgB,uBAAuB,CACrC,IAAI,GAAE,MAAqB,EAC3B,GAAG,GAAE,MAAM,CAAC,UAAwB,GACnC,eAAe,CAoBjB;AAGD,qEAAqE;AACrE,wBAAgB,4BAA4B,CAAC,GAAG,EAAE,MAAM,GAAG,SAAS,GAAG,IAAI,GAAG,MAAM,GAAG,IAAI,CAK1F;AAID;;;;;GAKG;AACH,wBAAgB,kBAAkB,CAAC,GAAG,EAAE,MAAM,GAAG,IAAI,GAAG,SAAS,GAAG,MAAM,GAAG,IAAI,CAqBhF;AAED,qBAAa,YAAY;IAKrB,OAAO,CAAC,QAAQ,CAAC,OAAO;IACxB,OAAO,CAAC,QAAQ,CAAC,KAAK;IALxB,OAAO,CAAC,IAAI,CAAY;IACxB,OAAO,CAAC,SAAS,CAAqG;gBAGnG,OAAO,EAAE,MAAM,EACf,KAAK,GAAE,eAA2C;IAKrE,OAAO,KAAK,IAAI,GAEf;IAED,OAAO,CAAC,IAAI;IAUZ,OAAO,CAAC,IAAI;IAMZ,GAAG,CAAC,UAAU,EAAE,MAAM,GAAG,cAAc,GAAG,SAAS;IAInD,wEAAwE;IACxE,MAAM,CAAC,UAAU,EAAE,MAAM,EAAE,GAAG,EAAE,MAAM,GAAG,IAAI,GAAG,SAAS,GAAG,IAAI;IAShE;;;;;OAKG;IACG,iBAAiB,IAAI,OAAO,CAAC;QAAE,QAAQ,EAAE,MAAM,CAAC;QAAC,OAAO,EAAE,MAAM,CAAA;KAAE,CAAC;IAyCzE,qEAAqE;IACrE,OAAO,CAAC,eAAe;IA0BvB;;;;OAIG;IACG,aAAa,IAAI,OAAO,CAAC;QAAE,MAAM,EAAE,GAAG,CAAC,MAAM,EAAE,MAAM,EAAE,CAAC,CAAC;QAAC,WAAW,EAAE,GAAG,CAAC,MAAM,EAAE,MAAM,CAAC,CAAA;KAAE,CAAC;CAsBpG;AAED;;;;;GAKG;AACH,wBAAgB,yBAAyB,CAAC,KAAK,EAAE,YAAY,EAAE,OAAO,EAAE,OAAO,GAAG,IAAI,CAarF;AA+BD,wFAAwF;AACxF,wBAAgB,iBAAiB,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,GAAG,IAAI,CAsB7D;AAID,kFAAkF;AAClF,wBAAgB,eAAe,CAAC,KAAK,EAAE,YAAY,GAAG,IAAI,GAAG,IAAI,CAEhE;AAED,wBAAgB,eAAe,IAAI,YAAY,GAAG,IAAI,CAErD"}