Language Folder: ISO 639-2?

[diacritics]

Language Folder: ISO 639-2?

[sixtyfive]

Hi julmot, I just see that the answer I sent you 6 days ago via email has not arrived here. I've just arrived at the meeting's venue and will be able to talk to everyone over the next two days.

[Mottie]

Well, I am totally clueless when it comes to this language. We are including non-Latin scripts, but since we're unfamiliar with the details of Aramaic and Arabic, we'll need help. Would you be able to help contribute a corrected language file? I know it's a lot of work, but there's no rush on completing it.

[Mottie]

Hi @sixtyfive!

The ISO 639-1 portion of the spec is in reference to the language folder. If the folder is missing, please let us know. Which language are you working on?

If you want to contribute to a language, use the IETF language tag. Here is a handy IETF reference if you need it.

[julmot]

@sixtyfive Please let us know if we can close this issue.

[sixtyfive]

@julmot, sorry, I have a dayjob and can tend to this second job only so often. I did write a reply do @Mottie this morning, though - unfortunately it seems not to have gotten through to here.

I opened the issue specifically because of the language folder. In simple words: do you allow ISO 639-2 codes in there? I'm asking because the language I'm working with does not have a ISO 639-1 code and your language folder does not have a single ISO 639-2 code in it...

[Mottie]

Which language? If we need to, we can update the spec.

[sixtyfive]

Specifically, it's a subvariant of a variant of Aramaic. The variant is mostly referred to as Syriac with both the ISO 639-2 and 639-3 codes being "syc".

[sixtyfive]

Since you added the support label - I've no need for anything atm, I just came here and took a look because I was invited to :)

[julmot]

@sixtyfive The support tag is used until a bug or feature request can be determined by the issue. Until now this wasn't the case. However, if this changes, the tag will change too.

I see that Syriac is a dialect of Aramaic, which is spoken by a minority. The countries in which this language is spoken have a different official language.

Anyway, the first question would be: Are there any diacritics that need mapping at all? And if so, we need to decide whether or not to include it. If I can remember @Mottie we've not included another language spoken by a minority before. In contrast to that, we've included Swiss German which is spoken by almost 5 million people. Syriac is spoken by about 3 million people. What do you think?

@sixtyfive Just for my interest: Who did invite you? Thanks for having a look around. You're invited by me and @Mottie to share your knowledge if you like -- it would be very much appreciated.

[Mottie]

Syriac is spoken by about 3 million people. What do you think?

Well, I think if a language include diacritics, it should probably be included in the database. I'm not sure what population cutoff we should set, but 3 million still sounds like a decent amount to me.

Who did invite you?

I invited him to look at the database and help out. @sixtyfive submitted a layout to my keyboard project which included a few diacritics. I didn't want to assume any specific language, so I left it open.

[sixtyfive]

Putting my linguist hat on, I'm wondering whether a better measure than number of active speakers might be activity of the language community. Some of the Aramaic dialects (including Suroyo [aka Turoyo]) have very active efforts underway to hold or increase the number of speakers. Another example would be the German dialect Ostfriesisch, which is even taught in school again.

As for the diacritics, I know of no writing system used for Aramaic that wouldn't have any. In the case of Syriac, three writing systems are used: Arabic and Latin scripts, which you already have all the diacritics catalogued for on first glance. And the Syriac writing system which comes in two flavours of vowels. I'm not an expert on Syriac script but if no urgent need arises on your side, I'll try and contribute it during my next trip to Berlin (early next year?) where I'll meet a bunch of people who are.

[julmot]

Hey @sixtyfive,

Has the meeting already taken place?

Thanks for your reply.

[sixtyfive]

@julmot, hello again. So it turns out all this time I had the info about the Surayt diacritica sitting right under my nose, in the first two chapters of the textbook that I've been working on.

Could you take a look at https://beta.surayt.com/en/Level%20A/1?content-fragment-id=1736%20ff. and https://beta.surayt.com/en/Level%20A/2.2?content-fragment-id=1748%20f. and let me know if that has all the info you need? If not, then please let me know what else you need and I'll ask for it.

[Mottie]

I put together this language file. From my understanding it could go into the ar folder and the file would be named ar-sy.json. Or, were you saying when opening this issue that there should be a syr folder and this file should be named syr.json?

