Page MenuHomePhorge
Files F114995

No OneTemporary
Actions

Size
15 KB
Referenced Files
None
Subscribers
None

View Options

diff --git a/lib/linkify/parser.ex b/lib/linkify/parser.ex
index af826c3..64d08db 100644
--- a/lib/linkify/parser.ex
+++ b/lib/linkify/parser.ex
@@ -1,362 +1,363 @@
defmodule Linkify.Parser do
@moduledoc """
Module to handle parsing the the input string.
"""
alias Linkify.Builder
@invalid_url ~r/(\.\.+)|(^(\d+\.){1,2}\d+$)/
@match_url ~r{^(?:\W*)?(?<url>(?:https?:\/\/)?[\w.-]+(?:\.[\w\.-]+)+[\w\-\._~%:\/?#[\]@!\$&'\(\)\*\+,;=.]+$)}u
@match_hostname ~r{^\W*(?<scheme>https?:\/\/)?(?:[^@\n]+\\w@)?(?<host>[^:#~\/\n?]+)}u
@match_ip ~r"^(([0-9]|[1-9][0-9]|1[0-9]{2}|2[0-4][0-9]|25[0-5])\.){3}([0-9]|[1-9][0-9]|1[0-9]{2}|2[0-4][0-9]|25[0-5])$"
# @user
# @user@example.com
+ # credo:disable-for-next-line
@match_mention ~r/^(?:\W*)?(?<long>@[a-zA-Z\d_-]+@[a-zA-Z0-9_-](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?(?:\.[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?)*)|^(?:\W*)?(?<short>@[a-zA-Z\d_-]+)/u
# https://www.w3.org/TR/html5/forms.html#valid-e-mail-address
@match_email ~r"^[a-zA-Z0-9.!#$%&'*+\/=?^_`{|}~-]+@[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?(?:\.[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?)*$"u
@match_hashtag ~r/^(?<tag>\#[[:word:]_]*[[:alpha:]_·][[:word:]_·\p{M}]*)/u
@match_skipped_tag ~r/^(?<tag>(a|code|pre)).*>*/
@delimiters ~r/[,.;:>]*$/
@prefix_extra [
"magnet:?",
"dweb://",
"dat://",
"gopher://",
"ipfs://",
"ipns://",
"irc://",
"ircs://",
"irc6://",
"mumble://",
"ssb://"
]
@tlds "./priv/tlds.txt" |> File.read!() |> String.split("\n", trim: true) |> MapSet.new()
@default_opts %{
url: true,
validate_tld: true
}
@doc """
Parse the given string, identifying items to link.
Parses the string, replacing the matching urls with an html link.
## Examples
iex> Linkify.Parser.parse("Check out google.com")
~s{Check out <a href="http://google.com">google.com</a>}
"""
@types [:url, :email, :hashtag, :extra, :mention]
def parse(input, opts \\ %{})
def parse(input, opts) when is_binary(input), do: {input, %{}} |> parse(opts) |> elem(0)
def parse(input, list) when is_list(list), do: parse(input, Enum.into(list, %{}))
def parse(input, opts) do
opts = Map.merge(@default_opts, opts)
{buffer, user_acc} = do_parse(input, opts, {"", [], :parsing})
if opts[:iodata] do
{buffer, user_acc}
else
{IO.iodata_to_binary(buffer), user_acc}
end
end
defp accumulate(acc, buffer),
do: [buffer | acc]
defp accumulate(acc, buffer, trailing),
do: [trailing, buffer | acc]
defp do_parse({"", user_acc}, _opts, {"", acc, _}),
do: {Enum.reverse(acc), user_acc}
defp do_parse(
{"<" <> text, user_acc},
%{hashtag: true} = opts,
{"#" <> _ = buffer, acc, :parsing}
) do
{buffer, user_acc} = link(buffer, opts, user_acc)
case Regex.run(@match_skipped_tag, text, capture: [:tag]) do
[tag] ->
text = String.trim_leading(text, tag)
