refactor(fase1-s3): split s3-controller into s3/ modules + guards

This commit is contained in:
asepharyana
2026-09-14 17:25:07 +07:00
parent 5d01a9405f
commit d323260ed5
8 changed files with 1875 additions and 1778 deletions
File diff suppressed because it is too large Load Diff
@@ -0,0 +1,111 @@
import { bucketRepository, fileRepository } from '../../../../infrastructure/di';
import { BucketNameSchema, parseOrNull } from '../../../../shared/validation/schemas';
import { bucketVersioningConfigurationXml, listBucketsXml, s3ErrorResponse } from '../../../s3/xml';
import { resolveBucketOr404, s3Response } from './s3-common';
// ─────── Bucket Operations ───────
/**
* Handles GET / — lists all buckets as an S3 ListAllMyBuckets XML response.
*
* @param reqId - The request identifier for S3 headers.
* @returns An S3 XML response with the bucket list.
*/
export const handleListBuckets = async (reqId: string): Promise<Response> => {
const buckets = await bucketRepository.list();
const xml = listBucketsXml(buckets, reqId);
return s3Response(xml, 200, reqId, { 'content-type': 'application/xml' });
};
/**
* Handles PUT /{bucket} — creates a new S3 bucket.
*
* Validates the bucket name format (BucketNameSchema, incl. the stricter M13
* rules: no consecutive dots, no IP format, no `xn--` prefix) and checks for
* duplicates.
*
* @param bucketName - The requested bucket name.
* @param reqId - The request identifier for S3 headers.
* @returns An S3 XML response indicating success or failure.
*/
export const handleCreateBucket = async (bucketName: string, reqId: string): Promise<Response> => {
if (parseOrNull(BucketNameSchema, bucketName) === null) {
return s3ErrorResponse(
'InvalidBucketName',
'The specified bucket is not valid.',
`/${bucketName}`,
400,
reqId,
);
}
const existing = await bucketRepository.findByName(bucketName);
if (existing) {
return s3ErrorResponse(
'BucketAlreadyExists',
'The requested bucket name is not available.',
`/${bucketName}`,
409,
reqId,
);
}
await bucketRepository.create(bucketName);
return s3Response(null, 200, reqId);
};
/**
* Handles HEAD /{bucket} — checks whether a bucket exists.
*
* @param bucketName - The bucket name to check.
* @param reqId - The request identifier for S3 headers.
* @returns A 200 response when the bucket exists, or an S3 XML error.
*/
export const handleHeadBucket = async (bucketName: string, reqId: string): Promise<Response> => {
const bucket = await resolveBucketOr404(bucketRepository, bucketName, `/${bucketName}`, reqId);
if (bucket instanceof Response) return bucket;
return s3Response(null, 200, reqId);
};
/**
* Handles DELETE /{bucket} — deletes a bucket.
*
* Fails with `BucketNotEmpty` if the bucket still contains objects.
*
* @param bucketName - The bucket name to delete.
* @param reqId - The request identifier for S3 headers.
* @returns A 204 response on success, or an S3 XML error.
*/
export const handleDeleteBucket = async (bucketName: string, reqId: string): Promise<Response> => {
const bucket = await resolveBucketOr404(bucketRepository, bucketName, `/${bucketName}`, reqId);
if (bucket instanceof Response) return bucket;
const objCount = await fileRepository.countByBucket(bucket.id);
if (objCount > 0) {
return s3ErrorResponse(
'BucketNotEmpty',
'The bucket you tried to delete is not empty.',
`/${bucketName}`,
409,
reqId,
);
}
await bucketRepository.delete(bucketName);
return s3Response(null, 204, reqId);
};
/**
* Handles GET /{bucket}?versioning — returns the bucket versioning
* configuration (always disabled in this implementation).
*
* @param bucketName - The bucket name.
* @param reqId - The request identifier for S3 headers.
* @returns An S3 XML response with the versioning configuration.
*/
export const handleGetBucketVersioning = async (
bucketName: string,
reqId: string,
): Promise<Response> => {
const bucket = await resolveBucketOr404(bucketRepository, bucketName, `/${bucketName}`, reqId);
if (bucket instanceof Response) return bucket;
return s3Response(bucketVersioningConfigurationXml(), 200, reqId, {
'content-type': 'application/xml',
});
};
@@ -0,0 +1,146 @@
import { nanoid } from 'nanoid';
import type { Bucket } from '../../../../domain/entities/bucket';
import type { MultipartUpload } from '../../../../domain/entities/multipart';
import type { IBucketRepository } from '../../../../domain/ports/bucket-repository';
import type { IMultipartRepository } from '../../../../domain/ports/multipart-repository';
import { config } from '../../../../env';
import { s3Headers } from '../../../s3/headers';
import { unsatisfiedContentRange } from '../../../s3/range';
import { s3ErrorResponse } from '../../../s3/xml';
/**
* The default S3 region returned when no region is explicitly configured.
*/
export const REGION = config.s3DefaultRegion || 'us-east-1';
/**
* Generates a unique request identifier for S3 responses.
*
* @returns A hex string suitable for x-amz-request-id and x-amz-id-2.
*/
export const REQUEST_ID = (): string => nanoid(16);
/**
* Builds a standard S3 response with the appropriate headers.
*
* @param body - The XML or empty response body.
* @param status - HTTP status code.
* @param reqId - The request identifier for S3 headers.
* @param extraHeaders - Optional extra response headers.
* @returns An S3-formatted Response.
*/
export const s3Response = (
body: string | null,
status: number,
reqId: string,
extraHeaders: Record<string, string> = {},
): Response => {
// Add content-type for empty 200-series responses (not 204 which has no body)
if (
body === null &&
status >= 200 &&
status < 300 &&
status !== 204 &&
!extraHeaders['content-type']
) {
extraHeaders['content-type'] = 'application/xml';
}
return new Response(body, { status, headers: s3Headers(reqId, extraHeaders) });
};
/**
* Resolves a bucket by name, returning a `NoSuchBucket` S3 error when missing.
*
* Replaces the ~10 identical `findByName` + `NoSuchBucket` blocks previously
* inlined in every bucket/object/listing/multipart handler.
*
* @param bucketRepo - The bucket repository to look up.
* @param bucket - The bucket name from the request path.
* @param path - The request path for the S3 error resource.
* @param reqId - The request identifier for S3 headers.
* @returns The bucket record, or an S3 error Response when not found.
*/
export const resolveBucketOr404 = async (
bucketRepo: IBucketRepository,
bucket: string,
path: string,
reqId: string,
): Promise<Bucket | Response> => {
const bucketRecord = await bucketRepo.findByName(bucket);
if (!bucketRecord) {
return s3ErrorResponse(
'NoSuchBucket',
'The specified bucket does not exist.',
path,
404,
reqId,
);
}
return bucketRecord;
};
/**
* Resolves an in-progress multipart upload, returning a `NoSuchUpload` S3
* error when missing.
*
* Replaces the 5 identical `findById` + `NoSuchUpload` blocks previously
* inlined in the multipart handlers. Pass `key` for the H5 key-match check
* (UploadPart / CompleteMultipartUpload); omit it for Abort / ListParts,
* which historically only checked existence.
*
* @param multipartRepo - The multipart repository to look up.
* @param uploadId - The upload identifier from `?uploadId=`.
* @param path - The request path for the S3 error resource.
* @param reqId - The request identifier for S3 headers.
* @param key - Optional object key the upload must belong to.
* @returns The upload record, or an S3 error Response when not found.
