| Author |
 Topic  |
|
n/a
deleted
593 Posts |
 Posted - 22 May 2002 : 17:55:26
|
Am not sure how to do this testing with Japanese (more because of my lack of understanding of codes and setting up a progam to execute...yeah, I need something like "123 for Dummies")...
Anycase, intuitively I see some problems with a test string for Japanese:
strLangTest = "There [%1]is|are[/%1] %s1% [%1]topic|topics[/%1] in %s2% [%2]forum|forums[/%2]."
1) There is no singular/plural distinction for verbs in Japanese...almost none and often verbs are not used - thus no clear way to distinctively express is|are
2) Same with some nouns like topic|tpics - same in JA chars... and only difference is when it is written in katakana (as done often for computer jargons, imported words)where you literally represent phonetically each char into katakana, thus, a difference "s" at the end is shown in plural. So is the case with forum/forums....
3)Like Turkish case mentioned above, some plurals are expressed with exact number associated or with "many""few""little", etc. a range expression... so typically with an expression like n forum for n forums, etc.
You may have similar issues with Chinese, though not exactly the same as in Japanese...not sure of Korean. I suspect there would be a slightly different set of rules for Far Eastern languages....
Sorry if this is too general...
LEORAT
i2AsiaFORUM
Edited by - LeoRat on 22 May 2002 18:24:46< |
 |
|
|
Deleted
deleted
4116 Posts |
Posted - 22 May 2002 : 18:17:25
|
quote:
Am not sure how to do this testing with Japanese (more because of my lack of understanding of codes and setting up a progam to execute...yeah, I need something like "123 for Dummies")...
Anycase, intuitively I see some problems with a test string for Japanese:
strLangTest = "There [%1]is|are[/%1] %s1% [%1]topic|topics[/%1] in %s2% [%2]forum|forums[/%2]."
1) There is no singular/plural distinction for verbs in Japanese...almost none and often verbs are not used - thus no clear way to distinctively express is|are
2) Same with some nouns like topic|tpics - same in JA chars... and only difference is when it is written in katakana (as done often for computer jargons, imported words)where you literally represent phonetically each char into katakana, thus, a difference "s" at the end is shown in plural. So is the case with forum/forums....
LEORAT
i2AsiaFORUM
Leorat, because a similar code will be in the base code, you will only play with the language string. That means you will use the it one time only like:
strLangTest = "There [%1]is|is[/%1] %s1% [%1]トピックス|トピックス[/%1] in %s2% [%2]フォーラム|フォーラム[/%2]."
(I don't know how you say the other words  )
Think Pink
==> Start Internationalization Here< |
|
|
|
n/a
deleted
593 Posts |
Posted - 22 May 2002 : 18:42:20
|
quote:
Leorat, because a similar code will be in the base code, you will only play with the language string. That means you will use the it one time only like:
strLangTest = "There [%1]is|is[/%1] %s1% [%1]トピックス|トピックス[/%1] in %s2% [%2]フォーラム|フォーラム[/%2]."
(I don't know how you say the other words )
Think Pink
==> Start Internationalization Here
You are right on the money. There is no way to express verbs by themselves. As mentioned, it will be something like this:
strLangTest = "There [%1]is|is[/%1] %s1% [%1]トピック|トピックス[/%1] in %s2% [%2]フォーラム|フォーラム[/%2]."
Notice Topic=トピック, Topics=トピックス, and Forum(s) will be the same in JA. But, in a formal Japanese expression, Topic is 課題 or 話題, depending on context it is used, and no plural expresion unless you use 2, many, little, etc. either number or range expression with it, like you mentioned about Turkish. eg: All Forums = 全てのフォーラム or more formally 全フォーラム. Forum has its own Japanese characters - 2 char combination as well, and same principle applies as Topic. So, again, it depends on how words/terminologies used... in case of Lang1041, many of expressions are left in Katakana because of this nuiance/contextual interpretation issue.
(Now getting a bit of headache....)
LEORAT
i2AsiaFORUM
< |
|
|
|
Deleted
deleted
4116 Posts |
Posted - 22 May 2002 : 18:53:01
|
Ehm. Is the context not given in our case? We are not saying "There are too many topics in few forums" at all .
