Initial commit

This commit is contained in:
Mikko Ahlroth 2015-02-14 20:57:07 +02:00
commit 234538f653
7 changed files with 363 additions and 0 deletions

5
.gitignore vendored Normal file
View file

@ -0,0 +1,5 @@
/ebin
/deps
erl_crash.dump
*.ez
_build/*

3
README.md Normal file
View file

@ -0,0 +1,3 @@
# Nurina
** TODO: Add description **

174
lib/nurina.ex Normal file
View file

@ -0,0 +1,174 @@
defmodule Nurina do
defmodule Info do
@moduledoc """
Contains a struct containing info about a given URL.
"""
defstruct scheme: nil,
hier: nil,
query: nil,
fragment: nil,
valid: true,
authority: nil,
path: nil,
host: nil,
port: nil,
userinfo: nil,
is_ipv6: false
end
@doc """
Parse an URI into components. Will return a Nurina.Info struct.
Tries to follow RFC 3986.
"""
def parse(uri) do
parsed = parse uri, %Info{}, "", :scheme
case parsed do
%{valid: true, port: nil} -> %{parsed | port: URI.default_port parsed.scheme}
_ -> parsed
end
end
# Hier part parsing
def parse(<< "//", rest :: binary >>, parsed, :hier_parse), do: parse(rest, parsed, :hier_auth)
def parse(hier, parsed, :hier_parse), do: parse(hier, parsed, :hier_no_auth)
def parse(hier, parsed, :hier_no_auth), do: %{parsed | path: nil_or hier}
def parse(hier, parsed, :hier_auth) do
parsed = parse(hier, parsed, "", :authority)
# Go inside authority to parse parts
if parsed.authority != nil do
parsed = parse(parsed.authority, parsed, "", :userinfo)
end
parsed
end
# Host part parsing
# Split into IPv6 parsing if needed, :host_4 will handle IPv4 and domains
def parse(<< "[", rest :: binary >>, parsed, :host), do: parse(rest, parsed, "", :host_6)
def parse(hier, parsed, :host), do: parse(hier, parsed, "", :host_4)
# Port part parsing
def parse(<< ":", rest :: binary >>, parsed, :port) do
case Integer.parse rest do
:error -> %{parsed | valid: false}
{_, remainder} when remainder != "" -> %{parsed | valid: false}
{port, _} -> %{parsed | port: port}
end
end
def parse(_, parsed, :port), do: %{parsed | port: nil, valid: false}
# Scheme part parsing
# If URI stops at scheme, it's not valid
def parse("", parsed, _, :scheme), do: %{parsed | valid: false}
def parse(<< ":", rest :: binary >>, parsed, current_part, :scheme) do
parsed = %{parsed | scheme: nil_or String.downcase current_part}
# Scheme must exist
case parsed.scheme do
nil -> %{parsed | valid: false}
_ -> parse(rest, parsed, "", :hier)
end
end
# Hier part parsing
# Hier is the hierarchical sequence of the URI. In RFC 3986 it is 'hier-part'.
def parse("", parsed, current_part, :hier) do
parsed = %{parsed | hier: nil_or current_part}
# Go inside hierarchy to parse parts
parse(current_part, parsed, :hier_parse)
end
def parse(<< "?", rest :: binary >>, parsed, current_part, :hier) do
parsed = %{parsed | hier: nil_or current_part}
# Go inside hierarchy to parse parts
parsed = parse(current_part, parsed, :hier_parse)
# If we have a query and
parse(rest, parsed, "", :query)
end
# Query part parsing
def parse("", parsed, current_part, :query), do: %{parsed | query: nil_or current_part}
def parse(<< "#", rest :: binary >>, parsed, current_part, :query) do
# All the rest is just fragment
%{parsed | query: nil_or(current_part), fragment: nil_or rest}
end
# Authority part parsing
def parse("", parsed, current_part, :authority), do: %{parsed | authority: nil_or current_part}
def parse(<< "/", rest :: binary >>, parsed, current_part, :authority) do
%{parsed | authority: nil_or(current_part), path: "/" <> rest}
end
# Userinfo part parsing
# If no userinfo was found, start from the beginning and look for host instead
def parse("", parsed, current_part, :userinfo), do: parse(current_part, parsed, :host)
def parse(<< "@", rest :: binary >>, parsed, current_part, :userinfo) do
parsed = %{parsed | userinfo: nil_or current_part}
parse(rest, parsed, :host)
end
# IPv6 host parsing
# Host must not end without closing ]
def parse("", parsed, _, :host_6), do: %{parsed | valid: false}
def parse(<< "]", rest :: binary >>, parsed, current_part, :host_6) do
parsed = %{parsed | is_ipv6: true, host: nil_or current_part}
parse(rest, parsed, :port)
end
# "Normal" host parsing
def parse("", parsed, current_part, :host_4), do: %{parsed | host: nil_or current_part}
def parse(<< ":", rest :: binary >>, parsed, current_part, :host_4) do
parsed = %{parsed | host: nil_or current_part}
parse(":" <> rest, parsed, :port)
end
# Default walking function for all parsing modes, just walk through all
# non recognised characters
def parse(<<char, rest :: binary>>, parsed, current_part, mode) do
current_part = current_part <> << char :: utf8 >>
parse(rest, parsed, current_part, mode)
end
# Convert "" into nil
defp nil_or(str) when str == "", do: nil
defp nil_or(str), do: str
end