*/
export const requireUploadOr404 = async (
multipartRepo: IMultipartRepository,
uploadId: string,
path: string,
reqId: string,
key?: string,
): Promise<MultipartUpload | Response> => {
const multipart = await multipartRepo.findById(uploadId);
if (!multipart || (key !== undefined && multipart.s3Key !== key)) {
return s3ErrorResponse(
'NoSuchUpload',
'The specified upload does not exist.',
path,
404,
reqId,
);
}
return multipart;
};
/**
* Returns the stable S3 etag for a file, falling back to a random ID when
* the record has no content hash yet.
*
* Replaces the 4 inline `file.fileHash || nanoid(16)` fallbacks (conditional
* headers, HeadObject, copy result, list entries). Callers add quotes where
* the transport needs them. The copy-source check keeps its own inline
* `telegramFileId` variant (M9) with a comment at the call site.
*
* @param fileHash - The stored SHA-256 content hash (may be null).
* @returns The hash, or a random 16-char fallback.
*/
export const etagOrFallback = (fileHash: string | null): string => fileHash || nanoid(16);
/**
* Builds the shared 416 response for unsatisfiable Range requests.
*
* Replaces the 3 identical `InvalidRange` blocks (chunked GET, regular GET,
* multipart GET).
*
* @param path - The request path for the S3 error resource.
* @param totalSize - The total object size for the Content-Range header.
* @param reqId - The request identifier for S3 headers.
* @returns A 416 S3 error Response.
*/
export const invalidRangeResponse = (path: string, totalSize: number, reqId: string): Response =>
s3ErrorResponse('InvalidRange', 'The requested range is not satisfiable.', path, 416, reqId, {
'content-range': unsatisfiedContentRange(totalSize),
});
@@ -0,0 +1,142 @@
import type { File as FileEntity } from '../../../../domain/entities/file';
import { bucketRepository, fileRepository } from '../../../../infrastructure/di';
import { clampMaxKeys } from '../../../../shared/validation/schemas';
import { listBucketResultXml, listBucketV2ResultXml } from '../../../s3/xml';
import { etagOrFallback, resolveBucketOr404, s3Response } from './s3-common';
/** Shape of an S3 list entry object. */
export type S3ListEntry = {
key: string;
sizeBytes: number;
etag: string;
lastModified: Date;
mimeType: string;
};
/**
* Maps a File entity to an S3 list entry object.
*
* @param file - The file entity from the repository.
* @returns An S3 list entry with key, size, etag, last modified, and MIME type.
*/
export const mapFileToListEntry = (file: FileEntity): S3ListEntry => ({
key: file.s3Key ?? '',
sizeBytes: file.sizeBytes,
etag: etagOrFallback(file.fileHash),
lastModified: file.createdAt instanceof Date ? file.createdAt : new Date(),
mimeType: file.mimeType,
});
/**
* Handles GET /{bucket} (ListObjectsV1 with query parameters).
*
* @param bucket - The bucket name.
* @param searchParams - URL query parameters (prefix, delimiter, max-keys,
* marker, encoding-type).
* @param reqId - The request identifier for S3 headers.
* @returns An S3 XML ListBucketResult response.
*/
export const handleListObjectsV1 = async (
bucket: string,
searchParams: URLSearchParams,
reqId: string,
): Promise<Response> => {
const bucketRecord = await resolveBucketOr404(bucketRepository, bucket, `/${bucket}`, reqId);
if (bucketRecord instanceof Response) return bucketRecord;
const prefix = searchParams.get('prefix') || '';
const delimiter = searchParams.get('delimiter') || null;
const maxKeys = clampMaxKeys(searchParams.get('max-keys'));
const marker = searchParams.get('marker') || null;
const encodingType = searchParams.get('encoding-type') || null;
const { objects, prefixes: commonPrefixes } = await fileRepository.listByPrefix(
bucketRecord.id,
prefix,
delimiter,
maxKeys,
marker,
);
const isTruncated = objects.length > maxKeys;
const displayObjects = objects.slice(0, maxKeys);
const nextMarker = isTruncated
? (displayObjects[displayObjects.length - 1]?.s3Key ?? null)
: null;
const xml = listBucketResultXml(
bucket,
displayObjects.map(mapFileToListEntry),
commonPrefixes,
isTruncated,
marker,
maxKeys,
prefix,
delimiter,
nextMarker,
reqId,
encodingType,
);
return s3Response(xml, 200, reqId, { 'content-type': 'application/xml' });
};
/**
* Handles GET /{bucket}?list-type=2 (ListObjectsV2).
*
* Small approved behavior fix: V2 previously used a bare
* `Math.min(parse, 1000)` which admitted 0, negatives, and NaN. It now uses
* the same `clampMaxKeys` ([1, 1000]) as V1.
*
* @param bucket - The bucket name.
* @param searchParams - URL query parameters (prefix, delimiter, max-keys,
* continuation-token, start-after, encoding-type).
* @param reqId - The request identifier for S3 headers.
* @returns An S3 XML ListBucketV2Result response.
*/
export const handleListObjectsV2 = async (
bucket: string,
searchParams: URLSearchParams,
reqId: string,
): Promise<Response> => {
const bucketRecord = await resolveBucketOr404(bucketRepository, bucket, `/${bucket}`, reqId);
if (bucketRecord instanceof Response) return bucketRecord;
const prefix = searchParams.get('prefix') || '';
const delimiter = searchParams.get('delimiter') || null;
const maxKeys = clampMaxKeys(searchParams.get('max-keys'));
const continuationToken = searchParams.get('continuation-token') || null;
const startAfter = searchParams.get('start-after') || null;
const encodingType = searchParams.get('encoding-type') || null;
const { objects, prefixes: commonPrefixes } = await fileRepository.listByPrefix(
bucketRecord.id,
prefix,
delimiter,
maxKeys,
continuationToken || startAfter,
);
const isTruncated = objects.length > maxKeys;
const displayObjects = objects.slice(0, maxKeys);
const nextContinuationToken = isTruncated
? (displayObjects[displayObjects.length - 1]?.s3Key ?? null)
: null;
const xml = listBucketV2ResultXml(
bucket,
displayObjects.map(mapFileToListEntry),
commonPrefixes,
isTruncated,
maxKeys,
prefix,
delimiter,
continuationToken,
nextContinuationToken,
displayObjects.length,
reqId,
encodingType,
);
return s3Response(xml, 200, reqId, { 'content-type': 'application/xml' });
};
@@ -0,0 +1,431 @@
import { createReadStream } from 'node:fs';
import { nanoid } from 'nanoid';
import { buildNewFile } from '../../../../domain/entities/file-factory';
import type { IMultipartRepository } from '../../../../domain/ports/multipart-repository';
import type { ForwardResult } from '../../../../domain/ports/telegram-service';
import { config } from '../../../../env';
import {
bucketRepository,
fileRepository,
multipartRepository,
} from '../../../../infrastructure/di';
import { botPool } from '../../../../infrastructure/telegram/bot-pool';
import { cleanupTempFile, DEFAULT_FILE_TYPE } from '../../../../shared/utils/file';
import {
CompletePartSchema,
clampMaxKeys,
PartNumberSchema,
parseOrNull,
} from '../../../../shared/validation/schemas';
import {
completeMultipartUploadXml,
initiateMultipartUploadXml,
listMultipartUploadsXml,
listPartsXml,
parseCompleteMultipartBody,
s3ErrorResponse,
} from '../../../s3/xml';
import { requireUploadOr404, resolveBucketOr404, s3Response } from './s3-common';
type MultipartRepo = IMultipartRepository;
// ─────── Multipart Upload ───────
/**
* Handles POST /{bucket}/{key}?uploads — initiates a multipart upload.
