matching tab with sed

Question

I am new to sed, so want to know how can I match tab. I have some text like this -

    line one {
            line two {
                    filler {
                            text value
                    }
                    line four
            }
    }

I am using the expression, which matches the above text if there are no tabs.

sed -i -e "/ line two {/,/}/s/text .*\$/text new/" $file

However, there are 2 tabs before "line two" and 3 tabs before "filler" and 4 tabs before "text value".

Any help?

Answer 1

In GNU sed, tabs are represented as \t. So, the substitution will be successfully made with the following:

sed -i -e "/\tline two {/,/}/s/text .*\$/text new/" "$file"

The only required change was to replace / line two {/ with /\tline two {/. The command /s/text .*\$/text new/ does not require any change because it does not attempt to match the white space that precedes text. If there might be a tab that follows text, then use:

sed -i -e "/\tline two {/,/}/s/text[[:space:]].*\$/text new/" "$file"

If you are unsure of what white space character will appear, the expression [[:space:]] can be used to match any kind of white space.