Improve performance (#38)

Add optimized `readbytes!` method which copies data by chunks instead of byte per byte.
Internally, avoid passing a `SubArray` to `readbytes!` as there is currently no optimized
method for them: instead, wrap the corresponding memory in an `Array`.
Increase the size of the buffer from 100 to 200 bytes, which appears to be a good tradeoff.
This makes loading a file about 10 times faster than before.

src/StringEncodings.jl

@@ -20,7 +20,7 @@ end
 using Base.Libc: errno, strerror, E2BIG, EINVAL, EILSEQ
 
 import Base: close, eachline, eof, flush, isreadable, iswritable,
-             open, readline, readlines, readuntil, show, write, read
+             open, read, readbytes!, readline, readlines, readuntil, show, write
 
 export StringEncoder, StringDecoder, encode, decode, encodings
 export StringEncodingError, OutputBufferError, IConvError
@@ -94,7 +94,7 @@ end
 
 ## StringEncoder and StringDecoder common functions
 
-const BUFSIZE = 100
+const BUFSIZE = 200
 
 mutable struct StringEncoder{F<:Encoding, T<:Encoding, S<:IO} <: IO
     stream::S
@@ -318,7 +318,13 @@ function fill_buffer!(s::StringDecoder)
         return i
     end
 
-    s.inbytesleft[] += readbytes!(s.stream, view(s.inbuf, Int(s.inbytesleft[]+1):BUFSIZE))
+    # readbytes! performance with SubArray was improved by JuliaLang/julia#36607
+    @static if VERSION >= v"1.6.0-DEV.438"
+        inbuf_view = view(s.inbuf, Int(s.inbytesleft[]+1):BUFSIZE)
+    else
+        inbuf_view = unsafe_wrap(Array, pointer(s.inbuf, s.inbytesleft[]+1), BUFSIZE)
+    end
+    s.inbytesleft[] += readbytes!(s.stream, inbuf_view)
     iconv!(s.cd, s.inbuf, s.outbuf, s.inbufptr, s.outbufptr, s.inbytesleft, s.outbytesleft)
 end
 
@@ -328,7 +334,7 @@ end
 #    data contains only state control sequences which may be converted to nothing)
 # 3) if not, reset iconv to initial state, which may generate data
 function eof(s::StringDecoder)
-    length(s.outbuf) - s.outbytesleft[] == s.skip &&
+    BUFSIZE - s.outbytesleft[] == s.skip &&
         fill_buffer!(s) == 0 &&
         iconv_reset!(s) == 0
 end
@@ -403,6 +409,28 @@ function open(fname::AbstractString, enc::Encoding, mode::AbstractString)
     wrap_stream(open(fname, mode), enc)
 end
 
+# optimized method adapted from Base but reading as many bytes
+# as the buffer contains on each iteration rather than a single one,
+# which increases performance dramatically
+function readbytes!(s::StringDecoder, b::AbstractArray{UInt8}, nb=length(b))
+    olb = lb = length(b)
+    nr = 0
+    while nr < nb && !eof(s)
+        nc = min(nb-nr, BUFSIZE - s.outbytesleft[])
+        if nr+nc > lb
+            lb = (nr+nc) * 2
+            resize!(b, lb)
+        end
+        copyto!(b, firstindex(b)+nr, s.outbuf, s.skip+1, nc)
+        s.skip += nc
+        nr += nc
+    end
+    if lb > olb
+        resize!(b, nr) # shrink to just contain input data if was resized
+    end
+    return nr
+end
+
 """
     read(stream::IO, [nb::Integer,] enc::Encoding)
     read(filename::AbstractString, [nb::Integer,] enc::Encoding)

test/runtests.jl

@@ -141,6 +141,8 @@ mktemp() do path, io
         write(io, s)
     end
 
+    nc = ncodeunits(first(s, 90))
+
     @test String(read(path, enc"ISO-2022-JP")) == s
     @test String(open(io->read(io, enc"ISO-2022-JP"), path)) == s
     @test String(open(io->read(io), path, enc"ISO-2022-JP")) == s
@@ -149,14 +151,48 @@ mktemp() do path, io
     @test String(open(io->read(io, 1000, enc"ISO-2022-JP"), path)) == s
     @test String(open(io->read(io, 1000), path, enc"ISO-2022-JP")) == s
 
-    @test String(read(path, 10, enc"ISO-2022-JP")) == first(s, 10)
-    @test String(open(io->read(io, 10, enc"ISO-2022-JP"), path)) == first(s, 10)
-    @test String(open(io->read(io, 10), path, enc"ISO-2022-JP")) == first(s, 10)
+    @test String(read(path, nc, enc"ISO-2022-JP")) == first(s, nc)
+    @test String(open(io->read(io, nc, enc"ISO-2022-JP"), path)) == first(s, nc)
+    @test String(open(io->read(io, nc), path, enc"ISO-2022-JP")) == first(s, nc)
 
     @test read(path, String, enc"ISO-2022-JP") == s
     @test open(io->read(io, String, enc"ISO-2022-JP"), path) == s
     @test open(io->read(io, String), path, enc"ISO-2022-JP") == s
 
+    b = zeros(UInt8, nc)
+    @test open(io->read!(io, b), path, enc"ISO-2022-JP") === b
+    @test String(b) == first(s, 90)
+
+    b = zeros(UInt8, nc)
+    @test open(io->readbytes!(io, b), path, enc"ISO-2022-JP") == ncodeunits(s)
+    @test String(b) == s
+
+    b = zeros(UInt8, 1000)
+    @test open(io->readbytes!(io, b), path, enc"ISO-2022-JP") == ncodeunits(s)
+    @test length(b) == 1000
+    @test String(b[1:ncodeunits(s)]) == s
+
+    b = UInt8[]
+    @test open(io->readbytes!(io, b), path, enc"ISO-2022-JP") == 0
+    @test length(b) == 0
+
+    b = zeros(UInt8, nc)
+    @test open(io->readbytes!(io, b, nc), path, enc"ISO-2022-JP") == nc
+    @test String(b) == first(s, 90)
+
+    b = zeros(UInt8, 1000)
+    @test open(io->readbytes!(io, b, nc), path, enc"ISO-2022-JP") == nc
+    @test length(b) == 1000
+    @test String(b[1:nc]) == first(s, 90)
+
+    b = UInt8[]
+    @test open(io->readbytes!(io, b, nc), path, enc"ISO-2022-JP") == nc
+    @test String(b) == first(s, 90)
+
+    b = UInt8[]
+    open(io->while !eof(io); push!(b, read(io, UInt8)) end, path, enc"ISO-2022-JP")
+    @test String(b) == s
+
     @test readuntil(path, enc"ISO-2022-JP", '\0') == "a string "
     @test open(io->readuntil(io, enc"ISO-2022-JP", '\0'), path) == "a string "
     @test open(io->readuntil(io, enc"ISO-2022-JP", '\0', keep=true), path) == "a string \0"