mirror of
https://github.com/moltbot/moltbot.git
synced 2026-03-21 16:41:56 +00:00
refactor(media): extract runner entry execution helpers
This commit is contained in:
Peter Steinberger
591
src/media-understanding/runner.entries.ts
Normal file
591
src/media-understanding/runner.entries.ts
Normal file
@@ -0,0 +1,591 @@
|
||||
import fs from "node:fs/promises";
|
||||
import os from "node:os";
|
||||
import path from "node:path";
|
||||
import type { MsgContext } from "../auto-reply/templating.js";
|
||||
import type { OpenClawConfig } from "../config/config.js";
|
||||
import type {
|
||||
MediaUnderstandingConfig,
|
||||
MediaUnderstandingModelConfig,
|
||||
} from "../config/types.tools.js";
|
||||
import type {
|
||||
MediaUnderstandingCapability,
|
||||
MediaUnderstandingDecision,
|
||||
MediaUnderstandingModelDecision,
|
||||
MediaUnderstandingOutput,
|
||||
MediaUnderstandingProvider,
|
||||
} from "./types.js";
|
||||
import { requireApiKey, resolveApiKeyForProvider } from "../agents/model-auth.js";
|
||||
import { applyTemplate } from "../auto-reply/templating.js";
|
||||
import { logVerbose, shouldLogVerbose } from "../globals.js";
|
||||
import { runExec } from "../process/exec.js";
|
||||
import { MediaAttachmentCache } from "./attachments.js";
|
||||
import {
|
||||
CLI_OUTPUT_MAX_BUFFER,
|
||||
DEFAULT_AUDIO_MODELS,
|
||||
DEFAULT_TIMEOUT_SECONDS,
|
||||
} from "./defaults.js";
|
||||
import { MediaUnderstandingSkipError } from "./errors.js";
|
||||
import { describeImageWithModel } from "./providers/image.js";
|
||||
import { getMediaUnderstandingProvider, normalizeMediaProviderId } from "./providers/index.js";
|
||||
import { resolveMaxBytes, resolveMaxChars, resolvePrompt, resolveTimeoutMs } from "./resolve.js";
|
||||
import { estimateBase64Size, resolveVideoMaxBase64Bytes } from "./video.js";
|
||||
|
||||
export type ProviderRegistry = Map<string, MediaUnderstandingProvider>;
|
||||
|
||||
function trimOutput(text: string, maxChars?: number): string {
|
||||
const trimmed = text.trim();
|
||||
if (!maxChars || trimmed.length <= maxChars) {
|
||||
return trimmed;
|
||||
}
|
||||
return trimmed.slice(0, maxChars).trim();
|
||||
}
|
||||
|
||||
function extractLastJsonObject(raw: string): unknown {
|
||||
const trimmed = raw.trim();
|
||||
const start = trimmed.lastIndexOf("{");
|
||||
if (start === -1) {
|
||||
return null;
|
||||
}
|
||||
const slice = trimmed.slice(start);
|
||||
try {
|
||||
return JSON.parse(slice);
|
||||
} catch {
|
||||
return null;
|
||||
}
|
||||
}
|
||||
|
||||
function extractGeminiResponse(raw: string): string | null {
|
||||
const payload = extractLastJsonObject(raw);
|
||||
if (!payload || typeof payload !== "object") {
|
||||
return null;
|
||||
}
|
||||
const response = (payload as { response?: unknown }).response;
|
||||
if (typeof response !== "string") {
|
||||
return null;
|
||||
}
|
||||
const trimmed = response.trim();
|
||||
return trimmed || null;
|
||||
}
|
||||
|
||||
function extractSherpaOnnxText(raw: string): string | null {
|
||||
const tryParse = (value: string): string | null => {
|
||||
const trimmed = value.trim();
|
||||
if (!trimmed) {
|
||||
return null;
|
||||
}
|
||||
const head = trimmed[0];
|
||||
if (head !== "{" && head !== '"') {
|
||||
return null;
|
||||
}
|
||||
try {
|
||||
const parsed = JSON.parse(trimmed) as unknown;
|
||||
if (typeof parsed === "string") {
|
||||
return tryParse(parsed);
|
||||
}
|
||||
if (parsed && typeof parsed === "object") {
|
||||
const text = (parsed as { text?: unknown }).text;
|
||||
if (typeof text === "string" && text.trim()) {
|
||||
return text.trim();
|
||||
}
|
||||
}
|
||||
} catch {}
|
||||
return null;
|
||||
};
|
||||
|
||||
const direct = tryParse(raw);
|
||||
if (direct) {
|
||||
return direct;
|
||||
}
|
||||
|
||||
const lines = raw
|
||||
.split("\n")
|
||||
.map((line) => line.trim())
|
||||
.filter(Boolean);
|
||||
for (let i = lines.length - 1; i >= 0; i -= 1) {
|
||||
const parsed = tryParse(lines[i] ?? "");
|
||||
if (parsed) {
|
||||
return parsed;
|
||||
}
|
||||
}
|
||||
return null;
|
||||
}
|
||||
|
||||
function commandBase(command: string): string {
|
||||
return path.parse(command).name;
|
||||
}
|
||||
|
||||
function findArgValue(args: string[], keys: string[]): string | undefined {
|
||||
for (let i = 0; i < args.length; i += 1) {
|
||||
if (keys.includes(args[i] ?? "")) {
|
||||
const value = args[i + 1];
|
||||
if (value) {
|
||||
return value;
|
||||
}
|
||||
}
|
||||
}
|
||||
return undefined;
|
||||
}
|
||||
|
||||
function hasArg(args: string[], keys: string[]): boolean {
|
||||
return args.some((arg) => keys.includes(arg));
|
||||
}
|
||||
|
||||
function resolveWhisperOutputPath(args: string[], mediaPath: string): string | null {
|
||||
const outputDir = findArgValue(args, ["--output_dir", "-o"]);
|
||||
const outputFormat = findArgValue(args, ["--output_format"]);
|
||||
if (!outputDir || !outputFormat) {
|
||||
return null;
|
||||
}
|
||||
const formats = outputFormat.split(",").map((value) => value.trim());
|
||||
if (!formats.includes("txt")) {
|
||||
return null;
|
||||
}
|
||||
