Awk Tab delimiter not working

pickup · February 18, 2014, 3:51am

Hi All,

I am having trouble with the AWK tab delimiter.

BEGIN {
FS = “\t”;
}

{
#This is where I set the total number of fields to iterate, in this case, it is
#231 because there are 33 rows and 7 columns
for(i=0; i<231; i++)
{
sum = 9+(i*7) #I iterate through starting at field 9
timefield = 8+(i*7) # This is to iterate through the time function
#if the sum isn’t a blank space, then I go ahead and print. If it is, I loop.
if ($sum != "") print($2, "\t", $timefield, "\t", $sum, "\t", "0")

}
}
END{

When I run this, my output looks like this:

However, if I remove one of the fields ("\t|, $timefield) then my output is correct with no new line being put in between the date and the next string, like so.

Any thoughts or help would be greatly appreciated. Thank you for your help!

DJ_Bazzie_Wazzie · February 18, 2014, 11:46am

It seems like you define your input field separator wrong. Here en example of how it will work:

set tabbedData to "id	field1	field2	field3	field4
20	john	doe	mainstreet	london
NULL	NULL	NULL	NULL	NULL
21	john	marksson	secondstreet	new york
22	mark	johnsson	thirdstreet		hong kong"

do shell script "awk -F " & quoted form of tab & " '
{
	if ($1 != \"NULL\" && NR != 1)
		print($2, $3, $4, $5)	  
}' <<<" & quoted form of tabbedData

To get the output tabbed separated, you only need to set the output field separator


set tabbedData to "id	field1	field2	field3	field4
20	john	doe	mainstreet	london
NULL	NULL	NULL	NULL	NULL
21	john	marksson	secondstreet	new york
22	mark	johnsson	thirdstreet	hong kong"

do shell script "awk -v 'OFS=" & tab & "' -F " & quoted form of tab & " '
{
	if ($1 != \"NULL\" && NR != 1)
		print($2, $3, $4, $5)	  
}' <<<" & quoted form of tabbedData

*because the question is not complete, this is only what I have.

pickup · February 18, 2014, 3:55pm

Thank you so much for your reply. I am pretty new to awk and am struggling to understand how this all works, so I apologize for the lack of information. Presently, I am using an awk script named: “awktest.awk”. In terminal, I am running the command (in the same directory):

administrator$ awk -f awktest.awk 022314.txt > test.txt

It seems that the way you are defining and building awk within Applescript might be the better way to go. The text file (022314.txt) I am parsing basically is this (slightly longer):

2014-02-23 2014-02-24 2014-02-25 2014-02-26 2014-02-27 2014-02-28
07:00:00 Western Tradition: Episode 37 (137) Dragon’s Tongue: Day by Day (02) Destinos: There Will Be Four of Us (43) “Connect With English: Lessons 47-48, Intro & Lesson 1 (47)” Fokus Deutsch: Lessons 22-24 with Review (22) French in Action: Lesson 46 (46)
07:30:00 Western Tradition: Episode 38 (138) Dragon’s Tongue: Getting Around (03) Destinos: A Promise and a Smile (44) French in Action: Lesson 47 (47)
08:00:00 “It’s A St. Louis Thing: New Media, Old Stories (130329)” Minnesingers Jubilee: Episode 809 (809) GED Connection: Episode 104 (104) Closer to Truth: The Mystery of Existence? (913) Drury Uncut: Episode 5 (105)
08:30:00 Minnesingers GED Connection: Episode 105 (105) Closer to Truth: Is God Perfect? (708)
09:00:00 Frames: 5th Annual Africa World Documentary Film Festival (1303)

I then run the awktest.awk which is this:

BEGIN {
FS = “\t”;
}

{
#This is where I set the total number of fields to iterate, in this case, it is
#231 because there are 33 rows and 7 columns
for(p=0; p<6; p++)
{
for(i=0; i<=NF; i++)
{
sum = (p+9+(i*7)) #I iterate through starting at field 9
timefield = 8+(i*7) # This is to iterate through the time function
datefield = (p+2)
#if the sum isn’t a blank space, then I go ahead and print. If it is, I loop.
if ($sum != "") print($datefield, "\t", $timefield, "\t", $sum, "\t", "0")

}
}
}
END{
}

output

Generally, the “\t” seems to work, but when I add the last field with “\t”, it seems to add a new line instead. What I don’t understand is why there is a newline after the date ($datefield) and time($timefield) when I have the tab delimiter (“\t”) after the fields.