Shell Programming and Scripting

How could i take an input file and split the numeric values from the alpha values (123 vs abc) to distinc columns, and if the source is blank to keep it blank (null) in both of the new columns:


So if the source file had a column like:

Value:

Code:

|1   |
|2.3|
|     |
|No|

I would create a new appended file containing a column with the numeric:
Numeric:

Code:

|1   |
|2.3|
|     |
|     |

And a column with the abc:

Code:

|     |
|     |
|     |
|No|

Thanks,
DL

Moderator's Comments:

Use code tags!

Code:

awk -F'|' '{if ($2 ~ /[^0-9. ]/) {$(NF+1)=$2;$2=" "}}NF++' OFS='|' myFile

vgresh thank you. i'm goign to see if i can use your code. The file in question actually has 1.3million rows (tab delimited). the source column is number 35, and I need to append the 123 and abc column as two new columns to the end of the file.

---------- Post updated at 02:27 PM ---------- Previous update was at 02:24 PM ----------

I should also add it is possible that something that would need to be parsed to the abc column is: < 3.2

---------- Post updated at 02:39 PM ---------- Previous update was at 02:27 PM ----------

vgresh! thank you this is almost there.

currently only the abc values are moved to a new column. I also need the 123 values to move to their own distinct column.

in addition I need to copy these values to new columns and retain them in the source column.

cheers! ty!

please provide a small representative sample of the ACTUAL data and the desired output. A 10 record/line representative sample should be enough.
Please either code tags OR attach a file to a thread.

hi vgresh,

Here is a representative sample attached (input worksheet).
Desired output is on the output worksheet.

assuming fields are pipe separated - if not change -F and OFS= to the appropriate characters:

Code:

awk -F'|' '{NF+=2;;$NF=$(NF-1)=""}{($1 ~ /[^0-9. ]/)?$(NF-1)=$1:$(NF)=$1}1' OFS='|' myFile

or better yet - dealing with empty input fileds AND specifying what field number (f) you want to check:

Code:

awk -F'|' '{NF+=(!NF)?3:2;$NF=$(NF-1)=""}{($f ~ /[^0-9. ]/)?$(NF-1)=$f:$(NF)=$f}1' f=1 OFS='|' myFile

Input file myFile, TAB separated fields, work with field#1, append two columns to each line:

Code:

awk 'BEGIN{FS=OFS="\t"} {spc=$1; gsub(/./," ",spc); if ($field~/[^0-9.]/) {print $0,$field,spc} else {print $0,spc,$field}}' field=1 myFile

It tries to keep the old field width. Otherwise it is just

Code:

awk 'BEGIN{FS=OFS="\t"; spc=""} {if ($field~/[^0-9.]/) {print $0,$field,spc} else {print $0,spc,$field}}'  field=1 myFile