*
* @param bucket - The bucket name.
* @param key - The object key being uploaded.
* @param _searchParams - URL query parameters (unused).
* @param reqId - The request identifier for S3 headers.
* @returns An S3 XML InitiateMultipartUpload response.
*/
export const handleCreateMultipartUpload = async (
bucket: string,
key: string,
_searchParams: URLSearchParams,
headers: Record<string, string>,
reqId: string,
): Promise<Response> => {
const bucketRecord = await resolveBucketOr404(
bucketRepository,
bucket,
`/${bucket}/${key}`,
reqId,
);
if (bucketRecord instanceof Response) return bucketRecord;
// NOTE: IMultipartRepository.create takes 3 args (bucketId, s3Key,
// initiatedBy). The content-type is intentionally dropped here — the
// complete step also falls back to 'application/octet-stream' since the
// in-progress upload record carries no contentType field.
void headers;
const uploadId = await (multipartRepository as MultipartRepo).create(bucketRecord.id, key, 's3');
const xml = initiateMultipartUploadXml(bucket, key, uploadId);
return s3Response(xml, 200, reqId, { 'content-type': 'application/xml' });
};
/**
* Handles PUT /{bucket}/{key}?uploadId=&partNumber= — uploads a single
* part of a multipart upload.
*
* @param bucket - The bucket name.
* @param key - The object key.
* @param searchParams - URL query parameters containing uploadId and
* partNumber.
* @param req - The incoming HTTP request with the part body.
* @param reqId - The request identifier for S3 headers.
* @returns An S3 response with the part etag, or an error.
*/
export const handleUploadPart = async (
bucket: string,
key: string,
searchParams: URLSearchParams,
req: Request,
reqId: string,
): Promise<Response> => {
const uploadId = searchParams.get('uploadId')!;
// M14: partNumber must be an integer 1–10000 (PartNumberSchema mirrors the
// old manual check; same InvalidArgument message preserved).
const partNumber = parseOrNull(PartNumberSchema, searchParams.get('partNumber'));
if (partNumber === null) {
return s3ErrorResponse(
'InvalidArgument',
'Part number must be an integer between 1 and 10000',
`/${bucket}/${key}`,
400,
reqId,
);
}
// H5: Verify both upload exists AND key matches.
const multipart = await requireUploadOr404(
multipartRepository,
uploadId,
`/${bucket}/${key}`,
reqId,
key,
);
if (multipart instanceof Response) return multipart;
// Stream the part body to temp — O(1) memory, safe for large parts
const tempPath = `/tmp/filedrop-mp-${nanoid()}`;
const writer = Bun.file(tempPath).writer();
const reader = (
req.body ??
new ReadableStream({
start(c) {
c.close();
},
})
).getReader();
const hasher = new Bun.CryptoHasher('sha256');
let sizeBytes = 0;
try {
while (true) {
const { done, value } = await reader.read();
if (done) break;
const chunk = Buffer.from(value);
sizeBytes += chunk.byteLength;
hasher.update(chunk);
writer.write(chunk);
}
await writer.end();
} catch (error) {
try {
writer.end();
} catch {
// ignore during error path
}
await cleanupTempFile(tempPath);
throw error;
} finally {
reader.releaseLock();
}
if (sizeBytes > config.telegramChunkSizeBytes) {
await cleanupTempFile(tempPath);
return s3ErrorResponse(
'EntityTooLarge',
`Your proposed upload part size (${sizeBytes} bytes) exceeds the maximum allowed part size (${config.telegramChunkSizeBytes} bytes) for this storage backend. Use smaller part sizes.`,
`/${bucket}/${key}`,
400,
reqId,
);
}
// M4: Ensure temp file cleanup even if forwardToStorage fails
let forwardResult: ForwardResult;
try {
forwardResult = await botPool.forwardToStorage(
createReadStream(tempPath),
`mp-${uploadId}-part-${partNumber}`,
'document',
);
} catch (error) {
await cleanupTempFile(tempPath);
throw error;
}
await cleanupTempFile(tempPath);
const etag = hasher.digest('hex');
await multipartRepository.insertPart({
uploadId,
partNumber,
telegramFileId: forwardResult.telegramFileId,
telegramFileUniqueId: forwardResult.telegramFileUniqueId,
storageMessageId: forwardResult.storageMessageId,
sizeBytes,
etag,
});
return s3Response(null, 200, reqId, { etag: `"${etag}"` });
};
/**
* Handles POST /{bucket}/{key}?uploadId= — completes a multipart upload.
*
* Validates the submitted part list (each part against CompletePartSchema —
* invalid → InvalidPart 400; all parts present, ascending order), creates
* the final file record, and marks the upload as completed.
*
* @param bucket - The bucket name.
* @param key - The object key.
* @param searchParams - URL query parameters containing uploadId.
* @param body - The raw XML request body containing the complete part list.
* @param reqId - The request identifier for S3 headers.
* @returns An S3 XML CompleteMultipartUpload response.
*/
export const handleCompleteMultipartUpload = async (
bucket: string,
key: string,
searchParams: URLSearchParams,
body: string,
reqId: string,
): Promise<Response> => {
const uploadId = searchParams.get('uploadId')!;
// H5: Verify both upload exists AND key matches (consistent with handleUploadPart)
const multipart = await requireUploadOr404(
multipartRepository,
uploadId,
`/${bucket}/${key}`,
reqId,
key,
);
if (multipart instanceof Response) return multipart;
const rawParts = parseCompleteMultipartBody(body);
// Validate each submitted part against CompletePartSchema; any invalid
// part → InvalidPart 400 (approved small fix; previously malformed parts
// were silently dropped by the regex parser and surfaced as count mismatch).
const parts: Array<{ partNumber: number; etag: string }> = [];
for (const raw of rawParts) {
const valid = parseOrNull(CompletePartSchema, raw);
if (valid === null) {
return s3ErrorResponse(
'InvalidPart',
'One or more specified parts could not be found.',
`/${bucket}/${key}`,
400,
reqId,
);
}
parts.push(valid);
}
const storedParts = await multipartRepository.listParts(uploadId);
// Validate ascending part order
const partNumbers = parts.map((p) => p.partNumber);
if (partNumbers.length > 1 && partNumbers.some((n, i) => i > 0 && n <= partNumbers[i - 1])) {
return s3ErrorResponse(
'InvalidPartOrder',
'The list of parts was not in ascending order.',
`/${bucket}/${key}`,
400,
reqId,
);
}
// H8: Verify count AND part numbers AND etags match stored parts
if (parts.length !== storedParts.length) {
return s3ErrorResponse(
'InvalidPart',
'One or more specified parts could not be found.',
`/${bucket}/${key}`,
400,
reqId,
);
}
// Build a map for O(1) part number lookup
const storedByNumber = new Map<number, (typeof storedParts)[0]>();
for (const sp of storedParts) {
storedByNumber.set(sp.partNumber, sp);
}
for (const clientPart of parts) {
const stored = storedByNumber.get(clientPart.partNumber);
if (!stored || stored.etag !== clientPart.etag) {
return s3ErrorResponse(
'InvalidPart',
'One or more specified parts could not be found. The etag or part number does not match.',
`/${bucket}/${key}`,
400,
reqId,
);
}
}
const totalSize = storedParts.reduce((sum, p) => sum + Number(p.sizeBytes), 0);
const combinedEtag = storedParts.map((p) => p.etag).join('-');
const publicId = nanoid();
// The upload record carries no content-type, so the assembled object
// defaults to application/octet-stream.
