Feat/formatting

[andy9a9]

• extend extension by formatting feature
  • use 80 chars as a default line length
    • show warnings for longer lines with comments
  • 8 tabs or 8 spaces
• format whole document only
  • maybe add range support in the future
  • try to use Linux DeviceTree standard for indents, comments, arrays, ...
    • tested on common setup, but there could be some issues for sure due to lack of tested variants :)
• add formatting unit-tests

[andy9a9]

Local OSX tests are ok:

DeviceTree Extension Test Suite
    ✔ Extension should be present
    ✔ Extension should activate
    Language Support Tests
      ✔ DeviceTree language should be registered
      ✔ DeviceTree files should be recognized
    Formatter Tests
      ✔ formats input.dts with tabs
      ✔ formats input.dts with spaces
  6 passing (39ms)

Windows has to be evaluated ...

// EDIT: fixed by enforcing line endings to LF

[mkessler001]

I added some comments.
Another thing that came to my mind: wouldn't it be possible to show warnings BEFORE a file is formatted?

[mkessler001]

One last thing before we merge this: Should the comment block at the end of a line also be considered for its length?
Let's say we specify the max. line length with 80, and we have the following line:

	pinctrl_ctrl_sleep_moci: ctrlsleepmocigrp {
		fsl,pins =
			<MX8MP_IOMUXC_SAI3_RXC__GPIO4_IO29		0x1c4>;	/* SODIMM 256 */
	}

The semicolon is in character position 79, and the space is in position 80. However, the diagnosis shows a warning as the line exceeds 80 characters (current 97). Is that valid?

[andy9a9]

	pinctrl_ctrl_sleep_moci: ctrlsleepmocigrp {
		fsl,pins =
			<MX8MP_IOMUXC_SAI3_RXC__GPIO4_IO29		0x1c4>;	/* SODIMM 256 */
	}

The semicolon is in character position 79, and the space is in position 80. However, the diagnosis shows a warning as the line exceeds 80 characters (current 97). Is that valid?

I think so it's correct, because...

/dts-v1/;

/ {
<tab8><tab8>pinctrl_ctrl_sleep_moci: ctrlsleepmocigrp { // line length: 8 + 8 + 23 + 1 +1 + 16 + 1 + = 59 (60)
<tab8><tab8>fsl,pins = // line length: 8 + 8 + 8 + 1 + 1 = 26 (27)
<tab8><tab8><tab8><MX8MP_IOMUXC_SAI3_RXC__GPIO4_IO29<tab6><tab8>0x1c4>;<tab1>/* SODIMM 256 */ // line length: 8 + 8 + 8 + 1 + 33 + 6 + 8 + 5 + 1 + 1 + 1 + 16 = 96 (97)
<tab8>} // line length: 8 + 1 = 9 (10)
};

So I would say yes, diagnostic is showing the correct number :). And of course if you change it to spaces, then the number is re-calculated correctly.

So are we ready?