Printing with Chinese Fonts, using base_report_unicode

bhikkhu · **Posted:** Wed Sep 02, 2009 1:54 pm

Hi everybody,
Have been struggling with reports in Chinese for some days now. Like to share with you my experiences. I chose the route suggested by Dukai on his blog (implemented in module base_report_unicode in addons-extras) as this seemed the most convenient. Unfortunately it doesn't work immediately for Chinese (and in version 5.0.2).
First I put the fonts SimSun and SimHei in the fonts directory of the module. Then I replaced the fontmap in __init__.py of that module by:

Code:

        fontmap = {
            'Times-Roman':                   'SimSun',
            'Times-BoldItalic':              'SimSun',
            'Times-Bold':                    'SimSun',
            'Times-Italic':                  'SimSun',

            'Helvetica':                     'SimHei',
            'Helvetica-BoldItalic':          'SimHei',
            'Helvetica-Bold':                'SimHei',
            'Helvetica-Italic':              'SimHei',
            'Helvetica-Oblique':             'SimHei',
            'Helvetica-BoldOblique':         'SimHei',

            'Courier':                       'SimHei',
            'Courier-Bold':                  'SimHei',
            'Courier-BoldItalic':            'SimHei',
            'Courier-Italic':                'SimHei',
            'Courier-Oblique':               'SimHei',
            'Courier-BoldOblique':           'SimHei',

            'Helvetica-ExtraLight':          'SimHei',

            'TimesCondensed-Roman':          'SimSun',
            'TimesCondensed-BoldItalic':     'SimSun',
            'TimesCondensed-Bold':           'SimSun',
            'TimesCondensed-Italic':         'SimSun',

            'HelveticaCondensed':            'SimHei',
            'HelveticaCondensed-BoldItalic': 'SimHei',
            'HelveticaCondensed-Bold':       'SimHei',
            'HelveticaCondensed-Italic':     'SimHei',
        }

Note that I added a few permutations. I know for sure that also Helvetica-Oblique is used in some of the reports.
But you also need to to solve a bug in the code that goes unnoticed if not using the Chinese fonts. The bug is here:

Code:

        for  old, new in fontmap.iteritems():
            data = data.replace(old, new)

This replaces the old fonts by the new fonts in the rml files. The problem is that a replacement is made also with partial matches. So the 'Helvetica' in 'Helvetica-Oblique' could very well be replaced by 'Simhei' resulting in the meaningless 'Simhei-Oblique' which results in an error.
I solved this by replacing the above code with:

Code:

   while len(fontmap)>0:
       ck=max(fontmap)
       data = data.replace(ck,fontmap.pop(ck))       

This did the trick! Hope you find it useful.

elicoidal · **Posted:** Sun Oct 11, 2009 2:25 pm

Thanks a lot for this how-to. Some precisions from my side after struggling one afternoon to set it up:

- Those 2 TTF files you mention can be found in Windows XP c:/windows/fonts directory

- I am used not to unzip the imported modules but in case of base_report_unicode, this leads to the problem which is the module cant find the 2 chinese TTF. So I recommand the following steps:
1. download the latest version of the base_report_unicode module from bazar
2. unzip the module in addons directory. Delete zip file in addons directory
3. place the 2 windows fonts file in ../addons/base_report_unicode/fonts with name "SimSun.ttf" and "SimHei.ttf"
3. change the __init__.py as recommanded by bhikkhu
4. Do not forget to reset your rights on addons folder and restart your server.

some information in this post : http://openobject.com/forum/topic9962.html

There is a module called base_report_cn as well, including part of the information of the current post but fonts mapping is not as complete as this one.

Eric