bicycle1885

using Bio.Seq


# Utils for Benchmark
# -------------------

function random_seq(n, nts, probs)
    cumprobs = cumsum(probs)
    x = Array(Char, n)
    for i in 1:n

bicycle1885

using Bio.Seq

let
    str = "ACGT"^10000
    seq = DNASequence(str)
    seq′ = DNASequence("AAA" * str)[4:end]
    @assert hash(seq) === hash(seq′)

    println("ASCIIString")
    hash(str)

bicycle1885

# Toy Regular Expression Engine Written in Julia
# ==============================================

# Example
# -------
#
# julia> include("re.jl")
# run (generic function with 1 method)
# 
# julia> ast = parse("foo*|ba*r")

bicycle1885

type PhyNode{I,T<:Tuple}
    id::I
    metadata::T
    children::Vector{PhyNode{I,T}}
    parent::PhyNode{I,T}

    # root
    function PhyNode(id, metadata, children)
        node = new(id, metadata, children)
        node.parent = node

bicycle1885

bicycle1885

bicycle1885

# Nested Containment List
# =======================
#
# Alekseyenko, A. V., & Lee, C. J. (2007). "Nested Containment List (NCList): A
# new algorithm for accelerating interval query of genome alignment and interval
# databases." Bioinformatics, 23(11), 1386–1393.
# doi:10.1093/bioinformatics/btl647

# An interval type (inclusive).
immutable Interval{T}

bicycle1885

bicycle1885

using Automa
using Automa.RegExp
const re = Automa.RegExp

length = re"[0-9]+\.[0-9]+"
name = re.rep1(re"[!-~]" \ re"[,:;()[\]]")
spaces = re" +"

const newick = compile(
    re"\(" => :(emit(:lparen)),

bicycle1885

import Automa
import Automa.RegExp: @re_str
const re = Automa.RegExp

# VCF v4.2
const vcf_machine = (function ()
    delim(x, sep) = re.opt(re.cat(x, re.rep(re.cat(sep, x))))

    nl = re"\n"
    newline = re"\r?" * nl