Highlighting

Highlighting enables you to obtain highlighted text fragments (referred to as snippets) from documents containing matching keywords.

The SQL HIGHLIGHT() function, the "highlight" property in JSON queries via HTTP, and the highlight() function in the PHP client all utilize the built-in document storage to retrieve the original field contents (enabled by default).

‹›
  • SQL
  • JSON
  • PHP
  • Python
  • Javascript
  • Java
  • C#
  • TypeScript
  • Go
📋
SELECT HIGHLIGHT() FROM books WHERE MATCH('try');
‹›
Response
+----------------------------------------------------------+
| highlight()                                              |
+----------------------------------------------------------+
| Don`t <strong>try</strong> to compete in childishness, said Bliss. |
+----------------------------------------------------------+
1 row in set (0.00 sec)

When using SQL for highlighting search results, you will receive snippets from various fields combined into a single string due to the limitations of the MySQL protocol. You can adjust the concatenation separators with the field_separator and snippet_separator options, as detailed below.

When executing JSON queries through HTTP or using the PHP client, there are no such constraints, and the result set includes an array of fields containing arrays of snippets (without separators).

Keep in mind that snippet generation options like limit, limit_words, and limit_snippets apply to each field individually by default. You can alter this behavior using the limits_per_field option, but it could lead to unwanted results. For example, one field may have matching keywords, but no snippets from that field are included in the result set because they didn't rank as high as snippets from other fields in the highlighting engine.

The highlighting algorithm currently prioritizes better snippets (with closer phrase matches) and then snippets with keywords not yet included in the result. Generally, it aims to highlight the best match for the query and to highlight all query keywords, as allowed by the limits. If no matches are found in the current field, the beginning of the document will be trimmed according to the limits and returned by default. To return an empty string instead, set the allow_empty option to 1.

Highlighting is performed during the so-called post limit stage, which means that snippet generation is deferred not only until the entire final result set is prepared but also after the LIMIT clause is applied. For instance, with a LIMIT 20,10 clause, the HIGHLIGHT() function will be called a maximum of 10 times.

Highlighting options

There are several optional highlighting options that can be used to fine-tune snippet generation, which are common to SQL, HTTP, and PHP clients.

before_match

A string to insert before a keyword match. The %SNIPPET_ID% macro can be used in this string. The first occurrence of the macro is replaced with an incrementing snippet number within the current snippet. Numbering starts at 1 by default but can be overridden with the start_snippet_id option. %SNIPPET_ID% restarts at the beginning of each new document. The default is <strong>.

after_match

A string to insert after a keyword match. The default is </strong>.

limit

The maximum snippet size, in symbols (codepoints). The default is 256. This is applied per-field by default, see limits_per_field.

limit_words

Limits the maximum number of words that can be included in the result. Note that this limit applies to all words, not just the matched keywords to highlight. For example, if highlighting Mary and a snippet Mary had a little lamb is selected, it contributes 5 words to this limit, not just 1. The default is 0 (no limit). This is applied per-field by default, see limits_per_field.

limit_snippets

Limits the maximum number of snippets that can be included in the result. The default is 0 (no limit). This is applied per-field by default, see limits_per_field.

limits_per_field

Determines whether limit, limit_words, and limit_snippets operate as individual limits in each field of the document being highlighted or as global limits for the entire document. Setting this option to 0 means that all combined highlighting results for one document must be within the specified limits. The downside is that you may have several snippets highlighted in one field and none in another if the highlighting engine decides they are more relevant. The default is 1 (use per-field limits).

around

The number of words to select around each matching keyword block. The default is 5.

use_boundaries

Determines whether to additionally break snippets by phrase boundary characters, as configured in table settings with the phrase_boundary directive. The default is 0 (don't use boundaries).

weight_order

Specifies whether to sort the extracted snippets in order of relevance (decreasing weight) or in order of appearance in the document (increasing position). The default is 0 (don't use weight order).

force_all_words

Ignores the length limit until the result includes all keywords. The default is 0 (don't force all keywords).

start_snippet_id

Sets the starting value of the %SNIPPET_ID% macro (which is detected and expanded in before_match, after_match strings). The default is 1.

html_strip_mode

Defines the HTML stripping mode setting. Defaults to index, meaning that table settings will be used. Other values include none and strip, which forcibly skip or apply stripping regardless of table settings; and retain, which retains HTML markup and protects it from highlighting. The retain mode can only be used when highlighting full documents and therefore requires that no snippet size limits are set. The allowed string values are none, strip, index, and retain.

allow_empty

Permits an empty string to be returned as the highlighting result when no snippets could be generated in the current field (no keyword match or no snippets fit the limit). By default, the beginning of the original text would be returned instead of an empty string. The default is 0 (don't allow an empty result).

snippet_boundary

Ensures that snippets do not cross a sentence, paragraph, or zone boundary (when used with a table that has the respective indexing settings enabled). The allowed values are sentence, paragraph, and zone.

emit_zones

Emits an HTML tag with the enclosing zone name before each snippet. The default is 0 (don't emit zone names).

force_snippets

Determines whether to force snippet generation even if limits allow highlighting the entire text. The default is 0 (don't force snippet generation).

‹›
  • SQL
  • JSON
  • PHP
  • Python
  • Javascript
  • Java
  • C#
  • TypeScript
  • Go
📋
SELECT HIGHLIGHT({limit=50}) FROM books WHERE MATCH('try|gets|down|said');
‹›
Response
+---------------------------------------------------------------------------+
| highlight({limit=50})                                                     |
+---------------------------------------------------------------------------+
|  ... , "It <strong>gets</strong> infantile pleasure  ...  to knock it <strong>down</strong>." |
| Don`t <strong>try</strong> to compete in childishness, <strong>said</strong> Bliss.           |
|  ...  a small room. Bander <strong>said</strong>, "Come, half-humans, I ...         |
+---------------------------------------------------------------------------+
3 rows in set (0.00 sec)

Highlighting via SQL

The HIGHLIGHT() function can be used to highlight search results. Here's the syntax:

HIGHLIGHT([options], [field_list], [query] )

By default, it works with no arguments.

