Code and untranslatable content
Given the huge amount of different templating and programming languages, we advise clients to remove any such language from the source content as it's hard to successfully identify such a range of different codes. We've also noticed that sometimes, parts in the code also need to be translated. Please don't do either of this! Use our key-value structure to give context and structure to your text and render the translations back into your code. Many languages and frameworks have tools to help with this (i.e. yaml files, xliff)
Encoding
All our requests are encoded UTF-8 and are interpreted as such.
Entities and HTML
HTML entities are decoded to UTF-8 (e.g. ü becomes ü) so that translators can work more easily with them. Our output will not contain HTML entities.
There is one exception for < and > (< and >) If a text is deemed to contain HTML, as opposed to plain text, any < or > character that is not part of a tag, will be returned as a HTML entity.
Newlines
In a well formatted text newlines denote the boundaries of a sentence or the start of a new paragraph. Many of our customers use them as such. Unfortunately, what we see often is that newlines might be the result of copy pasting or faulty user input. Since we can not reliably tell the difference, we advise clients to remove all newlines that are not meant to differentiate sentences. This way, we can ensure that sentences are segmented correctly with the the best translations as a result.