CODEWITHSUNDEEP
bashawkgrep
Bleakley
Bleakley

Reputation: 733

grep lines that appear partly in one file to another

I have a file #1 that has lines as such:

SSJ4_00089026.dpx SSJ4_00089134.dpx SSJ4_00090253.dpx SSJ4_00090267.dpx SSJ4_00090379.dpx SSJ4_00090392.dpx SSJ4_00090492.dpx SSJ4_00094320.dpx

and another file #2 that has lines as such:

693d092aba91a6acd2b8d73903fe4f78 SSJ4_00089026.dpx 0333b53c486de6a7214c05fd583f9745 SSJ4_00089134.dpx fdc713709251e0538a84549355627e65 SSJ4_00090253.dpx a8766ba5d6e1ee854bf7db10c05c7e55 SSJ4_00090267.dpx 733f6564ca442974555a76b91d506cff SSJ4_00090379.dpx f6154ddf1d7ae7c8076f08c25c9fe2c3 SSJ4_00094320.dpx

All of the lines in #1 are in #2, but they don't have the hash. Not all of the lines in #2 are in #1.

I want to pull all the lines that are #1 from #2 including the hash and put them into a new file, #3. So that #3 would look like:

ccfd092aba91a6acd2b8d73903fe4f78 SSJ4_00089026.dpx 056nb53c486de6a7214c05fd583f9745 SSJ4_00089134.dpx 45g313709251e0538a84549355627e65 SSJ4_00090253.dpx

etc

I'm thinking some command with grep such as grep -v -f -i file#1 file#2 >> file#3 or maybe awk

Upvotes: 2

Views: 3838

Answers (3)

shellter
shellter

Reputation: 37268

grep -iFf file1 file2 > file

you need to tell grep that it is in fgrep mode with the -F option, then the -f specifies what file to read from.

Note that I have changed your >> redirect (append) to > (create). You'll trip yourself up in testing using >>, as your first tests will always appear at the top of the file, and if you're rushing you won't think it is working. Use > for development and if you really need append mode, then add it after you are certain you basic cmd is working as required.

Finally, I'd use the -i (ignore case) option sparingly. If you really need to match lower case versions of your target strings, better to include those in your file1, so your process is self documenting.

IHTH

Upvotes: 4

BlooB
BlooB

Reputation: 965

NOTE: It was pointed out in the comments correctly that grep will have problems if you stick to having fields.

do this grep -Fwf file2 file1 > out . ,or use

awk 'NR==FNR{pats[$0]; next} $2 in pats' File2 File1

Upvotes: 4

Hakan Baba
Hakan Baba

Reputation: 2025

cat file1 | xargs -I@ -n1 grep @ file2 | tee file3

Print file1. Using xargs make each line of file 1 the pattern to grep call. Write the output to file3

Input file1:

SSJ4_00086400.dpx
SSJ4_00086403.dpx

Input file2:

693d092aba91a6acd2b8d73903fe4f78  SSJ4_00086400.dpx
0333b53c486de6a7214c05fd583f9745  SSJ4_00086401.dpx
fdc713709251e0538a84549355627e65  SSJ4_00086402.dpx
a8766ba5d6e1ee854bf7db10c05c7e55  SSJ4_00086403.dpx
733f6564ca442974555a76b91d506cff  SSJ4_00086404.dpx
f6154ddf1d7ae7c8076f08c25c9fe2c3  SSJ4_00086405.dpx

Output file3:

693d092aba91a6acd2b8d73903fe4f78  SSJ4_00086400.dpx
a8766ba5d6e1ee854bf7db10c05c7e55  SSJ4_00086403.dpx

Upvotes: 1

Related Questions