‹›
  • SQL
SQL
📋
SELECT HIGHLIGHT() FROM books WHERE MATCH('before');
‹›
Response
+-----------------------------------------------------------+
| highlight()                                               |
+-----------------------------------------------------------+
| A door opened <strong>before</strong> them, revealing a small room. |
+-----------------------------------------------------------+
1 row in set (0.00 sec)

HIGHLIGHT() retrieves all available full-text fields from document storage and highlights them against the provided query. Field syntax in queries is supported. Field text is separated by field_separator, which can be modified in the options.

‹›
  • SQL
SQL
📋
SELECT HIGHLIGHT() FROM books WHERE MATCH('@title one');
‹›
Response
+-----------------+
| highlight()     |
+-----------------+
| Book <strong>one</strong> |
+-----------------+
1 row in set (0.00 sec)

Optional first argument in HIGHLIGHT() is the list of options.

‹›
  • SQL
SQL
📋
SELECT HIGHLIGHT({before_match='[match]',after_match='[/match]'}) FROM books WHERE MATCH('@title one');
‹›
Response
+------------------------------------------------------------+
| highlight({before_match='[match]',after_match='[/match]'}) |
+------------------------------------------------------------+
| Book [match]one[/match]                                    |
+------------------------------------------------------------+
1 row in set (0.00 sec)

The optional second argument is a string containing a single field or a comma-separated list of fields. If this argument is present, only the specified fields will be fetched from document storage and highlighted. An empty string as the second argument signifies "fetch all available fields."

‹›
  • SQL
SQL
📋
SELECT HIGHLIGHT({},'title,content') FROM books WHERE MATCH('one|robots');
‹›
Response
+---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| highlight({},'title,content')                                                                                                                                                         |
+---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| Book <strong>one</strong> | They followed Bander. The <strong>robots</strong> remained at a polite distance, but their presence was a constantly felt threat.                                             |
| Bander ushered all three into the room. <strong>One</strong> of the <strong>robots</strong> followed as well. Bander gestured the other <strong>robots</strong> away and entered itself. The door closed behind it. |
+---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
2 rows in set (0.00 sec)

Alternatively, you can use the second argument to specify a string attribute or field name without quotes. In this case, the supplied string will be highlighted against the provided query, but field syntax will be ignored.

‹›
  • SQL
SQL
📋
SELECT HIGHLIGHT({}, title) FROM books WHERE MATCH('one');
‹›
Response
+---------------------+
| highlight({},title) |
+---------------------+
| Book <strong>one</strong>     |
| Book five           |
+---------------------+
2 rows in set (0.00 sec)

The optional third argument is the query. This is used to highlight search results against a query different from the one used for searching.

‹›
  • SQL
SQL
📋
SELECT HIGHLIGHT({},'title', 'five') FROM books WHERE MATCH('one');
‹›
Response
+-------------------------------+
| highlight({},'title', 'five') |
+-------------------------------+
| Book one                      |
| Book <strong>five</strong>              |
+-------------------------------+
2 rows in set (0.00 sec)

Although HIGHLIGHT() is designed to work with stored full-text fields and string attributes, it can also be used to highlight arbitrary text. Keep in mind that if the query contains any field search operators (e.g., @title hello @body world), the field part of them is ignored in this case.

‹›
  • SQL
SQL
📋
SELECT HIGHLIGHT({},TO_STRING('some text to highlight'), 'highlight') FROM books WHERE MATCH('@title one');
‹›
Response
+----------------------------------------------------------------+
| highlight({},TO_STRING('some text to highlight'), 'highlight') |
+----------------------------------------------------------------+
| some text to <strong>highlight</strong>                                  |
+----------------------------------------------------------------+
1 row in set (0.00 sec)

Several options are relevant only when generating a single string as a result (not an array of snippets). This applies exclusively to the SQL HIGHLIGHT() function:

snippet_separator

A string to insert between snippets. The default is ....

field_separator

A string to insert between fields. The default is |.

Another way to highlight text is to use the CALL SNIPPETS statement. This mostly duplicates the HIGHLIGHT() functionality but cannot use built-in document storage. However, it can load source text from files.

Highlighting via HTTP

To highlight full-text search results in JSON queries via HTTP, field contents must be stored in document storage (enabled by default). In the example, full-text fields content and title are fetched from document storage and highlighted against the query specified in the query clause.

Highlighted snippets are returned in the highlight property of the hits array.

‹›
  • JSON
  • PHP
  • Python
  • Javascript
  • Java
  • C#
  • TypeScript
  • Go
📋
POST /search
{
  "index": "books",
  "query": { "match": { "*": "one|robots" } },
  "highlight":
  {
    "fields": ["content"]
  }
}
‹›
Response
{
  "took": 0,
  "timed_out": false,
  "hits": {
    "total": 1,
    "hits": [
      {
        "_id": "1",
        "_score": 2788,
        "_source": {
          "title": "Books one",
          "content": "They followed Bander. The robots remained at a polite distance, but their presence was a constantly felt threat. Bander ushered all three into the room. One of the robots followed as well. Bander gestured the other robots away and entered itself. The door closed behind it. "
        },
        "highlight": {
          "content": [
            "They followed Bander. The <strong>robots</strong> remained at a polite distance, ",
            " three into the room. <strong>One</strong> of the <strong>robots</strong> followed as well. Bander",
            " gestured the other <strong>robots</strong> away and entered itself. The"
          ]
        }
      }
    ]
  }
}

To highlight all possible fields, pass an empty object as the highlight property.

