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(slicing_syntax)]
18 extern crate collections;
20 use std::ascii::{AsciiExt, OwnedAsciiExt};
21 use std::cmp::Ordering::{mod, Less, Greater, Equal};
22 use std::collections::HashMap;
23 use std::sync::mpsc::{channel, Sender, Receiver};
24 use std::mem::replace;
28 use std::string::IntoString;
29 use std::thread::Thread;
31 fn f64_cmp(x: f64, y: f64) -> Ordering {
32 // arbitrarily decide that NaNs are larger than everything.
35 } else if x.is_nan() {
46 // given a map, print a sorted version of it
47 fn sort_and_fmt(mm: &HashMap<Vec<u8> , uint>, total: uint) -> String {
48 fn pct(xx: uint, yy: uint) -> f64 {
49 return (xx as f64) * 100.0 / (yy as f64);
52 // sort by key, then by value
53 fn sortKV(mut orig: Vec<(Vec<u8> ,f64)> ) -> Vec<(Vec<u8> ,f64)> {
54 orig.sort_by(|&(ref a, _), &(ref b, _)| a.cmp(b));
55 orig.sort_by(|&(_, a), &(_, b)| f64_cmp(b, a));
59 let mut pairs = Vec::new();
62 for (key, &val) in mm.iter() {
63 pairs.push(((*key).clone(), pct(val, total)));
66 let pairs_sorted = sortKV(pairs);
68 let mut buffer = String::new();
69 for &(ref k, v) in pairs_sorted.iter() {
70 buffer.push_str(format!("{} {:0.3}\n",
71 k.to_ascii_uppercase(),
78 // given a map, search for the frequency of a pattern
79 fn find(mm: &HashMap<Vec<u8> , uint>, key: String) -> uint {
80 let key = key.into_ascii_lowercase();
81 match mm.get(key.as_bytes()) {
82 option::Option::None => { return 0u; }
83 option::Option::Some(&num) => { return num; }
87 // given a map, increment the counter for a key
88 fn update_freq(mm: &mut HashMap<Vec<u8> , uint>, key: &[u8]) {
89 let key = key.to_vec();
90 let newval = match mm.pop(&key) {
94 mm.insert(key, newval);
97 // given a Vec<u8>, for each window call a function
98 // i.e., for "hello" and windows of size four,
99 // run it("hell") and it("ello"), then return "llo"
100 fn windows_with_carry(bb: &[u8], nn: uint, it: |window: &[u8]|) -> Vec<u8> {
104 while ii < len - (nn - 1u) {
109 return bb[len - (nn - 1u)..len].to_vec();
112 fn make_sequence_processor(sz: uint,
113 from_parent: &Receiver<Vec<u8>>,
114 to_parent: &Sender<String>) {
115 let mut freqs: HashMap<Vec<u8>, uint> = HashMap::new();
116 let mut carry = Vec::new();
117 let mut total: uint = 0u;
119 let mut line: Vec<u8>;
123 line = from_parent.recv().unwrap();
124 if line == Vec::new() { break; }
126 carry.push_all(line.as_slice());
127 carry = windows_with_carry(carry.as_slice(), sz, |window| {
128 update_freq(&mut freqs, window);
133 let buffer = match sz {
134 1u => { sort_and_fmt(&freqs, total) }
135 2u => { sort_and_fmt(&freqs, total) }
136 3u => { format!("{}\t{}", find(&freqs, "GGT".to_string()), "GGT") }
137 4u => { format!("{}\t{}", find(&freqs, "GGTA".to_string()), "GGTA") }
138 6u => { format!("{}\t{}", find(&freqs, "GGTATT".to_string()), "GGTATT") }
139 12u => { format!("{}\t{}", find(&freqs, "GGTATTTTAATT".to_string()), "GGTATTTTAATT") }
140 18u => { format!("{}\t{}", find(&freqs, "GGTATTTTAATTTATAGT".to_string()),
141 "GGTATTTTAATTTATAGT") }
142 _ => { "".to_string() }
145 to_parent.send(buffer);
148 // given a FASTA file on stdin, process sequence THREE
150 use std::io::{stdio, MemReader, BufferedReader};
152 let rdr = if os::getenv("RUST_BENCH").is_some() {
153 let foo = include_bytes!("shootout-k-nucleotide.data");
154 box MemReader::new(foo.to_vec()) as Box<Reader>
156 box stdio::stdin() as Box<Reader>
158 let mut rdr = BufferedReader::new(rdr);
160 // initialize each sequence sorter
161 let sizes = vec!(1u,2,3,4,6,12,18);
162 let mut streams = Vec::from_fn(sizes.len(), |_| Some(channel::<String>()));
163 let mut from_child = Vec::new();
164 let to_child = sizes.iter().zip(streams.iter_mut()).map(|(sz, stream_ref)| {
166 let stream = replace(stream_ref, None);
167 let (to_parent_, from_child_) = stream.unwrap();
169 from_child.push(from_child_);
171 let (to_child, from_parent) = channel();
173 Thread::spawn(move|| {
174 make_sequence_processor(sz, &from_parent, &to_parent_);
178 }).collect::<Vec<Sender<Vec<u8> >> >();
181 // latch stores true after we've started
182 // reading the sequence of interest
183 let mut proc_mode = false;
185 for line in rdr.lines() {
186 let line = line.unwrap().as_slice().trim().to_string();
188 if line.len() == 0u { continue; }
190 match (line.as_bytes()[0] as char, proc_mode) {
192 // start processing if this is the one
194 match line.as_slice().slice_from(1).find_str("THREE") {
195 option::Option::Some(_) => { proc_mode = true; }
196 option::Option::None => { }
200 // break our processing
201 ('>', true) => { break; }
203 // process the sequence for k-mers
205 let line_bytes = line.as_bytes();
207 for (ii, _sz) in sizes.iter().enumerate() {
208 let lb = line_bytes.to_vec();
209 to_child[ii].send(lb);
219 for (ii, _sz) in sizes.iter().enumerate() {
220 to_child[ii].send(Vec::new());
223 // now fetch and print result messages
224 for (ii, _sz) in sizes.iter().enumerate() {
225 println!("{}", from_child[ii].recv().unwrap());