dandri/meshcore-bot
1
0 Fork 0
mirror of https://github.com/agessaman/meshcore-bot.git synced 2026-11-04 11:57:16 +00:00
Code Issues Packages Projects Releases Wiki Activity
Files
7159169eb37f075a70abe75286a6c2ca014f51e9
meshcore-bot/modules/commands/wx_command.py
T
agessaman f994171818 feat(weather): [Weather] always_show_location names the place in every wx/gwx reply 
wx and gwx name the place only when it adds information: a city outside default_state or default_country, or a position. always_show_location (default false) names every place a lookup finds instead, including cities in the bot's own state and ZIP codes.

For a ZIP code, gwx now does a reverse lookup when the option is on (it skipped that since the name was never shown), and wx reuses the reverse lookup it already did for a companion position. A failed lookup leaves the reply unnamed rather than repeating the ZIP code. The prefix comes out of the message budget as before. The option is in the wx settings schema for the web viewer, the example config, the command reference and the changelog.
2026-10-02 07:55:18 -07:00

2928 lines
136 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
#!/usr/bin/env python3
"""
Weather command for the MeshCore Bot
Provides weather information using zip codes and NOAA APIs
"""
import asyncio
import re
import threading
import xml.dom.minidom
from datetime import datetime, timedelta, timezone
from typing import Callable, Optional, ParamSpec, TypeVar
import requests
from .. import alert_format
from ..clients.mqtt_weather import ( # noqa: F401 re-exported
get_mqtt_weather_topic,
load_mqtt_weather_format_config,
mqtt_weather_display_for_topic,
)
# First-party modules with only required dependencies; they always import.
from ..clients.wxsim_parser import WXSIMParser
from ..http_retry import make_retry_session
from ..location import get_bot_lat_lon, get_companion_lat_lon
from ..models import MeshMessage
from ..nws_alerts import WX_SPECIAL_RULES, entry_nws_headline, entry_summary, entry_title, parse_alert_fields
from ..nws_coverage import NWSNoCoverageCache
from ..utils import (
format_temperature_high_low,
geocode_city_sync,
geocode_zipcode_sync,
get_nominatim_geocoder,
normalize_us_state,
)
from ..weather_common import _ARROWS_8, _COMPASS_16, WeatherCommandMixin
from .alternatives.wx_international import GlobalWxCommand
from .base_command import BaseCommand
from .rain_command import nws_http_means_no_coverage
# Kept for code that checked them; these imports used to be optional.
WX_INTERNATIONAL_AVAILABLE = True
WXSIM_PARSER_AVAILABLE = True
# Multiday: plain digits (e.g. 7), 7day/7-day, or suffix form 7d/10d (min 2, max below).
WX_MULTIDAY_MAX_DAYS = 16
_P = ParamSpec("_P")
_T = TypeVar("_T")
_WEEKDAYS_LOWER = ('monday', 'tuesday', 'wednesday', 'thursday', 'friday', 'saturday', 'sunday')
# Multi-day line labels, Monday first (date.weekday() order); the catalog's
# commands.wx.day_abbrev translates them, and these are the English fallback.
_DAY_ABBREVS = ('M', 'T', 'W', 'Th', 'F', 'Sa', 'Su')
_WEEKDAY_NAMES = ('Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday')
# Forecast-text patterns for the extract_* readers, tried in order.
# extract_humidity: "humidity 45%" or "45% humidity"
_HUMIDITY_PATTERNS = (
r'humidity\s+(\d+)%',
r'(\d+)%\s+humidity',
r'relative humidity\s+(\d+)%',
r'(\d+)%\s+relative humidity',
)
# extract_precip_chance: "20% chance" or "chance of rain 30%"
_PRECIP_CHANCE_PATTERNS = (
r'(\d+)%\s+chance',
r'chance\s+of\s+\w+\s+(\d+)%',
r'(\d+)%\s+probability',
r'probability\s+of\s+\w+\s+(\d+)%',
)
# extract_uv_index
_UV_INDEX_PATTERNS = (
r'uv\s+index\s+(\d+)',
r'uv\s+(\d+)',
r'ultraviolet\s+index\s+(\d+)',
)
# extract_dew_point
_DEW_POINT_PATTERNS = (
r'dew point\s+(\d+)',
r'dewpoint\s+(\d+)',
r'dew\s+point\s+(\d+)°',
)
# extract_visibility
_VISIBILITY_PATTERNS = (
r'visibility\s+(\d+)\s+miles',
r'visibility\s+(\d+)\s+mi',
r'(\d+)\s+mile\s+visibility',
r'(\d+)\s+mi\s+visibility',
)
# extract_precip_probability
_PRECIP_PROBABILITY_PATTERNS = (
r'(\d+)%\s+chance\s+of\s+(?:rain|precipitation|showers)',
r'chance\s+of\s+(?:rain|precipitation|showers)\s+(\d+)%',
r'chance\s+of\s+(?:rain|precipitation|showers)\s+is\s+(\d+)%', # NOAA: "Chance of precipitation is 60%."
r'(\d+)%\s+probability\s+of\s+(?:rain|precipitation|showers)',
r'probability\s+of\s+(?:rain|precipitation|showers)\s+(\d+)%',
r'(\d+)%\s+chance',
r'chance\s+(\d+)%',
)
# extract_wind_gusts
_WIND_GUST_PATTERNS = (
r'gusts\s+to\s+(\d+)\s+mph',
r'gusts\s+up\s+to\s+(\d+)\s+mph',
r'wind\s+gusts\s+to\s+(\d+)\s+mph',
r'wind\s+gusts\s+up\s+to\s+(\d+)\s+mph',
r'gusts\s+(\d+)\s+mph',
r'wind\s+gusts\s+(\d+)\s+mph',
r'gusts\s+as\s+high\s+as\s+(\d+)\s+mph', # NOAA: "with gusts as high as 25 mph."
)
# extract_wind_gusts in SI forecast text ("Wind gusts up to 48 km/h")
_WIND_GUST_KMH_PATTERNS = tuple(p.replace(r"\s+mph", r"\s*km/h") for p in _WIND_GUST_PATTERNS)
# extract_pressure
_PRESSURE_PATTERNS = (
r'pressure\s+(\d+)\s*hpa',
r'pressure\s+(\d+)\s*mb',
r'barometric\s+pressure\s+(\d+)\s*hpa',
r'barometric\s+pressure\s+(\d+)\s*mb',
r'(\d+)\s*hpa',
r'(\d+)\s*mb\s+pressure',
)
def _has_temp(value) -> bool:
"""Whether a NOAA temperature is present; 0° is a temperature (common in Celsius)."""
return value is not None and value != ''
def _first_match(text: str, patterns: tuple[str, ...], low: int | None = None, high: int | None = None) -> str:
"""Group 1 of the first pattern found in the lowercased *text*, within [low, high] when given.
Each pattern is tried at its first match only; a value out of range moves on
to the next pattern, not to a later match of the same one.
"""
if not text:
return ""
lowered = text.lower()
for pattern in patterns:
match = re.search(pattern, lowered)
if match:
value = match.group(1)
if low is None or high is None:
return value
try:
if low <= int(value) <= high:
return value
except ValueError:
continue
return ""
class WxCommand(WeatherCommandMixin, BaseCommand):
"""Handles weather commands with zipcode support"""
# Plugin metadata
# Read-only informational output; safe for scheduled {cmd:...} rendering.
render_safe = True
name = "wx"
translation_ns = "commands.wx"
keywords = ['wx', 'weather', 'wxa', 'wxalert']
description = "Get weather information for a zip code (usage: wx 12345)"
category = "weather"
cooldown_seconds = 5 # 5 second cooldown per user to prevent API abuse
# NOAA/geocoding need the network, but custom WXSIM/MQTT sources may be LAN-only; check connectivity inside execute paths.
requires_internet = False
# Documentation
short_description = "Get weather for a US location using NOAA weather data"
usage = "wx <zipcode|city> [tomorrow|<N>d|hourly|alerts]"
examples = ["wx 98101", "wx seattle", "wx 90210 7d"]
parameters = [
{"name": "location", "description": "US zip code or city name"},
{"name": "option", "description": "tomorrow, Nd (e.g. 7d, 10d), hourly, or alerts (optional)"}
]
# Web-viewer settings schema (see modules/settings_schema.py)
settings_schema = [
{
"key": "temperature_unit",
"label": "Temperature unit",
"type": "enum",
"options": [
{"value": "fahrenheit", "label": "Fahrenheit (°F)"},
{"value": "celsius", "label": "Celsius (°C)"},
],
"default": "fahrenheit",
"help": "Unit used when reporting temperatures.",
},
{
"key": "wind_speed_unit",
"label": "Wind speed unit",
"type": "enum",
"options": [
{"value": "mph", "label": "Miles per hour (mph)"},
{"value": "kmh", "label": "Kilometers per hour (km/h)"},
{"value": "ms", "label": "Meters per second (m/s)"},
{"value": "kn", "label": "Knots (kn)"},
],
"default": "mph",
"help": "Unit used when reporting wind speed.",
},
{"key": "weather_provider", "label": "Weather provider", "type": "enum", "section": "Weather",
"options": [
{"value": "noaa", "label": "NOAA (US, includes alerts)"},
{"value": "openmeteo", "label": "Open-Meteo (global)"},
],
"default": "noaa", "help": "API used for forecasts. Shared weather setting."},
{"key": "default_city", "label": "Default city", "type": "str", "section": "Weather",
"default": "", "help": "City used for a bare 'wx' when no location is given. Shared weather setting."},
{"key": "default_state", "label": "Default state", "type": "str", "section": "Weather",
"default": "", "help": "2-letter state for city disambiguation (e.g. WA). Shared weather setting."},
{"key": "default_country", "label": "Default country", "type": "str", "section": "Weather",
"default": "US", "help": "2-letter country code (e.g. US). Shared weather setting."},
{"key": "always_show_location", "label": "Always name the location", "type": "bool", "section": "Weather",
"default": False,
"help": "Name the place in every wx/gwx reply a place is found for, not only when it is outside the "
"default state or country. Costs message length, and a reverse lookup for ZIP codes. "
"Shared weather setting."},
]
# Error constants
NO_DATA_NOGPS = "No GPS data available"
ERROR_FETCHING_DATA = "Error fetching weather data"
NO_ALERTS = "No weather alerts"
# Floor for the forecast body once a location prefix has been reserved out of
# the frame budget, so a long "City, State: " can never squeeze the forecast
# down to nothing (or negative, which the -10/-20 arithmetic in the period
# formatters would turn into dropped fields).
MIN_BODY_BUDGET = 40
def __init__(self, bot):
super().__init__(bot)
self.wx_enabled = self.get_config_value('Wx_Command', 'enabled', fallback=True, value_type='bool')
self.wxsim_parser = WXSIMParser()
# Check weather provider setting - delegate to international command if using Open-Meteo
weather_provider = bot.config.get('Weather', 'weather_provider', fallback='noaa').lower()
if weather_provider == 'openmeteo':
# Delegate to international weather command
self.delegate_command = GlobalWxCommand(bot)
# Use wx triggers plus any [Wx_Command] aliases loaded by BaseCommand.
self.delegate_command.keywords = list(self.keywords)
self.delegate_command.description = "Get weather information for any location (usage: wx Tokyo)"
self.logger.info("Weather provider set to 'openmeteo', delegating wx command to wx_international")
else:
self.delegate_command = None
# Only initialize NOAA-specific attributes if not delegating
if self.delegate_command is None:
self.url_timeout = 8 # seconds (reduced from 10 for faster failure detection)
self.forecast_duration = 3 # days
self.num_wx_alerts = 2 # number of alerts to show
self.use_metric = False # Use imperial units by default
self.zulu_time = False # Use local time by default
# Get default location/state/country from config for fallback/disambiguation
self.default_city = self.bot.config.get('Weather', 'default_city', fallback='').strip()
self.default_state = self.bot.config.get('Weather', 'default_state', fallback='')
self.default_country = self.bot.config.get('Weather', 'default_country', fallback='US')
self.always_show_location = self.bot.config.getboolean('Weather', 'always_show_location', fallback=False)
# Initialize geocoder (will use rate-limited helpers for actual calls)
# Keep geolocator for backwards compatibility, but prefer rate-limited helpers
self.geolocator = get_nominatim_geocoder()
# Get database manager for geocoding cache
self.db_manager = bot.db_manager
# Create a retry-enabled session for NOAA API calls
# This makes the API more resilient to timeouts and transient errors
self.noaa_session = self._create_retry_session()
# requests.Session and WXSIM parser instances are shared by this command.
# Keep provider calls serialized as they were before moving them to worker
# threads. The lock is acquired inside the worker, never on the event loop.
self._sync_provider_lock = threading.Lock()
self._nws_no_coverage = NWSNoCoverageCache()
def _unit_setting(self, key: str, fallback: str) -> str:
"""A unit setting: [Wx_Command] when it overrides, else the shared [Weather] one.
Read directly rather than through get_config_value, whose [Weather] fallback
logs a notice to move the setting into [Wx_Command]; for units, [Weather] is
where they belong.
"""
config = self.bot.config
for section in ('Wx_Command', 'Weather'):
if config.has_section(section) and config.has_option(section, key):
return str(config.get(section, key)).strip().lower()
return fallback
def _noaa_units(self) -> tuple[str, str]:
"""Configured (temperature_unit, wind_speed_unit) for NOAA replies.
Same lookup as the WXSIM path: [Wx_Command], falling back to [Weather].
Invalid values fall back to fahrenheit/mph, as gwx does.