‹›
  • JSON
  • PHP
  • Python
  • Javascript
  • Java
  • C#
  • TypeScript
  • Go
📋
POST /search
{
  "index": "books",
  "query": { "match": { "*": "one|robots" } },
  "highlight": {}
}
‹›
Response
{
  "took": 0,
  "timed_out": false,
  "hits": {
    "total": 1,
    "hits": [
      {
        "_id": "1",
        "_score": 2788,
        "_source": {
          "title": "Books one",
          "content": "They followed Bander. The robots remained at a polite distance, but their presence was a constantly felt threat. Bander ushered all three into the room. One of the robots followed as well. Bander gestured the other robots away and entered itself. The door closed behind it. "
        },
        "highlight": {
          "title": [
            "Books <strong>one</strong>"
          ],
          "content": [
            "They followed Bander. The <strong>robots</strong> remained at a polite distance, ",
            " three into the room. <strong>One</strong> of the <strong>robots</strong> followed as well. Bander",
            " gestured the other <strong>robots</strong> away and entered itself. The"
          ]
        }
      }
    ]
  }
}

In addition to common highlighting options, several synonyms are available for JSON queries via HTTP:

fields

The fields object contains attribute names with options. It can also be an array of field names (without any options).

Note that by default, highlighting attempts to highlight the results following the full-text query. In a general case, when you don't specify fields to highlight, the highlight is based on your full-text query. However, if you specify fields to highlight, it highlights only if the full-text query matches the selected fields.

encoder

The encoder can be set to default or html. When set to html, it retains HTML markup when highlighting. This works similarly to the html_strip_mode=retain option.

highlight_query

The highlight_query option allows you to highlight against a query other than your search query. The syntax is the same as in the main query.

‹›
  • JSON
  • PHP
  • Python
  • Javascript
  • Java
  • C#
  • TypeScript
  • Go
📋
POST /search
{
  "index": "books",
  "query": { "match": { "content": "one|robots" } },
  "highlight":
  {
    "fields": [ "content"],
    "highlight_query": { "match": { "*":"polite distance" } }
   }
}
‹›
Response
{'aggregations': None,
 'hits': {'hits': [{u'_id': u'1',
                    u'_score': 1788,
                    u'_source': {u'content': u'They followed Bander. The robots remained at a polite distance, but their presence was a constantly felt threat. Bander ushered all three into the room. One of the robots followed as well. Bander gestured the other robots away and entered itself. The door closed behind it. ',
                                 u'title': u'Books one'},
                    u'highlight': {u'content': [u'. The robots remained at a <strong>polite distance</strong>, but their presence was a']}}],
          'max_score': None,
          'total': 1},
 'profile': None,
 'timed_out': False,
 'took': 0}

pre_tags and post_tags

pre_tags and post_tags set the opening and closing tags for highlighted text snippets. They function similarly to the before_match and after_match options. These are optional, with default values of <strong> and </strong>.

‹›
  • JSON
  • PHP
  • Python
  • Javascript
  • Java
  • C#
  • TypeScript
  • Go
📋
POST /search
{
  "index": "books",
  "query": { "match": { "*": "one|robots" } },
  "highlight":
  {
    "fields": [ "content", "title" ],
    "pre_tags": "before_",
    "post_tags": "_after"
   }
}
‹›
Response
Document: 1
title : Books one
content : They followed Bander. The robots remained at a polite distance, but their presence was a constantly felt threat. Bander ushered all three into the room. One of the robots followed as well. Bander gestured the other robots away and entered itself. The door closed behind it.
Highlight for content:
- They followed Bander. The before_robots_after remained at a polite distance,
-  three into the room. before_One_after of the before_robots_after followed as well. Bander
-  gestured the other before_robots_after away and entered itself. The
Highlight for title:
- Books before_one_after

no_match_size

no_match_size functions similarly to the allow_empty option. If set to 0, it acts as allow_empty=1, allowing an empty string to be returned as a highlighting result when a snippet could not be generated. Otherwise, the beginning of the field will be returned. This is optional, with a default value of 1.

‹›
  • JSON
  • PHP
  • Python
  • Javascript
  • Java
  • C#
  • TypeScript
  • Go
📋
POST /search
{
  "index": "books",
  "query": { "match": { "*": "one|robots" } },
  "highlight":
  {
    "fields": [ "content", "title" ],
    "no_match_size": 0
  }
}
‹›
Response
Document: 1
title : Books one
content : They followed Bander. The robots remained at a polite distance, but their presence was a constantly felt threat. Bander ushered all three into the room. One of the robots followed as well. Bander gestured the other robots away and entered itself. The door closed behind it.
Highlight for content:
- They followed Bander. The <strong>robots</strong> remained at a polite distance,
-  three into the room. <strong>One</strong> of the <strong>robots</strong> followed as well. Bander
-  gestured the other <strong>robots</strong> away and entered itself. The
Highlight for title:
- Books <strong>one</strong>

order

order sets the sorting order of extracted snippets. If set to "score", it sorts the extracted snippets in order of relevance. This is optional and works similarly to the weight_order option.

‹›
  • JSON
  • PHP
  • Python
  • Javascript
  • Java
  • C#
  • TypeScript
  • Go
📋
POST /search
{
  "index": "books",
  "query": { "match": { "*": "one|robots" } },
  "highlight":
  {
    "fields": [ "content", "title" ],
    "order": "score"
  }
}
‹›
Response
Document: 1
title : Books one
content : They followed Bander. The robots remained at a polite distance, but their presence was a constantly felt threat. Bander ushered all three into the room. One of the robots followed as well. Bander gestured the other robots away and entered itself. The door closed behind it.
Highlight for content:
-  three into the room. <strong>One</strong> of the <strong>robots</strong> followed as well. Bander
-  gestured the other <strong>robots</strong> away and entered itself. The
- They followed Bander. The <strong>robots</strong> remained at a polite distance,
Highlight for title:
- Books <strong>one</strong>

fragment_size

fragment_size sets the maximum snippet size in symbols. It can be global or per-field. Per-field options override global options. This is optional, with a default value of 256. It works similarly to the limit option.

