No?
1)
find(1) searches at any level for any name containing
63 AND n.
2)
find(1) searches for names containing a
n at each level of the directories that have a
63 in theirs.
In the second case you have:
[(
number of names matching 63) + (
number of names matching n in directories matching 63)
] unique result lines.
So, perhaps you got 34 directories which match
63 and contain a filename with a
n?
What does the following return?
Code:
find -name '*63*' -exec bash -c 'shopt -s nullglob; for r in "$1"/*; do [[ ${r##*/} = *n* ]] && echo "$r"; done' _ {} \; | wc -l