#158 Cannot enter certain character with built-in IME

[Michael van den Berg]

Discovered and Reported by Oliver Emery (thrymgjol)

#1. IME problem with double entries in input-pinyin.utf-8 --> I fixed this with code, but I could also remove all duplicate entries from input-pinyin.utf-8
#1. Try entering dian or jian using the ime. You will get 1 option with two characters. This is because in input-pinyin.utf8, there is an entry for dian(1 character) and an entry for dian(2 charcaters). The parser records the first one, but overwrites it when it reaches the second.

Originally reported on: http://www.chinese-forums.com/index.php?/topic/44618-zdt-102-released/?p=344725

[Michael van den Berg]

Reproduction:

• Turn on ZDT's built-in IME with Edit->IME Mode or CTRL-I
• Open a Category by double clicking on it
• Add a new Entry with CTRL-N

• Type 'dian'. Only one option shows up:"堤岸". It should show a series of options, such as 点, 电, 店, 垫 or 典.
  It will be virtually impossible for the user to input for example '点'.

• The problem also with 'jian' (only option is '堤岸'), bian (only option '彼岸'), and it is not ruled out whether this are the only ones or not.

[Anonymous]

Fixed and uploaded. We could also modify input-pinyin.utf-8 and remove the double entries.

[Michael van den Berg]

Bug retested and fix verified.

• Turn on Ime Mode (Edit -> IME Mode or CTRL-I)
• Open a Category by double clicking on it.
• Open the Add a new Entry dialog with CTRL-N.
• Typing 'jian' gives as the first four option 间见建 and 件, 'dian' gives 电点店 and典, and bian gives as first three options 变便 and 边.
  Result: fixed, but now it is not possible to enter double characters.

Proposal: approve the current fix for release in v1.0.3. (See discussion point below).

Tested with: rebuild (by kaya) "Flashcard Plug-in version 1.0.3" and "zdt.jar" - replaced these files in respectively 'zdt\plugins\net.sourceforge.zdt.core_1.0.2' and 'zdt\'. (supposedly up to build 1241).

[Michael van den Berg]

A discussion question related to the solution:
Since the double character options have been removed from the results in order to fix this, a user is not able anymore to enter such double character at once (they can do it one by one).
Question: Shouldn't a user be able to choose double entries? Some common double character words might be more common than some other rare single character. Are we actually able to know which one is more common?
And, can our (IME) module actually know which entry is more common? Or is this more of feature request - most people will probably use IME that comes with the PC anyway?

In any case, the current solution is better, as it allows to select the simple characters, which are more common than the complexer double ones, perhaps often only used by expert users, who probably have access to other tools as well.

Proposal: approve the current fix for release in v1.0.3 and see if we can think of a solution.