‹›
  • JSON
  • PHP
  • Python
  • Javascript
  • Java
  • C#
  • TypeScript
  • Go
📋
POST /search
{
  "index": "books",
  "query": { "match": { "*": "one|robots" } },
  "highlight":
  {
    "fields": [ "content", "title" ],
    "fragment_size": 100
  }
}
‹›
Response
Document: 1
title : Books one
content : They followed Bander. The robots remained at a polite distance, but their presence was a constantly felt threat. Bander ushered all three into the room. One of the robots followed as well. Bander gestured the other robots away and entered itself. The door closed behind it.
Highlight for content:
-  the room. <strong>One</strong> of the <strong>robots</strong> followed as well
- Bander gestured the other <strong>robots</strong> away and entered
Highlight for title:
- Books <strong>one</strong>

number_of_fragments

number_of_fragments limits the maximum number of snippets in the result. Like fragment_size, it can be global or per-field. This is optional, with a default value of 0 (no limit). It works similarly to the limit_snippets option.

‹›
  • JSON
  • PHP
  • Python
  • Javascript
  • Java
  • C#
  • TypeScript
  • Go
📋
POST /search
{
  "index": "books",
  "query": { "match": { "*": "one|robots"  } },
  "highlight":
  {
    "fields": [ "content", "title" ],
    "number_of_fragments": 10
  }
}
‹›
Response
Document: 1
title : Books one
content : They followed Bander. The robots remained at a polite distance, but their presence was a constantly felt threat. Bander ushered all three into the room. One of the robots followed as well. Bander gestured the other robots away and entered itself. The door closed behind it.
Highlight for content:
- They followed Bander. The <strong>robots</strong> remained at a polite distance,
-  three into the room. <strong>One</strong> of the <strong>robots</strong> followed as well. Bander
-  gestured the other <strong>robots</strong> away and entered itself. The
Highlight for title:
- Books <strong>one</strong>

limit, limit_words, limit_snippets

Options like limit, limit_words, and limit_snippets can be set as global or per-field options. Global options are used as per-field limits unless per-field options override them. In the example, the title field is highlighted with default limit settings, while the content field uses a different limit.

‹›
  • JSON
  • PHP
  • Python
  • Javascript
  • Java
  • C#
  • TypeScript
  • Go
📋
POST /search
{
  "index": "books",
  "query": { "match": { "*": "one|robots"  } },
      "highlight":
      {
        "fields":
        {
            "title": {},
            "content" : { "limit": 50 }
        }
      }
}
‹›
Response
Document: 1
title : Books one
content : They followed Bander. The robots remained at a polite distance, but their presence was a constantly felt threat. Bander ushered all three into the room. One of the robots followed as well. Bander gestured the other robots away and entered itself. The door closed behind it.
Highlight for content:
-  into the room. <strong>One</strong> of the <strong>robots</strong> followed as well
Highlight for title:
- Books <strong>one</strong>

limits_per_field

Global limits can also be enforced by specifying limits_per_field=0. Setting this option means that all combined highlighting results must be within the specified limits. The downside is that you may get several snippets highlighted in one field and none in another if the highlighting engine decides that they are more relevant.

‹›
  • JSON
  • PHP
  • Python
  • Javascript
  • Java
  • C#
  • TypeScript
  • Go
📋
POST /search
{
  "index": "books",
  "query": { "match": { "content": "and first" } },
      "highlight":
      {
        "limits_per_field": false,
        "fields":
        {
            "content" : { "limit": 50 }
        }
      }
}
‹›
Response
Document: 1
title : Books one
content : They followed Bander. The robots remained at a polite distance, but their presence was a constantly felt threat. Bander ushered all three into the room. One of the robots followed as well. Bander gestured the other robots away and entered itself. The door closed behind it.
Highlight for content:
-  gestured the other robots away <strong>and</strong> entered itself. The door closed

CALL SNIPPETS

The CALL SNIPPETS statement builds a snippet from provided data and query using specified table settings. It can't access built-in document storage, which is why it's recommended to use the HIGHLIGHT() function instead.

The syntax is:

CALL SNIPPETS(data, table, query[, opt_value AS opt_name[, ...]])

data

data serves as the source from which a snippet is extracted. It can either be a single string or a list of strings enclosed in curly brackets.

table

table refers to the name of the table that provides the text processing settings for snippet generation.

query

query is the full-text query used to build the snippets.

opt_value and opt_name

opt_value and opt_name represent the snippet generation options.

‹›
  • SQL
SQL
📋
CALL SNIPPETS(('this is my document text','this is my another text'), 'forum', 'is text', 5 AS around, 200 AS limit);
‹›
Response
+----------------------------------------+
| snippet                                |
+----------------------------------------+
| this <strong>is</strong> my document <strong>text</strong> |
| this <strong>is</strong> my another <strong>text</strong>  |
+----------------------------------------+
2 rows in set (0.02 sec)

Most options are the same as in the HIGHLIGHT() function. There are, however, several options that can only be used with CALL SNIPPETS.

The following options can be used to highlight text stored in separate files:

load_files

This option, when enabled, treats the first argument as file names instead of data to extract snippets from. The specified files on the server side will be loaded for data. Up to max_threads_per_query worker threads per request will be used to parallelize the work when this flag is enabled. Default is 0 (no limit). To distribute snippet generation between remote agents, invoke snippets generation in a distributed table containing only one(!) local agent and several remotes. The snippets_file_prefix option is used to generate the final file name. For example, when searchd is configured with snippets_file_prefix = /var/data_ and text.txt is provided as a file name, snippets will be generated from the content of /var/data_text.txt.

load_files_scattered

This option only works with distributed snippets generation with remote agents. Source files for snippet generation can be distributed among different agents, and the main server will merge all non-erroneous results. For example, if one agent of the distributed table has file1.txt, another agent has file2.txt, and you use CALL SNIPPETS with both of these files, searchd will merge agent results, so you will get results from both file1.txt and file2.txt. Default is 0.

