Transforming text using table file.

[carl1905]

I attached simple text that is written in Japanese(01 Original text) and Korean(02 Transformed text).

It contains just 1 word that means "Quantum Mechanics". The question is how can I transform 01 to 02 using table.txt file?

sample.zip

Edited April 23, 2016 by carl1905

[jchd]

I don't believe there can exist such one-to-one correspondance between those codepages. Nevertheless it's still possible to achieve what you want.

Build a table of 64k entries. Entry address in the table is the Japanese character code, read as a 16-bit integer. Entry value is the corresponding Korean character code.

Note that not all characters in JIS use 16-bit encoding; you have to decode the text flow to decide whether the next char is 8- or 16-bit wide. See https://en.wikipedia.org/wiki/Shift_JIS for a primer.

 

[carl1905]

45 minutes ago, jchd said:

I don't believe there can exist such one-to-one correspondance between those codepages. Nevertheless it's still possible to achieve what you want.

Build a table of 64k entries. Entry address in the table is the Japanese character code, read as a 16-bit integer. Entry value is the corresponding Korean character code.

Note that not all characters in JIS use 16-bit encoding; you have to decode the text flow to decide whether the next char is 8- or 16-bit wide. See https://en.wikipedia.org/wiki/Shift_JIS for a primer.

 

Yes, there is no correspondence.  By the way, the final goal is that there is some kind of system that only Shift-JIS is allowed.

In this kind of system, I won't able to write EUC-KR. Well, the procedure that I want to do is..

1. Extract Shift-JIS text from binary file and write it in UTF-16LE. (Solved)

2. Translation into Korean. (Solved. Actually, it doesn't need code. )

3. Reading text from UTF-16LE txt file and put it back to binary file. (I can handle it.)

My question is, in 3rd procedure, I want to change Korean words to Japanese words according to table.txt file.

Because the system won't read EUC-KR, so system will read the binary file as if it is Shift-JIS. But actually I change the array of the Shit-JIS to my own corresponding table.

 

For example,

1) Original text is 'ABC' (=0x414243)

2) Deleting ABC and write BAC. (=0x424143)

3) Making a table and write 0x41=B, 0x42=C, 0x43=A.

4) Change the text file according to table and put it back to binary file.

5) Then the result is still BAC according to table, but the binary array is 0x414342.

 

Edited April 23, 2016 by carl1905

[jchd]

So use the 64K array to "translate" as you want.

[carl1905]

29 minutes ago, jchd said:

So use the 64K array to "translate" as you want.

Ok, thank you! I will try it.

Edited April 23, 2016 by carl1905