Re: [Gramps-bugs] GRAMP BUG [win32]

[Benny M. <ben...@gm...>]

Isaac,

your username on windows is indeed the problem. The error GRAMPS throws up
is:
UnicodeDecodeError: 'utf8' codec can't decode bytes in position 33-35:
invalid data
and position 33-35 is exactly the accent in your name.

It looks like the code to open a database environment will have to take into
account filesystem encoding.
Can you file a bug at our bug tracker with the information that is in this
mail, you find it at http://bugs.gramps-project.org/my_view_page.php

As a workaround so you can use GRAMPS:
look for the file const.py in the install dir of gramps. You find the code
piece in it:

#
# Determine the home directory. According to Wikipedia, most UNIX like
# systems use HOME. I'm assuming that this would apply to OS X as well.
# Windows apparently uses USERPROFILE
#
#-------------------------------------------------------------------------
if os.environ.has_key('USERPROFILE'):
    user_home = os.environ['USERPROFILE']
    home_dir = os.path.join(user_home,'gramps')
else:
    user_home = os.environ['HOME']
    home_dir = os.path.join(user_home,'.gramps')

And change this so user_home is a directory that has no strange accents. Eg:


#
# Determine the home directory. According to Wikipedia, most UNIX like
# systems use HOME. I'm assuming that this would apply to OS X as well.
# Windows apparently uses USERPROFILE
#
#-------------------------------------------------------------------------
if os.environ.has_key('USERPROFILE'):
    user_home = 'C:\Documents and Settings'
    home_dir = os.path.join(user_home,'gramps')
else:
    user_home = os.environ['HOME']
    home_dir = os.path.join(user_home,'.gramps')

That should allow you to save grdb files.
Note that once you did the above change, any previously existing grdb can no
longer be opened. Grdb is not a portable format, it is tied to the
environment directory. Export to .gramps regularly to obtain backups.