If the load_files option is also enabled, the request will return an error if any of the files is not available anywhere. Otherwise (if load_files is not enabled), it will return empty strings for all absent files. Searchd does not pass this flag to agents, so agents do not generate a critical error if the file does not exist. If you want to be sure that all source files are loaded, set both load_files_scattered and load_files to 1. If the absence of some source files on some agent is not critical, set only load_files_scattered to 1.

‹›
  • SQL
SQL
📋
CALL SNIPPETS(('data/doc1.txt','data/doc2.txt'), 'forum', 'is text', 1 AS load_files);
‹›
Response
+----------------------------------------+
| snippet                                |
+----------------------------------------+
| this <strong>is</strong> my document <strong>text</strong> |
| this <strong>is</strong> my another <strong>text</strong>  |
+----------------------------------------+
2 rows in set (0.02 sec)

Sorting and ranking

Query results can be sorted by full-text ranking weight, one or more attributes or expressions.

Full-text queries return matches sorted by default. If nothing is specified, they are sorted by relevance, which is equivalent to ORDER BY weight() DESC in SQL format.

Non-full-text queries do not perform any sorting by default.

Advanced sorting

Extended mode is automatically enabled when you explicitly provide sorting rules by adding the ORDER BY clause in SQL format or using the sort option via HTTP JSON.

Sorting via SQL

General syntax:

SELECT ... ORDER BY
{attribute_name | expr_alias | weight() | random() } [ASC | DESC],
...
{attribute_name | expr_alias | weight() | random() } [ASC | DESC]

In the sort clause, you can use any combination of up to 5 columns, each followed by asc or desc. Functions and expressions are not allowed as arguments for the sort clause, except for the weight() and random() functions (the latter can only be used via SQL in the form of ORDER BY random()). However, you can use any expression in the SELECT list and sort by its alias.

‹›
  • SQL
SQL
📋
select *, a + b alias from test order by alias desc;
‹›
Response
+------+------+------+----------+-------+
| id   | a    | b    | f        | alias |
+------+------+------+----------+-------+
|    1 |    2 |    3 | document |     5 |
+------+------+------+----------+-------+

Sorting via JSON

"sort" specifies an array where each element can be an attribute name or _score if you want to sort by match weights. In that case, the sort order defaults to ascending for attributes and descending for _score.

‹›
  • JSON
  • PHP
  • Python
  • javascript
  • Java
  • C#
  • typescript
  • go
📋
{
  "index":"test",
  "query":
  {
    "match": { "title": "Test document" }
  },
  "sort": [ "_score", "id" ],
  "_source": "title",
  "limit": 3
}
‹›
Response
    {
      "took": 0,
      "timed_out": false,
      "hits": {
        "total": 5,
        "total_relation": "eq",
        "hits": [
          {
            "_id": "5406864699109146628",
            "_score": 2319,
            "_source": {
              "title": "Test document 1"
            }
          },
          {
            "_id": "5406864699109146629",
            "_score": 2319,
            "_source": {
              "title": "Test document 2"
            }
          },
          {
            "_id": "5406864699109146630",
            "_score": 2319,
            "_source": {
              "title": "Test document 3"
            }
          }
        ]
      }
    }

You can also specify the sort order explicitly:

  • asc: sort in ascending order
  • desc: sort in descending order
‹›
  • JSON
  • PHP
  • Python
  • javascript
  • Java
  • C#
  • typescript
  • go
📋
{
  "index":"test",
  "query":
  {
    "match": { "title": "Test document" }
  },
  "sort":
  [
    { "id": "desc" },
    "_score"
  ],
  "_source": "title",
  "limit": 3
}
‹›
Response
    {
      "took": 0,
      "timed_out": false,
      "hits": {
        "total": 5,
        "total_relation": "eq",
        "hits": [
          {
            "_id": "5406864699109146632",
            "_score": 2319,
            "_source": {
              "title": "Test document 5"
            }
          },
          {
            "_id": "5406864699109146631",
            "_score": 2319,
            "_source": {
              "title": "Test document 4"
            }
          },
          {
            "_id": "5406864699109146630",
            "_score": 2319,
            "_source": {
              "title": "Test document 3"
            }
          }
        ]
      }
    }

You can also use another syntax and specify the sort order via the order property:

‹›
  • JSON
  • PHP
  • Python
  • javascript
  • Java
  • C#
  • typescript
  • go
📋
{
  "index":"test",
  "query":
  {
    "match": { "title": "Test document" }
  },
  "sort":
  [
    { "id": { "order":"desc" } }
  ],
  "_source": "title",
  "limit": 3
}
‹›
Response
    {
      "took": 0,
      "timed_out": false,
      "hits": {
        "total": 5,
        "total_relation": "eq",
        "hits": [
          {
            "_id": "5406864699109146632",
            "_score": 2319,
            "_source": {
              "title": "Test document 5"
            }
          },
          {
            "_id": "5406864699109146631",
            "_score": 2319,
            "_source": {
              "title": "Test document 4"
            }
          },
          {
            "_id": "5406864699109146630",
            "_score": 2319,
            "_source": {
              "title": "Test document 3"
            }
          }
        ]
      }
    }

Sorting by MVA attributes is also supported in JSON queries. Sorting mode can be set via the mode property. The following modes are supported:

  • min: sort by minimum value
  • max: sort by maximum value
‹›
  • JSON
  • PHP
  • Python
  • javascript
  • Java
  • C#
  • typescript
  • go
📋
{
  "index":"test",
  "query":
  {
    "match": { "title": "Test document" }
  },
  "sort":
  [
    { "attr_mva": { "order":"desc", "mode":"max" } }
  ],
  "_source": "title",
  "limit": 3
}
‹›
Response
    {
      "took": 0,
      "timed_out": false,
      "hits": {
        "total": 5,
        "total_relation": "eq",
        "hits": [
          {
            "_id": "5406864699109146631",
            "_score": 2319,
            "_source": {
              "title": "Test document 4"
            }
          },
          {
            "_id": "5406864699109146629",
            "_score": 2319,
            "_source": {
              "title": "Test document 2"
            }
          },
          {
            "_id": "5406864699109146628",
            "_score": 2319,
            "_source": {
              "title": "Test document 1"
            }
          }
        ]
      }
    }

