kokoye2007 / waitzar

Automatically exported from code.google.com/p/waitzar
Other
0 stars 1 forks source link

Fixing Ayar inconsistencies #109

Closed GoogleCodeExporter closed 9 years ago

GoogleCodeExporter commented 9 years ago
#WZ 1.8 demo

Besides the easy-to-spot errors (@track_capslock not supported, etc.) the
following inconsistencies exist in the first Ayar keyboard:

1) Typing "0-" will give us "cannot match on a rule that is already done"
2) Shifted number keys don't work.
3) Typing "u`E" is required ("u`e" will just stack another letter...). Need
to check the source for this.
4) Typing "uU<back>" works, but "u`u<back>" leaves the "`". (I think
KeyMagic assumes that some letters are not combined when typed...)
5) Typing "," and "." is impossible with WZ's forwarding.

This is the first batch of errors; I'll have more to report on after fixing
this round.

Original issue reported on code.google.com by seth.h...@gmail.com on 11 Apr 2010 at 3:37

GoogleCodeExporter commented 9 years ago
Fixed 5.

Original comment by seth.h...@gmail.com on 11 Apr 2010 at 9:53

GoogleCodeExporter commented 9 years ago
Fixed 1, and a side problem with switches.

Original comment by seth.h...@gmail.com on 11 Apr 2010 at 4:19

GoogleCodeExporter commented 9 years ago
Manually entering the ayar_tests.txt file, here are the ones that fail, in 
Zawgyi.
NOTE: There will be additional errors when we try to do these tests 
automatically;
this bug isn't concerned with those (since they'll trigger a warning.)

General errors:
"uiF" => "ကငၤ"   actual: "ကၤင"   //This is a general problem with 
"kinzi" in Ayar4WZ
"!" => "ဋ"       actual: (lakaung)
"@" => "ဍ"       actual: "႒"
"#" => "ဎ"       actual: "~"
")" => "၀"       actual: "ဧ"

Keys not supported?
"[" => "ဟ"      actual: "ဩ"
"{" => "ဧ"       actual: U+86
"]" => "ဉ"       actual: U+96
"}" => "ဦ"       actual: U+87
"\\" => "ဿ"     actual: U+85
"|" => "၎င်း"    actual: U+88
";" => "း"        actual: U+89
"'" => "ဒ"       actual: U+8A
"\"" => "ဓ"      actual: U+8C
"/" => "။"        actual: U+92
"?" => "၊"        actual: U+93
"u<" => "က္ဓ"     actual: U+1000 U+94
"u>" => "က္ဆ"    actual: U+1000 U+95
"_" => "ြသ"       actual: U+8F
"+" => "ေြသာ်"     actual: U+9F
"$" => "ဣ"      actual: U+7F
"%" => "ဈ"       actual: U+80
"^" => "/"       actual: U+92
"&" => "ရ"       actual: U+82
"*" => "ဂ"       actual: U+83

Summary: For a lot of these the hotkey just isn't triggered right. There are 
some
ordering issues with kinzi (and others) that are probably more related to our
converter. Since there are so many problems anyway, I'm considering redoing the
keypress API (it's incorrect at best.)

Original comment by seth.h...@gmail.com on 4 Jun 2010 at 12:50

GoogleCodeExporter commented 9 years ago
Note to self: All hotkeys are properly registered. So it's all internal. 

Original comment by seth.h...@gmail.com on 4 Jun 2010 at 12:56

GoogleCodeExporter commented 9 years ago
Gah, I was converting from "Ayar" to "Unicode" improperly. Will re-run tests 
later, after trying to fix a few of the symbol-based errors.

Original comment by seth.h...@gmail.com on 14 Jun 2010 at 6:13

GoogleCodeExporter commented 9 years ago
Note: Some of the above tests are wrong: 
! => U100B (ဋ), NOT lakaung. Will re-check...

Original comment by seth.h...@gmail.com on 14 Jun 2010 at 6:17

GoogleCodeExporter commented 9 years ago
Some updates. All errors stated above fixed except.

//These errors remain, but are actually part of the Ayar.kms file. 
( => ''     Actual: ဦ
} => 'ဦ'    Actual: '႒'

Original comment by seth.h...@gmail.com on 14 Jun 2010 at 9:02

GoogleCodeExporter commented 9 years ago
Also:

% => ဈ

Should we output \u1005\u103B or \u1008? I prefer \u1005\u103B, but let's check 
with the Ayar team.

Original comment by seth.h...@gmail.com on 14 Jun 2010 at 10:14

GoogleCodeExporter commented 9 years ago
Also:

) => ၀

Actual output is "wa" (101D) but I think 1040 is better. Again, have to check.

Original comment by seth.h...@gmail.com on 14 Jun 2010 at 10:16

GoogleCodeExporter commented 9 years ago
In addition:

ဩ & ဪ are output as special characters... but why should we do this? To 
allow users to output "composed" versions? That might make sense, but I'll 
check with the Ayar team anyway.

Original comment by seth.h...@gmail.com on 14 Jun 2010 at 10:17

GoogleCodeExporter commented 9 years ago
(Before we go to the Ayar team, I have to fix the "auto-test cases" failing on 
"K", etc. Not sure why this is....

Original comment by seth.h...@gmail.com on 14 Jun 2010 at 10:38

GoogleCodeExporter commented 9 years ago
All except these fail; see above for justification. Auto-tests work too:

(in Zawgyi encoding)

Test results
-----------
} => ဦ
   Actual result: ႒
% => ဈ
   Actual result: ဈ
) => ဝ
   Actual result: ဝ
_ => ျသ
   Actual result: ဩ
+ => ေျသာ္
   Actual result: ဪ

Original comment by seth.h...@gmail.com on 15 Jun 2010 at 1:50

GoogleCodeExporter commented 9 years ago
All tests passed.

Original comment by seth.h...@gmail.com on 27 Sep 2010 at 2:58