When B>1, fsNext failed because HL was changed in the Seek call.

On build, pack `parts/z80` into a CFS and embed it into the emulated
zasm executable as an fsdev. This will allow for the upcoming include
directive to have something to go to.

For now, this is useless except for inflating the emulated zasm's size.

This 'apps' folder is going to disappear. Everything is going to be a
"part" to be assembled through recipes. 'apps' has no meaning.

Instead of buffering input in memory one line at a time, we go in "just
in time" mode and always read contents directly from I/O, without
buffering.

It forces us to implement a `ioPutback` scheme, but on the other hand it
greatly simplifies cases where multiple tokens are on the same line
(when a label is directly followed by an instruction).

The end result feels much more solid and less hackish.

We'll need it for an upcoming "local labels" registry.

Will need it soon.

It was previously skipped on a local label.

That is, labels with code following right afterwards.

With local labels, these two will have to be decoupled.

Facilitates debugging

Gives clearer output on errors. Easier debugging...

`xor c` didn't do what I thought it did at the time. Newbie mistake,
there are probably a lot like this scattered around...

Also, add warning on attempts to write to ROM.

scas didn't warn about this, but it seems that `jr` calls were out of
range. Seek and Tell was broken. Replacing `jr` with `jp` does the
trick.

With the help of the newly-introduced cfspack tool, we can mount a
filesystem in our emulated shell and play around. Read-only for now.
Unpacking incoming.

Also, confirm in tests that we can refer to other symbols in `.equ`.

Also, add a port for rewinding stdin.