Offline mode new boards+components.

[tyeth]

1. This adds the xiao and espressif s3 devkit boards, the sense camera addon (SD), and expansion board (rtc/sd/oled/grove).

2. Adds the drivers for the omron d6t, plus PCF8563 RTC on xiao grove expansion board (part of RTCLib already)

3. Adds the INA228, and does the INA 260/237/238 refactor, linked to Ina238 + 237 via a refactor of INA260 too. #765

4. Updates the build CI step inline with main, for the online job, and the offline job has been similarly updated.

5. Adds a build define for wokwi, to allow switching the platformio sketch file (src/sketch) between network adapters.

6. Updates the autoconfig flag to also support globally disabling the auto config. Reason being it can crash repeatedly when attempting to initialise undefined components so the user never sees the error nor has time to remove the config file (except fast repeated shell commands to delete/remove the file).

Hey Brent, the auto config routine tried to initialise something (an unknown i2c rtc) as a vcnl4200 then rebooted. I couldn't disable the auto config globally and it scans then runs for found+unused i2c addresses.
So I've added support for autoConfig: false at the exportedDeviceConfig object level, so it can be specified per component or globally.
The XIAO has an rtc that's nearly the same so I've added that too (PCF8563 instead of -23), including now avoiding scanning that address (0x51) like the other rtc and mux address checks (0x68 + 0x70).

Maybe we could be smarter about the scanning of those addresses and still do so if the user has an already equivalent RTC/Mux component defined (e.g. if already defined soft rtc then they could auto scan 0x68 and 0x51, if defined pcf8523 then that frees up 0x51 for scanning, if a mux at not 0x70 then we can scan 0x70 looking for normal device components safe in the knowledge the only mux is at 0x71 or whatever the user defined)

[tyeth]

Interesting to note that the wokwi tests don't run on external contributor PRs.

[tyeth]

I'd missed the HDC302x, found in testing. Now all running as expected, SCD40 (0x62), HDC3022 (0x47), INA228 +237 +238 + 260.
Run on Xiao S3, with local platformIO build.

{"i2c_address":"0x44","timestamp":1755184910,"value":3.319726,"si_unit":"V"}
{"i2c_address":"0x44","timestamp":1755184910,"value":3.032684,"si_unit":"mA"}
{"i2c_address":"0x40","timestamp":1755184910,"value":3.31875,"si_unit":"V"}
{"i2c_address":"0x40","timestamp":1755184910,"value":2.13623,"si_unit":"mA"}
{"i2c_address":"0x41","timestamp":1755184910,"value":3.31875,"si_unit":"V"}
{"i2c_address":"0x41","timestamp":1755184910,"value":3.051758,"si_unit":"mA"}
{"i2c_address":"0x45","timestamp":1755184910,"value":3.3175,"si_unit":"V"}
{"i2c_address":"0x45","timestamp":1755184910,"value":2.5,"si_unit":"mA"}
{"i2c_address":"0x62","timestamp":1755184912,"value":32.10574,"si_unit":"C"}
{"i2c_address":"0x62","timestamp":1755184912,"value":89.79034,"si_unit":"F"}
{"i2c_address":"0x62","timestamp":1755184912,"value":43.6408,"si_unit":"%"}
{"i2c_address":"0x62","timestamp":1755184912,"value":625,"si_unit":"ppm"}
{"i2c_address":"0x47","timestamp":1755184915,"value":29.18975,"si_unit":"C"}
{"i2c_address":"0x47","timestamp":1755184915,"value":84.57999,"si_unit":"F"}
{"i2c_address":"0x47","timestamp":1755184915,"value":53.8323,"si_unit":"%"}

config.json

[brentru]

Few comments, overall a very strong PR!

[brentru]

LGTM

[brentru]

Oh, check that one failing check before merging