"""
temp = self._unit_setting('temperature_unit', 'fahrenheit')
wind = self._unit_setting('wind_speed_unit', 'mph')
if temp not in ('fahrenheit', 'celsius'):
temp = 'fahrenheit'
if wind not in ('mph', 'kmh', 'ms', 'kn'):
wind = 'mph'
return temp, wind
@property
def _noaa_metric_distance(self) -> bool:
"""Kilometers instead of miles, following the temperature unit as gwx does."""
return self._noaa_units()[0] == 'celsius'
def _noaa_units_url(self, url: str) -> str:
"""A NOAA forecast URL asking for SI units (°C, km/h, also in the forecast text) when Celsius is configured."""
if self._noaa_units()[0] != 'celsius':
return url
return f"{url}{'&' if '?' in url else '?'}units=si"
def _noaa_wind_convert(self, number: str, wind_speed_text: str) -> str:
"""A NOAA wind number in the configured wind_speed_unit (NOAA sends mph, or km/h in SI mode)."""
source = 'kmh' if 'km/h' in wind_speed_text else 'mph'
target = self._noaa_units()[1]
if source == target:
return number
mph = int(number) if source == 'mph' else int(number) / 1.609344
value = {'mph': mph, 'kmh': mph * 1.609344, 'ms': mph * 0.44704, 'kn': mph * 0.868976}[target]
return str(int(round(value)))
@staticmethod
def _noaa_period_temp_symbol(period: dict) -> str:
u = (period.get("temperatureUnit") or "F").upper()
return "°F" if u == "F" else "°C"
def _create_retry_session(self) -> requests.Session:
"""Create a requests session with retry logic for NOAA API calls"""
return make_retry_session()
def _run_sync_provider(
self,
operation: Callable[_P, _T],
*args: _P.args,
**kwargs: _P.kwargs,
) -> _T:
"""Run one shared synchronous provider operation under its worker lock."""
with self._sync_provider_lock:
return operation(*args, **kwargs)
async def _run_sync_provider_async(
self,
operation: Callable[_P, _T],
*args: _P.args,
**kwargs: _P.kwargs,
) -> _T:
"""Move blocking provider work off-loop without concurrent Session use."""
return await asyncio.to_thread(self._run_sync_provider, operation, *args, **kwargs)
def get_help_text(self) -> str:
"""Get help text, delegating to international command if using Open-Meteo"""
if self.delegate_command:
return self.delegate_command.get_help_text()
return self.translate('commands.wx.description')
def matches_keyword(self, message: MeshMessage) -> bool:
"""Check if message starts with a weather keyword"""
if self.delegate_command:
return self.delegate_command.matches_keyword(message)
return self._cleaned_content_matches(
message,
lambda content_lower: any(
content_lower.startswith(keyword + ' ') or content_lower == keyword
for keyword in self.keywords
),
)
def can_execute(self, message: MeshMessage, skip_channel_check: bool = False) -> bool:
"""Override to delegate or use base class cooldown"""
# Check if wx command is enabled
if not self.wx_enabled:
return False
if self.delegate_command:
# Enforce [Wx_Command] channels first; delegate uses skip_channel_check
# so [Wx_Command] channels override is honored when using Open-Meteo
if not self.is_channel_allowed(message):
return False
return self.delegate_command.can_execute(message, skip_channel_check=True)
# Use base class for cooldown and other checks
return super().can_execute(message)
def get_remaining_cooldown(self, user_id: Optional[str] = None) -> int:
"""Get remaining cooldown time for a specific user"""
if self.delegate_command:
return self.delegate_command.get_remaining_cooldown(user_id)
# Use base class method
return super().get_remaining_cooldown(user_id)
def _get_companion_location(self, message: MeshMessage) -> Optional[tuple[float, float]]:
"""Get companion/sender location from the contact-tracking database."""
return get_companion_lat_lon(self.bot, message, self.logger, error_level="warning", trace=True)
def _get_bot_location(self) -> Optional[tuple[float, float]]:
"""Get bot location from config ([Bot] bot_latitude, bot_longitude)."""
return get_bot_lat_lon(self.bot, self.logger)
def _get_custom_wxsim_source(self, location: Optional[str] = None) -> Optional[str]:
"""Get custom WXSIM source URL from config.
Looks for keys in [Weather] section with pattern: custom.wxsim.<name> = <url>
Similar to how Channels_List handles dotted keys.
Args:
location: Location name or None for default source
Returns:
Optional[str]: Source URL or None if not found
"""
section = 'Weather'
if not self.bot.config.has_section(section):
self.logger.debug(f"Config section '{section}' does not exist")
return None
if location:
# Strip whitespace and normalize
location = location.strip()
location_lower = location.lower()
self.logger.debug(f"Checking for WXSIM source for location: '{location}' (normalized: '{location_lower}')")
# Look for keys matching custom.wxsim.<location> pattern
prefix = 'custom.wxsim.'
for key, value in self.bot.config.items(section):
if key.startswith(prefix):
# Extract the location name from the key (e.g., "custom.wxsim.lethbridge" -> "lethbridge")
key_location = key[len(prefix):].strip()
if key_location.lower() == location_lower:
self.logger.debug(f"Found WXSIM source: {key} = {value}")
return value
self.logger.debug(f"No WXSIM source found for location '{location}'")
else:
# Check for default source: custom.wxsim.default
default_key = 'custom.wxsim.default'
if self.bot.config.has_option(section, default_key):
url = self.bot.config.get(section, default_key)
self.logger.debug(f"Found default WXSIM source: {url}")
return url
self.logger.debug("No default WXSIM source configured")
return None
def _get_wxsim_weather(self, source_url: str, forecast_type: str = "default",
num_days: int = 7, message: MeshMessage = None,
location_name: Optional[str] = None) -> str:
"""Get and format weather from WXSIM source.
Args:
source_url: URL to WXSIM plaintext.txt file
forecast_type: "default", "tomorrow", or "multiday"
num_days: Number of days for multiday forecast
message: The MeshMessage for dynamic length calculation
location_name: Optional location name for display
Returns:
str: Formatted weather string
"""
if forecast_type in ("hourly", "alerts"):
return self.translate("commands.wx.source_option_not_available")
# Fetch WXSIM data
text = self.wxsim_parser.fetch_from_url(source_url, timeout=self.url_timeout)
if not text:
return self.translate('commands.wx.error', error="Failed to fetch WXSIM data")
# Parse the data
forecast = self.wxsim_parser.parse(text)
# Validate forecast is not stale
is_stale, stale_reason = self.wxsim_parser.is_forecast_stale(forecast, max_age_hours=48)
if is_stale:
self.logger.warning(f"WXSIM forecast appears stale: {stale_reason}")
# Still return the forecast, but log the warning
# Optionally, we could return an error message here instead
# Get unit preferences from config. Canonical section is [Wx_Command];
# get_config_value falls back to legacy [Weather] for existing setups.
temp_unit = self._unit_setting('temperature_unit', 'fahrenheit')
wind_unit = self._unit_setting('wind_speed_unit', 'mph')
# Format based on forecast type
if forecast_type == "tomorrow":
# Get tomorrow's forecast
if len(forecast.periods) > 1:
tomorrow = forecast.periods[1]
high = self.wxsim_parser._convert_temp(tomorrow.high_temp, temp_unit) if tomorrow.high_temp is not None else None
low = self.wxsim_parser._convert_temp(tomorrow.low_temp, temp_unit) if tomorrow.low_temp is not None else None
temp_symbol = "°F" if temp_unit == 'fahrenheit' else "°C"
result = f"Tomorrow: {tomorrow.conditions}"
hl = self._format_high_low(high, low, temp_symbol)
if hl:
result += f" {hl}"
if tomorrow.precip_chance and tomorrow.precip_chance > 30:
result += f" {tomorrow.precip_chance}% PoP"
if location_name:
return f"{location_name}: {result}"
return result
else:
return self.translate('commands.wx.error', error="Tomorrow forecast not available")
elif forecast_type == "multiday":
# Format multiday forecast
summary = self.wxsim_parser.format_forecast_summary(forecast, num_days, temp_unit, wind_unit)
if location_name:
return f"{location_name}:\n{summary}"
return summary
else:
# Default: current conditions + today's forecast
current = self.wxsim_parser.format_current_conditions(forecast, temp_unit, wind_unit)
# Add today's high/low if available (use first period as "today")
if forecast.periods:
today = forecast.periods[0]
high = self.wxsim_parser._convert_temp(today.high_temp, temp_unit) if today.high_temp is not None else None
low = self.wxsim_parser._convert_temp(today.low_temp, temp_unit) if today.low_temp is not None else None
temp_symbol = "°F" if temp_unit == 'fahrenheit' else "°C"
hl_today = self._format_high_low(high, low, temp_symbol)
if hl_today:
current += f" | {hl_today}"
# Add tomorrow if available (second period)
if len(forecast.periods) > 1:
tomorrow = forecast.periods[1]
tomorrow_high = self.wxsim_parser._convert_temp(tomorrow.high_temp, temp_unit) if tomorrow.high_temp is not None else None
tomorrow_low = self.wxsim_parser._convert_temp(tomorrow.low_temp, temp_unit) if tomorrow.low_temp is not None else None
hl_tom = self._format_high_low(tomorrow_high, tomorrow_low, temp_symbol)
if hl_tom:
current += f" | Tomorrow: {hl_tom}"
if location_name:
return f"{location_name}: {current}"
return current
async def _get_wxsim_weather_async(
self,
source_url: str,
forecast_type: str = "default",
num_days: int = 7,
message: MeshMessage = None,
location_name: Optional[str] = None,
) -> str:
return await self._run_sync_provider_async(
self._get_wxsim_weather,
source_url,
forecast_type,
num_days,
message,
location_name,
)
def _coordinates_to_location_string(self, lat: float, lon: float) -> Optional[str]:
"""Convert coordinates to a location string (city name) using reverse geocoding.
Args:
lat: Latitude.
lon: Longitude.
Returns:
Optional[str]: Location string (city name) or None if geocoding fails.
"""
try:
from ..utils import rate_limited_nominatim_reverse_sync
result = rate_limited_nominatim_reverse_sync(self.bot, f"{lat}, {lon}", timeout=10)
if result and hasattr(result, 'raw'):
# Extract city name from address
address = result.raw.get('address', {})
city = (address.get('city') or
address.get('town') or
address.get('village') or
address.get('municipality') or
address.get('county', ''))
state = address.get('state', '')
# Normalize state to abbreviation
if state:
state_abbr, _ = normalize_us_state(state)
if state_abbr:
state = state_abbr
if city:
if state:
return f"{city}, {state}"
return city
return None
except Exception as e:
self.logger.debug(f"Error reverse geocoding coordinates {lat}, {lon}: {e}")
return None
async def _coordinates_to_location_string_async(
self, lat: float, lon: float
) -> Optional[str]:
return await self._run_sync_provider_async(
self._coordinates_to_location_string, lat, lon
)
async def execute(self, message: MeshMessage) -> bool:
"""Execute the weather command"""
# Delegate to international command if using Open-Meteo provider
if self.delegate_command:
return await self.delegate_command.execute(message)
content = message.content.strip()
# Parse the command to extract location and forecast type
# Support formats: "wx 12345", "wx seattle", "wx paris, tx", "weather everett", "wxa bellingham"
# New formats: "wx 12345 tomorrow", "wx 12345 7", "wx 12345 7d", "wx 12345 7day", "wx 12345 alerts"
parts = content.split()
