Read in data from 100,000+ files via command line

I had originally asked the following question:
https://www.experts-exchange.com/questions/28329553/Read-in-data-from-100-000-files-via-command-line.html

This at first seemed like it work exactly the way I needed, however I just discovered that lines with spaces were not correctly read in.

Below is the last iteration of the code:

#!/bin/sh

echo `date`
find . -name "*.arf" | while read f; do
  newpath="$(basename $(dirname "$f"))"
#/$(basename $f)"
  cat "$f" | gawk -v p="$newpath" '{ 
    attname=substr($1,1,length($1)-1); nlist=nlist"`, `"attname;
    attvalue= substr($2,2,length($2)-2); vlist=vlist", '\''"attvalue"'\''";
  }
  END { 
    printf "insert into `mydatabase`.`archives` (`NEWPATH%s`) values ('\''%s'\''%s);\n", nlist, p, vlist;
  }' >> myinsertfile.sql
#| tee -a myinsertfile.sql
  cnt=$((cnt+1))
  [ $(($cnt%100)) -eq 0 ] && echo "File #$cnt: $f"
done

echo "Total Files: $cnt"

echo `date`

Select allOpen in new window

For the following .arf file:

FILEID: "TIF490336"
PATH: "/optical/incoming/TIF490336"
TYPE: "TIF"
SECLEV: "10"
STATID: ""
USRID: "admin"
REQDATE: "08/02/2012"
REQTIME: "09:02:32"
GENDATE: "08/03/2012"
GENTIME: "09:02:32"
PROGID: ""
GROUPID: "Check Stubs"
DESC: "August"

Select allOpen in new window

It produced the following SQL statement:

insert into `mydatabase`.`archives` (NEWPATH,FILEID,PATH,TYPE,SECLEV,STATID,USRID,REQDATE,REQTIME,GENDATE,GENTIME,PROGID,GROUPID,`DESC`) values ('TIF18','TIF490336','/optical/incoming/TIF490336','TIF','10','','admin','08/02/2012','09:02:32','08/03/2012','09:02:32','','Chec','August');

Select allOpen in new window

Which resulted in the GROUPID column containing incorrect values.  How can the code above be adjusted to process spaces as well?