We have a FetchXml query with a couple of link entities, one of which is to the annotation entity in order to retrieve an attached image's base64 encoded content. This is so we can display an image inline on a page using data uris.
The problem is that the image displays trimmed because the documentbody returned is actually cut off at 2,000 characters.
The problem ended up being an issue with the query itself. We had distinct="true" set on the fetch query, removing this attribute then caused the result to contain the entire documentbody of the image. Very strange.
来源：https://stackoverflow.com/questions/16719349/fetchxml-query-returns-trimmed-documentbody