Help to Create a keyboard for IAST and Vedic Accents (A part of ISO 15919)

[Martin Gluckman]

Greetings,

I need a Ukelele expert to help us creat a keyboard for typing IAST (Sanskrit transliteration) as well as the Vedic accents (used for Vedic texts) which are part of ISO 15919. All will be generated by either 2 or 3 keystrokes when the respective mode is enabled for example -a will generate ā as the output and so forth. It is a visual and intuitive system of "drawing" the diacritics.

If you have expereince with Ukelele and would like to help (we can offer payment for this), please PM me.

Thank you!

Martin Gluckman

[Sorin Paliga]

If you do not find an expert in Sanskrit, I may be implied, after Sept. 7 (I will be out at a conference). Meanwhile, I would need a complete list of the chars in vue. For example, if the accents used for the Vedic texts are already included in Combining Diacritical Marks area, than it is quite easy to add them to an existing keyboard layout, at least I hope so. What such keylayouts do you currently use for these languages?

--
You received this message because you are subscribed to the Google Groups "Ukelele Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ukelele-user...@googlegroups.com.
To post to this group, send email to ukelel...@googlegroups.com.
Visit this group at http://groups.google.com/group/ukelele-users.
For more options, visit https://groups.google.com/d/optout.

[Martin Gluckman]

Dear Sorin,

I have the needed knowledge in Sanskrit so I could guide the requirement quite easily, I would provide a spreadsheet with the input keystrokes the user would type (when the specific keyboard we create is enabled) and then the output that would be expected. Then you (or whoever helps) would need to simply get Ukelele to do the needful.

Kindest Wishes,

Martin Gluckman

--
You received this message because you are subscribed to a topic in the Google Groups "Ukelele Users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/ukelele-users/uXTg6yF95fs/unsubscribe.
To unsubscribe from this group and all its topics, send an email to ukelele-user...@googlegroups.com.

[Sorin Paliga]

I have colleagues familiar with Sanskrit, so I may co-operate with them. Perhaps this might be of interest for them too. 

If you do not find someone closer, let us begin a more serious discussion after Sept. 5 when I come back from the conference.

Send the e-mail directly to

sorin[dot]paliga[at]gmail[dot]com

[Geke]

I assume that you have already had a look at Mac standard keyboard layout "US Extended"?
For non-vedic texts I find it quite enough.

If you know: I was wondering how currently is the situation (support in browsers and other software, I mean) with those vedic accents that weren’t part of Unicode until recently, or are still in the pipeline?
Oh, wait, that’s with Devanagari, not transliteration.

I can promise my help already for free – if it won’t take too much time...
Here’s my info page on the topic:
http://geke.yolasite.com/sanskrit-transliteration.php

[Sorin Paliga]

This is correct. As I once explained in the Read-me file accompanying my US Academic, U.S. Extended has two major flaws:

- it does not cover all the needs for transliteration and dialectal texts; CMD’s are not covered, for example, but not only;

- it is not mnemotechnic for many sequences, some are different from the traditional ones. 

Now, of course, the best key layout is the one somebody is accustomed to. I remember I began creating US Academic just because I was not satisfied with the solution included in OS X. Meanwhile, the Unicode Consortium has added some new chars, but I have been lazy to update my key layout, I do not need them for current use, but I will update it one good day.

Any transliteration may be achieved with some patience, it depends on the number of chars and diacriticals used in a text or in a series of texts. This requires some exercise and, for the things to be smooth, some mnemotechnic principles. 

[Martin Gluckman]

Dear Geke,

 

This is the system we have developed (it is working really well on the Pc and we have many student susing it and appreciating its simplicity and not having to learn anything new as it is intuitve).

 

So we need to get this equivalent on the Mac using Ukelele. Help is much appreciated!

