1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
/// Lift over coordinates using the CIGAR string
pub mod bamlift;
/// Add and remove base modifications from a bam record
pub mod basemods;
/// Center fiberseq information around a reference position
pub mod center;
/// Command line interface for fibertools-rs.
pub mod cli;
#[cfg(feature = "cnn")]
/// m6A prediction with a CNN
pub mod cnn;
/// Extract fiberseq data into plain text formats
pub mod extract;
/// m6A prediction
pub mod predict_m6a;
/// m6A prediction with a XGBoost
pub mod xgb;

use anyhow::Result;
use itertools::Itertools;
use lazy_static::lazy_static;
use rust_htslib::{bam, bam::Read};
use std::collections::HashMap;
use std::env;
use std::fs::File;
use std::io::{self, stdout, BufWriter, Write};
use std::path::PathBuf;
use std::process::exit;

const VERSION: &str = env!("CARGO_PKG_VERSION");
const BUFFER_SIZE: usize = 32 * 1024;
const PROGRESS_STYLE: &str =
    "[{elapsed_precise:.yellow}] {bar:50.cyan/blue} {human_pos:>5.cyan}/{human_len:.blue} {percent:>3.green}% {per_sec:<10.cyan}";

/// Get a buffered output writer from stdout or a file
fn get_output(path: Option<PathBuf>) -> Result<Box<dyn Write + Send + 'static>> {
    let writer: Box<dyn Write + Send + 'static> = match path {
        Some(path) => {
            if path.as_os_str() == "-" {
                Box::new(BufWriter::with_capacity(BUFFER_SIZE, stdout()))
            } else {
                Box::new(BufWriter::with_capacity(BUFFER_SIZE, File::create(path)?))
            }
        }
        None => Box::new(BufWriter::with_capacity(BUFFER_SIZE, stdout())),
    };
    Ok(writer)
}

/// unzip a vector of tuples
pub fn unzip_to_vectors<T, U>(vec: Vec<(T, U)>) -> (Vec<T>, Vec<U>) {
    vec.into_iter().unzip()
}

/// join a vector with commas
pub fn join_by_str<'a, I, Z>(vals: I, sep: &str) -> String
where
    I: IntoIterator<Item = Z>,
    Z: ToString + 'a,
{
    vals.into_iter().map(|v| v.to_string() + sep).collect()
}

/// Write to stdout if - or the file specified by a path
pub fn writer(filename: &str) -> Result<Box<dyn Write>> {
    //let ext = Path::new(filename).extension();
    let path = PathBuf::from(filename);
    let buffer = get_output(Some(path))?; //.expect("Error: cannot create output file");
    Ok(buffer)
}

/// Open bam file
pub fn bam_reader(bam: &str, threads: usize) -> bam::Reader {
    let mut bam = if bam == "-" {
        bam::Reader::from_stdin().unwrap_or_else(|_| panic!("Failed to open bam from stdin"))
    } else {
        bam::Reader::from_path(bam).unwrap_or_else(|_| panic!("Failed to open {}", bam))
    };
    bam.set_threads(threads).unwrap();
    bam
}

/// Write to a bam file.
pub fn bam_writer(out: &str, template_bam: &bam::Reader, threads: usize) -> bam::Writer {
    let mut header = bam::Header::from_template(template_bam.header());

    // add to the header
    let header_string = String::from_utf8_lossy(&header.to_bytes()).to_string();
    let ft_count = header_string.matches("PP:ft").count();
    let mut header_rec = bam::header::HeaderRecord::new(b"PG");
    header_rec.push_tag(b"ID", &format!("ft.{}", ft_count + 1));
    header_rec.push_tag(b"PP", &format!("ft.{}", ft_count));
    header_rec.push_tag(b"PN", &"fibertools-rs");
    header_rec.push_tag(b"VN", &VERSION);
    let cli = env::args().join(" ");
    header_rec.push_tag(b"CL", &cli);
    header.push_record(&header_rec);
    log::trace!("{:?}", String::from_utf8_lossy(&header.to_bytes()));

    // make the writer
    let mut out = if out == "-" {
        bam::Writer::from_stdout(&header, bam::Format::Bam).unwrap()
    } else {
        bam::Writer::from_path(out, &header, bam::Format::Bam).unwrap()
    };
    out.set_threads(threads).unwrap();
    out
}

pub struct FiberOut {
    pub m6a: Option<Box<dyn Write>>,
    pub cpg: Option<Box<dyn Write>>,
    pub msp: Option<Box<dyn Write>>,
    pub nuc: Option<Box<dyn Write>>,
    pub all: Option<Box<dyn Write>>,
    pub reference: bool,
    pub simplify: bool,
    pub quality: bool,
    pub min_ml_score: u8,
    pub full_float: bool,
}