# Track if we're using companion location (so we always show location in response)
using_companion_location = False
# An option alone ("wx hourly") applies to the no-location fallbacks below.
parts, option_word, option_type, option_days = self._split_option_only(parts, WX_MULTIDAY_MAX_DAYS)
# If no location specified, check custom MQTT then WXSIM default sources
if len(parts) < 2:
mqtt_topic = self._get_custom_mqtt_weather_topic(None)
if mqtt_topic:
try:
self.record_execution(message.sender_id)
weather_data = self._mqtt_weather_line(mqtt_topic, option_type, None)
return bool(await self.send_response(message, weather_data))
except Exception as e:
self.logger.error(f"Error reading MQTT weather: {e}")
await self.send_response(message, self.translate("commands.wx.error", error=str(e)))
return True
wxsim_source = self._get_custom_wxsim_source(None) # Check for default
if wxsim_source:
# Use custom WXSIM default source
try:
self.record_execution(message.sender_id)
weather_data = await self._get_wxsim_weather_async(
wxsim_source, option_type, option_days, message
)
if option_type == "multiday":
return await self._send_multiday_forecast(message, weather_data)
return bool(await self.send_response(message, weather_data))
except Exception as e:
self.logger.error(f"Error fetching WXSIM weather: {e}")
await self.send_response(message, self.translate('commands.wx.error', error=str(e)))
return True
# No custom source, try companion location
companion_location = self._get_companion_location(message)
if companion_location:
# Use coordinates directly to avoid re-geocoding issues
location_str = self._coordinates_query(*companion_location)
parts = [parts[0], location_str]
using_companion_location = True
# Get city name for display
display_name = await self._coordinates_to_location_string_async(
companion_location[0], companion_location[1]
)
if display_name:
self.logger.info(f"Using companion location: {display_name} ({companion_location[0]}, {companion_location[1]})")
else:
self.logger.info(f"Using companion coordinates: {location_str}")
else:
# No companion location: use default city if configured, then bot location fallback
if self.default_city:
location_parts = [self.default_city]
if self.default_state:
location_parts.append(self.default_state)
if self.default_country:
location_parts.append(self.default_country)
location_str = ", ".join(location_parts)
parts = [parts[0], location_str]
self.logger.info(f"Using default city (no args): {location_str}")
else:
# No default city: optionally use bot's configured coordinates
use_bot = self.get_config_value(
'Wx_Command',
'use_bot_location_when_no_location',
fallback=False,
value_type='bool',
)
bot_loc = self._get_bot_location() if use_bot else None
if bot_loc:
location_str = self._coordinates_query(*bot_loc)
parts = [parts[0], location_str]
display_name = await self._coordinates_to_location_string_async(
bot_loc[0], bot_loc[1]
)
if display_name:
self.logger.info(
f"Using bot location (no args): {display_name} ({bot_loc[0]}, {bot_loc[1]})"
)
else:
self.logger.info(f"Using bot coordinates (no args): {location_str}")
else:
if use_bot:
self.logger.debug(
"use_bot_location_when_no_location enabled but bot_latitude/bot_longitude "
"not set; showing usage"
)
else:
self.logger.debug("No companion/default city location found, showing usage")
await self.send_response(message, self.translate('commands.wx.usage'))
return True
if option_word:
parts.append(option_word)
# Check for "alerts" keyword first (special handling)
show_full_alerts = False
if len(parts) > 2 and parts[-1].lower() == "alerts":
show_full_alerts = True
location_parts = parts[1:-1] # Remove "alerts" from location
else:
location_parts = parts[1:]
forecast_type = "alerts" if show_full_alerts else "default"
num_days = 7 # Default for multi-day forecast
if not show_full_alerts and not (len(parts) == 2 and self._is_custom_source_name(parts[1])):
location_parts, forecast_type, num_days = self._parse_forecast_suffix(
location_parts, WX_MULTIDAY_MAX_DAYS, allow_hourly=True
)
# Join remaining parts to handle "city, state" format
location = ' '.join(location_parts).strip()
if not location:
await self.send_response(message, self.translate('commands.wx.usage'))
return True
# Custom MQTT before WXSIM
mqtt_topic = self._get_custom_mqtt_weather_topic(location)
if mqtt_topic:
self.logger.info(f"Using custom MQTT weather topic for location '{location}': {mqtt_topic}")
try:
self.record_execution(message.sender_id)
weather_data = self._mqtt_weather_line(
mqtt_topic, forecast_type, location
)
if forecast_type == "multiday":
return await self._send_multiday_forecast(message, weather_data)
return bool(await self.send_response(message, weather_data))
except Exception as e:
self.logger.error(f"Error reading MQTT weather: {e}")
await self.send_response(message, self.translate("commands.wx.error", error=str(e)))
return True
# Check for custom WXSIM source first (before checking location type)
wxsim_source = self._get_custom_wxsim_source(location)
if wxsim_source:
self.logger.info(f"Using custom WXSIM source for location '{location}': {wxsim_source}")
# Use custom WXSIM source
try:
self.record_execution(message.sender_id)
weather_data = await self._get_wxsim_weather_async(
wxsim_source,
forecast_type,
num_days,
message,
location_name=location,
)
if forecast_type == "multiday":
return await self._send_multiday_forecast(message, weather_data)
return bool(await self.send_response(message, weather_data))
except Exception as e:
self.logger.error(f"Error fetching WXSIM weather: {e}")
await self.send_response(message, self.translate('commands.wx.error', error=str(e)))
return True
else:
self.logger.debug(f"No custom WXSIM source found for location '{location}', using normal weather API")
# Check if it's coordinates, zipcode, or city name
if re.match(r'^\s*-?\d+\.?\d*\s*,\s*-?\d+\.?\d*\s*$', location):
# It's coordinates (lat,lon format)
location_type = "coordinates"
elif re.match(r'^\s*\d{5}\s*$', location):
# It's a zipcode (allow surrounding whitespace; strip later in geocode)
location_type = "zipcode"
else:
# It's a city name (possibly with state)
location_type = "city"
try:
# Record execution for this user
self.record_execution(message.sender_id)
# Special handling for "alerts" command
if show_full_alerts:
# Get alerts only (no weather forecast)
lat, lon = None, None
if location_type == "coordinates":
try:
lat_str, lon_str = location.split(',')
lat = float(lat_str.strip())
lon = float(lon_str.strip())
if not (-90 <= lat <= 90) or not (-180 <= lon <= 180):
await self.send_response(message, self.translate('commands.wx.error', error="Invalid coordinates"))
return True
except ValueError:
await self.send_response(message, self.translate('commands.wx.error', error=f"Invalid coordinates format: {location}"))
return True
elif location_type == "zipcode":
lat, lon = await self._zipcode_to_lat_lon_async(location)
if lat is None or lon is None:
await self.send_response(message, self.translate('commands.wx.no_location_zipcode', location=location))
return True
else: # city
result = await self._city_to_lat_lon_async(location)
if len(result) == 3:
lat, lon, address_info = result
else:
lat, lon = result
if lat is None or lon is None:
region = self.default_state or self.default_country
await self.send_response(message, self.translate('commands.wx.no_location_city', location=location, state=region))
return True
# Get and display full alert list
return await self._send_full_alert_list(message, lat, lon)
# Get weather data for the location
weather_data = await self.get_weather_for_location(location, location_type, forecast_type, num_days, message, using_companion_location=using_companion_location)
# Check if we need to send multiple messages
if isinstance(weather_data, tuple) and weather_data[0] == "multi_message":
# Send weather data first; if it was refused (rate limit, send failure),
# do not send its second part on its own.
if not await self.send_response(message, weather_data[1]):
return False
# Wait for bot TX rate limiter to allow next message
rate_limit = self.bot.config.getfloat('Bot', 'bot_tx_rate_limit_seconds', fallback=1.0)
# Use a conservative sleep time to avoid rate limiting
sleep_time = max(rate_limit + 1.0, 2.0) # At least 2 seconds, or rate_limit + 1 second
await self._pace_reply(message, sleep_time)
# Send the special weather statement (already formatted with prioritization)
alert_text = weather_data[2]
# Second part of the same reply: the reply limiter already let the first through.
return bool(await self.send_response(message, alert_text, skip_user_rate_limit=True))
if forecast_type == "multiday":
# Use message splitting for multi-day forecasts
return await self._send_multiday_forecast(message, weather_data)
# Send single message as usual
return bool(await self.send_response(message, weather_data))
except Exception as e:
self.logger.error(f"Error in weather command: {e}")
await self.send_response(message, self.translate('commands.wx.error', error=str(e)))
return True
async def get_weather_for_location(self, location: str, location_type: str, forecast_type: str = "default", num_days: int = 7, message: MeshMessage = None, using_companion_location: bool = False) -> str:
"""Run the ordered synchronous geocode/NOAA workflow off the event loop."""
return await self._run_sync_provider_async(
self._get_weather_for_location_sync,
location,
location_type,
forecast_type,
num_days,
message,
using_companion_location,
)
def _get_weather_for_location_sync(self, location: str, location_type: str, forecast_type: str = "default", num_days: int = 7, message: MeshMessage = None, using_companion_location: bool = False) -> str:
"""Get weather data for a location (coordinates, zipcode, or city)
Args:
location: The location (coordinates "lat,lon", zipcode, or city name)
location_type: "coordinates", "zipcode", or "city"
forecast_type: "default", "tomorrow", "multiday", or "hourly"
num_days: Number of days for multiday forecast (2–16)
message: The MeshMessage for dynamic length calculation
using_companion_location: If True, always include location prefix even if same state
"""
try:
# Convert location to lat/lon based on type
if location_type == "coordinates":
# Parse coordinates from "lat,lon" format
try:
lat_str, lon_str = location.split(',')
lat = float(lat_str.strip())
lon = float(lon_str.strip())
# Validate coordinate ranges
if not (-90 <= lat <= 90):
return self.translate('commands.wx.error', error=f"Invalid latitude: {lat}")
if not (-180 <= lon <= 180):
return self.translate('commands.wx.error', error=f"Invalid longitude: {lon}")
# Get address_info for location display via reverse geocoding
location_str = self._coordinates_to_location_string(lat, lon)
if location_str:
# Parse the location string to get city and state for address_info
parts = location_str.split(',')
if len(parts) >= 2:
city = parts[0].strip()
state = parts[1].strip()
address_info = {'city': city, 'state': state}
else:
address_info = {'city': location_str}
else:
address_info = {}
except ValueError:
return self.translate('commands.wx.error', error=f"Invalid coordinates format: {location}")
elif location_type == "zipcode":
lat, lon = self.zipcode_to_lat_lon(location)
if lat is None or lon is None:
return self.translate('commands.wx.no_location_zipcode', location=location)
address_info = None
else: # city
result = self.city_to_lat_lon(location)
if len(result) == 3:
lat, lon, address_info = result
else:
lat, lon = result
address_info = None
if lat is None or lon is None:
region = self.default_state or self.default_country
return self.translate('commands.wx.no_location_city', location=location, state=region)
# Check if the found city is in a different state than default
actual_city = location
actual_state = self.default_state or self.default_country
if address_info:
# Try to get the best city name from various address fields
actual_city = (address_info.get('city') or
address_info.get('town') or
address_info.get('village') or
address_info.get('hamlet') or
address_info.get('municipality') or
location)
actual_state = address_info.get('state', self.default_state)
# Convert full state name to abbreviation if needed using the us library
if len(actual_state) > 2:
state_abbr, _ = normalize_us_state(actual_state)
if state_abbr:
actual_state = state_abbr
# Also check if the default state needs to be converted for comparison
default_state_full = self.default_state
if len(self.default_state) == 2:
# Convert abbreviation to full name for comparison
_, default_state_full = normalize_us_state(self.default_state)
if not default_state_full:
default_state_full = self.default_state
# Add location info if city is in a different state than default, or if using companion location
location_prefix = ""
if location_type == "coordinates" and address_info:
# For coordinates, always show location if we have address info
city = address_info.get('city', '')
state = address_info.get('state', '')
if city and state:
# Normalize state to abbreviation
state_abbr, _ = normalize_us_state(state)
if state_abbr:
state = state_abbr
location_prefix = f"{city}, {state}: "
elif city:
location_prefix = f"{city}: "
elif location_type == "city" and address_info:
# Compare states (handle both full names and abbreviations)
states_different = (actual_state != self.default_state and
actual_state != default_state_full)
# Always show location if using companion location, or if state is different
if using_companion_location or states_different or self.always_show_location:
location_prefix = f"{actual_city}, {actual_state}: " if actual_state else f"{actual_city}: "
elif location_type == "zipcode" and (using_companion_location or self.always_show_location):
# For zipcode with companion location, try to get city name from reverse geocoding
location_str = self._coordinates_to_location_string(lat, lon)
if location_str:
location_prefix = f"{location_str}: "
# Get max message length dynamically
max_length = self.get_max_message_length(message) if message else 130
# location_prefix is prepended to the formatted body below, so it has to
# come out of the same frame budget. Formatting the body to the full
# budget and then prefixing it overruns the frame by the prefix's own
# length -- "Lockhart, Texas: " alone is 17 bytes, enough to push an
# emoji-dense forecast past the 160-byte firmware limit. The alert text
# goes out as its own message and keeps the full budget.
body_max_length = max(
max_length - self._count_display_width(location_prefix), self.MIN_BODY_BUDGET
)
# Get weather forecast based on type
if forecast_type == "tomorrow":
forecast_periods, points_data = self.get_noaa_weather(lat, lon, return_periods=True, max_length=body_max_length)
if forecast_periods == self.ERROR_FETCHING_DATA:
return self.translate('commands.wx.error_fetching')
weather = self.format_tomorrow_forecast(forecast_periods, max_length=body_max_length)
elif forecast_type == "multiday":
forecast_periods, points_data = self.get_noaa_weather(lat, lon, return_periods=True, max_length=body_max_length)
if forecast_periods == self.ERROR_FETCHING_DATA:
return self.translate('commands.wx.error_fetching')
weather = self.format_multiday_forecast(forecast_periods, num_days, max_length=body_max_length)
elif forecast_type == "hourly":
hourly_periods, points_data = self.get_noaa_hourly_weather(lat, lon)
if hourly_periods == self.ERROR_FETCHING_DATA:
return self.translate('commands.wx.error_fetching')
weather = self.format_hourly_forecast(hourly_periods, max_length=body_max_length)
else: # default
weather, points_data = self.get_noaa_weather(lat, lon, max_length=body_max_length)
if weather == self.ERROR_FETCHING_DATA:
return self.translate('commands.wx.error_fetching')
# Note: Current conditions are now integrated directly into the current period
# via _add_period_details() using observation station data
# Get weather alerts (only for default forecast type to avoid cluttering)
if forecast_type == "default":
alerts_result = self.get_weather_alerts_noaa(lat, lon, return_full_data=False)
if alerts_result == self.ERROR_FETCHING_DATA or alerts_result == self.NO_ALERTS:
pass
else:
full_alert_text, abbreviated_alert_text, alert_count = alerts_result
if alert_count > 0:
# Get full alert data for prioritized formatting
alerts_full_result = self.get_weather_alerts_noaa(lat, lon, return_full_data=True)
if alerts_full_result not in [self.ERROR_FETCHING_DATA, self.NO_ALERTS]:
alerts_list, _ = alerts_full_result
# Format with prioritization and summary
formatted_alert_text = self._format_alerts_compact_summary(alerts_list, alert_count, max_length=max_length)
else:
# Fallback to old format
formatted_alert_text = full_alert_text
# Always send weather first, then alerts in separate message
self.logger.info(f"Found {alert_count} alerts - using two-message mode")
return ("multi_message", f"{location_prefix}{weather}", formatted_alert_text, alert_count)
return f"{location_prefix}{weather}"
except Exception as e:
self.logger.error(f"Error getting weather for {location_type} {location}: {e}")
return self.translate('commands.wx.error', error=str(e))
def zipcode_to_lat_lon(self, zipcode: str) -> tuple:
"""Convert zipcode to latitude and longitude"""
try:
lat, lon = geocode_zipcode_sync(self.bot, zipcode, timeout=10)
return lat, lon
except Exception as e:
self.logger.error(f"Error geocoding zipcode {zipcode}: {e}")
return None, None
async def _zipcode_to_lat_lon_async(self, zipcode: str) -> tuple:
return await self._run_sync_provider_async(self.zipcode_to_lat_lon, zipcode)
def city_to_lat_lon(self, city: str) -> tuple:
"""Convert city name to latitude and longitude using default state"""
try:
# Use shared geocode_city_sync function with address info
default_country = self.bot.config.get('Weather', 'default_country', fallback='US')
lat, lon, address_info = geocode_city_sync(
self.bot, city, default_state=self.default_state,
default_country=default_country,
include_address_info=True, timeout=10
)
# Explicit None checks: 0 is a valid latitude (equator) or longitude (prime meridian).
if lat is not None and lon is not None:
return lat, lon, address_info or {}
else:
return None, None, None
except Exception as e:
self.logger.error(f"Error geocoding city {city}: {e}")
return None, None, None
async def _city_to_lat_lon_async(self, city: str) -> tuple:
return await self._run_sync_provider_async(self.city_to_lat_lon, city)
def _noaa_fetch(self, url: str, what: str):
"""GET *url* through the NOAA session; None, after a warning naming *what*,
on an HTTP error status, a timeout or a connection error.
