some bash shell scripting questions...

[3m00]

Okie, here they go...

1) How can I perform the same action to all the files in my current directory?

2) How can I ask for user input? (y/n answers)

[error27]

1)
for i in * ; do
echo foo $i
done

2)
read foo
echo $foo

[root.veg]

Follow up questions for anyone willing (sorry to piggyback, but this thread just answered my first question):

I want to run a couple of hundred MS Word documents and templates through antiword and save the results as plain text files with the same name, just without the .doc suffix - so can anyone fill in the blanks?:

for i in ./*.do? ; do
antiword [blank] >> [blank]
done

This means I need to know a method for:

(1) referring to the filename I currently want to look at. At a command prompt I would use a pipe, something like

cat ./*.do? | antiword

but I'm stumped when needing to save the output to to separate fielnames which depend on the current filename.

(2) Stripping the last n characters off the current filename from (1).

These seem like such basic questions, but using man, info and websearches hasn't been useful because I seem to need to know the exact command name...

[bwkaz]

for i in * ; do
newfilename=$(echo $i | sed -e 's/.doc/.txt/')
antiword $i >> $newfilename
done Of course, you'll have issues if any of your Word filename have spaces in them. If that's the case, you'll have to save IFS, replace IFS with $'\n', do the "for i in" stuff, replace $i with "$i" everywhere, and replace $newfilename with "$newfilename". Then set IFS to the saved value.

[root.veg]

Blimey. I was way off the mark there! Will read the man pages for sed and then I should understand properly what you mean.

And yes, there are loads of spaces in the filenames, but I'll try what you recommend and post back if it works...

Thanks!

[root.veg]

works a treat, thanks bwkaz: this is how it turned out:

keptvalue=$IFS

IFS=$'\n'

for i in *.doc ; do
newfilename=$(echo "$i" | sed -e 's/.doc/.txt/')
antiword "$i" >> "$newfilename"
done

IFS=keptvalue


Now I know more about sed, but still need to read on about the difference between single and double quotes, and what IFS is... time to rtfm. Joy!

[bwkaz]

For that last line, you'll probably want IFS=$keptvalue instead. ;)

IFS is the shell variable that contains the characters that separate words (the letters stand for Internal Field Separator, I think). It's talked about somewhere in bash's man page.

By default on most distros, IFS is set to $' \t\n', which is space, tab and newline. By setting it to $'\n', you ignore spaces and tabs, so that the shell doesn't think that spaces or tabs separate words.

The $'....' construct lets you use most C escape sequences between the single quotes.

[root.veg]

Many thanks for the concise explanation. Saved me a lot of time!