My suggestion is to build a community to make the task lighter I assume it's a big map since of the patch. Similar to Github.
For a start you may want to begin with the simplest, war3map.wts & war3mapMisc.txt editable via text editor that preserves letters/symbols that are non-english. The usual 2 files that are easily translated by 1 person if wts is not that lengthy.
...as well as mapname editable via Hex Editor. war3map.w3a/u/t/h are either translated via community of translators or via machine translation. The javascript program war3toJSON may be useful for since JSON format is openable via simple text editor. If you can't find those files (abilities,units,items,buffs) you may to dewidgetize it first. Simple ones use SLK tool but I heard there are other tools for more complex map protection methods. Check out more advanced guildes or ask other cool wc3edit fellas or similar social platforms about it.
Lastly hardest to translate is the war3map.j
I use notepad++ 7.5 w/ translate plugin (i prefer free google translate since latest version of notepad++ needs paid Microsoft Translate API code)
If one could simply make a program that repeatably searches for non-latin (e.g. korean, russian, chinese...) REGEX pattern of
Code:
(\b((?:(?![a-zA-Z])\pL){1,})\b)
then press Ctrl+Alt+Z by routine to translate, that would be really easier.
Also you might want to find & replace also via Regular Expression
Code:
(\|c\S\S\S\S\S\S\S\S)
then replace it with
Code:
\1
since machine translation ignores stuffs like
Quote:
|cff440029Cacke
but translate fine with
Quote:
|cff8855O0 Hapyto
I know those are latin transliterals but I hope you get my point (just imagine their cyrillic counterpart).
A couple of honorable mentions for Machine Translation are Microsoft Office's MS WORD 2010 with translate document feature. Some software even has offline file translate such as with PROMT's.