A problem of letter conversion

[Gelez]

Hello,

I work under MI 9.5.1.
I meet a problem of letter conversion. On a French forum (I'm from
Paris, France), no one was able to answer my questions.

Here's the problem:
I hold an Excel file of about 3.000 lines with data for Central Europe
villages. I want to open it with MI and conserve all particular
letters in the names of these villages (Č, č, Ć, ć, Đ, đ, Š, š, Ž, ž).
If I open directly the file, only Đ, Š, š, Ž, ž are preserved; but the
"c" with accentuation are not kept: “d” and "c" is there, but there is
no accent. It seems that migration is run with a code that does not
take them in account (perhaps a WindowsLatin1-type code, which knows
Đ, Š, š, Ž, ž). MI does not suggest any other code for migration.

Under Excel, I thus decided to save my .xls in Text UNICODE (.txt),
the only format that keeps on the accents. When opening the .txt file,
MI gives a list of code to read it. I tried several times with
different codes, but no one is able to read my “d” and "c" with
accents. What is worse: apparently, MI does not know UNICODE (UTF-8)
code and related .txt files, but only simple .txt: actually, the
resulting .TAB has only one column where you find strange letters.
Nothing better when I erase the title line.

I thus decided to manage the problem in the frame of MI. As I first
worked under MI 7.5, I kept my .TAB with a WindowsLatin1 charset, and
transformed my letters as following:

Č = È
č = è
Ć = Æ
ć = æ
Đ = Đ
đ = [I did not find the corresponding letter]
Š = Š
š = š
Ž = Ž
ž = ž

Then, for labeling, I had to choose the "Arial CE" or similar CE
police, and I had my accents. This worked pretty well, until I
upgraded MI to 9.5.1 version. It does not work any more.

Under 9.5.1., I tried to write directly in the table with my specific
letters. So I changed the charset of my table (and of the
workspace .WOR) in WindowsLatin2. Nothing.

Now, I don’t know how to get on with this file!

Thanks for help!

Philippe

[eric.bl...@pb.com]

Phillipe,

MapInfo Professional only fully supports one language group or
character set at a time and that is determined by the "non-Unicode
language" that is set in the Control Panel, Regional settings dialog
in Windows. So assuming that your machine is set for France,
Professional can ony fully support the Western European languages. To
have Professional support the Central European languages (Latin2) you
would have to reset your machine on this "language for non-Unicode
programs" setting for one of the Central European languages. Then all
the characters would be supported.

Eric Blasenheim
Chief Product Architect
Pitney Bowes Business Insight (MapInfo)

[Warren Vick]

Hi Eric,

The lack of support for full Unicode seems to be an increasing problem with MI Pro. A few years ago it was just a handful of people who were trying to "bend" and International English Pro to work in Japanese, Arabic, etc., but there have been at least two queries here on the -L' this year so far and quite a few people ask me about the tricks to get it working (which you kindly taught me). It causes huge problems (e.g. data loss) when integrating with any external DB from a modest Access/Jet DB to Oracle. Could this be regarded as an issue of increasing significance and put into the road map, if it's not there already?

Regards,
Warren Vick
Europa Technologies Ltd.
http://www.europa.uk.com