mjhjmtu

I've a file which has the following output:

The dn: can can have more rdcPositions.

I only need the dn: which has an rdcPositions contains acme#6#

The result should print the cn and also the rdcPosition

dn: cn=00fa69bd-bede-4918-a017-b59b0901bb3d,ou=Named,ou=Identities,ou=Active,o
 u=Vault,o=acme
rdcPosition: cn=1950,ou=Entities,ou=Active,ou=Vault,o=acme#6#<position><cn>8946
 702990</cn><reqdate>1529318977</reqdate><startdate>1529318977</startdate><end
 date>1924902000</enddate><lifecycle><change><previousstatus/><newstatus>1</ne
 wstatus><date>1529318977</date></change><change><date>1529319116</date><previ
 ousstatus>1</previousstatus><newstatus>3</newstatus></change><change><date>15
 29481285</date><previousstatus>3</previousstatus><newstatus>6</newstatus></ch
 ange></lifecycle></position>

dn: cn=010903cd-e92d-4307-bffc-4921379153c0,ou=Named,ou=Identities,ou=Active,o
 u=Vault,o=acme
rdcPosition: cn=922445,ou=Entities,ou=Active,ou=Vault,o=acme#5#<position><cn>42
 79084890</cn><reqdate>1429014997</reqdate><startdate>1429014997</startdate><e
 nddate>1924902000</enddate><lifecycle><change><previousstatus/><newstatus>1</
 newstatus><date>1429014997</date></change><change><date>1429023084</date><pre
 viousstatus>1</previousstatus><newstatus>3</newstatus></change><change><date>
 1525107741</date><previousstatus>3</previousstatus><newstatus>6</newstatus></
 change><change><date>1525126716</date><previousstatus>6</previousstatus><news
 tatus>5</newstatus></change></lifecycle></position>
rdcPosition: cn=311982,ou=Entities,ou=Active,ou=Vault,o=acme#6#<position><cn>97
 26910833</cn><reqdate>1528120494</reqdate><startdate>1528120494</startdate><e
 nddate>1924902000</enddate><lifecycle><change><previousstatus/><newstatus>1</
 newstatus><date>1528120494</date></change><change><date>1528123478</date><pre
 viousstatus>1</previousstatus><newstatus>3</newstatus></change></lifecycle></
 position>

dn: cn=01126aa4-af80-401b-8713-29e360868999,ou=Named,ou=Identities,ou=Active,o
 u=Vault,o=acme
rdcPosition: cn=914570,ou=Entities,ou=Active,ou=Vault,o=acme#6#<position><cn>20
 68839799</cn><reqdate>1406284665</reqdate><startdate>1406284665</startdate><e
 nddate>1924902000</enddate><lifecycle><change><previousstatus/><newstatus>0</
 newstatus><date>1406284665</date></change><change><date>1406284666</date><pre
 viousstatus>1</previousstatus><newstatus>3</newstatus></change><change><date>
 1435847283</date><previousstatus>3</previousstatus><newstatus>6</newstatus></
 change></lifecycle></position>
rdcPosition: cn=999546,ou=Entities,ou=Active,ou=Vault,o=acme#6#<position><cn>76
 03071057</cn><reqdate>1400325753</reqdate><startdate>1400325753</startdate><e
 nddate>1924902000</enddate><lifecycle><change><previousstatus/><newstatus>0</
 newstatus><date>1400325753</date></change><change><date>1400325754</date><pre
 viousstatus>1</previousstatus><newstatus>3</newstatus></change><change><date>
 1449224475</date><previousstatus>3</previousstatus><newstatus>6</newstatus></
 change></lifecycle></position>
rdcPosition: cn=3513,ou=Entities,ou=Active,ou=Vault,o=acme#6#<position><cn>2802
 042129</cn><reqdate>1406284761</reqdate><startdate>1406284761</startdate><end
 date>1924902000</enddate><lifecycle><change><previousstatus/><newstatus>0</ne
 wstatus><date>1406284761</date></change><change><date>1406284762</date><previ
 ousstatus>1</previousstatus><newstatus>3</newstatus></change><change><date>14
 49224599</date><previousstatus>3</previousstatus><newstatus>6</newstatus></ch
 ange></lifecycle></position>
rdcPosition: cn=312936,ou=Entities,ou=Active,ou=Vault,o=acme#3#<position><cn>19
 23461515</cn><reqdate>1449217172</reqdate><startdate>1449217172</startdate><e
 nddate>1924902000</enddate><lifecycle><change><previousstatus/><newstatus>1</
 newstatus><date>1449217172</date></change><change><date>1449225081</date><pre
 viousstatus>1</previousstatus><newstatus>3</newstatus></change></lifecycle></
 position>

The input seems to be LDIF as specified in RFC 2849.