"""
try:
response = self.noaa_session.get(url, timeout=self.url_timeout)
if not response.ok:
self.logger.warning(f"Error fetching {what} from NOAA: HTTP {response.status_code}")
return None
except (requests.exceptions.Timeout, requests.exceptions.ConnectionError) as e:
self.logger.warning(f"Timeout/connection error fetching {what} from NOAA: {e}")
return None
return response
def get_noaa_weather(self, lat: float, lon: float, return_periods: bool = False, max_length: int = 130) -> tuple:
"""Get weather forecast from NOAA and return both weather string and points data
Args:
lat: Latitude
lon: Longitude
return_periods: If True, return forecast periods array instead of formatted string
max_length: Maximum message length (default 130 for backwards compatibility)
Returns:
Tuple of (weather_string_or_periods, points_data)
"""
try:
# Round coordinates to 4 decimal places to avoid API redirects
lat_rounded = round(lat, 4)
lon_rounded = round(lon, 4)
# Get weather data from NOAA
weather_api = f"https://api.weather.gov/points/{lat_rounded},{lon_rounded}"
# Get the forecast URL (with retry logic)
weather_data = self._noaa_fetch(weather_api, "weather data")
if weather_data is None:
return self.ERROR_FETCHING_DATA, None
weather_json = weather_data.json()
forecast_url = weather_json['properties']['forecast']
# Get the forecast (with retry logic)
forecast_data = self._noaa_fetch(self._noaa_units_url(forecast_url), "weather forecast")
if forecast_data is None:
return self.ERROR_FETCHING_DATA, None
forecast_json = forecast_data.json()
forecast = forecast_json['properties']['periods']
# If return_periods is True, return the periods array directly
if return_periods:
if not forecast:
return self.ERROR_FETCHING_DATA, None
return forecast, weather_json
# Format the forecast - focus on current conditions and key info
if not forecast:
return "No forecast data available", weather_json
current = forecast[0]
day_name = self._noaa_period_display_name(current)
temp = current.get('temperature', 'N/A')
temp_unit = current.get('temperatureUnit', 'F')
short_forecast = current.get('shortForecast', 'Unknown')
wind_speed = current.get('windSpeed', '')
wind_direction = current.get('windDirection', '')
detailed_forecast = current.get('detailedForecast', '')
# Extract additional useful info from detailed forecast
self.extract_humidity(detailed_forecast)
precip_chance = self.extract_precip_chance(detailed_forecast)
# Create compact but complete weather string with emoji
weather_emoji = self.get_weather_emoji(short_forecast)
weather = f"{day_name}: {weather_emoji}{short_forecast} {temp}°{temp_unit}"
# Add wind info if available
if wind_speed and wind_direction:
wind_match = re.search(r'(\d+)', wind_speed)
if wind_match:
wind_num = self._noaa_wind_convert(wind_match.group(1), wind_speed)
wind_dir = self.abbreviate_wind_direction(wind_direction)
if wind_dir:
weather += f" {wind_dir}{wind_num}"
# PRIORITIZE: Add all available details to current period first
# Get observation station data for more accurate current conditions
observation_data = self.get_observation_data(weather_json)
# Use most of the max_length limit (max_length - 10 chars) to ensure current period gets full details
# Additional periods will only be added if there's remaining space
# Pass observation_data to use real-time station data instead of parsing from text
current_period_max = max_length - 10
weather = self._add_period_details(weather, detailed_forecast, 0, max_length=current_period_max, observation_data=observation_data)
# Also add precipitation chance if available (not in helper function)
if precip_chance and self._count_display_width(weather) < current_period_max:
weather += f" 🌦️{precip_chance}%"
# Also add UV index if available (not in helper function)
uv_index = self.extract_uv_index(detailed_forecast)
if uv_index and self._count_display_width(weather) < current_period_max:
weather += f" UV{uv_index}"
# Add next period (Today, Tonight) and Tomorrow if available
# First, find Today, Tonight, and Tomorrow periods
today_period = None
tonight_period = None
tomorrow_period = None
current_period_name = current.get('name', '').lower()
is_current_tonight = 'tonight' in current_period_name
is_current_night = any(word in current_period_name for word in ['tonight', 'overnight', 'night'])
# Check if current period is a night period (Overnight, Tonight, etc.)
# If so, we should prioritize showing the upcoming daytime period (Today)
for i, period in enumerate(forecast):
period_name = period.get('name', '').lower()
# Look for "Today" period (daytime forecast)
if 'today' in period_name and today_period is None and i > 0:
# Make sure it's not a night period
if 'night' not in period_name and 'tonight' not in period_name:
today_period = (i, period)
elif 'tonight' in period_name and tonight_period is None:
tonight_period = (i, period)
elif 'tomorrow' in period_name and tomorrow_period is None:
tomorrow_period = (i, period)
# If current is a night period and we haven't found Today yet, look for next daytime period
if is_current_night and not today_period:
# Look for the next period that's not a night period
for i, period in enumerate(forecast):
if i > 0: # Skip current period
period_name = period.get('name', '').lower()
# Look for daytime periods (Today, or day names without "night")
if 'today' in period_name and 'night' not in period_name:
today_period = (i, period)
break
# Also check for day names that aren't night periods
day_names = ['monday', 'tuesday', 'wednesday', 'thursday', 'friday', 'saturday', 'sunday']
if any(day in period_name for day in day_names) and 'night' not in period_name:
today_period = (i, period)
break
# If current is Tonight and we haven't found Tomorrow yet, look for next day's periods
if is_current_tonight and not tomorrow_period:
# If today_period is a day name (not "Today"), look for the next period after it
if today_period:
period_name_lower = today_period[1].get('name', '').lower()
day_names = ['monday', 'tuesday', 'wednesday', 'thursday', 'friday', 'saturday', 'sunday']
if any(day in period_name_lower for day in day_names) and 'today' not in period_name_lower:
# today_period is actually tomorrow's daytime period - look for the night period after it
today_period_index = today_period[0]
# Look for the next period after today_period (should be the night period for that day)
for i, period in enumerate(forecast):
if i > today_period_index: # Look for periods after today_period
period_name = period.get('name', '').lower()
# Look for the night period for the same day, or the next day
if any(word in period_name for word in ['night', 'tomorrow', 'monday', 'tuesday', 'wednesday', 'thursday', 'friday', 'saturday', 'sunday']):
tomorrow_period = (i, period)
break
# If we didn't find a night period, use today_period as tomorrow_period
if not tomorrow_period:
tomorrow_period = today_period
else:
# Look for periods after Tonight (next day)
for i, period in enumerate(forecast):
if i > 0: # Skip current period
period_name = period.get('name', '').lower()
# Skip if this period is already set as today_period (avoid duplicates)
if today_period and today_period[0] == i:
continue
# Look for tomorrow, next day, or day names
if any(word in period_name for word in ['tomorrow', 'monday', 'tuesday', 'wednesday', 'thursday', 'friday', 'saturday', 'sunday']):
tomorrow_period = (i, period)
break
else:
# Look for periods after Tonight (next day)
for i, period in enumerate(forecast):
if i > 0: # Skip current period
period_name = period.get('name', '').lower()
# Look for tomorrow, next day, or day names
if any(word in period_name for word in ['tomorrow', 'monday', 'tuesday', 'wednesday', 'thursday', 'friday', 'saturday', 'sunday']):
tomorrow_period = (i, period)
break
# If current is a night period, prioritize adding Today (the upcoming daytime)
# When today_period is a day name (like "Tuesday"), we still add it as tomorrow's daytime period
if is_current_night and today_period:
period = today_period[1]
# Always add today_period - it represents tomorrow's daytime when current is Tonight
period_detailed = period.get('detailedForecast', '')
period_head = self._noaa_period_str(period)
if period_head:
period_str = self._noaa_period_wind(weather, period_head, period, max_length - 10, max_length)
# Add additional details (humidity, dew point, visibility, etc.)
# But only if current period isn't too long - prioritize current period details
current_weather_len = self._count_display_width(weather)
# Only add details to additional periods if current period is under max_length - 20 chars
# This ensures we prioritize current period details first
if current_weather_len < max_length - 20:
period_str = self._add_period_details(period_str, period_detailed, current_weather_len, max_length=max_length)
# Only add if we have space (using display width)
# Be more conservative - only add if current period is reasonable length
if current_weather_len < max_length - 20 and self._count_display_width(weather + period_str) <= max_length:
weather += period_str
# Add Tonight if it's the immediate next period (and current is not already Tonight)
# If we already added Today, we can still add Tonight if it's the next period after Today
if tonight_period and not is_current_tonight:
# Only add if it's the immediate next period, or if current is night and we haven't added Today yet
should_add_tonight = False
if is_current_night and today_period:
# If current is night and we added Today, check if Tonight comes after Today
if tonight_period[0] > today_period[0]:
should_add_tonight = True
elif tonight_period[0] == 1:
# If current is not night, Tonight should be the immediate next period
should_add_tonight = True
if should_add_tonight:
period = tonight_period[1]
period_detailed = period.get('detailedForecast', '')
period_head = self._noaa_period_str(period)
if period_head:
period_str = self._noaa_period_wind(weather, period_head, period, max_length - 10, max_length)
# Add additional details (humidity, dew point, visibility, etc.)
# But only if current period isn't too long - prioritize current period details
current_weather_len = self._count_display_width(weather)
# Only add details to additional periods if current period is under max_length - 20 chars
# This ensures we prioritize current period details first
if current_weather_len < max_length - 20:
period_str = self._add_period_details(period_str, period_detailed, current_weather_len, max_length=max_length)
# Only add if we have space (using display width)
# Be more conservative - only add if current period is reasonable length
if current_weather_len < max_length - 20 and self._count_display_width(weather + period_str) <= max_length:
weather += period_str
# Always try to add Tomorrow if available (especially if current is Tonight)
# Prioritize adding Tomorrow when current is Tonight to use more of the available message length
if tomorrow_period:
period = tomorrow_period[1]
period_detailed = period.get('detailedForecast', '')
period_short = period.get('shortForecast', '')
night = is_current_tonight or is_current_night
period_head = None
if period.get('temperature', '') and period_short:
# Shorten long forecast text (especially when current is a night period)
forecast_text = (
self._abbreviate_noaa_forecast(period_short) if night and len(period_short) > 20 else period_short
)
period_head = self._noaa_period_str(period, forecast_text)
if period_head:
# Be more aggressive about adding wind when current is a night period
wind_threshold = 115 if night else 120
period_str = self._noaa_period_wind(weather, period_head, period, wind_threshold, max_length)
# Add additional details (humidity, dew point, visibility, etc.)
# But only if current period isn't too long - prioritize current period details
current_weather_len = self._count_display_width(weather)
# Only add details to additional periods if current period is under max_length - 20 chars
# This ensures we prioritize current period details first
if current_weather_len < max_length - 20:
max_chars = max_length - 2 if (is_current_tonight or is_current_night) else max_length
period_str = self._add_period_details(period_str, period_detailed, current_weather_len, max_chars)
# Only add if we have space (using display width, prioritize current period)
# Be more aggressive about adding tomorrow_period when current is Tonight and we have space
max_chars = max_length - 2 if (is_current_tonight or is_current_night) else max_length
# If current is Tonight and we have plenty of space, be more lenient with the length check
if is_current_tonight or is_current_night:
# Allow adding tomorrow_period if we're under max_length - 10 chars (more lenient)
if current_weather_len < max_length - 10 and self._count_display_width(weather + period_str) <= max_chars:
weather += period_str
else:
# For non-night periods, use the stricter check
if current_weather_len < max_length - 20 and self._count_display_width(weather + period_str) <= max_chars:
weather += period_str
return weather, weather_json
except Exception as e:
self.logger.error(f"Error fetching NOAA weather: {e}")
return self.ERROR_FETCHING_DATA, None
def _noaa_period_str(self, period: dict, forecast_text: Optional[str] = None) -> Optional[str]:
"""``" | Name: <emoji><forecast> <high/low or temp°>"`` for a forecast period.
