高级会员
注册日期: 06-11
帖子: 14579
精华: 1
现金: 224494 标准币
资产: 234494 标准币
|
【转帖】roubles With Japanese Kanji Characters
troubles with japanese kanji characters
troubles with japanese kanji characters
hi,
i have two questions relating to the handling of code pages:
my first question relates to the use of oddbdatabase()->getdwgcodepage() function.
the attached test file contains japanese kanji characters. the opendwg
"adshortcodepagetoascii(adhd->dwgcodepage,codepagestr)" function returns a a codepage value of dos932, which is what i would expect with 932 being the japanese codepage.
we can use opendwg to display the kanji text string successfully on a japanese localised machine. this matches the text string that is displayed when opening the dfx file in notepad on the same machine.
however, oddbdatabase()->getdwgcodepage() function returns p_ansi_1252, which is the codepage for western eurpope. this value apears to be hard coded in the sysvardef.h file. is there an alternative method for getting the codepage of the database?
my second question relates to how the currproperties field of oddbtextiterator doesn't appear be to be getting set properly.
the documentation says that:
"the returned character will be a unicode character in all cases, except for when the binbigfont flag is set in the currproperties() value. in this case, the returned character will be mbcs (corresponding to a \m+nxxxx character in the original string)."
attached below is some code to iterate through the input string, which is: \m+18343\m+18393\m+18374\m+18348\m+1837d\m+18365\m +18342\m+1834e\m+18358
i am finding that the currprops.binbigfont is never set in the following code, using the attached test file. am i doing something wrong?
many thanks,
peter.
//////////////////////////////////////////////////////////////////////////////////////
oddbstring strtext=ptext->textstring();
odcodepageid codepage=cp_dos932;
odgitextstyle textstyle;
odstring strfontname=ptextstylerec->filename();
textstyle.setcodepage(codepage);
odstring strtypeface;
bool bbold=false, bitalic=false;
int ncharset=0, npitchandfamily=0;
ptextstylerec->font(strtypeface,bbold,bitalic,ncharset,npitchand family);
textstyle.setfont(strtypeface.c_str(),bbold,bitali c,ncharset,npitchandfamily);
odstring strbigfontfilename=ptextstylerec->bigfontfilename();
double dtextsize=ptextstylerec->textsize();
double dxscale=ptextstylerec->xscale();
double dobliquingangle=ptextstylerec->obliquingangle();
bool bisbackward=ptextstylerec->isbackwards();
bool bisupsidedown=ptextstylerec->isupsidedown();
bool bisvertical=ptextstylerec->isvertical();
textstyle.set(strfontname,strbigfontfilename,dtext size,dxscale,
dobliquingangle,50.0,bisbackward,bisupsidedown,
bisvertical,false,false);
oddbtextiteratorptr ptextiter=oddbtextiterator::createobject(strtext,s trtext.getlength(),false,codepage,&textstyle);
const odcharacterproperties& currprops=ptextiter->currproperties();
if (currprops.binbigfont)
{
trace("big font true\n");
}
const odchar* pchar=ptextiter->currpos();
while (*pchar!=0)
{
oduint16 nchar=ptextiter->nextchar();
// do something with the character here.
pchar=ptextiter->currpos();
}
attached files
quote:
originally posted by petera
however, oddbdatabase()->getdwgcodepage() function returns p_ansi_1252, which is the codepage for western eurpope. this value apears to be hard coded in the sysvardef.h file. is there an alternative method for getting the codepage of the database?
the sysvardef.h file contains the default value(for example: for create new file). the function oddbdatabase()->getdwgcodepage() returns value which dwgdirect loads from file. oddbhostappservices->readfile() method has parameter ballowcpconversion. in case it flag is true dwgdirect converts symbols in system code page and sets dwgcodepage equal system code page. unfortunately it isn't work with dxf file(there is bug in dwgdirect v.1.12). i am not able reproduce your situation. for attached file oddbdatabase()->getdwgcodepage() returns dos932. i have checked it in v.1.12.
quote:
originally posted by petera
the documentation says that:
"the returned character will be a unicode character in all cases, except for when the binbigfont flag is set in the currproperties() value. in this case, the returned character will be mbcs (corresponding to a \m+nxxxx character in the original string)."
the binbigfont flag is set when text style contains big font and this font contains current character. there is no big font in text style in attached file.
--
best regards,
sergey zaitcev
|