I have these as input files
file1
Red + Yellow = Orange; Yellow + Blue = Green; Blue + Red = Violet
file2
Red;1
Yellow;2
Blue;3
I am doing this in unix.
Output
file3
1 + 2 = Orange; 2 + 3 = Green; 3 + 1 = Violet
2 Answers
awk 'NR == FNR { # First file
split($0,a,/;/)
#print "change " a[1] "to " a[2]
change[a[1]]=a[2]
}
NR != FNR {
for (i=1;i<NF;i++) {
if ($i in change) {
$i=change[$i]
}
}
print
}' file2 file1
Use the awk idiom NR != FNR to tell if this is the first file or the second. If it is the first file then split the line on ; and store the mapping in the change array. If it is the second then loop over the input field, if any match then replace with the correct change. At the end print the result.
-
Thank you for that. What if I change my file 2 into something like this? Red;1 2 Yellow;2 3 Blue;3 1 And the output file will look like this 1 2 + 2 3 = Orange; 2 3 + 3 1 = Green; 3 1 + 1 2 = Violet which is essentially just replacing the strings in file1 with the corresponding string in the 2nd col of file2. Will this still work?peon– peon2016-12-06 07:17:53 +00:00Commented Dec 6, 2016 at 7:17
-
@peon please edit your question to show the data files, and try the commands. If I understand your comment correctly the answer is
yes.icarus– icarus2016-12-06 13:07:40 +00:00Commented Dec 6, 2016 at 13:07
#!/bin/bash
IFS=";"
while read NAME VALUE
do
sed -i "s/${NAME}/${VALUE}/g" file1
done < file2
-
This would also replace the words when they occur as substrings, as in
LightBlue.2019-04-20 16:08:52 +00:00Commented Apr 20, 2019 at 16:08
stringsall single words, space delimited?