Adds parsing of Format16 bytedata

`from_record` is now called `with_record` as the library standard
suggests

.vscode/launch.json

@@ -0,0 +1,47 @@
+{
+    // Use IntelliSense to learn about possible attributes.
+    // Hover to view descriptions of existing attributes.
+    // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
+    "version": "0.2.0",
+    "configurations": [
+        {
+            "type": "lldb",
+            "request": "launch",
+            "name": "Debug executable 'wfdb_corrosion'",
+            "cargo": {
+                "args": [
+                    "build",
+                    "--bin=wfdb_corrosion",
+                    "--package=wfdb_corrosion"
+                ],
+                "filter": {
+                    "name": "wfdb_corrosion",
+                    "kind": "bin"
+                }
+            },
+            "args": [
+                "./test_data/s1_high_resistance_bike.hea"
+            ],
+            "cwd": "${workspaceFolder}"
+        },
+        {
+            "type": "lldb",
+            "request": "launch",
+            "name": "Debug unit tests in executable 'wfdb_corrosion'",
+            "cargo": {
+                "args": [
+                    "test",
+                    "--no-run",
+                    "--bin=wfdb_corrosion",
+                    "--package=wfdb_corrosion"
+                ],
+                "filter": {
+                    "name": "wfdb_corrosion",
+                    "kind": "bin"
+                }
+            },
+            "args": [],
+            "cwd": "${workspaceFolder}"
+        }
+    ]
+}

src/headparse.rs

@@ -0,0 +1,499 @@
+use std::{collections::HashMap, hash::Hash, u64, vec};
+
+use crate::SignalFormat;
+
+/// Holds the essential header information about the record, like the amount of
+/// signals, and the sampling frequency.
+/// 
+/// Only the `name`, `signal_count`, and `sampling_freq` values are required by 
+/// default.
+#[derive(Debug, Clone)]
+struct Record {
+    name: String,
+    seg_num: Option<u64>,
+    signal_count: usize,
+    sampling_freq: u64,
+    counter_freq: Option<u64>,
+    base_counter_val: Option<u64>,
+    sample_num: Option<u64>,
+    basetime: Option<String>, // I dont thing we really need to care much about these
+    basedate: Option<String>
+}
+
+impl Record {
+    /// Attempts to generate the record information from a string of the argument 
+    /// line that's provided by the WFDB header file
+    /// 
+    /// Returns a `Result<Record, &str>`, with the error field potentially
+    /// describing why the parsing operation failed
+    pub fn from_str(argument_line: &str) -> Result<Record, &str> {
+        let args: Vec<&str> = argument_line.split(' ').collect();
+        if  args.len() < 2 {
+            return Err("error: header file provided lacks sufficient record arguments.");
+        }
+        let name = args[0].to_string();
+        let seg_num: Option<u64>;
+        let sig_count: usize;
+        let sampling_freq: u64;
+        // Everything else is initialized with None, this is for sake of me not getting a stroke
+        let mut counter_freq: Option<u64> = None;
+        let mut base_counter_val: Option<u64> = None;
+        let mut sample_num: Option<u64> = None;
+        let mut basetime: Option<String> = None;
+        let mut basedate: Option<String> = None;
+
+
+        // Signals and segments are kept in a single argument organized as signal/segment
+        // segments are optional and we need to check this too
+        {
+            let seg_sig: Vec<&str> = args[1].split('/').collect();
+            if seg_sig.len() == 2 {
+                match seg_sig[1].parse::<u64>() {
+                    Ok(value) => {
+                        seg_num = Some(value);
+                    }
+                    Err(_) => {
