keymanapp / keyboards

Open Source Keyman keyboards
138 stars 256 forks source link

HEX 2019 for ISO-15919 Avagraha transcription not available please add #2539

Closed gregjay closed 4 months ago

gregjay commented 4 months ago

itrans_roman.keyboard purports to support both IAST and ISO-15919. However while "'" is used in IAST for Avagraha. HEX 2019 is used in ISO-15919. Please provide in a future version a simple way to type Avagraha HEX 2019 for use in ISO-15919 transliteration.

Thank you

Greg

mcdurdin commented 4 months ago

Would you be willing to submit a pull request that implements this change? I assume you mean Unicode character U+2019 () representing Devangari Sign Avagraha (, U+093D)?

https://help.keyman.com/developer/keyboards/ has some instructions on how to submit a pull request

gregjay commented 4 months ago

I don't know what a pull request is. But sure I'm willing to do it. Yes, I want to use your keyboard for ISO-15919 input but I need U+2019 which is the ISO transliteration of Avagraha (U+093D).

I just went through your help file for your keyboard and it states that transliteration of Avagraha (U+093D) to Latin letters is U+0027 for both IAST and ISO. This is incorrect. It is correct transliteration for IAST but NOT for ISO. I'm just suggesting you add a simple way to type U+2019 which is the actual Latin transliteration of Avagraha (U+093D) according to ISO-15919.

Cheers

Greg

On Tue, Feb 6, 2024 at 5:36 PM Marc Durdin @.***> wrote:

Would you be willing to submit a pull request that implements this change? I assume you mean Unicode character U+2019 (’) representing Devangari Sign Avagraha (ऽ, U+093D)?

https://help.keyman.com/developer/keyboards/ has some instructions on how to submit a pull request

— Reply to this email directly, view it on GitHub https://github.com/keymanapp/keyboards/issues/2539#issuecomment-1931217499, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACBE5Z6POK3JUTUWLJCZBSLYSLZCXAVCNFSM6AAAAABC5CYJOCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMZRGIYTONBZHE . You are receiving this because you authored the thread.Message ID: @.***>

Meng-Heng commented 4 months ago

Hello @gregjay,

We will make changes and create a pull request on your behalf. You can guide us through what needs to change? If not mistaken, the section shown below is where the change needs to be? image Feel free to show the correct form if the changes below are incorrect. image We went for this change instead of adding a dedicated key for U+2019 because of the convenience of typing (U+0027) > ’ > ʼ > (U+0027).

Changes are made for the keyboard too, here are the sequences:

The extra ʼ(U+02BC) is not going to represent anything, it appears to look the same but has a different Unicode.

Let us know whether this change suits your usage for ISO.

gregjay commented 4 months ago

Thank you

A. These solutions are very good. I need only U+0027 for IAST and U+2019 for ISO. B. I understand why you want to add U+02BC but I feel that it will be confused by some for U+2019 as it appears to be the same glyph even though it is not the same Unicode point. C. As for typing 4 times to return to U+0027 I think this is unnecessary as 1 time will be enough.

Thus I feel that since the purpose of your keyboard is to represent IAST and ISO diacritics only Solution A. above is required. Your ideas for B. and C. above are adding confused and redundancy in my opinion, but it's your keyboard and even adding these ideas would not be a problem for me. Solution B. might be a problem for others who want to take a text created with your keyboard and transcribing it automatically with a site like Aksharamukha or another utility. If they are unable to see the difference between U+2019 and U+02BC then there might be problems changing to Indic scripts primarily Devanagari.

Thanks for all your help and suggestions. If you can let me know when the next version is available I would be very grateful.

Greg

On Wed, Feb 7, 2024 at 7:49 PM MengHeng @.***> wrote:

Hello @gregjay https://github.com/gregjay,

We will make changes and create a pull request on your behalf. You can guide us through what needs to change? If not mistaken, the section shown below is where the change needs to be? image.png (view on web) https://github.com/keymanapp/keyboards/assets/90595388/3ad30068-3b82-41c4-bf58-16ee5e25a39a Feel free to show the correct form if the changes below are incorrect. image.png (view on web) https://github.com/keymanapp/keyboards/assets/90595388/156c2a04-fd32-40a6-9189-beac8858ea27 We went for this change instead of adding a dedicated key for U+2019 because of the convenience of typing (U+0027) > ’ > ʼ > (U+0027).

