Upload to aws s3 using cli without folder path

Question

I'd like to upload a file.txt on aws s3 that is located in something like main/part1/part2/file.txt , where part1 and part2 are unknown (those folders always change). I can do that with the command aws s3 cp./main s3://mybucket --exclude "*" --include "*.txt" --recursive , but then in my bucket I have the file located in part1/part2/file.txt . I'd like file.txt to be at the base of the bucket, not inside part1/part2

Is that possible given that part1 and part2 are constantly changing?

Answer 1

for dir1 in $(ls main); do
    for dir2 in $(ls main/$dir1); do
        aws s3 cp ./main/$dir1/$dir2/ s3://my-bucket --exclude "*" --include "*.txt" --recursive
    done
done

upload: main/part1/part2/file.txt to s3://my-bucket/file.txt
upload: main/part11/part22/file2.txt to s3://my-bucket/file2.txt

Answer 2

The following will work if main will never contain more than 1 subdirectory at a time ( part1 ) & that subdirectory in-turn will never contain more than 1 subdirectory at a time ( part2 ):

aws s3 cp ./main/*/*/ s3://my-bucket --exclude "*" --include "*.txt" --recursive

upload: main/part1/part2/file.txt to s3://my-bucket/file.txt