+                        return Err("error: failed to parse segment number from header.");
+                    }
+                }
+            } else {
+                seg_num = None;
+            }
+            match seg_sig[0].parse::<usize>() {
+                Ok(value) => {
+                    sig_count = value;
+                }
+                Err(_) => {
+                    return Err("error: failed to parse signal number from header.");
+                }
+            }
+        }
+        
+        // Parse everything else, if present
+        loop { // This is apparently the way to get safe correct goto-like behaviour
+            if args.len() <= 2 { // Sampling frequency and counter frequency
+                sampling_freq = 250; // Default value
+                break;
+            }
+            
+            {
+                let freq_dual: Vec<&str> = args[2].split('/').collect();
+                if freq_dual.len() == 2 {
+                    match freq_dual[1].parse::<u64>() {
+                        Ok(value) => {
+                            counter_freq = Some(value);
+                        }
+                        Err(_) => {
+                            return Err("error: failed to parse counter frequency from header.");
+                        }
+                    }
+                }
+                match freq_dual[0].parse::<u64>() {
+                    Ok(value) => {
+                        sampling_freq = value;
+                    }
+                    Err(_) => {
+                        return Err("error: failed to parse sampling frequency from header.");
+                    }
+                }
+            }
+
+            if args.len() <= 3 {break;}
+
+            match args[3].parse::<u64>() {
+                Ok(value) => {
+                    base_counter_val = Some(value);
+                }
+                Err(_) => {}
+            }
+            
+            if args.len() <= 4 {break;}
+
+            match args[4].parse::<u64>() {
+                Ok(value) => {
+                    sample_num = Some(value);
+                }
+                Err(_) => {}
+            }
+
+            if args.len() <= 5 {break;}
+
+            basetime = Some(args[5].to_string());
+
+            if args.len() <= 6 {break;}
+
+            basedate = Some(args[6].to_string());
+
+            break;
+        }
+        Ok(Record { 
+            name: name, 
+            seg_num: seg_num,
+            signal_count: sig_count, 
+            sampling_freq: sampling_freq, 
+            counter_freq: counter_freq, 
+            base_counter_val: base_counter_val, 
+            sample_num: sample_num, 
+            basetime: basetime, 
+            basedate: basedate 
+        })
+    }
+}
+
+/// Keys for all the data that can possibly be parsed from the values appended 
+/// to the ADC block
+#[derive(Debug, Eq, Hash, PartialEq)]
+enum AdcBlockKeys {
+    Gain,
+    Baseline,
+    Units
+}
+
+/// Holds the information relevant to individual signals and their specifications
+/// 
+/// By minimum, only carries the `filename`, `format`, and `adc_zero` attributes
+#[derive(Debug, Clone)]
+pub struct SignalSpec {
+    filename: String,
+    pub format: SignalFormat,
+    samples_frame: Option<u64>,
+    skew: Option<u64>,
+    offset: Option<u64>,
+    adc_gain: Option<f64>,
+    baseline: Option<i64>,
+    units: Option<String>,
+    adc_resolution: Option<u64>,
+    adc_zero: i64,
+    initial_val: Option<i64>,
+    checksum: Option<i64>,
+    blocksize: Option<u64>,
+    desc: Option<String>
+}
+
+impl SignalSpec {
+    /// Attempts to generate a valid signal specification struct from a supplied
+    /// string.
+    /// Returns a `Result<SignalSpec, &str>` containing possible error information
+    /// relevant to parsing failures.
