Skip to content

Commit

Permalink
[8.x] Update semantic_text query to use highlighting option (#205795) (…
Browse files Browse the repository at this point in the history 
…#207125)

# Backport

This will backport the following commits from `main` to `8.x`:
- [Update semantic_text query to use highlighting option
(#205795)](#205795)

<!--- Backport version: 9.4.3 -->

### Questions ?
Please refer to the [Backport tool
documentation](https://github.com/sqren/backport)

<!--BACKPORT [{"author":{"name":"Saikat
Sarkar","email":"[email protected]"},"sourceCommit":{"committedDate":"2025-01-17T18:42:15Z","message":"Update
semantic_text query to use highlighting option (#205795)\n\nThis PR
addresses
[this\r\nissue](elastic/search-team#8928) by
replacing\r\nthe current semantic_text implementation, which uses
inner_hit, with\r\nsemantic_text
highlighting.\r\n\r\n\r\nhttps://github.com/user-attachments/assets/bac8abf7-ec50-4463-b0ad-d3152872253a\r\n\r\n---------\r\n\r\nCo-authored-by:
Joseph McElroy <[email protected]>\r\nCo-authored-by: Elastic
Machine
<[email protected]>","sha":"575d57e8fa7c75e1cf5549c839acb5a7d901611a","branchLabelMapping":{"^v9.0.0$":"main","^v8.18.0$":"8.x","^v(\\d+).(\\d+).\\d+$":"$1.$2"}},"sourcePullRequest":{"labels":["release_note:skip","v9.0.0","Team:EnterpriseSearch","backport:prev-minor","v8.18.0"],"title":"Update
semantic_text query to use highlighting
option","number":205795,"url":"https://github.com/elastic/kibana/pull/205795","mergeCommit":{"message":"Update
semantic_text query to use highlighting option (#205795)\n\nThis PR
addresses
[this\r\nissue](elastic/search-team#8928) by
replacing\r\nthe current semantic_text implementation, which uses
inner_hit, with\r\nsemantic_text
highlighting.\r\n\r\n\r\nhttps://github.com/user-attachments/assets/bac8abf7-ec50-4463-b0ad-d3152872253a\r\n\r\n---------\r\n\r\nCo-authored-by:
Joseph McElroy <[email protected]>\r\nCo-authored-by: Elastic
Machine
<[email protected]>","sha":"575d57e8fa7c75e1cf5549c839acb5a7d901611a"}},"sourceBranch":"main","suggestedTargetBranches":["8.x"],"targetPullRequestStates":[{"branch":"main","label":"v9.0.0","branchLabelMappingKey":"^v9.0.0$","isSourceBranch":true,"state":"MERGED","url":"https://github.com/elastic/kibana/pull/205795","number":205795,"mergeCommit":{"message":"Update
semantic_text query to use highlighting option (#205795)\n\nThis PR
addresses
[this\r\nissue](elastic/search-team#8928) by
replacing\r\nthe current semantic_text implementation, which uses
inner_hit, with\r\nsemantic_text
highlighting.\r\n\r\n\r\nhttps://github.com/user-attachments/assets/bac8abf7-ec50-4463-b0ad-d3152872253a\r\n\r\n---------\r\n\r\nCo-authored-by:
Joseph McElroy <[email protected]>\r\nCo-authored-by: Elastic
Machine
<[email protected]>","sha":"575d57e8fa7c75e1cf5549c839acb5a7d901611a"}},{"branch":"8.x","label":"v8.18.0","branchLabelMappingKey":"^v8.18.0$","isSourceBranch":false,"state":"NOT_CREATED"}]}]
BACKPORT-->

Co-authored-by: Saikat Sarkar <[email protected]>
  • Loading branch information
@kibanamachine @saikatsarkar056
kibanamachine and saikatsarkar056 authored Jan 17, 2025
1 parent 11ca9a3 commit 7ad184b
Show file tree
Hide file tree
Showing 7 changed files with 104 additions and 145 deletions.
11 changes: 6 additions & 5 deletions ...layground/public/components/view_code/examples/__snapshots__/py_lang_client.test.tsx.snap
View file

Some generated files are not rendered by default. Learn more about how customized files appear on GitHub.