do_parse({text, user_acc}, opts, {"", accumulate(acc, buffer, "<#{tag}"), :skip})
nil ->
do_parse({text, user_acc}, opts, {"<", acc, {:open, 1}})
end
end
defp do_parse({"<br" <> text, user_acc}, opts, {buffer, acc, :parsing}) do
{buffer, user_acc} = link(buffer, opts, user_acc)
do_parse({text, user_acc}, opts, {"", accumulate(acc, buffer, "<br"), {:open, 1}})
end
defp do_parse({"<a" <> text, user_acc}, opts, {buffer, acc, :parsing}),
do: do_parse({text, user_acc}, opts, {"", accumulate(acc, buffer, "<a"), :skip})
defp do_parse({"<pre" <> text, user_acc}, opts, {buffer, acc, :parsing}),
do: do_parse({text, user_acc}, opts, {"", accumulate(acc, buffer, "<pre"), :skip})
defp do_parse({"<code" <> text, user_acc}, opts, {buffer, acc, :parsing}),
do: do_parse({text, user_acc}, opts, {"", accumulate(acc, buffer, "<code"), :skip})
defp do_parse({"</a>" <> text, user_acc}, opts, {buffer, acc, :skip}),
do: do_parse({text, user_acc}, opts, {"", accumulate(acc, buffer, "</a>"), :parsing})
defp do_parse({"</pre>" <> text, user_acc}, opts, {buffer, acc, :skip}),
do: do_parse({text, user_acc}, opts, {"", accumulate(acc, buffer, "</pre>"), :parsing})
defp do_parse({"</code>" <> text, user_acc}, opts, {buffer, acc, :skip}),
do: do_parse({text, user_acc}, opts, {"", accumulate(acc, buffer, "</code>"), :parsing})
defp do_parse({"<" <> text, user_acc}, opts, {"", acc, :parsing}),
do: do_parse({text, user_acc}, opts, {"<", acc, {:open, 1}})
defp do_parse({">" <> text, user_acc}, opts, {buffer, acc, {:attrs, _level}}),
do: do_parse({text, user_acc}, opts, {"", accumulate(acc, buffer, ">"), :parsing})
defp do_parse({<<ch::8>> <> text, user_acc}, opts, {"", acc, {:attrs, level}}) do
do_parse({text, user_acc}, opts, {"", accumulate(acc, <<ch::8>>), {:attrs, level}})
end
defp do_parse({text, user_acc}, opts, {buffer, acc, {:open, level}}) do
do_parse({text, user_acc}, opts, {"", accumulate(acc, buffer), {:attrs, level}})
end
defp do_parse(
{<<char::bytes-size(1), text::binary>>, user_acc},
opts,
{buffer, acc, state}
)
when char in [" ", "\r", "\n"] do
{buffer, user_acc} = link(buffer, opts, user_acc)
do_parse(
{text, user_acc},
opts,
{"", accumulate(acc, buffer, char), state}
)
end
defp do_parse({<<ch::8>>, user_acc}, opts, {buffer, acc, state}) do
{buffer, user_acc} = link(buffer <> <<ch::8>>, opts, user_acc)
do_parse(
{"", user_acc},
opts,
{"", accumulate(acc, buffer), state}
)
end
defp do_parse({<<ch::8>> <> text, user_acc}, opts, {buffer, acc, state}),
do: do_parse({text, user_acc}, opts, {buffer <> <<ch::8>>, acc, state})
def check_and_link(:url, buffer, opts, _user_acc) do
if url?(buffer, opts) do
case @match_url |> Regex.run(buffer, capture: [:url]) |> hd() do
^buffer ->
link_url(buffer, opts)
url ->
link = link_url(url, opts)
restore_stripped_symbols(buffer, url, link)
end
else
:nomatch
end
end
def check_and_link(:email, buffer, opts, _user_acc) do
