guide on compile document in build time

[himself65]

Ask from CityJS, someone asks about this. I know we have generate-data script in create-llama, but I'm not sure if we should do this as a module or give a guide on this

[himself65]

I added one ad-hoc plugin in our document here

LlamaIndexTS/apps/next/source.config.ts

Lines 52 to 112 in efb7e1b

	() => {
	return (_, file, next) => {
	const metadata = file.data.frontmatter as Record<string, unknown>;
	const title = metadata.title as string;
	const description = metadata.description as string;
	let content: string;
	if (file.value instanceof Uint8Array) {
	content = file.value.toString();
	} else {
	content = file.value;
	}
	if (file.path.includes("content/docs/cloud/api")) {
	// skip cloud api docs
	return next();
	}
	// eslint-disable-next-line turbo/no-undeclared-env-vars
	if (process.env.NODE_ENV === "development") {
	// skip development
	return next();
	}
	if (!title || !description) {
	throw new Error(`Missing title or description in ${file.path}`);
	}
	const id = relative(baseDir, file.path);
	if (
	// eslint-disable-next-line turbo/no-undeclared-env-vars
	process.env.LLAMA_CLOUD_UPSERT_PIPELINE_DOCUMENTS === "true" &&
	// eslint-disable-next-line turbo/no-undeclared-env-vars
	process.env.LLAMA_CLOUD_PIPELINE_ID !== undefined
	) {
	PipelinesService.upsertBatchPipelineDocumentsApiV1PipelinesPipelineIdDocumentsPut(
	{
	baseUrl: "https://api.cloud.llamaindex.ai/",
	body: [
	{
	metadata: {
	title,
	description,
	documentUrl: id,
	},
	text: content,
	id,
	},
	],
	path: {
	// eslint-disable-next-line turbo/no-undeclared-env-vars
	pipeline_id: process.env.LLAMA_CLOUD_PIPELINE_ID,
	},
	throwOnError: true,
	headers: {
	// eslint-disable-next-line turbo/no-undeclared-env-vars
	Authorization: `Bearer ${process.env.LLAMA_CLOUD_API_KEY}`,
	},
	},
	).catch((error) => {
	console.error(error);
	});
	}
	return next();
	};