const mimeType = 'application/octet-stream';
await fileRepository.create(
buildNewFile({
publicId,
telegramFileId: storedParts[0]!.telegramFileId,
telegramFileUniqueId: storedParts[0]!.telegramFileUniqueId,
storageChatId: config.storageChatId,
storageMessageId: storedParts[0]!.storageMessageId,
fileName: key.split('/').pop() || 'file',
mimeType,
sizeBytes: totalSize,
fileType: DEFAULT_FILE_TYPE,
uploaderId: 0,
bucketId: multipart.bucketId,
s3Key: key,
storageBackend: 'telegram',
multipartUploadId: uploadId,
}),
);
await multipartRepository.complete(uploadId);
const location = `${config.baseUrl}/${bucket}/${key}`;
const xml = completeMultipartUploadXml(bucket, key, combinedEtag, location);
return s3Response(xml, 200, reqId, { 'content-type': 'application/xml' });
};
/**
* Handles GET /{bucket}?uploads — lists in-progress multipart uploads.
*
* @param bucket - The bucket name.
* @param searchParams - URL query parameters (max-uploads, key-marker).
* @param reqId - The request identifier for S3 headers.
* @returns An S3 XML ListMultipartUploadsResult response.
*/
export const handleListMultipartUploads = async (
bucket: string,
searchParams: URLSearchParams,
reqId: string,
): Promise<Response> => {
const bucketRecord = await resolveBucketOr404(bucketRepository, bucket, `/${bucket}`, reqId);
if (bucketRecord instanceof Response) return bucketRecord;
const maxUploads = clampMaxKeys(searchParams.get('max-uploads'));
const keyMarker = searchParams.get('key-marker') || null;
const { uploads, isTruncated, nextKeyMarker } = await multipartRepository.listByBucket(
bucketRecord.id,
maxUploads,
keyMarker,
);
const xml = listMultipartUploadsXml(
bucket,
uploads.map((u) => ({
key: u.s3Key,
uploadId: u.uploadId,
initiatedAt: u.initiatedAt,
initiatedBy: u.initiatedBy,
})),
maxUploads,
isTruncated,
nextKeyMarker,
reqId,
);
return s3Response(xml, 200, reqId, { 'content-type': 'application/xml' });
};
/**
* Handles DELETE /{bucket}/{key}?uploadId= — aborts a multipart upload.
*
* @param bucket - The bucket name.
* @param key - The object key.
* @param searchParams - URL query parameters containing uploadId.
* @param reqId - The request identifier for S3 headers.
* @returns A 204 response on success, or an S3 XML error.
*/
export const handleAbortMultipartUpload = async (
bucket: string,
key: string,
searchParams: URLSearchParams,
reqId: string,
): Promise<Response> => {
const uploadId = searchParams.get('uploadId')!;
const multipart = await requireUploadOr404(
multipartRepository,
uploadId,
`/${bucket}/${key}`,
reqId,
);
if (multipart instanceof Response) return multipart;
await multipartRepository.abort(uploadId);
return s3Response(null, 204, reqId);
};
/**
* Handles GET /{bucket}/{key}?uploadId= — lists uploaded parts of a
* multipart upload.
*
* @param bucket - The bucket name.
* @param key - The object key.
* @param searchParams - URL query parameters containing uploadId and
* optional max-parts.
* @param reqId - The request identifier for S3 headers.
* @returns An S3 XML ListPartsResult response.
*/
export const handleListParts = async (
bucket: string,
key: string,
searchParams: URLSearchParams,
reqId: string,
): Promise<Response> => {
const uploadId = searchParams.get('uploadId')!;
const multipart = await requireUploadOr404(
multipartRepository,
uploadId,
`/${bucket}/${key}`,
reqId,
);
if (multipart instanceof Response) return multipart;
const parts = await multipartRepository.listParts(uploadId);
const maxParts = clampMaxKeys(searchParams.get('max-parts'));
const xml = listPartsXml(
bucket,
key,
uploadId,
parts.map((p) => ({
partNumber: p.partNumber,
etag: p.etag,
sizeBytes: p.sizeBytes,
createdAt: p.createdAt,
})),
maxParts,
false,
reqId,
);
return s3Response(xml, 200, reqId, { 'content-type': 'application/xml' });
};
@@ -0,0 +1,358 @@
import type { File as FileEntity } from '../../../../domain/entities/file';
import {
bucketRepository,
chunkedStorage,
fileRepository,
multipartRepository,
} from '../../../../infrastructure/di';
import { botPool } from '../../../../infrastructure/telegram/bot-pool';
import { buildTelegramFileUrl } from '../../../../infrastructure/telegram/file-url';
import logger from '../../../../shared/logger/index';
import { getErrorMessage } from '../../../../shared/utils/file';
import { s3Headers } from '../../../s3/headers';
import { createGetObjectResponse, type ObjectPartSource } from '../../../s3/object-stream';
import { parseRangeHeader } from '../../../s3/range';
import { s3ErrorResponse } from '../../../s3/xml';
import { etagOrFallback, invalidRangeResponse, resolveBucketOr404, s3Response } from './s3-common';
// ─────── Conditional Headers Helper ──────
/**
* S3-compatible response for 304 Not Modified.
*/
export const notModifiedResponse = (
reqId: string,
etag: string,
mimeType: string,
sizeBytes: number,
lastModified: Date,
): Response =>
new Response(null, {
status: 304,
headers: s3Headers(reqId, {
etag,
'content-type': mimeType,
'content-length': String(sizeBytes),
'last-modified': lastModified.toUTCString(),
'x-amz-version-id': 'null',
}),
});
/**
* S3-compatible response for 412 Precondition Failed.
*/
export const preconditionFailedResponse = (path: string, reqId: string): Response =>
s3ErrorResponse(
'PreconditionFailed',
'At least one of the pre-conditions you specified did not hold.',
path,
412,
reqId,
);
/**
* Checks conditional headers (If-Match, If-None-Match, If-Modified-Since,
* If-Unmodified-Since) and returns a prepared Response if the condition
* is not satisfied, or `null` to let the request proceed.
*
* @returns A 304 / 412 Response when a condition fails, or `null` to continue.
*/
export const checkConditionalHeaders = (
headers: Record<string, string>,
file: {
mimeType: string;
sizeBytes: number;
fileHash: string | null;
createdAt: Date | string | number;
},
path: string,
reqId: string,
): Response | null => {
const etag = `"${etagOrFallback(file.fileHash)}"`;
const lastModified = file.createdAt instanceof Date ? file.createdAt : new Date(file.createdAt);
// If-Match
const ifMatch = headers['if-match'];
if (ifMatch && ifMatch !== '*' && ifMatch !== etag) {
return preconditionFailedResponse(path, reqId);
}
// If-None-Match
const ifNoneMatch = headers['if-none-match'];
if (ifNoneMatch && ifNoneMatch === etag) {
return notModifiedResponse(reqId, etag, file.mimeType, file.sizeBytes, lastModified);
}
// If-Modified-Since
const ifModifiedSince = headers['if-modified-since'];
if (ifModifiedSince) {
const since = new Date(ifModifiedSince);
if (!Number.isNaN(since.getTime()) && lastModified.getTime() <= since.getTime()) {
return notModifiedResponse(reqId, etag, file.mimeType, file.sizeBytes, lastModified);
}
}
// If-Unmodified-Since
const ifUnmodifiedSince = headers['if-unmodified-since'];
if (ifUnmodifiedSince) {
const since = new Date(ifUnmodifiedSince);
if (!Number.isNaN(since.getTime()) && lastModified.getTime() > since.getTime()) {
return preconditionFailedResponse(path, reqId);
}
}
return null;
};
// ─────── Object Operations ───────
/**
* Handles GET /{bucket}/{key} — retrieves an S3 object.