if email?(buffer, opts), do: link_email(buffer, opts), else: :nomatch
end
def check_and_link(:mention, buffer, opts, user_acc) do
buffer
|> match_mention
|> link_mention(buffer, opts, user_acc)
end
def check_and_link(:hashtag, buffer, opts, user_acc) do
buffer
|> match_hashtag
|> link_hashtag(buffer, opts, user_acc)
end
def check_and_link(:extra, "xmpp:" <> handle = buffer, opts, _user_acc) do
if email?(handle, opts), do: link_extra(buffer, opts), else: :nomatch
end
def check_and_link(:extra, buffer, opts, _user_acc) do
if String.starts_with?(buffer, @prefix_extra), do: link_extra(buffer, opts), else: :nomatch
end
defp strip_parens(buffer) do
buffer
|> String.trim_leading("(")
|> String.trim_trailing(")")
end
defp strip_punctuation(buffer), do: String.replace(buffer, @delimiters, "")
def url?(buffer, opts) do
valid_url?(buffer) && Regex.match?(@match_url, buffer) && valid_tld?(buffer, opts)
end
def email?(buffer, opts) do
valid_url?(buffer) && Regex.match?(@match_email, buffer) && valid_tld?(buffer, opts)
end
defp valid_url?(url), do: !Regex.match?(@invalid_url, url)
@doc """
Validates a URL's TLD. Returns a boolean.
Will return `true` if `:validate_tld` option set to `false`.
Will skip validation and return `true` if `:validate_tld` set to `:no_scheme` and the url has a scheme.
"""
def valid_tld?(url, opts) do
[scheme, host] = Regex.run(@match_hostname, url, capture: [:scheme, :host])
cond do
opts[:validate_tld] == false ->
true
ip?(host) ->
true
# don't validate if scheme is present
opts[:validate_tld] == :no_scheme and scheme != "" ->
true
true ->
tld = host |> String.split(".") |> List.last()
MapSet.member?(@tlds, tld)
end
end
def ip?(buffer), do: Regex.match?(@match_ip, buffer)
def match_mention(buffer) do
case Regex.run(@match_mention, buffer, capture: [:long, :short]) do
[mention, ""] -> mention
["", mention] -> mention
_ -> nil
end
end
def match_hashtag(buffer) do
case Regex.run(@match_hashtag, buffer, capture: [:tag]) do
[hashtag] -> hashtag
_ -> nil
end
end
def link_hashtag(nil, _buffer, _, _user_acc), do: :nomatch
def link_hashtag(hashtag, buffer, %{hashtag_handler: hashtag_handler} = opts, user_acc) do
hashtag
|> hashtag_handler.(buffer, opts, user_acc)
|> maybe_update_buffer(hashtag, buffer)
end
def link_hashtag(hashtag, buffer, opts, _user_acc) do
hashtag
|> Builder.create_hashtag_link(buffer, opts)
|> maybe_update_buffer(hashtag, buffer)
end
def link_mention(nil, _buffer, _, _user_acc), do: :nomatch
def link_mention(mention, buffer, %{mention_handler: mention_handler} = opts, user_acc) do
mention
|> mention_handler.(buffer, opts, user_acc)
|> maybe_update_buffer(mention, buffer)
end
def link_mention(mention, buffer, opts, _user_acc) do
mention
|> Builder.create_mention_link(buffer, opts)
|> maybe_update_buffer(mention, buffer)
end
defp maybe_update_buffer(out, match, buffer) when is_binary(out) do
