ZH to EN translations?

[jmgilman]

Would love some assistance here if anyone has figured this out: how would one translate the zh_name field on all of the items to their English counterpart? My first attempt was looking up the actual string found in zh_name in the respective gettext/zh files, however, I only find results for about half of the total items.

Some of the cases that fail seem to be when the name contains ASCII characters, like this one:

衍生无人机 - LMH-M-1 - B

That string does not exist in the zh files, however, if we do a little reverse engineering we can find that the likely locality id is 14029 which is Metadrones - LMH-M-1 - B in English and 衍生无人机 - 露姆赫-姆-1 - 比 in Chinese. It's not clear to me why in the item file there is ASCII characters but in the locality file, there isn't. I believe I was getting like 29k items in total but only 16k of them could make a direct match back to the Chinese locality data.

Other than reverse looking up like I tried above, it's not clear how one is supposed to tie an item name to its translated name in the locality files. I tried searching around for various locality ID's and most only show up in the locality files and nowhere else. Am I missing data or am I looking at all of this wrong?

[xforce]

Translation is possible using the provided script at https://github.com/xforce/eve-echoes-tools/blob/main/scripts/text_lookup.py

Basically you have to hash the chinese text, then do a a division and index lookup to get to the locale id and lookup the text using that.

Hope that makes it clear how to do it.

[jmgilman]

Should have done more digging! That is very helpful. What's interesting is that even using this method I'm still only getting a little over half of the items. Here is an example that results in a KeyError using your script:

  "44530000009": {
    "base_price": 22160000,
    "can_be_jettisoned": false,
    "icon_id": 44503000000,
    "zh_desc": "$$用于修复损毁的舰船残件。",
    "zh_name": "$$米玛塔尔9级零件"
  },

If you try to lookup the zh_name for that item the hash doesn't exist in the message index.

$ python3 text_convert.py '$$米玛塔尔9级零件' en
Traceback (most recent call last):
  File "text_convert.py", line 18, in <module>
    msg_id = msg_index[str(key)]

Is this normal?

One more example:

"99001000018": {
    "can_be_jettisoned": false,
    "zh_desc": "g85tr",
    "zh_name": "加压蒸汽"
  }

[xforce]

If the index doesn't exists, then the client will also not show a translation in-game, there are a few places where I have seen chinese text in-game. I guess that is just the way it is... There are also a few left over dev/qa items in all the data, I suspect those won't have translations, because why bother. The second example does look like a thing that is not supposed to show in-game anyways, because the description indicating g85tr where g85 is a reference to the internal identifier this game runs as in NetEase.

[xforce]

Anyways, yes. There will be many things that don't have translations, either internal test items for qa, for development, just unfinished things that will be released in the future, or just a missing translation. Either of those could apply.

For now, I would just go with what you get, while possible that I may have missed something, I think that is rather unlikely and will only look further into the translations if/when I find an actual in-game accessible item that can't be translated.

[jmgilman]

No worries, you're way more familiar with the data so I just needed a second set of eyes. Thanks for the help!