*
* Supports chunked objects (streaming multi-part response), multipart
* objects (assembled from a completed multipart upload), and regular
* Telegram-stored objects (proxy streaming or 302 redirect depending
* on configuration). HTTP Range headers are respected when present.
*
* @param bucket - The bucket name.
* @param key - The object key.
* @param _searchParams - URL query parameters (unused for GET).
* @param headers - The request headers (used for Range and etag checks).
* @param reqId - The request identifier for S3 headers.
* @returns An S3 response with the object content or an error.
*/
export const handleGetObject = async (
bucket: string,
key: string,
_searchParams: URLSearchParams,
headers: Record<string, string>,
reqId: string,
): Promise<Response> => {
const bucketRecord = await resolveBucketOr404(
bucketRepository,
bucket,
`/${bucket}/${key}`,
reqId,
);
if (bucketRecord instanceof Response) return bucketRecord;
const file = await fileRepository.findByBucketAndKey(bucketRecord.id, key);
if (!file)
return s3ErrorResponse(
'NoSuchKey',
'The specified key does not exist.',
`/${bucket}/${key}`,
404,
reqId,
);
// H3: Conditional headers — If-Match / If-None-Match / If-Modified-Since / If-Unmodified-Since
const conditionResult = checkConditionalHeaders(headers, file, `/${bucket}/${key}`, reqId);
if (conditionResult) {
return conditionResult;
}
// Chunked storage object
if (file.storageBackend === 'chunked') {
const totalSize = Number(file.sizeBytes);
const range = parseRangeHeader(headers.range || null, totalSize);
if (range.type === 'invalid') {
return invalidRangeResponse(`/${bucket}/${key}`, totalSize, reqId);
}
try {
return await chunkedStorage.createChunkedObjectResponse({ file, range, reqId });
} catch (error) {
logger.warn('Chunked object content fetch failed', { key, error: getErrorMessage(error) });
return s3ErrorResponse(
'InternalError',
'Failed to fetch object content from storage',
`/${bucket}/${key}`,
502,
reqId,
);
}
}
// Multipart upload assembled object
if (file.multipartUploadId) {
return handleGetMultipartObject(file, bucket, key, headers, reqId);
}
// Regular Telegram object
const fileInfo = await botPool.getFileInfo(file.telegramFileId);
const telegramUrl = buildTelegramFileUrl(fileInfo.file_path, fileInfo.bot_token);
const totalSize = file.sizeBytes;
const range = parseRangeHeader(headers.range || null, totalSize);
if (range.type === 'invalid') {
return invalidRangeResponse(`/${bucket}/${key}`, totalSize, reqId);
}
// H1: Always proxy S3 GETs to avoid leaking the Telegram bot token
// in redirect URLs. The 302 redirect path is removed because the
// URL contains the bot_token — exposing it to clients is a security risk.
const part: ObjectPartSource = {
telegramFileId: file.telegramFileId,
telegramUrl,
sizeBytes: file.sizeBytes,
partNumber: 1,
};
try {
return await createGetObjectResponse({
reqId,
contentType: file.mimeType,
etag: file.fileHash || '',
lastModified: file.createdAt instanceof Date ? file.createdAt : new Date(file.createdAt),
totalSize: file.sizeBytes,
parts: [part],
range,
});
} catch (error) {
logger.warn('Telegram content fetch failed', {
fileId: file.telegramFileId,
error: getErrorMessage(error),
});
return s3ErrorResponse(
'InternalError',
'Failed to fetch object content from storage',
`/${bucket}/${key}`,
502,
reqId,
);
}
};
/**
* Handles GET for objects assembled from a completed multipart upload.
*
* Resolves the Telegram CDN URLs for each part and builds a multi-part
* streaming response, respecting HTTP Range headers.
*
* @param file - The file entity with a `multipartUploadId` reference.
* @param bucket - The bucket name.
* @param key - The object key.
* @param headers - The request headers (for Range parsing).
* @param reqId - The request identifier for S3 headers.
* @returns An S3 response streaming the assembled object content.
*/
export const handleGetMultipartObject = async (
file: FileEntity,
bucket: string,
key: string,
headers: Record<string, string>,
reqId: string,
): Promise<Response> => {
const uploadId = file.multipartUploadId!;
const parts = await multipartRepository.listParts(uploadId);
if (parts.length === 0) {
return s3ErrorResponse(
'InternalError',
'Multipart object has no parts.',
`/${bucket}/${key}`,
500,
reqId,
);
}
const totalSize = parts.reduce((sum, p) => sum + Number(p.sizeBytes), 0);
const range = parseRangeHeader(headers.range || null, totalSize);
if (range.type === 'invalid') {
return invalidRangeResponse(`/${bucket}/${key}`, totalSize, reqId);
}
const sources: ObjectPartSource[] = [];
// Resolve all part CDN URLs concurrently (independent getFile calls) so
// assembly latency is ~1 round-trip instead of N.
sources.push(
...(await Promise.all(
parts.map(async (part) => {
const fileInfo = await botPool.getFileInfo(part.telegramFileId);
return {
telegramFileId: part.telegramFileId,
telegramUrl: buildTelegramFileUrl(fileInfo.file_path, fileInfo.bot_token),
sizeBytes: part.sizeBytes,
partNumber: part.partNumber,
};
}),
)),
);
// H1: Always proxy — never expose bot token in redirect URL
try {
return await createGetObjectResponse({
reqId,
contentType: file.mimeType,
etag: file.fileHash || parts.map((p) => p.etag).join('-'),
lastModified: file.createdAt instanceof Date ? file.createdAt : new Date(file.createdAt),
totalSize,
parts: sources,
range,
});
} catch (error) {
logger.warn('Telegram multipart content fetch failed', {
uploadId: file.multipartUploadId,
error: getErrorMessage(error),
});
return s3ErrorResponse(
'InternalError',
'Failed to fetch object content from storage',
`/${bucket}/${key}`,
502,
reqId,
);
}
};
/**
* Handles HEAD /{bucket}/{key} — returns object metadata without the body.
*
* @param bucket - The bucket name.
* @param key - The object key.
* @param reqId - The request identifier for S3 headers.
* @returns An S3 response with object metadata headers.
