Parsing three-letter language names as culture info

Question

I'm working with an API that returns some information about audio streams in a file, more specifically the audio language in its three-letter ISO name (ISO 639-2) representation.

I would like to parse this information into a new CultureInfo object, but there is no constructor that takes a three-letter code. I could of course write an enormous select statement (switch for you C# people), but I figured it would be more cost-efficient to ask around for a better way first. So am I out of luck or is there a secret way to create a CultureInfo object using three letter names?

I don't think there is a good way to go in the opposite direction. For example, **eng** can range anywhere from **en-029** to **en-ZW**. How would you expect this to behave if you read **eng**? — myermian, Mar 23 '12 at 15:00

Balazs Tihanyi · Accepted Answer · 2012-03-23T15:28:33.817

10

EDIT: sorry, I've used the wrong property:

public static CultureInfo FromISOName(string name)
{
    return CultureInfo
        .GetCultures(CultureTypes.NeutralCultures)
        .FirstOrDefault(c => c.ThreeLetterISOLanguageName == name);
}

However, there are still duplicates in the list and no support for "dut".

edited Mar 23 '12 at 15:28

answered Mar 23 '12 at 15:04

Balazs Tihanyi

6,659
5
23
24

Close, but no cigar. :( It fails when there is more than one letter code for the same language (example: both "dut" and "nld" are used for Dutch, the framework doesn't recognise "dut"). Apparently Microsoft chose to include only one letter code per language, very unfortunate. – Steven Liekens Mar 23 '12 at 15:23
I'll mark this as the answer as it's the closest it will get with the current framework implementation of ISO 639-2 codes. – Steven Liekens Mar 23 '12 at 16:28
I've discovered that there are only 21 languages that have both a bibliographic code and a terminology code. It's not too much effort to convert between the two and this method works fine after conversion. – Steven Liekens Mar 23 '12 at 16:45

score 2 · Answer 2 · answered Mar 23 '12 at 15:12

2

I would go for Balazs solution, but it would be better in your case to use CultureTypes.NeutralCultures as you don't seem to be concerned with region/country data.

It would always return a single CultureInfo with no need of FirstOrDefault

answered Mar 23 '12 at 15:12

user484189

191
5

score 0 · Answer 3 · answered Aug 14 '13 at 09:53

Here is an extension method for the Silverlight System.Globalization.CultureInfo class, that gets the ISO 639-2 three-letter code for the language of the current System.Globalization.CultureInfo object. It uses ISO-639-2 table, that is retrieved via Utils.GetResourceStream() custom method.

The implementation also relies on a custom String.NthIndexOf() method.

public static string ThreeLetterISOLanguageName(this CultureInfo cultureInfo)
{
    const string separator = "|";

    using (var reader = new StreamReader(Utils.GetResourceStream("ISO-639-2_utf-8.txt")))
    {
        while (!reader.EndOfStream)
        {
            string line = reader.ReadLine();
            //two-letter ISO code is in the third column, i.e. after the second separator character
            string twoLetterISOCode = line.Substring(line.NthIndexOf(separator, 1) + separator.Length, 2);
            if (!twoLetterISOCode.Equals(cultureInfo.TwoLetterISOLanguageName)) continue;
            return line.Substring(0, 3);
        }
    }

    return null;
}

Full Gist

score 0 · Answer 4 · answered Mar 23 '12 at 14:47

0

There is nothing built in that will help with such parsing.

Instead of a select, you can create a Dictionary(Of string, CultureInfo) to map from one to the other. This is more as an ease of use suggestion.

answered Mar 23 '12 at 14:47

Oded

489,969
99
883
1,009

Parsing three-letter language names as culture info

4 Answers4