Anyway, please verify the following data:

ar-sy.json

{
  "metadata": {
    "alphabet": "Syrc",
    "continent": [
      "AS"
    ],
    "language": "Syriac",
    "languageNative": "ܠܫܢܐ ܣܘܪܝܝܐ",
    "source": [
      "https://beta.surayt.com/en/Level%20A/1?content-fragment-id=1736%20ff",
      "https://beta.surayt.com/en/Level%20A/2.2?content-fragment-id=1748%20f"
    ]
  },
  "data": {
    "ä": {
      "case": "lower",
      "mapping": {
        "base": "a"
      }
    },
    "č": {
      "case": "lower",
      "mapping": {
        "base": "c"
      }
    },
    "ḏ": {
      "case": "lower",
      "mapping": {
        "base": "d"
      }
    },
    "ë": {
      "case": "lower",
      "mapping": {
        "base": "e"
      }
    },
    "ġ": {
      "case": "lower",
      "mapping": {
        "base": "g"
      }
    },
    "ḥ": {
      "case": "lower",
      "mapping": {
        "base": "h"
      }
    },
    "ṣ": {
      "case": "lower",
      "mapping": {
        "base": "s"
      }
    },
    "š": {
      "case": "lower",
      "mapping": {
        "base": "s"
      }
    },
    "ṭ": {
      "case": "lower",
      "mapping": {
        "base": "t"
      }
    },
    "ṯ": {
      "case": "lower",
      "mapping": {
        "base": "t"
      }
    },
    "ž": {
      "case": "lower",
      "mapping": {
        "base": "z"
      }
    },
    "Ä": {
      "case": "upper",
      "mapping": {
        "base": "A"
      }
    },
    "Č": {
      "case": "upper",
      "mapping": {
        "base": "C"
      }
    },
    "Ḏ": {
      "case": "upper",
      "mapping": {
        "base": "D"
      }
    },
    "Ë": {
      "case": "upper",
      "mapping": {
        "base": "E"
      }
    },
    "Ġ": {
      "case": "upper",
      "mapping": {
        "base": "G"
      }
    },
    "Ḥ": {
      "case": "upper",
      "mapping": {
        "base": "H"
      }
    },
    "Ṣ": {
      "case": "upper",
      "mapping": {
        "base": "S"
      }
    },
    "Š": {
      "case": "upper",
      "mapping": {
        "base": "S"
      }
    },
    "Ṭ": {
      "case": "upper",
      "mapping": {
        "base": "T"
      }
    },
    "Ṯ": {
      "case": "upper",
      "mapping": {
        "base": "T"
      }
    },
    "Ž": {
      "case": "upper",
      "mapping": {
        "base": "Z"
      }
    }
  }
}

[sixtyfive]

I put together this language file. From my understanding it could go into the ar folder and the file would be named ar-sy.json.

No! Completely different thing! ar-sy would signify the Syrian dialect of Arabic.

The Syriac dialect of Aramaic on the other hand is a more difficult topic. Without going into too much depth, in your above scheme of lang-dialect it would probably be most appropriately represented as syr-tru (ISO-639-* seem to consider Syriac to be unrelated to Aramaic. Thus syr for lang and then tru [stands for Turoyo which is another name of Surayt] for dialect).

Or, were you saying when opening this issue that there should be a syr folder and this file should be named syr.json?

I don't know how your project prefers to organize things. I guess if there's arguments against syr-tru and the lang-dialect scheme is not a requirement then just syr or just tru would be okay as well. In that case, tru would be more precise.

Anyway, please verify the following data:

Your list only contains Latin transliteration characters. Looking at https://github.com/diacritics/database/blob/master/src/ar/ar.json I thought that this project was interested in non-Latin scripts as well? If so, then it would probably be a good idea to start out there and then afterwards add the Latin-script variant?

[sixtyfive]

Glancing over your ar.json that looks fine. Don't worry about not having anything for any of the Arabic dialects - the only more or less formal ways of putting them into writing are in use in Academia and are very many even nowadays (native speakers of the dialects just improvise all day long - basically the same as if you were putting Scottish into writing on WhatsApp with your mate from Edinburgh).

As for Surayt, I'll have to ask around if there's someone on the project who'd be willing to write one of these 2.5kLOC JSON files...

[julmot]

Any update in this regard?

[julmot]

Closing due to inactivity.