*/
export const handleHeadObject = async (
bucket: string,
key: string,
headers: Record<string, string>,
reqId: string,
): Promise<Response> => {
const bucketRecord = await resolveBucketOr404(
bucketRepository,
bucket,
`/${bucket}/${key}`,
reqId,
);
if (bucketRecord instanceof Response) return bucketRecord;
const file = await fileRepository.findByBucketAndKey(bucketRecord.id, key);
if (!file)
return s3ErrorResponse(
'NoSuchKey',
'The specified key does not exist.',
`/${bucket}/${key}`,
404,
reqId,
);
// H3: Conditional headers for HEAD — If-Match / If-None-Match / If-Modified-Since / If-Unmodified-Since
const headConditionResult = checkConditionalHeaders(headers, file, `/${bucket}/${key}`, reqId);
if (headConditionResult) {
return headConditionResult;
}
return s3Response(null, 200, reqId, {
'content-type': file.mimeType,
'content-length': String(file.sizeBytes),
etag: `"${etagOrFallback(file.fileHash)}"`,
'last-modified':
file.createdAt instanceof Date ? file.createdAt.toUTCString() : new Date().toUTCString(),
'accept-ranges': 'bytes',
'cache-control': 'public, max-age=31536000',
'x-amz-version-id': 'null',
});
};
@@ -0,0 +1,442 @@
import { createReadStream } from 'node:fs';
import { nanoid } from 'nanoid';
import { buildNewFile } from '../../../../domain/entities/file-factory';
import type { ForwardResult } from '../../../../domain/ports/telegram-service';
import { config } from '../../../../env';
import { bucketRepository, chunkedStorage, fileRepository } from '../../../../infrastructure/di';
import { botPool } from '../../../../infrastructure/telegram/bot-pool';
import { cleanupTempFile, DEFAULT_FILE_TYPE, ensureExtension } from '../../../../shared/utils/file';
import { streamToTemp } from '../../../../shared/utils/temp-stream';
import { DeleteObjectsBodySchema, parseOrNull } from '../../../../shared/validation/schemas';
import { verifyBodyHash } from '../../../s3/auth';
import {
copyObjectResultXml,
deleteResultXml,
parseDeleteObjectsBody,
s3ErrorResponse,
} from '../../../s3/xml';
import { etagOrFallback, resolveBucketOr404, s3Response } from './s3-common';
/**
* Streams the request body to a temporary file while computing its SHA-256
* and MD5 hashes.
*
* Unlike `req.arrayBuffer()`, this approach uses O(1) memory regardless of
* file size, making it safe for multi-GB Docker registry layer blobs.
*
* MD5 is computed alongside SHA-256 so that Content-MD5 verification (when
* the header is present) does not need to re-read the entire file.
*
* @param body - The ReadableStream from the HTTP request body.
* @returns The temp file path, SHA-256 hash, MD5 hash (base64), total size, and signature bytes.
*/
export const streamBodyToTemp = async (
body: ReadableStream<Uint8Array> | null,
): Promise<{
tempPath: string;
fileHash: string;
md5Hash?: string;
sizeBytes: number;
signatureBuffer: Buffer;
}> => {
const reader = (
body ??
new ReadableStream({
start(c) {
c.close();
},
})
).getReader() as ReadableStreamDefaultReader<Uint8Array>;
return streamToTemp(reader, { computeMd5: true, prefix: '/tmp/filedrop-s3-' });
};
/**
* Handles PUT /{bucket}/{key} — uploads an S3 object.
*
* Streams the request body directly to a temporary file to avoid buffering
* the entire payload in memory. This is essential for supporting large
* Docker registry layer blobs (100MB–2GB+).
*
* Supports regular binary uploads, copy-object via `x-amz-copy-source`,
* and tag operations. Large files are stored as chunked objects (across
* multiple Telegram messages), while smaller files use a single Telegram
* message.
*
* @param bucket - The bucket name.
* @param key - The object key.
* @param searchParams - URL query parameters.
* @param headers - The request headers.
* @param req - The incoming HTTP request with the object body.
* @param reqId - The request identifier for S3 headers.
* @returns An S3 response with the object etag or an error.
*/
export const handlePutObject = async (
bucket: string,
key: string,
searchParams: URLSearchParams,
headers: Record<string, string>,
req: Request,
reqId: string,
): Promise<Response> => {
const bucketRecord = await resolveBucketOr404(
bucketRepository,
bucket,
`/${bucket}/${key}`,
reqId,
);
if (bucketRecord instanceof Response) return bucketRecord;
// Tag operations are idempotent no-ops
if (searchParams.has('tagging')) {
return s3Response(null, 204, reqId);
}
// Copy-object path
const copySource = headers['x-amz-copy-source'];
if (copySource) {
return handleCopyObject(bucket, key, copySource, headers, bucketRecord.id, reqId);
}
// Stream body to temp file — O(1) memory, safe for multi-GB blobs
const contentType = headers['content-type'] || 'application/octet-stream';
const streamed = await streamBodyToTemp(req.body);
// H4: Verify body hash against x-amz-content-sha256
const bodyHashError = verifyBodyHash(streamed.fileHash, headers);
if (bodyHashError) {
await cleanupTempFile(streamed.tempPath);
return s3ErrorResponse(
bodyHashError.errorCode || 'BadDigest',
'The x-amz-content-sha256 you specified did not match what we received.',
`/${bucket}/${key}`,
400,
reqId,
);
}
// Content-Length validation: ensure actual body size matches header
const contentLengthHeader = headers['content-length'];
if (contentLengthHeader) {
const declaredLength = Number.parseInt(contentLengthHeader, 10);
if (Number.isFinite(declaredLength) && declaredLength !== streamed.sizeBytes) {
await cleanupTempFile(streamed.tempPath);
return s3ErrorResponse(
'IncompleteBody',
'You did not provide the number of bytes specified by the Content-Length HTTP header.',
`/${bucket}/${key}`,
400,
reqId,
);
}
}
// Content-MD5 validation: use pre-computed MD5 from streaming (no OOM re-read)
const contentMd5 = headers['content-md5'];
if (contentMd5 && contentMd5 !== streamed.md5Hash) {
await cleanupTempFile(streamed.tempPath);
return s3ErrorResponse(
'BadDigest',
'The Content-MD5 you specified did not match what we received.',
`/${bucket}/${key}`,
400,
reqId,
);
}
// M12: Reject oversized bodies
if (streamed.sizeBytes > config.maxRequestBodyBytes) {
await cleanupTempFile(streamed.tempPath);
return s3ErrorResponse(
'EntityTooLarge',
'Your proposed upload exceeds the maximum allowed object size.',
`/${bucket}/${key}`,
400,
reqId,
);
}
// Idempotent PUT: if the object already exists, skip upload
try {
const existing = await fileRepository.findByBucketAndKey(bucketRecord.id, key);
if (existing) {
await cleanupTempFile(streamed.tempPath);
return s3Response(null, 200, reqId, { etag: `"${streamed.fileHash}"` });
}
return await storeFileFromTemp(streamed, key, bucketRecord, contentType, reqId);
} catch (error) {
await cleanupTempFile(streamed.tempPath);
throw error;
}
};
/**
* Stores a streamed file to Telegram storage as an S3 object.
*
* Accepts the result of `streamBodyToTemp` (temp path + hash + size) instead
* of a raw Buffer, enabling O(1) memory usage for multi-GB Docker layer blobs.
*
* Handles both chunked (large files) and single-message (small files) paths.
*
* @param streamed - The streamed file result (temp path, hash, size, signature).
* @param key - The S3 object key.
* @param bucketRecord - The resolved bucket record (id and name).
* @param contentType - The MIME type from the request Content-Type header.
* @param reqId - The request identifier for S3 headers.
* @returns An S3 response with the etag of the stored object.
