Page Comparison

During indexing the files are stripped and only the text content is extracted. Some files contain a lot of metadata or embedded resources (e.g. photos) so a 10MB 20MB .pdf may only contain 1MB 2MB of actual data.

Once you implement the code in Visual Studio , build your solution and you will also have to reindex the index you are using from Administrator → Search Indexes → Action → Reindex

Expected results

Now if you Inspect your frontend page you should be able to see the title of your large document, but not the content. There will be no content field as well in the XHR search → results → document fields

using System;
using System.Collections.Generic;
using System.Linq;
using Hawksearch.Search;
using Telerik.Sitefinity.Services.Search.Data;
using Telerik.Sitefinity.Configuration;
using Telerik.Sitefinity.Services.Search.Model;
using Hawksearch.Configuration;
using Hawksearch.SDK.Indexing;
using Field = Telerik.Sitefinity.Services.Search.Publishing.Field;

namespace SitefinityWebApp.Search
{
    public class CustomSearchService : HawksearchService
    {
        private const string DocumentContentType = "Telerik.Sitefinity.Libraries.Model.Document";

        protected override List<SubmitDocument> AdaptDocuments(IEnumerable<IDocument> documents)
        {
            var doc = documents.ToList().FirstOrDefault();
            var documentList = new List<IDocument>(documents);

            if (doc != null)
            {
                var contentTypeField = doc.Fields.FirstOrDefault(f => f.Name == "ContentType");

                if (contentTypeField != null)
                {
                    if (string.Equals(contentTypeField.Value.ToString(), DocumentContentType, StringComparison.InvariantCultureIgnoreCase))
                    {
                        var configManager = ConfigManager.GetManager();
                        var hawkConfig = configManager.GetSection<HawkSearchConfig>();
                        documentList.Clear();

                        foreach (var document in documents)
                        {
                            var modifiedDocument = document;
                            var documentSize = this.CalculateDocumentSize(document);

                            if (documentSize > hawkConfig.DocumentSizeLimit)
                            {
                                modifiedDocument = this.ModifyDocument(document);
                            }

                            documentList.Add(modifiedDocument);
                        }
                    }
                }
            }

            return base.AdaptDocuments(documentList);
        }

        private double CalculateDocumentSize(IDocument document)
        {
            var documentSize = 0.0;

            foreach (var field in document.Fields)
            {
                if (field.Value != null)
                {
                    documentSize += System.Text.Encoding.Unicode.GetByteCount(field.Value.ToString()) / 1024.0;
                }
            }

            return documentSize;
        }

        private IDocument ModifyDocument(IDocument document)
        {
            var wordLimit = 500;
            var fields = new List<IField>(document.Fields);
            var contentField = document.Fields.FirstOrDefault(f => f.Name == "Content");
            fields.Remove(contentField);
            contentField = this.ExtractFieldContent(contentField, wordLimit);
            fields.Add(contentField);

            var modifiedDocument = new Document(fields, document.IdentityField.Name);

            return modifiedDocument;
        }

        private IField ExtractFieldContent(IField contentField, int wordLimit)
        {
            var fieldValue = contentField.Value.ToString();

            if (!string.IsNullOrWhiteSpace(fieldValue))
            {
                var modifiedContent = string.Join(" ", fieldValue.Split(' ').Take(wordLimit).ToArray());

                contentField = new Field
                {
                    Name = "Content",
                    Value = modifiedContent
                };
            }

            return contentField;
        }
    }
}

Once you implement the code in Visual Studio , build your solution and you will also have to reindex the index you are using from Administrator → Search Indexes → Action → Reindex

Expected results

Now if you Inspect your frontend page you should be able to see the content with first 500 symbols of your large document. There will be also a content field with 500 symbols in the XHR search → results → document fields