How i can merge two text files?

[Francesco_13]

Hi,
i have two text files with table names,for example:

file1 file2
--------|-------------
tab1 tab6
tab2 tab2
tab3 tab5
tab4 tab7
tab5 | tab4

How i can merge file1 and file2 and produce file3 with all table names , with NO duplicate table names?
thanks
Best regards.
Francesco

[James R. Ferguson · Accepted solution]

Hi:

# cat file1 file2 | sort -u > file3

Regards!

...JRF...

[A. Clay Stephenson]

sort -u file1 file2 > mymergedfile

[Francesco_13]

i'm sorry , i forget the last action:
for the same table name in file1 and file2, i'm do it to write the table name from file2 only.(the duplicate table_name to eliminate it's the table_name from file1).

more thanks and regards

Francesco

[James R. Ferguson]

Hi (again):

You wrote, "for the same table name in file1 and file2, i'm do it to write the table name from file2 only.(the duplicate table_name to eliminate it's the table_name from file1).".

I'm missing something. What does it matter if they are the same?

Regards!

...JRF...

[Francesco_13]

Hi,
the file1 and file2 have the same table names, but file1 it's from one old version and file2 from new version of the same table. I need to produce the file3 with all table_names and know if the occurence it's from old or new version of same table.
(the same table have different data depending of version and i need to resume if the table have old or new data
example:
file3
tab1 > file1
tab2 > file1
tab3 > file1
tab4 > file2
tab5 > file2
tab6 > file2

i hope it's clear explained for my poor english, and sorry if i have changed the original question!
+ thanks and regards

Francesco

[Hein van den Heuvel]

You'll need a little script.
Here is a perl suggestion and sample usage

$ cat x.pl
while ($file=shift @ARGV) {
open FILE, "<$file";
while () {
chomp;
$tables{$_} = $file;
}
}
foreach (sort keys %tables) {
print "$_ > $tables{$_}\n";
}$ cat x
aap
noot
mies
$ cat xx
noot
mies
teun
piet
$ cat xxx
teun
vuur$ perl x.pl x xx xxx
aap > x
mies > xx
noot > xx
piet > xx
teun > xxx
vuur > xxx
$ perl x.pl xxx x xx
aap > x
mies > xx
noot > xx
piet > xx
teun > xx
vuur > xxx


Cheers,
Hein.

[spex]

Hi Francesco,

I tested this under FreeBSD, as I don't have access to HP-UX at the moment. HP-UX solution should be similar.

$ cat f1 && echo '====' && cat f2
tab1
tab1
tab2
tab3
tab4
====
tab2
tab4
tab6
tab7
$ grep -H '' * | sort -t: +1 | uniq
f1:tab1
f1:tab2
f2:tab2
f1:tab3
f1:tab4
f2:tab4
f2:tab6
f2:tab7

PCS

[Francesco_13]

Hi,
thanks for script, but it's too hard for me.(i don´t know use of perl) i will try to use it.
A traditionally unix script would be more easy for me.

thanks.
Best regards.
Francesco

[Dennis Handly]

From Hein's example you can use awk:
#!/usr/bin/ksh
TMP=/var/tmp/merge.$$
for i in $* ; do
awk -v NAME=$i '
{ print $0, ">", NAME
} ' $i >> $TMP
done
sort -u -k1,1 $TMP
rm -f $TMP

The output for the 3 files:
$ merge x*
aap > x
mies > x
noot > x
piet > xx
teun > xx
vuur > xxx

[Hein van den Heuvel]

Actually, with AWK there is a really 'short' solution:

$ awk '{a[$1]=FILENAME} END {for (k in a) { print k " > " a[k] }}' file1 file2

With the same sample files

$ grep "." x*
x:aap
x:noot
x:mies
xx:noot
xx:mies
xx:teun
xx:piet
xxx:teun
xxx:vuur

$ awk '{a[$1]=FILENAME} END {for (k in a) { print k " > " a[k] }}' x xx xxx
teun > xxx
vuur > xxx
mies > xx
aap > x
noot > xx
piet > xx

$ awk '{a[$1]=FILENAME} END {for (k in a) { print k " > " a[k] }}' xxx xx x
teun > xx
vuur > xxx
mies > x
aap > x
noot > x
piet > xx
Hein

[Sandman!]

Try the awk construct below:

awk '{
if (x[$1]) {
if (FILENAME=="file2")
x[$1]=FILENAME
} else
x[$1]=FILENAME
} END{for (i in x) print i" > "x[i]}' file1 file2

[Francesco_13]

Hi,
thanks to all! with the last solution i have produced the correct file.
Best regards.
Francesco