find files which contain a string

[bio_informatics]

Hi members,

Pardon me if this looks a trivial doubt.
I know this has been asked several times on stackoverflow, unix and many other forums.

I'm having a hard time, finding file using 'find'. I have tens of folders and each of them have another tens of folders, and so on.
The name of folders and the file I am interested to find are way to long.

Isolate_96_CN_11_B21_M1_C3_P2_GGATTAGG_L001_R2_001.fastq.gz

Example:

I want to find file which has "96_CN_11_B21_M1_C3_P2".

find . -name "*.fastq.*" | xargs grep "96_CN"

 find . -name "*.fastq.gz" -exec egrep -Hn "96_CN_" {} \;

Above commands take ages. I have waited more than 90 mins for the output and it still was processing.

Please guide here.
Again, sorry for this small query.

[GenoMax]

I assume you are looking to locate file names/paths?

If you are going to be needing to do this frequently perhaps creating a database of file/folder names like so

$ updatedb --require-visibility 0 -U /folder_hierarchy_to_search -o mydb

Then you can use "locate" command to find file names/paths very rapidly

$ locate -d mydb filename_to_search

Creating the database could take a significant amount of time but would be worth the returns.

[bio_informatics]

Hi Genomax,
Thank you for your reply.
I would be working on files for few months, then new data - new files names.
Creating database is a nice idea but not worth when new data floods in on regular basis.

I will post it on some other forum. (Will mention this post as well)

[GenoMax]

This is not a regular database and may be the fastest way of finding files (http://linux-sxs.org/utilities/updatedb.html) and http://en.wikipedia.org/wiki/Locate_%28Unix%29.

You could look into running this as a cron job each day so you would not need to worry about it.

[bio_informatics]

Thank you for your valuable suggestions. I didn't know if anything of this type existed.
I'm on a cluster and do not have much rights.
I will definitely take these suggestions into consideration.

[bio_informatics]

I'm doing wrong here:

find . -name "*.fastq.*" | xargs grep "96_CN"

It finds .fastq files and in them, it tries to look/grep "96_CN".
fastq.gz are binary files and it would definitely take ages to grep.

Didn't try this on small directories first.

--
The correct one I got as:

time find . -name "*96_CN_11_B21_M1_C3_P2*"

Got path and file names needed with time.

real 0m0.434s
user 0m0.037s
sys 0m0.152s

[GenoMax]

Can you try this?

$ find . -type f -name "*.fastq.*" | grep "96_CN"

[bio_informatics]

Genomax:
I was making a horrible error in my command.
Morning with fresh mind picked it up instantly.

[GenoMax]

Consider adding "-type f" to your find command since you are only looking for files.

[bio_informatics]

That worked.

Thanks much for your help and time.