How to generate a text report in Linux?

Question

I have a command I'm using to find all of the *.xml files on our server. I'm using:

find -type f -exec grep -il "xml" {} \;

This is great, but there are maybe a thousand files like this on the server, and my window fills up.

I would like to be able to generate a report and have it save somewhere on the server or even email to me.

I would also like to specify to look only in the last 5 days when it generates this report.

Answer 1

Let's start with saving the file somewhere first.

Take your command:

find -type f -exec grep -il "xml" {} \;

and read through: What are the shell's control and redirection operators? to where it says:

> : Directs the output of a command into a file.

to make your command something like:

find -type f -exec grep -il "xml" {} \; > /tmp/report.txt

I'll take this opportunity to recommend a couple tweaks to your current find command:

1. Might be a personal preference (or legacy memory on my part), but I like to specify the starting directory for find, like so:

   find / -type f ....

so that it doesn't matter where you start the command or script from, it will always produce the expected results.

2. Your find command is not searching for "*.xml" files, it's searching for files that contain the (case-insensitive) string "xml". If contents-searching is what you want, then stop reading this paragraph. If you actually want to find files that have "xml" in their filename, then look for something like: find / -type f -iname '*xml*' -- where iname tells find to match for filenames that contain (case-insensitive) "xml". If you only care about "xml" being at the end of the filename, then use find / -type f -iname '*xml' or ... -iname '*.xml, if you want a period before the xml.

2a. If you are wanting to search the contents of the files, consider changing your general-case grep to one that tells grep that you're searching for a fixed string: ... grep -Fil xml ... to speed things up a tiny bit.

3. To tell find to only report files that were modified in the past 5 days, use -mtime -5.

4. Since you tagged the Q as Linux, I'll assume that you have GNU find (man find and look for "GNU" in the Description to confirm). In that case, can can speed the process up a bit more by telling find to pass in multiple files to each exec by using this syntax: find / -type f -mtime -5 -exec grep -Fil xml {} + > /tmp/report.txt