None when the period has no temperature or no short forecast.
``forecast_text`` replaces the short forecast in the text (the emoji
still comes from the short forecast).
"""
period_name = self._noaa_period_display_name(period)
period_temp = period.get('temperature', '')
period_short = period.get('shortForecast', '')
if not (_has_temp(period_temp) and period_short):
return None
period_high_low = self.extract_high_low(
period.get('detailedForecast', ''), self._noaa_period_temp_symbol(period)
)
period_emoji = self.get_weather_emoji(period_short)
text = period_short if forecast_text is None else forecast_text
if period_high_low:
return f" | {period_name}: {period_emoji}{text} {period_high_low}"
return f" | {period_name}: {period_emoji}{text} {period_temp}°"
def _noaa_period_wind(self, weather: str, period_str: str, period: dict, threshold: int, max_length: int) -> str:
"""Append the period's wind to ``period_str`` when ``weather + period_str`` is under ``threshold``
display columns and the result still fits ``max_length``."""
period_wind_speed = period.get('windSpeed', '')
period_wind_direction = period.get('windDirection', '')
if period_wind_speed and period_wind_direction:
test_str = weather + period_str
if self._count_display_width(test_str) < threshold:
wind_match = re.search(r'(\d+)', period_wind_speed)
if wind_match:
wind_num = self._noaa_wind_convert(wind_match.group(1), period_wind_speed)
wind_dir = self.abbreviate_wind_direction(period_wind_direction)
if wind_dir:
wind_info = f" {wind_dir}{wind_num}"
if self._count_display_width(test_str + wind_info) <= max_length:
return period_str + wind_info
return period_str
@staticmethod
def _abbreviate_noaa_forecast(period_short: str) -> str:
"""Shorten a long short-forecast ("A then B and C") to its main conditions."""
abbreviated_forecast = period_short
# Try to shorten forecast text to fit more info
# Remove transitional words and keep meaningful conditions
words = period_short.split()
# Transitional words to skip
transitions = {'then', 'and', 'or', 'becoming', 'followed', 'by', 'with'}
# If there's a "then" pattern, take first condition and last significant condition
if 'then' in words:
then_index = words.index('then')
# Take first condition (before "then")
first_part = words[:then_index]
# Take last significant condition (after "then", skip small words)
if then_index + 1 < len(words):
last_part = [w for w in words[then_index + 1:] if w.lower() not in transitions]
# Combine: first condition + last significant condition (max 2 words)
if last_part:
abbreviated_forecast = ' '.join(first_part)
if len(last_part) <= 2:
abbreviated_forecast += ' ' + ' '.join(last_part)
else:
# Take last 2 words of the last part
abbreviated_forecast += ' ' + ' '.join(last_part[-2:])
else:
abbreviated_forecast = ' '.join(first_part)
else:
abbreviated_forecast = ' '.join(first_part)
else:
# Filter out transitional words and take first meaningful words
meaningful_words = [w for w in words if w.lower() not in transitions]
if len(meaningful_words) > 3:
abbreviated_forecast = ' '.join(meaningful_words[:3])
else:
abbreviated_forecast = ' '.join(meaningful_words)
return abbreviated_forecast
def get_noaa_hourly_weather(self, lat: float, lon: float) -> tuple:
"""Get hourly weather forecast from NOAA
Args:
lat: Latitude
lon: Longitude
Returns:
Tuple of (hourly_periods_list, points_data)
"""
try:
# Round coordinates to 4 decimal places to avoid API redirects
lat_rounded = round(lat, 4)
lon_rounded = round(lon, 4)
# Get weather data from NOAA
weather_api = f"https://api.weather.gov/points/{lat_rounded},{lon_rounded}"
# Get the forecast URL (with retry logic)
weather_data = self._noaa_fetch(weather_api, "weather data")
if weather_data is None:
return self.ERROR_FETCHING_DATA, None
weather_json = weather_data.json()
hourly_forecast_url = weather_json['properties'].get('forecastHourly')
if not hourly_forecast_url:
self.logger.warning("Hourly forecast not available for this location")
return self.ERROR_FETCHING_DATA, None
# Get the hourly forecast (with retry logic)
hourly_data = self._noaa_fetch(self._noaa_units_url(hourly_forecast_url), "hourly forecast")
if hourly_data is None:
return self.ERROR_FETCHING_DATA, None
hourly_json = hourly_data.json()
hourly_periods = hourly_json['properties']['periods']
if not hourly_periods:
self.logger.warning("No hourly periods returned from NOAA")
return self.ERROR_FETCHING_DATA, None
return hourly_periods, weather_json
except Exception as e:
self.logger.error(f"Error fetching NOAA hourly weather: {e}")
return self.ERROR_FETCHING_DATA, None
def format_hourly_forecast(self, hourly_periods: list, max_length: int = 130) -> str:
"""Format hourly forecast to fit as many hours as possible in max_length bytes
Args:
hourly_periods: List of hourly forecast periods from NOAA
max_length: Maximum message length (default 130 for backwards compatibility)
Returns:
Formatted string with one hour per line
"""
try:
if not hourly_periods:
return self.translate('commands.wx.hourly_not_available')
future_periods = self._future_hourly_periods(hourly_periods)
if not future_periods:
return "No future hourly periods available"
return self._pack_hourly_lines(
(self._hourly_line(period) for period in future_periods), max_length
)
except Exception as e:
self.logger.error(f"Error formatting hourly forecast: {e}")
return f"Error formatting hourly forecast: {str(e)}"
@staticmethod
def _parse_noaa_start_time(start_time_str: str) -> datetime:
"""Parse a NOAA ISO ``startTime``; a trailing ``Z`` means UTC."""
if 'Z' in start_time_str:
return datetime.fromisoformat(start_time_str.replace('Z', '+00:00'))
return datetime.fromisoformat(start_time_str)
def _future_hourly_periods(self, hourly_periods: list) -> list:
"""Periods starting after now; unparseable or missing times are kept.
NOAA times carry the location's UTC offset, so they are compared as
absolute times. Dropping the offset compared the location's wall clock
with the bot's, which kept past hours or dropped future ones whenever
the two were in different time zones.
"""
now_utc = datetime.now(timezone.utc)
now_naive = datetime.now()
future_periods = []
for period in hourly_periods:
start_time_str = period.get('startTime', '')
if not start_time_str:
future_periods.append(period)
continue
try:
start_time = self._parse_noaa_start_time(start_time_str)
except (ValueError, TypeError):
future_periods.append(period)
continue
now = now_utc if start_time.tzinfo else now_naive
if start_time > now:
future_periods.append(period)
return future_periods
def _hourly_line(self, period: dict) -> str:
"""One hour as "10AM: 🌦️ 26% Chance Light Rain 49° SSW5"."""
temp = period.get('temperature', '')
short_forecast = period.get('shortForecast', '')
wind_speed = period.get('windSpeed', '')
wind_direction = period.get('windDirection', '')
precip_prob = period.get('probabilityOfPrecipitation', {}).get('value')
time_str = self._hour_label(period.get('startTime', ''))
emoji = self.get_weather_emoji(short_forecast)
forecast_short = self._short_hourly_description(short_forecast)
line_parts = []
if time_str:
line_parts.append(f"{time_str}:")
line_parts.append(emoji)
if precip_prob is not None and precip_prob > 0:
line_parts.append(f"{precip_prob}%")
line_parts.append(forecast_short)
if _has_temp(temp):
line_parts.append(f"{temp}°")
if wind_speed and wind_direction:
wind_match = re.search(r'(\d+)', wind_speed)
if wind_match:
wind_dir_abbrev = self._without_arrow(self.abbreviate_wind_direction(wind_direction))
line_parts.append(f"{wind_dir_abbrev}{self._noaa_wind_convert(wind_match.group(1), wind_speed)}")
return " ".join(line_parts)
@staticmethod
def _period_date(period: dict):
"""The local calendar date a NOAA period starts on, from its startTime, or None.
NOAA writes startTime in the location's own offset ("2026-10-01T18:00:00-07:00"),
so its date part is the local date there, whatever the bot's clock says.
"""
start = period.get('startTime') if isinstance(period, dict) else None
if not isinstance(start, str) or len(start) < 10:
return None
try:
return datetime.strptime(start[:10], '%Y-%m-%d').date()
except ValueError:
return None
def _forecast_today(self, forecast: list):
"""Today's date at the forecast's location, or the bot's date when no period has an offset.
The location's "now" is the current instant in the UTC offset NOAA writes on
its period times. The first period's own date is not enough: at 1 AM the
first period can still be the night that started yesterday evening.
"""
for period in forecast:
start = period.get('startTime') if isinstance(period, dict) else None
if not isinstance(start, str):
continue
try:
start_dt = self._parse_noaa_start_time(start)
except (ValueError, TypeError):
continue
if start_dt.tzinfo is not None:
return datetime.now(timezone.utc).astimezone(start_dt.tzinfo).date()
return datetime.now().date()
def _period_dates(self, forecast: list) -> list:
"""Each period's local date: its startTime date, or inferred from its neighbors when it has none.
NOAA periods alternate day and night in order, so an undated period is the
same date as a preceding daytime period when it is a night, else the day after.
"""
dates = [self._period_date(p) for p in forecast]
def is_night(period) -> bool:
day = period.get('isDaytime') if isinstance(period, dict) else None
if day is not None:
return not day
return 'night' in str(period.get('name', '') if isinstance(period, dict) else '').lower()
for i in range(1, len(dates)):
if dates[i] is None and dates[i - 1] is not None:
same_day = not is_night(forecast[i - 1]) and is_night(forecast[i])
dates[i] = dates[i - 1] if same_day else dates[i - 1] + timedelta(days=1)
for i in range(len(dates) - 2, -1, -1):
if dates[i] is None and dates[i + 1] is not None:
same_day = not is_night(forecast[i]) and is_night(forecast[i + 1])
dates[i] = dates[i + 1] if same_day else dates[i + 1] - timedelta(days=1)
return dates
def _find_tomorrow_periods(self, forecast: list) -> list:
"""Tomorrow's NOAA periods: those dated the day after today at the location;
without dates, those named "Tomorrow", else named for tomorrow's weekday, else the
(up to two) periods after today's.
"""
# Dated periods: tomorrow is the day after today at the location.
if any(self._period_date(p) for p in forecast):
tomorrow = self._forecast_today(forecast) + timedelta(days=1)
dated = [p for p, d in zip(forecast, self._period_dates(forecast), strict=True) if d == tomorrow]
if dated:
return dated
# No usable startTime: fall back to period names and the bot's clock.
tomorrow_day_name = (datetime.now() + timedelta(days=1)).strftime('%A')
tomorrow_periods = [p for p in forecast if 'tomorrow' in p.get('name', '').lower()]
if tomorrow_periods:
return tomorrow_periods
for period in forecast:
period_name_lower = period.get('name', '').lower()
if tomorrow_day_name.lower() in period_name_lower:
# A name like "Monday" can also be today's; skip those.
today_day_name = datetime.now().strftime('%A')
if today_day_name.lower() not in period_name_lower:
tomorrow_periods.append(period)
if tomorrow_periods:
return tomorrow_periods
# Generic names: take the periods after today's (Today, This Afternoon,
# This Evening, Tonight), usually tomorrow's day and night.
found_tonight = False
for period in forecast:
period_name = period.get('name', '').lower()
if any(word in period_name for word in ['today', 'this afternoon', 'this evening', 'tonight']):
found_tonight = True
continue
if found_tonight:
tomorrow_periods.append(period)
if len(tomorrow_periods) >= 2:
break
return tomorrow_periods
def format_tomorrow_forecast(self, forecast: list, max_length: int = 130) -> str:
"""Format a detailed forecast for tomorrow"""
try:
tomorrow_periods = self._find_tomorrow_periods(forecast)
if not tomorrow_periods:
return self.translate('commands.wx.tomorrow_not_available')
# Build detailed forecast for tomorrow
parts = []
for period in tomorrow_periods:
period_name = self._noaa_period_display_name(period)
temp = period.get('temperature', '')
temp_unit = period.get('temperatureUnit', 'F')
short_forecast = period.get('shortForecast', '')
detailed_forecast = period.get('detailedForecast', '')
wind_speed = period.get('windSpeed', '')
wind_direction = period.get('windDirection', '')
if not _has_temp(temp) or not short_forecast:
continue
# Create period string
emoji = self.get_weather_emoji(short_forecast)
period_str = f"{period_name}: {emoji}{short_forecast} {temp}°{temp_unit}"
# Wind info, which goes first when the reply is too long
wind = ""
if wind_speed and wind_direction:
wind_match = re.search(r'(\d+)', wind_speed)
if wind_match:
wind_num = self._noaa_wind_convert(wind_match.group(1), wind_speed)
wind_dir = self.abbreviate_wind_direction(wind_direction)
if wind_dir:
wind = f" {wind_dir}{wind_num}"
# Try to extract high/low
high_low = self.extract_high_low(
detailed_forecast, self._noaa_period_temp_symbol(period)
)
if high_low and '°' not in (period_str + wind).split()[-1]: # Avoid duplicate temp
period_str = period_str.replace(f" {temp}°{temp_unit}", f" {high_low}")
parts.append((period_str + wind, period_str))
if not parts:
return self.translate('commands.wx.tomorrow_not_available')
return self._fit_tomorrow_parts(parts, max_length)
except Exception as e:
self.logger.error(f"Error formatting tomorrow forecast: {e}")
return self.translate('commands.wx.tomorrow_error')
def _fit_tomorrow_parts(self, parts: list[tuple[str, str]], max_length: int) -> str:
"""Join tomorrow's periods within *max_length* UTF-8 bytes.