const base = path.parse(mediaPath).name;
|
||||
return path.join(outputDir, `${base}.txt`);
|
||||
}
|
||||
|
||||
function resolveWhisperCppOutputPath(args: string[]): string | null {
|
||||
if (!hasArg(args, ["-otxt", "--output-txt"])) {
|
||||
return null;
|
||||
}
|
||||
const outputBase = findArgValue(args, ["-of", "--output-file"]);
|
||||
if (!outputBase) {
|
||||
return null;
|
||||
}
|
||||
return `${outputBase}.txt`;
|
||||
}
|
||||
|
||||
async function fileExists(filePath?: string | null): Promise<boolean> {
|
||||
if (!filePath) {
|
||||
return false;
|
||||
}
|
||||
try {
|
||||
await fs.stat(filePath);
|
||||
return true;
|
||||
} catch {
|
||||
return false;
|
||||
}
|
||||
}
|
||||
|
||||
async function resolveCliOutput(params: {
|
||||
command: string;
|
||||
args: string[];
|
||||
stdout: string;
|
||||
mediaPath: string;
|
||||
}): Promise<string> {
|
||||
const commandId = commandBase(params.command);
|
||||
const fileOutput =
|
||||
commandId === "whisper-cli"
|
||||
? resolveWhisperCppOutputPath(params.args)
|
||||
: commandId === "whisper"
|
||||
? resolveWhisperOutputPath(params.args, params.mediaPath)
|
||||
: null;
|
||||
if (fileOutput && (await fileExists(fileOutput))) {
|
||||
try {
|
||||
const content = await fs.readFile(fileOutput, "utf8");
|
||||
if (content.trim()) {
|
||||
return content.trim();
|
||||
}
|
||||
} catch {}
|
||||
}
|
||||
|
||||
if (commandId === "gemini") {
|
||||
const response = extractGeminiResponse(params.stdout);
|
||||
if (response) {
|
||||
return response;
|
||||
}
|
||||
}
|
||||
|
||||
if (commandId === "sherpa-onnx-offline") {
|
||||
const response = extractSherpaOnnxText(params.stdout);
|
||||
if (response) {
|
||||
return response;
|
||||
}
|
||||
}
|
||||
|
||||
return params.stdout.trim();
|
||||
}
|
||||
|
||||
type ProviderQuery = Record<string, string | number | boolean>;
|
||||
|
||||
function normalizeProviderQuery(
|
||||
options?: Record<string, string | number | boolean>,
|
||||
): ProviderQuery | undefined {
|
||||
if (!options) {
|
||||
return undefined;
|
||||
}
|
||||
const query: ProviderQuery = {};
|
||||
for (const [key, value] of Object.entries(options)) {
|
||||
if (value === undefined) {
|
||||
continue;
|
||||
}
|
||||
query[key] = value;
|
||||
}
|
||||
return Object.keys(query).length > 0 ? query : undefined;
|
||||
}
|
||||
|
||||
function buildDeepgramCompatQuery(options?: {
|
||||
detectLanguage?: boolean;
|
||||
punctuate?: boolean;
|
||||
smartFormat?: boolean;
|
||||
}): ProviderQuery | undefined {
|
||||
if (!options) {
|
||||
return undefined;
|
||||
}
|
||||
const query: ProviderQuery = {};
|
||||
if (typeof options.detectLanguage === "boolean") {
|
||||
query.detect_language = options.detectLanguage;
|
||||
}
|
||||
if (typeof options.punctuate === "boolean") {
|
||||
query.punctuate = options.punctuate;
|
||||
}
|
||||
if (typeof options.smartFormat === "boolean") {
|
||||
query.smart_format = options.smartFormat;
|
||||
}
|
||||
return Object.keys(query).length > 0 ? query : undefined;
|
||||
}
|
||||
|
||||
function normalizeDeepgramQueryKeys(query: ProviderQuery): ProviderQuery {
|
||||
const normalized = { ...query };
|
||||
if ("detectLanguage" in normalized) {
|
||||
normalized.detect_language = normalized.detectLanguage as boolean;
|
||||
delete normalized.detectLanguage;
|
||||
}
|
||||
if ("smartFormat" in normalized) {
|
||||
normalized.smart_format = normalized.smartFormat as boolean;
|
||||
delete normalized.smartFormat;
|
||||
}
|
||||
return normalized;
|
||||
}
|
||||
|
||||
function resolveProviderQuery(params: {
|
||||
providerId: string;
|
||||
config?: MediaUnderstandingConfig;
|
||||
entry: MediaUnderstandingModelConfig;
|
||||
}): ProviderQuery | undefined {
|
||||
const { providerId, config, entry } = params;
|
||||
const mergedOptions = normalizeProviderQuery({
|
||||
...config?.providerOptions?.[providerId],
|
||||
...entry.providerOptions?.[providerId],
|
||||
});
|
||||
if (providerId !== "deepgram") {
|
||||
return mergedOptions;
|
||||
}
|
||||
const query = normalizeDeepgramQueryKeys(mergedOptions ?? {});
|
||||
const compat = buildDeepgramCompatQuery({ ...config?.deepgram, ...entry.deepgram });
|
||||
for (const [key, value] of Object.entries(compat ?? {})) {
|
||||
if (query[key] === undefined) {
|
||||
query[key] = value;
|
||||
}
|
||||
}
|
||||
return Object.keys(query).length > 0 ? query : undefined;
|
||||
}
|
||||
|
||||
export function buildModelDecision(params: {
|
||||
entry: MediaUnderstandingModelConfig;
|
||||
entryType: "provider" | "cli";
|
||||
outcome: MediaUnderstandingModelDecision["outcome"];
|
||||
reason?: string;
|
||||
}): MediaUnderstandingModelDecision {
|
||||
if (params.entryType === "cli") {
|
||||
const command = params.entry.command?.trim();
|
||||
return {
|
||||
type: "cli",
|
||||
provider: command ?? "cli",
|
||||
model: params.entry.model ?? command,
|
||||
outcome: params.outcome,
|
||||
reason: params.reason,
|
||||
};
|
||||
}
|
||||
const providerIdRaw = params.entry.provider?.trim();
|
||||
const providerId = providerIdRaw ? normalizeMediaProviderId(providerIdRaw) : undefined;