I'd strongly recommend not to use the usual awk/sed/grep tool chain for processing LDIF because of the following reasons:

• Long attribute values lines (including dn:) are wrapped with a single space indicating the line-continuation.


• Attribute values containing non-ASCII chars will be base64-encoded.

The best solution is to use a decent LDIF parser for your favorite scripting language.

E.g. for Python use the module ldif in python-ldap:

See docs: ldif -- LDIF parser and generator

Your desired output is not quite clear. How far would this get you:

 awk '
 while (match($0, /rdcPosition: [^ ]*acme#6#[^ ]*/)) print substr ($0, RSTART, RLENGTH)
 $0 = substr ($0, RSTART + RLENGTH);
 
 
 ' file 
rdcPosition: cn=1950,ou=Entities,ou=Active,ou=Vault,o=acme#6#8946
rdcPosition: cn=311982,ou=Entities,ou=Active,ou=Vault,o=acme#6#97
rdcPosition: cn=914570,ou=Entities,ou=Active,ou=Vault,o=acme#6#20
rdcPosition: cn=999546,ou=Entities,ou=Active,ou=Vault,o=acme#6#76
rdcPosition: cn=3513,ou=Entities,ou=Active,ou=Vault,o=acme#6#2802

For your changed request in your comment, how far would this get you? If not satisfied, please become way more specific defining your desired output.

awk '
 DN = $1 FS $2
 while (match($0, /rdcPosition: [^ ]*acme#6#[^ ]*/)) print DN, substr ($0, RSTART, RLENGTH)
 $0 = substr ($0, RSTART + RLENGTH);
 
 
' file
dn: cn=00fa69bd-bede-4918-a017-b59b0901bb3d,ou=Named,ou=Identities,ou=Active,ou=Vault,o=acme rdcPosition: cn=1950,ou=Entities,ou=Active,ou=Vault,o=acme#6#8946
dn: cn=010903cd-e92d-4307-bffc-4921379153c0,ou=Named,ou=Identities,ou=Active,ou=Vault,o=acme rdcPosition: cn=311982,ou=Entities,ou=Active,ou=Vault,o=acme#6#97
dn: cn=01126aa4-af80-401b-8713-29e360868999,ou=Named,ou=Identities,ou=Active,ou=Vault,o=acme rdcPosition: cn=914570,ou=Entities,ou=Active,ou=Vault,o=acme#6#20
dn: cn=01126aa4-af80-401b-8713-29e360868999,ou=Named,ou=Identities,ou=Active,ou=Vault,o=acme rdcPosition: cn=999546,ou=Entities,ou=Active,ou=Vault,o=acme#6#76
dn: cn=01126aa4-af80-401b-8713-29e360868999,ou=Named,ou=Identities,ou=Active,ou=Vault,o=acme rdcPosition: cn=3513,ou=Entities,ou=Active,ou=Vault,o=acme#6#2802

Using the following sed script (it assumes we're running it with sed -n):

/^dn:/ # this is a "dn" line
 N; # append the next line
 s/n //; # remove the newline and the space
 x; # exchange pattern space with hold space
 /o=acme#6#/p; # print if pattern space contains our string
 d; # delete from pattern space, start next cycle

/^rdcPosition:/ # this is a "rdcPosition" line
 :again; # define label for loop
 N; # append the next line
 s/n //; # remove the newline and the space
 #</position>#!b again; # if the end tag "</position>" was not read, loop
 /o=acme#6#/H; # append to hold space if matching what we're looking for

$ # at the very end of input
 x; # exchange pattern and hold space
 /o=acme#6#/p; # print if pattern space contains our string

What the script does is essentially to build up a string in the sed "hold space" (general purpose buffer that survives between cycles). The string will start with the dn line and then it will be appended with the rdcPosition lines that contain the particular string that we are interested in.