Each part is (with wind, without wind). The later periods' wind goes
first, then the first period's, then the later periods themselves. When
even the first period alone does not fit, it is sent anyway (the send
path splits it) rather than leaving the reply empty.
"""
full = [with_wind for with_wind, _ in parts]
bare = [without for _, without in parts]
candidates = [full]
if len(parts) > 1:
candidates.append(full[:1] + bare[1:])
candidates += [bare, full[:1], bare[:1]]
for candidate in candidates:
text = " | ".join(candidate)
if self._count_display_width(text) <= max_length:
return text
return bare[0]
def format_multiday_forecast(self, forecast: list, num_days: int = 7, max_length: int = 130) -> str:
"""Format a less detailed multi-day forecast summary"""
try:
if any(self._period_date(p) for p in forecast):
parts = self._multiday_lines_by_date(forecast, num_days)
if not parts:
return self.translate('commands.wx.multiday_not_available', num_days=num_days)
return "\n".join(parts)
# No usable startTime: group by weekday name against the bot's clock.
# One entry per weekday; a day period wins over a night one.
days: dict[str, dict] = {}
for period in forecast:
period_name_lower = period.get('name', '').lower()
day_name = self._multiday_day_name(period_name_lower)
if not day_name:
continue
# Prefer the high/low from the detailed text over the bare temperature.
temp = period.get('temperature', '')
detailed_forecast = period.get('detailedForecast', '')
high_low = self.extract_high_low(
detailed_forecast, self._noaa_period_temp_symbol(period)
)
if high_low:
temp_str = high_low
elif _has_temp(temp):
temp_str = f"{temp}°"
else:
continue
short_forecast = period.get('shortForecast', '')
if not short_forecast:
continue
# "tonight" contains "night", so this also covers Tonight.
is_day = 'night' not in period_name_lower
if day_name not in days or is_day or not days[day_name]['is_day']:
days[day_name] = {'temp': temp_str, 'forecast': short_forecast, 'is_day': is_day}
if not days:
return self.translate('commands.wx.multiday_not_available', num_days=num_days)
# Format as compact summary
parts = []
day_order = ['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday']
# Get today's day name to start ordering
today = datetime.now().strftime('%A')
# Reorder days starting from today
if today in day_order:
start_idx = day_order.index(today)
ordered_days = day_order[start_idx:] + day_order[:start_idx]
else:
ordered_days = day_order
# Limit to requested number of days
# Map day names to short abbreviations in the reply's language
day_abbrev_map = {name: self._day_abbrev(i) for i, name in enumerate(_WEEKDAY_NAMES)}
# Collect days up to num_days, starting from tomorrow (skip today)
days_collected = 0
for day in ordered_days[1:]: # Skip today, start from tomorrow
if days_collected >= num_days:
break
if day in days:
day_data = days[day]
day_abbrev = day_abbrev_map.get(day, day[:2]) # Use 2-letter abbrev
emoji = self.get_weather_emoji(day_data['forecast'])
# Abbreviate forecast text
forecast_short = self.abbreviate_noaa(day_data['forecast'])
# Further shorten if needed to fit on one line (but be less aggressive)
if len(forecast_short) > 25:
forecast_short = forecast_short[:22] + "..."
parts.append(f"{day_abbrev}: {emoji}{forecast_short} {day_data['temp']}")
days_collected += 1
if not parts:
return self.translate('commands.wx.multiday_not_available', num_days=num_days)
# Join with newlines instead of pipes
result = "\n".join(parts)
return result
except Exception as e:
self.logger.error(f"Error formatting {num_days}-day forecast: {e}")
return self.translate('commands.wx.multiday_error', num_days=num_days)
def _multiday_lines_by_date(self, forecast: list, num_days: int) -> list[str]:
"""One line per local date after today at the location, up to *num_days*.
Grouping by startTime date (instead of weekday name against the bot's
clock) keeps holiday-named periods ("Christmas Day") and does not depend
on the bot sharing the location's time zone. A daytime period wins over
that date's night period.
"""
today = self._forecast_today(forecast)
days: dict = {}
for period, day in zip(forecast, self._period_dates(forecast), strict=True):
if day is None or day <= today:
continue
temp = period.get('temperature', '')
high_low = self.extract_high_low(
period.get('detailedForecast', ''), self._noaa_period_temp_symbol(period)
)
if high_low:
temp_str = high_low
elif _has_temp(temp):
temp_str = f"{temp}°"
else:
continue
short_forecast = period.get('shortForecast', '')
if not short_forecast:
continue
is_day = period.get('isDaytime')
if is_day is None:
is_day = 'night' not in period.get('name', '').lower()
if day not in days or (is_day and not days[day]['is_day']):
days[day] = {'temp': temp_str, 'forecast': short_forecast, 'is_day': bool(is_day)}
parts = []
for day in sorted(days)[:num_days]:
data = days[day]
forecast_short = self.abbreviate_noaa(data['forecast'])
if len(forecast_short) > 25:
forecast_short = forecast_short[:22] + "..."
abbrev = self._day_abbrev(day.weekday())
parts.append(f"{abbrev}: {self.get_weather_emoji(data['forecast'])}{forecast_short} {data['temp']}")
return parts
def _day_abbrev(self, weekday: int) -> str:
"""The multi-day label for a weekday (0 = Monday), translated (English "M", "Th", "Sa")."""
key = f'commands.wx.day_abbrev.{_WEEKDAY_NAMES[weekday]}'
label = self.translate(key)
return label if isinstance(label, str) and label and label != key else _DAY_ABBREVS[weekday]
@staticmethod
def _multiday_day_name(period_name_lower: str) -> str | None:
"""The weekday a NOAA period belongs to, or None to skip it.
Part-of-day periods (Tonight, This Afternoon) count only when they name a
weekday; otherwise Today and Tomorrow map to the current and next weekday.
"""
for day in _WEEKDAYS_LOWER:
if day in period_name_lower:
return day.capitalize()
if any(word in period_name_lower for word in ['tonight', 'afternoon', 'morning', 'evening']):
return None
if 'tomorrow' in period_name_lower:
return (datetime.now() + timedelta(days=1)).strftime('%A')
if 'today' in period_name_lower:
return datetime.now().strftime('%A')
return None
def _add_period_details(self, period_str: str, detailed_forecast: str, current_weather_length: int, max_length: int = 130, observation_data: dict = None) -> str:
"""Add additional details (humidity, dew point, visibility, etc.) to a period string
Args:
period_str: The base period string (e.g., " | Today: ☀️Sunny 75°")
detailed_forecast: The detailed forecast text to extract info from
current_weather_length: Current length of the weather string (to check total length)
max_length: Maximum total length allowed (default 130)
observation_data: Optional dict with observation station data (humidity, dew_point, visibility, wind_gusts, pressure)
Returns:
Updated period string with additional details if space allows
"""
# Prefer the station observation (more accurate); fall back to parsing the forecast text.
observed = observation_data or {}
humidity = observed.get('humidity') or self.extract_humidity(detailed_forecast)
dew_point = observed.get('dew_point') or self.extract_dew_point(detailed_forecast)
visibility = observed.get('visibility') or self.extract_visibility(detailed_forecast)
wind_gusts = observed.get('wind_gusts') or self._forecast_text_gusts(detailed_forecast)
pressure = observed.get('pressure') or self.extract_pressure(detailed_forecast)
# Precipitation probability only comes from the forecast text.
precip_prob = self.extract_precip_probability(detailed_forecast)
# Add each available detail, in this order, as long as the total still fits.
result = period_str
for value, template in (
(humidity, " {}%RH"),
(dew_point, " 💧{}°"),
(visibility, " 👁️{}km" if self._noaa_metric_distance else " 👁️{}mi"),
(precip_prob, " 🌦️{}%"),
(wind_gusts, " 💨{}"),
(pressure, " 📊{}hPa"),
):
if value:
piece = template.format(value)
if self._count_display_width(result + piece) + current_weather_length <= max_length:
result += piece
return result
def get_weather_alerts_noaa(self, lat: float, lon: float, return_full_data: bool = False) -> tuple:
"""Get weather alerts from NOAA with full metadata extraction and prioritization
Args:
lat: Latitude
lon: Longitude
return_full_data: If True, return list of alert dicts instead of formatted strings
Returns:
If return_full_data=False: (full_first_alert_text, abbreviated_first_alert_text, alert_count)
If return_full_data=True: (list of alert dicts, alert_count)
"""
try:
if self._nws_no_coverage.is_unavailable(lat, lon):
self.logger.debug("Skipping NWS weather alerts for cached point %s,%s outside NWS coverage", round(lat, 2), round(lon, 2))
return self.ERROR_FETCHING_DATA
# Round coordinates to 4 decimal places to avoid API redirects
lat_rounded = round(lat, 4)
lon_rounded = round(lon, 4)
alert_url = f"https://api.weather.gov/alerts/active.atom?point={lat_rounded},{lon_rounded}"
try:
alert_data = self.noaa_session.get(alert_url, timeout=self.url_timeout)
if not alert_data.ok:
if nws_http_means_no_coverage(alert_data.status_code):
if self._nws_no_coverage.mark_unavailable(lat, lon):
self.logger.warning(
"NWS weather alerts unavailable (HTTP %s); NOAA alerts are US-only; "
"point %s,%s is outside NWS coverage",
alert_data.status_code, round(lat, 2), round(lon, 2),
)
else:
self.logger.warning(
f"Error fetching weather alerts from NOAA: HTTP {alert_data.status_code}"
)
return self.ERROR_FETCHING_DATA
except (requests.exceptions.Timeout, requests.exceptions.ConnectionError) as e:
self.logger.warning(f"Timeout/connection error fetching weather alerts from NOAA: {e}")
return self.ERROR_FETCHING_DATA
self._nws_no_coverage.mark_available(lat, lon)
alerts = [] # Store structured alert data
alertxml = xml.dom.minidom.parseString(alert_data.text)
for entry in alertxml.getElementsByTagName("entry"):
try:
title = entry_title(entry)
summary = entry_summary(entry)
nws_headline = entry_nws_headline(entry)
alerts.append(parse_alert_fields(entry, title, summary, nws_headline, WX_SPECIAL_RULES))
except Exception as e:
self.logger.warning(f"Error parsing alert entry: {e}")
# Fallback: just use title
if title:
alerts.append({
'title': title,
'summary': '',
'nws_headline': '',
'event': title.split()[0] if title else "",
'event_type': 'Unknown',
'severity': 'Unknown',
'urgency': 'Unknown',
'certainty': 'Unknown',
'effective': '',
'expires': '',
'area_desc': '',
'office': ''
})
if not alerts:
return self.NO_ALERTS
# Post-process alerts to differentiate duplicate Special Statements
# If multiple statements have the same event, add distinguishing details
alerts = self._differentiate_duplicate_statements(alerts)
# Prioritize alerts using hybrid scoring
alerts = self._prioritize_alerts(alerts)
if return_full_data:
return alerts, len(alerts)
# Format for compact display (backward compatibility)
# Return first alert formatted, plus count
first_alert = alerts[0]
full_first_alert_text = self._format_alert_compact(first_alert, include_details=True)
abbreviated_first_alert_text = self._format_alert_compact(first_alert, include_details=False)
return full_first_alert_text, abbreviated_first_alert_text, len(alerts)
except Exception as e:
self.logger.error(f"Error fetching NOAA weather alerts: {e}")
return self.ERROR_FETCHING_DATA
async def _get_weather_alerts_noaa_async(
self, lat: float, lon: float, return_full_data: bool = False
) -> tuple:
return await self._run_sync_provider_async(
self.get_weather_alerts_noaa,
lat,
lon,
return_full_data,
)
def _differentiate_duplicate_statements(self, alerts: list) -> list:
"""Differentiate Special Statements that have the same event type by adding unique details
Args:
alerts: List of alert dicts
Returns:
List of alerts with differentiated event names for duplicate statements
"""
# Group alerts by event type and event name
statement_groups = {}
for alert in alerts:
if alert.get('event_type') == 'Statement':
event = alert.get('event', 'Special')
if event not in statement_groups:
statement_groups[event] = []
statement_groups[event].append(alert)
# For each group with multiple statements, differentiate them
for event, group in statement_groups.items():
if len(group) > 1:
# Multiple statements with same event - need to differentiate
for i, alert in enumerate(group):
nws_headline = alert.get('nws_headline', '')
summary = alert.get('summary', '')
effective = alert.get('effective', '')
# Try to extract unique distinguishing details
distinguishing_detail = ""
# Strategy 1: Extract unique keywords from headline that aren't in other headlines
if nws_headline:
headline_lower = nws_headline.lower()
# Look for unique time references
if 'today' in headline_lower or 'now' in headline_lower:
distinguishing_detail = " (Today)"
elif 'week' in headline_lower or 'past week' in headline_lower:
distinguishing_detail = " (Week)"
elif 'continues' in headline_lower or 'remains' in headline_lower:
distinguishing_detail = " (Ongoing)"
# Look for unique severity/impact words
if not distinguishing_detail:
if 'increased' in headline_lower or 'increasing' in headline_lower:
distinguishing_detail = " (Increased)"
elif 'new' in headline_lower:
distinguishing_detail = " (New)"
elif 'update' in headline_lower:
distinguishing_detail = " (Update)"
# Strategy 2: Use timing to differentiate (morning vs afternoon vs evening)
if not distinguishing_detail and effective:
try:
from datetime import datetime
# Try to parse effective time
if 'T' in effective:
dt = datetime.fromisoformat(effective.replace('Z', '+00:00'))
hour = dt.hour
if 5 <= hour < 12:
distinguishing_detail = " (AM)"
elif 12 <= hour < 17:
distinguishing_detail = " (PM)"
elif 17 <= hour < 21:
distinguishing_detail = " (Eve)"
else:
distinguishing_detail = " (Night)"
except:
pass
# Strategy 3: Extract unique topic from summary if headline didn't help
if not distinguishing_detail and summary:
summary_lower = summary.lower()
# Look for secondary topics that might be unique
# Check for specific locations, conditions, or impacts
if 'burn' in summary_lower or 'burned area' in summary_lower:
distinguishing_detail = " (Burn)"
elif 'coastal' in summary_lower:
distinguishing_detail = " (Coastal)"
elif 'urban' in summary_lower:
distinguishing_detail = " (Urban)"
elif 'mountain' in summary_lower or 'cascade' in summary_lower:
distinguishing_detail = " (Mtn)"
# Strategy 4: Use index as last resort (but make it subtle)
if not distinguishing_detail:
distinguishing_detail = f" ({i+1})"
# Update the event name with distinguishing detail
alert['event'] = event + distinguishing_detail
return alerts
def _prioritize_alerts(self, alerts: list) -> list:
"""Prioritize alerts using hybrid scoring system
Scoring:
- Severity: Extreme=100, Severe=75, Moderate=50, Minor=25, Unknown=0
- Urgency: Immediate=50, Expected=30, Future=10, Past=0
- Event Type: Warning=40, Watch=30, Advisory=20, Statement=10
- Time: (hours until expiration) * -5 (sooner = higher score)
Returns sorted list (highest priority first)
"""
def calculate_score(alert):
score = 0
# Severity score
severity_scores = {
'Extreme': 100,
'Severe': 75,
'Moderate': 50,
'Minor': 25,
'Unknown': 0
}
score += severity_scores.get(alert.get('severity', 'Unknown'), 0)
# Urgency score
urgency_scores = {
'Immediate': 50,
'Expected': 30,
'Future': 10,
'Past': 0,
'Unknown': 0
}
score += urgency_scores.get(alert.get('urgency', 'Unknown'), 0)
# Event type score
event_type_scores = {
'Warning': 40,
'Watch': 30,
'Advisory': 20,
'Statement': 10,
'Unknown': 0
}
score += event_type_scores.get(alert.get('event_type', 'Unknown'), 0)
# Time urgency (estimate hours until expiration)
expires = alert.get('expires', '')
expires_hours = 999 # Default to far future
if expires:
try:
# Try to parse expiration time
if 'at' in expires.lower():
# Rough estimate: if it says "6:00AM" assume it's today or tomorrow
time_match = re.search(r'(\d+):?(\d+)?(AM|PM)', expires, re.IGNORECASE)
if time_match:
# For simplicity, assume alerts expire within 48 hours
expires_hours = 24 # Default estimate
except:
pass
# Time score: sooner expiration = higher priority
# Subtract hours (sooner = higher score)
score += max(0, 50 - expires_hours)
return score
# Sort by score (descending), then by event type, then by title
sorted_alerts = sorted(alerts, key=lambda a: (
-calculate_score(a), # Negative for descending
{'Warning': 0, 'Watch': 1, 'Advisory': 2, 'Statement': 3, 'Unknown': 4}.get(a.get('event_type', 'Unknown'), 4),
a.get('title', '')
))
return sorted_alerts
def _format_alert_compact(self, alert: dict, include_details: bool = True) -> str:
"""Format a single alert compactly
Shares its formatting with the proactive WeatherService broadcasts via
``modules.alert_format``, so both localize from one code path.