When sorting on an attribute, match weight (score) calculation is disabled by default (no ranker is used). You can enable weight calculation by setting the track_scores property to true:

‹›
  • JSON
  • PHP
  • Python
  • javascript
  • Java
  • C#
  • typescript
  • go
📋
{
  "index":"test",
  "track_scores": true,
  "query":
  {
    "match": { "title": "Test document" }
  },
  "sort":
  [
    { "attr_mva": { "order":"desc", "mode":"max" } }
  ],
  "_source": "title",
  "limit": 3
}
‹›
Response
    {
      "took": 0,
      "timed_out": false,
      "hits": {
        "total": 5,
        "total_relation": "eq",
        "hits": [
          {
            "_id": "5406864699109146631",
            "_score": 2319,
            "_source": {
              "title": "Test document 4"
            }
          },
          {
            "_id": "5406864699109146629",
            "_score": 2319,
            "_source": {
              "title": "Test document 2"
            }
          },
          {
            "_id": "5406864699109146628",
            "_score": 2319,
            "_source": {
              "title": "Test document 1"
            }
          }
        ]
      }
    }

Ranking overview

Ranking (also known as weighting) of search results can be defined as a process of computing a so-called relevance (weight) for every given matched document regards to a given query that matched it. So relevance is, in the end, just a number attached to every document that estimates how relevant the document is to the query. Search results can then be sorted based on this number and/or some additional parameters, so that the most sought-after results would appear higher on the results page.

There is no single standard one-size-fits-all way to rank any document in any scenario. Moreover, there can never be such a way, because relevance is subjective. As in, what seems relevant to you might not seem relevant to me. Hence, in general cases, it's not just hard to compute; it's theoretically impossible.

So ranking in Manticore is configurable. It has a notion of a so-called ranker. A ranker can formally be defined as a function that takes a document and a query as its input and produces a relevance value as output. In layman's terms, a ranker controls exactly how (using which specific algorithm) Manticore will assign weights to the documents.

Available built-in rankers

Manticore ships with several built-in rankers suited for different purposes. Many of them use two factors: phrase proximity (also known as LCS) and BM25. Phrase proximity works on keyword positions, while BM25 works on keyword frequencies. Essentially, the better the degree of phrase match between the document body and the query, the higher the phrase proximity (it maxes out when the document contains the entire query as a verbatim quote). And BM25 is higher when the document contains more rare words. We'll save the detailed discussion for later.

The currently implemented rankers are:

  • proximity_bm25, the default ranking mode that uses and combines both phrase proximity and BM25 ranking.
  • bm25, a statistical ranking mode that uses BM25 ranking only (similar to most other full-text engines). This mode is faster but may result in worse quality for queries containing more than one keyword.
  • none, a no-ranking mode. This mode is obviously the fastest. A weight of 1 is assigned to all matches. This is sometimes called boolean searching, which just matches the documents but does not rank them.
  • wordcount, ranking by the keyword occurrences count. This ranker computes the per-field keyword occurrence counts, then multiplies them by field weights, and sums the resulting values.
  • proximity returns the raw phrase proximity value as a result. This mode is internally used to emulate SPH_MATCH_ALL queries.
  • matchany returns rank as it was computed in SPH_MATCH_ANY mode earlier and is internally used to emulate SPH_MATCH_ANY queries.
  • fieldmask returns a 32-bit mask with the N-th bit corresponding to the N-th full-text field, numbering from 0. The bit will only be set when the respective field has any keyword occurrences satisfying the query.
  • sph04 is generally based on the default 'proximity_bm25' ranker, but additionally boosts matches when they occur at the very beginning or the very end of a text field. Thus, if a field equals the exact query, sph04 should rank it higher than a field that contains the exact query but is not equal to it. (For instance, when the query is "Hyde Park", a document titled "Hyde Park" should be ranked higher than one titled "Hyde Park, London" or "The Hyde Park Cafe".)
  • expr allows you to specify the ranking formula at runtime. It exposes several internal text factors and lets you define how the final weight should be computed from those factors. You can find more details about its syntax and a reference of available factors in a subsection below.

The ranker name is case-insensitive. Example:

SELECT ... OPTION ranker=sph04;

Quick summary of the ranking factors

Name Level Type Summary
max_lcs query int maximum possible LCS value for the current query
bm25 document int quick estimate of BM25(1.2, 0)
bm25a(k1, b) document int precise BM25() value with configurable K1, B constants and syntax support
bm25f(k1, b, {field=weight, ...}) document int precise BM25F() value with extra configurable field weights
field_mask document int bit mask of matched fields
query_word_count document int number of unique inclusive keywords in a query
doc_word_count document int number of unique keywords matched in the document
lcs field int Longest Common Subsequence between query and document, in words
user_weight field int user field weight
hit_count field int total number of keyword occurrences
word_count field int number of unique matched keywords
tf_idf field float sum(tf*idf) over matched keywords == sum(idf) over occurrences
min_hit_pos field int first matched occurrence position, in words, 1-based
min_best_span_pos field int first maximum LCS span position, in words, 1-based
exact_hit field bool whether query == field
min_idf field float min(idf) over matched keywords
max_idf field float max(idf) over matched keywords
sum_idf field float sum(idf) over matched keywords
exact_order field bool whether all query keywords were a) matched and b) in query order
min_gaps field int minimum number of gaps between the matched keywords over the matching spans
lccs field int Longest Common Contiguous Subsequence between query and document, in words
wlccs field float Weighted Longest Common Contiguous Subsequence, sum(idf) over contiguous keyword spans
atc field float Aggregate Term Closeness, log(1+sum(idf1idf2pow(distance, -1.75)) over the best pairs of keywords

