DSED(1)

NAME

dsed - A command-line edit of partitioned dataset members and sequential datasets or zFS files.

SYNOPSIS

dsed [-dfhjJilrsv] [-c <encoding>] [-e <command>]+ <cmd> <dataset>

-c
    Dataset encoding. Default is IBM-1047.

-d
    Print debug information.

-e
    A sed command to execute before <cmd>.

-f
    Force open. Open dataset member in DISP=SHR mode.
    Default is DISP=OLD mode when not provided.

-h
    Display this syntax message.

-j
    Print a JSON object in this format:
    { "data": {"cmd": "<command>", "found": <integer>, "changed": <boolean> } }

-J
    Print JSON output in a readable format. This option enables -j if it was not specified.

-l
    Obtain exclusive lock for the dataset.

-r
    Use back references in regular expressions.

-s
    Stop executing sed commands if current command was successful.

-v
    Print verbose information.

DESCRIPTION

Replace word (global replace):

dsed 's/<regex>/<line>/g'    dest

Replace line global:

dsed '/<regex>/c\<line>'     dest

Replace nth line:

dsed '<n> c\<line>'          dest

Replace line first match n=1,last match n=$:

dsed '/<regex>/c\<line>/<n>' dest

Insert after(greedy match and insert):

dsed '/<regex>/a\<line>'     dest

Insert before(greedy match and insert):

dsed '/<regex>/i\<line>'     dest

Insert after(non-greedy match and insert,first match n=1,last match n=$):

dsed '/<regex>/a\<line>/<n>' dest

Insert before(non-greedy match and insert,first match n=1,last match n=$):

dsed '/<regex>/i\<line>/<n>' dest

Insert BOF:

dsed '1 i\<line>'            dest

Append EOF:

dsed '$ a\<line>'            dest

Append to nth line:

dsed '<n> a\<line>'          dest

Insert before nth line:

dsed '<n> i\<line>'          dest

Delete regex existed line:

dsed '/<regex>/d'            dest

Multiple sed commands:

dsed -e <command1> -e <command2>    dest

Modify target file, print debug information:

dsed -d <command>            dest

Modify ASCII dataset:

dsed -c ISO8859-1 <command>  dest

Multiple sed commands, will not execute command2 if command1 was successful:

dsed -s -e <command1> -e <command2> dest

NOTES:

'#' can be used as the delimiter instead of '/'
Do not escape curly braces in regular expressions.

dsed supports operations over generation datasets (GDS) that are referenced by relative name, but does not protect the generation data group base during execution. To serialize the group, use a batch job or mvscmd. Relative names are useful only when modifying sequential datasets. To reference a PDS or PDSE member from a GDS, use its absolute name.

EXAMPLES

Change all occurrences of '1st' to 'change':

dsed  's/1st/change/g'  IBM.TEXT

Replace lines containing '1st' with 'this is changed':

dsed  '/1st/c\\this is changed'  IBM.TEXT

Replace the last line containing '1st' with 'this is changed':

dsed  '/1st/c\\this is changed/$'  IBM.TEXT

Insert 'this is appended' after last line containing '1st':

dsed  '/1st/a\\this is appended/$'  IBM.TEXT

Insert 'this is inserted' before any line containing '1st':

dsed  '/1st/i\\this is inserted'  IBM.TEXT

Delete any line containing '1st':

dsed  '/1st/d' IBM.TEXT

Delete all the lines containing a capital letter followed by 3 digits:

dsed /[A-Z][0-9]{3}/d IBM.TEXT

Add a new line to beginning of the file:

dsed  '1 i\\this is an added line' IBM.TEXT

Insert a new line after the 2nd line:

dsed  '2 a\\this is an added line' IBM.TEXT

Change all occurrences of '1st' to 'change' (using '#' as delimiter):

dsed  's#1st#change#g'  IBM.TEXT

EXIT VALUES

0
    dsed successfully applied the command to the dataset.

other
    Error occurred. See error messages for details.