Args:
alert: Alert dict with event, event_type, severity, expires, office, etc.
include_details: If True, include expiration time and office
Returns:
Formatted alert string, e.g. "🟠High Wind Warn til 6AM by NWS SEA"
"""
return alert_format.format_alert_compact(
alert,
self.response_translator,
include_details=include_details,
# !wx alerts has never shown the area description; the proactive
# broadcast does, so the shared helper keeps it optional.
include_location=False,
)
def _format_alerts_compact_summary(self, alerts: list, alert_count: int, max_length: int = 130) -> str:
"""Format multiple alerts with prioritized first alert and summary of others
Args:
alerts: List of prioritized alert dicts
alert_count: Total number of alerts
max_length: Maximum message length (default 130 for backwards compatibility)
Returns:
Compact formatted string: "4 alerts: 🟠High Wind Warn til 6AM | +3: 🌊Flood Watch, ❄️Freeze Adv, 🌫️Dense Fog Adv"
"""
if not alerts:
return f"{alert_count} alerts"
# Format first (highest priority) alert with details
first_alert = alerts[0]
first_alert_text = self._format_alert_compact(first_alert, include_details=True)
# If only one alert, return it
if alert_count == 1:
return f"{alert_count} alert: {first_alert_text}"
# Build summary of remaining alerts
remaining_alerts = alerts[1:]
remaining_count = len(remaining_alerts)
# Format remaining alerts as event types only
remaining_parts = []
for alert in remaining_alerts[:5]: # Limit to 5 to avoid overflow
event = alert.get('event', '')
event_type = alert.get('event_type', '')
# Get emoji for event type
event_emoji = self._get_event_emoji(event, event_type)
# Build compact event string. Trimmed harder than the lead alert:
# first word only, since these are a comma-joined tail.
event_short = alert_format.shorten_event(event, limit=12, max_words=1)
event_type_abbrev = alert_format.event_type_abbrev(event_type, self.response_translator)
if event:
remaining_parts.append(f"{event_emoji}{event_short} {event_type_abbrev}")
else:
remaining_parts.append(f"{event_emoji}{event_type_abbrev}")
# Build summary
if remaining_count > 5:
remaining_summary = f"+{remaining_count}: {', '.join(remaining_parts[:5])}..."
else:
remaining_summary = f"+{remaining_count}: {', '.join(remaining_parts)}"
# Combine: first alert + summary
result = f"{alert_count} alerts: {first_alert_text} | {remaining_summary}"
# Check if it fits in max_length chars, truncate if needed
if self._count_display_width(result) > max_length:
# Try shorter first alert
first_alert_text_short = self._format_alert_compact(first_alert, include_details=False)
result = f"{alert_count} alerts: {first_alert_text_short} | {remaining_summary}"
# If still too long, truncate remaining summary
if self._count_display_width(result) > max_length:
max_remaining = 3
while max_remaining > 0 and self._count_display_width(result) > max_length:
if remaining_count > max_remaining:
remaining_summary = f"+{remaining_count}: {', '.join(remaining_parts[:max_remaining])}..."
else:
remaining_summary = f"+{remaining_count}: {', '.join(remaining_parts[:max_remaining])}"
result = f"{alert_count} alerts: {first_alert_text_short} | {remaining_summary}"
max_remaining -= 1
return result
def _get_event_emoji(self, event: str, event_type: str) -> str:
"""Get emoji for event type"""
event_lower = event.lower() if event else ""
# Weather event emojis
if any(word in event_lower for word in ['flood', 'flooding']):
return '🌊'
elif any(word in event_lower for word in ['wind', 'gale']):
return '💨'
elif any(word in event_lower for word in ['snow', 'winter', 'blizzard']):
return '❄️'
elif any(word in event_lower for word in ['fog', 'smoke', 'haze']):
return '🌫️'
elif any(word in event_lower for word in ['heat', 'excessive heat']):
return '🌡️'
elif any(word in event_lower for word in ['freeze', 'frost']):
return '🧊'
elif any(word in event_lower for word in ['thunderstorm', 'tornado']):
return '⛈️'
elif any(word in event_lower for word in ['fire', 'red flag']):
return '🔥'
elif any(word in event_lower for word in ['hurricane', 'tropical']):
return '🌀'
elif any(word in event_lower for word in ['tsunami']):
return '🌊'
else:
# Default by event type
return {
'Warning': '⚠️',
'Watch': '👁️',
'Advisory': 'ℹ️',
'Statement': '📢'
}.get(event_type, '⚠️')
def _format_alert_full(self, alert: dict, index: int = None) -> str:
"""Format a single alert with full details for multi-message display
Args:
alert: Alert dict
index: Optional alert number (1-based)
Returns:
Formatted alert string with start/stop times
"""
translator = self.response_translator
parts = []
if index is not None:
parts.append(f"{index}.")
# No trimming here — this form is sent across as many messages as it needs.
parts.append(
alert_format.severity_emoji(alert.get('severity', 'Unknown'))
+ alert_format.format_event_label(
alert.get('event', ''), alert.get('event_type', ''), translator, limit=None
)
)
window = alert_format.format_alert_window(alert, translator)
if window:
parts.append(window)
# The full form is not length-capped, so it keeps the longer fallback.
office = alert_format.format_office(alert.get('office', ''), translator, limit=15)
if office:
parts.append(office)
return " ".join(parts)
async def _send_full_alert_list(self, message: MeshMessage, lat: float, lon: float) -> bool:
"""Send full list of alerts with details, splitting across multiple messages if needed.
Returns whether every message was sent; a refused first message ends the list.
"""
# Get full alert data
alerts_result = await self._get_weather_alerts_noaa_async(
lat, lon, return_full_data=True
)
if alerts_result == self.ERROR_FETCHING_DATA:
return bool(await self.send_response(message, self.translate('commands.wx.error_fetching')))
elif alerts_result == self.NO_ALERTS:
return bool(await self.send_response(message, "No weather alerts"))
alerts, alert_count = alerts_result
if not alerts:
return bool(await self.send_response(message, "No weather alerts"))
# Format each alert with full details
alert_lines = []
for i, alert in enumerate(alerts, 1):
alert_line = self._format_alert_full(alert, index=i)
alert_lines.append(alert_line)
# Send alerts, splitting into multiple messages if needed
rate_limit = self.bot.config.getfloat('Bot', 'bot_tx_rate_limit_seconds', fallback=1.0)
sleep_time = max(rate_limit + 1.0, 2.0)
# Get max message length dynamically
max_length = self.get_max_message_length(message)
# Group alerts into messages that fit within max_length chars
current_message = f"{alert_count} alerts:"
messages = []
for line in alert_lines:
# Check if adding this line would exceed limit
test_message = current_message + "\n" + line if current_message else line
if self._count_display_width(test_message) > max_length:
# Current message is full, start new one
if current_message:
messages.append(current_message)
current_message = line
else:
# Add to current message
if current_message:
current_message += "\n" + line
else:
current_message = line
# Add last message
if current_message:
messages.append(current_message)
# Send all messages (per-user rate limit applies only to first; skip for continuations)
for i, msg in enumerate(messages):
if not await self.send_response(message, msg, skip_user_rate_limit=(i > 0)):
return False
if i < len(messages) - 1:
await self._pace_reply(message, sleep_time)
return True
def abbreviate_city_name(self, city: str) -> str:
"""Abbreviate city names for compact display (e.g., Seattle -> SEA)"""
return alert_format.abbreviate_city_name(city)
def compact_time(self, time_str: str) -> str:
"""Compact time format: '6:00AM' -> '6AM', 'December 16 at 3:12PM' -> 'Dec 16 3:12PM'
Also handles ISO format: '2025-12-17T01:00:00-08:00' -> 'Dec 17 1AM'"""
return alert_format.compact_time(time_str, self.response_translator)
def abbreviate_wind_direction(self, direction: str) -> str:
"""Abbreviate wind direction to emoji + 2-3 characters"""
if not direction:
return ""
direction = direction.upper()
# NOAA sends 16-point abbreviations ("WNW"); keep them, with the nearest
# 8-point arrow. Before, they fell through to the 2-character fallback
# below, which turned "WNW" into "WN" and dropped the arrow.
if direction in _COMPASS_16:
arrow = _ARROWS_8[int(_COMPASS_16.index(direction) / 2 + 0.5) % 8]
return f"{arrow}{self._wind_letters(direction)}"
replacements = {
"NORTHWEST": "NW",
"NORTHEAST": "NE",
"SOUTHWEST": "SW",
"SOUTHEAST": "SE",
"NORTH": "N",
"EAST": "E",
"SOUTH": "S",
"WEST": "W"
}
for full, point in replacements.items():
if full in direction:
arrow = _ARROWS_8[_COMPASS_16.index(point) // 2]
return f"{arrow}{self._wind_letters(point)}"
# If no match, return first 2 characters with generic wind emoji
return f"💨{direction[:2]}" if len(direction) >= 2 else f"💨{direction}"
def _wind_letters(self, point: str) -> str:
"""A 16-point compass abbreviation in the reply's language ("NE" is "NO" in German), as gwx shows it."""
key = f"common.wind_directions.{point}"
label = self.translate(key)
return label if isinstance(label, str) and label and label != key else point
def extract_humidity(self, text: str) -> str:
"""Extract humidity percentage from forecast text"""
return _first_match(text, _HUMIDITY_PATTERNS)
def extract_precip_chance(self, text: str) -> str:
"""Extract precipitation chance from forecast text"""
return _first_match(text, _PRECIP_CHANCE_PATTERNS)
def extract_high_low(self, text: str, units_str: str = "°F") -> str:
"""Extract high/low temperatures from forecast text; format via [Weather] templates."""