Benny


2008/2/8, Isaac Sánchez Barrera <isa...@te...>:
>
>  Benny,
> This is the Windows XP output:
>
> Python 2.5.1 <r251:54863, Apr 18 2007, 08:51:08> [MSC v.1310  32 bit
> <Intel>] on win32
> Type "help", "copyright", "credits" or "license" for more information.
> >>> import sys
> >>> print sys.getfilesystemencoding()
> mbcs
> >>> import os
> >>> user_home = os.environ['USERPROFILE']
> >>> home_dir = os.path.join(user_home,'gramps')
> >>> env_dir = os.path.join(home_dir,"env"
> >>> print env_dir
> C:\Documents and Settings\Isaac Sβnchez\gramps\env
> >>> newfilename = 'myfilename'
> >>> drive, tmp_name = os.path.splitdrive(newfilename)
> >>> tmp_name = tmp_name.lstrip(os.sep)
> >>> print tmp_name
> myfilename
> >>> env_name = os.path.join(os.path.expanduser(env_dir), tmp_name)
> >>> print env_name
> C:\Documents and Settings\Isaac Sβnchez\gramps\env\myfilename
> >>>
>
> Maybe the problem is there because my username has got accents, but there
> aren't any errors when saving on .gramps or .ged (and, as I told you on my
> last message, I save the genealogical data in a file without accents nor
> spaces on its name and path).
>
> I've tried GRAMPS on my Ubuntu again and it saves in ".grdb" correctly, so
> I haven't tested the commands on it. If you need me to run those commands on
> my Ubuntu, just tell me and I will.
>
> Isaac
>
>
>
> Benny Malengier wrote:
>
> Jim, you think this is an encoding problem?
>
> Isaac,
>
> First, this code is changed for upcoming  3.0 version, so there it should
> not happen.
> Please test if it is a problem for linux too, this is an os specific piece
> of code.
>
> Could you test some commands in python?
> Just start a python session in windows, then:
>
> python
> Python 2.5.1 (r251:54863, May  2 2007, 16:27:44)
> [GCC 4.1.2 (Ubuntu 4.1.2-0ubuntu4)] on linux2      --> in your case this
> should be win32
> Type "help", "copyright", "credits" or "license" for more information.
> >>> import sys
> >>> print sys.getfilesystemencoding()
> >>> import os
> >>> user_home = os.environ['USERPROFILE']
> >>> home_dir = os.path.join(user_home,'gramps')
> >>> env_dir = os.path.join(home_dir,"env")
> >>> print env_dir
> >>> newfilename = '_WRITE_HERE_THE_FILENAME '
> >>> drive, tmp_name = os.path.splitdrive(newfilename)
> >>> tmp_name = tmp_name.lstrip(os.sep)
> >>> print tmp_name
> >>> env_name = os.path.join(os.path.expanduser(env_dir), tmp_name)
> >>> print env_name
>
> GRAMPS crashes on that second to last line, so the crash is in a python
> module.
> I would like to have the output you obtain from the print statements
> above. Can you paste them here?
> Now, the general problem is that if you write here a filename, it is in
> the encoding of how you type strings, however, in GRAMPS, the filename comes
> from a filechooser dialog, so it is given in the encoding of the filesystem.
>
> Benny
>
> PS, in linux it is user_home = os.environ['HOME'], and the join is with
> .gramps, not gramps
>
> 2008/2/7, Isaac Sánchez Barrera <isa...@te...>:
> >
> > When I save in GRAMPS XML format or GEDCOM, it saves correctly. If I try
> > to save the file in a ".grdb" format, I get the error I told you on the last
> > message.
> > I'm currently using English filenames without spaces (of the format
> > x:\foldername\treefilename.ged OR x:\foldername\treefilename.gramps) and, if
> > I'm not wrong, the GEDCOM (and the GRAMPS file) are using the UTF-8 charset,
> > but there are surnames and individuals with Spanish characters.
> >
> > OS: Windows XP Professional [Win32 5.1]
> >
> > I've also got a Ubuntu-Studio installation and I think the same problem
> > appears when trying to export to a .grdb file.
> >
> > Thanks for your fast answer,
> >
> > Isaac
> >
> > PS: I've just signed up on the Windows list with my gmail account.
> >
> >
> > Benny Malengier wrote:
> >
> > Isaac, this looks like a windows bug with the name of a database.
> >
> > Is it correct you did a Save As, and gave a name with eg spanish
> > characters? Could you give the name you gave?
> >
> > It should work if you limit yourself to english language characters, but
> > we would be interested to know what is not working on Windows. Could you
> > give the Windows version too, and if possible, what charset you work with? I
> > don't use windows so unfortunately I cannot tell you where to find that
> > info, but we have a gramps windows mailing list (
> > http://www.gramps-project.org/wiki/index.php?title=Contact&grampsd_wiki_session=1ac7cb6147a0c92333b0f1cd4112ff65)
> > where you can ask that.
> >
> > Benny
> >
> > PS: export to .gramps format regularly on Windows
> >
> > 2008/2/7, Isaac Sánchez Barrera <isa...@te...>:
> > >
> > > User Information:
> > > ===================
> > >
> > > I was saving on grdb
> > >
> > >
> > > Error Details:
> > > ===================
> > >
> > > 35655: ERROR: dbloader.py: line 488: Failed to open database.
> > > Traceback (most recent call last):
> > >   File "C:\Archivos de programa\gramps\DbLoader.py", line 485, in
> > > open_saved_as
> > >     self.uistate.pulse_progressbar)
> > >   File "C:\Archivos de programa\gramps\GrampsDb\_GrampsBSDDB.py", line
> > > 465, in load_from
> > >     self.load(filename,callback)
> > >   File "C:\Archivos de programa\gramps\GrampsDb\_GrampsBSDDB.py", line
> > > 354, in load
> > >     tmp_name)
> > >   File "C:\Python25\lib\ntpath.py", line 102, in join
> > >     path += "\\" + b
> > > UnicodeDecodeError: 'utf8' codec can't decode bytes in position 33-35:
> > > invalid data
> > >
> > > System Information:
> > > ===================
> > >
> > > Python version: 2.5.1 (r251:54863, Apr 18 2007, 08:51:08) [MSC v.131032 bit (Intel)]
> > > BSDDB version: 4.4.5.2
> > > Gramps version: 2.2.10-1
> > > LANG: es
> > > OS: win32
> > > Distribution:
> > >
> > >
> > >
> > > -------------------------------------------------------------------------
> > > This SF.net email is sponsored by: Microsoft
> > > Defy all challenges. Microsoft(R) Visual Studio 2008.
> > > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> > > _______________________________________________
> > > Gramps-bugs mailing list
> > > Gra...@li...
> > > https://lists.sourceforge.net/lists/listinfo/gramps-bugs
> > >
> > >
> >
>