zaari / nmea-parser Goto Github PK

I notice that the code isn't formatted with rustfmt, the official Rust code formatter. I was wondering if you'd be open to a PR that formatted the code with it? The benefit in my eyes is that you can just let the tool take care of it, knowing that it will be consistent with other code in the Rust ecosystem. I notice that in some places you've formatted code a particular way (tables of bits for example). If I were to format the project I'd try to make sure that things like this were preserved.

I have a receiver that sends multiple messages with GPS/UTC time. When parsing the timestamp field returns the correct h/m/s, but not the correct date in most cases. (Appends the local computer date) This is most apparent when testing on a machine where the actual date is not the same as the real GPS/UTC date.

The RMC message does parse and return the correct date in the timestamp while the GLL and GGA messages ignore the date. The NMEA string appears to have the same UTC format, but the message parsers here use different utility functions. There may be additional instances in messages I'm not receiving.

Issue:
When parsing an AIS frame. you sometime might want to access the raw data.

example:
VesselDynamicData.nav_status is an enum constructed from an u8 field. If you need to use the parser as an intermediate parser for other tool, you might want to use the underlying u8 and not manipulate the enum

solutions:

• for fields store the raw value in the object and only parse the enum when requested.
• do not store the raw value and impl "to_value" for enums and accept dataloss for enums mapping different values to one enum field -> ex: ShipType ' 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | 49 => ShipType::HighSpeedCraft,'

It seems you reverted this merge changes
( #12 )
on release 0.8.0
mainly on file:
https://github.com/zaari/nmea-parser/commits/master/src/ais/mod.rs

Any reason for it ?

I need to interact with a proprietary GPS logger that sends sentences like:

$PMTK182,8,0001F800,FFFFFFFFFFFFFF…FFFFF*52

Now I could wrap the parse_sentence function and handle those proprietary functions in the wrapper but I would have to duplicate code like checksum verification.

Possible solutions:

• Include the whole sentence as a struct field in ParseError::UnsupportedSentenceType so that wrapper can attempt to handle it after running parse_sentence.
• Return Ok(ParsedMessage::Proprietary(String)) from parse_sentence instead of ParseError::UnsupportedSentenceType.
• Factor out the verify_and_strip_checksum function and make it public.
• Add NmeaParserFull<ProprietaryMessagesType> (NmeaParser would become NmeaParserFull<()>) and a method like NmeaParserFull::register_parser_for_prefix<H>(prefix: &str, handler: H) where H: Fn(&str) -> Option<ProprietaryMessagesType>

At some point it could be interesting to use either thiserror ( https://github.com/dtolnay/thiserror )
Or manually implementing std::error::Error for the errors returned by the library.

The parser panics on incomplete sentences. For example, parsing $WIMWV,295.4,T, would panic.
Returning an error would be much nicer. I've already fixed this for a couple of sentences in my fork (https://github.com/johann2/nmea-parser). If this approach works for you, I could fix this for other sentences as well.

Code panics when parsing messages with invalid payload:
ex: Type26
!AIVDM,1,1,,,Jl@bhbmCU`:lwOd0,0*48

( I'm using this Lib to parse huge files, thus I need to avoid panic and correctly propagate errors)

But this isn't the only type it can panic on. All messages with valid decode but with invalid payload length will panic because of the way byteslices are extracted:

BitVec::from_bitslice(...
Which panics.

Although this message is invalid, valid information can be extracted from it:
If you paste it here: http://ais.tbsalling.dk/ you get everything up to the data.

Proposals:

• I could just pr fixes on each message for which I get a panic and go from there...
• Quick fix: add test on length when parsing ( wrapp all ::from_bitslice with a length check and throw error on invalid length )
• Refactor with a parsing lib such as https://github.com/Geal/nom

I could spend some time working on that but I guess it really depend on the status of the lib.
How do you this this lib future ? Have any plan for it ?

error[E0635]: unknown feature `stdsimd`
  --> /home/drebbe/.cargo/registry/src/index.crates.io-6f17d22bba15001f/ahash-0.7.6/src/lib.rs:33:42
   |
33 | #![cfg_attr(feature = "stdsimd", feature(stdsimd))]
   |                                          ^^^^^^^

For more information about this error, try `rustc --explain E0635`.

rust-lang/hashbrown#499

Hashbrown crate needs to have support for ahash 0.8.7 or newer.