|
||||
return {
|
||||
type: "provider",
|
||||
provider: providerId ?? providerIdRaw,
|
||||
model: params.entry.model,
|
||||
outcome: params.outcome,
|
||||
reason: params.reason,
|
||||
};
|
||||
}
|
||||
|
||||
export function formatDecisionSummary(decision: MediaUnderstandingDecision): string {
|
||||
const total = decision.attachments.length;
|
||||
const success = decision.attachments.filter(
|
||||
(entry) => entry.chosen?.outcome === "success",
|
||||
).length;
|
||||
const chosen = decision.attachments.find((entry) => entry.chosen)?.chosen;
|
||||
const provider = chosen?.provider?.trim();
|
||||
const model = chosen?.model?.trim();
|
||||
const modelLabel = provider ? (model ? `${provider}/${model}` : provider) : undefined;
|
||||
const reason = decision.attachments
|
||||
.flatMap((entry) => entry.attempts.map((attempt) => attempt.reason).filter(Boolean))
|
||||
.find(Boolean);
|
||||
const shortReason = reason ? reason.split(":")[0]?.trim() : undefined;
|
||||
const countLabel = total > 0 ? ` (${success}/${total})` : "";
|
||||
const viaLabel = modelLabel ? ` via ${modelLabel}` : "";
|
||||
const reasonLabel = shortReason ? ` reason=${shortReason}` : "";
|
||||
return `${decision.capability}: ${decision.outcome}${countLabel}${viaLabel}${reasonLabel}`;
|
||||
}
|
||||
|
||||
export async function runProviderEntry(params: {
|
||||
capability: MediaUnderstandingCapability;
|
||||
entry: MediaUnderstandingModelConfig;
|
||||
cfg: OpenClawConfig;
|
||||
ctx: MsgContext;
|
||||
attachmentIndex: number;
|
||||
cache: MediaAttachmentCache;
|
||||
agentDir?: string;
|
||||
providerRegistry: ProviderRegistry;
|
||||
config?: MediaUnderstandingConfig;
|
||||
}): Promise<MediaUnderstandingOutput | null> {
|
||||
const { entry, capability, cfg } = params;
|
||||
const providerIdRaw = entry.provider?.trim();
|
||||
if (!providerIdRaw) {
|
||||
throw new Error(`Provider entry missing provider for ${capability}`);
|
||||
}
|
||||
const providerId = normalizeMediaProviderId(providerIdRaw);
|
||||
const maxBytes = resolveMaxBytes({ capability, entry, cfg, config: params.config });
|
||||
const maxChars = resolveMaxChars({ capability, entry, cfg, config: params.config });
|
||||
const timeoutMs = resolveTimeoutMs(
|
||||
entry.timeoutSeconds ??
|
||||
params.config?.timeoutSeconds ??
|
||||
cfg.tools?.media?.[capability]?.timeoutSeconds,
|
||||
DEFAULT_TIMEOUT_SECONDS[capability],
|
||||
);
|
||||
const prompt = resolvePrompt(
|
||||
capability,
|
||||
entry.prompt ?? params.config?.prompt ?? cfg.tools?.media?.[capability]?.prompt,
|
||||
maxChars,
|
||||
);
|
||||
|
||||
if (capability === "image") {
|
||||
if (!params.agentDir) {
|
||||
throw new Error("Image understanding requires agentDir");
|
||||
}
|
||||
const modelId = entry.model?.trim();
|
||||
if (!modelId) {
|
||||
throw new Error("Image understanding requires model id");
|
||||
}
|
||||
const media = await params.cache.getBuffer({
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
maxBytes,
|
||||
timeoutMs,
|
||||
});
|
||||
const provider = getMediaUnderstandingProvider(providerId, params.providerRegistry);
|
||||
const result = provider?.describeImage
|
||||
? await provider.describeImage({
|
||||
buffer: media.buffer,
|
||||
fileName: media.fileName,
|
||||
mime: media.mime,
|
||||
model: modelId,
|
||||
provider: providerId,
|
||||
prompt,
|
||||
timeoutMs,
|
||||
profile: entry.profile,
|
||||
preferredProfile: entry.preferredProfile,
|
||||
agentDir: params.agentDir,
|
||||
cfg: params.cfg,
|
||||
})
|
||||
: await describeImageWithModel({
|
||||
buffer: media.buffer,
|
||||
fileName: media.fileName,
|
||||
mime: media.mime,
|
||||
model: modelId,
|
||||
provider: providerId,
|
||||
prompt,
|
||||
timeoutMs,
|
||||
profile: entry.profile,
|
||||
preferredProfile: entry.preferredProfile,
|
||||
agentDir: params.agentDir,
|
||||
cfg: params.cfg,
|
||||
});
|
||||
return {
|
||||
kind: "image.description",
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
text: trimOutput(result.text, maxChars),
|
||||
provider: providerId,
|
||||
model: result.model ?? modelId,
|
||||
};
|
||||
}
|
||||
|
||||
const provider = getMediaUnderstandingProvider(providerId, params.providerRegistry);
|
||||
if (!provider) {
|
||||
throw new Error(`Media provider not available: ${providerId}`);
|
||||
}
|
||||
|
||||
if (capability === "audio") {
|
||||
if (!provider.transcribeAudio) {
|
||||
throw new Error(`Audio transcription provider "${providerId}" not available.`);
|
||||
}
|
||||
const media = await params.cache.getBuffer({
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
maxBytes,
|
||||
timeoutMs,
|
||||
});
|
||||
const auth = await resolveApiKeyForProvider({
|
||||
provider: providerId,
|
||||
cfg,
|
||||
profileId: entry.profile,
|
||||
preferredProfile: entry.preferredProfile,
|
||||
agentDir: params.agentDir,
|
||||
});
|
||||
const apiKey = requireApiKey(auth, providerId);
|
||||
const providerConfig = cfg.models?.providers?.[providerId];
|
||||
const baseUrl = entry.baseUrl ?? params.config?.baseUrl ?? providerConfig?.baseUrl;
|
||||
const mergedHeaders = {