Also the "formality" of the language should be same in the translation, I think. In German files, there are "du" and "Sie" versions. "Sie" is the "polite" (therefore formal?) version.
If somebody forms a "metalica" site, he/she would not like our translations, and would perhaps like to go to the language file and replace "posted" with "****ed" . This would be the "heavy" version.
Think Pink
==> Start Internationalization Here< |
|
|
|
n/a
deleted
593 Posts |
Posted - 22 May 2002 : 23:58:03
|
Well, my old Pentium machine/Win98 w/PSW has totally gone nuts today and crashed....I was hoping to run a little test on that to see how it goes, but alas, no more. (Probably going into a Trash. Anyway, maybe you can test/try it with the following (again UTF8 format)
strLangTest = "There [%1]܂|܂[/%1] %s1% [%1]gsbN|gsbNX[/%1] in %s2% [%2]tH[|tH[[/%2]."
strLangTest = "There [%1]܂|܂[/%1] %s1% [%1]b|b[/%1] in %s2% [%2]tH[|tH[[/%2]."
for testing purposes, "a number of" expression for "are" is used so that it an be expressed in X number exists. If this testing supposedly generate a sentense like "There are too many topics in few forums"
then, a word "ga" needs to be added after a subject/object, such as
NN topics exists in XX forums
or 1: mm̃gsbNXww̃tH[ɂ܂
or 2: XX̃tH[ɂmm̃gsbNX܂
where in
1: NN Topics in XX Forums exist
2: XX Forums have NN Topics
Most likely for Japanese, "Normal" "Ordinary" expression is much more suited..but there are more "contempoary" expressions to do with computers/internet...so there are lots of "general guideline" issue to be looked at for translation. One of the issues I encountered with Lang1041 is not so much with this issue but style issues related to something like the above example. Original translator used certain styles and some mixed expressions of ordinary/polite/contempoary etc. besides some small but important usage of word like "ga", "wa", "ha" etc. goes with subject...
Anyway, if you can try with the above test sample string, please do.
(Sorry that I don't have a facility to do it well).
(woops... JA chars showing properly in Shift-JIS (IE) here....... I better check what I am doing with NotePad2002).
@
LEORAT
i2AsiaFORUM
Edited by - LeoRat on 23 May 2002 00:01:54< |
|
|
|
Deleted
deleted
4116 Posts |
Posted - 01 October 2002 : 18:52:44
|
I recently e-mailed survivor about this and informed him about the current 9 LangSP calls. It seems that there is no solution for Russian  because there are too many special cases related to nouns (yeah, not verbs).
It seems that it will be best to use simpler language constructs like
Topic(s): %s1.
instead of using
There [%1]is|are[/%1] currently %s1% [%1]topic|topics[/%1].
Any idea on this?
PS: According to this resource there are 7.5 million Internet users at the end of year 2000 (15th on the earth). A more recent resource (Hmm. CIA again) claims that there are 9.2 million users (going up to 9th place).