impl FiberOut {
    #[allow(clippy::too_many_arguments)]
    pub fn new(
        m6a: &Option<String>,
        cpg: &Option<String>,
        msp: &Option<String>,
        nuc: &Option<String>,
        all: &Option<String>,
        reference: bool,
        simplify: bool,
        quality: bool,
        min_ml_score: u8,
        full_float: bool,
    ) -> Result<Self> {
        let m6a = match m6a {
            Some(m6a) => Some(writer(m6a)?),
            None => None,
        };
        let cpg = match cpg {
            Some(cpg) => Some(writer(cpg)?),
            None => None,
        };
        let msp = match msp {
            Some(msp) => Some(writer(msp)?),
            None => None,
        };
        let nuc = match nuc {
            Some(nuc) => Some(writer(nuc)?),
            None => None,
        };
        let all = match all {
            Some(all) => Some(writer(all)?),
            None => None,
        };
        // set to zero
        let mut min_ml_score = min_ml_score;
        if full_float {
            min_ml_score = 0;
        }

        Ok(FiberOut {
            m6a,
            cpg,
            msp,
            nuc,
            all,
            reference,
            simplify,
            quality,
            min_ml_score,
            full_float,
        })
    }
}

/// write to a file, but don't error on broken pipes
pub fn write_to_file(out: &str, file: &mut Box<dyn Write>) {
    let out = write!(file, "{}", out);
    if let Err(err) = out {
        if err.kind() == io::ErrorKind::BrokenPipe {
            exit(0);
        } else {
            panic!("Error: {}", err);
        }
    }
}

pub fn write_to_stdout(out: &str) {
    let mut out_f = Box::new(std::io::stdout()) as Box<dyn Write>;
    write_to_file(out, &mut out_f);
}

// This is a bam chunk reader
use colored::Colorize;
use std::time::Instant;

struct BamChunk<'a> {
    bam: bam::Records<'a, bam::Reader>,
    chunk_size: usize,
}

// The `Iterator` trait only requires a method to be defined for the `next` element.
impl<'a> Iterator for BamChunk<'a> {
    // We can refer to this type using Self::Item
    type Item = Vec<bam::Record>;

    // The return type is `Option<T>`:
    //     * When the `Iterator` is finished, `None` is returned.
    //     * Otherwise, the next value is wrapped in `Some` and returned.
    // We use Self::Item in the return type, so we can change
    // the type without having to update the function signatures.
    fn next(&mut self) -> Option<Self::Item> {
        let start = Instant::now();
        let mut cur_vec = vec![];
        for r in self.bam.by_ref().take(self.chunk_size) {
            cur_vec.push(r.unwrap())
        }
        // return
        if cur_vec.is_empty() {
            None
        } else {
            let duration = start.elapsed().as_secs_f64();
            log::info!(
                "Read {} bam records at {}.",
                format!("{:}", cur_vec.len()).bright_magenta().bold(),
                format!("{:.2?} reads/s", cur_vec.len() as f64 / duration)
                    .bright_cyan()
                    .bold(),
            );
            Some(cur_vec)
        }
    }
}

#[derive(Clone)]
pub enum PbChem {
    Two,
    TwoPointTwo,
}

pub fn find_pb_polymerase(header: &bam::Header) -> PbChem {
    lazy_static! {
        static ref CHEMISTRY_MAP: HashMap<String, PbChem> = HashMap::from([
            // regular 2.0
            ("101-789-500".to_string(), PbChem::Two),
            // this is actually 2.1 we didn't have training data for it
            ("101-820-500".to_string(), PbChem::Two),
            // regular 2.2
            ("101-894-200".to_string(), PbChem::TwoPointTwo),
            // is really 3.1 but has polymerase of 2.1, and we need to make that 2.0
            ("102-194-200".to_string(), PbChem::Two),
            // is really 3.2 but has polymerase of 2.2
            ("102-194-100".to_string(), PbChem::TwoPointTwo),
            // Revio has kinetics most similar to 2.2
            ("102-739-100".to_string(), PbChem::TwoPointTwo)
        ]);
    }
    lazy_static! {
        static ref MM_DS: regex::Regex =
            regex::Regex::new(r".*READTYPE=([^;]+);.*BINDINGKIT=([^;]+);").unwrap();
    }
    let z = header.to_hashmap();
    let rg = z.get("RG").expect("RG tag missing from bam file");
    let mut read_type = "";
    let mut binding_kit = "";
    for tag in rg {
        for (tag, val) in tag {
            if tag == "DS" {
                for cap in MM_DS.captures_iter(val) {
                    read_type = cap.get(1).map_or("", |m| m.as_str());
                    binding_kit = cap.get(2).map_or("", |m| m.as_str());
                }
            }
        }
    }
    // make sure read-type is CCS
    assert_eq!(read_type, "CCS");
    // grab chemistry
    let chemistry = CHEMISTRY_MAP.get(binding_kit).unwrap_or_else(|| {
        log::warn!(
            "Polymerase for BINDINGKIT={} not found. Defaulting to ML model made with 2.2",
            binding_kit
        );
        &PbChem::TwoPointTwo
    });
    chemistry.clone()
}