bug with the Type3 font

[imported from SourceForge]
http://sourceforge.net/tracker/index.php?group_id=78314&atid=552832&aid=1086372
Originally submitted by jinfeng_wang on 2004-12-16 03:49.

hi, ben , today i found another bug withe the type3
font.

you know, the type3 font is used to display some other
special "chatacter" exception the ASCII font, for
example, the maths formula, the CJK chararter in some
PDF file. in the PDFont.encode(), it will revoke the
following :
retval = encoding.getCharacter( getCodeFromArray( c,
offset, length ) );

in the static Encoding.getCharacter( COSName name ),
it should return the charcter according to the cosname.
however, in the encoding.getCharacte() , it will return
the cosname when meeting a "special character".

i think the following code should be comment out:
if( character == null )

{ // it will return the Name to replace the correct chacter. character = name.getName(); }

[comment on SourceForge]
Originally sent by jinfeng_wang.
Logged In: YES
user_id=1145721

the PDF file using Type3 Font:

http://www.census.gov/population/socdemo/race/indian/cp-3-
7/tab01.pdf - the dash character is Type3 font.
http://www.math.pku.edu.cn/course/probability/download/11.
pdf - all the Chinese is Type3 font.
http://ee-wcl.tamu.edu/itw2004/program/yeung_inv.pdf

you can refer the PDF Reference version 1.5 page 386.
http://partners.adobe.com/public/developer/pdf/index_referen
ce.html

you can try to copy the text from the adobe reader , and
comparing to the result of the PDFBox.

certainly, if the Encoding.getCharacter() can return null as i
have modified, the code in PDFStreamEngine.showString()
should be corrected as following :

if (c!=null)

{ stringResult.append( c ); log.info("char="+c); }

else

{ continue; }

[comment on SourceForge]
Originally sent by benlitchfield.
Logged In: YES
user_id=601708

Please attach or mail me a pdf with this.

Thanks
ben@csh.rit.edu