feat(ai): derive generated model types from JSON
This commit is contained in:
Armin Ronacher
@@ -23,7 +23,7 @@ import {
|
||||
createModelDataManifest,
|
||||
type ModelDataStructure,
|
||||
MODEL_DATA_MANIFEST_FILE,
|
||||
readModelDataStructure,
|
||||
readModelDataProviderIds,
|
||||
validateGeneratedModelData,
|
||||
validateModelDataDirectory,
|
||||
} from "./model-data.ts";
|
||||
@@ -2482,41 +2482,30 @@ async function generateModels() {
|
||||
|
||||
const serializeJson = (value: unknown) => `${JSON.stringify(value, null, generatorOptions.pretty ? 2 : undefined)}\n`;
|
||||
const writeJson = (path: string, value: unknown) => writeFileSync(path, serializeJson(value));
|
||||
let generatedDataProviderIds = sortedProviderIds;
|
||||
let generatedDataProviders = jsonProviders;
|
||||
let modelDataStructure: ModelDataStructure = Object.fromEntries(
|
||||
sortedProviderIds.map((providerId) => [
|
||||
providerId,
|
||||
Object.fromEntries(
|
||||
Object.entries(jsonProviders[providerId]).map(([modelId, model]) => [modelId, model.api]),
|
||||
),
|
||||
]),
|
||||
);
|
||||
const generatedDataProviderIds = generatorOptions.dataOnly
|
||||
? readModelDataProviderIds(packageRoot)
|
||||
: sortedProviderIds;
|
||||
const missingProviderIds = generatedDataProviderIds.filter((providerId) => !jsonProviders[providerId]);
|
||||
if (missingProviderIds.length > 0) {
|
||||
throw new Error(`Cannot hydrate missing providers: ${missingProviderIds.join(", ")}`);
|
||||
}
|
||||
|
||||
if (generatorOptions.dataOnly) {
|
||||
modelDataStructure = readModelDataStructure(packageRoot);
|
||||
generatedDataProviderIds = Object.keys(modelDataStructure);
|
||||
const hydratedProviders: typeof jsonProviders = {};
|
||||
const hydrationErrors: string[] = [];
|
||||
for (const [providerId, expectedModels] of Object.entries(modelDataStructure)) {
|
||||
hydratedProviders[providerId] = {};
|
||||
for (const [modelId, expectedApi] of Object.entries(expectedModels)) {
|
||||
const model = jsonProviders[providerId]?.[modelId];
|
||||
if (!model) {
|
||||
hydrationErrors.push(`missing ${providerId}/${modelId}`);
|
||||
continue;
|
||||
}
|
||||
if (model.api !== expectedApi) {
|
||||
hydrationErrors.push(`${providerId}/${modelId} uses ${model.api}, expected ${expectedApi}`);
|
||||
continue;
|
||||
}
|
||||
hydratedProviders[providerId][modelId] = model;
|
||||
// Only the ignored internal data is grouped by API for type derivation. Public JSON catalog output stays flat.
|
||||
const generatedDataProviders: Record<string, Record<string, Record<string, Model<Api>>>> = {};
|
||||
const modelDataStructure: ModelDataStructure = {};
|
||||
for (const providerId of generatedDataProviderIds) {
|
||||
const models = jsonProviders[providerId];
|
||||
generatedDataProviders[providerId] = {};
|
||||
modelDataStructure[providerId] = {};
|
||||
const apiIds = Array.from(new Set(Object.values(models).map((model) => model.api))).sort();
|
||||
for (const api of apiIds) {
|
||||
generatedDataProviders[providerId][api] = {};
|
||||
for (const [modelId, model] of Object.entries(models)) {
|
||||
if (model.api !== api) continue;
|
||||
generatedDataProviders[providerId][api][modelId] = model;
|
||||
modelDataStructure[providerId][modelId] = api;
|
||||
}
|
||||
}
|
||||
if (hydrationErrors.length > 0) {
|
||||
throw new Error(`Cannot hydrate the committed model catalog:\n${hydrationErrors.map((error) => ` - ${error}`).join("\n")}`);
|
||||
}
|
||||
generatedDataProviders = hydratedProviders;
|
||||
}
|
||||
|
||||
if (!generatorOptions.jsonOnly) {
|
||||
@@ -2526,7 +2515,7 @@ async function generateModels() {
|
||||
const stagingRoot = mkdtempSync(join(providersDir, ".model-generation-"));
|
||||
const stagedDataDir = join(stagingRoot, "data");
|
||||
const previousDataDir = join(stagingRoot, "previous-data");
|
||||
let restoreStructuralCatalog: (() => void) | undefined;
|
||||
let restoreGeneratedCatalog: (() => void) | undefined;
|
||||
try {
|
||||
mkdirSync(stagedDataDir, { recursive: true });
|
||||
const fileContents: Record<string, string> = {};
|
||||
@@ -2543,7 +2532,6 @@ async function generateModels() {
|
||||
validateModelDataDirectory(modelDataStructure, stagedDataDir);
|
||||
|
||||
if (!generatorOptions.dataOnly) {
|
||||
// Generate TypeScript structural catalogs only after the model data is complete and valid.