*/
export const storeFileFromTemp = async (
streamed: { tempPath: string; fileHash: string; sizeBytes: number; signatureBuffer: Buffer },
key: string,
bucketRecord: { id: string; name: string },
contentType: string,
reqId: string,
): Promise<Response> => {
const fileName = key.split('/').pop() || 'file';
const { fileName: finalFileName, mimeType } = ensureExtension(
fileName,
streamed.signatureBuffer,
contentType,
);
const bucketId = bucketRecord.id;
const partFileNamePrefix = `s3-${bucketRecord.name}-${key.replace(/\//g, '_')}`;
if (streamed.sizeBytes > config.telegramChunkSizeBytes) {
const file = await chunkedStorage.storeFileInTelegramChunks({
tempPath: streamed.tempPath,
partFileNamePrefix,
fileName: finalFileName,
mimeType,
sizeBytes: streamed.sizeBytes,
fileType: DEFAULT_FILE_TYPE,
uploaderId: 0,
bucketId,
s3Key: key,
});
await cleanupTempFile(streamed.tempPath);
return s3Response(null, 200, reqId, { etag: `"${file.fileHash}"` });
}
const fileStream = createReadStream(streamed.tempPath);
let forwardResult: ForwardResult;
try {
forwardResult = await botPool.forwardToStorage(fileStream, partFileNamePrefix, 'document');
} catch (error) {
fileStream.destroy();
throw error;
}
fileStream.destroy();
const publicId = nanoid();
await fileRepository.create(
buildNewFile({
publicId,
telegramFileId: forwardResult.telegramFileId,
telegramFileUniqueId: forwardResult.telegramFileUniqueId,
storageChatId: config.storageChatId,
storageMessageId: forwardResult.storageMessageId,
fileName: finalFileName,
mimeType,
sizeBytes: streamed.sizeBytes,
fileType: DEFAULT_FILE_TYPE,
uploaderId: 0,
fileHash: streamed.fileHash,
bucketId,
s3Key: key,
storageBackend: 'telegram',
}),
);
await cleanupTempFile(streamed.tempPath);
return s3Response(null, 200, reqId, { etag: `"${streamed.fileHash}"` });
};
/**
* Handles PUT /{bucket}/{key} with an `x-amz-copy-source` header.
*
* Creates a new file record referencing the same Telegram-stored data as
* the source object. Chunked source objects are not supported for copy.
*
* @param _destBucket - The destination bucket name (unused — bucket record
* already resolved).
* @param destKey - The destination object key.
* @param rawCopySource - The raw `x-amz-copy-source` header value.
* @param headers - The request headers (for conditional copy checks).
* @param destBucketId - The UUID of the destination bucket.
* @param reqId - The request identifier for S3 headers.
* @returns An S3 XML response with the copy result or an error.
*/
export const handleCopyObject = async (
_destBucket: string,
destKey: string,
rawCopySource: string,
headers: Record<string, string>,
destBucketId: string,
reqId: string,
): Promise<Response> => {
const copySource = decodeURIComponent(rawCopySource);
const sourcePath = copySource.startsWith('/') ? copySource.slice(1) : copySource;
const parts = sourcePath.split('/');
const sourceBucket = parts[0];
const sourceKey = parts.slice(1).join('/');
const sourceBucketRecord = await bucketRepository.findByName(sourceBucket);
if (!sourceBucketRecord)
return s3ErrorResponse(
'NoSuchBucket',
'The specified bucket does not exist.',
copySource,
404,
reqId,
);
const sourceFile = await fileRepository.findByBucketAndKey(sourceBucketRecord.id, sourceKey);
if (!sourceFile)
return s3ErrorResponse(
'NoSuchKey',
'The specified key does not exist.',
copySource,
404,
reqId,
);
// Chunked objects cannot be copied yet
if (sourceFile.storageBackend === 'chunked') {
return s3ErrorResponse(
'NotImplemented',
'Copying chunked objects is not yet implemented.',
copySource,
501,
reqId,
);
}
// Conditional copy: if-match / if-none-match checks
// M9: Use stable etag (telegramFileId fallback when fileHash is null) —
// kept inline (not etagOrFallback) because this variant falls back to
// telegramFileId while the read/list paths fall back to nanoid(16).
const sourceEtag = sourceFile.fileHash || sourceFile.telegramFileId;
const ifMatch = headers['x-amz-copy-source-if-match'];
const ifNoneMatch = headers['x-amz-copy-source-if-none-match'];
if (ifMatch && ifMatch !== '*' && ifMatch !== `"${sourceEtag}"`) {
return s3ErrorResponse(
'PreconditionFailed',
'The preconditions you specified did not hold.',
copySource,
412,
reqId,
);
}
if (ifNoneMatch && ifNoneMatch === `"${sourceEtag}"`) {
return s3ErrorResponse(
'PreconditionFailed',
'The preconditions you specified did not hold.',
copySource,
412,
reqId,
);
}
const publicId = nanoid();
await fileRepository.create(
buildNewFile({
publicId,
telegramFileId: sourceFile.telegramFileId,
telegramFileUniqueId: sourceFile.telegramFileUniqueId,
storageChatId: sourceFile.storageChatId,
storageMessageId: sourceFile.storageMessageId,
fileName: sourceFile.fileName,
mimeType: sourceFile.mimeType,
sizeBytes: sourceFile.sizeBytes,
fileType: sourceFile.fileType,
uploaderId: 0,
fileHash: sourceFile.fileHash,
bucketId: destBucketId,
s3Key: destKey,
storageBackend: 'telegram',
}),
);
// copyObjectResultXml quotes the etag itself.
const xml = copyObjectResultXml(etagOrFallback(sourceFile.fileHash), new Date());
return s3Response(xml, 200, reqId, { 'content-type': 'application/xml' });
};
/**
* Handles DELETE /{bucket}/{key} — soft-deletes an S3 object.
*
* @param bucket - The bucket name.
* @param key - The object key to delete.
* @param reqId - The request identifier for S3 headers.
* @returns A 204 response on success, or an S3 XML error.
*/
export const handleDeleteObject = async (
bucket: string,
key: string,
reqId: string,
): Promise<Response> => {
const bucketRecord = await resolveBucketOr404(
bucketRepository,
bucket,
`/${bucket}/${key}`,
reqId,
);
if (bucketRecord instanceof Response) return bucketRecord;
await fileRepository.softDelete(bucketRecord.id, key);
return s3Response(null, 204, reqId);
};
/**
* Handles POST /{bucket}?delete — batch-deletes multiple S3 objects.
*
* Parses the XML Delete request body, soft-deletes each key, and returns
* an XML delete result. The parsed body is validated with
* DeleteObjectsBodySchema (malformed output → MalformedXML 400; the schema
* also enforces the M11 S3 limit of 1000 keys).
*
* @param bucket - The bucket name.
* @param body - The raw XML request body.
* @param reqId - The request identifier for S3 headers.
* @returns An S3 XML response listing deleted keys.
*/
export const handleDeleteObjects = async (
bucket: string,
body: string,
reqId: string,
): Promise<Response> => {
const bucketRecord = await resolveBucketOr404(bucketRepository, bucket, `/${bucket}`, reqId);
if (bucketRecord instanceof Response) return bucketRecord;
const parsed = parseOrNull(DeleteObjectsBodySchema, parseDeleteObjectsBody(body));
// M11: S3 spec limits batch delete to 1000 keys (also enforced by schema)
if (parsed === null || parsed.keys.length > 1000) {
return s3ErrorResponse(
'MalformedXML',
'The XML you provided was not well-formed or did not validate against our published schema. Max 1000 keys per request.',
`/${bucket}`,
400,
reqId,
);
}
const { keys, quiet } = parsed;
const deletedKeys: string[] = [];
const errors: Array<{ key: string; code: string; message: string }> = [];
for (const key of keys) {
const ok = await fileRepository.softDelete(bucketRecord.id, key);
if (ok) {
deletedKeys.push(key);
} else {
// Per S3 spec, deleting a non-existent key is idempotent — report as success
deletedKeys.push(key);
}
}
const xml = quiet ? deleteResultXml([], []) : deleteResultXml(deletedKeys, errors);
return s3Response(xml, 200, reqId, { 'content-type': 'application/xml' });
};
@@ -0,0 +1,238 @@
import { config } from '../../../../env';
import logger from '../../../../shared/logger/index';
import { getErrorMessage } from '../../../../shared/utils/file';
import { verifyPresignedUrl, verifySignature } from '../../../s3/auth';
import { S3_CORS_HEADERS } from '../../../s3/headers';
import { s3ErrorResponse } from '../../../s3/xml';
import {
handleCreateBucket,
handleDeleteBucket,
handleGetBucketVersioning,
handleHeadBucket,
handleListBuckets,
} from './s3-bucket-handlers';
import { REGION, REQUEST_ID, s3Response } from './s3-common';
import { handleListObjectsV1, handleListObjectsV2 } from './s3-listing';
import {
handleAbortMultipartUpload,
handleCompleteMultipartUpload,
handleCreateMultipartUpload,
handleListMultipartUploads,
handleListParts,
handleUploadPart,
} from './s3-multipart-handlers';
import { handleGetObject, handleHeadObject } from './s3-object-read';
import { handleDeleteObject, handleDeleteObjects, handlePutObject } from './s3-object-write';
/**
* Builds an S3 OPTIONS preflight response with CORS headers.
