What Does This Regular Expression Mean?

[James889]

Howdy,

I came across a regular expression i couldn't get my head around.

' there \([^ ]*\)'

echo "Howdy there neighbor" | sed 's/there \([^ ]*\)//'

returns howdy.

It's the subgroup that's a bit confusing.

match any sentence which contains banana then a space and then a non-space character.

Is this the correct way of interpreting this regular expression ?

 

[gsal]

So, basically, it matches "there " (the word 'there' followed with a blank space) followed with as many consecutive non-blank spaces as it can find "[^ ]*" and replaces that with nothing.

You can test that only replaces what I said, if you test it with "Howdy there neighbor what up?"

Oh, the back slashes are there to escape the parenthesis within the double quotes

 

[James889]

.. followed with as many consecutive non-blank spaces as it can find "[^ ]*" and replaces that with nothing.

Is this the same as saying 'match as much as possible up until a white space is found' ?

 

[Ibix]

"White space" would normally include tab characters. This will eat everything up until the first space character; that detail aside, yes.

 

[gsal]

Sorry, I guess I need to be more correct, like Ibix says.

The expression "[^ ]*" will consume consecutive character after character until it finds a "blank space" character.

A "blank space" character itself is not the same as "white space" in general...it is a subset. If the regular expression is looking for "white space" then, "blank space" and "tab" characters qualify...but if you are looking for "blank space", then a "tab" is a totally different character.