Flux.jl/src/data/cmudict.jl

50 lines
1.2 KiB
Julia
Raw Normal View History

2017-11-01 16:01:34 +00:00
module CMUDict
export cmudict
using ..Data: deps
const version = "0.7b"
2018-03-06 16:29:44 +00:00
const cache_prefix = "https://cache.julialang.org"
2017-11-01 16:01:34 +00:00
function load()
2018-03-06 16:29:44 +00:00
suffixes = ["", ".phones", ".symbols"]
if isdir(deps("cmudict"))
2018-04-15 19:09:08 +00:00
if all(isfile(deps("cmudict", "cmudict$x")) for x in suffixes)
2018-03-06 16:29:44 +00:00
return
end
end
2018-08-11 13:50:11 +00:00
@info "Downloading CMUDict dataset"
2017-11-01 16:01:34 +00:00
mkpath(deps("cmudict"))
2018-03-06 16:29:44 +00:00
for x in suffixes
2018-04-15 19:09:08 +00:00
download("$cache_prefix/http://svn.code.sf.net/p/cmusphinx/code/trunk/cmudict/cmudict-$version$x",
deps("cmudict", "cmudict$x"))
2017-11-01 16:01:34 +00:00
end
end
function phones()
load()
2018-07-18 07:01:06 +00:00
Symbol.(first.(split.(split(read(deps("cmudict", "cmudict.phones"),String),
2018-07-18 13:39:20 +00:00
"\n", keepempty = false), "\t")))
2017-11-01 16:01:34 +00:00
end
function symbols()
load()
2018-07-18 07:01:06 +00:00
Symbol.(split(read(deps("cmudict", "cmudict.symbols"),String),
2018-07-18 13:39:20 +00:00
"\n", keepempty = false))
2017-11-01 16:01:34 +00:00
end
function rawdict()
load()
Dict(String(xs[1]) => Symbol.(xs[2:end]) for xs in
2018-07-18 07:01:06 +00:00
filter(!isempty, split.(split(read(deps("cmudict", "cmudict"),String), "\n"))))
2017-11-01 16:01:34 +00:00
end
2018-07-18 07:01:06 +00:00
validword(s) = isascii(s) && occursin(r"^[\w\-\.]+$", s)
2017-11-01 16:01:34 +00:00
2018-07-18 13:39:20 +00:00
cmudict() = filter(p -> validword(p.first), rawdict())
2017-11-01 16:01:34 +00:00
alphabet() = ['A':'Z'..., '0':'9'..., '_', '-', '.']
end