way to catch if file is simple text based

[Jovaras]

hello,
I'm filtering files.
I'm looking for a way to detect if a file is simple text based(not encoded).
Of course I could define all the extensions, but is there a quick way to see if file is simple text(for e.g.: *.php) or encoded (for e.g.: *.exe) ?
Thanks in advance!

[Dark]

I believe I did something similar to what you're asking about earlier today. However I did mine in Swing. I'm pretty sure it belongs to IO so you should be able to use it in non-swing applications.

Using a buffered reader I opened up a selected file. Using the file information it collected I used

if(file.getAbsolutePath().contains(".txt")){//Code}

You could try something similar to accomplish that. Reading through a folder, directory or other collection of files may require a little bit more than this seeing as mine asked you to input the file location.

[Norm]

way to detect if a file is simple text based

Read it and check that all bytes/chars read are text or controls like tab and newline
Not all files have extensions.

[Dark]

I think he's trying to distinguish between files that do have extensions, at least that's what I got from it.

[JosAH]

Read it and check that all bytes/chars read are text or controls like tab and newline
Not all files have extensions.

That'd only catch English/American text, i.e. ASCII representable text. Any text outside of that (limited) range would fail that test. No need to worry though, because the good old 'file' Unix utility would fail as well ;-) There is no one-size-fits-all cheap test to classify a file base on its contents nor on its name nor on its name extension.

kind regards,

Jos

[Jovaras]

hmmm, so how do famous programs(like searching in files) handle reading files and excluding executables?
Or it simply can not be done using Java?

[Norm]

Any programming language that can read the bytes of a file can check them using an algorithm to determine if the whole file is made of text. Define what "text" is and I can write a program to test for it. The problem as JosAH pointed out is there are lots of different "text" character sets.