How to remove a particular character after a specific word

Question

How do I remove the , character after a specific word?

For example, I want to delete the , character after the word IBM

  IBM, Inc.:APP-VM-Core:4.4.2.0:04,

the expected results should be

  IBM Inc.:APP-VM-Core:4.4.2.0:04,

Meanwhile I do the following

   echo “IBM, Inc.:APP-VM-Core:4.4.2.0:04,” | sed s'/,//'

but the problem is that it deletes every , character, no matter what the first word was before the ,. In my case I want to delete the , only after IBM.

Example:

echo "APPLE, Inc.:APP-VM-Core:4.4.2.0:04,"| sed s'/,//'

APPLE Inc.:APP-VM-Core:4.4.2.0:04,

but should be as

APPLE, Inc.:APP-VM-Core:4.4.2.0:04, 

Answer 1

Try:

$ echo "IBM, Inc.:APP-VM-Core:4.4.2.0:04," | sed '/^ *IBM/s/,//'
IBM Inc.:APP-VM-Core:4.4.2.0:04,

or:

$ echo "IBM, Inc.:APP-VM-Core:4.4.2.0:04," | sed 's/^\( *IBM\),/\1/'
IBM Inc.:APP-VM-Core:4.4.2.0:04,

Answer 2

If you are trying to modify a file you'd rather use ex or ed but definitely not sed. Why not sed?

sed is a Stream EDitor and you'd modify the data coming from a stream with it.

Using ex you could do the following

ex -sc '%s/IBM,/IBM/g|x' file wheren file is your filename.

As alternative. If you want to remove leading spaces or tabs you will need more than ^ *. For example. The solution would be using awk as follow:

awk '{sub(/^[ \t]+IBM,/,"IBM");};1' input_file > output_file

or

awk '{sub(/^[[:blank:]]*IBM,/,"IBM");};1' input_file > output_file

Mind you. ^ * will not be able to detect tabs.

You also can use ex as explained earlier but now including the regular expression to match only the line beginning with spaces and including the pattern IBM

ex -sc '%s/^[[:blank:]]*IBM,/IBM/g|x' input_file