|
||||
const previousShardContents = new Map(
|
||||
readdirSync(providersDir)
|
||||
.filter((entry) => entry.endsWith(".models.ts"))
|
||||
@@ -2551,7 +2539,7 @@ async function generateModels() {
|
||||
);
|
||||
const aggregatorPath = join(packageRoot, "src/models.generated.ts");
|
||||
const previousAggregator = readFileSync(aggregatorPath, "utf8");
|
||||
restoreStructuralCatalog = () => {
|
||||
restoreGeneratedCatalog = () => {
|
||||
for (const entry of readdirSync(providersDir)) {
|
||||
if (entry.endsWith(".models.ts")) rmSync(join(providersDir, entry));
|
||||
}
|
||||
@@ -2568,21 +2556,12 @@ async function generateModels() {
|
||||
const catalogConstName = (providerId: string) =>
|
||||
`${providerId.toUpperCase().replace(/[^A-Z0-9]+/g, "_")}_MODELS`;
|
||||
const generatedShardFiles = new Set<string>();
|
||||
|
||||
function emitModelShape(model: Model<any>, indent: string): string {
|
||||
return `${indent}${JSON.stringify(model.id)}: Model<${JSON.stringify(model.api)}> & {\n${indent}\tid: ${JSON.stringify(model.id)};\n${indent}\tprovider: ${JSON.stringify(model.provider)};\n${indent}};\n`;
|
||||
}
|
||||
|
||||
for (const providerId of sortedProviderIds) {
|
||||
const models = providers[providerId];
|
||||
let output = generatedHeader;
|
||||
output += `import values from "./data/${providerId}.json" with { type: "json" };\n`;
|
||||
output += `import type { Model } from "../types.ts";\n\n`;
|
||||
output += `export const ${catalogConstName(providerId)} = values as {\n`;
|
||||
for (const modelId of Object.keys(models).sort()) {
|
||||
output += emitModelShape(models[modelId], "\t");
|
||||
}
|
||||
output += `};\n`;
|
||||
output += `import { flattenModelCatalog, type ModelCatalog } from "../model-catalog.ts";\n\n`;
|
||||
output += `export const ${catalogConstName(providerId)}: ModelCatalog<typeof values, ${JSON.stringify(providerId)}> =\n`;
|
||||
output += `\tflattenModelCatalog(${JSON.stringify(providerId)}, values);\n`;
|
||||
const filename = `${providerId}.models.ts`;
|
||||
generatedShardFiles.add(filename);
|
||||
writeFileSync(join(providersDir, filename), output);
|
||||
@@ -2590,19 +2569,22 @@ async function generateModels() {
|
||||
for (const entry of readdirSync(providersDir)) {
|
||||
if (entry.endsWith(".models.ts") && !generatedShardFiles.has(entry)) rmSync(join(providersDir, entry));
|
||||
}
|
||||
console.log(`Generated ${sortedProviderIds.length} catalog structures under src/providers/`);
|
||||
|
||||
let output = generatedHeader;
|
||||
for (const providerId of sortedProviderIds) {
|
||||
output += `import { ${catalogConstName(providerId)} } from "./providers/${providerId}.models.ts";\n`;
|
||||
}
|
||||
output += `\nexport const MODELS = {\n`;
|
||||
output += `\nexport const MODELS: {\n`;
|
||||
for (const providerId of sortedProviderIds) {
|
||||
output += `\treadonly ${JSON.stringify(providerId)}: typeof ${catalogConstName(providerId)};\n`;
|
||||
}
|
||||
output += `} = {\n`;
|
||||
for (const providerId of sortedProviderIds) {
|
||||
output += `\t${JSON.stringify(providerId)}: ${catalogConstName(providerId)},\n`;
|
||||
}
|
||||
output += `} as const;\n`;
|
||||
output += `};\n`;
|
||||
writeFileSync(aggregatorPath, output);
|
||||
console.log("Generated src/models.generated.ts");
|
||||
console.log("Generated provider catalogs and src/models.generated.ts");
|