|
||||
...providerConfig?.headers,
|
||||
...params.config?.headers,
|
||||
...entry.headers,
|
||||
};
|
||||
const headers = Object.keys(mergedHeaders).length > 0 ? mergedHeaders : undefined;
|
||||
const providerQuery = resolveProviderQuery({
|
||||
providerId,
|
||||
config: params.config,
|
||||
entry,
|
||||
});
|
||||
const model = entry.model?.trim() || DEFAULT_AUDIO_MODELS[providerId] || entry.model;
|
||||
const result = await provider.transcribeAudio({
|
||||
buffer: media.buffer,
|
||||
fileName: media.fileName,
|
||||
mime: media.mime,
|
||||
apiKey,
|
||||
baseUrl,
|
||||
headers,
|
||||
model,
|
||||
language: entry.language ?? params.config?.language ?? cfg.tools?.media?.audio?.language,
|
||||
prompt,
|
||||
query: providerQuery,
|
||||
timeoutMs,
|
||||
});
|
||||
return {
|
||||
kind: "audio.transcription",
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
text: trimOutput(result.text, maxChars),
|
||||
provider: providerId,
|
||||
model: result.model ?? model,
|
||||
};
|
||||
}
|
||||
|
||||
if (!provider.describeVideo) {
|
||||
throw new Error(`Video understanding provider "${providerId}" not available.`);
|
||||
}
|
||||
const media = await params.cache.getBuffer({
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
maxBytes,
|
||||
timeoutMs,
|
||||
});
|
||||
const estimatedBase64Bytes = estimateBase64Size(media.size);
|
||||
const maxBase64Bytes = resolveVideoMaxBase64Bytes(maxBytes);
|
||||
if (estimatedBase64Bytes > maxBase64Bytes) {
|
||||
throw new MediaUnderstandingSkipError(
|
||||
"maxBytes",
|
||||
`Video attachment ${params.attachmentIndex + 1} base64 payload ${estimatedBase64Bytes} exceeds ${maxBase64Bytes}`,
|
||||
);
|
||||
}
|
||||
const auth = await resolveApiKeyForProvider({
|
||||
provider: providerId,
|
||||
cfg,
|
||||
profileId: entry.profile,
|
||||
preferredProfile: entry.preferredProfile,
|
||||
agentDir: params.agentDir,
|
||||
});
|
||||
const apiKey = requireApiKey(auth, providerId);
|
||||
const providerConfig = cfg.models?.providers?.[providerId];
|
||||
const result = await provider.describeVideo({
|
||||
buffer: media.buffer,
|
||||
fileName: media.fileName,
|
||||
mime: media.mime,
|
||||
apiKey,
|
||||
baseUrl: providerConfig?.baseUrl,
|
||||
headers: providerConfig?.headers,
|
||||
model: entry.model,
|
||||
prompt,
|
||||
timeoutMs,
|
||||
});
|
||||
return {
|
||||
kind: "video.description",
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
text: trimOutput(result.text, maxChars),
|
||||
provider: providerId,
|
||||
model: result.model ?? entry.model,
|
||||
};
|
||||
}
|
||||
|
||||
export async function runCliEntry(params: {
|
||||
capability: MediaUnderstandingCapability;
|
||||
entry: MediaUnderstandingModelConfig;
|
||||
cfg: OpenClawConfig;
|
||||
ctx: MsgContext;
|
||||
attachmentIndex: number;
|
||||
cache: MediaAttachmentCache;
|
||||
config?: MediaUnderstandingConfig;
|
||||
}): Promise<MediaUnderstandingOutput | null> {
|
||||
const { entry, capability, cfg, ctx } = params;
|
||||
const command = entry.command?.trim();
|
||||
const args = entry.args ?? [];
|
||||
if (!command) {
|
||||
throw new Error(`CLI entry missing command for ${capability}`);
|
||||
}
|
||||
const maxBytes = resolveMaxBytes({ capability, entry, cfg, config: params.config });
|
||||
const maxChars = resolveMaxChars({ capability, entry, cfg, config: params.config });
|
||||
const timeoutMs = resolveTimeoutMs(
|
||||
entry.timeoutSeconds ??
|
||||
params.config?.timeoutSeconds ??
|
||||
cfg.tools?.media?.[capability]?.timeoutSeconds,
|
||||
DEFAULT_TIMEOUT_SECONDS[capability],
|
||||
);
|
||||
const prompt = resolvePrompt(
|
||||
capability,
|
||||
entry.prompt ?? params.config?.prompt ?? cfg.tools?.media?.[capability]?.prompt,
|
||||
maxChars,
|
||||
);
|
||||
const pathResult = await params.cache.getPath({
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
maxBytes,
|
||||
timeoutMs,
|
||||
});
|
||||
const outputDir = await fs.mkdtemp(path.join(os.tmpdir(), "openclaw-media-cli-"));
|
||||
const mediaPath = pathResult.path;
|
||||
const outputBase = path.join(outputDir, path.parse(mediaPath).name);
|
||||
|
||||
const templCtx: MsgContext = {
|
||||
...ctx,
|
||||
MediaPath: mediaPath,
|
||||
MediaDir: path.dirname(mediaPath),
|
||||
OutputDir: outputDir,
|
||||
OutputBase: outputBase,
|
||||
Prompt: prompt,
|
||||
MaxChars: maxChars,
|
||||
};
|
||||
const argv = [command, ...args].map((part, index) =>
|
||||
index === 0 ? part : applyTemplate(part, templCtx),
|
||||
);
|
||||
try {
|
||||
if (shouldLogVerbose()) {
|
||||
logVerbose(`Media understanding via CLI: ${argv.join(" ")}`);
|
||||
}
|
||||
const { stdout } = await runExec(argv[0], argv.slice(1), {
|
||||
timeoutMs,
|
||||
maxBuffer: CLI_OUTPUT_MAX_BUFFER,
|
||||
});
|
||||
const resolved = await resolveCliOutput({
|
||||
command,
|
||||
args: argv.slice(1),
|
||||
stdout,
|
||||
mediaPath,
|
||||
});
|
||||
const text = trimOutput(resolved, maxChars);
|
||||
if (!text) {
|
||||
return null;
|
||||
}
|
||||
return {
|
||||
kind: capability === "audio" ? "audio.transcription" : `${capability}.description`,
