metar-taf-parser 0.4.2

METAR and TAF parsing library
Documentation

metar_taf_parser

⚠️ Status: Active development – current version 0.4.2

A modern, strongly-typed METAR and TAF parser library written in Rust.

This project is now focused on the single library crate (metar-taf-parser) and is a Rust port of the original Python project: https://github.com/mivek/python-metar-taf-parser

Porting and parity work also tracks the maintained fork: https://github.com/umpire274/python-metar-taf-parser.


✈️ Goals

  • Provide a robust Rust-native API for METAR/TAF parsing.
  • Reach feature parity with the referenced Python fork.
  • Keep the public API ergonomic and strongly typed for embedding in other apps.
  • During porting, parser groups are being migrated using the same regex-first approach used in the Python fork to preserve behavior.
  • Porting process follows a regex-alignment policy documented in docs/PORTING_REGEX_POLICY.md.

🧱 Architecture

The repository now exposes a single library crate at the root:

metar_taf_parser/
├── src/
├── tests/
├── resources/
├── Cargo.toml
├── README.md
└── CHANGELOG.md

Core library (metar-taf-parser)

  • Token-based parsing
  • Strongly typed domain models
  • Designed to be embedded in other applications
  • No CLI assumptions

📚 Library usage

Add the core crate to your Cargo.toml:

[dependencies]

metar-taf-parser = "0.4.2"

METAR example

use metar_taf_parser::parse_metar;

let metar = parse_metar(
    "LIRF 121250Z 18012KT 9999 FEW030 SCT080 18/12 Q1015"
)?;

assert_eq!(metar.station, "LIRF");
assert!(metar.wind.is_some());
assert!(metar.temperature.is_some());

TAF example (tolerant mode)

use metar_taf_parser::parse_taf;

let taf = parse_taf(
    "TAF LIRF 121100Z 1212/1318 18010KT 9999 SCT020 TX18/1214Z TN08/1304Z TEMPO 1218/1222 4000 -RA"
)?;

assert_eq!(taf.station, "LIRF");
assert!(!taf.forecasts.is_empty());
// In tolerant mode, unsupported tokens are preserved in `unparsed_groups`.

METAR strict mode example

use metar_taf_parser::parse_metar_strict;

let strict_result = parse_metar_strict(
    "LIRF 121250Z 18010KT 9999 FEW030 UNKNOWN 18/12 Q1015"
);

assert!(strict_result.is_err());

TAF strict mode example

use metar_taf_parser::parse_taf_strict;

let strict_result = parse_taf_strict(
    "TAF LIRF 121100Z 1212/1318 18010KT 9999 SCT020 UNKNOWN_TOKEN"
);

assert!(strict_result.is_err());

Natural language description — field access (METAR)

Use describe_metar when you need to inspect individual fields programmatically.

use metar_taf_parser::{parse_metar, describe_metar, Language};

let metar = parse_metar(
    "LIRF 121250Z 18010KT 9999 FEW030 18/12 Q1015"
)?;

let desc = describe_metar(&metar, Language::En);

println!("{}", desc.wind.unwrap());
// "wind from 180° at 10 kt"

println!("{}", desc.visibility.unwrap());
// "visibility greater than 10 km"

println!("{}", desc.clouds[0]);
// "few clouds at 3000 ft"

println!("{}", desc.temperature.unwrap());
// "temperature 18°C, dew point 12°C"

println!("{}", desc.pressure.unwrap());
// "QNH 1015 hPa"

Natural language description — full formatted output (METAR)

Use format_metar (or println!("{}", desc)) to print the complete report in one call. Only the fields actually present in the message are included.

use metar_taf_parser::{parse_metar, format_metar, Language};

let metar = parse_metar(
    "LIRF 121250Z 18010KT 9999 FEW030 -RA 18/12 Q1015 NOSIG"
)?;

println!("{}", format_metar(&metar, Language::En));
METAR LIRF
  Time:        Day 12 at 12:50Z
  Wind:        wind from 180° at 10 kt
  Visibility:  visibility greater than 10 km
  Weather:     light rain
  Clouds:      few clouds at 3000 ft
  Temperature: temperature 18°C, dew point 12°C
  Pressure:    QNH 1015 hPa
  Trend:       no significant change