||||
}
|
||||
|
||||
const hadPreviousData = existsSync(dataDir);
|
||||
@@ -2615,14 +2597,14 @@ async function generateModels() {
|
||||
if (hadPreviousData && existsSync(previousDataDir)) renameSync(previousDataDir, dataDir);
|
||||
throw error;
|
||||
}
|
||||
restoreStructuralCatalog = undefined;
|
||||
restoreGeneratedCatalog = undefined;
|
||||
console.log(
|
||||
generatorOptions.dataOnly
|
||||
? "Hydrated JSON model values under src/providers/data/"
|
||||
: "Generated JSON model values under src/providers/data/",
|
||||
);
|
||||
} catch (error) {
|
||||
restoreStructuralCatalog?.();
|
||||
restoreGeneratedCatalog?.();
|
||||
throw error;
|
||||
} finally {
|
||||
rmSync(stagingRoot, { recursive: true, force: true });
|
||||
|
||||
@@ -2,7 +2,7 @@ import { createHash } from "node:crypto";
|
||||
import { existsSync, readFileSync, readdirSync, statSync } from "node:fs";
|
||||
import { join } from "node:path";
|
||||
|
||||
export const MODEL_DATA_SCHEMA_VERSION = 1;
|
||||
export const MODEL_DATA_SCHEMA_VERSION = 2;
|
||||
export const MODEL_DATA_MANIFEST_FILE = ".manifest.json";
|
||||
|
||||
export type ModelDataStructure = Record<string, Record<string, string>>;
|
||||
@@ -13,21 +13,13 @@ export interface ModelDataManifest {
|
||||
files: Record<string, string>;
|
||||
}
|
||||
|
||||
const JSON_STRING_PATTERN = '"(?:\\\\.|[^"\\\\])*"';
|
||||
const MODEL_SHAPE_PATTERN = new RegExp(`^\\t(${JSON_STRING_PATTERN}): Model<(${JSON_STRING_PATTERN})> & \\{$`);
|
||||
const MODEL_ID_PATTERN = new RegExp(`^\\t\\tid: (${JSON_STRING_PATTERN});$`);
|
||||
const MODEL_PROVIDER_PATTERN = new RegExp(`^\\t\\tprovider: (${JSON_STRING_PATTERN});$`);
|
||||
const MODEL_DATA_IMPORT_PATTERN =
|
||||
/^import \{ [A-Z][A-Z0-9_]*_MODELS \} from "\.\/providers\/([^"/]+)\.models\.ts";$/gm;
|
||||
|
||||
function sha256(value: string): string {
|
||||
return createHash("sha256").update(value).digest("hex");
|
||||
}
|
||||
|
||||
function parseJsonString(value: string, description: string): string {
|
||||
const parsed: unknown = JSON.parse(value);
|
||||
if (typeof parsed !== "string") throw new Error(`${description} is not a string`);
|
||||
return parsed;
|
||||
}
|
||||
|
||||
function sortedRecord<T>(entries: Iterable<readonly [string, T]>): Record<string, T> {
|
||||
return Object.fromEntries(Array.from(entries).sort(([a], [b]) => (a < b ? -1 : a > b ? 1 : 0)));
|
||||
}
|
||||
@@ -46,84 +38,6 @@ function describeSetDifference(expected: readonly string[], actual: readonly str
|
||||
.join("; ");
|
||||
}
|
||||
|
||||
function parseProviderStructure(path: string, providerId: string): Record<string, string> {
|
||||
const source = readFileSync(path, "utf8");
|
||||
const expectedImport = `import values from "./data/${providerId}.json" with { type: "json" };`;
|
||||
if (!source.includes(expectedImport)) {
|
||||
throw new Error(`${path} does not import ${providerId}.json`);
|
||||
}
|
||||
|
||||
const models = new Map<string, string>();
|
||||
const lines = source.split("\n");
|
||||
for (let index = 0; index < lines.length; index++) {
|
||||
const shapeMatch = MODEL_SHAPE_PATTERN.exec(lines[index]);
|
||||
if (!shapeMatch) continue;
|
||||
|
||||
const idMatch = MODEL_ID_PATTERN.exec(lines[index + 1] ?? "");
|
||||
const providerMatch = MODEL_PROVIDER_PATTERN.exec(lines[index + 2] ?? "");
|
||||