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
text,
|
||||
provider: "cli",
|
||||
model: command,
|
||||
};
|
||||
} finally {
|
||||
await fs.rm(outputDir, { recursive: true, force: true }).catch(() => {});
|
||||
}
|
||||
}
|
||||
@@ -16,13 +16,12 @@ import type {
|
||||
MediaUnderstandingOutput,
|
||||
MediaUnderstandingProvider,
|
||||
} from "./types.js";
|
||||
import { requireApiKey, resolveApiKeyForProvider } from "../agents/model-auth.js";
|
||||
import { resolveApiKeyForProvider } from "../agents/model-auth.js";
|
||||
import {
|
||||
findModelInCatalog,
|
||||
loadModelCatalog,
|
||||
modelSupportsVision,
|
||||
} from "../agents/model-catalog.js";
|
||||
import { applyTemplate } from "../auto-reply/templating.js";
|
||||
import { logVerbose, shouldLogVerbose } from "../globals.js";
|
||||
import { runExec } from "../process/exec.js";
|
||||
import { MediaAttachmentCache, normalizeAttachments, selectAttachments } from "./attachments.js";
|
||||
@@ -30,27 +29,21 @@ import {
|
||||
AUTO_AUDIO_KEY_PROVIDERS,
|
||||
AUTO_IMAGE_KEY_PROVIDERS,
|
||||
AUTO_VIDEO_KEY_PROVIDERS,
|
||||
CLI_OUTPUT_MAX_BUFFER,
|
||||
DEFAULT_AUDIO_MODELS,
|
||||
DEFAULT_IMAGE_MODELS,
|
||||
DEFAULT_TIMEOUT_SECONDS,
|
||||
} from "./defaults.js";
|
||||
import { isMediaUnderstandingSkipError, MediaUnderstandingSkipError } from "./errors.js";
|
||||
import { describeImageWithModel } from "./providers/image.js";
|
||||
import { isMediaUnderstandingSkipError } from "./errors.js";
|
||||
import {
|
||||
buildMediaUnderstandingRegistry,
|
||||
getMediaUnderstandingProvider,
|
||||
normalizeMediaProviderId,
|
||||
} from "./providers/index.js";
|
||||
import { resolveModelEntries, resolveScopeDecision } from "./resolve.js";
|
||||
import {
|
||||
resolveMaxBytes,
|
||||
resolveMaxChars,
|
||||
resolveModelEntries,
|
||||
resolvePrompt,
|
||||
resolveScopeDecision,
|
||||
resolveTimeoutMs,
|
||||
} from "./resolve.js";
|
||||
import { estimateBase64Size, resolveVideoMaxBase64Bytes } from "./video.js";
|
||||
buildModelDecision,
|
||||
formatDecisionSummary,
|
||||
runCliEntry,
|
||||
runProviderEntry,
|
||||
} from "./runner.entries.js";
|
||||
|
||||
export type ActiveMediaModel = {
|
||||
provider: string;
|
||||
@@ -220,49 +213,6 @@ function extractGeminiResponse(raw: string): string | null {
|
||||
return trimmed || null;
|
||||
}
|
||||
|
||||
function extractSherpaOnnxText(raw: string): string | null {
|
||||
const tryParse = (value: string): string | null => {
|
||||
const trimmed = value.trim();
|
||||
if (!trimmed) {
|
||||
return null;
|
||||
}
|
||||
const head = trimmed[0];
|
||||
if (head !== "{" && head !== '"') {
|
||||
return null;
|
||||
}
|
||||
try {
|
||||
const parsed = JSON.parse(trimmed) as unknown;
|
||||
if (typeof parsed === "string") {
|
||||
return tryParse(parsed);
|
||||
}
|
||||
if (parsed && typeof parsed === "object") {
|
||||
const text = (parsed as { text?: unknown }).text;
|
||||
if (typeof text === "string" && text.trim()) {
|
||||
return text.trim();
|
||||
}
|
||||
}
|
||||
} catch {}
|
||||
return null;
|
||||
};
|
||||
|
||||
const direct = tryParse(raw);
|
||||
if (direct) {
|
||||
return direct;
|
||||
}
|
||||
|
||||
const lines = raw
|
||||
.split("\n")
|
||||
.map((line) => line.trim())
|
||||
.filter(Boolean);
|
||||
for (let i = lines.length - 1; i >= 0; i -= 1) {
|
||||
const parsed = tryParse(lines[i] ?? "");
|
||||
if (parsed) {
|
||||
return parsed;
|
||||
}
|
||||
}
|
||||
return null;
|
||||
}
|
||||
|
||||
async function probeGeminiCli(): Promise<boolean> {
|
||||
const cached = geminiProbeCache.get("gemini");
|
||||
if (cached) {
|
||||
@@ -591,482 +541,6 @@ async function resolveActiveModelEntry(params: {
|
||||
};
|
||||
}
|
||||
|
||||
function trimOutput(text: string, maxChars?: number): string {
|
||||
const trimmed = text.trim();
|
||||
if (!maxChars || trimmed.length <= maxChars) {
|
||||
return trimmed;
|
||||
}
|
||||
return trimmed.slice(0, maxChars).trim();
|
||||
}
|
||||
|
||||
function commandBase(command: string): string {
|
||||
return path.parse(command).name;
|
||||
}
|
||||
|
||||
function findArgValue(args: string[], keys: string[]): string | undefined {
|
||||
for (let i = 0; i < args.length; i += 1) {
|
||||
if (keys.includes(args[i] ?? "")) {
|
||||
const value = args[i + 1];
|
||||
if (value) {
|
||||
return value;
|
||||
}
|
||||
}
|
||||
}
|
||||
return undefined;
|
||||
}
|
||||
|
||||
function hasArg(args: string[], keys: string[]): boolean {
|
||||
return args.some((arg) => keys.includes(arg));
|
||||
}
|
||||
|
||||
function resolveWhisperOutputPath(args: string[], mediaPath: string): string | null {
|
||||
const outputDir = findArgValue(args, ["--output_dir", "-o"]);
|
||||
const outputFormat = findArgValue(args, ["--output_format"]);
|
||||
if (!outputDir || !outputFormat) {
|
||||
return null;
|
||||
}
|
||||
const formats = outputFormat.split(",").map((value) => value.trim());
|
||||
if (!formats.includes("txt")) {
|
||||
return null;
|
||||
}
|
||||
const base = path.parse(mediaPath).name;
|
||||
return path.join(outputDir, `${base}.txt`);
|
||||
}
|
||||
|
||||