Whenever a new dn line is found, or when we are at the end of the input, the hold space is conditionally printed if it contains our string (it may not contain it if none of the rdcPosition lines for the current dn line matched).

Testing it:

$ sed -n -f script.sed file
dn: cn=00fa69bd-bede-4918-a017b59b0901bb3d,ou=Named,ou=Identities,ou=Active,ou=Vault,o=acme
rdcPosition: cn=1950,ou=Entities,ou=Active,ou=Vault,o=acme#6#<position><cn>8946702990</cn><reqdate>1529318977</reqdate><startdate>1529318977</startdate><enddate>1924902000</enddate><lifecycle><change><previousstatus/><newstatus>1</newstatus><date>1529318977</date></change><change><date>1529319116</date><previousstatus>1</previousstatus><newstatus>3</newstatus></change><change><date>1529481285</date><previousstatus>3</previousstatus><newstatus>6</newstatus></change></lifecycle></position>
dn: cn=010903cd-e92d-4307-bffc-4921379153c0,ou=Named,ou=Identities,ou=Active,ou=Vault,o=acme
rdcPosition: cn=311982,ou=Entities,ou=Active,ou=Vault,o=acme#6#<position><cn>9726910833</cn><reqdate>1528120494</reqdate><startdate>1528120494</startdate><enddate>1924902000</enddate><lifecycle><change><previousstatus/><newstatus>1</newstatus><date>1528120494</date></change><change><date>1528123478</date><previousstatus>1</previousstatus><newstatus>3</newstatus></change></lifecycle></position>
dn: cn=01126aa4-af80-401b-8713-29e360868999,ou=Named,ou=Identities,ou=Active,ou=Vault,o=acme
rdcPosition: cn=914570,ou=Entities,ou=Active,ou=Vault,o=acme#6#<position><cn>2068839799</cn><reqdate>1406284665</reqdate><startdate>1406284665</startdate><enddate>1924902000</enddate><lifecycle><change><previousstatus/><newstatus>0</newstatus><date>1406284665</date></change><change><date>1406284666</date><previousstatus>1</previousstatus><newstatus>3</newstatus></change><change><date>1435847283</date><previousstatus>3</previousstatus><newstatus>6</newstatus></change></lifecycle></position>
rdcPosition: cn=999546,ou=Entities,ou=Active,ou=Vault,o=acme#6#<position><cn>7603071057</cn><reqdate>1400325753</reqdate><startdate>1400325753</startdate><enddate>1924902000</enddate><lifecycle><change><previousstatus/><newstatus>0</newstatus><date>1400325753</date></change><change><date>1400325754</date><previousstatus>1</previousstatus><newstatus>3</newstatus></change><change><date>1449224475</date><previousstatus>3</previousstatus><newstatus>6</newstatus></change></lifecycle></position>
rdcPosition: cn=3513,ou=Entities,ou=Active,ou=Vault,o=acme#6#<position><cn>2802042129</cn><reqdate>1406284761</reqdate><startdate>1406284761</startdate><enddate>1924902000</enddate><lifecycle><change><previousstatus/><newstatus>0</newstatus><date>1406284761</date></change><change><date>1406284762</date><previousstatus>1</previousstatus><newstatus>3</newstatus></change><change><date>1449224599</date><previousstatus>3</previousstatus><newstatus>6</newstatus></change></lifecycle></position>

A logically equivalent awk script which produces the same output as the sed code above:

/^dn:/ 
 if (hold ~ "o=acme#6#")
 print hold

 hold = $0;
 getline
 hold = hold substr($0, 2)
 next


/^rdcPosition:/ 
 line = $0
 while (line !~ "</position>") 
 getline
 line = line substr($0, 2)
 

 if (line ~ "o=acme#6#")
 hold = hold ORS line


END 
 if (hold ~ "o=acme#6#")
 print hold

The substr($0, 2) calls will strip the leading space off of the broken lines in the input.

Both scripts assumes that the dn line is broken into exactly two lines.