Re: search text file

This shell command almost works, but not quite: it actually counts the number of lines each string matches, so if a string can occur more than once in a line you'll get a wrong answer.

( while read; do grep -cFe "$REPLY" secondfile; done ) < firstfile
[download]

Comment on Re: search text file Download Code

Replies are listed 'Best First'.
Re^2: search text file by jwkrahn (Abbot) on Jul 27, 2011 at 10:45 UTC
To count all the matches from the command line: `grep -oF -f firstfile secondfile \| sort \| uniq -c` [download]	[reply] [d/l]
Re^3: search text file by ambrus (Abbot) on Jul 27, 2011 at 11:51 UTC
Ah, good idea using `grep -o`. That does indeed find multiple matches in a single line. That, however, won't work correctly if some of the matches are overlapping. Eg. if the second file has `abcdef` and the first file has the two strings `abcd` and `cdef`, grep will only find the `abde` part. As a workaround, you could run grep once for each string in the first file. Thus, we get (I think) `( while read; do grep -oFe "$REPLY" secondfile \| wc -l; done ) < first +file` [download]	[reply] [d/l] [select]