Selective Replace of characters
I have a file where each line represents a database record. Each field is fixed width.
I need to add a "prefix" to one of the numeric fields. So, in one file, all the records will be added by 1000, on a second file by 2000, etc (I know for a fact that the numbers have less digits than the amount I'm adding).
I've made many manipulations with the file already using sed. But can't seem to find the way to do this one. For instance, suppose the lines are of the form shown below. I'm trying to change the 10th char to my prefix and then need to replace all consecutive spaces after that to 0. It's this replacement I'm having trouble with.
I know how to identify the fields I want:
sed 's/^\(.\{9\}\) \([ ]*\)\(.*\)$/\11\2\3/g' #doesn't work because it doesn't replace spaces, just the prefix
I know how to replace spaces with 0:
sed 's/ /0/g' #which doesn't work either as it replaces all spaces and not just the ones I want.
I've done quite a few manipulations with these files using sed but can't seem to do this one. I'm probably just being dense, and can't think out of the box right now. Anybody's got any suggestions? (It doesn't need to use sed at all, if there's a better alternative)
I need to add a "prefix" to one of the numeric fields. So, in one file, all the records will be added by 1000, on a second file by 2000, etc (I know for a fact that the numbers have less digits than the amount I'm adding).
I've made many manipulations with the file already using sed. But can't seem to find the way to do this one. For instance, suppose the lines are of the form shown below. I'm trying to change the 10th char to my prefix and then need to replace all consecutive spaces after that to 0. It's this replacement I'm having trouble with.
I know how to identify the fields I want:
sed 's/^\(.\{9\}\) \([ ]*\)\(.*\)$/\11\2\3/g' #doesn't work because it doesn't replace spaces, just the prefix
I know how to replace spaces with 0:
sed 's/ /0/g' #which doesn't work either as it replaces all spaces and not just the ones I want.
I've done quite a few manipulations with these files using sed but can't seem to do this one. I'm probably just being dense, and can't think out of the box right now. Anybody's got any suggestions? (It doesn't need to use sed at all, if there's a better alternative)
Sample text:
Field01 2 3Field04
Field01 23 5Field04
Field01 5 3Field04
Field01 2 4Field04
should become
Field01 1002 3Field04
Field01 1023 5Field04
Field01 1005 3Field04
Field01 1002 4Field04ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.