if (!idMatch || !providerMatch || lines[index + 3] !== "\t};") {
|
||||
throw new Error(`${path}:${index + 1} has a malformed generated model declaration`);
|
||||
}
|
||||
|
||||
const key = parseJsonString(shapeMatch[1], `${path}:${index + 1} model key`);
|
||||
const api = parseJsonString(shapeMatch[2], `${path}:${index + 1} model API`);
|
||||
const id = parseJsonString(idMatch[1], `${path}:${index + 2} model ID`);
|
||||
const provider = parseJsonString(providerMatch[1], `${path}:${index + 3} provider ID`);
|
||||
if (id !== key) throw new Error(`${path}:${index + 1} declares key ${key} with ID ${id}`);
|
||||
if (provider !== providerId) {
|
||||
throw new Error(`${path}:${index + 1} declares provider ${provider} instead of ${providerId}`);
|
||||
}
|
||||
if (models.has(key)) throw new Error(`${path} declares model ${key} more than once`);
|
||||
models.set(key, api);
|
||||
index += 3;
|
||||
}
|
||||
|
||||
if (models.size === 0) throw new Error(`${path} contains no generated model declarations`);
|
||||
return sortedRecord(models);
|
||||
}
|
||||
|
||||
export function readModelDataStructure(packageRoot: string): ModelDataStructure {
|
||||
const providersDir = join(packageRoot, "src", "providers");
|
||||
const shardProviderIds = readdirSync(providersDir)
|
||||
.filter((entry) => entry.endsWith(".models.ts"))
|
||||
.map((entry) => entry.slice(0, -".models.ts".length))
|
||||
.sort();
|
||||
if (shardProviderIds.length === 0) throw new Error(`No generated provider shards found under ${providersDir}`);
|
||||
|
||||
const aggregator = readFileSync(join(packageRoot, "src", "models.generated.ts"), "utf8");
|
||||
const importedProviderIds = Array.from(
|
||||
aggregator.matchAll(/^import \{ [A-Z0-9_]+_MODELS \} from "\.\/providers\/([^"/]+)\.models\.ts";$/gm),
|
||||
(match) => match[1],
|
||||
).sort();
|
||||
if (!sameStrings(shardProviderIds, importedProviderIds)) {
|
||||
throw new Error(
|
||||
`Generated model aggregator and provider shards do not match (${describeSetDifference(shardProviderIds, importedProviderIds)})`,
|
||||
);
|
||||
}
|
||||
|
||||
return sortedRecord(
|
||||
shardProviderIds.map((providerId) => [
|
||||
providerId,
|
||||
parseProviderStructure(join(providersDir, `${providerId}.models.ts`), providerId),
|
||||
] as const),
|
||||
);
|
||||
}
|
||||
|
||||
export function modelDataStructureHash(structure: ModelDataStructure): string {
|
||||
return sha256(JSON.stringify(structure));
|
||||
}
|
||||
|
||||
export function createModelDataManifest(
|
||||
structure: ModelDataStructure,
|
||||
fileContents: Readonly<Record<string, string>>,
|
||||
): ModelDataManifest {
|
||||
return {
|
||||
schemaVersion: MODEL_DATA_SCHEMA_VERSION,
|
||||
structureHash: modelDataStructureHash(structure),
|
||||
files: sortedRecord(Object.entries(fileContents).map(([file, content]) => [file, sha256(content)] as const)),
|
||||
};
|
||||
}
|
||||
|
||||
function isRecord(value: unknown): value is Record<string, unknown> {
|
||||
return typeof value === "object" && value !== null && !Array.isArray(value);
|
||||
}
|
||||
@@ -143,6 +57,76 @@ function readJsonObject(path: string, description: string, errors: string[]): Re
|
||||
return parsed;
|
||||
}
|
||||
|
||||
function readProviderStructure(path: string, providerId: string): Record<string, string> {
|
||||
const errors: string[] = [];
|
||||