Changes are made for the keyboard too, here are the sequences:

  • type ', it would output '(U+0027).
  • continue to type ' two times, it would output ’(U+2019).
  • continue to type ' three times , it would output ʼ(U+02BC).
  • type the fourth time, it would go back to '(U+0027).

The extra ʼ(U+02BC) is not going to represent anything, it appears to look the same but has a different Unicode.

Let us know whether this change suits your usage for ISO.

— Reply to this email directly, view it on GitHub https://github.com/keymanapp/keyboards/issues/2539#issuecomment-1933403861, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACBE5ZYINA7J75XCCBPL3B3YSRRNVAVCNFSM6AAAAABC5CYJOCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMZTGQYDGOBWGE . You are receiving this because you were mentioned.Message ID: @.***>

DavidLRowe commented 4 months ago

IIUC the keyboard already cycles through U+0027 and U+02BC. Perhaps a simpler solution would be to change the U+02BC to U+2019. (As you note, often the two are visually similar or identical in a given font, and it could be a source of confusion to have both.)

Can you point to an online reference to ISO-15919 that could be cited in the notes for this change? Thanks!

gregjay commented 4 months ago

I use as my reference for ISO-15919 use of U+2019 is: https://web.archive.org/web/20160418005422/http://homepage.ntlworld.com/stone-catend/triunltc.htm

Hope that helps

When using a converter like https://aksharamukha.appspot.com

(See his Indic/Roman matrix at https://aksharamukha.appspot.com/roman which conforms to the above linked reference)

Uses U+0027 for the transliteration of Avagraha in IAST system And uses U+2019 for transliteration of Avagraha in ISO system

Your present Indic Romanization from ITRANS keyboard makes no distinction, mapping U+0027 for transliteration of Avagraha for both systems. All I am asking for is correction of that mistake and a simple key sequence to produce U+2019 for use with the ISO system.

Thanks

Greg

P.S. On a different issue the Keyman program doesn't seem to work consistently with Mac OSX Sonoma 14.2.1 Seems like the basic program needs to be updated not just the keyboard above.

On Thu, Feb 8, 2024 at 8:17 AM DavidLRowe @.***> wrote:

IIUC the keyboard already cycles through U+0027 and U+02BC. Perhaps a simpler solution would be to change the U+02BC to U+2019. (As you note, often the two are visually similar or identical in a given font, and it could be a source of confusion to have both.)

Can you point to an online reference to ISO-15919 that could be cited in the notes for this change? Thanks!

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: @.***>

LornaSIL commented 4 months ago

Fixed by #2546

gregjay commented 4 months ago

When will Indic Keyboard for Itrans be updated? After receiving this message stating Fixed I deleted the keyboard then downloaded it again and installed it again. There was no change. Even though the keyboard help file on the net says it has been fixed.

Greg

On Wed, Feb 14, 2024 at 7:30 AM Lorna Evans @.***> wrote:

Fixed by #2546 https://github.com/keymanapp/keyboards/pull/2546

— Reply to this email directly, view it on GitHub https://github.com/keymanapp/keyboards/issues/2539#issuecomment-1944288631, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACBE5Z7HFTV45TMPMSYNZRTYTTYELAVCNFSM6AAAAABC5CYJOCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSNBUGI4DQNRTGE . You are receiving this because you were mentioned.Message ID: @.***>

mcdurdin commented 4 months ago

@gregjay investigating why the keyboard has not yet been published. We just did a major upgrade of our websites so it's probably related to that. Will update once I have some detail.

mcdurdin commented 4 months ago

@gregjay The deployment problem has been resolved and version 1.2.0 should be available now.

gregjay commented 4 months ago

Thank you. I tried the new deployment version 1.2.0 seems great. I'll let you know if I have any further suggestions

Greg

On Wed, Feb 14, 2024 at 4:55 PM Marc Durdin @.***> wrote:

@gregjay https://github.com/gregjay The deployment problem has been resolved and version 1.2.0 should be available now.

— Reply to this email directly, view it on GitHub https://github.com/keymanapp/keyboards/issues/2539#issuecomment-1945285789, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACBE5Z3EE562IZVBT6KWFKTYTV2LBAVCNFSM6AAAAABC5CYJOCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSNBVGI4DKNZYHE . You are receiving this because you were mentioned.Message ID: @.***>