Add per-field metadata. (elastic#49419)

This PR adds per-field metadata that can be set in the mappings and is later
returned by the field capabilities API. This metadata is completely opaque to
Elasticsearch but may be used by tools that index data in Elasticsearch to
communicate metadata about fields with tools that then search this data. A
typical example that has been requested in the past is the ability to attach
a unit to a numeric field.

In order to not bloat the cluster state, Elasticsearch requires that this
metadata be small:
 - keys can't be longer than 20 chars,
 - values can only be numbers or strings of no more than 50 chars - no inner
   arrays or objects,
 - the metadata can't have more than 5 keys in total.

Given that metadata is opaque to Elasticsearch, field capabilities don't try to
do anything smart when merging metadata about multiple indices, the union of
all field metadatas is returned.

Here is how the meta might look like in mappings:

```json
{
  "properties": {
    "latency": {
      "type": "long",
      "meta": {
        "unit": "ms"
      }
    }
  }
}
```

And then in the field capabilities response:

```json
{
  "latency": {
    "long": {
      "searchable": true,
      "aggreggatable": true,
      "meta": {
        "unit": [ "ms" ]
      }
    }
  }
}
```

When there are no conflicts, values are arrays of size 1, but when there are
conflicts, Elasticsearch includes all unique values in this array, without
giving ways to know which index has which metadata value:

```json
{
  "latency": {
    "long": {
      "searchable": true,
      "aggreggatable": true,
      "meta": {
        "unit": [ "ms", "ns" ]
      }
    }
  }
}
```

Closes elastic#33267

client/rest-high-level/src/test/java/org/elasticsearch/client/SearchIT.java

@@ -1229,11 +1229,11 @@ public void testFieldCaps() throws IOException {
         assertEquals(2, ratingResponse.size());
 
         FieldCapabilities expectedKeywordCapabilities = new FieldCapabilities(
-            "rating", "keyword", true, true, new String[]{"index2"}, null, null);
+            "rating", "keyword", true, true, new String[]{"index2"}, null, null, Collections.emptyMap());
         assertEquals(expectedKeywordCapabilities, ratingResponse.get("keyword"));
 
         FieldCapabilities expectedLongCapabilities = new FieldCapabilities(
-            "rating", "long", true, true, new String[]{"index1"}, null, null);
+            "rating", "long", true, true, new String[]{"index1"}, null, null, Collections.emptyMap());
         assertEquals(expectedLongCapabilities, ratingResponse.get("long"));
 
         // Check the capabilities for the 'field' field.
@@ -1242,7 +1242,7 @@ public void testFieldCaps() throws IOException {
         assertEquals(1, fieldResponse.size());
 
         FieldCapabilities expectedTextCapabilities = new FieldCapabilities(
-            "field", "text", true, false);
+            "field", "text", true, false, Collections.emptyMap());
         assertEquals(expectedTextCapabilities, fieldResponse.get("text"));
     }
 

docs/reference/mapping/params.asciidoc

@@ -8,23 +8,24 @@ parameters that are used by <<mapping-types,field mappings>>:
 The following mapping parameters are common to some or all field datatypes:
 
 * <<analyzer,`analyzer`>>
-* <<normalizer, `normalizer`>>
 * <<mapping-boost,`boost`>>
 * <<coerce,`coerce`>>
 * <<copy-to,`copy_to`>>
 * <<doc-values,`doc_values`>>
 * <<dynamic,`dynamic`>>
+* <<eager-global-ordinals,`eager_global_ordinals`>>
 * <<enabled,`enabled`>>
 * <<fielddata,`fielddata`>>
-* <<eager-global-ordinals,`eager_global_ordinals`>>
+* <<multi-fields,`fields`>>
 * <<mapping-date-format,`format`>>
 * <<ignore-above,`ignore_above`>>
 * <<ignore-malformed,`ignore_malformed`>>
 * <<index-options,`index_options`>>
 * <<index-phrases,`index_phrases`>>
 * <<index-prefixes,`index_prefixes`>>
 * <<mapping-index,`index`>>
-* <<multi-fields,`fields`>>
+* <<mapping-field-meta,`meta`>>
+* <<normalizer, `normalizer`>>
 * <<norms,`norms`>>
 * <<null-value,`null_value`>>
 * <<position-increment-gap,`position_increment_gap`>>
@@ -37,8 +38,6 @@ The following mapping parameters are common to some or all field datatypes:
 
 include::params/analyzer.asciidoc[]
 
-include::params/normalizer.asciidoc[]
-
 include::params/boost.asciidoc[]
 
 include::params/coerce.asciidoc[]
@@ -49,10 +48,10 @@ include::params/doc-values.asciidoc[]
 
 include::params/dynamic.asciidoc[]
 
-include::params/enabled.asciidoc[]
-
 include::params/eager-global-ordinals.asciidoc[]
 
