Highlighting Usageedit

Allows to highlight search results on one or more fields. The implementation uses either the lucene highlighter, fast-vector-highlighter or postings-highlighter.

See the Elasticsearch documentation on highlighting for more detail.

Fluent DSL exampleedit

s => s
.Query(q => q
    .Match(m => m
        .Field(f => f.Name.Suffix("standard"))
        .Query("Upton Sons Shield Rice Rowe Roberts")
    )
)
.Highlight(h => h
    .PreTags("<tag1>")
    .PostTags("</tag1>")
    .Fields(
        fs => fs
            .Field(p => p.Name.Suffix("standard"))
            .Type("plain")
            .ForceSource()
            .FragmentSize(150)
            .NumberOfFragments(3)
            .NoMatchSize(150),
        fs => fs
            .Field(p => p.LeadDeveloper.FirstName)
            .Type(HighlighterType.Fvh)
            .PreTags("<name>")
            .PostTags("</name>")
            .BoundaryMaxScan(50)
            .HighlightQuery(q => q
                .Match(m => m
                    .Field(p => p.LeadDeveloper.FirstName)
                    .Query("Kurt Edgardo Naomi Dariana Justice Felton")
                )
            ),
        fs => fs
            .Field(p => p.State.Suffix("offsets"))
            .Type(HighlighterType.Postings)
            .PreTags("<state>")
            .PostTags("</state>")
            .HighlightQuery(q => q
                .Terms(t => t
                    .Field(f => f.State.Suffix("offsets"))
                    .Terms(
                        StateOfBeing.Stable.ToString().ToLowerInvariant(),
                        StateOfBeing.BellyUp.ToString().ToLowerInvariant()
                    )
                )
            )
    )
)

Object Initializer syntax exampleedit

new SearchRequest<Project>
{
    Query = new MatchQuery
    {
        Query = "Upton Sons Shield Rice Rowe Roberts",
        Field = "name.standard"
    },
    Highlight = new Highlight
    {
        PreTags = new[] { "<tag1>" },
        PostTags = new[] { "</tag1>" },
        Fields = new Dictionary<Field, IHighlightField>
        {
            { "name.standard", new HighlightField
                {
                    Type = HighlighterType.Plain,
                    ForceSource = true,
                    FragmentSize = 150,
                    NumberOfFragments = 3,
                    NoMatchSize = 150
                }
            },
            { "leadDeveloper.firstName", new HighlightField
                {
                    Type = "fvh",
                    BoundaryMaxScan = 50,
                    PreTags = new[] { "<name>"},
                    PostTags = new[] { "</name>"},
                    HighlightQuery = new MatchQuery
                    {
                        Field = "leadDeveloper.firstName",
                        Query = "Kurt Edgardo Naomi Dariana Justice Felton"
                    }
                }
            },
            { "state.offsets", new HighlightField
                {
                    Type = HighlighterType.Postings,
                    PreTags = new[] { "<state>"},
                    PostTags = new[] { "</state>"},
                    HighlightQuery = new TermsQuery
                    {
                        Field = "state.offsets",
                        Terms = new [] { "stable", "bellyup" }
                    }
                }
            }
        }
    }
}

Example json output. 

{
  "query": {
    "match": {
      "name.standard": {
        "query": "Upton Sons Shield Rice Rowe Roberts"
      }
    }
  },
  "highlight": {
    "pre_tags": [
      "<tag1>"
    ],
    "post_tags": [
      "</tag1>"
    ],
    "fields": {
      "name.standard": {
        "type": "plain",
        "force_source": true,
        "fragment_size": 150,
        "number_of_fragments": 3,
        "no_match_size": 150
      },
      "leadDeveloper.firstName": {
        "type": "fvh",
        "boundary_max_scan": 50,
        "pre_tags": [
          "<name>"
        ],
        "post_tags": [
          "</name>"
        ],
        "highlight_query": {
          "match": {
            "leadDeveloper.firstName": {
              "query": "Kurt Edgardo Naomi Dariana Justice Felton"
            }
          }
        }
      },
      "state.offsets": {
        "type": "postings",
        "pre_tags": [
          "<state>"
        ],
        "post_tags": [
          "</state>"
        ],
        "highlight_query": {
          "terms": {
            "state.offsets": [
              "stable",
              "bellyup"
            ]
          }
        }
      }
    }
  }
}

