Selection

When documents are generated from scratch on the server, without using templates or &#9584; TX Text Control .NET Server &#9584; DocumentServer Namespace &#9584; MailMerge Class The MailMerge class is a .NET component that can be used to effortlessly merge template documents with database content in .NET projects, such as ASP.NET web applications, web services or Windows services. " data-tippy-interactive="true" data-tippy-placement="bottom" data-tooltip-width="wide"

>Mail Merge , small implementation details can have a measurable impact. In this type of workflow, the application creates every paragraph, heading, bullet list, table cell, and formatted text fragment programmatically. If each fragment is inserted first, selected again, and then formatted through the live selection, the document engine has to perform additional selection work for every operation.

This benchmark compares two ways to create the same generated document with &#9584; TX Text Control .NET Server &#9584; TXTextControl Namespace &#9584; ServerTextControl Class The ServerTextControl class implements a component that provide high-level text processing features for server-based applications. " data-tippy-interactive="true" data-tippy-placement="bottom" data-tooltip-width="wide"

>Server Text Control . The first pattern creates a &#9584; TX Text Control .NET Server &#9584; TXTextControl Namespace &#9584; Selection Class The Selection class describes and handles the attributes of a text selection. " data-tippy-interactive="true" data-tippy-placement="bottom" data-tooltip-width="wide"

>Selection object, sets text and formatting on that object, and assigns it once. The comparison pattern inserts text first, selects the inserted range, and then changes formatting through the live &#9584; TX Text Control .NET Server &#9584; TXTextControl Namespace &#9584; Selection Class The Selection class describes and handles the attributes of a text selection. " data-tippy-interactive="true" data-tippy-placement="bottom" data-tooltip-width="wide"

>Selection .

Selection Object Performance

In this benchmark, prebuilding TXTextControl.Selection objects is consistently about factor 2 faster than inserting text, selecting it, and formatting the live selection afterwards.

The Faster Pattern: Build the Selection First

The recommended pattern prepares the complete selection object before it is assigned to the current document position. Text and formatting are applied to the temporary object, and the assignment to Selection inserts the prepared content in one step.

AppendPrebuilt(control, new Selection { Text = paragraph + "\r\n", FontName = "Calibri", FontSize = Twips(10) });

static void AppendPrebuilt(ServerTextControl control, Selection selection) { control.Selection = selection; } The same approach is used for headings, normal paragraphs, and bullet text. Formatting such as font name, font size, bold, and italic is applied directly to the new Selection instance before it is inserted.

The Slower Pattern: Insert, Select, and Format

The comparison path performs more document-selection work. It inserts the text, selects the inserted range in the live document, applies formatting to the active selection, and then moves the selection back to the end of the inserted text.

int start = control.Selection.Start + control.Selection.Length;

control.Selection.Text = text; control.Select(start, text.Length);

control.Selection.FontName = "Calibri"; control.Selection.FontSize = Twips(10);

control.Select(start + text.Length, 0); This pattern is easy to write because it resembles interactive editing. But in generated documents it repeats selection and formatting operations for every fragment. The more paragraphs, sections, and list items the document contains, the more visible this overhead becomes.

Benchmark Setup

The benchmark is a .NET console application using TXTextControl.TextControl.Core.SDK and ServerTextControl . Both test paths create the same logical document content. The generated document contains a title, subtitle, sections, section headings, paragraphs, and bullet lists.

Each profile is warmed up before timing. The measured result is reported as average milliseconds per generated document. PDF files are written after the benchmark run so that disk I/O and PDF export do not affect the timing numbers.

Two dimensions were tested separately:

A static number of sections with an increasing number of paragraphs per section. A linear increase in the number of sections with a static number of paragraphs and bullets per section.

Results: Static Sections, Variable Paragraphs

The first test keeps the document structure stable with 10 sections and 4 bullets per section. Only the number of paragraphs per section changes from 10 to 100.

Sections Paragraphs/Section Bullets/Section Elements Prebuilt ms/doc Select ms/doc Speedup 10 10 4 152 5.7 10.6 1.86x 10 20 4 252 9.2 16.5 1.80x 10 30 4 352 10.6 21.6 2.04x 10 40 4 452 14.2 29.3 2.06x 10 50 4 552 18.2 37.2 2.04x 10 60 4 652 22.2 46.7 2.10x 10 70 4 752 27.2 56.7 2.09x 10 80 4 852 34.7 68.9 1.99x 10 90 4 952 37.3 85.3 2.29x 10 100 4 1052 42.9 88.9 2.08x

The paragraph test shows a clear scaling trend. As the number of generated text fragments grows, both approaches take longer, but the live-selection approach grows faster. Across the test range, the prebuilt selection pattern stays close to twice as fast.

Results: Variable Sections

The second test keeps the content inside each section stable with 10 paragraphs and 4 bullets per section. The section count increases linearly from 2 to 20.

Sections Paragraphs/Section Bullets/Section Elements Prebuilt ms/doc Select ms/doc Speedup 2 10 4 32 1.4 3.0 2.18x 4 10 4 62 2.3 4.6 2.02x 6 10 4 92 3.2 6.4 2.03x 8 10 4 122 4.4 8.1 1.85x 10 10 4 152 5.2 9.1 1.75x 12 10 4 182 5.3 10.4 1.98x 14 10 4 212 6.1 12.3 2.02x 16 10 4 242 7.2 14.1 1.95x 18 10 4 272 8.0 16.2 2.02x 20 10 4 302 9.0 18.3 2.04x

This section-based test isolates document structure growth. The number of sections increases in equal steps, while the amount of content per section remains unchanged. Again, the result is consistent: prebuilt Selection objects avoid repeated live-selection operations and remain approximately twice as fast.

What the Results Show

The important point is not a single timing value. Timings vary by machine, runtime, and document complexity. The important result is the shape of the data. Both benchmarks show that prebuilding the Selection object scales more efficiently as the number of generated content elements increases.

For applications that generate contracts, reports, invoices, statements, or other business documents, this pattern is especially useful when many formatted text fragments are inserted programmatically. Instead of thinking like an editor that inserts text and then selects it, think like a document generator: prepare the content fragment first and insert it once.

Conclusion

When generating documents with TX Text Control .NET Server, creating and formatting a Selection object before assigning it to control.Selection is the more efficient pattern. In this benchmark, it reduced document generation time by approximately factor 2 compared to inserting text first, selecting the inserted range, and formatting the live selection.

The code change is small, but the effect is visible as documents grow. For high-volume document generation or templates with many formatted fragments, using prebuilt Selection objects is a practical optimization that keeps the code simple and improves throughput.

TX Text Control .NET Server

DS Server

Server & Web

Windows Forms

WPF

ActiveX

Core Technologies

Blog

Core Technologies

Blog

Authors

Latest Posts

Support Resources

Documentation

Support

Getting Started

Other Resources

About Us

Industry events

Company

Contact

Newsletter

Legal

Sign in

Create account

Text Control Account

Selection

Post tagged with Selection in 2026

Server & Web

Windows Forms

WPF

ActiveX

Core Technologies

​

Latest Posts

​

Support

Getting Started

​

Other Resources

Company

Contact

Newsletter

Legal

Text Control Account

​

​

Selection

Post tagged with Selection in 2026 ​

July 2026

Post tagged with Selection in 2026