Shell Programming and Scripting

Depending on the original text, you may want to create the two output files in separate processes.
In one process drop only the obvious English characters, and in the second drop only the obvious Assamese.

Thank u alister it is working.

but with some blank line. For small size text file it is OK but for large text file if text of one language[suppose English ] is very dense then in the newly created file other language [suppose Assamese] file, there are same[ i.e dense English text nos ] number of blank lines are printing.

Hmmm. In case I misunderstood the problem, I assumed that each newline terminated string was one language or the other. If sub-line-level granularity is desired, my awk solution is unsuitable.

Regards,
Alister

---------- Post updated at 09:45 AM ---------- Previous update was at 09:42 AM ----------

You're very welcome. Yeah, that code will always interpret blank lines as English. If you want to discard them, perhaps you can use:

Code:

awk '!NF {next} /\xe0/ {print > "assamese"; next} {print > "english"}' data

alister's code, as I read it, will write a full line of mixed text to single file, never splitting words by language.

The regex tests for the existence of \xe0, which will always be true, based on the example. In order for awk to work correctly, you have to be in a locale that supports both languages, and test each character for assamese-ness or ASCII-ness. Because we are splitting on a per character/word basis.

Correct. It is a line-based solution. For some reason, I jumped to the conclusion that that's what was needed. Perhaps I flashed back to https://www.unix.com/shell-programmin...xcel-file.html :P