function resolveWhisperCppOutputPath(args: string[]): string | null {
|
||||
if (!hasArg(args, ["-otxt", "--output-txt"])) {
|
||||
return null;
|
||||
}
|
||||
const outputBase = findArgValue(args, ["-of", "--output-file"]);
|
||||
if (!outputBase) {
|
||||
return null;
|
||||
}
|
||||
return `${outputBase}.txt`;
|
||||
}
|
||||
|
||||
async function resolveCliOutput(params: {
|
||||
command: string;
|
||||
args: string[];
|
||||
stdout: string;
|
||||
mediaPath: string;
|
||||
}): Promise<string> {
|
||||
const commandId = commandBase(params.command);
|
||||
const fileOutput =
|
||||
commandId === "whisper-cli"
|
||||
? resolveWhisperCppOutputPath(params.args)
|
||||
: commandId === "whisper"
|
||||
? resolveWhisperOutputPath(params.args, params.mediaPath)
|
||||
: null;
|
||||
if (fileOutput && (await fileExists(fileOutput))) {
|
||||
try {
|
||||
const content = await fs.readFile(fileOutput, "utf8");
|
||||
if (content.trim()) {
|
||||
return content.trim();
|
||||
}
|
||||
} catch {}
|
||||
}
|
||||
|
||||
if (commandId === "gemini") {
|
||||
const response = extractGeminiResponse(params.stdout);
|
||||
if (response) {
|
||||
return response;
|
||||
}
|
||||
}
|
||||
|
||||
if (commandId === "sherpa-onnx-offline") {
|
||||
const response = extractSherpaOnnxText(params.stdout);
|
||||
if (response) {
|
||||
return response;
|
||||
}
|
||||
}
|
||||
|
||||
return params.stdout.trim();
|
||||
}
|
||||
|
||||
type ProviderQuery = Record<string, string | number | boolean>;
|
||||
|
||||
function normalizeProviderQuery(
|
||||
options?: Record<string, string | number | boolean>,
|
||||
): ProviderQuery | undefined {
|
||||
if (!options) {
|
||||
return undefined;
|
||||
}
|
||||
const query: ProviderQuery = {};
|
||||
for (const [key, value] of Object.entries(options)) {
|
||||
if (value === undefined) {
|
||||
continue;
|
||||
}
|
||||
query[key] = value;
|
||||
}
|
||||
return Object.keys(query).length > 0 ? query : undefined;
|
||||
}
|
||||
|
||||
function buildDeepgramCompatQuery(options?: {
|
||||
detectLanguage?: boolean;
|
||||
punctuate?: boolean;
|
||||
smartFormat?: boolean;
|
||||
}): ProviderQuery | undefined {
|
||||
if (!options) {
|
||||
return undefined;
|
||||
}
|
||||
const query: ProviderQuery = {};
|
||||
if (typeof options.detectLanguage === "boolean") {
|
||||
query.detect_language = options.detectLanguage;
|
||||
}
|
||||
if (typeof options.punctuate === "boolean") {
|
||||
query.punctuate = options.punctuate;
|
||||
}
|
||||
if (typeof options.smartFormat === "boolean") {
|
||||
query.smart_format = options.smartFormat;
|
||||
}
|
||||
return Object.keys(query).length > 0 ? query : undefined;
|
||||
}
|
||||
|
||||
function normalizeDeepgramQueryKeys(query: ProviderQuery): ProviderQuery {
|
||||
const normalized = { ...query };
|
||||
if ("detectLanguage" in normalized) {
|
||||
normalized.detect_language = normalized.detectLanguage as boolean;
|
||||
delete normalized.detectLanguage;
|
||||
}
|
||||
if ("smartFormat" in normalized) {
|
||||
normalized.smart_format = normalized.smartFormat as boolean;
|
||||
delete normalized.smartFormat;
|
||||
}
|
||||
return normalized;
|
||||
}
|
||||
|
||||
function resolveProviderQuery(params: {
|
||||
providerId: string;
|
||||
config?: MediaUnderstandingConfig;
|
||||
entry: MediaUnderstandingModelConfig;
|
||||
}): ProviderQuery | undefined {
|
||||
const { providerId, config, entry } = params;
|
||||
const mergedOptions = normalizeProviderQuery({
|
||||
...config?.providerOptions?.[providerId],
|
||||
...entry.providerOptions?.[providerId],
|
||||
});
|
||||
if (providerId !== "deepgram") {
|
||||
return mergedOptions;
|
||||
}
|
||||
let query = normalizeDeepgramQueryKeys(mergedOptions ?? {});
|
||||
const compat = buildDeepgramCompatQuery({ ...config?.deepgram, ...entry.deepgram });
|
||||
for (const [key, value] of Object.entries(compat ?? {})) {
|
||||
if (query[key] === undefined) {
|
||||
query[key] = value;
|
||||
}
|
||||
}
|
||||
return Object.keys(query).length > 0 ? query : undefined;
|
||||
}
|
||||
|
||||
function buildModelDecision(params: {
|
||||
entry: MediaUnderstandingModelConfig;
|
||||
entryType: "provider" | "cli";
|
||||
outcome: MediaUnderstandingModelDecision["outcome"];
|
||||
reason?: string;
|
||||
}): MediaUnderstandingModelDecision {
|
||||
if (params.entryType === "cli") {
|
||||
const command = params.entry.command?.trim();
|
||||
return {
|
||||
type: "cli",
|
||||
provider: command ?? "cli",
|
||||
model: params.entry.model ?? command,
|
||||
outcome: params.outcome,
|
||||
reason: params.reason,
|
||||
};
|
||||
}
|
||||
const providerIdRaw = params.entry.provider?.trim();
|
||||
const providerId = providerIdRaw ? normalizeMediaProviderId(providerIdRaw) : undefined;
|
||||
return {
|
||||
type: "provider",
|
||||
provider: providerId ?? providerIdRaw,
|
||||
model: params.entry.model,
|
||||
outcome: params.outcome,
|
||||
reason: params.reason,
|
||||
};
|
||||
}
|
||||
|
||||
function formatDecisionSummary(decision: MediaUnderstandingDecision): string {
|
||||
const total = decision.attachments.length;
|
||||
const success = decision.attachments.filter(
