正则表达式：以特定事物结束

Question

I am trying to find an expression that begins with Hello and ends in one of two ways: with nothing after the "Hello", or if it has something else after it needs to be preceded with "//". After the //, anything goes.

I tried: grep '^Hello(//.*)?$' but this does not work.
There is something wrong with the last part: (//.*)?$

Sample Input:
Hello
Hello blah
Hi
Hello //
Hello // blah blah
Hello //blah

Sample Output using egrep:
Hello
Hello //
Hello // blah blah
Hello //blah

Answer 1

This is pretty straightforward with egrep:

egrep '^Hello(\s*\/\/.*)?$' input.txt

That is:

• ^ ... - Force the match to start at the beginning of the line.
• Hello - Definitely match the required phrase Hello .
• (\\s* ... ) - Allow optional whitespace to follow Hello .
• ( ... \\/\\/ ... ) - Match the forward slashes, escaping them (because some shells can do funny things to your regexes).
• ( ... .*) - Allow anything after the slashes.
• ( ... )? - The question mark indicates the parenthesized part is optional.
• ... $ - Force the regex to only match if it consumes through the end of the line.

You were using grep instead of egrep . Plain grep uses a much simpler regex syntax that doesn't allow some of the operators you might like to use. Notably, in plain grep , parentheses and ? are just plain characters, not special meta-characters for grouping, so plain grep was searching for literal ( and ) in your file. When in doubt, prefer egrep .

(And yes, for the pedantic folks in the audience, egrep is indeed just an alternate name for grep -E or grep --extended-regexp , but it's much easier to remember and type egrep than either of the other two "native" forms.)

Answer 2

Given:

$ echo "$txt" 
Hello
Hello blah
Hi
Hello //
Hello // blah blah
Hello //blah

With grep :

$ echo "$txt" | grep -E '^Hello$|^Hello[[:space:]]+//'
Hello
Hello //
Hello // blah blah
Hello //blah

Or with awk :

$ echo "$txt" | awk '/^Hello$/ || /Hello[[:space:]]+\/\//'
Hello
Hello //
Hello // blah blah
Hello //blah

Or if you want to make sure there is something after the // :

$ grep -E '^Hello$|^Hello[[:space:]]+//[^[:space:]]+'

Answer 3

Use the "match only whole lines" option of egrep ( -x ).
Then look for optional whitespace ( [[:space:]]* ), two escaped / ( \\/\\/ ) followed by anything or nothing, .* .
Use the optional specifier and parenthesis to allow the special ending, not require it ( (...)? ).

egrep -x "Hello([[:space:]]*\/\/.*)?"

Answer 4

Another awk proposal. The first part /^Hello$/ matches just a singel Hello and the second part matches // and in this case prints the entire line.

awk '/^Hello$/||/\/\//' file

Hello
Hello //
Hello // blah blah
Hello //blah