Document does not have any linebreak for Chinese language

[iDreamCXK]

MigraDoc does not add any linebreak for Chinese language,as shown in the figure.
However,the English paragraph can be dealt correctly.

After debugging，I have found something wrong with the method "internal override void VisitDocumentObjectCollection(DocumentObjectCollection elements)" in the project MigraDoc.DocumentObjectModel"

internal override void VisitDocumentObjectCollection(DocumentObjectCollection elements)
{
        List<int> textIndices = new List<int>();
        if (elements is ParagraphElements)
        {
               for (int idx = 0; idx < elements.Count; ++idx)
              {
                    if (elements[idx] is Text)
                        textIndices.Add(idx);
              }
        }

     int[] indices = (int[])textIndices.ToArray();
     if (indices != null)
     {
          int insertedObjects = 0;
          foreach (int idx in indices)
          {
              Text text = (Text)elements[idx + insertedObjects];
              string currentString = "";
              foreach (char ch in text.Content)
              {
                   // TODO Add support for other breaking spaces (en space, em space, &c.).
                   switch (ch)
                   {
                       case ' ':
                       case '\r':
                       case '\n':
                       case '\t':
                       if (currentString != "")
                       {
                           elements.InsertObject(idx + insertedObjects, new Text(currentString));
                           ++insertedObjects;
                           currentString = "";
                        }
                       elements.InsertObject(idx + insertedObjects, new Text(" "));
                       ++insertedObjects;
                       break;
		       ...

The method replaces every blank for a DocumentObject ,and then MigraDoc can deal with every English word correctly ! Howevery ,thers is no blank for Chinese paragraph , such as "圣诞节（Christmas）又称耶诞节、耶稣诞辰，译名为“基督弥撒”，是西方传统节日，起源于基督教，在每年公历12月25日。弥撒是教会的一种礼拜仪式。圣诞节是一个宗教节，因为把它当作耶稣的诞辰来庆祝，故名“耶诞节”。"

I add a patch for my source code , if the text contains a Chines word , the method will replace it with a DocumentObject ,as shown below.

         foreach (char ch in text.Content)
         {
		if ( ch >= 0x4e00 && ch <= 0x9fbb )
		{
			if (currentString != "")
			{
				elements.InsertObject(idx + insertedObjects, new Text(currentString));
				++insertedObjects;
				currentString = "";
			}
			elements.InsertObject(idx + insertedObjects, new Text(ch.ToString()));**
			++insertedObjects;**
			continue;
		}

                // TODO Add support for other breaking spaces (en space, em space, &c.).
                switch (ch)
                 {
                      case ' ':
                      case '\r':
                      case '\n':
                      case '\t':
                      .....

And the MigraDoc performs well

However , I think the way to solve the problem is Low-Performance and not thorough. Other languages ,like Japanese,will also meet the same problem.Does the author have considered this problem ? The document do not have any way to resolve the problem.

[TH-Soft]

Does the author have considered this problem?

The authors know very little about CJK and RTL, so they are not surprised about issues with non-Latin non-LTR languages. They even wrote that in the FAQ.

[iDreamCXK]

Thank you for your reminding , I hope the issues showed above will be helpful for authors , and resovle it further .

[TureeZhang]

Thanks, spend so much time to looking for the reason, you are right, finally got your answer.

Then I also find a Stackoverflow answer says:

Text will break at spaces, at hyphens, at soft hyphens, and at zero-width non-joiners.

If you do not care where the linebreak occurs, just insert a zero-width non-joiner between each pair of characters or at suitable locations (e.g. for URLs after each slash or dot).

Soft hyphens between syllables will look better for human-readable text.

---User: I liked the old Stack Overflow

So I copy the ZeroWidth NonJoiner from here : https://unicode-explorer.com/c/200C

then write these simple code :

private readonly char _zeroWidthNonJoiner = '‌'; // ATTENSION: a Zero Width NonJoiner in this var
public string ZeroWidthNonJoinerString(string chinese)
{
    if (string.IsNullOrEmpty(chinese))
        return chinese;

    StringBuilder @string = new StringBuilder();
    foreach (char item in chinese)
    {
        @string.Append(item);
        @string.Append(this._zeroWidthNonJoiner);
    }

    return @string.ToString();
}

use the method return string then give it to Migradoc then line breaks works right:

[iDreamCXK]

You are so Handsome~

[Erhushenshou]

why I use version 1.50.5147 and render the font to be "Arial Unicode MS" still can't print Chinese?

[phoebusryan]

@iDreamCXK Which Version of Migradoc and font etc do you use to successfully draw chinese characters? I use the font "Arial Unicode MS" on a windows 10 machine and only get squares.. :/

I use the nuget-package "PDFsharp-MigraDoc-wpf" version="1.50.5147"

[iDreamCXK]

You can you Microsoft YaHei.It's support of Chinese

[phoebusryan]

@iDreamCXK When I use "Microsoft YaHei" my code throws an Exception in "RenderDocument()":
"TrueType collection fonts are not yet supported by PDFsharp"

[iDreamCXK]

Maybe you can get help from here
https://stackoverflow.com/questions/52791080/migradoc-pdfsharp-throwing-exceptions-with-chinese-yahei-font

[ThomasHoevel]

Find a True Type Font (.ttf), not a True Type Collection (.ttc).

[phoebusryan]

@ThomasHoevel That's tricky on a windows 10 machine. I just found Microsoft Yahei as a ttf but can't use it on Windows 10 because I already have a TTC with the same name installed. I can't uninstall the preinstalled version because it's a system font.. bah.

So i need to rename this font (don't know how) or find a different font with chinese characters....

[iDreamCXK]

Maybe  you can use 宋体在 2023年8月23日，17:13，Thomas Kaegi ***@***.***> 写道： @ThomasHoevel That's tricky on a windows 10 machine. I just found Microsoft Yahei as a ttf but can't use it on Windows 10 because I already have a TTC with the same name installed. I can't uninstall the preinstalled version because it's a system font.. bah. So i need to rename this font (don't know how) or find a different font with chinese characters.... —Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you were mentioned.Message ID: ***@***.***>

[phoebusryan]

I used the font "Noto Sans CJK SC" and it worked... damn... thanks guys!