15
lib/speedtest.ex Normal file
View file

@ -0,0 +1,15 @@
defmodule Nurina.Speedtest do
def run(_, iterations, _) when iterations == 0, do: nil
def run(url, iterations, :nurina) do
Nurina.parse(url)
run url, iterations - 1, :nurina
end
def run(url, iterations, :uri) do
URI.parse(url)
run url, iterations - 1, :uri
end
end

21
mix.exs Normal file
View file

@ -0,0 +1,21 @@
defmodule Nurina.Mixfile do
use Mix.Project
def project do
[ app: :nurina,
version: "0.1.0",
elixir: "~> 1.0.3",
deps: deps ]
end
# Configuration for the OTP application
def application do
[]
end
# Returns the list of dependencies in the format:
# { :foobar, "~> 0.1", git: "https://github.com/elixir-lang/foobar.git" }
defp deps do
[]
end
end

144
test/nurina_test.exs Normal file
View file

@ -0,0 +1,144 @@
defmodule NurinaTest do
use ExUnit.Case
test :parse_http do
assert %Nurina.Info{scheme: "http", host: "foo.com", path: "/path/to/something",
query: "foo=bar&bar=foo", fragment: "fragment", port: 80,
authority: "foo.com", userinfo: nil,
hier: "//foo.com/path/to/something", valid: true} ==
Nurina.parse("http://foo.com/path/to/something?foo=bar&bar=foo#fragment")
end
test :parse_https do
assert %Nurina.Info{scheme: "https", host: "foo.com", authority: "foo.com",
query: nil, fragment: nil, port: 443, path: nil, userinfo: nil,
hier: "//foo.com", valid: true} ==
Nurina.parse("https://foo.com")
end
test :parse_file do
assert %Nurina.Info{scheme: "file", host: nil, path: "/foo/bar/baz", userinfo: nil,
query: nil, fragment: nil, port: nil, authority: nil,
hier: "///foo/bar/baz", valid: true} ==
Nurina.parse("file:///foo/bar/baz")
end
test :parse_ftp do
assert %Nurina.Info{scheme: "ftp", host: "private.ftp-servers.example.com",
userinfo: "user001:secretpassword", authority: "user001:secretpassword@private.ftp-servers.example.com",
path: "/mydirectory/myfile.txt", query: nil, fragment: nil,
port: 21,
hier: "//user001:secretpassword@private.ftp-servers.example.com/mydirectory/myfile.txt", valid: true} ==
Nurina.parse("ftp://user001:secretpassword@private.ftp-servers.example.com/mydirectory/myfile.txt")
end
test :parse_sftp do
assert %Nurina.Info{scheme: "sftp", host: "private.ftp-servers.example.com",
userinfo: "user001:secretpassword", authority: "user001:secretpassword@private.ftp-servers.example.com",
path: "/mydirectory/myfile.txt", query: nil, fragment: nil,
port: 22,
hier: "//user001:secretpassword@private.ftp-servers.example.com/mydirectory/myfile.txt", valid: true} ==
Nurina.parse("sftp://user001:secretpassword@private.ftp-servers.example.com/mydirectory/myfile.txt")
end
test :parse_tftp do
assert %Nurina.Info{scheme: "tftp", host: "private.ftp-servers.example.com",
userinfo: "user001:secretpassword", authority: "user001:secretpassword@private.ftp-servers.example.com",
path: "/mydirectory/myfile.txt", query: nil, fragment: nil, port: 69,
hier: "//user001:secretpassword@private.ftp-servers.example.com/mydirectory/myfile.txt", valid: true} ==
Nurina.parse("tftp://user001:secretpassword@private.ftp-servers.example.com/mydirectory/myfile.txt")
end
test :parse_ldap do
assert %Nurina.Info{scheme: "ldap", host: nil, authority: nil, userinfo: nil,
path: "/dc=example,dc=com", query: "?sub?(givenName=John)",
fragment: nil, port: 389,
hier: "///dc=example,dc=com", valid: true} ==
Nurina.parse("ldap:///dc=example,dc=com??sub?(givenName=John)")