const groups = readJsonObject(path, `${providerId}.json`, errors);
|
||||
if (!groups) throw new Error(errors.join("\n"));
|
||||
|
||||
const models = new Map<string, string>();
|
||||
for (const [api, value] of Object.entries(groups)) {
|
||||
if (!isRecord(value)) throw new Error(`${path} API group ${JSON.stringify(api)} must be an object`);
|
||||
for (const modelId of Object.keys(value)) {
|
||||
if (models.has(modelId)) throw new Error(`${path} contains model ${modelId} in more than one API group`);
|
||||
models.set(modelId, api);
|
||||
}
|
||||
}
|
||||
if (models.size === 0) throw new Error(`${path} contains no generated model data`);
|
||||
return sortedRecord(models);
|
||||
}
|
||||
|
||||
export function readModelDataProviderIds(packageRoot: string): string[] {
|
||||
const aggregatorPath = join(packageRoot, "src", "models.generated.ts");
|
||||
const aggregator = readFileSync(aggregatorPath, "utf8");
|
||||
const providerIds = Array.from(aggregator.matchAll(MODEL_DATA_IMPORT_PATTERN), (match) => match[1]).sort();
|
||||
if (providerIds.length === 0) throw new Error(`No generated provider imports found in ${aggregatorPath}`);
|
||||
if (new Set(providerIds).size !== providerIds.length) {
|
||||
throw new Error(`Generated model aggregator contains duplicate provider imports: ${aggregatorPath}`);
|
||||
}
|
||||
return providerIds;
|
||||
}
|
||||
|
||||
export function readModelDataStructure(packageRoot: string): ModelDataStructure {
|
||||
const providersDir = join(packageRoot, "src", "providers");
|
||||
const dataDir = join(providersDir, "data");
|
||||
const providerIds = readModelDataProviderIds(packageRoot);
|
||||
const expectedShards = providerIds.map((providerId) => `${providerId}.models.ts`).sort();
|
||||
const actualShards = readdirSync(providersDir)
|
||||
.filter((entry) => entry.endsWith(".models.ts"))
|
||||
.sort();
|
||||
if (!sameStrings(expectedShards, actualShards)) {
|
||||
throw new Error(
|
||||
`Generated model aggregator and provider shards do not match (${describeSetDifference(expectedShards, actualShards)})`,
|
||||
);
|
||||
}
|
||||
|
||||
return sortedRecord(
|
||||
providerIds.map((providerId) => [
|
||||
providerId,
|
||||
readProviderStructure(join(dataDir, `${providerId}.json`), providerId),
|
||||
]),
|
||||
);
|
||||
}
|
||||
|
||||
export function modelDataStructureHash(structure: ModelDataStructure): string {
|
||||
const normalized = sortedRecord(
|
||||
Object.entries(structure).map(
|
||||
([providerId, models]) => [providerId, sortedRecord(Object.entries(models))] as const,
|
||||
),
|
||||
);
|
||||
return sha256(JSON.stringify(normalized));
|
||||
}
|
||||
|
||||
export function createModelDataManifest(
|
||||
structure: ModelDataStructure,
|
||||
fileContents: Readonly<Record<string, string>>,
|
||||
): ModelDataManifest {
|
||||
return {
|
||||
schemaVersion: MODEL_DATA_SCHEMA_VERSION,
|
||||
structureHash: modelDataStructureHash(structure),
|
||||
files: sortedRecord(Object.entries(fileContents).map(([file, content]) => [file, sha256(content)] as const)),
|
||||
};
|
||||
}
|
||||
|
||||
function validateModelValue(
|
||||
value: unknown,
|
||||
providerId: string,
|
||||
@@ -209,7 +193,7 @@ export function validateModelDataDirectory(structure: ModelDataStructure, dataDi
|
||||
.filter((entry) => entry.endsWith(".json") && entry !== MODEL_DATA_MANIFEST_FILE)
|
||||
