mathisbullinger / leganto-mono Goto Github PK

When translating a document into multiple languages, the number of possible combinations (disregarding order of languages) is $\sum\limits_{i=1}^{n} \frac{n!}{i!(n-i)!} = 2^{n-1}$, where n is the number of languages the document is translated to.
E.g. for a document that is translated to German, English, and Spanish, the possible combinations are de, en, es, de,en, de,es, en,es, and de,en,es.

To store these combinations in a denormalized system, certain limits must be set. A first step might be to limit the number of languages that can be displayed at any given time, resulting in the number of combinations being $\sum\limits_{i=1}^{x} \frac{n!}{i!(n-i)!}$, where x is the maximum number of languages that can be displayed. This result in the following number of combinations:

Clearly, just limiting the number of languages per combination is not flexible enough if the number of resulting combinations should be kept reasonably low.

A better approach may be setting a limit on the allowed combinations per document and adjusting the number of languages that can be displayed concurrently y based on the number of languages the document is translated to x to fit the limit:

Setting the limit at 1200 combinations, this would result in documents translated in up to 19 languages being viewable in three languages at a time, and documents being translated in to up to 48 languages being viewable in two languages at a time, setting an effective limit for the number of languages a document can be translated to.