Handling Responsesedit

response.ShouldBeValid();

foreach (var highlightsInEachHit in response.Hits.Select(d=>d.Highlights))
{
    foreach (var highlightField in highlightsInEachHit)
    {
        if (highlightField.Key == "name.standard")
        {
            foreach (var highlight in highlightField.Value.Highlights)
            {
                highlight.Should().Contain("<tag1>");
                highlight.Should().Contain("</tag1>");
            }
        }
        else if (highlightField.Key == "leadDeveloper.firstName")
        {
            foreach (var highlight in highlightField.Value.Highlights)
            {
                highlight.Should().Contain("<name>");
                highlight.Should().Contain("</name>");
            }
        }
        else if (highlightField.Key == "state.offsets")
        {
            foreach (var highlight in highlightField.Value.Highlights)
            {
                highlight.Should().Contain("<state>");
                highlight.Should().Contain("</state>");
            }
        }
        else
        {
            Assert.True(false, $"highlights contains unexpected key {highlightField.Key}");
        }
    }
}

Unified highlighteredit

The unified highlighter can extract offsets from either postings, term vectors, or via re-analyzing text. Under the hood it uses Lucene UnifiedHighlighter which picks its strategy depending on the field and the query to highlight.

Note

Unified highlighter is available only in Elasticsearch 5.3.0+

Warning

This functionality is experimental and may be changed or removed completely in a future release. Elastic will take a best effort approach to fix any issues, but experimental features are not subject to the support SLA of official GA features.

Fluent DSL exampleedit

s => s
.Query(q => q
    .Match(m => m
        .Field(f => f.Name.Suffix("standard"))
        .Query("Upton Sons Shield Rice Rowe Roberts")
    )
)
.Highlight(h => h
    .Fields(
        fs => fs
            .Field(p => p.LeadDeveloper.LastName)
            .Type(HighlighterType.Unified)
            .PreTags("<name>")
            .PostTags("</name>")
            .HighlightQuery(q => q
                .Match(m => m
                    .Field(p => p.LeadDeveloper.LastName)
                    .Query(LastNameSearch)
                )
            )
    )
)

Object Initializer syntax exampleedit

new SearchRequest<Project>
{
    Query = new MatchQuery
    {
        Query = "Upton Sons Shield Rice Rowe Roberts",
        Field = "name.standard"
    },
    Highlight = new Highlight
    {
        Fields = new Dictionary<Field, IHighlightField>
        {
            { "leadDeveloper.lastName", new HighlightField
                {
                    Type = HighlighterType.Unified,
                    PreTags = new[] { "<name>"},
                    PostTags = new[] { "</name>"},
                    HighlightQuery = new MatchQuery
                    {
                        Field = "leadDeveloper.lastName",
                        Query = LastNameSearch
                    }
                }
            }
        }
    }
}

Example json output. 

{
  "query": {
    "match": {
      "name.standard": {
        "query": "Upton Sons Shield Rice Rowe Roberts"
      }
    }
  },
  "highlight": {
    "fields": {
      "leadDeveloper.lastName": {
        "type": "unified",
        "pre_tags": [
          "<name>"
        ],
        "post_tags": [
          "</name>"
        ],
        "highlight_query": {
          "match": {
            "leadDeveloper.lastName": {
              "query": "Stokes"
            }
          }
        }
      }
    }
  }
}

Handling Responsesedit

response.ShouldBeValid();

foreach (var highlightsInEachHit in response.Hits.Select(d => d.Highlights))
{
    foreach (var highlightField in highlightsInEachHit)
    {
        if (highlightField.Key == "leadDeveloper.lastName")
        {
            foreach (var highlight in highlightField.Value.Highlights)
            {
                highlight.Should().Contain("<name>");
                highlight.Should().Contain("</name>");
            }
        }
        else
        {
            Assert.True(false, $"highlights contains unexpected key {highlightField.Key}");
        }
    }
}