Messages

Yeah, now it looks alright

The description is OK, but you have problem with code-page. Extra chars are not displayed correctly. You need to use HTML entity (&something;).

F.e. ñ ї

The classes look alright.

There is no point license information in Czech language, but there is difficulty that I have description of that 30 (or littler over) dictionaries in Czech and there is no time to retype it into English at the moment.

Now, I can write licencions are totaly free expect Český tezaurus, Česko-brněnský slovník, Česko-valašský slovník, Česko-staroslověnský and Česko-slezský slovník.

No, I have done no changes. This is still the latest version.

I checked licensions of dictionaries, but I have no time to translate information about dictionaries from Czech into English. Now, I can write notes about licensions into Czech text on my web and I think it is has zerro value for you.

I would be wondered if there was someday a possibility to choose defined letters from the box like in the map of chars known from windows. Nowadays, if people has dictionary with non-latin font, he can use transcription or somehow try to type these letters on mobile and it is not always possible.

Probably it would be needed to use bitmap font all the time and f.e. by pressing '*' show text box with defined groups of letters or can assign to keys of mobile other characters.

[T]

Well, there are many differents standard. For example, look at wikipedia.
//Now I see that properly name of first transcription standard is GOST 1971.

It seems that NormationRus does only filtration of interpunctions and spaces. Also it seems that NormationNor and NormationEng does nothing and NormationEng2 is only duplicate of NormationLat with lots of missing special letters.

Well, I can comment only classes what I have done. (I didn't made class NormationRus.)

NormationRus2 allows you to search words both in azbuka and latin transcription (according to the GOSP 1971 - but yards are 'x' and there are used no apostrophes).
NormationRusC allows you to search words both in azbuka and latin transcription (according to the Czech ISO norm - but yards are 'x', there are used no diacritics and ë is transcripted as 'jo').

NormationUkr allows you to search words both in azbuka and latin transcription (according to the GOSP 1971 - but yards are 'x' and there are used no apostrophes).
NormationUkrC allows you to search words both in azbuka and latin transcription (according to the Czech ISO norm - but yards are 'x', there are used no diacritics, ë is transcripted as 'jo' and ï as 'ji'.).

NormationLat is only upgrades today's version of it.
Czech translation was also only improved.

You can write full my name, but I don't want to have posted there my e-mail because of spam. If it is possible to make PHP form without visible e-mail (or hidden in source code visible at client), so do it (I have solved e-mail forms in this way). Otherwise, you can use link to the e-mail form on the my webpage, where are posted newly generated DfMs' dictionaries (http://java-ap.wz.cz/index.php?str=mail).

Well, I've uploaded JAR archive with source code of new /edited files.

NormationLat: edited; extended character set
NormationRus2 & NormationUkr: new; typing azbuka by transcription into latin (+/- according the norm GOSP 1971, but no apostrophes and the yards are replaced by 'x')
NormationRusC & NormationUkr2: new; also typing azbuka by transcription into latin but according to +/- Czech ISO norm (letters are without diacritics, yards are replaced by 'x', ë is replaced by 'jo')
+ enhaced translation of environment of application into Czech language.

Czech language: tested; NormationLat, NormationRusC & NormationUkrC: tested; NormationRus2 & NormationUkr: tested in previous version (there was made no changes)

Link for files: http://java-ap.wz.cz/V7.ZIP

P.S. ZIP archive also contains empty JAR file.

Thank you for classes. Now I can say the classes work.

Now I have tried again compile in Netbeans and now by miracle I can compile without any errors. And I researched that single compiled classes don't work (even that from you-I get java-lang-thrown-error). But if I compile whole project, the java classes work.

Working example:
http://java-ap.wz.cz/MobileApplication1.jar

F.e. You can try search 'алкоголь' by typing 'alkogolx'.

Now I again try to build project with addition of dots and twice used strings.

EDIT: I hope that now I have fully working release of classes. The searching of dots assumes impossible. So the yards are replaced by 'x' again.

Of course, it is no problem. I will test 'em in afternoon.

I'm sorry. Althought past release of my classes worked, this release doesn't work because of that:

- error in NormationLat: founded, corrected (non-updated name of newly added string function)
- errors in NormationRus2 & NormationUkr: removed controversial (error occuring) parts

I hope now it will be working.

Link: http://java-ap.wz.cz/NormationV3.zip

//In NormationRus2 & NormationUkr there were error somewhere in this part of script:

Code Select Expand


...
else if (defaultNormatedWord.charAt(charPos) == 'й') {
normatedWord.append("y");
}
...
else if (defaultNormatedWord.charAt(charPos) == 'ъ') {
normatedWord.append(".");
}
else if (defaultNormatedWord.charAt(charPos) == 'ы') {
normatedWord.append("y");
}
else if (defaultNormatedWord.charAt(charPos) == 'ь') {
normatedWord.append("."); //maybe needed backslash?
}
...
Now, there are not the same strings and no interpunction.

[So if only you (or anyone else) were so kind and once more you recompiled the project.]

Well, my fault. However, it is very simple:

I sent you classes NormationLat v1.0, NormationRus2 v1.0 and NormationUkr v1.0. These I have compiled from you.

I tested them on real dictionaries by way you have described and I detected that they worked but with some objections. So I had to make edited version v2.0.

There is the problem. Now I have again only the *.JAVA files and I am not able compile the project, althought I can edit existed Normation or make new one. This is not difficult, this only logic. But I am not java programmer and I tried many ways how to compile the project in Netbeans and cmd.exe and every time it fails because some type of error. These programms are not intuitive enough for me, I think.
--> So if only you (or anyone else) were so kind and once more you recompiled the project. When I have these three *.CLASS files, I can finish it.
Link for files: http://java-ap.wz.cz/NormationV2.zip

One half of dictionaries declared as BETA are waiting for new Normation classes, because its contained languages have not fully suported alphabet. For example when I was testing submited classes with CzechTurkish dictionary I found out that in Turkish alphabet i without dot undet it is missing even in my edited NormationLat.

I know that it is possible to use DfM with non-open source dictionaries. But I have to think about used sources and read carefully licences to define which dictionaries were made from fully open sources and can be added here.

Also, I have some new dictionaries. For now they are posted at http://java-ap.wz.cz/index.php?str=mydfm. Now I have to validate if all of them are from really open-source sources and to wait at almost one half when will be added new version of DfM with new Normation classes written by me. They are need 'cause application cannnot find without them correct word in those languages that their character set is not fully supported now.

//I hope it is written understandable enough because I don't have excellent English.