if not text:
return ""
# NOAA writes cold values as plain negatives ("Low around -5."); the ranges
# cover the coldest and hottest forecasts it issues.
def _single_ok(val: int) -> bool:
if units_str == "°C":
return -55 <= val <= 55
return -65 <= val <= 130
def _pair_ok(hi: int, lo: int) -> bool:
return _single_ok(hi) and _single_ok(lo) and hi > lo
pair_patterns = [
r'high\s+near\s+(-?\d+).*?low\s+around\s+(-?\d+)',
r'high\s+(-?\d+).*?low\s+(-?\d+)',
r'(-?\d+)\s+to\s+(-?\d+)\s+degrees',
r'temperature\s+(-?\d+)\s+to\s+(-?\d+)',
r'high\s+near\s+(-?\d+).*?temperatures\s+falling\s+to\s+around\s+(-?\d+)',
]
for pattern in pair_patterns:
match = re.search(pattern, text.lower())
if match and len(match.groups()) == 2:
high, low = match.groups()
try:
high_val = int(high)
low_val = int(low)
if _pair_ok(high_val, low_val):
return format_temperature_high_low(
self.bot.config, high_val, low_val, units_str, self.logger,
translator=self.response_translator,
)
except ValueError:
continue
low_match = re.search(r'low\s+around\s+(-?\d+)', text.lower())
if low_match:
try:
low_val = int(low_match.group(1))
if _single_ok(low_val):
return format_temperature_high_low(
self.bot.config, None, low_val, units_str, self.logger,
translator=self.response_translator,
)
except ValueError:
pass
high_match = re.search(r'high\s+near\s+(-?\d+)', text.lower())
if high_match:
try:
high_val = int(high_match.group(1))
if _single_ok(high_val):
return format_temperature_high_low(
self.bot.config, high_val, None, units_str, self.logger,
translator=self.response_translator,
)
except ValueError:
pass
return ""
def extract_uv_index(self, text: str) -> str:
"""Extract UV index from forecast text"""
return _first_match(text, _UV_INDEX_PATTERNS, 0, 15)
def extract_dew_point(self, text: str) -> str:
"""Extract dew point temperature from forecast text"""
return _first_match(text, _DEW_POINT_PATTERNS, -20, 80)
def extract_visibility(self, text: str) -> str:
"""Extract visibility from forecast text"""
return _first_match(text, _VISIBILITY_PATTERNS, 0, 20)
def extract_precip_probability(self, text: str) -> str:
"""Extract precipitation probability from forecast text"""
return _first_match(text, _PRECIP_PROBABILITY_PATTERNS, 0, 100)
def extract_wind_gusts(self, text: str) -> str:
"""Extract wind gusts from forecast text"""
return _first_match(text, _WIND_GUST_PATTERNS, 10, 100)
def _forecast_text_gusts(self, text: str) -> str:
"""Gusts from the forecast text ("gusts up to 30 mph" or "48 km/h") in the configured unit."""
mph = self.extract_wind_gusts(text)
if mph:
return self._noaa_wind_convert(mph, "mph")
kmh = _first_match(text, _WIND_GUST_KMH_PATTERNS, 16, 161)
if kmh:
return self._noaa_wind_convert(kmh, "km/h")
return ""
def extract_pressure(self, text: str) -> str:
"""Extract barometric pressure from forecast text"""
return _first_match(text, _PRESSURE_PATTERNS, 600, 1100)
def get_observation_data(self, points_data: dict) -> dict:
"""Get observation station data from NOAA and return as a dict
Returns:
Dict with keys: humidity, dew_point, visibility, wind_gusts, pressure
Values are strings ready for display, or None if not available
"""
try:
if not points_data:
return {}
weather_json = points_data
station_url = weather_json['properties'].get('observationStations')
if not station_url:
return {}
# Get the nearest station (with retry logic)
# Use shorter timeout for optional observation data to avoid blocking main response
obs_timeout = min(self.url_timeout, 5) # Cap at 5 seconds for optional data
try:
stations_data = self.noaa_session.get(station_url, timeout=obs_timeout)
if not stations_data.ok:
return {}
except (requests.exceptions.Timeout, requests.exceptions.ConnectionError):
return {}
stations_json = stations_data.json()
if not stations_json.get('features'):
return {}
# Get current observations from the nearest station (with retry logic)
station_id = stations_json['features'][0]['properties']['stationIdentifier']
obs_url = f"https://api.weather.gov/stations/{station_id}/observations/latest"
try:
obs_data = self.noaa_session.get(obs_url, timeout=obs_timeout)
if not obs_data.ok:
return {}
except (requests.exceptions.Timeout, requests.exceptions.ConnectionError):
return {}
obs_json = obs_data.json()
if not obs_json.get('properties'):
return {}
props = obs_json['properties']
obs_data_dict = {}
# Extract useful current conditions
# Check for None explicitly to handle cases where value exists but is None
humidity_val = props.get('relativeHumidity', {}).get('value')
if humidity_val is not None:
humidity = int(humidity_val)
obs_data_dict['humidity'] = str(humidity)
temp_unit, wind_unit = self._noaa_units()
dewpoint_val = props.get('dewpoint', {}).get('value')
if dewpoint_val is not None:
if temp_unit == 'celsius':
dewpoint = int(round(dewpoint_val))
else:
dewpoint = int(dewpoint_val * 9/5 + 32) # Convert C to F
obs_data_dict['dew_point'] = str(dewpoint)
visibility_val = props.get('visibility', {}).get('value')
if visibility_val is not None and str(props.get('visibility', {}).get('unitCode', '')).endswith(':km'):
visibility_val *= 1000 # NOAA normally reports meters; handle km too
if visibility_val is not None:
if self._noaa_metric_distance:
visibility = int(visibility_val / 1000) # Convert m to km
else:
visibility = int(visibility_val * 0.000621371) # Convert m to miles
if visibility > 0:
obs_data_dict['visibility'] = str(visibility)
wind_gust_val = props.get('windGust', {}).get('value')
if wind_gust_val is not None:
# NOAA observations declare their unit; station gusts are usually km/h, not m/s.
if 'km_h' in str(props.get('windGust', {}).get('unitCode', '')):
wind_gust_val = wind_gust_val / 3.6
# Shown only above 10 mph, whatever unit it is shown in.
if int(wind_gust_val * 2.237) > 10:
factor = {'mph': 2.237, 'kmh': 3.6, 'ms': 1.0, 'kn': 1.944}[wind_unit]
obs_data_dict['wind_gusts'] = str(int(wind_gust_val * factor))
pressure_val = props.get('barometricPressure', {}).get('value')
if pressure_val is not None:
pressure = int(pressure_val / 100) # Convert Pa to hPa
obs_data_dict['pressure'] = str(pressure)
return obs_data_dict
except Exception as e:
self.logger.debug(f"Error getting observation data: {e}")
return {}
def get_weather_emoji(self, condition: str) -> str:
"""Get emoji for weather condition"""
if not condition:
return ""
condition_lower = condition.lower()
# Weather condition emojis
# Order matters: thunderstorms outrank everything, even sun ("Sunny then Slight
# Chance Showers And Thunderstorms"); precipitation outranks cloud ("Chance Rain
# Showers then Mostly Cloudy" is a rain forecast); and a specific phrase has to be
# tested before a word it contains ("partly cloudy" before "cloudy").
if any(word in condition_lower for word in ['thunderstorm', 't-storm']):
return "⛈️"
elif any(word in condition_lower for word in ['sunny', 'clear']):
return "☀️"
elif any(word in condition_lower for word in ['heavy rain', 'heavy showers', 'excessive rain']):
return "🌧️" # Cloud with rain - more rain, less sun
elif any(word in condition_lower for word in ['rain', 'showers']):
return "🌦️"
elif any(word in condition_lower for word in ['snow', 'snow showers']):
return "❄️"
elif any(word in condition_lower for word in ['partly cloudy', 'mostly cloudy']):
return "⛅"
elif any(word in condition_lower for word in ['cloudy', 'overcast']):
return "☁️"
elif any(word in condition_lower for word in ['fog', 'mist', 'haze']):
return "🌫️"
elif any(word in condition_lower for word in ['smoke']) or any(word in condition_lower for word in ['windy', 'breezy']):
return "💨"
else:
return "🌤️" # Default weather emoji
# NOAA sometimes names forecast periods after federal holidays (e.g. "Washington's Birthday")
# instead of the weekday. Match these so we can resolve to weekday via startTime.
_NOAA_HOLIDAY_NAME_PATTERNS = (
"washington's birthday", "presidents day", "president's day",
"martin luther king", "mlk day", "memorial day", "labor day",
"independence day", "juneteenth", "columbus day", "veterans day",
"thanksgiving", "christmas day", "new year's day", "new year's eve",
)
def _noaa_period_display_name(self, period: dict) -> str:
"""Return display label for a NOAA forecast period. Resolves holiday names to weekday."""
name = period.get('name', '') or ''
start_time_str = period.get('startTime')
name_lower = name.lower()
is_holiday = any(p in name_lower for p in self._NOAA_HOLIDAY_NAME_PATTERNS)
if is_holiday and start_time_str:
try:
# startTime is ISO 8601, e.g. 2025-02-17T08:00:00-08:00
dt = datetime.fromisoformat(start_time_str.replace('Z', '+00:00'))
# Python weekday(): Mon=0 .. Sun=6
weekdays = ('Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun')
day_abbrev = weekdays[dt.weekday()]
if 'night' in name_lower or 'overnight' in name_lower:
return f"{day_abbrev} Night"
return day_abbrev
except (ValueError, TypeError):
pass
return self.abbreviate_noaa(name)
def abbreviate_noaa(self, text: str) -> str:
"""Replace long strings with shorter ones for display"""
replacements = {
"monday": "Mon",
"tuesday": "Tue",
"wednesday": "Wed",
"thursday": "Thu",
"friday": "Fri",
"saturday": "Sat",
"sunday": "Sun",
"northwest": "NW",
"northeast": "NE",
"southwest": "SW",
"southeast": "SE",
"north": "N",
"south": "S",
"east": "E",
"west": "W",
"precipitation": "precip",
"showers": "shwrs",
"thunderstorms": "t-storms",
"thunderstorm": "t-storm",
"quarters": "qtrs",
"quarter": "qtr",
"january": "Jan",
"february": "Feb",
"march": "Mar",
"april": "Apr",
"may": "May",
"june": "Jun",
"july": "Jul",
"august": "Aug",
"september": "Sep",
"october": "Oct",
"november": "Nov",
"december": "Dec",
"degrees": "°",
"percent": "%",
"department": "Dept.",
"amounts less than a tenth of an inch possible.": "< 0.1in",
"temperatures": "temps.",
"temperature": "temp.",
}
line = text
for key, value in replacements.items():
# Case insensitive replace
line = line.replace(key, value).replace(key.capitalize(), value).replace(key.upper(), value)
return line
async def get_weather_for_zipcode(self, zipcode: str) -> str:
"""Get weather data for a specific zipcode (legacy method)"""
return await self.get_weather_for_location(zipcode, "zipcode")
def abbreviate_alert_title(self, title: str) -> str:
"""Abbreviate alert title for brevity"""
# Common alert type abbreviations
replacements = {
"warning": "Warn",
"watch": "Watch",
"advisory": "Adv",
"statement": "Stmt",
"severe thunderstorm": "SvrT-Storm",
"tornado": "Tornado",
"flash flood": "FlashFlood",
"flood": "Flood",
"winter storm": "WinterStorm",
"blizzard": "Blizzard",
"ice storm": "IceStorm",
"freeze": "Freeze",
"frost": "Frost",
"heat": "Heat",
"excessive heat": "ExHeat",
"extreme heat": "ExtHeat",
"wind": "Wind",
"high wind": "HighWind",
"wind advisory": "WindAdv",
"fire weather": "FireWx",
"red flag": "RedFlag",
"dense fog": "DenseFog",
"issued": "iss",
"until": "til",
"effective": "eff",
"expires": "exp",
"dense smoke": "DenseSmoke",
"air quality": "AirQuality",
"coastal flood": "CoastalFlood",
"lakeshore flood": "LakeshoreFlood",
"rip current": "RipCurrent",
"high surf": "HighSurf",
"hurricane": "Hurricane",
"tropical storm": "TropStorm",
"tropical depression": "TropDep",
"storm surge": "StormSurge",
"tsunami": "Tsunami",
"earthquake": "Earthquake",
"volcano": "Volcano",
"avalanche": "Avalanche",
"landslide": "Landslide",
"debris flow": "DebrisFlow",
"dust storm": "DustStorm",
"sandstorm": "Sandstorm",
"blowing dust": "BlwDust",
"blowing sand": "BlwSand"
}
result = title
for key, value in replacements.items():
# Case insensitive replace
result = result.replace(key, value).replace(key.capitalize(), value).replace(key.upper(), value)
# Limit to reasonable length
if len(result) > 30:
result = result[:27] + "..."
return result
def get_current_conditions(self, points_data: dict) -> str:
"""Get additional current conditions data from NOAA using existing points data (legacy method)"""
obs_data = self.get_observation_data(points_data)
if not obs_data:
return ""
conditions = []
# Build conditions list in priority order
if 'humidity' in obs_data:
conditions.append(f"{obs_data['humidity']}%RH")
if 'dew_point' in obs_data:
conditions.append(f"💧{obs_data['dew_point']}°")
if 'visibility' in obs_data:
conditions.append(f"👁️{obs_data['visibility']}{'km' if self._noaa_metric_distance else 'mi'}")
if 'wind_gusts' in obs_data:
conditions.append(f"💨{obs_data['wind_gusts']}")
if 'pressure' in obs_data:
conditions.append(f"📊{obs_data['pressure']}hPa")
return " ".join(conditions[:3]) # Limit to 3 conditions to avoid overflow
Reference in new issue View Git Blame Copy Permalink