Find, grep and then create table

Question

I am trying to find all *.md at ${PWD}, then grep for lines beginning with title: and print it along with the absolute filepath -H and the line number of the pattern match -n, then I try to output the result in a table with separator -s as :

find ${PWD} -type f -name "*.md" -exec bash -c 'i="$1"; grep -HnE "^title:" $1; | column -t -s :;' _ {} \;

This is working till grep but I am unable to create a table.

Also, from the correction suggested by @steeldriver, the exec seems to be happening on each find result one by one, thereby table is not being created?

---

Sample Output

/Admin/Specification/Specification.md    2  title   Specification
/Admin/GraphicCard/index.md              2  title   "GraphicCard"

Answer 1

You might be overthinking things. Rather than trying to tabulate the results of grep, why not let awk, a prolific table-maker, make the table for you?

find . -type f -name \*.md -print0 | xargs -0 awk 'BEGIN { OFS="\t" } /^title:/ { print FILENAME, FNR, $0 }'

find will locate all the md files you're interested in, and pass them all as arguments to awk by way of xargs. We then just have awk spit out, on lines that match your criterium, the name of the file (which find will have passed with the full relative path), the record number in the file (i. e. the line number), and the line as seen in the file; all separated by tabs as defined in OFS, the Output Field Separator.

If you want to be a little more exact with the alignment of your output, you could instead do something like:

/^title:/ { printf( "%35s %4d %s\n", FILENAME, FNR, $0 ) }

Also, if you don't want to find | xargs awk but would rather find -exec, this also works:

find . -name \*.md -type f -exec awk '/expiry/ { printf( "%30s %4d %s\n", FILENAME, FNR, $0 ) }' "{}" +