 

For IAST this is the mapping:

 

If the user types	Then output	Notes
-a	ā	IAST
-i	ī	IAST
-u	ū	IAST
-r.	ṝ	IAST
r.	ṛ	IAST
-l.	ḹ	IAST
l.	ḷ	IAST
.m	ṁ	IAST
m.	ṃ	IAST
h.	ḥ	IAST
.n	ṅ	IAST
~n	ñ	IAST
t.	ṭ	IAST
d.	ḍ	IAST
n.	ṇ	IAST
's	ś	IAST
s.	ṣ	IAST
\/	√	Root Symbol

 

Then for the Vedic accents in addition to the above we add:

 

If the user types	Then output	Notes
'a	á	Vedic Accents
'-a	ā́	Vedic Accents
'i	í	Vedic Accents
'-i	ī́	Vedic Accents
'u	ú	Vedic Accents
'-u	ū́	Vedic Accents
'r.	ṛ́	Vedic Accents
'-r.	ṝ́	Vedic Accents
'l.	ḷ́	Vedic Accents
'-l.	ḹ́	Vedic Accents
'e	é	Vedic Accents
'o	ó	Vedic Accents
`a	à	Vedic Accents
`-a	ā̀	Vedic Accents
`i	ì	Vedic Accents
`-i	ī̀	Vedic Accents
`u	ù	Vedic Accents
`-u	ū̀	Vedic Accents
`r.	ṛ̀	Vedic Accents
`-r.	ṝ̀	Vedic Accents
`l.	ḷ̀	Vedic Accents
`-l.	ḹ̀	Vedic Accents
`e	è	Vedic Accents
`o	ò	Vedic Accents
a-	a̱	Vedic Accents
-a-	ā̱	Vedic Accents
i-	i̱	Vedic Accents
-i-	ī̱	Vedic Accents
u-	u̱	Vedic Accents
-u-	ū̱	Vedic Accents
r.-	ṛ̱	Vedic Accents
-r.-	ṝ̱	Vedic Accents
l.-	ḷ̱	Vedic Accents
-l.-	ḹ̱	Vedic Accents
e-	e̱	Vedic Accents
o-	o͟	Vedic Accents

 

I understand this needs to be done with the dead keys freature, some like -l.- would need multiple dead keys in a row (is this possible?).

 

Any help to produce these appreciated. I would make two new keyboards:

 

1.US Keyboard + IAST

2.US Keyboard + IAST + Vedic Accents.

 

Kindest Wishes,

 

Martin Gluckman

--
You received this message because you are subscribed to a topic in the Google Groups "Ukelele Users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/ukelele-users/uXTg6yF95fs/unsubscribe.
To unsubscribe from this group and all its topics, send an email to ukelele-user...@googlegroups.com.

[Geke]

Dear Martin,

I'm sorry: for your IAST table alone, your system can probably be done in Ukelele, but certainly not for IAST + vedic accents.
So it may be better to implement both using a different approach, only that is beyond my capacities.

Anyone reading here knows what this would take? Maybe a custom input method?

My curiosity would like to know how you have implemented this in Windows?

In the first sentence, I wrote "probably", because some questions remain, like:
- Should this system work with many fonts, or is one font enough? And: in one program or system-wide? MS Word's "Autocorrect" feature could get you a long way.
- Which keystrokes are to be used for typing the marks? In other words: will the keystrokes for typing the marks - . ~ ' \ / ` be different from the keystrokes for hyphen, full stop, etc.? And if not, how should the system "know" what the user wants to type, e.g., n + full stop or a retroflex n?

Op maandag 31 augustus 2015 19:11:53 UTC+2 schreef Martin Gluckman:

[Martin Gluckman]

Dear Geke,

Take a look at our tool on windows which demos the IAST version of it and the root symbol is included there:

http://www.vedicsociety.org/download/SanskritWriterSetup.exe

We implemented this with AutoHotkey, we are working on a version with the Vedic Accents using the same system and it should work fine on Windows.

In this software you simply toggle modes between SANSKRIT DISABLED | IAST | DEVANAGARI by using the hotkey Ctrl + Alt + S. This is how the system knows what you are typing when in IAST mode -a will yield ā otherwise it will be -a as per usual.

If you have Windows running feel free to give the system a try, it also works in Parallels fine. The core concept behind our method was to create a method of typing Sanskrit without having to relearn any new system so using your already memory mapped knowledge of your keyboard. Then the only rule the person needs to know is to visually draw the characters from top to bottom so thence ṝ is rendered as -r.

We have tested this on a number of students and within 1 minute of explanation they can touch type IAST. We are using this for Sanskrit but this method could be applied for any diacritics in theory.

Our software has a Devanagari output mode too where you type and draw the IAST characters but it outputs in Devanagari.