assert %Nurina.Info{scheme: "ldap", host: "ldap.example.com", authority: "ldap.example.com",
userinfo: nil, path: "/cn=John%20Doe,dc=example,dc=com", fragment: nil,
port: 389, query: nil,
hier: "//ldap.example.com/cn=John%20Doe,dc=example,dc=com", valid: true} ==
Nurina.parse("ldap://ldap.example.com/cn=John%20Doe,dc=example,dc=com")
end
test :parse_mailto do
assert %Nurina.Info{scheme: "mailto", host: nil, authority: nil, userinfo: nil,
path: "foo@foo.com", query: nil, fragment: nil, port: nil, hier: "foo@foo.com", valid: true} ==
Nurina.parse("mailto:foo@foo.com")
end
test :parse_splits_authority do
assert %Nurina.Info{scheme: "http", host: "foo.com", path: nil,
query: nil, fragment: nil, port: 4444,
authority: "foo:bar@foo.com:4444",
userinfo: "foo:bar",
hier: "//foo:bar@foo.com:4444", valid: true} ==
Nurina.parse("http://foo:bar@foo.com:4444")
assert %Nurina.Info{scheme: "https", host: "foo.com", path: nil,
query: nil, fragment: nil, port: 443,
authority: "foo:bar@foo.com", userinfo: "foo:bar",
hier: "//foo:bar@foo.com", valid: true} ==
Nurina.parse("https://foo:bar@foo.com")
assert %Nurina.Info{scheme: "http", host: "foo.com", path: nil,
query: nil, fragment: nil, port: 4444,
authority: "foo.com:4444",
userinfo: nil,
hier: "//foo.com:4444", valid: true} ==
Nurina.parse("http://foo.com:4444")
end
test :parse_bad_uris do
%Nurina.Info{valid: false} = Nurina.parse("https??@?F?@#>F//23/")
%Nurina.Info{valid: false} = Nurina.parse("")
%Nurina.Info{valid: false} = Nurina.parse(":https")
%Nurina.Info{valid: false} = Nurina.parse("https")
%Nurina.Info{valid: false} = Nurina.parse("http://example.com:what/")
end
test :ipv6_addresses do
addrs = [
"::", # undefined
"::1", # loopback
"1080::8:800:200C:417A", # unicast
"FF01::101", # multicast
"2607:f3f0:2:0:216:3cff:fef0:174a", # abbreviated
"2607:f3F0:2:0:216:3cFf:Fef0:174A", # mixed hex case
"2051:0db8:2d5a:3521:8313:ffad:1242:8e2e", # complete
"::00:192.168.10.184" # embedded IPv4
]
Enum.each addrs, fn(addr) ->
simple_uri = Nurina.parse("http://[#{addr}]/")
assert simple_uri.host == addr
userinfo_uri = Nurina.parse("http://user:pass@[#{addr}]/")
assert userinfo_uri.host == addr
assert userinfo_uri.userinfo == "user:pass"
port_uri = Nurina.parse("http://[#{addr}]:2222/")
assert port_uri.host == addr
assert port_uri.port == 2222
userinfo_port_uri = Nurina.parse("http://user:pass@[#{addr}]:2222/")
assert userinfo_port_uri.host == addr
assert userinfo_port_uri.userinfo == "user:pass"
assert userinfo_port_uri.port == 2222
end
end
test :downcase_scheme do
assert Nurina.parse("hTtP://google.com").scheme == "http"
end
#test :to_string do
# assert to_string(Nurina.parse("http://google.com")) == "http://google.com"
# assert to_string(Nurina.parse("http://google.com:443")) == "http://google.com:443"
# assert to_string(Nurina.parse("https://google.com:443")) == "https://google.com"
# assert to_string(Nurina.parse("http://lol:wut@google.com")) == "http://lol:wut@google.com"
# assert to_string(Nurina.parse("http://google.com/elixir")) == "http://google.com/elixir"
# assert to_string(Nurina.parse("http://google.com?q=lol")) == "http://google.com?q=lol"
# assert to_string(Nurina.parse("http://google.com?q=lol#omg")) == "http://google.com?q=lol#omg"
#end
end

1
test/test_helper.exs Normal file
View file

@ -0,0 +1 @@
ExUnit.start