Retargeting to PiOS/Python3/PyQt5

[n0bel]

Thanks to @SerBrynden we're moving down the road to python3/PyQt5. I'm also moving to PiOS (bullseye) (PiOS starting 2021-10-30)

The upgrade path may be difficult for those on older versions of Raspbian/python2. Not sure how to handle these issues yet, to make it easy to upgrade without starting over.

The python3/PyQt5 testing branch is Python3-SerBryden

/K

[taftr1]

Using the latest version of Raspberry Pi OS (raspberrypi 5.10.92-v8+ #1514) I'm using the python3-SerBrynden Branch. After working out the changes between the install of the Master Branch and this one, I am getting the following when I attempt to do a First Start. It is possible I missed something... I did Blank out my Keys and location in this output

pi@raspberrypi:~/PiClock $ sh startup.sh -n -s Disabling screen blanking.... Setting sound to max (assuming Monitor Tv controls volume).... Checking for NeoPixels Ambilight... 1493 1496 Checking for GPIO Buttons... Checking for Temperature Sensors... Starting PiClock.... logging to screen. libpng warning: iCCP: known incorrect sRGB profile libpng warning: iCCP: known incorrect sRGB profile libpng warning: iCCP: known incorrect sRGB profile map base url for radar1: http://maps.googleapis.com/maps/api/staticmap?key=KEYBLANKED&center=MYLAT,-MYLON&zoom=7&size=266x244&maptype=hybrid map base url for radar2: http://maps.googleapis.com/maps/api/staticmap?key=KEYBLANKED&center=MYLAT,-MYLON&zoom=11&size=266x244&maptype=hybrid map base url for radar3: http://maps.googleapis.com/maps/api/staticmap?key=KEYBLANKED&center=MYLAT,-MYLON&zoom=7&size=622x622&maptype=hybrid map base url for radar4: http://maps.googleapis.com/maps/api/staticmap?key=KEYBLANKED&center=MYLAT,-MYLON&zoom=11&size=622x622&maptype=hybrid libpng warning: iCCP: known incorrect sRGB profile metar url: https://tgftp.nws.noaa.gov/data/observations/metar/stations/KFWN.TXT getting current: Thu Feb 17 23:00:04 2022 https://api.climacell.co/v3/weather/realtime?apikey=KEYBLANKED&lat=MYLAT&lon=MYLON&unit_system=us&fields=temp,weather_code,feels_like,humidity,wind_speed,wind_direction,wind_gust,baro_pressure getting hourly: Thu Feb 17 23:00:04 2022 https://api.climacell.co/v3/weather/forecast/hourly?apikey=KEYBLANKED&lat=MYLAT&lon=MYLON&unit_system=us&fields=temp,precipitation,precipitation_type,precipitation_probability,weather_code getting daily: Thu Feb 17 23:00:04 2022 https://api.climacell.co/v3/weather/forecast/daily?apikey=KEYBLANKED&lat=MYLAT&lon=MYLON&unit_system=us&fields=temp,precipitation_accumulation,precipitation_probability,weather_code wxstart for radar1 wxstart for radar2 Traceback (most recent call last): File "/home/pi/PiClock/Clock/PyQtPiClock.py", line 793, in wxfinished_cc wxdata = json.loads(wxstr) File "/usr/lib/python3.9/json/init.py", line 346, in loads return _default_decoder.decode(s) File "/usr/lib/python3.9/json/decoder.py", line 337, in decode obj, end = self.raw_decode(s, idx=_w(s, 0).end()) File "/usr/lib/python3.9/json/decoder.py", line 355, in raw_decode raise JSONDecodeError("Expecting value", s, err.value) from None json.decoder.JSONDecodeError: Expecting value: line 1 column 1 (char 0) Aborted

[SerBrynden]

Climacell.co changed to Tomorrow.io and their v3 API was replaced with the v4 API. I don't think the v3 API works anymore and, last I checked, the PiClock code hasn't been updated to support the v4 API. I use OpenWeatherMap.org.

[taftr1]

@SerBrynden

That was it.... Thanks for your Help

[taftr1]

So seeing 2 issues now, one is brand new. The first is there seems to be some problem displaying the Radar Overlays. They come up all garbled and mirrored is the best way I can describe it. I am seeing "libpng warning: iCCP: known incorrect sRGB profile" running through the logs in sets of 3.

This morning a New error popped up. This causes PiClock to Crash. Its referencing "Compass" in the error. Near as I can tell its something to do with decoding the METAR. Originally I thought it had something to do with the METAR I was using not being available or missing. However I was able to pull the file up and restarting PiClock caused the same issue. I tried using a Different Airport code, utilizing the nearest international airport, thinking its format would not have the same error. However the same problem happens. the code referenced is Line 1178 and is "wd = f.wind_dir.compass()" Unfortunately my programing knowledge is limited here and cant seem to find where is defined.