|
||||
(entry) => entry.chosen?.outcome === "success",
|
||||
).length;
|
||||
const chosen = decision.attachments.find((entry) => entry.chosen)?.chosen;
|
||||
const provider = chosen?.provider?.trim();
|
||||
const model = chosen?.model?.trim();
|
||||
const modelLabel = provider ? (model ? `${provider}/${model}` : provider) : undefined;
|
||||
const reason = decision.attachments
|
||||
.flatMap((entry) => entry.attempts.map((attempt) => attempt.reason).filter(Boolean))
|
||||
.find(Boolean);
|
||||
const shortReason = reason ? reason.split(":")[0]?.trim() : undefined;
|
||||
const countLabel = total > 0 ? ` (${success}/${total})` : "";
|
||||
const viaLabel = modelLabel ? ` via ${modelLabel}` : "";
|
||||
const reasonLabel = shortReason ? ` reason=${shortReason}` : "";
|
||||
return `${decision.capability}: ${decision.outcome}${countLabel}${viaLabel}${reasonLabel}`;
|
||||
}
|
||||
|
||||
async function runProviderEntry(params: {
|
||||
capability: MediaUnderstandingCapability;
|
||||
entry: MediaUnderstandingModelConfig;
|
||||
cfg: OpenClawConfig;
|
||||
ctx: MsgContext;
|
||||
attachmentIndex: number;
|
||||
cache: MediaAttachmentCache;
|
||||
agentDir?: string;
|
||||
providerRegistry: ProviderRegistry;
|
||||
config?: MediaUnderstandingConfig;
|
||||
}): Promise<MediaUnderstandingOutput | null> {
|
||||
const { entry, capability, cfg } = params;
|
||||
const providerIdRaw = entry.provider?.trim();
|
||||
if (!providerIdRaw) {
|
||||
throw new Error(`Provider entry missing provider for ${capability}`);
|
||||
}
|
||||
const providerId = normalizeMediaProviderId(providerIdRaw);
|
||||
const maxBytes = resolveMaxBytes({ capability, entry, cfg, config: params.config });
|
||||
const maxChars = resolveMaxChars({ capability, entry, cfg, config: params.config });
|
||||
const timeoutMs = resolveTimeoutMs(
|
||||
entry.timeoutSeconds ??
|
||||
params.config?.timeoutSeconds ??
|
||||
cfg.tools?.media?.[capability]?.timeoutSeconds,
|
||||
DEFAULT_TIMEOUT_SECONDS[capability],
|
||||
);
|
||||
const prompt = resolvePrompt(
|
||||
capability,
|
||||
entry.prompt ?? params.config?.prompt ?? cfg.tools?.media?.[capability]?.prompt,
|
||||
maxChars,
|
||||
);
|
||||
|
||||
if (capability === "image") {
|
||||
if (!params.agentDir) {
|
||||
throw new Error("Image understanding requires agentDir");
|
||||
}
|
||||
const modelId = entry.model?.trim();
|
||||
if (!modelId) {
|
||||
throw new Error("Image understanding requires model id");
|
||||
}
|
||||
const media = await params.cache.getBuffer({
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
maxBytes,
|
||||
timeoutMs,
|
||||
});
|
||||
const provider = getMediaUnderstandingProvider(providerId, params.providerRegistry);
|
||||
const result = provider?.describeImage
|
||||
? await provider.describeImage({
|
||||
buffer: media.buffer,
|
||||
fileName: media.fileName,
|
||||
mime: media.mime,
|
||||
model: modelId,
|
||||
provider: providerId,
|
||||
prompt,
|
||||
timeoutMs,
|
||||
profile: entry.profile,
|
||||
preferredProfile: entry.preferredProfile,
|
||||
agentDir: params.agentDir,
|
||||
cfg: params.cfg,
|
||||
})
|
||||
: await describeImageWithModel({
|
||||
buffer: media.buffer,
|
||||
fileName: media.fileName,
|
||||
mime: media.mime,
|
||||
model: modelId,
|
||||
provider: providerId,
|
||||
prompt,
|
||||
timeoutMs,
|
||||
profile: entry.profile,
|
||||
preferredProfile: entry.preferredProfile,
|
||||
agentDir: params.agentDir,
|
||||
cfg: params.cfg,
|
||||
});
|
||||
return {
|
||||
kind: "image.description",
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
text: trimOutput(result.text, maxChars),
|
||||
provider: providerId,
|
||||
model: result.model ?? modelId,
|
||||
};
|
||||
}
|
||||
|
||||
const provider = getMediaUnderstandingProvider(providerId, params.providerRegistry);
|
||||
if (!provider) {
|
||||
throw new Error(`Media provider not available: ${providerId}`);
|
||||
}
|
||||
|
||||
if (capability === "audio") {
|
||||
if (!provider.transcribeAudio) {
|
||||
throw new Error(`Audio transcription provider "${providerId}" not available.`);
|
||||
}
|
||||
const media = await params.cache.getBuffer({
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
maxBytes,
|
||||
timeoutMs,
|
||||
});
|
||||
const auth = await resolveApiKeyForProvider({
|
||||
provider: providerId,
|
||||
cfg,
|
||||
profileId: entry.profile,
|
||||
preferredProfile: entry.preferredProfile,
|
||||
agentDir: params.agentDir,
|
||||
});
|
||||
const apiKey = requireApiKey(auth, providerId);
|
||||
const providerConfig = cfg.models?.providers?.[providerId];
|
||||
const baseUrl = entry.baseUrl ?? params.config?.baseUrl ?? providerConfig?.baseUrl;
|
||||
const mergedHeaders = {
|
||||
...providerConfig?.headers,
|
||||
...params.config?.headers,
|
||||
...entry.headers,
|
||||
};
|
||||
const headers = Object.keys(mergedHeaders).length > 0 ? mergedHeaders : undefined;
|
||||
const providerQuery = resolveProviderQuery({
|
||||
providerId,
|
||||
config: params.config,
|
||||
entry,
|
||||
});
|
||||
const model = entry.model?.trim() || DEFAULT_AUDIO_MODELS[providerId] || entry.model;