Natural language description — field access (TAF)

use metar_taf_parser::{parse_taf, describe_taf, Language};

let taf = parse_taf(
    "TAF LIRF 121100Z 1212/1318 18010KT 9999 SCT020 TX18/1214Z TN08/1304Z TEMPO 1218/1222 4000 -RA"
)?;

let desc = describe_taf(&taf, Language::En);

println!("{}", desc.validity.unwrap());
// "12/12Z to 13/18Z"

for forecast in &desc.forecasts {
    println!("[{}]", forecast.kind);
    if let Some(wind) = &forecast.wind { println!("  {}", wind); }
    if let Some(vis)  = &forecast.visibility { println!("  {}", vis); }
    for cloud in &forecast.clouds { println!("  {}", cloud); }
}
// [Base forecast]
//   wind from 180° at 10 kt
//   visibility greater than 10 km
//   scattered clouds at 2000 ft
// [Temporary]
//   visibility 4000 m
//   light rain

Natural language description — full formatted output (TAF)

use metar_taf_parser::{parse_taf, format_taf, Language};

let taf = parse_taf(
    "TAF LIRF 121100Z 1212/1318 18010KT 9999 SCT020 TX18/1214Z TN08/1304Z TEMPO 1218/1222 4000 -RA"
)?;

println!("{}", format_taf(&taf, Language::En));
TAF LIRF
  Issued:    Day 12 at 11:00Z
  Validity:  12/12Z to 13/18Z
  [Base forecast]
    Wind:        wind from 180° at 10 kt
    Visibility:  visibility greater than 10 km
    Clouds:      scattered clouds at 2000 ft
    Max temp:    maximum temperature 18°C on day 12 at 14:00Z
    Min temp:    minimum temperature 8°C on day 13 at 04:00Z
  [Temporary]  12/18Z to 12/22Z
    Visibility:  visibility 4000 m
    Weather:     light rain

Structured METAR remarks (RMK section)

The RMK section is parsed into typed variants instead of a raw string. Access the raw text via remarks.raw; inspect parsed groups via remarks.items.

use metar_taf_parser::parse_metar;
use metar_taf_parser::metar::models::remark::{AutoStationKind, Remark};

let metar = parse_metar(
    "KORD 121750Z 18010KT 9999 FEW030 18/12 A2992 RMK AO2 SLP132 T02560178"
)?;

for remark in &metar.remarks.items {
    match remark {
        Remark::AutoStation(AutoStationKind::AO2) => {
            println!("automated station with precipitation discriminator");
        }
        Remark::SeaLevelPressure(hpa) => {
            println!("SLP: {:.1} hPa", hpa); // "SLP: 1013.2 hPa"
        }
        Remark::HourlyTemperature { temperature, dewpoint } => {
            println!("T: {:.1}°C  Td: {:.1}°C", temperature, dewpoint);
        }
        _ => {}
    }
}

// Raw text is always preserved:
assert_eq!(metar.remarks.raw, "AO2 SLP132 T02560178");

// Unrecognised tokens end up in remarks.unparsed:
let m2 = parse_metar("UOOO 181500Z 07002MPS CAVOK M25/M28 Q1014 RMK QFE746/0995")?;
assert_eq!(m2.remarks.unparsed, vec!["QFE746/0995"]);

Recognised remark variants: PeakWind, WindShift, SeaLevelPressure, PrecipitationAmount, HourlyTemperature, MaxMinTemperature, PressureTendency, AutoStation, Lightning, Virga, MaintenanceIndicator, PressureRisingRapidly, PressureFallingRapidly, SensorStatus.

The nosig field on Metar is true whenever a NOSIG trend group is present.


Typical parser use cases

  • Parse a single METAR and inspect typed fields (wind, visibility, clouds, temperature, pressure).
  • Parse a TAF and iterate forecast sections (BASE, FM, BECMG, TEMPO, PROB).
  • Enable strict TAF mode in validation pipelines where unknown groups must fail fast.
  • Use tolerant mode when you want best-effort parsing plus visibility on unsupported tokens.

📘 Documentation


🧪 Testing

cargo test


📄 License

MIT License.