There are a few more renames to make. But not just renames: the "V2" models should be collapsed with their v1 (no suffix, that is) counterparts and original @odata.type retained for round-tripping. When new ones are created, the "V2" equivalent should always be created because the v1s will cause the service to fail when using the new api-version.
| Old name |
New name |
Notes |
| EdgeNGramTokenFilterV2 |
EdgeNGramTokenFilter |
Collapse with existing EdgeNGramTokenFilter |
| KeywordTokenizerV2 |
KeywordTokenizer |
Collapse with existing KeywordTokenizer (will have separate properties for v1 and "V2") |
| LuceneStandardTokenizerV2 |
LuceneStandardTokenizer |
Collapse with existing LucenseStandardTokenizer |
| NGramTokenFilterV2 |
NGramTokenFilter |
Collapse with existing NGramTokenFilter |
| PathHierarchyTokenizerV2 |
PathHierarchyTokenizer |
(There is no v1) |
| Similarity |
SimilarityAlgorithm |
Do not rename derivative types |
| Suggester |
SearchSuggester |
|
There are a few more renames to make. But not just renames: the "V2" models should be collapsed with their v1 (no suffix, that is) counterparts and original
@odata.typeretained for round-tripping. When new ones are created, the "V2" equivalent should always be created because the v1s will cause the service to fail when using the new api-version.