Document-level Ranking Factors

A document-level factor is a numeric value computed by the ranking engine for every matched document with regards to the current query. So it differs from a plain document attribute in that the attribute does not depend on the full text query, while factors might. These factors can be used anywhere in the ranking expression. Currently implemented document-level factors are:

  • bm25 (integer), a document-level BM25 estimate (computed without keyword occurrence filtering).
  • max_lcs (integer), a query-level maximum possible value that the sum(lcs*user_weight) expression can ever take. This can be useful for weight boost scaling. For instance, MATCHANY ranker formula uses this to guarantee that a full phrase match in any field ranks higher than any combination of partial matches in all fields.
  • field_mask (integer), a document-level 32-bit mask of matched fields.
  • query_word_count (integer), the number of unique keywords in a query, adjusted for the number of excluded keywords. For instance, both (one one one one) and (one !two) queries should assign a value of 1 to this factor, because there is just one unique non-excluded keyword.
  • doc_word_count (integer), the number of unique keywords matched in the entire document.

Field-level Ranking Factors

A field-level factor is a numeric value computed by the ranking engine for every matched in-document text field regards to the current query. As more than one field can be matched by a query, but the final weight needs to be a single integer value, these values need to be folded into a single one. To achieve that, field-level factors can only be used within a field aggregation function, they can not be used anywhere in the expression. For example, you cannot use (lcs+bm25) as your ranking expression, as lcs takes multiple values (one in every matched field). You should use (sum(lcs)+bm25) instead, that expression sums lcs over all matching fields, and then adds bm25 to that per-field sum. Currently implemented field-level factors are:

  • lcs (integer), the length of a maximum verbatim match between the document and the query, counted in words. LCS stands for Longest Common Subsequence (or Subset). Takes a minimum value of 1 when only stray keywords were matched in a field, and a maximum value of query keywords count when the entire query was matched in a field verbatim (in the exact query keywords order). For example, if the query is 'hello world' and the field contains these two words quoted from the query (that is, adjacent to each other, and exactly in the query order), lcs will be 2. For example, if the query is 'hello world program' and the field contains 'hello world', lcs will be 2. Note that any subset of the query keyword works, not just a subset of adjacent keywords. For example, if the query is 'hello world program' and the field contains 'hello (test program)', lcs will be 2 just as well, because both 'hello' and 'program' matched in the same respective positions as they were in the query. Finally, if the query is 'hello world program' and the field contains 'hello world program', lcs will be 3. (Hopefully that is unsurprising at this point.)

  • user_weight (integer), the user specified per-field weight (refer to OPTION field_weights in SQL). The weights default to 1 if not specified explicitly.

  • hit_count (integer), the number of keyword occurrences that matched in the field. Note that a single keyword may occur multiple times. For example, if 'hello' occurs 3 times in a field and 'world' occurs 5 times, hit_count will be 8.

  • word_count (integer), the number of unique keywords matched in the field. For example, if 'hello' and 'world' occur anywhere in a field, word_count will be 2, regardless of how many times both keywords occur.

  • tf_idf (float), the sum of TF/IDF over all the keywords matched in the field. IDF is the Inverse Document Frequency, a floating point value between 0 and 1 that describes how frequent the keyword is (basically, 0 for a keyword that occurs in every document indexed, and 1 for a unique keyword that occurs in just a single document). TF is the Term Frequency, the number of matched keyword occurrences in the field. As a side note, tf_idf is actually computed by summing IDF over all matched occurrences. That's by construction equivalent to summing TF*IDF over all matched keywords.

  • min_hit_pos (integer), the position of the first matched keyword occurrence, counted in words

    Therefore, this is a relatively low-level, "raw" factor that you'll likely want to adjust before using it for ranking. The specific adjustments depend heavily on your data and the resulting formula, but here are a few ideas to start with: (a) any min_gaps-based boosts could be simply ignored when word_count<2;

    (b) non-trivial min_gaps values (i.e., when word_count>=2) could be clamped with a certain "worst-case" constant, while trivial values (i.e., when min_gaps=0 and word_count<2) could be replaced by that constant;

    (c) a transfer function like 1/(1+min_gaps) could be applied (so that better, smaller min_gaps values would maximize it, and worse, larger min_gaps values would fall off slowly); and so on.

  • lccs (integer). Longest Common Contiguous Subsequence. The length of the longest subphrase common between the query and the document, computed in keywords.

    The LCCS factor is somewhat similar to LCS but more restrictive. While LCS can be greater than 1 even if no two query words are matched next to each other, LCCS will only be greater than 1 if there are exact, contiguous query subphrases in the document. For example, (one two three four five) query vs (one hundred three hundred five hundred) document would yield lcs=3, but lccs=1, because although the mutual dispositions of 3 keywords (one, three, five) match between the query and the document, no 2 matching positions are actually adjacent.

    Note that LCCS still doesn't differentiate between frequent and rare keywords; for that, see WLCCS.

  • wlccs (float). Weighted Longest Common Contiguous Subsequence. The sum of IDFs of the keywords of the longest subphrase common between the query and the document.

    WLCCS is calculated similarly to LCCS, but every "suitable" keyword occurrence increases it by the keyword IDF instead of just by 1 (as with LCS and LCCS). This allows ranking sequences of rarer and more important keywords higher than sequences of frequent keywords, even if the latter are longer. For example, a query (Zanzibar bed and breakfast) would yield lccs=1 for a (hotels of Zanzibar) document, but lccs=3 against (London bed and breakfast), even though "Zanzibar" is actually somewhat rarer than the entire "bed and breakfast" phrase. The WLCCS factor addresses this issue by using keyword frequencies.

  • atc (float). Aggregate Term Closeness. A proximity-based measure that increases when the document contains more groups of more closely located and more important (rare) query keywords.

    WARNING: you should use ATC with OPTION idf='plain,tfidf_unnormalized' (see below); otherwise, you may get unexpected results.

