Issue with font family on chinese characters

#27039

Hi

We were using PD4ML version 361, now we have upgraded to latest one

Now we have a problem with the font

for example,
for the paragraphs which contain both the english and chinese characters are rendered with the Mingliu font we are using for chinese

the below is our piece of code.

WENJUN Money 育暮趣所及定平関島指形由幹設 Laundering in UNITED KINGDOM IS NOT functional now

style used is,

.match-text{font-size: 11pt;font-family: Times New Roman,MingLiu;}

In the old version of PD4ML, we were getting the english in times new roman and only the chinese characters in the Mingliu font. But in the latest pd4ml version, we are seeing everything on the Mingliu font version and not getting the correct style we expected.

Please help us to resolve this issue

Regards
Sarin

#29814

Do you have a font mapping for Times New Roman in pd4fonts.properties? Does it load the .ttf file during a conversion?

Times New Roman=times.ttf

#29815

Yes, we have defined this in pd4fonts.properties

Arial=arial.ttf
Arial Bold=arialbd.ttf
Arial Bold Italic=arialbi.ttf
Arial Italic=ariali.ttf
Courier New=cour.ttf
Courier New Bold=courbd.ttf
Courier New Bold Italic=courbi.ttf
Courier New Italic=couri.ttf
Times New Roman=times.ttf
Times New Roman Bold=timesbd.ttf
Times New Roman Bold Italic=timesbi.ttf
Times New Roman Italic=timesi.ttf
Calibri Bold=calibribd.ttf
Calibri=calibri.ttf
Calibri Italic=Calibrii.ttf
MingLiU=mingliu.ttc_0

The issue happens only when the english and chinese characters inside one span or tag, If the chinese in in another span, it is working properly. But in our case, it will come together inside one span, please help us to resolve this. Otherwise it would a big blow for us as the existing functionality will be broken with the upgrade

Regards
Sarin

#29816

We plan to address the issue on Monday. I’ll let you know asap.

#29817

Ok, thanks, we will wait for your reply

#29818

Hello PD4ML support, please can you respond on this issue as this is blocking our timelines and resources.

#29819

Sorry it takes that long. We’ve been still working on the issue. I’ll let you know.

#29820

In current DEV build we implemented a solution for the issue. It is a very sensitive part of the rendering engine and the patched version still cannot pass QA by “exotic” test cases (a mix of Chinese and Arabic scripts), so we made the patch conditionally switchable on/off. Currently by default it is switched off.

In order to enable it you should either run your Java (app server?) with the following JVM parameter:

-Dpd4ml.fix.20150816=true

Alternatively you may trigger PD4ML API to achieve the same effect:

HashMap map = new HashMap();
map.put( PD4Constants.PD4ML_FIX20150816, “true” );
pd4ml.setDynamicParams(map);

Hopefully the workaround is ok for you.

We currently work on a totally refactored PD4ML v4.x, which should implicitly solve such kind of issues.

We’ll send you an updated library by request to support pd4ml com

#29821

Hi,

thank you for your solution. Would like to know,

Whether we have any workaround without upgrading the library, we are using the v397 licenced version and not easy to change the JAR again. Do we have any workaround where we can solve this.

we really appreciate your support, please give us some suggestions how we can fix this

Regards
Sarin

#29822

A workaround is to preprocess input HTML and to isolate CJK portions of text to separate spans.

WENJUN Money 育暮趣所及定平関島指形由幹設 Laundering in UNITED KINGDOM IS NOT functional now

to

WENJUN Money 育暮趣所及定平関島指形由幹設 Laundering in UNITED KINGDOM IS NOT functional now

Is it doable on your side?

#29823

thanks for your support
we have changed it to call this function before showing the content

import java.util.regex.Matcher;
import java.util.regex.Pattern;

public class CJKFinder {

public static void main(String[] args) {
String str = ” 育暮 1 Great Work M R 12 13 14 育暮趣所及定平関島指形由幹設世戸 122 育暮趣所$及定平関 島指形由幹設世戸 hello. aiguë διαίρεσις”;
System.out.println(applyCJKStyle(str));
}

public static String applyCJKStyle(String text){
Pattern pattern = Pattern.compile(“([u4E00-u9FFF]{1,})”);
Matcher matcher = pattern.matcher(text);
StringBuffer output = new StringBuffer();
while (matcher.find()){
matcher.appendReplacement(output, “$1“);
}
matcher.appendTail(output);
return output.toString();
}

}

[Author]

Posts