+    /// 
+    /// ## Arguments
+    /// - `argument_line` - argument line for the signal specification, taken 
+    /// from header file
+    pub fn from_str(argument_line: &str) -> Result<SignalSpec, &str> {
+        let args: Vec<&str> = argument_line.split(' ').collect();
+
+        if  args.len() < 2 {
+            return Err("error: signal provided by header file lacks sufficient arguments.");
+        }
+        let name = args[0].to_string();
+        let sigformat: SignalFormat;
+        // Optional args
+        let samples_frame: Option<u64> = None;
+        let skew: Option<u64> = None;
+        let offset: Option<u64> = None;
+        let mut adc_gain: Option<f64> = None;
+        let mut baseline: Option<i64> = None;
+        let mut units: Option<String> = None;
+        let mut adc_resolution: Option<u64> = None;
+        let mut adc_zero:i64 = 0;
+        let mut initial_val: Option<i64> = None;
+        let mut checksum: Option<i64> = None;
+        let mut blocksize: Option<u64> = None;
+        let mut desc: Option<String> = None;
+
+        // TODO: implement samplesperframe, skew, and offset
+        match args[1].parse::<u64>() {
+            Ok(value) => {
+                sigformat = SignalSpec::parse_format(value);
+            },
+            Err(_) => {
+                return Err("error: unable to parse format from signal");
+            } 
+        }
+
+        loop {
+            if args.len() <= 2 {
+                break;
+            }
+
+            {
+                let mut results: HashMap<AdcBlockKeys, String> = HashMap::new();
+                let mut buffer: String = String::new();
+                let mut found_baseline = false;
+                let mut found_units = false;
+
+                // Parse through all the characters and get the individual parts
+                for character in args[2].chars() {
+                    loop {
+                        if !found_baseline && !found_units && character == '(' {
+                            found_baseline = true;
+                            results.insert(AdcBlockKeys::Gain, buffer);
+                            buffer = String::new();
+                            break;
+                        }
+                        else if found_baseline && character == ')' {
+                            results.insert(AdcBlockKeys::Baseline, buffer);
+                            buffer = String::new();
+                            break;
+                        }
+                        else if !found_units && character == '/' {
+                            found_units = true;
+                            if !found_baseline {
+                                results.insert(AdcBlockKeys::Gain, buffer);
+                            }
+                            buffer = String::new();
+                            break;
+                        }
+                        buffer.push(character);
+                        break;
+                    }
+                }
+                if found_units {
+                    results.insert(AdcBlockKeys::Units, buffer);
+                }
+                
+                // Try and get the actual values out of the results
+                match results.get(&AdcBlockKeys::Gain) {
+                    Some(value ) => {
+                        match value.parse::<f64>() {
+                            Ok(parsed) => {
+                                adc_gain = Some(parsed);
+                            },
+                            Err(_) => {
+                                return Err("error: Signal contains adc gain block with invalid entry");
+                            }
+                        }
+                    },
+                    None => {
+                        return Err("error: Signal contains adc gain block but the block is empty.");
+                    }
+                }
+
+                match results.get(&AdcBlockKeys::Baseline) {
+                    Some(value) => {
+                        match value.parse::<i64>() {
+                            Ok(parsed) => {
+                                baseline = Some(parsed);
+                            },
+                            Err(_) => {
+                                return Err("error: Unable to parse baseline from signal");
+                            }
+                        }
+                    },
+                    None => {}
+                }
+                match results.get(&AdcBlockKeys::Units) {
+                    Some(value) => {
+                        units = Some(value.clone());
+                    },
+                    None => {}
+                }
+            }
+            
+            if args.len() <= 3 {
+                break;
+            }
+            match args[3].parse::<u64>() {
+                Ok(value) => {
+                    adc_resolution = Some(value);
+                },
+                Err(_) => {
+                    return Err("error: unable to parse resolution from signal");
+                } 
+            }
+
+            if args.len() <= 4 {
+                break;
+            }
+            match args[4].parse::<i64>() {
+                Ok(value) => {
+                    adc_zero = value;
+                },
+                Err(_) => {
+                    adc_zero = 0; // Standard: If this field is missing, a value of zero is assumed.
+                } 
+            }
+
+            if args.len() <= 5 {
+                break;
+            }
+            match args[5].parse::<i64>() {
+                Ok(value) => {
+                    initial_val = Some(value);
+                },
+                // Standard: If this field is missing, 
+                // the adc_zero value is presumed.
+                // We do this on the data parsing part.
