ASCII Data in Text Archive Files
When a table that contains only ASCII characters is exported to a text archive file, the .idt file adheres to the basic archive file format. If the table contains non-ASCII information, the format of the archive file is extended to include code page information.
When a table that contains only ASCII characters is exported to an archive file, the .idt file is in the basic archive file format. Each stream in the table is exported as a file with an .ibd file name extension. The .ibd files are stored in a folder with the same name as the table. For example, consider the export of the following Binary table.
The directory structure after exporting this table is as follows. The information in the database table is exported to Binary.idt. The two streams of binary data are exported to Book.ibd and Cars.ibd saved in the folder named Binary.
Binary.idt [Binary] Books.ibd Cars.ibd
The Binary.idt archive file is in the basic archive file format and would look as follows.
Name Data s72 v0 Binary Name Books Books.ibd Cars Cars.ibd
If the file contains non-ASCII data, the basic archive file format of the .idt file is extended to include code page information. The third row in the .idt table is the numeric code page followed by the table name and primary key column names separated by tabs.
Note An .idt file that contains non-ASCII information should be saved in the ASCII format. For example, a text archive file can contain the column and table names encoded as UTF-8, but the archive file itself should be ASCII.
The following ActionText table localized into French would contain non-ASCII information. The numeric code page used for French strings is 1252.
|ADVERTISE||Publication d'informations sur l'application|
The exported archive file, ActionText.idt, would be as follows.
Action Description Template s72 L0 L0 1252 ActionText Action Advertise Publication d'informations sur l'application
Note If a text archive file contains non-ASCII data, the archive file includes code page information. Archive files with code page information can only be imported back into a database of that exact code page or into a language neutral database. In the case of a language neutral database, the code page is set to the code page of the archive file. For more information about how Windows Installer handles code pages see the section Code Page Handling (Windows Installer).