1 // Copyright 2012-2014 The Rust Project Developers. See the COPYRIGHT
2 // file at the top-level directory of this distribution and at
3 // http://rust-lang.org/COPYRIGHT.
5 // Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
6 // http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
7 // <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
8 // option. This file may not be copied, modified, or distributed
9 // except according to those terms.
11 // ignore-android: FIXME(#10393)
12 // ignore-pretty very bad with line comments
14 // multi tasking k-nucleotide
16 #![feature(box_syntax)]
18 use std::ascii::{AsciiExt, OwnedAsciiExt};
19 use std::cmp::Ordering::{self, Less, Greater, Equal};
20 use std::collections::HashMap;
21 use std::mem::replace;
25 use std::sync::mpsc::{channel, Sender, Receiver};
26 use std::thread::Thread;
28 fn f64_cmp(x: f64, y: f64) -> Ordering {
29 // arbitrarily decide that NaNs are larger than everything.
32 } else if x.is_nan() {
43 // given a map, print a sorted version of it
44 fn sort_and_fmt(mm: &HashMap<Vec<u8> , uint>, total: uint) -> String {
45 fn pct(xx: uint, yy: uint) -> f64 {
46 return (xx as f64) * 100.0 / (yy as f64);
49 // sort by key, then by value
50 fn sortKV(mut orig: Vec<(Vec<u8> ,f64)> ) -> Vec<(Vec<u8> ,f64)> {
51 orig.sort_by(|&(ref a, _), &(ref b, _)| a.cmp(b));
52 orig.sort_by(|&(_, a), &(_, b)| f64_cmp(b, a));
56 let mut pairs = Vec::new();
59 for (key, &val) in mm.iter() {
60 pairs.push(((*key).clone(), pct(val, total)));
63 let pairs_sorted = sortKV(pairs);
65 let mut buffer = String::new();
66 for &(ref k, v) in pairs_sorted.iter() {
67 buffer.push_str(format!("{:?} {:0.3}\n",
68 k.to_ascii_uppercase(),
75 // given a map, search for the frequency of a pattern
76 fn find(mm: &HashMap<Vec<u8> , uint>, key: String) -> uint {
77 let key = key.into_ascii_lowercase();
78 match mm.get(key.as_bytes()) {
79 option::Option::None => { return 0u; }
80 option::Option::Some(&num) => { return num; }
84 // given a map, increment the counter for a key
85 fn update_freq(mm: &mut HashMap<Vec<u8> , uint>, key: &[u8]) {
86 let key = key.to_vec();
87 let newval = match mm.remove(&key) {
91 mm.insert(key, newval);
94 // given a Vec<u8>, for each window call a function
95 // i.e., for "hello" and windows of size four,
96 // run it("hell") and it("ello"), then return "llo"
97 fn windows_with_carry<F>(bb: &[u8], nn: uint, mut it: F) -> Vec<u8> where
103 while ii < len - (nn - 1u) {
108 return bb[len - (nn - 1u)..len].to_vec();
111 fn make_sequence_processor(sz: uint,
112 from_parent: &Receiver<Vec<u8>>,
113 to_parent: &Sender<String>) {
114 let mut freqs: HashMap<Vec<u8>, uint> = HashMap::new();
115 let mut carry = Vec::new();
116 let mut total: uint = 0u;
118 let mut line: Vec<u8>;
122 line = from_parent.recv().unwrap();
123 if line == Vec::new() { break; }
125 carry.push_all(line.as_slice());
126 carry = windows_with_carry(carry.as_slice(), sz, |window| {
127 update_freq(&mut freqs, window);
132 let buffer = match sz {
133 1u => { sort_and_fmt(&freqs, total) }
134 2u => { sort_and_fmt(&freqs, total) }
135 3u => { format!("{}\t{}", find(&freqs, "GGT".to_string()), "GGT") }
136 4u => { format!("{}\t{}", find(&freqs, "GGTA".to_string()), "GGTA") }
137 6u => { format!("{}\t{}", find(&freqs, "GGTATT".to_string()), "GGTATT") }
138 12u => { format!("{}\t{}", find(&freqs, "GGTATTTTAATT".to_string()), "GGTATTTTAATT") }
139 18u => { format!("{}\t{}", find(&freqs, "GGTATTTTAATTTATAGT".to_string()),
140 "GGTATTTTAATTTATAGT") }
141 _ => { "".to_string() }
144 to_parent.send(buffer).unwrap();
147 // given a FASTA file on stdin, process sequence THREE
149 use std::io::{stdio, MemReader, BufferedReader};
151 let rdr = if os::getenv("RUST_BENCH").is_some() {
152 let foo = include_bytes!("shootout-k-nucleotide.data");
153 box MemReader::new(foo.to_vec()) as Box<Reader>
155 box stdio::stdin() as Box<Reader>
157 let mut rdr = BufferedReader::new(rdr);
159 // initialize each sequence sorter
160 let sizes = vec!(1u,2,3,4,6,12,18);
161 let mut streams = range(0, sizes.len()).map(|_| {
162 Some(channel::<String>())
163 }).collect::<Vec<_>>();
164 let mut from_child = Vec::new();
165 let to_child = sizes.iter().zip(streams.iter_mut()).map(|(sz, stream_ref)| {
167 let stream = replace(stream_ref, None);
168 let (to_parent_, from_child_) = stream.unwrap();
170 from_child.push(from_child_);
172 let (to_child, from_parent) = channel();
174 Thread::spawn(move|| {
175 make_sequence_processor(sz, &from_parent, &to_parent_);
179 }).collect::<Vec<Sender<Vec<u8> >> >();
182 // latch stores true after we've started
183 // reading the sequence of interest
184 let mut proc_mode = false;
186 for line in rdr.lines() {
187 let line = line.unwrap().as_slice().trim().to_string();
189 if line.len() == 0u { continue; }
191 match (line.as_bytes()[0] as char, proc_mode) {
193 // start processing if this is the one
195 match line.as_slice().slice_from(1).find_str("THREE") {
196 Some(_) => { proc_mode = true; }
201 // break our processing
202 ('>', true) => { break; }
204 // process the sequence for k-mers
206 let line_bytes = line.as_bytes();
208 for (ii, _sz) in sizes.iter().enumerate() {
209 let lb = line_bytes.to_vec();
210 to_child[ii].send(lb).unwrap();
220 for (ii, _sz) in sizes.iter().enumerate() {
221 to_child[ii].send(Vec::new()).unwrap();
224 // now fetch and print result messages
225 for (ii, _sz) in sizes.iter().enumerate() {
226 println!("{:?}", from_child[ii].recv().unwrap());