Unconvertable characters from Stata13 to Stata16

[Anna Bonetto]

Dear Statalisters,

I worked on Stata13 the past few years, creating dofiles and datasets with French characters/accents. I have been using Stata16 for the past few weeks and it seems that these characters have been replaced by little squares with question marks. Do you know how I could recover the previous characters?

Thanks!

[William Lisowski]

In Stata 13 your French characters and accents were created, I think, using Extended ASCII characters (those from 129-256).

Stata 14 and later (congratulations on your upgrade!) support Unicode characters.

If you don't know much about Unicode, then you should start with the output of help unicode and in particular notice the following paragraph.

Second, if you have datasets, do-files, ado-files, or other files that you used with Stata 13 or earlier and those files contain characters other than plain ASCII such as accented characters, Chinese, Japanese, or Korean (CJK) characters, Cyrillic characters, and the like, you should read [D] unicode translate.

For comprehensive documentation, Stata recommends you start with section 12.4.2 Handling Unicode strings in the Stata User's Guide PDF. Although it's largely about working with data containing Unicode characters, it gives you the background you will need to understand. And then read the unicode command entries in the Stata Data Management Reference Manual PDF. Both PDFs are included in your Stata installation and accessible from Stata's Help menu.