sed command to replace all occurances of a word in a file

[ukbhan]

hi ,
I have a text file in which i need to replace all occurances of the word "prod" with "test". I should not replace for the substirng "prod" in a word.
for example "product" should not be replaced to "testuct"

I am using the command sed -e 's/prod/test/g' to replace all occurances of the word "prod" with "test".

but it is replacing "product" to "testuct" which i dont want.

So how should i change my sed command.


Thanks in advance

[Zeroangel]

hi ,
I have a text file in which i need to replace all occurances of the word "prod" with "test". I should not replace for the substirng "prod" in a word.
for example "product" should not be replaced to "testuct"

I am using the command sed -e 's/prod/test/g' to replace all occurances of the word "prod" with "test".

but it is replacing "product" to "testuct" which i dont want.

So how should i change my sed command.


Thanks in advance

Is there a common character that comes after 'prod', such as a space, line break, or quotation mark? If so, then that's your answer right there.

You could also run another sed command to replace testuct with product.

[mvenkatapathy]

Hi ukbhan,

sed "s/ prod / test /g" will solve your problem.

[blinky]

try (sed -e "s/\<prod\>/test/g") eg:

echo "prod product cattleprod prod." | sed -e "s/\<prod\>/test/g

[mvenkatapathy]

try (sed -e "s/\<prod\>/test/g")

eg:

echo "prod product cattleprod prod." | sed -e "s/\<prod\>/test/g

What does "<" and ">" brackets in "\<prod\>" actually mean here?
I am sure "\" means you are escaping the brackets.

[blinky]

\< = starts with and \> = ends with , used together whole word

[mvenkatapathy]

Thanks blinky.