We have a number of students requesting the same for the Mac hence I am exploring ways of doing this without hardcoding a software. AutoHotkey does not seem to have an equivalent on the Mac and Ukelele seems to be a simple way to implement this but you are saying that Ukelele can not manage cases like ṝ̱ which would be inputed as -r.-

Kindest Wishes,

Martin Gluckman

[Sorin Paliga]

It is absolutely obvious that any similar method should as mnemotechnic as possible, e.g. a dot below r should be somewhat done with either option-R or option-dot. This depends on the number of chars you need for a given language or a set of languages.

You received this message because you are subscribed to the Google Groups "Ukelele Users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ukelele-user...@googlegroups.com.

[Martin Gluckman]

Dear Sorin,

 

In this case the concept of the user being able to “draw” the character mentally helps them to learn the system instantly.

 

 

 

Kindest Wishes,

 

Martin

From: ukelel...@googlegroups.com [mailto:ukelel...@googlegroups.com] On Behalf Of Sorin Paliga

Sent: 01 September 2015 11:22 AM
To: ukelel...@googlegroups.com

[Tom Gewecke]

On Sep 1, 2015, at 1:38 AM, Martin Gluckman wrote:

>
>
> We implemented this with AutoHotkey

There are a number of similar apps for OS X, like Keyboard Maestro. But I have no experience with them and don't know whether you can do the same thing exactly. Perhaps someone else does.

Or Apple's table-driven custom input method:

https://support.apple.com/kb/PH18456

[Geke]

Thanks, Tom!

Would you know where I can find an existing input method for a non-CJK script, that is: with only one output string per input string?
I'd like to see how such an input method works in practice, I mean what exactly happens while you type.

I'm also going to try Keyboard Maestro; it's shareware, looks like.

[Geke]

Thanks for that pointer, Tom, I'll check it out.

In the mean time I tried Keyboard Maestro and it seems to do exactly what is needed.
It took me a few tries to figure out the correct order of combining marks, but as a test I got `-l.  ->  ḹ̀ to work.

How I think it functions: after each keystroke, it adds it to a buffer and checks if the (ending of the) buffer contents match one of the search entries in a two-column list of search-and-replace strings. At a match, some backspaces are output to delete those keystrokes and the replace string is output: one or more characters to replace the search string.

Amazingly, there are almost no cases where this simple approach fails. One of them is r.n – which could either be output as ṛn or rṅ – so I wonder how to handle such things. (I haven't tried the Windows utility yet.)

This may all sound a bit off-topic on the Ukelele forum, but Keyboard Maestro looks like a good solution for all those people asking here for stuff like changing arrow keys or other non-Ukelele features.

[Detlef Schulz]

Hi all,

maybe the small app "aText" is sufficient as well, with only 5 Euro in the AppStore cheaper then other solutions. And you can substitute any text even without separator signs, so you can type away without hitting the space bar all the time to trigger the substitution as with the OS X Text service.

Detlef

--

You received this message because you are subscribed to the Google Groups "Ukelele Users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ukelele-user...@googlegroups.com.

[Geke]

Yes, a quick test seems to indicate that "aText" will do the trick as well.
Thanks, Detlef!

More testing is needed however, also of Keyboard Maestro, e.g. to see if one can share a list of abbreviations with other users.

More suggestions? aText can "import data from TextExpander, TypeIt4Me, SpellCatcherX, Automaton, CSV file". Maybe some candidates there...

Geke

Op woensdag 2 september 2015 17:14:22 UTC+2 schreef Detlef Schulz:

[Martin Gluckman]

Wow! I've just created the functionality in 5 minutes using aText, works perfectly! I've done IAST and will input the Vedic accent typing system too shortly.

Thank you for this suggestion! I will look into how to make this a standalone application with an installer next.

Martin

--
You received this message because you are subscribed to a topic in the Google Groups "Ukelele Users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/ukelele-users/uXTg6yF95fs/unsubscribe.
To unsubscribe from this group and all its topics, send an email to ukelele-user...@googlegroups.com.

[Geke]

Congratulations, Martin!

"Standalone": do you mean installing the program + importing the search+replace array? Maybe you can get the developer to create a custom package for you, once you have the array finalized.

Otherwise installing it is very straightforward from the .dmg file, no?

Op woensdag 2 september 2015 21:06:15 UTC+2 schreef Martin Gluckman: