Update scope search index manager to include analyze document #133
Conversation
| @@ -1563,6 +1563,7 @@ The SDK should not try to intercept a "bucket.scope.index" name and extract "ind | |||
| The implementation of `ScopeSearchIndexManager` is identical to `SearchIndexManagement`, except for the endpoints and some error handling details: | |||
|
|
|||
| * `GetIndex`, `UpsertIndex`, `DropIndex`: use endpoint `/api/bucket/{bucketName}/scope/{scopeName}/index/{indexName}` with GET, PUT and DELETE respectively. | |||
| * `AnalyzeDocument`: use endpoint `/api/index/{bucketName}.{scopeName}.{indexName}/analyzeDoc` with POST, there is no scoped endpoint for analyze document. | |||
There was a problem hiding this comment.
Ah, really? May have a bug in the Java SDK then.
There was a problem hiding this comment.
Apparently so, from elsewhere (and from someone else)
They told me:
so there isn’t a scoped endpoint for analyseDoc
you’ll have to provide the fully qualified index name
There was a problem hiding this comment.
This just seems to be a server bug - why have a scoped index endpoint for every operation except analyzeDoc?
|
@programmatix Is the change suggested by this PR something we want to implement or do we want to still use the |
Someone needs to raise it with FTS I think. It seems at the very least downright odd that only one endpoint is missing a scoped variant. |
I've filed an MB: https://issues.couchbase.com/browse/MB-60643 |
|
The FTS team have now added a scoped endpoint for
I think we can now close this PR. |
No description provided.