11 changes: 6 additions & 5 deletions .../search/plugins/search_playground/public/components/view_code/examples/py_lang_client.tsx
View file
Original file line number Diff line number Diff line change
Expand Up @@ -40,11 +40,12 @@ def get_elasticsearch_results():
def create_openai_prompt(results):
context = ""
for hit in results:
inner_hit_path = f"{hit['_index']}.{index_source_fields.get(hit['_index'])[0]}"
## For semantic_text matches, we need to extract the text from the inner_hits
if 'inner_hits' in hit and inner_hit_path in hit['inner_hits']:
context += '\\n --- \\n'.join(inner_hit['_source']['text'] for inner_hit in hit['inner_hits'][inner_hit_path]['hits']['hits'])
## For semantic_text matches, we need to extract the text from the highlighted field
if "highlight" in hit:
highlighted_texts = []
for values in hit["highlight"].values():
highlighted_texts.extend(values)
context += "\\n --- \\n".join(highlighted_texts)
else:
source_field = index_source_fields.get(hit["_index"])[0]
hit_context = hit["_source"][source_field]
Expand Down
56 changes: 24 additions & 32 deletions x-pack/solutions/search/plugins/search_playground/public/utils/create_query.test.ts
View file
Original file line number Diff line number Diff line change
Expand Up @@ -516,20 +516,9 @@ describe('create_query', () => {
{
standard: {
query: {
nested: {
inner_hits: {
_source: ['field2.inference.chunks.text'],
name: 'index1.field2',
size: 2,
},
path: 'field2.inference.chunks',
query: {
sparse_vector: {
field: 'field2.inference.chunks.embeddings',
inference_id: 'model2',
query: '{query}',
},
},
semantic: {
field: 'field2',
query: '{query}',
},
},
},
Expand All @@ -542,6 +531,15 @@ describe('create_query', () => {
],
},
},
highlight: {
fields: {
field2: {
number_of_fragments: 2,
order: 'score',
type: 'semantic',
},
},
},
});
});

Expand Down Expand Up @@ -638,24 +636,9 @@ describe('create_query', () => {
{
standard: {
query: {
nested: {
inner_hits: {
_source: ['field2.inference.chunks.text'],
name: 'index1.field2',
size: 2,
},
path: 'field2.inference.chunks',
query: {
knn: {
field: 'field2.inference.chunks.embeddings',
query_vector_builder: {
text_embedding: {
model_id: 'model2',
model_text: '{query}',
},
},
},
},
semantic: {
field: 'field2',
query: '{query}',
},
},
},
Expand All @@ -668,6 +651,15 @@ describe('create_query', () => {
],
},
},
highlight: {
fields: {
field2: {
number_of_fragments: 2,
order: 'score',
type: 'semantic',
},
},
},
});
});

Expand Down
109 changes: 54 additions & 55 deletions x-pack/solutions/search/plugins/search_playground/public/utils/create_query.ts
View file
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,7 @@
* 2.0.
*/

import { RetrieverContainer } from '@elastic/elasticsearch/lib/api/types';
import { RetrieverContainer, SearchHighlight } from '@elastic/elasticsearch/lib/api/types';
import { IndicesQuerySourceFields, QuerySourceFields } from '../types';

export type IndexFields = Record<string, string[]>;
Expand Down Expand Up @@ -36,6 +36,8 @@ const SUGGESTED_SOURCE_FIELDS = [
'text_field',
];

const SEMANTIC_FIELD_TYPE = 'semantic';

interface Matches {
queryMatches: any[];
knnMatches: any[];
Expand All @@ -52,7 +54,7 @@ export function createQuery(
rerankOptions: ReRankOptions = {
rrf: true,
}
): { retriever: RetrieverContainer } {
): { retriever: RetrieverContainer; highlight?: SearchHighlight } {
const indices = Object.keys(fieldDescriptors);
const boolMatches = Object.keys(fields).reduce<Matches>(
(acc, index) => {
Expand All @@ -64,60 +66,8 @@ export function createQuery(

const semanticMatches = indexFields.map((field) => {
const semanticField = indexFieldDescriptors.semantic_fields.find((x) => x.field === field);
const isSourceField = sourceFields[index].includes(field);

// this is needed to get the inner_hits for the source field
// we cant rely on only the semantic field
// in future inner_hits option will be added to semantic
if (semanticField && isSourceField) {
if (semanticField.embeddingType === 'dense_vector') {
const filter =
semanticField.indices.length < indices.length
? { filter: { terms: { _index: semanticField.indices } } }
: {};

return {
nested: {
path: `${semanticField.field}.inference.chunks`,
query: {
knn: {
field: `${semanticField.field}.inference.chunks.embeddings`,
...filter,
query_vector_builder: {
text_embedding: {
model_id: semanticField.inferenceId,
model_text: '{query}',
},
},
},
},
inner_hits: {
size: 2,
name: `${index}.${semanticField.field}`,
_source: [`${semanticField.field}.inference.chunks.text`],
},
},
};
} else if (semanticField.embeddingType === 'sparse_vector') {
return {
nested: {
path: `${semanticField.field}.inference.chunks`,
query: {
sparse_vector: {
inference_id: semanticField.inferenceId,
field: `${semanticField.field}.inference.chunks.embeddings`,
query: '{query}',
},
},
inner_hits: {
size: 2,
name: `${index}.${semanticField.field}`,
_source: [`${semanticField.field}.inference.chunks.text`],
},
},
};
}
} else if (semanticField) {
if (semanticField) {
return {
semantic: {
field: semanticField.field,
Expand Down Expand Up @@ -241,12 +191,34 @@ export function createQuery(