.sort();
|
||||
if (!sameStrings(expectedFiles, actualFiles)) {
|
||||
errors.push(`provider data files do not match the structural catalog (${describeSetDifference(expectedFiles, actualFiles)})`);
|
||||
errors.push(`provider data files do not match the generated catalog (${describeSetDifference(expectedFiles, actualFiles)})`);
|
||||
}
|
||||
|
||||
const manifestPath = join(dataDir, MODEL_DATA_MANIFEST_FILE);
|
||||
@@ -221,7 +205,7 @@ export function validateModelDataDirectory(structure: ModelDataStructure, dataDi
|
||||
}
|
||||
const expectedStructureHash = modelDataStructureHash(structure);
|
||||
if (manifest?.structureHash !== expectedStructureHash) {
|
||||
errors.push("model data generation stamp does not match the structural catalog");
|
||||
errors.push("model data generation stamp does not match the generated catalog");
|
||||
}
|
||||
const manifestFiles = isRecord(manifest?.files) ? manifest.files : undefined;
|
||||
if (!manifestFiles) errors.push("model data manifest has no file hashes");
|
||||
@@ -240,15 +224,37 @@ export function validateModelDataDirectory(structure: ModelDataStructure, dataDi
|
||||
if (manifestFiles && manifestFiles[filename] !== sha256(content)) {
|
||||
errors.push(`${filename} does not match its manifest hash`);
|
||||
}
|
||||
const values = readJsonObject(path, filename, errors);
|
||||
if (!values) continue;
|
||||
const expectedModelIds = Object.keys(expectedModels).sort();
|
||||
const actualModelIds = Object.keys(values).sort();
|
||||
if (!sameStrings(expectedModelIds, actualModelIds)) {
|
||||
errors.push(`${filename} model IDs do not match the structural catalog (${describeSetDifference(expectedModelIds, actualModelIds)})`);
|
||||
const groups = readJsonObject(path, filename, errors);
|
||||
if (!groups) continue;
|
||||
|
||||
const actualModels = new Map<string, string>();
|
||||
for (const [api, value] of Object.entries(groups)) {
|
||||
if (!isRecord(value)) {
|
||||
errors.push(`${filename} API group ${JSON.stringify(api)} must be an object`);
|
||||
continue;
|
||||
}
|
||||
for (const [modelId, model] of Object.entries(value)) {
|
||||
if (actualModels.has(modelId)) {
|
||||
errors.push(`${providerId}/${modelId} appears in more than one API group`);
|
||||
continue;
|
||||
}
|
||||
actualModels.set(modelId, api);
|
||||
validateModelValue(model, providerId, modelId, api, errors);
|
||||
}
|
||||
}
|
||||
for (const [modelId, api] of Object.entries(expectedModels)) {
|
||||
if (modelId in values) validateModelValue(values[modelId], providerId, modelId, api, errors);
|
||||
|
||||
const expectedModelIds = Object.keys(expectedModels).sort();
|
||||
const actualModelIds = Array.from(actualModels.keys()).sort();
|
||||
if (!sameStrings(expectedModelIds, actualModelIds)) {
|
||||
errors.push(`${filename} model IDs do not match the generated catalog (${describeSetDifference(expectedModelIds, actualModelIds)})`);
|
||||
}
|
||||
for (const [modelId, expectedApi] of Object.entries(expectedModels)) {
|
||||
const actualApi = actualModels.get(modelId);
|
||||
if (actualApi !== undefined && actualApi !== expectedApi) {
|
||||
errors.push(
|
||||
`${providerId}/${modelId} is grouped under API ${JSON.stringify(actualApi)}, expected ${JSON.stringify(expectedApi)}`,
|
||||
);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
Reference in New Issue
Block a user