[RELEASE] PSPedict V0.3 (Japanese Dictionary)

[PixelDragon]

Hi Everyone

Back again with the latest update to PSPedict, now at V0.3.

The biggest change is the inclusion of Tanaka Corpus example sentences.
There are example sentences for approximately 20,500 unique words.

As the Tanaka Corpus is considered a work in progress, included with the
download is a Windows application called "EdictUpdate" which can be
used to update the example sentences in the EDICT files. Further
instructions on how to use this tool can be found by clicking the "About"
button when the application is open. The EDICT files in the download
have already been updated. I have only tested the application on Vista.

A quick disclaimer regarding the example sentences. Please do not expect
each sentence to be a perfect match. Sometimes the sense of the word
is not correct, and other times the Japanese or English sentence may
contian errors. They should only be used as a rough guide or for context.

Download:
Download PSPedict V0.3

I'll release the new source code when I clean it up a little.

Links to previous version threads:
[RELEASE] PSPedict V0.2
[RELEASE] PSPedict V0.1

Changelog for V0.3:
- Included support for Tanaka Corpus example sentences.
- Fixed a bug that cut some English description strings short.
- The last English search string will be the default text in the
On Screen Keyboard(OSK). I was unable to impliment this for
Japanese text. If anyone has luck with this, I'd appreciate an
e-mail/post about how to impliment this.
- Included an application file "Edict Update". This can be used
by the user to update Edict files if the Tanaka Corpus examples
are updated.

- V0.2a (02/11/09)
- Fixed a bug where if you had an empty Study List or Found Set and you
pressed the 'X' button, the PSP hangs. Thanks to QJ.net forum member
batanen for reporting this bug.


Installation
============
See the readme in the download file for full information.
Simply copy the contents of the folder "PSPedict" to the
game folder of your psp, e.g. "X:\PSP\GAME" where X: is
the drive letter for your PSP. If you are asked to overwrite
any files, select yes/yes to all.

Please note, installing a new version will not overwrite your
saved found sets or study lists.


Todo list
=========
- More accurately incorporate the example sentences and support the more recent
Edict format that includes word sense numbers.
- I may look at supporting UTF-8 strings as a number of people have wanted to create
their own Edict files in languages that use extended ASCII/special characters.

Jim Breen & Monash University for EDICT, KANJIDIC and modified Tanaka Corpus.
WWWJDIC: Word Search
Brunni for OSLib. home - OldSchool Library
Sakya for OSLib Mod. Sakya's Homepage
BenHur for Intrafont. intraFont and other PSP Releases by BenHur
DAX for consistently pumping out quality CFW. Legend.
Lastly, anyone who has downloaded, supported, given feedback, etc...

Pictures:








As always, comments/suggestions welcomed.

Hope you enjoy this :)

[batanen]

Truly an amazing update! Nice that you were able to implement the Tanaka Corpus and solve the memory issue for that. The input-remembering functionality was great, though of course for the Japanese it would have been the most needed, but as you explained it was not possible at the moment. Other than that I don't know what I could suggest anymore (than I have mentioned before). PSPedict works so well. There aren't any major funcionalities that would be essential. However, if you're interested in trying something new, I can think of a couple of things.

First is of course the memory issue you're aware of with the large files (Enamdict). Then a text zooming function with the unused arrow buttons (font size decrease)<- ->(font size increase). By zooming one could see the more complex kanjis more preciesly (and by doing this PSPedict increases the accessibility for the visually impaired users; yes, I am in an accessibility course in this period.).

Thank you very much for the latest release!

[RyanF]

Thanks for the update! Really useful app

[batanen]

Hi again! I am not sure but I might have a solution for the Japanese search string. I just happened to read a C++ study book and there was just mentioned a way to store 16-bit characters. Instead of a string you should use a wstring (wcin, wcout). I don't know how you have stored the English string as I cannot see the source code yet. Maybe wstring is a worth to look into? I have never used it though, just learned it exists and for what it is used for.

[PixelDragon]

Hi batanen.
Sorry for the late reply. I previously looked at the source code for the OSK used in OSLib Mod and noticed that it uses the type of unsigned short (2 bytes) for the initial text string and the output (returned) string. I have assumed that internally the Japanese character string is possibly in UCS2 format (each character represented by 2 bytes). This is the format the returned output string is in. If I try to use the exact same string that was previously returned as the new initial text string, the OSK does show some text, but it is not in Japanese. I know the returned text is in UCS2, because I later convert it from UCS2 to SJIS. To be honest, I'm not good enough to understand why it doesn't work. Hopefully I'll stumble across an answer sometime. Thanks for trying.

[Hell-Raiser-85]

Thanks for the update PixelDragon

Great App

[batanen]

Hi batanen.
Sorry for the late reply. I previously looked at the source code for the OSK used in OSLib Mod and noticed that it uses the type of unsigned short (2 bytes) for the initial text string and the output (returned) string. I have assumed that internally the Japanese character string is possibly in UCS2 format (each character represented by 2 bytes). This is the format the returned output string is in. If I try to use the exact same string that was previously returned as the new initial text string, the OSK does show some text, but it is not in Japanese. I know the returned text is in UCS2, because I later convert it from UCS2 to SJIS. To be honest, I'm not good enough to understand why it doesn't work. Hopefully I'll stumble across an answer sometime. Thanks for trying.

That's ok. Thank you for replying!

Hmm... that sounds problematic. As it's quite clear, I don't know anything about PSP programming (once I compiled "hello world", though). However, there is still one method I can think of. Would converting the UCS2-format to SJIS internally work? I can only assume, but if that's the case, maybe intraFont could help. If I understood right, the sample-file had an example how the encoding can be changed to SJIS. EDIT: Apparently you knew about the Intrafont already. Nevermind then.

Merry Christmas holidays!

[PixelDragon]

Hi batanen

Thanks for trying again.
Have a Merry Christmas and a good holiday too :)

[batanen]

Hi PixelDragon!

I have been trying to contact you for a few days, but haven't yet been able to hear about you. I would be interested in having the latest source code. It would be very nice if you could upload it. I did already check the V0.2 code. Almost 8000 lines in one file is a little difficult to comprehend, but maybe I can improve PSPedict a little bit if you're not wanting to develop it anymore.

By the way, the code for V0.2 does not compile because oslCccStrlenUCS2-function seems not to exist (tried with the latest Oslib Mod 1.1.1). I did comment this function and wrote length = *instr * 2, but the results of course we're a little messy. I wish the source code for V0.3 will compile without errors (I guess warnings can't be avoided...).

In the case you are still interested in working with this I can share my ideas:

- Unlimited number of dictionary files
* At start-up fetch every file from dictionary folder (needed if you want to study from your own dictionary-files).

- Improved study lists
* Support for multiple study lists.
* A possibility to hide/show kana/translation one by one or all at once (for studying the words both ways)
* Randomize study list

- Stroke Count & SKIP-code search with Japanese input
* If type of string is int and size = 1, show all kanjis that has a stroke count of the given int.
* When search contains numbers this way <type of pattern> <stroke count 1> <stroke count 2>, use SKIP search. SKIP-codes have four different patterns. If given pattern is 0, show results with all patterns.

- Kanji Drawing
* Easiest way would probably be trying to implement it similarly as in pspkanji (latest source should be available in SourceForge).

(Sorry Qj, if this is too old thread to reply. Then what is the time limit?)