Remove nGram and edgeNGram token filter names
#38911
Conversation
cbuescher
added
:Search/Analysis
|
Pinging @elastic/es-search |
| @@ -1,9 +1,9 @@ | |||
| [[analysis-edgengram-tokenfilter]] | |||
| === Edge NGram Token Filter | |||
|
|
|||
| A token filter of type `edgeNGram`. | |||
| A token filter of type `edge_ngram`. | |||
There was a problem hiding this comment.
These doc-changes should go back to 7.0 as well.
| @@ -81,7 +81,7 @@ public void testNgramHighlightingWithBrokenPositions() throws IOException { | |||
| .put("analysis.tokenizer.autocomplete.max_gram", 20) | |||
| .put("analysis.tokenizer.autocomplete.min_gram", 1) | |||
| .put("analysis.tokenizer.autocomplete.token_chars", "letter,digit") | |||
| .put("analysis.tokenizer.autocomplete.type", "nGram") | |||
| .put("analysis.tokenizer.autocomplete.type", "ngram") | |||
There was a problem hiding this comment.
This is about the tokenizer, not the filter, but I think we also shouldn't use the camel case version of that one anymore.
| @@ -133,7 +101,7 @@ | |||
| text: "foobar" | |||
| explain: true | |||
| tokenizer: | |||
| type: nGram | |||
| type: ngram | |||
There was a problem hiding this comment.
same here, this is about the tokenizer, but we shouldn't use the camel case name here regardless
Will do. I was wondering if we also need to start throwing errors starting with 7.0 when token filters are used via "nGram" or "edgeNGram" in new indices, so essentially throwing an error where since 6.4 we have issued the deprecation warning, or if we can remove the filter in 8.0 only on the basis of having deprecated it in 6.4? The problem with token filters is AFAIK we currently cannot easily throw errors on e.g. index creation really but only when the filter is used (e.g. an index or analyze operation). Wdyt? |
+1, we should throw an error if the deprecated name is used on an index created in 7.0+. |
* master: Address some CCR REST test case flakiness (elastic#38975) Edits to text in Completion Suggester doc (elastic#38980) SQL: doc polishing [DOCS] Fixes broken formatting SQL: Polish the rest chapter (elastic#38971) Remove `nGram` and `edgeNGram` token filter names (elastic#38911) Add an exception throw if waiting on transport port file fails (elastic#37574) Improve testcluster distribution artifact handling (elastic#38933) Advance max_seq_no before add operation to Lucene (elastic#38879) Reduce global checkpoint sync interval in disruption tests (elastic#38931) [test] disable packaging tests for suse boxes Relax testStressMaybeFlushOrRollTranslogGeneration (elastic#38918) [DOCS] Edits warning in put watch API (elastic#38582) Fix serialization bug in ShardFollowTask after cutting this class over to extend from ImmutableFollowParameters. [DOCS] Updates methods for upgrading machine learning (elastic#38876)
In elastic#30209 we deprecated the camel case `nGram` filter name in favour of `ngram` and did the same for `edgeNGram` and `edge_ngram`. Using these names has been deprecated since 6.4 and is issuing deprecation warnings since then. I think we can remove these filters in 8.0. In a backport of this PR I would change what was a dreprecation warning from 6.4. to an error starting with new indices created in 7.0.
In #30209 we deprecated the camel case `nGram` filter name in favour of `ngram` and did the same for `edgeNGram` and `edge_ngram` and we are removing those names in 8.0. This change disallows using the deprecated names for new indices created in 7.0 by throwing an error if these filters are used. Relates to #38911
In #30209 we deprecated the camel case `nGram` filter name in favour of `ngram` and did the same for `edgeNGram` and `edge_ngram` and we are removing those names in 8.0. This change disallows using the deprecated names for new indices created in 7.0 by throwing an error if these filters are used. Relates to #38911
In #30209 we deprecated the camel case
nGramfilter name in favour ofngramanddid the same for
edgeNGramandedge_ngram. Using these names has been deprecatedsince 6.4 and is issuing deprecation warnings since then.
I think we can remove these filters in 8.0. In a backport of this PR I would change what was a
dreprecation warning from 6.4. to an error starting with new indices created in 7.0.