Add bots files handling logic

src/harness/providers/openai-compatible/embed.ts

@@ -0,0 +1,68 @@
+import { AppData } from "@/src/data/app-data";
+import HarnessError from "../../harness-error";
+import type { HarnessAdapter } from "../../types";
+import grabProviderPreset from "../grab-provider-preset";
+import providerRequest from "./provider-request";
+
+/**
+ * Embeddings at the provider's `embeddings_path` (the OpenAI `/embeddings`
+ * shape). Batched, since a large file's chunks would otherwise go in one
+ * request the provider may refuse.
+ */
+const embed: HarnessAdapter["embed"] = async ({ candidate, input, signal }) => {
+    const { provider, model } = candidate;
+    const preset = grabProviderPreset({ kind: provider.kind });
+    const vectors: number[][] = [];
+
+    /**
+     * Every OpenAI-compatible API puts embeddings at `/embeddings`.
+     */
+    const path = "/embeddings";
+
+    for (let start = 0; start < input.length; start += AppData.EmbeddingBatchSize) {
+        const batch = input
+            .slice(start, start + AppData.EmbeddingBatchSize)
+            .map((text) => text.slice(0, AppData.EmbeddingMaxInputChars));
+
+        const res = await providerRequest({
+            provider,
+            path,
+            json: { model: model.slug, input: batch },
+            signal,
+            candidate_key: candidate.key,
+        });
+
+        const json = await res.json().catch(() => null);
+        const items: any[] = Array.isArray(json?.data)
+            ? json.data
+            : Array.isArray(json?.embeddings)
+              ? json.embeddings
+              : [];
+
+        if (items.length != batch.length) {
+            throw new HarnessError({
+                message: `${provider.name || preset.title}: ${items.length} of ${batch.length} texts got a vector.`,
+                retryable: true,
+                candidate_key: candidate.key,
+            });
+        }
+
+        for (const item of items) {
+            const vector = Array.isArray(item) ? item : item?.embedding;
+
+            if (!Array.isArray(vector) || !vector.length) {
+                throw new HarnessError({
+                    message: `${provider.name || preset.title} sent no embedding back.`,
+                    retryable: true,
+                    candidate_key: candidate.key,
+                });
+            }
+
+            vectors.push(vector.map(Number));
+        }
+    }
+
+    return { vectors };
+};
+
+export default embed;