Problems with mixed code page upgrade to Unicode

From SuperMemopedia
Jump to: navigation, search

From: Bruno
Country: {{{Country}}}
Sent: Aug 2, 2017
Subject: Help with upgrade to SuperMemo 17


I am not sure how to upgrade my Supermemo 2008 collection to SuperMemo 16 on my way to SuperMemo 17.

My collections have been created on a French Windows system. I use French and Cyrillic characters (German an Spanish as well but I guess using the French code page). Those could be mixed in the same item (Question in French/Answer in Russian). I tried to convert the collection to SuperMemo 16 indicating a Cyrillic code page. As you can imagine the Russian part was fine but the French, German and Spanish special characters where converted to their Cyrillic equivalent.

By the way many thanks for the awesome product, I've been using for quite a some time now. Never missed a day!

Best Regards Bruno

SuperMemo user since 06/2000

Build=14.03 Release date: Nov 25, 2009


  • For your type of upgrade, there is an important function missing in SuperMemo: extracting the code page from the template for conversions to Unicode. Currently, all upgrades are done along a single code page. That code page information is not wired in the text. It is only stored in the font or in the template.
  • The above limitation means that for correct conversion you need to use template functions to separate texts for upgrade, or for manual conversion (e.g. with search&replace). This is pretty complex.
  • XML format can be used as an intermediary for items that can be converted with search&replace
  • If you are able to read your items, and the number of conversions is not too large (e.g. as in Spanish), you might consider doing the whole job incrementally, i.e. first convert to Unicode using Russian code page, then fix individual Spanish mistakes during repetitions
  • If this is overwhelming, you can send your collection for help with conversions
  • As we no longer receive requests for upgrade help, writing a function for Unicode conversion based on old text templates is not a priority. This means that it would be wasteful to wait for the implementation