proof of concept for ocaml/RFCs#19
The multicat.sh tool is like cat, but provides options to encode/decode various things.
$ echo "hello\nworld\n🤗" > foo
$ ./multicat.sh foo foo foo -rot13
uryyb\ajbeyq\a🤗
uryyb\ajbeyq\a🤗
uryyb\ajbeyq\a🤗
$ ./multicat.sh foo foo foo -rot13 > bar # save it into a file
$ ./multicat.sh bar -rot13 -rot13 -rot13 # triple decode!
hello\nworld\n🤗
hello\nworld\n🤗
hello\nworld\n🤗The following example reads a file, translates it via "rot13", then uses the http1.1 chunked encoding; then the rest of the pipeline decodes the chunked encoding and rot13.
Note that the pipeline runs in constant memory, it does not need to read the whole file to do its translations.
$ cat src/IO_helpers.ml
… # some content
$ ./multicat.sh src/IO_helpers.ml -rot13 -chunk
… # content, but less readable
$ ./multicat.sh src/IO_helpers.ml -rot13 -chunk | ./multicat.sh -unchunk -rot13
… # initial content of the file
-zip and -unzip don't work yet, I haven't updated them.