maybe_update_buffer({out, nil}, match, buffer)
end
defp maybe_update_buffer({out, user_acc}, match, buffer)
when match != buffer and out != buffer do
out = String.replace(buffer, match, out)
{out, user_acc}
end
defp maybe_update_buffer(out, _match, _buffer), do: out
@doc false
def link_url(buffer, opts) do
Builder.create_link(buffer, opts)
end
@doc false
def link_email(buffer, opts) do
Builder.create_email_link(buffer, opts)
end
def link_extra(buffer, opts) do
Builder.create_extra_link(buffer, opts)
end
defp link(buffer, opts, user_acc) do
Enum.reduce_while(@types, {buffer, user_acc}, fn type, _ ->
if opts[type] == true do
check_and_link_reducer(type, buffer, opts, user_acc)
else
{:cont, {buffer, user_acc}}
end
end)
end
defp check_and_link_reducer(type, buffer, opts, user_acc) do
str =
buffer
|> String.split("<")
|> List.first()
|> strip_punctuation()
|> strip_parens()
case check_and_link(type, str, opts, user_acc) do
:nomatch ->
{:cont, {buffer, user_acc}}
{link, user_acc} ->
{:halt, {restore_stripped_symbols(buffer, str, link), user_acc}}
link ->
{:halt, {restore_stripped_symbols(buffer, str, link), user_acc}}
end
end
defp restore_stripped_symbols(buffer, buffer, link), do: link
defp restore_stripped_symbols(buffer, stripped_buffer, link) do
buffer
|> String.split(stripped_buffer)
|> Enum.intersperse(link)
end
end
diff --git a/mix.exs b/mix.exs
index 93b7934..f78671d 100644
--- a/mix.exs
+++ b/mix.exs
@@ -1,44 +1,44 @@
defmodule Linkify.Mixfile do
use Mix.Project
@version "0.2.0"
def project do
[
app: :linkify,
version: @version,
elixir: "~> 1.8",
build_embedded: Mix.env() == :prod,
start_permanent: Mix.env() == :prod,
deps: deps(),
docs: [extras: ["README.md"]],
package: package(),
name: "Linkify",
description: """
Linkify is a basic package for turning website names into links.
"""
]
end
# Configuration for the OTP application
def application do
# Specify extra applications you'll use from Erlang/Elixir
[extra_applications: [:logger]]
end
# Dependencies can be Hex packages:
defp deps do
[
{:ex_doc, "~> 0.20", only: :dev, runtime: false},
- {:credo, "~> 1.4.0", only: [:dev, :test], runtime: false}
+ {:credo, "~> 1.5", only: [:dev, :test], runtime: false}
]
end
defp package do
[
licenses: ["MIT"],
links: %{"GitLab" => "https://git.pleroma.social/pleroma/elixir-libraries/linkify"},
files: ~w(lib priv README.md mix.exs LICENSE)
]
end
end
diff --git a/mix.lock b/mix.lock
index be2b77f..a658f27 100644
--- a/mix.lock
+++ b/mix.lock
@@ -1,11 +1,12 @@
%{
"bunt": {:hex, :bunt, "0.2.0", "951c6e801e8b1d2cbe58ebbd3e616a869061ddadcc4863d0a2182541acae9a38", [:mix], [], "hexpm", "7af5c7e09fe1d40f76c8e4f9dd2be7cebd83909f31fee7cd0e9eadc567da8353"},
- "credo": {:hex, :credo, "1.4.0", "92339d4cbadd1e88b5ee43d427b639b68a11071b6f73854e33638e30a0ea11f5", [:mix], [{:bunt, "~> 0.2.0", [hex: :bunt, repo: "hexpm", optional: false]}, {:jason, "~> 1.0", [hex: :jason, repo: "hexpm", optional: false]}], "hexpm", "1fd3b70dce216574ce3c18bdf510b57e7c4c85c2ec9cad4bff854abaf7e58658"},