// for single Elser support to make it easy to read - skips bool query
if (boolMatches.queryMatches.length === 1 && boolMatches.knnMatches.length === 0) {
const semanticField = boolMatches.queryMatches[0].semantic?.field ?? null;

let isSourceField = false;
indices.forEach((index) => {
if (sourceFields[index].includes(semanticField)) {
isSourceField = true;
}
});

return {
retriever: {
standard: {
query: boolMatches.queryMatches[0],
},
},
...(isSourceField
? {
highlight: {
fields: {
[semanticField]: {
type: SEMANTIC_FIELD_TYPE,
number_of_fragments: 2,
order: 'score',
},
},
},
}
: {}),
};
}

Expand Down Expand Up @@ -285,12 +257,39 @@ export function createQuery(
};
});

const semanticFields = matches
.filter((match) => match.semantic)
.map((match) => match.semantic.field)
.filter((field) => {
let isSourceField = false;
indices.forEach((index) => {
if (sourceFields[index].includes(field)) {
isSourceField = true;
}
});
return isSourceField;
});

return {
retriever: {
rrf: {
retrievers,
},
},
...(semanticFields.length > 0
? {
highlight: {
fields: semanticFields.reduce((acc, field) => {
acc[field] = {
type: SEMANTIC_FIELD_TYPE,
number_of_fragments: 2,
order: 'score',
};
return acc;
}, {}),
},
}
: {}),
};
}

Expand Down
14 changes: 1 addition & 13 deletions x-pack/solutions/search/plugins/search_playground/server/lib/conversational_chain.test.ts
View file
Original file line number Diff line number Diff line change
Expand Up @@ -237,19 +237,7 @@ describe('conversational chain', () => {
{
_index: 'index',
_id: '1',
inner_hits: {
'index.field': {
hits: {
hits: [
{
_source: {
text: 'value',
},
},
],
},
},
},
highlight: { field: ['value'] },
},
],
expectedDocs: [
Expand Down
41 changes: 11 additions & 30 deletions ...utions/search/plugins/search_playground/server/utils/get_value_for_selected_field.test.ts
View file
Original file line number Diff line number Diff line change
Expand Up @@ -78,49 +78,30 @@ describe('getValueForSelectedField', () => {
expect(getValueForSelectedField(hit, 'bla.sources')).toBe('');
});

test('should return when its a chunked passage', () => {
test('should return when it has highlighted messages', () => {
const hit = {
_index: 'sample-index',
_index: 'books',
_id: '8jSNY48B6iHEi98DL1C-',
_score: 0.7789394,
_source: {
test: 'The Shawshank Redemption',
test: 'The Big Bang and Black Holes',
metadata: {
source:
'Over the course of several years, two convicts form a friendship, seeking consolation and, eventually, redemption through basic compassion',
'This book explores the origins of the universe, beginning with the Big Bang—an immense explosion that created space, time, and matter. It delves into how black holes, regions of space where gravity is so strong that not even light can escape, play a crucial role in the evolution of galaxies and the universe as a whole. Stephen Hawking’s groundbreaking discoveries about black hole radiation, often referred to as Hawking Radiation, are also discussed in detail.',
},
},
inner_hits: {
'sample-index.test': {
hits: {
hits: [
{
_source: {
text: 'Over the course of several years',
},
},
{
_source: {
text: 'two convicts form a friendship',
},
},
{
_source: {
text: 'seeking consolation and, eventually, redemption through basic compassion',
},
},
],
},
},
highlight: {
test: [
'This book explores the origins of the universe.',
'The beginning with the Big Bang—an immense explosion that created space, time, and matter. It delves into how black holes, regions of space where gravity is so strong that not even light can escape, play a crucial role in the evolution of galaxies and the universe as a whole. Stephen Hawking’s groundbreaking discoveries about black hole radiation, often referred to as Hawking Radiation, are also discussed in detail.',
],
},
};

expect(getValueForSelectedField(hit as any, 'test')).toMatchInlineSnapshot(`
"Over the course of several years
---
two convicts form a friendship
"This book explores the origins of the universe.
---
seeking consolation and, eventually, redemption through basic compassion"
The beginning with the Big Bang—an immense explosion that created space, time, and matter. It delves into how black holes, regions of space where gravity is so strong that not even light can escape, play a crucial role in the evolution of galaxies and the universe as a whole. Stephen Hawking’s groundbreaking discoveries about black hole radiation, often referred to as Hawking Radiation, are also discussed in detail."
`);
});

Expand Down
7 changes: 2 additions & 5 deletions ...k/solutions/search/plugins/search_playground/server/utils/get_value_for_selected_field.ts
View file
Original file line number Diff line number Diff line change
Expand Up @@ -14,11 +14,8 @@ export const getValueForSelectedField = (hit: SearchHit, path: string): string =
}

// for semantic_text matches
const innerHitPath = `${hit._index}.${path}`;
if (!!hit.inner_hits?.[innerHitPath]) {
return hit.inner_hits[innerHitPath].hits.hits
.map((innerHit) => innerHit._source.text)
.join('\n --- \n');
if (hit.highlight && hit.highlight[path]) {
return hit.highlight[path].flat().join('\n --- \n');
}

return has(hit._source, `${path}.text`)
Expand Down

0 comments on commit 7ad184b

Please sign in to comment.