I have the following API defined for the RTK query:
export const postsApi = createApi({
reducerPath: "postsApi",
baseQuery: fetchBaseQuery({ baseUrl: 'https://myservice.co/api/v2/' }),
tagTypes: ["Posts"],
endpoints: (builder) => ({
getAllPosts: builder.query({
query: () => ({ method: "GET", url: "/posts" }),
transformResponse: (response) => response.posts,
providesTags: (result) =>
result
? [
...result.map(({ id }) => ({ type: "Posts", id })),
{ type: "Posts", id: "LIST" },
]
: [{ type: "Posts", id: "LIST" }],
}),
updatePost: builder.mutation({
query: ({ postId, ...body }) => ({
url: `/posts/${postId}`,
method: "POST",
config: { body },
}),
invalidatesTags: (_, __, arg) => [{ type: "Posts", id: arg.id }],
}),
getPost: builder.query({
query: (postId) => ({
method: "GET",
url: `/posts/${postId}`,
}),
providesTags: (_, __, id) => [{ type: "Posts", id }],
}),
}),
});
export const { useGetAllPostsQuery, useUpdatePostMutation, useGetPostQuery } = postsApi;
What I would like this to do is that when updatePost is called succesfully, it would only invalidate cache for a single post and use the getPost query to refetch the information instead the getAllPosts. Is this anyway possible? The posts are shown in a table fetched with the getAllPosts query.
No. RTK-Query is a document cache (full response = document), not a normalized cache. It does not know anything about the contents of the cached responses and their structure - and it will never try to "stitch" anything in there by itself.
You can do that manually with an optimistic update, but everything RTK-Q does automatically is refetch, which should be more than enough in most use cases.