*
* @returns A 204 No Content Response.
*/
export const s3OptionsResponse = (): Response =>
new Response(null, { status: 204, headers: S3_CORS_HEADERS });
/**
* Parses an S3 pathname into bucket and key components.
*
* Supports path-style URLs such as `/bucket-name/key/with/prefix`.
*
* @param pathname - The URL pathname.
* @returns An object with the extracted bucket and key (both may be null).
*/
export const parseS3Path = (pathname: string): { bucket: string | null; key: string | null } => {
const parts = pathname.split('/').filter(Boolean);
if (parts.length === 0) return { bucket: null, key: null };
if (parts.length === 1) return { bucket: parts[0], key: null };
// Decode URI components to match virtual-hosted behavior (H10)
const key = parts
.slice(1)
.map((segment) => decodeURIComponent(segment))
.join('/');
return { bucket: parts[0], key };
};
/**
* Converts a Request's headers into a plain key-value record (all keys
* lowercased) for SigV4 signature verification.
*
* @param req - The incoming HTTP request.
* @returns A record of lowercased header key-value pairs.
*/
export const headersToRecord = (req: Request): Record<string, string> => {
const record: Record<string, string> = {};
for (const [key, value] of req.headers.entries()) {
record[key.toLowerCase()] = value;
}
return record;
};
/**
* Main S3 request dispatcher.
*
* Parses the request (method, path, query parameters, headers), validates
* the SigV4 signature or presigned URL, and dispatches to the appropriate
* bucket, object, or multipart operation handler.
*
* Supports both path-style (`/bucket/key`) and virtual-hosted-style
* (`bucket.example.com/key`) addressing.
*
* @param req - The incoming S3 HTTP request.
* @param virtualHostBucket - When the request was routed through a
* virtual-hosted domain, the extracted bucket
* name; otherwise `null`.
* @returns An S3-formatted Response.
*/
export const handleS3Request = async (
req: Request,
virtualHostBucket: string | null = null,
): Promise<Response> => {
const method = req.method;
const url = new URL(req.url);
const pathname = url.pathname;
const { bucket, key } = virtualHostBucket
? {
bucket: virtualHostBucket,
key: pathname === '/' ? null : decodeURIComponent(pathname.slice(1)),
}
: parseS3Path(pathname);
const headers = headersToRecord(req);
const searchParams = url.searchParams;
const reqId = REQUEST_ID();
// Handle CORS preflight
if (method === 'OPTIONS') {
return s3OptionsResponse();
}
// SigV4 authentication
const isPresigned = searchParams.has('X-Amz-Signature');
const authResult = isPresigned
? await verifyPresignedUrl({
url: req.url,
method,
headers,
s3AccessKey: config.s3AccessKey,
s3SecretKey: config.s3SecretKey,
region: REGION,
})
: await verifySignature(
method,
req.url,
headers,
null,
config.s3AccessKey,
config.s3SecretKey,
REGION,
);
if (!authResult.isValid) {
const status = authResult.errorCode === 'NotImplemented' ? 501 : 403;
const message =
authResult.errorCode === 'NotImplemented'
? 'aws-chunked streaming payloads are not supported.'
: isPresigned
? 'Presigned URL verification failed'
: 'Authentication required';
return s3ErrorResponse(
authResult.errorCode || 'AccessDenied',
message,
pathname,
status,
reqId,
);
}
try {
// ── Root: ListBuckets / Service-level operations ──
if (!bucket) {
if (method === 'GET') {
return handleListBuckets(reqId);
}
return s3ErrorResponse(
'MethodNotAllowed',
'The specified method is not allowed against this resource.',
'/',
405,
reqId,
);
}
// ── Bucket-level operations ──
if (!key) {
if (method === 'GET') {
if (searchParams.has('versioning')) {
return handleGetBucketVersioning(bucket, reqId);
}
if (searchParams.has('uploads')) {
return handleListMultipartUploads(bucket, searchParams, reqId);
}
const listType = searchParams.get('list-type');
if (listType === '2') {
return handleListObjectsV2(bucket, searchParams, reqId);
}
return handleListObjectsV1(bucket, searchParams, reqId);
}
if (method === 'PUT') return handleCreateBucket(bucket, reqId);
if (method === 'HEAD') return handleHeadBucket(bucket, reqId);
if (method === 'DELETE') return handleDeleteBucket(bucket, reqId);
if (method === 'POST') {
if (searchParams.has('delete')) {
const body = await req.text();
return handleDeleteObjects(bucket, body, reqId);
}
if (searchParams.has('tagging')) {
return s3Response(null, 204, reqId);
}
}
return s3ErrorResponse(
'MethodNotAllowed',
'The specified method is not allowed against this resource.',
`/${bucket}`,
405,
reqId,
);
}
// ── Object-level: Multipart operations ──
if (searchParams.has('uploads') && method === 'POST') {
return handleCreateMultipartUpload(bucket, key, searchParams, headers, reqId);
}
if (searchParams.has('uploadId') && searchParams.has('partNumber') && method === 'PUT') {
return handleUploadPart(bucket, key, searchParams, req, reqId);
}
if (searchParams.has('uploadId') && method === 'POST') {
const body = await req.text();
return handleCompleteMultipartUpload(bucket, key, searchParams, body, reqId);
}
if (searchParams.has('uploadId') && method === 'DELETE') {
return handleAbortMultipartUpload(bucket, key, searchParams, reqId);
}
if (searchParams.has('uploadId') && method === 'GET') {
return handleListParts(bucket, key, searchParams, reqId);
}
// ── Standard object operations ──
if (method === 'GET') return handleGetObject(bucket, key, searchParams, headers, reqId);
if (method === 'HEAD') return handleHeadObject(bucket, key, headers, reqId);
if (method === 'PUT') return handlePutObject(bucket, key, searchParams, headers, req, reqId);
if (method === 'DELETE') return handleDeleteObject(bucket, key, reqId);
return s3ErrorResponse(
'MethodNotAllowed',
'The specified method is not allowed against this resource.',
`/${bucket}/${key}`,
405,
reqId,
);
} catch (error: unknown) {
logger.error('S3 operation error', { bucket, key, error: getErrorMessage(error) });
return s3ErrorResponse(
'InternalError',
'We encountered an internal error. Please try again.',
pathname,
500,
reqId,
);
}
};