< |
 Stop the WAR! |
|
|
|
n/a
deleted
593 Posts |
|
|
GauravBhabu
Advanced Member
4288 Posts |
Posted - 02 October 2002 : 10:09:42
|
Try this function:
function SingularPlural(strlangTest, strNumerals)
dim arrlangTest, arrNumerals, strTemp, arrTemp, strSP, arrSP, i, j, k
arrlangTest = split(strlangTest, "%||%")
arrNumerals = split(strNumerals, "||")
for i = 0 to ubound(arrlangTest)
strTemp = arrlangTest(i)
arrTemp = split(strTemp, "%sp%")
k = 1
if arrNumerals(i) = 1 then k = 0
for j = 0 to ubound(arrTemp)
if Instr(arrTemp(j),"|") then
arrSP = split(arrTemp(j), "|")
strSP = arrSP(k)
arrTemp(j) = strSP
end if
next
strTemp = Join(arrTemp)
strTemp = Replace(strTemp, "%N%", arrNumerals(i))
arrlangTest(i) = strTemp
next
strOut = Join(arrlangTest)
Response.Write strOut
end function
And call it as below
strNumerals = "2||1"
strLangTest = "there %sp%is|are%sp% %N% %sp%topic|topics%sp% in %||% %N% %sp%forum|forums%sp%"
Call SingularPlural(strLangTest, strNumerals) < |
Edited by - GauravBhabu on 02 October 2002 10:10:42 |
|
|
|
Deleted
deleted
4116 Posts |
Posted - 02 October 2002 : 10:42:47
|
GB, the problem is that we cannot find a way to write an algorithm for Russian. This is the info I've got from survivor some time ago:
quote:
English version works fine, but in russian not such well... Russian language have some form of words in both singular and plural that must be dealed separately:
strLangTest = "There [%1]is|are[/%1] %s1% [%1]topic|topics[/%1] in %s2% [%2]forum|forums[/%2]." - original
strLangTest = " %s1% [%1]|[/%1] %s2% [%2]|[/%2]." - in russian
for "topic/topics" in russian
When number ended with 1 (example: 1, 21, 51, etc) it must be writed as "" (in russian)
When number ended with 2 to 4 (example: 2, 23, 54, etc) it must be writed as ""
When number ended with 5 to 10 (example: 0, 5, 27, 60, etc) it must be writed as ""
for "forum/forums"
When number ended with 1 (example: 1, 21, 51, etc) it must be writed as "" (in russian)
When number ended with 2 to 4 (example: 2, 23, 54, etc) it must be writed as ""
When number ended with 5 to 10 and from 11 to 20 (example: 0, 5, 27, 60, etc) it must be writed as ""
So... your code doesn't work with russian ...
The problem is with the nouns, the rule differs from noun to noun...
< |
Stop the WAR! |
|
|
|
GauravBhabu
Advanced Member
4288 Posts |
Posted - 02 October 2002 : 11:38:57
|
| I will try to work that out. However what do you think about the function posted by me. It gives the same output as the function currently being tested. How the two functions compare performance wise.< |
|
|
|
GauravBhabu
Advanced Member
4288 Posts |
Posted - 02 October 2002 : 11:44:04
|
| So incase of russian we have to choose from three strings instead of one< |
|
|
|
pweighill
Junior Member
United Kingdom
453 Posts |
Posted - 02 October 2002 : 12:02:50
|
Is it possible to use the function as is, and for Russian, use a generic phrase which would pass through the function without change.
e.g There are Topic(s) %1 in %2 Forum(s) - (Obviously in Russian not English)< |
|
|
|
Deleted
deleted
4116 Posts |
Posted - 02 October 2002 : 12:08:31
|
Well, actually I didn't looked at it or tested it throughly. But I'll do...
The problem with Russian is not there are 2 or 3 numbers. The rule changes there ! Compare these two (taken from examples above):
quote:
When number ended with 5 to 10 (example: 0, 5, 27, 60, etc) it must be writed as ""
When number ended with 5 to 10 and from 11 to 20 (example: 0, 5, 27, 60, etc) it must be writed as ""
< |
Stop the WAR! |
|
|
|
GauravBhabu
Advanced Member
4288 Posts |
Posted - 02 October 2002 : 12:11:35
|
Try this. Change the values in query string
http://68.37.63.21:2910/files/lang.asp?nums=5||7
make sure the querystring values are like this. The link above is not including the part from where || appear
Nums=1||9
or
Nums = 5||2< |
Edited by - GauravBhabu on 02 October 2002 12:22:07 |
|
|
|
GauravBhabu
Advanced Member
4288 Posts |
Posted - 02 October 2002 : 12:33:08
|
quote:
Originally posted by bozden
Well, actually I didn't looked at it or tested it throughly. But I'll do...
The problem with Russian is not there are 2 or 3 numbers. The rule changes there ! Compare these two (taken from examples above):
quote:
When number ended with 5 to 10 (example: 0, 5, 27, 60, etc) it must be writed as ""
When number ended with 5 to 10 and from 11 to 20 (example: 0, 5, 27, 60, etc) it must be writed as ""
So it means 11 to 20 rule applies to forum/forums only and not to topic/topics???< |
|
|
|
Topic |
|
|
|
Forum Locked
