Cfxtrjtrk

Look at these two files:

• File 1


• File 2

They look to me entirely identical, but File 1 has 41 bytes and File 2 has 44 bytes.

Now, I would not care about 3 extra chars, however, I think these chars are also responsible to a strange error: when I clone and run these files, File 1 runs OK while File 2 returns an error on line 3: "./BasicCheck.sh: line 3: cd: case0 : No such file or directory".

So: what are these extra chars? Why do they cause errors? And what should I do to fix these errors?

My guess is that they are newlines - perhaps File 1 has Linux-based newlines and File 2 has Windows-based newlines. Is it correct? And why should it cause such errors?

The file command points out the issue:

% file file*

file1: Bourne-Again shell script, ASCII text executable

file2: Bourne-Again shell script, ASCII text executable, with CRLF line terminators

The second script is in "DOS" format. and so will not work as expected. You can convert it with dos2unix and the results are now the same

% dos2unix file2

dos2unix: converting file file2 to Unix format ...



% ls -l file1 file2

-rw-r--r-- 1 sweh sweh 41 Mar 19 11:23 file1

-rw-r--r-- 1 sweh sweh 41 Mar 19 11:25 file2

When you try to use a DOS format file with Unix commands, there is a trailing control-M at the end of every line. So a line such as

a=b

will really set a to b followed by control-M, shorthand ^M:

a=b^M

Every command will be similarly impacted; eg a simple line with the two-character command ls would try to run the three-character command ls<control-M>, which doesn't exist.