+                Err(_) => {} 
+            }
+
+            if args.len() <= 6 {
+                break;
+            }
+            match args[6].parse::<i64>() {
+                Ok(value) => {
+                   checksum = Some(value);
+                },
+                Err(_) => {
+                    return Err("error: unable to parse checksum from signal");
+                } 
+            }
+
+            if args.len() <= 7 {
+                break;
+            }
+            match args[7].parse::<u64>() {
+                Ok(value) => {
+                    blocksize = Some(value);
+                },
+                Err(_) => {
+                    return Err("error: unable to parse block size from signal");
+                } 
+            }
+
+            if args.len() <= 8 {
+                break;
+            }
+            desc = Some(args[8].to_string());
+
+            break;
+        }
+        
+        Ok(
+            SignalSpec { filename: name,
+                format: sigformat, 
+                samples_frame: samples_frame, 
+                skew: skew, offset: offset, 
+                adc_gain: adc_gain, 
+                baseline: baseline, 
+                units: units, 
+                adc_resolution: adc_resolution, 
+                adc_zero: adc_zero, 
+                initial_val: initial_val, 
+                checksum: checksum, 
+                blocksize: blocksize, 
+                desc: desc }
+        )
+    }
+
+    /// Parses the WFDB signal format from the format number, if it exists and
+    /// is implemented
+    fn parse_format(formatnum: u64) -> SignalFormat {
+        match formatnum {
+            16 => SignalFormat::Format16,
+            212 => SignalFormat::Format212,
+            0..=u64::MAX => SignalFormat::Unimpl
+        }
+    }
+}
+
+/// Holds all the possible data from the WFDB header file, those being the record
+/// line data and all the possible signal specifications present on the following
+/// lines
+#[derive(Debug, Clone)]
+pub struct Header {
+    record: Option<Record>,
+    signal_specs: Vec<SignalSpec>
+}
+
+impl Header {
+    /// Creates a completely empty, not fully initialized header
+    /// 
+    /// This is a workaround because I couldn't figure out how to make the compiler
+    /// accept non-initializing the value for the first pass of a for loop in 
+    /// `parse_header()`
+    fn new() -> Header {
+        Header { record: None, signal_specs: vec![] }
+    }
+
+    /// Creates a Header from a supplied Record struct. 
+    /// 
+    /// Initializes the signal_specs vector with the correct capacity provided
+    /// by the record.
+    fn with_record(record: Record) -> Header {
+        let capacity = record.signal_count;
+        Header { record: Some(record), signal_specs: Vec::with_capacity(capacity) }
+    }
+
+    fn add_signal_spec(&mut self, spec: SignalSpec) {
+        self.signal_specs.push(spec);
+    }
+
+    pub fn is_empty(&self) -> bool {
+        match self.record {
+            Some(_) => false,
+            None => true
+        }
+    }
+
+    pub fn signals_per_file(&self) -> HashMap<String, u64> {
+        let mut map: HashMap<String, u64> = HashMap::new();
+        for signal in &self.signal_specs {
+            map.entry(signal.filename.clone()).and_modify(|val: &mut u64| *val += 1 ).or_insert(1);
+        }
+        map
+    }
+
+    pub fn signals_in_file(&self, filename: String) -> Vec<&SignalSpec> {
+        let mut buffer: Vec<&SignalSpec> = Vec::new();
+        for signal in &self.signal_specs {
+            if signal.filename == filename {
+                buffer.push(signal);
+            }
+        }
+        buffer
+    }
+}
+
+/// Attempts to parse the header file. 
+/// 
+/// Returns a Result either containing the `Header` struct, or the
+/// error string describing why we couldn't parse the header.