|
||||
const result = await provider.transcribeAudio({
|
||||
buffer: media.buffer,
|
||||
fileName: media.fileName,
|
||||
mime: media.mime,
|
||||
apiKey,
|
||||
baseUrl,
|
||||
headers,
|
||||
model,
|
||||
language: entry.language ?? params.config?.language ?? cfg.tools?.media?.audio?.language,
|
||||
prompt,
|
||||
query: providerQuery,
|
||||
timeoutMs,
|
||||
});
|
||||
return {
|
||||
kind: "audio.transcription",
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
text: trimOutput(result.text, maxChars),
|
||||
provider: providerId,
|
||||
model: result.model ?? model,
|
||||
};
|
||||
}
|
||||
|
||||
if (!provider.describeVideo) {
|
||||
throw new Error(`Video understanding provider "${providerId}" not available.`);
|
||||
}
|
||||
const media = await params.cache.getBuffer({
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
maxBytes,
|
||||
timeoutMs,
|
||||
});
|
||||
const estimatedBase64Bytes = estimateBase64Size(media.size);
|
||||
const maxBase64Bytes = resolveVideoMaxBase64Bytes(maxBytes);
|
||||
if (estimatedBase64Bytes > maxBase64Bytes) {
|
||||
throw new MediaUnderstandingSkipError(
|
||||
"maxBytes",
|
||||
`Video attachment ${params.attachmentIndex + 1} base64 payload ${estimatedBase64Bytes} exceeds ${maxBase64Bytes}`,
|
||||
);
|
||||
}
|
||||
const auth = await resolveApiKeyForProvider({
|
||||
provider: providerId,
|
||||
cfg,
|
||||
profileId: entry.profile,
|
||||
preferredProfile: entry.preferredProfile,
|
||||
agentDir: params.agentDir,
|
||||
});
|
||||
const apiKey = requireApiKey(auth, providerId);
|
||||
const providerConfig = cfg.models?.providers?.[providerId];
|
||||
const result = await provider.describeVideo({
|
||||
buffer: media.buffer,
|
||||
fileName: media.fileName,
|
||||
mime: media.mime,
|
||||
apiKey,
|
||||
baseUrl: providerConfig?.baseUrl,
|
||||
headers: providerConfig?.headers,
|
||||
model: entry.model,
|
||||
prompt,
|
||||
timeoutMs,
|
||||
});
|
||||
return {
|
||||
kind: "video.description",
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
text: trimOutput(result.text, maxChars),
|
||||
provider: providerId,
|
||||
model: result.model ?? entry.model,
|
||||
};
|
||||
}
|
||||
|
||||
async function runCliEntry(params: {
|
||||
capability: MediaUnderstandingCapability;
|
||||
entry: MediaUnderstandingModelConfig;
|
||||
cfg: OpenClawConfig;
|
||||
ctx: MsgContext;
|
||||
attachmentIndex: number;
|
||||
cache: MediaAttachmentCache;
|
||||
config?: MediaUnderstandingConfig;
|
||||
}): Promise<MediaUnderstandingOutput | null> {
|
||||
const { entry, capability, cfg, ctx } = params;
|
||||
const command = entry.command?.trim();
|
||||
const args = entry.args ?? [];
|
||||
if (!command) {
|
||||
throw new Error(`CLI entry missing command for ${capability}`);
|
||||
}
|
||||
const maxBytes = resolveMaxBytes({ capability, entry, cfg, config: params.config });
|
||||
const maxChars = resolveMaxChars({ capability, entry, cfg, config: params.config });
|
||||
const timeoutMs = resolveTimeoutMs(
|
||||
entry.timeoutSeconds ??
|
||||
params.config?.timeoutSeconds ??
|
||||
cfg.tools?.media?.[capability]?.timeoutSeconds,
|
||||
DEFAULT_TIMEOUT_SECONDS[capability],
|
||||
);
|
||||
const prompt = resolvePrompt(
|
||||
capability,
|
||||
entry.prompt ?? params.config?.prompt ?? cfg.tools?.media?.[capability]?.prompt,
|
||||
maxChars,
|
||||
);
|
||||
const pathResult = await params.cache.getPath({
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
maxBytes,
|
||||
timeoutMs,
|
||||
});
|
||||
const outputDir = await fs.mkdtemp(path.join(os.tmpdir(), "openclaw-media-cli-"));
|
||||
const mediaPath = pathResult.path;
|
||||
const outputBase = path.join(outputDir, path.parse(mediaPath).name);
|
||||
|
||||
const templCtx: MsgContext = {
|
||||
...ctx,
|
||||
MediaPath: mediaPath,
|
||||
MediaDir: path.dirname(mediaPath),
|
||||
OutputDir: outputDir,
|
||||
OutputBase: outputBase,
|
||||
Prompt: prompt,
|
||||
MaxChars: maxChars,
|
||||
};
|
||||
const argv = [command, ...args].map((part, index) =>
|
||||
index === 0 ? part : applyTemplate(part, templCtx),
|
||||
);
|
||||
try {
|
||||
if (shouldLogVerbose()) {
|
||||
logVerbose(`Media understanding via CLI: ${argv.join(" ")}`);
|
||||
}
|
||||
const { stdout } = await runExec(argv[0], argv.slice(1), {
|
||||
timeoutMs,
|
||||
maxBuffer: CLI_OUTPUT_MAX_BUFFER,
|
||||
});
|
||||
const resolved = await resolveCliOutput({
|
||||
command,
|
||||
args: argv.slice(1),
|
||||
stdout,
|
||||
mediaPath,
|
||||
});
|
||||
const text = trimOutput(resolved, maxChars);
|
||||
if (!text) {
|
||||
return null;
|
||||
}
|
||||
return {
|
||||
kind: capability === "audio" ? "audio.transcription" : `${capability}.description`,
|
||||
attachmentIndex: params.attachmentIndex,
|
||||
text,
|
||||
provider: "cli",
|
||||
model: command,
|
||||
};
|
||||
} finally {
|
||||
await fs.rm(outputDir, { recursive: true, force: true }).catch(() => {});
|
||||
}
|
||||
}
|
||||
|
||||
async function runAttachmentEntries(params: {
|
||||
capability: MediaUnderstandingCapability;
|
||||
cfg: OpenClawConfig;
|
||||
|
||||
Reference in New Issue
Block a user