Bump copyright year

src: Import and run lang_main ...

This is the bare minimum needed to run this on real hardware.

Clock: Add ability to convert from microseconds to kernel time ...

This functionality allows us to ask the kernel to schedule events
based on a high resolution real-time clock.

An example in main waits until the next second.

Clock: Combine clocks to create a high resolution real-time clock ...

The kernel clock is high resolution but only tracks uptime.
The RTC clock is low resolution but tracks real time.
The NTP clock is high resolution but costly to access.

We can combine the kernel clock with a real-time clock of any type
by sporadically calculating the time the system was booted at then
adding the kernel's uptime to it whenever we need a new clock.

Clock: Clean up and document the code

Clock: Use microseconds instead of struct timeval ...

timevals are a little annoying to work with compared to just microseconds,
so switch to using them for our clock API.

Clock: Print microseconds in ISO 8601 strings

Clock: Add _now suffix and variants for clock functions ...

Currently the clock code implicitly grabs the current time.
Add variants that let you pass in your own time value, and specify
extra _now functions that use the current time.

main: Disable color tracing ...

Color tracing was really messing with grabserial and minicom.

NetworkHandler: make status callback private, reorder functions ...

The status callback function now gets privately declared at the earliest
possible moment, to both allow for reordering of code and avoid calling
this function directly from any other part of the code

* void changed to static void in status_callback declaration
* this now makes it possible to define network_init before
status_callback, as status_callback is declared much earlier

main: Setup and sync clock when 3 presses are read

Clock: Add our clock ...

This clock is intended to be used to get a timeval with sub-second
accuracy and convert it back and forth between kernel and real time.

NetworkHandler: All access to the underlying NetworkInterface ...

We need this to actually use the network.

NetworkHandler: replace redundant include of EthernetInterface ...

The NetworkHandler header already includes EthernetInterface.h, so this
was redundant. It's also better in the future to not have the
NetworkHandler code directly require EthernetInterface.h, to allow for
versatile usage.

NetworkHandler: added a new function to check weather we're connected ...

This now uses an atomic boolean to be thread safe.

Added networking support ...

Networking is handled by the NetworkHandler portion of the code, which
provides a few functions for use by other parts of the program, and a
status callback to monitor for connectivity changes. Connectivity is
also handled in an asynchronous mode, to avoid freezing the entire OS
while actions are performed.

* network_connect should be called by any part of the program wishing to
establish connectivity
* network_disconnect should be called to disconnect the network once
done
* network_init is a special function which should only be called by
main. It is responsible for setting the network interface in
asynchronous mode as well as attaching the status monitor callback. No
other part of the program should ever call this.

Lwip is used for the network stack, and IPv6 support has been enabled,
and prefered over IPv4 if available.

main: Only enable tracing if it's enabled in Mbed ...

GCC will whine about unused variables if we try to set up tracing and
Mbed implements it as a no-op. Wrap this in an a compile-time check.

Make use of the mbed tracing subsystem instead of basic printing ...

This enables use of the mbed tracing subsystem for the entire source
code of the project. All the uses of printfs are now appropriately
tagged and assigned a trace level, as in the following example:

[INFO][FS]: Unmounting the filesystem...
[INFO][FS]: OK
[INFO][USBM]: Switching to the usb mass storage mode...
[INFO][USBM]: Disconnecting the usb mass storage device...
[INFO][USBM]: Switched usb mass storage mode off.
[INFO][FS]: Mounting the filesystem...
[INFO][FS]: OK

This was also done in a way that the tracing is thread safe, using a
mutex to lock and unlock as needed.

Note that this cannot be turned on or off yet during build. It is
currently always enabled.

src/main: reworked the main loop ...

With the latest code cleanup, issues were introduced regarding the mounting and
unmounting of the filesystem.

* Sometimes the board would wake up from sleep and perform mounting and
unmounting of the filesystem in a more or less short loop before going
back to sleep. I suspect this might be due to the sleep getting
interrupted without any button press, aka every 10 minutes or so.
* This code also created a bug where it would create then mount a new
filesystem on the flash after erasing it, breaking the purpose for that
function in the first place which is to erase the flash and nothing
else.

Main is now calling mountFilesystem and unmountFilesystem where
appropriate:

* Before running doUSBMSD, main unmounts the filesystem. When doUSBMSD
is done, main once again mounts the filesystem.
* Before erasing the flash with doERASE, main unmounts the filesystem,
and never mounts it again.
* As a tidying up step, the filesystem gets unmounted, same as the
button thread terminating.

main: Unmount filesystem on shutdown

ButtonThread: Use a more descriptive error message

src: Don't pass FileSystem to doErase and doUSBMSD ...

They don't need the filesystem now, so get rid of this arg.

main: Handle mounting and unmounting for erase and MSD ...

Having the functions handle the filesystem creates an dependency
on the filesystem layer when the functions don't even use it.

Instead, have main handle mounting and unmounting the filesystems.

Filesystem: Refer to mbed::FileSystem not LittleFileSystem2 ...

Functions don't need to know about what kind of filesystem we pass it,
so just tell them it's a mbed::FileSystem.

ButtonThread: Remvoe unused flag

Added SPDX license informations to all files. ...

The only files where this isn't present are the LICENSE.md for obvious
reasons and the mbed_app.json, as comments are not allowed in json.

---
Changes v1 -> v2:
- fixed capitalisation of the copyright word in the script directory and in the cmake list
- modified the source code and header files so the SPDX info is
in a multi-line comment block

Code format

Initial commit