Braille application.

cvp

This list is correct.

I'm surprised, for instance,
are you're sure that ぎ is

I suppose that the list of existing dots is ok but the generated image would be different...

cvp

@shinya.ta Please confirm that this list and their graphical representation is correct
the dots are prefix 5 or 6 | normal dots

か が 5|16
き ぎ 5|126
く ぐ 5|146
け げ 5|1246
こ ご 5|246
さ ざ 5|156
し じ 5|1256
す ず 5|1456
せ ぜ 5|12456
そ ぞ 5|2456
た だ 5|135
ち ぢ 5|1235
つ づ 5|1345
て で 5|12345
と ど 5|2345
は ば 5|136
ひ び 5|1236
ふ ぶ 5|1346
へ べ 5|12346
ほ ぼ 5|2346
は ぱ 6|136
ひ ぴ 6|1236
ふ ぷ 6|1346
へ ぺ 6|12346
ほ ぽ 6|2346

for those who would be intellectually interested, I have generated that by a little script,
after having discovered that the only difference between this 6-dots prefix is the last half-byte to which is added 1 or a 2 (ex: x8b -> x8c)

def p(dots,prefix,ch):											# ex: 5,か
	d = 1 if prefix == '5' else 2				# ex: 5 -> 1           6 -> 2
	b = ch.encode('utf-8')							# ex: b'\xe3\x81\x8b'
	n = b[:-1] + bytes([int(b[-1])+d])	# ex: b'\xe3\x81\x8c'
	c = str(n,'utf-8')									# ex: か -> が          へ -> ぺ 
	print(ch,c,dots)
	
for ele in Japanese_Braille.keys():
	if '|' in ele:											# ex: 5|1345
		prefix = ele[0]										# ex: 5
		k = ele[2:]												# ex:	1345
		ch = Japanese_Braille[k]					# ex: つ
		p(ele,prefix,ch)

cvp

New Version 0.7

# Version 0.7
# - bug corrected: sqlite3 operationalerror accessing HiraganaToKanji.db
#                  was not intercepted
# - new: cursor left and right move buttons
# - new: support Hirgana with one/dot prefix for (han)dakuten

To input characters with dakuten or handakuten like が or ぴ

tap the prefix dot 5 or 6
tap ✅ to say that the dots are ended
the program recognizes this as a prefix and waits for a second dots set
tap all dots of normal character like 1236 for ひ
tap ✅
the accentuated hirgana is generated like ぴ

This version does not yet support speech, for this, I wait that you explain when and what has to be spoken...ex: the Kanji when it is sent to the TextField

Perhaps, one day 😀,you will also ask me to support Yō-on which is supported in Japanese Braille

shinya.ta

@cvp

I took a test.
It's great.

And I'd like to have a text field and a speech on the conversion list.

cvp

@shinya.ta Finish for today, I have a dinner outside.
Tomorrow, mothers' Day, thus not a lot free
Monday should be ok to go with the program, sorry for the delay

Did you check the words with prefix?
Did you test with custom keyboard in another app?

shinya.ta

@cvp

I tested the prefix.

It functioned normally.
The test for other applications is not yet done.
Just a moment, please.

shinya.ta

@cvp

http://www.yoihari.com/tenji/tyou.htm

http://www.yoihari.com/tenji/tsuji.htm

http://www.yoihari.com/tenji/tkigo.htm

This Braille is also necessary.
There are some other details, but I don't remember well, so the necessary Braille is only this.

cvp

@shinya.ta It becomes really complicated, let me some time to try to understand...

cvp

@shinya.ta The Yō-on table that you link is a katakana list, not hirgana

For instance, are you sur that you want キャ instead of きゃ

It is important to decide before I begin this very complex part...

shinya.ta

@cvp

The list is displayed in Katakana, but there is no distinction between Hiragana and Katakana.
Because, for visually-impaired people, the distinction between Katakana and Hiragana is irrelevant to reading a document.

