View source for Talk:Dhivehi and English/Contrastive Grammar

އަހަރެން{{tag|prn}}{{tag|pers}}{{tag|p1}}{{tag|sg}}{{tag|std}}{{tag|dir}} އެކަނި {{tag|adj}} ދުކޮހް {{tag|v}}{{tag|tv}}{{tag|act}}{{tag|pres><p3><imp> ލާ{{tag|mod}}

https://en.wikipedia.org/wiki/Template:Lrm
https://www.w3.org/TR/WCAG20-TECHS/H34.html

How does machine-based translation really work for right-to-left (RTL) languages?
Does the machine see the RTL character first and then start reading from the other side? If so, does it start by reading the tags, but backwards? Does it read the tags forwards? Or, does it read the entire word? Does it just ignore the RTL character and read the word "backwards?"
When adding tags "before" (when reading RTL) words, we experience weird out-of-placement issues. We can do this with a very brute-force tag-placeholder-set-up using a text editor like libre office or text edit or another, but we are wondering if there is a better way to deal with this.