NEST Elasticsearch查询中file.filename返回null，需返回指定字段

Hey there! Let's work through your issues step by step—since you're new to Elasticsearch and NEST, these are totally common hurdles, so no need to stress.

1. Why file.filename is returning null #

The most likely culprit here is a mismatch between your C# class definition and how the data is structured in Elasticsearch. NEST relies on proper object mapping to deserialize nested fields correctly. Here's what you need to check:

Fix your document class structure #

If your Elasticsearch document has a nested file object with a filename field, your C# class needs to mirror that nested structure. For example:

// Your main document class
public class SearchDocument
{
    public string Content { get; set; }
    
    // Define a nested class to represent the `file` object in ES
    public FileMetadata File { get; set; }
}

// Nested class for the file details
public class FileMetadata
{
    // Match the field name exactly as it exists in Elasticsearch
    // If your ES field uses snake_case (like `file_name`), add the PropertyName attribute:
    // [PropertyName("file_name")]
    public string Filename { get; set; }
}

Ensure you're including the nested field in your query's source #

Even if the mapping is correct, you might be accidentally excluding the file.filename field from the response. Explicitly specify which fields to return using the Source method:

var response = client.Search<SearchDocument>(s => s
    .Query(q => /* Your query here */)
    .Source(src => src
        .Include(f => f.Content, f => f.File.Filename)
    )
);

Quick sanity check #

First, run a raw DSL query in Kibana or Elasticsearch's Dev Tools to confirm the file.filename field actually has values in your documents. If the DSL response shows the field is populated, the issue is definitely in your C# mapping or query setup.

2. Why use a term query? #

Great question—this depends entirely on what you're trying to search for and how your content field is mapped in Elasticsearch:

• Term queries perform exact matches. They work best on keyword type fields (or the .keyword sub-field of a text field) where you want to match an exact value (e.g., a specific tag or code).
• Match queries are for full-text search on text type fields. Elasticsearch will tokenize your search query and match against the tokenized terms in the field, which is what you want if you're searching for words or phrases in a longer content field.

The GitHub example you referenced probably used a term query because it was targeting a keyword field. If your content field is a text type (for full-text search), you should swap to a match query instead:

// Full-text search on content field
.Query(q => q
    .Match(m => m
        .Field(f => f.Content)
        .Query("your search phrase here")
    )
)

Full Working Example #

Putting it all together, here's a complete query that retrieves both content and file.filename:

var client = new ElasticClient(settings); // Your existing client setup

var searchResponse = client.Search<SearchDocument>(s => s
    .Index("your-index-name") // Specify your index
    .Query(q => q
        .Match(m => m
            .Field(f => f.Content)
            .Query("example search text")
        )
    )
    .Source(src => src
        .Include(f => f.Content, f => f.File.Filename)
    )
);

// Access the results
if (searchResponse.IsValid)
{
    foreach (var hit in searchResponse.Hits)
    {
        Console.WriteLine($"Content snippet: {hit.Source.Content}");
        Console.WriteLine($"Filename: {hit.Source.File?.Filename ?? "No filename found"}");
    }
}
else
{
    Console.WriteLine($"Query failed: {searchResponse.ServerError.Error.Reason}");
}

内容的提问来源于stack exchange，提问作者abdul qayyum