v5.1.1

Patch Changes

• 5ced9fa: Hygiene: CI now tests Node 20, 22 and 24; README notes the project is personal and not accepting contributions; removed redundant .npmignore.

v5.1.0

Minor Changes

• 027d0b8: Add queryListItems (CAML queries via GetItems, for lists above the 5,000-item view threshold, with pagingInfo support), batch (several REST calls in one /_api/$batch round trip on SharePoint 2016+, consecutive writes grouped into a changeset, one BatchResponse per request), and Retry-After handling: Failure.retryAfterMs carries the server's requested wait, and the retry logic honours it — or stops early when it exceeds maxDelayMs. parseRetryAfter is exported. New types: QueryListItemsOptions, BatchOptions, BatchRequest, BatchResponse, BatchMethod, ItemsAction, ItemsActionFactory.

v5.0.0

Major Changes

• 24325b0: Remove the APIs deprecated during 4.x.Breaking changes- accessToken has been removed from every write action. Use requestDigest, or omit it and let the connection manage the digest.- addListItem.payload and updateListItem.patch no longer accept a JSON string; pass a plain object. The entity type is resolved automatically.- addListItem resolves to the created item (T & { Id: number }, via a new type parameter) instead of a raw JSON string. Replace JSON.parse(res.data).Id with res.data.Id. Both the nometadata and verbose response shapes are accepted; a response without an Id is a PARSE failure.- The NTLM transport (get, post, auth, request, and the message builders) is no longer exported. SpRestError and parseSharePointError remain public.See "Migrating to 5.0" in the README. Everything added in 4.x — structuredFailure codes, signal, agentOptions, retry, typed reads, paging, themanaged digest, and streams — is unchanged.

v4.3.0

Minor Changes

• f4d50fe: Managed request digest, object payloads, and streaming uploads/downloads.- requestDigest is now optional on every write. When omitted, the connection fetches a digest, caches it until a minute before SharePoint says it expires, shares one fetch between concurrent writes, and on a 403 refreshes it and retries the write once. Pass requestDigest to bypass the cache. Standalone factories accept a ConnectionContext (from the new createConnectionContext) as a second argument to share the cache.- addListItem.payload and updateListItem.patch accept plain objects. The list's entity type is looked up once per list and cached; include __metadata.type yourself to skip the lookup. The JSON-string form still works but is deprecated and will be removed in 5.0.- Streams. addDocumentToLibrary accepts a Readable as payload (plus optional contentLength; chunked otherwise). New getDocumentStreamFromLibrary returns a Readable instead of a Buffer. A stream upload is not retried on a stale managed digest.- New exports: createConnectionContext, getDocumentStreamFromLibrary, and types ConnectionContext, DigestCache, EntityTypeCache, ListItemPayload, GetDocumentStreamFromLibraryOptions.- Quick start and examples simplified accordingly.

v4.2.0

Minor Changes

• fc36ffd: Typed reads, paging helpers, and opt-in retry.- Type parameters on reads. getListContents<T>, getFolderContents<T> and getListItem<T> return T[] / T instead of unknown. The default is still unknown, so existing code compiles unchanged.- Paging. A successful getListContents / getFolderContents result now carries nextLink when SharePoint has more items. New actions follow it: getNextPage({ nextLink }) (same-origin links only), getAllListContents / getAllFolderContents (every page in one array; a failed page fails the whole call), and iterateListContents / iterateFolderContents (async iterators yielding one PageResult per page). All accept maxPages.- Retry. SiteConnectionOptions.retry enables exponential-backoff retry (full jitter) of read actions that fail with NETWORK, TIMEOUT, or HTTP 429/502/503/504. Writes are never retried. isTransientFailure is exported as the default predicate.- New exported types: Page, PageResult, PagedAction, TypedAction, PagedIterator, RetryOptions, GetNextPageOptions, PagingOptions, GetAllListContentsOptions, GetAllFolderContentsOptions.

v4.1.0

Minor Changes

• f0642b4: Structured failures, request cancellation, TLS options and startup validation.- Failure is now structured. Every failure carries code ('HTTP' | 'AUTH' | 'TIMEOUT' | 'ABORTED' | 'NETWORK' | 'PARSE' | 'UNKNOWN'), plus status and details ({ code, message } parsed from SharePoint's error body) for HTTP responses. Branch on code instead of matching error text. code is a required property of Failure; code that constructs Failure objects by hand must add it.- Error messages no longer include a redundant Error: prefix and now read Error in <action>: HTTP <status>: <SharePoint message> for server responses.- signal?: AbortSignal is accepted by every action; aborting yields a Failure with code ABORTED. timeout is documented as a socket inactivity timeout.- agentOptions?: https.AgentOptions on SiteConnectionOptions passes ca, rejectUnauthorized, client certificates, lookup, etc. to the HTTP agent, so internal CAs no longer need NODE_TLS_REJECT_UNAUTHORIZED=0.- createSiteConnection validates its options and throws a TypeError for a protocol or path inside site, a serverRelativeUrl without a leading slash or with a trailing slash, or an unknown protocol.- SpRestError and parseSharePointError are exported.- Supported Node.js range widened to >=20.- Published with npm provenance.
• f0642b4: Structured failures, request cancellation, TLS options and startup validation.

v1.0.1

Patch Changes

• 4ddb1a3: Resolve type error

v1.0.0

Major Changes

• 0cd91e1: Move to stable

v0.6.0

Minor Changes

• 8643fc7: Added support to set timeout and binary in configuration

v0.5.2

Patch Changes

• a475240: Improved error messaging