    ATC essentially operates as follows. For each keyword occurrence in the document, we compute the so-called term closeness. To do this, we examine all the other closest occurrences of all the query keywords (including the keyword itself) to the left and right of the subject occurrence, calculate a distance dampening coefficient as k = pow(distance, -1.75) for these occurrences, and sum the dampened IDFs. As a result, for every occurrence of each keyword, we obtain a "closeness" value that describes the "neighbors" of that occurrence. We then multiply these per-occurrence closenesses by their respective subject keyword IDF, sum them all, and finally compute a logarithm of that sum.

In other words, we process the best (closest) matched keyword pairs in the document, and compute pairwise "closenesses" as the product of their IDFs scaled by the distance coefficient:

pair_tc = idf(pair_word1) * idf(pair_word2) * pow(pair_distance, -1.75)

We then sum such closenesses, and compute the final, log-dampened ATC value:

atc = log(1+sum(pair_tc))

Note that this final dampening logarithm is precisely the reason you should use OPTION idf=plain because, without it, the expression inside the log() could be negative.

Having closer keyword occurrences contributes much more to ATC than having more frequent keywords. Indeed, when the keywords are right next to each other, distance=1 and k=1; when there's just one word in between them, distance=2 and k=0.297, with two words between, distance=3 and k=0.146, and so on. At the same time, IDF attenuates somewhat slower. For example, in a 1 million document collection, the IDF values for keywords that match in 10, 100, and 1000 documents would be respectively 0.833, 0.667, and 0.500. So a keyword pair with two rather rare keywords that occur in just 10 documents each but with 2 other words in between would yield pair_tc = 0.101 and thus barely outweigh a pair with a 100-doc and a 1000-doc keyword with 1 other word between them and pair_tc = 0.099. Moreover, a pair of two unique, 1-doc keywords with 3 words between them would get a pair_tc = 0.088 and lose to a pair of two 1000-doc keywords located right next to each other and yielding a pair_tc = 0.25. So, basically, while ATC does combine both keyword frequency and proximity, it still somewhat favors proximity.

Ranking factor aggregation functions

A field aggregation function is a single-argument function that accepts an expression with field-level factors, iterates over all matched fields, and computes the final results. The currently implemented field aggregation functions include:

  • sum, which adds the argument expression over all matched fields. For example sum(1) should return the number of matched fields.
  • top, which returns the highest value of the argument across all matched fields.

Formula expressions for all the built-in rankers

Most other rankers can actually be emulated using the expression-based ranker. You just need to provide an appropriate expression. While this emulation will likely be slower than using the built-in, compiled ranker, it may still be interesting if you want to fine-tune your ranking formula starting with one of the existing ones. Additionally, the formulas describe the ranker details in a clear, readable manner.

  • proximity_bm25 (default ranker) = sum(lcs*user_weight)*1000+bm25
  • bm25 = sum(user_weight)*1000+bm25
  • none = 1
  • wordcount = sum(hit_count*user_weight)
  • proximity = sum(lcs*user_weight)
  • matchany = sum((word_count+(lcs-1)*max_lcs)*user_weight)
  • fieldmask = field_mask
  • sph04 = sum((4*lcs+2*(min_hit_pos==1)+exact_hit)*user_weight)*1000+bm25

Configuration of IDF formula

The historically default IDF (Inverse Document Frequency) in Manticore is equivalent to OPTION idf='normalized,tfidf_normalized', and those normalizations may cause several undesired effects.

First, idf=normalized causes keyword penalization. For instance, if you search for the | something and the occurs in more than 50% of the documents, then documents with both keywords the and[something will get less weight than documents with just one keyword something. Using OPTION idf=plain avoids this.

Plain IDF varies in [0, log(N)] range, and keywords are never penalized; while the normalized IDF varies in [-log(N), log(N)] range, and too frequent keywords are penalized.

Second, idf=tfidf_normalized causes IDF drift over queries. Historically, we additionally divided IDF by query keyword count, so that the entire sum(tf*idf) over all keywords would still fit into [0,1] range. However, that means that queries word1 and word1 | nonmatchingword2 would assign different weights to the exactly same result set, because the IDFs for both word1 and nonmatchingword2 would be divided by 2. OPTION idf='tfidf_unnormalized' fixes that. Note that BM25, BM25A, BM25F() ranking factors will be scale accordingly once you disable this normalization.

IDF flags can be mixed; plain and normalized are mutually exclusive;tfidf_unnormalized and tfidf_normalized are mutually exclusive; and unspecified flags in such a mutually exclusive group take their defaults. That means that OPTION idf=plain is equivalent to a complete OPTION idf='plain,tfidf_normalized' specification.

Pagination of search results

Manticore Search returns the top 20 matched documents in the result set by default.

SQL

In SQL, you can navigate through the result set using the LIMIT clause.

LIMIT can accept either one number as the size of the returned set with a zero offset, or a pair of offset and size values.

HTTP JSON

When using HTTP JSON, the nodes offset and limit control the offset of the result set and the size of the returned set. Alternatively, you can use the pair size and from instead.

‹›
  • SQL
  • JSON
📋
SELECT  ... FROM ...  [LIMIT [offset,] row_count]
SELECT  ... FROM ...  [LIMIT row_count][ OFFSET offset]

Result set window

By default, Manticore Search uses a result set window of 1000 best-ranked documents that can be returned in the result set. If the result set is paginated beyond this value, the query will end in error.

This limitation can be adjusted with the query option max_matches.

Increasing the max_matches to very high values should only be done if it's necessary for the navigation to reach such points. A high max_matches value requires more memory and can increase the query response time. One way to work with deep result sets is to set max_matches as the sum of the offset and limit.

Lowering max_matches below 1000 has the benefit of reducing the memory used by the query. It can also reduce the query time, but in most cases, it might not be a noticeable gain.

‹›
  • SQL
  • JSON
📋
SELECT  ... FROM ...   OPTION max_matches=<value>