Using awk to print CLI-version of kaddressbook ?

Yalla-One · 11-03-2006, 03:49 AM

Hi,

I recently read latest article on Linux.com on "CLI Magic: Command-line contact management" and am trying to make the example there work for a CLI-based reader of KDE's addressbook.

Currently I can grep the ~/.kde/share/apps/kabc/std.vcf, but is it possible to use awk to grep for a name or email address, and then have it print the entire record between the tags BEGIN:VCARD and END:VCARD ?

I tried this without the greatest success:

Code:

awk 'BEGIN { RS = "VCARD" } /${SEARCHNAME}/' std.vcf

It's a start, but I'm a bit stuck in getting further, ie making it print:

Name: firstname lastname
Email: emailaddress
Phone: 1234
Address: somestreet, etc etc
Birthday: ssss

Or does such an awk-script already exist under some obscure non-Googleable name?

-Y1

druuna · 11-03-2006, 04:02 AM

Hi,

I'm not familiar with the layout of these vcf files. Could you post an example (2 or 3 records are enough)? Would possibly make it a lot easier to help you

AnanthaP · 11-03-2006, 04:39 AM

Can u try the following pseudo-code? (A possible algorithm). Assumes that the input data format is text (not having ascii > 127) and on knwn format.

(1) sequentially read the data file until EOF.
(2) If BEGIN:VCARD, then append the line to a variable.
(3) If email or name matches, then set a flag.
(4) If END:VCARD, examine the flag and if on (search criteria was matched), then dump the array. Initialise the array, reset the flag and loop back to step 1.
(5) on EOF, examine the flag and if on, dump the array.

Steps 2, 3 4 are done for each record.

End

Yalla-One · 11-03-2006, 05:26 AM

Quote:

Originally Posted by druuna

I'm not familiar with the layout of these vcf files. Could you post an example (2 or 3 records are enough)? Would possibly make it a lot easier to help you

Here comes:

~>less .kde/share/apps/kabc/std.vcf

Code:

BEGIN:VCARD
ADR;TYPE=work:;;P.O. Box 11111;Dubai;;;United Arab Emirates
CATEGORIES:Personal
CLASS:PUBLIC
EMAIL:email@domain.com
FN:Donald Duck
N:Duck;Donald;;;
ORG:Duckster, Inc.
REV:2005-11-01T15:31:08Z
TEL;TYPE=WORK:+971 1 333 5555
TEL;TYPE=FAX:+971 4 330 5555
TEL;TYPE=CELL:+971 5 333 6666
TEL;TYPE=FAX;TYPE=WORK:+971 3 333 3456
UID:0CdjO4uH12gO7
VERSION:3.0
X-KPILOT-RecordID:1061634854
END:VCARD

BEGIN:VCARD
ADR;TYPE=work:;;;Paris;;0111 ;France
ADR;TYPE=home:;;Private Street 5;Paris;;0111;France
BDAY:1971-02-01T00:00:00Z
CATEGORIES:Personal
CLASS:PUBLIC
EMAIL:daffy@duck.com
FN:Daffy Duck
N:Duck;Daffy Mohamed;;;
ORG:Ducksters Son&Co
REV:2001-11-11T15:31:13Z
TEL;TYPE=WORK:+1 111 1111 1111
TEL;TYPE=CELL:+2 222 2222 2222
UID:0dp98Yt75An46g
VERSION:3.0
X-KPILOT-RecordID:106173131323
END:VCARD

BEGIN:VCARD
.....etc...
END:VCARD

druuna · 11-03-2006, 06:47 AM

Hi,

I wrote a general paragraph grep script a while ago:

Code:

#!/bin/bash
# usage: rgrep <string> <file>
# paragraph's should be seperated by a blank line.

pattern=$1
shift

awk -v pattern="$pattern" \
  'BEGIN { FS="\n" ; RS=""; ORS="\n\n" }
  $0 ~ pattern { print $0 }
  ' $*

This will also return the BEGIN and END tags, which could be filtered out later.

Sample run:

Code:

$ rgrep "Donald Duck" infile
BEGIN:VCARD
ADR;TYPE=work:;;P.O. Box 11111;Dubai;;;United Arab Emirates
CATEGORIES:Personal
CLASS:PUBLIC
EMAIL:email@domain.com
FN:Donald Duck
N:Duck;Donald;;;
ORG:Duckster, Inc.
REV:2005-11-01T15:31:08Z
TEL;TYPE=WORK:+971 1 333 5555
TEL;TYPE=FAX:+971 4 330 5555
TEL;TYPE=CELL:+971 5 333 6666
TEL;TYPE=FAX;TYPE=WORK:+971 3 333 3456
UID:0CdjO4uH12gO7
VERSION:3.0
X-KPILOT-RecordID:1061634854
END:VCARD

Hope this is good enough.

[edit]
fixed a mistake in the above script
[/edit]

Yalla-One · 11-05-2006, 02:33 PM

Thanks much!

Don't know what I did wrong, because the script simply prints all the records, quite similar to doing cat on the file..

-Y1

druuna · 11-05-2006, 02:51 PM

Hi,

I did find a mistake in my script. This: /bin/awk should be awk. I already changed it in the previous post.

But that does not explain the cat like behavior. Did you copy/paste my code or typed it by hand? Seems to be working here after I copy/pasted it.