Question

How can I compute the intersection between two text files in terms of raw text? It doesn’t matter whether the solution uses a shell command or is expressed in Python, Elisp, or other common scripting languages.

I know comm and grep -Fxv -f file1 file2. Both assume that I am interested in the intersection of lines, whereas I am interested in the intersection of characters (with a minimum on the number of characters necessary to count as a match).

Bonus points for efficiency.

Example

If file 1 contains

foo bar baz-fee

and file 2 contains

fee foo bar-faa

then I would like to see

• foo bar
• fee

assuming a minimum match length of 3.

Answer 1

You’re looking for Python’s difflib module (in the standard library), and in particular difflib.SequenceMatcher.