Why SpellCheck always marks words from additional dictionary (utf-8, utf-8 with BOM, UTF-16) as bad?

Question

I have this control made by Hans Passant, that I grabbed from here (Trying to use the C# SpellCheck class) and modified to load additional dictionary:

using System;
using System.Windows;
using System.ComponentModel;
using System.Windows.Controls;
using System.Windows.Forms.Integration;

namespace SoAL_TextExtractor
{
    class SpellBox : ElementHost
    {
        public SpellBox()
        {
            box = new TextBox();
            base.Child = box;
            box.TextChanged += (s, e) => OnTextChanged(EventArgs.Empty);
            Uri lex_file = new Uri(System.Windows.Forms.Application.StartupPath + "\\Russian.lex");
            box.SpellCheck.CustomDictionaries.Add(lex_file);
            box.SpellCheck.IsEnabled = true;
            box.VerticalScrollBarVisibility = ScrollBarVisibility.Auto;
            this.Size = new System.Drawing.Size(100, 20);
        }
        public override string Text
        {
            get { return box.Text; }
            set { box.Text = value; }
        }
        [DefaultValue(false)]
        public bool Multiline
        {
            get { return box.AcceptsReturn; }
            set { box.AcceptsReturn = value; }
        }
        [DefaultValue(false)]
        public bool WordWrap
        {
            get { return box.TextWrapping != TextWrapping.NoWrap; }
            set { box.TextWrapping = value ? TextWrapping.Wrap : TextWrapping.NoWrap; }
        }
        [DesignerSerializationVisibility(DesignerSerializationVisibility.Hidden)]
        public new System.Windows.UIElement Child
        {
            get { return base.Child; }
            set { /* Do nothing to solve a problem with the serializer !! */ }
        }
        private TextBox box;
    }
}

This dictionary:

enter image description here

And, this result:

enter image description here

I tried different encodings for *.lex but words still marked as bad.

lex has the LID as needed:

#LID 1049
абажур
абажура
...

What's up?

You can't just copy somebody else's code and repost it as though it is your own. Attribution is required, look at the bottom of the page for details. — Hans Passant, Sep 04 '14 at 09:52
@HansPassant - excuse me? Where did I said it's my code? All I said is "I have this control". I know this is grabbed from another answer, but I never told this is mine. I using it, I have problems with it and I showing here what I have. — Kosmo零, Sep 04 '14 at 13:42
The license doesn't say anything about that. It is an exceedingly simple rule, when you repost somebody else's work then you **must** attribute. Always. Please take care of it. — Hans Passant, Sep 04 '14 at 14:11
Oh huh, so you are the author. I added credits to my questions. Also, question solved by removing LID from lex. — Kosmo零, Sep 04 '14 at 15:06
@Kosmos I tested removing the LID and that didn't solve the problem. It resulted in none of the words being underlined as misspelled. — Loathing, Sep 05 '14 at 18:38
@Loathing - Well, this is the differences that we probably have: 1 - **I do not set: `box.Language = /*to anything*/;`.** 2 - **My lex is saved in UTF-16, while you said it's in 1252 code page**. On msdn said that: **If the locale is not specified, the dictionary applies to all languages.**... So, i don't know how, but the SpellCheck is work for me as expected. — Kosmo零, Sep 06 '14 at 06:27
I confirm UTF-16 with no language setting and no LID works. Surprising how giving more information causes it to not work as one would expect. Thanks. — Loathing, Sep 06 '14 at 06:42

score 2 · Accepted Answer · edited May 23 '17 at 11:57

I was able to get French working by doing the following:

saving file using 1252 encoding
setting the language on the text box
setting the #LID to 1036 in the lex file

Testing using a file containing: ééé

e.g.

box.Language = System.Windows.Markup.XmlLanguage.GetLanguage("fr");
String text = System.IO.File.ReadAllText(lex_file.AbsolutePath, Encoding.UTF8);
System.IO.File.WriteAllText(lex_file.AbsolutePath, text, Encoding.GetEncoding(1252));

Unfortunately, this doesn't work for Russian. It appears only 4 languages are supported by default. See Does SpellCheck .Net class support russian language? for more information.

Why SpellCheck always marks words from additional dictionary (utf-8, utf-8 with BOM, UTF-16) as bad?

1 Answers1