http://www.yoihari.com/tenji/omotegazo.html

cvp

@shinya.ta Ok, understood, but for the display, what do you want? Because I have to display it to generate a voice

cvp

@shinya.ta Digits process is also very complex, due to this rule

Words immediately follow numbers, unless they begin with a vowel or with r-. 
Because the syllables a i u e o and ra ri ru re ro are homographic with the digits 0–9, 
a hyphen is inserted to separate them.
 Thus 6人 "six people" (6 nin) is written without a hyphen, ⠼⠋⠇⠴ ⟨6nin⟩, 
but 6円 "six yen" (6 en) is written with a hyphen, ⠼⠋⠤⠋⠴ ⟨6-en⟩, 
because ⠼⠋⠋⠴ would be read as ⟨66n⟩.

Thus I'll need some time to program that...

All of this would normally be done next week as I have a feast today...

I hope that you understand that this kind of program is not written very quickly 😀

shinya.ta

@cvp

I think you can put Katakana into the conversion list of Kanji.

I won't rush.
Slowly, please.

cvp

@shinya.ta I have a question for digits.

The hyphen that needs to be inserted in some cases after a digit, is it your wife who has to do it?
If yes, the program does not need to have this special process, but the blind user?
Am I right ?

shinya.ta

@cvp

I use it sometimes.
I'll use the figures.

cvp

@shinya.ta New version 0.8

# Version 0.8
# - new: support Yō-on, displayed as Katakana
# - new: support arab digits with special process:
#				 prefix is valable until next char is not a digit 
#          or an hyphen which is not displayed.
#				 But the following process has to be done by user:
#         "Words immediately follow numbers, unless they begin with a vowel or 
#          with r-. 
#          Because the syllables a i u e o and ra ri ru re ro are homographic
#          with the digits 0–9, an hyphen is inserted to separate them.
#          Thus 6人 "six people" (6 nin) is written w/o hyphen'' ⠼⠋⠇⠴ ⟨6nin)
#          but 6円 "six yen" (6 en) is written with a hyphen, ⠼⠋⠤⠋⠴ ⟨6-en⟩,
#          because ⠼⠋⠋⠴ would be read as ⟨66n⟩."
#  - new: support Japanese punctuation of one single dots set: -。？！、・

This version does not yet support:

complex punctuation with start and end characters, like （・・・）, 「・・・」
- I don't understand all in http://www.yoihari.com/tenji/tkigo.htm
speech

I also have some questions:

how do you want your wife inserts a space (blank)?
how does your wife scroll the list of possible conversions?
- and thus how speech the content of the list
do you want the speech functionality only on
- the selected Kanji inserted in the TextField

shinya.ta

@cvp

If you swipe with three fingers, the conversion list will scroll.
So, the conversion list is more likely to react if you display it a little bigger.

My wife doesn't use much space when making mail.

The voice recognition function requires not only the text field but also the voice guidance of the Chinese character list.

shinya.ta

@cvp

I'm sorry.
It's a mistake in kanji.

Chinese.→kanji

cvp

@shinya.ta two questions:

scroll list with 3 fingers swipe, do I need to program that or is that standard with VoiceOver?
Edit: just discovered that this swipe is a standard VoiceOver feature
you want speech the Kanji's list, I don't understand how to do that, because as soon a Kanji is selected, it is going to the TextField. Which element of the list could I speech?

For instance, I could try to say (speech) the first visible item on the list.
Then when you scroll one line, the first will change and would be said.
Of course, I should need to add some blank lines on the list so even the last one the first visible... Hoping I'm clear enough.

If your wife is able to see all buttons like delete, close, ok, conversion,
I could add 3 buttons at right of list,

down one in the list, program would down one the list and say the element
up one in the list, program would up one in the list and say the element
select current, program would select the element and send it to the TextField

Please, discuss it with your wife and give me a feedback

This figure is not yet programmed but only to show how it could be, also bigger font for the list

shinya.ta

@cvp

The button layout like the one in the example is the best.

I hope there will be an explanation when you scroll a line.

My wife knows the position of the button properly.

It is exactly approaching the ideal keyboard.