+include::params/enabled.asciidoc[]
+
 include::params/fielddata.asciidoc[]
 
 include::params/format.asciidoc[]
@@ -69,8 +68,12 @@ include::params/index-phrases.asciidoc[]
 
 include::params/index-prefixes.asciidoc[]
 
+include::params/meta.asciidoc[]
+
 include::params/multi-fields.asciidoc[]
 
+include::params/normalizer.asciidoc[]
+
 include::params/norms.asciidoc[]
 
 include::params/null-value.asciidoc[]

docs/reference/mapping/params/meta.asciidoc

@@ -0,0 +1,31 @@
+[[mapping-field-meta]]
+=== `meta`
+
+Metadata attached to the field. This metadata is opaque to Elasticsearch, it is
+only useful for multiple applications that work on the same indices to share
+meta information about fields such as units
+
+[source,console]
+------------
+PUT my_index
+{
+  "mappings": {
+    "properties": {
+      "latency": {
+        "type": "long",
+        "meta": {
+          "unit": "ms"
+        }
+      }
+    }
+  }
+}
+------------
+// TEST
+
+NOTE: Field metadata enforces at most 5 entries, that keys have a length that
+is less than or equal to 20, and that values are strings whose length is less
+than or equal to 50.
+
+NOTE: Field metadata is updatable by submitting a mapping update. The metadata
+of the update will override the metadata of the existing field.

docs/reference/mapping/types/boolean.asciidoc

@@ -120,3 +120,6 @@ The following parameters are accepted by `boolean` fields:
     the <<mapping-source-field,`_source`>> field. Accepts `true` or `false`
     (default).
 
+<<mapping-field-meta,`meta`>>::
+
+    Metadata about the field.

docs/reference/mapping/types/date.asciidoc

@@ -137,3 +137,7 @@ The following parameters are accepted by `date` fields:
     Whether the field value should be stored and retrievable separately from
     the <<mapping-source-field,`_source`>> field. Accepts `true` or `false`
     (default).
+
+<<mapping-field-meta,`meta`>>::
+
+    Metadata about the field.

docs/reference/mapping/types/keyword.asciidoc

@@ -115,6 +115,10 @@ The following parameters are accepted by `keyword` fields:
     when building a query for this field.
     Accepts `true` or `false` (default).
 
+<<mapping-field-meta,`meta`>>::
+
+    Metadata about the field.
+
 NOTE: Indexes imported from 2.x do not support `keyword`. Instead they will
 attempt to downgrade `keyword` into `string`. This allows you to merge modern
 mappings with legacy mappings. Long lived indexes will have to be recreated

docs/reference/mapping/types/numeric.asciidoc

@@ -149,6 +149,10 @@ The following parameters are accepted by numeric types:
     the <<mapping-source-field,`_source`>> field. Accepts `true` or `false`
     (default).
 
+<<mapping-field-meta,`meta`>>::
+
+    Metadata about the field.
+
 [[scaled-float-params]]
 ==== Parameters for `scaled_float`
 

docs/reference/mapping/types/text.asciidoc

@@ -143,3 +143,7 @@ The following parameters are accepted by `text` fields:
 
     Whether term vectors should be stored for an <<mapping-index,`analyzed`>>
     field. Defaults to `no`.
+
+<<mapping-field-meta,`meta`>>::
+
+    Metadata about the field.

docs/reference/search/field-caps.asciidoc

@@ -78,6 +78,12 @@ include::{docdir}/rest-api/common-parms.asciidoc[tag=index-ignore-unavailable]
   The list of indices where this field is not aggregatable, or null if all 
   indices have the same definition for the field.
 
+`meta`::
+  Merged metadata across all indices as a map of string keys to arrays of values.
+  A value length of 1 indicates that all indices had the same value for this key,
+  while a length of 2 or more indicates that not all indices had the same value
+  for this key.
+
 
 [[search-field-caps-api-example]]
 ==== {api-examples-title}

modules/mapper-extras/src/test/java/org/elasticsearch/index/mapper/ScaledFloatFieldMapperTests.java

@@ -27,6 +27,7 @@
 import org.elasticsearch.common.xcontent.XContentFactory;
 import org.elasticsearch.common.xcontent.XContentType;
 import org.elasticsearch.index.IndexService;
+import org.elasticsearch.index.mapper.MapperService.MergeReason;
 import org.elasticsearch.plugins.Plugin;
 import org.elasticsearch.test.ESSingleNodeTestCase;
 import org.elasticsearch.test.InternalSettingsPlugin;
@@ -35,6 +36,7 @@
 import java.io.IOException;
 import java.util.Arrays;
 import java.util.Collection;
+import java.util.Collections;
 import java.util.List;
 