+pub fn parse_header(header_data: &str) -> Result<Header, &str> {
+    let header_lines: Vec<&str> = header_data.split("\n").collect();
+    let mut found_record: bool = false;
+    let mut header: Header = Header::new();
+    let mut specs_read: usize = 0;
+    let mut specs_max: usize = 0;
+    for line in header_lines {
+        // Ignore commented lines
+        if line.starts_with("#") {
+            continue;
+        }
+
+        if !found_record {
+            let possible_record = Record::from_str(line);
+            match possible_record {
+                Ok(rec) => {
+                    specs_max = rec.signal_count;
+                    header = Header::with_record(rec);
+                    found_record = true;
+                    continue;
+                }
+                Err(e) => {
+                    return Err(e)
+                }
+            }
+        }
+
+        let possible_spec = SignalSpec::from_str(line);
+        match possible_spec {
+            Ok(spec) => {
+               header.add_signal_spec(spec);
+            }
+            Err(e) => {
+                return Err(e);
+            }
+        }
+
+        specs_read += 1;
+        if specs_read >= specs_max {
+            break;
+        }
+    }
+
+    if header.is_empty() {
+        return Err("Unable to parse valid header information");
+    }
+    Ok(header)
+}

src/main.rs

@@ -1,3 +1,154 @@
-fn main() {
+use std::{env::{self}, os::linux::raw, path::Path, str::FromStr};
-    println!("Hello, world!");
+use std::fs;
+
+pub mod headparse; // The HEAder parsing
+pub mod signal_data;
+
+use crate::{headparse::{Header, SignalSpec}, signal_data::SignalData};
+
+/// Use for handling possible formats of the WFDB data
+#[derive(Debug, Clone, Copy, PartialEq)]
+enum SignalFormat {
+    Format16 = 16,
+    Format212 = 212,
+    Invalid = -1,
+    Unimpl = 0
 }
+
+fn main() -> Result<(), String>{
+    let args: Vec<String> = env::args().collect();
+
+    if args.len() <= 1 || args.contains(&"help".to_string()) {
+        help();
+        return Ok(());
+    }
+
+    let filepath = Path::new(&args[1]);
+    if !filepath.is_file() {
+        println!("Path argument provided is not a valid file");
+        return Ok(());
+    }
+    if filepath.extension().unwrap() != "hea" {
+        println!("File provided is not a .hea file");
+        return Ok(());
+    }
+
+    // Parse the header information
+    let header: Header;
+    {
+    let hea_file_result = fs::read_to_string(filepath);
+        match hea_file_result {
+            Ok(file_data) => {
+                match headparse::parse_header(file_data.as_str()) {
+                    Ok(h) => {header = h}
+                    Err(e) => {return Err(e.to_string())}
+                }
+            }
+            Err(e) => {return Err("error: Provided file cannot be opened".to_string())}
+        }
+    }
+    
+    let path_parent: String;
+    {
+        // This is cursed lol
+        match filepath.parent() {
+            Some(p) => {
+                match p.to_str() {
+                    Some(str) => {
+                        path_parent = str.to_string();
+                    }
+                    None => {path_parent = "".to_string()}
+                }
+                
+            }
+            None => {path_parent = "".to_string()}
+        }
+    }
+
+    let parsed_signal_result = get_all_data(header, path_parent);
+
+    println!("Hello, world!");
+    Ok(())
+}
+
+fn help() {
+    println!("Conversion of WFDB files to a more human readable format. By default to a CSV.");
+    println!("\nUse in the format \"wfdb_corrosion (.hea filename)\"")
+}
+
+fn get_signal_data(spec: &SignalSpec, offset: u64, data_length: usize, raw: &Vec<u64>) -> SignalData {
+    todo!()
+}
+
+fn get_all_data(header: Header, header_root: String) -> Result<Vec<SignalData>, String> {
+    let signal_counts = header.signals_per_file();
+    for (fname, sigcount) in signal_counts {
+        let mut full = header_root.to_owned(); // Consider refactoring these to avoid this massive amount of ownership bs
+        full.push('/');
+        let fname = fname.to_owned();
+
+        full.push_str(&fname);
+        let filepath = Path::new(&full);
+        if !filepath.is_file() {
+            return Err("Error: Could not open signal .dat file. Check if the
+            header points to the right .dat file and if files are present".to_string());
+        }
+        if filepath.extension().unwrap() != "dat" {
+            println!("warn: Header file defines signal files with an extension 
+            other than .dat. 
