Request for German language support (de)

Hi team,

I would like to ask if there is interest in adding German language support to WFinfo. Currently, only English and Korean are supported via Tesseract OCR. I am willing to contribute and work on integrating German (de) language support, including providing the necessary de.traineddata file and adapting the codebase where needed.

Is German localization and OCR support something the project would welcome? If so, I can start preparing the required changes and submit a pull request.

[D1firehail]

I don't see any issues with adding German OCR support. UI localization would involve a lot more work, so I'm hesitant about that.

#168 and #231 are good references. This recent addition is the only thing I can currently think of that isn't covered by the content or comments of either.
If you want to reorganize the language-dependent checks to be easier to find, you're welcome (but not expected) to do so in my opinion.

I think the .traineddata files are taken from the official tesseract repos, although not sure if it's the "normal", best, or fast ones. Since German is not in our libs branch currently, a temporary solution to get started would be to manually download and rename whichever you choose, then add a bypass here.

@D1firehail

I am currently translating the names of the blueprints from English into German to compare them more easily.

One issue I’ve noticed is that the word “Blueprint” is missing from all Warframe components in market_items.json and name_data.json, even though it is present in the main blueprints. I’m not entirely sure why that’s the case.

I’m currently at the following function:

public int LevenshteinDistanceGerman(string s, string t)
{
    // s == English name
    // t == German name

    //s = s.Replace("Blueprint", "").Trim();
    //t = t.Replace("Blaupause", "").Trim();

    s = GetLocaleNameData(s);

    return LevenshteinDistanceDefault(s, t);
}

[...]

public string GetLocaleNameData(string s)
{
    string localeName = "";

    lock (marketItemsLock)
    {
        if (marketItems != null) // Add null check
        {
            foreach (var marketItem in marketItems)
            {
                if (marketItem.Key == "version")
                    continue;

                string[] split = marketItem.Value.ToString().Split('|');
                if (split[0] == s)
                {
                    localeName = split.Length > 2 ? split[2] : "";
                    break;
                }
            }
        }
    }

    return localeName;
}

For example:
"Baruuk Prime Systems Blueprint" returns an empty string ("") because in market_items.json the entry

"639a8d9374b02f003889eb7e": "Baruuk Prime Systems|baruuk_prime_systems_blueprint|Baruuk Prime: Systeme Blaupause",

is stored and the "Blueprint" part seems to be missing in the relevant fields.

[D1firehail]

Seems like the issue is GetPartName calling LevenshteinDistance using the key from nameData instead of the value. It doesn't currently run into issues because the english (default) levenshtein distance method never calls GetLocaleNameData.

Korean runs into the same issue you did, but seemingly nobody is using the korean language support so the issue hasn't been reported.

Thanks! That was the problem.