 import static org.hamcrest.Matchers.containsString;
@@ -353,4 +355,33 @@ public void testRejectIndexOptions() throws IOException {
         MapperParsingException e = expectThrows(MapperParsingException.class, () -> parser.parse("type", new CompressedXContent(mapping)));
         assertThat(e.getMessage(), containsString("index_options not allowed in field [foo] of type [scaled_float]"));
     }
+
+    public void testMeta() throws Exception {
+        String mapping = Strings.toString(XContentFactory.jsonBuilder().startObject().startObject("_doc")
+                .startObject("properties").startObject("field").field("type", "scaled_float")
+                .field("meta", Collections.singletonMap("foo", "bar"))
+                .field("scaling_factor", 10.0)
+                .endObject().endObject().endObject().endObject());
+
+        DocumentMapper mapper = indexService.mapperService().merge("_doc",
+                new CompressedXContent(mapping), MergeReason.MAPPING_UPDATE);
+        assertEquals(mapping, mapper.mappingSource().toString());
+
+        String mapping2 = Strings.toString(XContentFactory.jsonBuilder().startObject().startObject("_doc")
+                .startObject("properties").startObject("field").field("type", "scaled_float")
+                .field("scaling_factor", 10.0)
+                .endObject().endObject().endObject().endObject());
+        mapper = indexService.mapperService().merge("_doc",
+                new CompressedXContent(mapping2), MergeReason.MAPPING_UPDATE);
+        assertEquals(mapping2, mapper.mappingSource().toString());
+
+        String mapping3 = Strings.toString(XContentFactory.jsonBuilder().startObject().startObject("_doc")
+                .startObject("properties").startObject("field").field("type", "scaled_float")
+                .field("meta", Collections.singletonMap("baz", "quux"))
+                .field("scaling_factor", 10.0)
+                .endObject().endObject().endObject().endObject());
+        mapper = indexService.mapperService().merge("_doc",
+                new CompressedXContent(mapping3), MergeReason.MAPPING_UPDATE);
+        assertEquals(mapping3, mapper.mappingSource().toString());
+    }
 }

rest-api-spec/src/main/resources/rest-api-spec/test/field_caps/10_basic.yml

@@ -317,4 +317,3 @@ setup:
   - match: {fields.misc.unmapped.searchable:            false}
   - match: {fields.misc.unmapped.aggregatable:          false}
   - match: {fields.misc.unmapped.indices:               ["test2", "test3"]}
-

rest-api-spec/src/main/resources/rest-api-spec/test/field_caps/20_meta.yml

@@ -0,0 +1,65 @@
+---
+"Merge metadata across multiple indices":
+
+  - skip:
+      version: " - 7.99.99"
+      reason: Metadata support was added in 7.6
+
+  - do:
+        indices.create:
+          index: test1
+          body:
+              mappings:
+                properties:
+                  latency:
+                    type: long
+                    meta:
+                      unit: ms
+                      metric_type: gauge
+
+  - do:
+        indices.create:
+          index: test2
+          body:
+              mappings:
+                properties:
+                  latency:
+                    type: long
+                    meta:
+                      unit: ns
+                      metric_type: gauge
+
+  - do:
+        indices.create:
+          index: test3
+
+  - do:
+      field_caps:
+        index: test3
+        fields: [latency]
+
+  - is_false: fields.latency.long.meta.unit
+
+  - do:
+      field_caps:
+        index: test1
+        fields: [latency]
+
+  - match: {fields.latency.long.meta.unit:              ["ms"]}
+  - match: {fields.latency.long.meta.metric_type:       ["gauge"]}
+
+  - do:
+      field_caps:
+        index: test1,test3
+        fields: [latency]
+
+  - match: {fields.latency.long.meta.unit:              ["ms"]}
+  - match: {fields.latency.long.meta.metric_type:       ["gauge"]}
+
+  - do:
+      field_caps:
+        index: test1,test2,test3
+        fields: [latency]
+
+  - match: {fields.latency.long.meta.unit:              ["ms", "ns"]}
+  - match: {fields.latency.long.meta.metric_type:       ["gauge"]}

rest-api-spec/src/main/resources/rest-api-spec/test/indices.put_mapping/10_basic.yml

@@ -108,3 +108,39 @@
 
   - match: { error.type: "illegal_argument_exception" }
   - match: { error.reason: "Types cannot be provided in put mapping requests, unless the include_type_name parameter is set to true." }
+
+---
+"Update per-field metadata":
+
+  - skip:
+      version:     " - 7.99.99"
+      reason:      "Per-field meta was introduced in 7.6"
+
+  - do:
+      indices.create:
+        index: test_index
+        body:
+          mappings:
+            properties:
+              foo:
+                type: keyword
+                meta:
+                  bar: baz
+
+  - do:
+      indices.put_mapping:
+        index: test_index
+        body:
+          properties:
+            foo:
+              type: keyword
+              meta:
+                baz: quux
+
+  - do:
+      indices.get_mapping:
+        index: test_index
+
+  - is_false: test_index.mappings.properties.foo.meta.bar
+  - match: { test_index.mappings.properties.foo.meta.baz: "quux" }
+