trajectories 0.0.0 → 0.1.1

package/legacy/src/gemini_embeddings.ts

@@ -0,0 +1,103 @@
+/**
+ * Gemini cloud embedding provider (gemini-embedding-2, 768-dim).
+ *
+ * Live-checked against the documented REST surface:
+ *   POST /v1beta/models/gemini-embedding-2:batchEmbedContents
+ *   requests: [{ model, content: {parts:[{text}]}, outputDimensionality }]
+ * Task prefixes per docs "Task types with Embeddings 2" (retrieval, asymmetric):
+ *   query:    "task: search result | query: {content}"
+ *   document: "title: {title} | text: {content}"   (title: none when absent)
+ */
+import type { EmbeddingProvider } from "./embeddings.js";
+import { readFileSync, existsSync } from "node:fs";
+import { homedir } from "node:os";
+import { join } from "node:path";
+
+export const GEMINI_EMBEDDING_MODEL = "gemini-embedding-2";
+export const GEMINI_EMBEDDING_DIM = 768;
+
+const API_BASE = "https://generativelanguage.googleapis.com/v1beta/models";
+const BATCH = 100; // requests per batchEmbedContents call
+const MAX_RETRIES = 3;
+
+export function resolveGeminiKey(): string | null {
+  if (process.env.GEMINI_API_KEY) return process.env.GEMINI_API_KEY;
+  if (process.env.LEDGER_GEMINI_KEY) return process.env.LEDGER_GEMINI_KEY;
+  const p = join(homedir(), ".air", "gemini-api-key");
+  if (existsSync(p)) return readFileSync(p, "utf8").trim();
+  return null;
+}
+
+export function formatQuery(text: string): string {
+  return `task: search result | query: ${text}`;
+}
+
+export function formatDocument(text: string, title?: string | null): string {
+  return `title: ${title ?? "none"} | text: ${text}`;
+}
+
+export class GeminiEmbeddingProvider implements EmbeddingProvider {
+  readonly dimension = GEMINI_EMBEDDING_DIM;
+  readonly name = "gemini-embedding-2";
+
+  private apiKey: string | null = null;
+  /** Set false to embed everything as documents (default indexing path). */
+  queryMode = false;
+
+  private key(): string | null {
+    this.apiKey ??= resolveGeminiKey();
+    return this.apiKey;
+  }
+
+  async embed(texts: string[]): Promise<Float32Array[]> {
+    if (texts.length === 0) return [];
+    const key = this.key();
+    if (!key) throw new Error("No Gemini API key (GEMINI_API_KEY or ~/.air/gemini-api-key)");
+
+    const out: Float32Array[] = [];
+    for (let start = 0; start < texts.length; start += BATCH) {
+      const slice = texts.slice(start, start + BATCH);
+      out.push(...(await this.embedBatch(slice, key)));
+    }
+    return out;
+  }
+
+  private async embedBatch(texts: string[], key: string): Promise<Float32Array[]> {
+    const requests = texts.map((t) => ({
+      model: `models/${GEMINI_EMBEDDING_MODEL}`,
+      content: { parts: [{ text: this.queryMode ? formatQuery(t) : formatDocument(t) }] },
+      outputDimensionality: GEMINI_EMBEDDING_DIM,
+    }));
+
+    let lastErr: unknown = null;
+    for (let attempt = 0; attempt < MAX_RETRIES; attempt++) {
+      try {
+        const res = await fetch(`${API_BASE}/${GEMINI_EMBEDDING_MODEL}:batchEmbedContents`, {
+          method: "POST",
+          headers: { "Content-Type": "application/json", "x-goog-api-key": key },
+          body: JSON.stringify({ requests }),
+        });
+        if (!res.ok) {
+          const body = await res.text();
+          throw new Error(`Gemini embedContent HTTP ${res.status}: ${body.slice(0, 300)}`);
+        }
+        const data = (await res.json()) as { embeddings?: Array<{ values: number[] }> };
+        const embs = data.embeddings ?? [];
+        if (embs.length !== texts.length) {
+          throw new Error(`Gemini returned ${embs.length}/${texts.length} embeddings`);
+        }
+        return embs.map((e) => {
+          const v = new Float32Array(GEMINI_EMBEDDING_DIM);
+          for (let i = 0; i < Math.min(v.length, e.values.length); i++) v[i] = e.values[i];
+          return v;
+        });
+      } catch (err) {
+        lastErr = err;
+        if (attempt < MAX_RETRIES - 1) {
+          await new Promise((r) => setTimeout(r, 500 * 2 ** attempt));
+        }
+      }
+    }
+    throw lastErr instanceof Error ? lastErr : new Error(String(lastErr));
+  }
+}

package/legacy/src/identity.ts

@@ -0,0 +1,9 @@
+/** Trajectory identities are native to a harness; Ledger keys include its source. */
+export function ledgerId(source: string, nativeId: string): string {
+  return `${source}:${nativeId}`;
+}
+
+export function nativeId(source: string, id: string): string {
+  const prefix = `${source}:`;
+  return id.startsWith(prefix) ? id.slice(prefix.length) : id;
+}

package/legacy/src/index.ts

@@ -0,0 +1,10 @@
+export * from "./types.js";
+export { CREATE_TABLES_SQL, SCHEMA_VERSION } from "./schema.js";
+export { SessionNotFoundError, StoreNotInitializedError, TombstonedSessionError, TrajectoryDb } from "./db.js";
+export { ReaderError, TrajectoryReader } from "./reader.js";
+export * from "./sync-truth.js";
+export * from "./normalizer.js";
+export * from "./scanner.js";
+export * from "./cli.js";
+export * from "./watch.js";
+export * from "./gemini_embeddings.js";