$ cargo tree -i ahash -p nmea-parser
ahash v0.7.6
└── hashbrown v0.12.3
    └── nmea-parser v0.10.0

Error when parsing ais static messages for shiptype: Tanker && Other

fixed in #27

Hi, thanks for creating this library!

I encounter panic when trying to parse certain messages, an example of the issue is included as follows

use nmea_parser::NmeaParser;

fn parsemsg(sentence: &str) {
    let mut parser = NmeaParser::new();

    if let Ok(payload) = parser.parse_sentence(sentence) {
        println!("did not panic yet {:?}", payload);
    }
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn this_works() {
        let sentence = "!AIVDM,1,1,,,19NS8eP01t2rTJugUtmqOoVL00R@,0*53";
        parsemsg(sentence);
        let sentence2 = "!AIVDM,not,a,valid,nmea,string,0*00";
        parsemsg(sentence2);
    }

    #[test]
    fn this_panics() {
        let sentence = "!AIVDM,1,1,,,;ie05s`0Kk6UvFiQ`IaUfW3iC8pB,0*02";
        parsemsg(sentence);
    }
}

Produces output

running 2 tests
test errcheck::tests::this_works ... ok
test errcheck::tests::this_panics ... FAILED

failures:

---- errcheck::tests::this_panics stdout ----
thread 'errcheck::tests::this_panics' panicked at 'No such local time', /home/matt/.cargo/registry/src/github.com-1ecc6299db9ec823/chrono-0.4.19/src/offset/mod.rs:173:34
stack backtrace:
   0: std::panicking::begin_panic
             at /rustc/1.57.0/library/std/src/panicking.rs:543:12
   1: chrono::offset::LocalResult<T>::unwrap
             at /home/matt/.cargo/registry/src/github.com-1ecc6299db9ec823/chrono-0.4.19/src/offset/mod.rs:173:34
   2: chrono::offset::TimeZone::ymd
             at /home/matt/.cargo/registry/src/github.com-1ecc6299db9ec823/chrono-0.4.19/src/offset/mod.rs:214:9
   3: nmea_parser::ais::vdm_t11::handle
             at /home/matt/.cargo/registry/src/github.com-1ecc6299db9ec823/nmea-parser-0.8.0/src/ais/vdm_t11.rs:31:17
   4: nmea_parser::NmeaParser::parse_sentence
             at /home/matt/.cargo/registry/src/github.com-1ecc6299db9ec823/nmea-parser-0.8.0/src/lib.rs:577:31
   5: rust_aisdb_lib::errcheck::parsemsg
             at ./src/errcheck.rs:6:26
   6: rust_aisdb_lib::errcheck::tests::this_panics
             at ./src/errcheck.rs:26:9
   7: rust_aisdb_lib::errcheck::tests::this_panics::{{closure}}
             at ./src/errcheck.rs:24:5
   8: core::ops::function::FnOnce::call_once
             at /rustc/1.57.0/library/core/src/ops/function.rs:227:5
   9: core::ops::function::FnOnce::call_once
             at /rustc/1.57.0/library/core/src/ops/function.rs:227:5
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.


failures:
    errcheck::tests::this_panics

test result: FAILED. 1 passed; 1 failed; 0 ignored; 0 measured; 3 filtered out; finished in 0.06s

error: test failed, to rerun pass '--lib'

see:
https://gpsd.gitlab.io/gpsd/AIVDM.html#_type_27_long_range_ais_broadcast_message

There was an inversion in lon lat values when generating the parsed sentence.

Hey, do you plan on including other nmea sentences like DPT and DBT?
I would offer to implement them.

For example ,MwvData, WhwData, etc.
Since I can't find any other way to access the data contained, I assume this is not intentional?