Search for a keyword and get its count occurrence

Question

I want to find out the occurrence of a keyword group by filename inside a directory.

Example:
Keyword - TEST (means TEST1, TEST2, etc. can be valid results)
Files - a.txt, b.txt etc. (inside a directory .. /tmp).

Output will be as follows.

a.txt - 4
b.txt - 5
c.txt - 0

I tried with grep -o "TEST" a.txt and this gives only TEST and its occurrences; like if a file has 3 occurrences (like TEST1, TEST_XXX, TESTXYES) the output will be

TEST
TEST
TEST

OR if multiple files search is not possible then OUTPUT should provide all the similar string starting TEST; for example,

TEST
TEST_1
TEST_2
TEST4

I tried the following, too.

grep -c "TEST" a.txt

but this gives only the count; it does not give what was the entire text starting with TEST.

Answer 1

The grep -c prints number of matching lines, so you won't get right result in combination with -o option. I would pipe the output of grep -o to uniq -c:

$ echo 'test1 test2 test3' >/tmp/file1
$ echo 'test1 test2 test3 test4' >/tmp/file2
$ grep -o test /tmp/file* | uniq -c
  3 /tmp/file1:test
  4 /tmp/file2:test

You can pipe it further to modify the output with some text-processing tools.