+            Check if the header points to the right files, reading will continue
+            but correct data is not guaranteed");
+        }
+
+        // Time to actually do the work
+        let signals = header.signals_in_file(fname);
+        let format = file_signal_format_check(signals);
+        if format == SignalFormat::Invalid {
+            return Err("Error: Signals in single file use multiple formats.
+            Is your header file correctly formatted?".to_string());
+        }
+        if format == SignalFormat::Unimpl {
+            println!("Warn: Signals in single file use an unimplemented format,
+            and will be ignored");
+            continue;
+        }
+
+        let raw_bytes = fs::read(filepath).unwrap();
+
+        match format {
+            SignalFormat::Format16 => {
+                let mut buffer: u64 = 0;
+                let mut do_append = false;
+                let mut raw_converted: Vec<u64> = Vec::with_capacity(raw_bytes.len() / 2);
+
+                for byte in raw_bytes {
+                    if do_append {
+                        buffer = (buffer & 0xFF) + (byte as u64) << 8;
+                        raw_converted.push(buffer);
+                        do_append = false;
+                    } else {
+                        buffer = byte as u64;
+                        do_append = true;
+                    }
+                }
+                dbg!(raw_converted);
+            },
+            _ => {
+                panic!("Somehow reached actual file and format processing with
+                unimplemented format. Please make a bug report")
+            }
+        }
+    }
+    todo!()
+}
+
+fn file_signal_format_check(signals: Vec<&SignalSpec>) -> SignalFormat {
+    let first_sig = signals[0];
+    let format = first_sig.format;
+    for sig in signals {
+        if sig.format != format {
+            return SignalFormat::Invalid;
+        }
+    }
+    return format;
+}

src/signal_data.rs

@@ -0,0 +1,53 @@
+/// Holds fully processed signal data already parsed from the .dat file, ready
+/// to be outputted in a practical way
+#[derive(Debug, Clone)]
+pub struct SignalData {
+    name: String,
+    values: Vec<f64>,
+    unit: String
+}
+
+impl SignalData {
+    /// Creates a [`SignalData`](crate::SignalData) with a name and an empty
+    /// `values` vector and `unit` string.
+    /// 
+    /// The `values` vector is initialized with no capacity, which causes this
+    /// struct to be quite a bit slower. This method is thus not suggested, use 
+    /// [`SignalData::with_name_and_capacity`](crate::SignalData::with_name_and_capacity)
+    /// when you know how much you want to allocate
+    /// 
+    /// # Arguments
+    /// * `name` - The name of the Signal
+    pub fn with_name(name: String) -> SignalData {
+        SignalData { name: name, values: vec![], unit: String::new() }
+    }
+
+    /// Creates a [`SignalData`](crate::SignalData) with a name and a
+    /// `values` vector of a certain capacity and an empty `unit` string.
+    /// 
+    /// # Arguments
+    /// 
+    /// * `name` - The name of the Signal
+    /// * `capacity` - The capacity with which the `values` vector will be 
+    /// initialized
+    pub fn with_name_and_capacity(name: String, capacity: usize) -> SignalData {
+        SignalData { name: name, values: Vec::with_capacity(capacity), unit: String::new() }
+    }
+
+    /// Appends a value to `values` inside the struct
+    pub fn push(&mut self, value: f64) {
+        self.values.push(value);
+    }
+
+    /// Appends data copied from the supplied `data` argument to the end of the
+    /// `values` vector inside the struct
+    pub fn append_copy(&mut self, data: Vec<f64>) {
+        let mut to_add: Vec<f64> = data;
+        to_add.append(&mut self.values);
+    }
+}
+
+pub struct SignalDataRaw {
+    name: String,
+    raw_data: Vec<u8>
+}