Sed

jimmyendless

Hi All,

I found this command that removes any white space at the start of a line.
sed -e 's/^ [ \t]*//'
It works grand but could anyone explain to me how it works?

Thanks

Khannie

it's a regular expression. They are worth books, but here's a quick overview of that one:

s/change this/to this/

The s bit is "substitute".
/ = start of match part
^ means match start of line
then there's a space
then there's a grouping of space or \t (tab)
the * means any number of times
/ = end of match part, start of substitute part
/ = end of substitute part

So, putting that all together: Do a substitution. Match the start of a line followed by a space, followed by any number of spaces or tabs, replace with nothing (i.e. delete).

jimmyendless

Thanks for that Khannie. It's pointless using it if I have no idea how it works.

Khannie

No problem. Regular expressions are a slight pain in the arse to learn, and intimidating to read, but they have saved me hundreds of hours over the years. They're an amazing tool.

Skrynesaver

There is only one book, Jeffrey Friedl's Mastering regular expressions

Aside from grep, sed and awk which use regexes to great effect at the command line they are also a "mini-game" within every programming language and are worth learning.

Because so many language ape Perl's regex library (PCRE, Perl Compatible Regular Expressions) it's probably worth learning how to use them in Perl (particularly if you need to manipulate text.

Khannie

Skrynesaver wrote: »

There is only one book, Jeffrey Friedl's Mastering regular expressions

I read (most of) it years ago. Really good book.

humbert

I found this brilliant for getting started:
http://www.codeproject.com/Articles/9099/The-30-Minute-Regex-Tutorial

10-10-20

Khannie wrote: »

it's a regular expression. They are worth books, but here's a quick overview of that one:

s/change this/to this/

The s bit is "substitute".
/ = start of match part
^ means match start of line
then there's a space
then there's a grouping of space or \t (tab)
the * means any number of times
/ = end of match part, start of substitute part
/ = end of substitute part

So, putting that all together: Do a substitution. Match the start of a line followed by a space, followed by any number of spaces or tabs, replace with nothing (i.e. delete).

I gotta say, that's the first time I have ever understood any statement which follows the name of the utility. Did you ever think of writing a book...?

Khannie

Haha. Glad to be of service. That approach to regex's (match this, then match this, then match this) is something I probably picked up from the book mentioned earlier in the thread.