libpng warning: iCCP: known incorrect sRGB profile libpng warning: iCCP: known incorrect sRGB profile libpng warning: iCCP: known incorrect sRGB profile map base url for radar1: https://api.mapbox.com/styles/v1/bcurley/cj712peyz0bwr2sqfndbggupb/static/LAT,LON,6,0,0/266x244?access_token=TOKEN&attribution=false&logo=false map overlay url for radar1: https://api.mapbox.com/styles/v1/bcurley/cj712r01c0bw62rm9isme3j8c/static/LAT,LON,6,0,0/266x244?access_token=TOKEN map base url for radar2: https://api.mapbox.com/styles/v1/bcurley/cj712peyz0bwr2sqfndbggupb/static/LAT,LON,10,0,0/266x244?access_token=TOKEN&attribution=false&logo=false map overlay url for radar2: https://api.mapbox.com/styles/v1/bcurley/cj712r01c0bw62rm9isme3j8c/static/LAT,LON,10,0,0/266x244?access_token=TOKEN map base url for radar3: https://api.mapbox.com/styles/v1/bcurley/cj712peyz0bwr2sqfndbggupb/static/LAT,LON,6,0,0/622x622?access_token=TOKEN&attribution=false&logo=false map overlay url for radar3: https://api.mapbox.com/styles/v1/bcurley/cj712r01c0bw62rm9isme3j8c/static/LAT,LON,6,0,0/622x622?access_token=TOKEN map base url for radar4: https://api.mapbox.com/styles/v1/bcurley/cj712peyz0bwr2sqfndbggupb/static/LAT,LON,10,0,0/622x622?access_token=TOKEN&attribution=false&logo=false map overlay url for radar4: https://api.mapbox.com/styles/v1/bcurley/cj712r01c0bw62rm9isme3j8c/static/LAT,LON,10,0,0/622x622?access_token=TOKEN libpng warning: iCCP: known incorrect sRGB profile libpng warning: iCCP: known incorrect sRGB profile libpng warning: iCCP: known incorrect sRGB profile libpng warning: iCCP: known incorrect sRGB profile metar url: https://tgftp.nws.noaa.gov/data/observations/metar/stations/K12N.TXT getting current and forecast: Fri Feb 25 10:04:33 2022 https://api.openweathermap.org/data/2.5/onecall?appid=KEY&lat=LAT&lon=LON&units=imperial&lang=en&r=0.4262510323217398 wxstart for radar1 wxstart for radar2 wxmetar: K12N 251454Z AUTO VRB05KT M01/M03 A2995 RMK AO1 SLP154 60017 T10111028 58037 QPixmap::scaled: Pixmap is a null pixmap QPixmap::scaled: Pixmap is a null pixmap Traceback (most recent call last): File "/home/pi/PiClock/Clock/PyQtPiClock.py", line 1178, in wxfinished_metar wd = f.wind_dir.compass() AttributeError: 'NoneType' object has no attribute 'compass' Aborted

[taftr1]

Tried again today, and the Crashing and alert about "Compass" have stopped. However the Image corruption continues

[n0bel]

I don't see why it has that error. K12N works fine for me in PiClock and PiClock3. Perhaps an upgrade...

sudo pip3 install python-metar --upgrade

[taftr1]

I ran the command you provided with the following results:

pi@raspberrypi:~ $ sudo pip3 install python-metar --upgrade Looking in indexes: https://pypi.org/simple, https://www.piwheels.org/simple Requirement already satisfied: python-metar in /usr/local/lib/python3.9/dist-packages (1.4.0) Collecting python-metar Using cached https://www.piwheels.org/simple/python-metar/python_metar-1.4.0-py3-none-any.whl (18 kB)

I restarted PiClock and it is working. However this error seems to happen after its been running for a while. I'll let you know should it fault again.

I still see the problem with the radar overlays being garbled. Its particularly visible on the second page where it looks like is random noise along with the radar overlays being mixed on both maps. it seems to get worse over time with images from the previous update being mixed together with new update s

[taftr1]

Still getting that same crash related to "compass" It seems that it happens when a Wind direction is not included in the Metar Such as: wxmetar: K12N 082154Z AUTO VRB03KT 07/M09 A3013 RMK AO1 SLP208 T00721089 QPixmap::scaled: Pixmap is a null pixmap QPixmap::scaled: Pixmap is a null pixmap Traceback (most recent call last): File "/home/pi/PiClock/Clock/PyQtPiClock.py", line 1178, in wxfinished_metar wd = f.wind_dir.compass() AttributeError: 'NoneType' object has no attribute 'compass' Aborted

[taftr1]

Following up. When the Metar changed to the following for Wind the crash stopped. Not sure how this is working with python-metar, as I'm not a Python programmer

2022/03/09 19:54 K12N 091954Z AUTO 00000KT 02/00 A2990 RMK AO1 SLP134 P0008 T00170000

[taftr1]

So After some research I am able to duplicate the error. Seems to be a Bug in the python-metar code relating to the wind_dir.compass variable. If the wind is designated as something like "VRB03KT" it throws that error. Its related to the VRB (variations in wind direction greater than or equal to 30 degrees).

Here is a little Test program I put together. If you change the Comment for the sample METAR you can make the error happen or not.

from __future__ import print_function

from metar import Metar

# A sample METAR report
code = (
#    "K12N 121754Z AUTO 36007G16KT 330V040 M01/M04 A2931 RMK AO1 SLP934 P0001 60042 T10111044 10056 21017 56042"
    "K12N 082154Z AUTO VRB03KT 07/M09 A3013 RMK AO1 SLP208 T00721089"
)

print("-----------------------------------------------------------------------")
print("METAR: ", code)
print("-----------------------------------------------------------------------")

# Initialize a Metar object with the coded report
obs = Metar.Metar(code)

# Print the individual data

# The 'station_id' attribute is a string.
print("station: %s" % obs.station_id)

print("Compass: " + obs.wind_dir.compass())

print("-----------------------------------------------------------------------\n")```

[taftr1]

So I opened a Issue with the Python Metar Github. Seems that it is indeed an issue with that code. They aren't quite sure yet how to fix it. But there is a Work around...

If you Replace the 2 occurrences of:

wd = f.wind_dir.compass()

with:

if f.wind_dir is None: wd = " " else: wd = f.wind_dir.compass()

That will set the wind direction to a blank in the cases that it isn't specified or VRB. This will stop the crashing at least until a more elegant method is developed.

I am still having issues with the Radar overlay. I'm not quite able to figure that out....