+ "credo": {:hex, :credo, "1.5.1", "4fe303cc828412b9d21eed4eab60914c401e71f117f40243266aafb66f30d036", [:mix], [{:bunt, "~> 0.2.0", [hex: :bunt, repo: "hexpm", optional: false]}, {:file_system, "~> 0.2.8", [hex: :file_system, repo: "hexpm", optional: false]}, {:jason, "~> 1.0", [hex: :jason, repo: "hexpm", optional: false]}], "hexpm", "0b219ca4dcc89e4e7bc6ae7e6539c313e738e192e10b85275fa1e82b5203ecd7"},
"earmark": {:hex, :earmark, "1.4.10", "bddce5e8ea37712a5bfb01541be8ba57d3b171d3fa4f80a0be9bcf1db417bcaf", [:mix], [{:earmark_parser, ">= 1.4.10", [hex: :earmark_parser, repo: "hexpm", optional: false]}], "hexpm", "12dbfa80810478e521d3ffb941ad9fbfcbbd7debe94e1341b4c4a1b2411c1c27"},
"earmark_parser": {:hex, :earmark_parser, "1.4.10", "6603d7a603b9c18d3d20db69921527f82ef09990885ed7525003c7fe7dc86c56", [:mix], [], "hexpm", "8e2d5370b732385db2c9b22215c3f59c84ac7dda7ed7e544d7c459496ae519c0"},
"ex_doc": {:hex, :ex_doc, "0.22.1", "9bb6d51508778193a4ea90fa16eac47f8b67934f33f8271d5e1edec2dc0eee4c", [:mix], [{:earmark, "~> 1.4.0", [hex: :earmark, repo: "hexpm", optional: false]}, {:makeup_elixir, "~> 0.14", [hex: :makeup_elixir, repo: "hexpm", optional: false]}], "hexpm", "d957de1b75cb9f78d3ee17820733dc4460114d8b1e11f7ee4fd6546e69b1db60"},
- "jason": {:hex, :jason, "1.2.1", "12b22825e22f468c02eb3e4b9985f3d0cb8dc40b9bd704730efa11abd2708c44", [:mix], [{:decimal, "~> 1.0", [hex: :decimal, repo: "hexpm", optional: true]}], "hexpm", "b659b8571deedf60f79c5a608e15414085fa141344e2716fbd6988a084b5f993"},
+ "file_system": {:hex, :file_system, "0.2.9", "545b9c9d502e8bfa71a5315fac2a923bd060fd9acb797fe6595f54b0f975fd32", [:mix], [], "hexpm", "3cf87a377fe1d93043adeec4889feacf594957226b4f19d5897096d6f61345d8"},
+ "jason": {:hex, :jason, "1.2.2", "ba43e3f2709fd1aa1dce90aaabfd039d000469c05c56f0b8e31978e03fa39052", [:mix], [{:decimal, "~> 1.0 or ~> 2.0", [hex: :decimal, repo: "hexpm", optional: true]}], "hexpm", "18a228f5f0058ee183f29f9eae0805c6e59d61c3b006760668d8d18ff0d12179"},
"makeup": {:hex, :makeup, "1.0.3", "e339e2f766d12e7260e6672dd4047405963c5ec99661abdc432e6ec67d29ef95", [:mix], [{:nimble_parsec, "~> 0.5", [hex: :nimble_parsec, repo: "hexpm", optional: false]}], "hexpm", "2e9b4996d11832947731f7608fed7ad2f9443011b3b479ae288011265cdd3dad"},
"makeup_elixir": {:hex, :makeup_elixir, "0.14.1", "4f0e96847c63c17841d42c08107405a005a2680eb9c7ccadfd757bd31dabccfb", [:mix], [{:makeup, "~> 1.0", [hex: :makeup, repo: "hexpm", optional: false]}], "hexpm", "f2438b1a80eaec9ede832b5c41cd4f373b38fd7aa33e3b22d9db79e640cbde11"},
"nimble_parsec": {:hex, :nimble_parsec, "0.6.0", "32111b3bf39137144abd7ba1cce0914533b2d16ef35e8abc5ec8be6122944263", [:mix], [], "hexpm", "27eac315a94909d4dc68bc07a4a83e06c8379237c5ea528a9acff4ca1c873c52"},
}

File Metadata

Mime Type
text/x-diff
Expires
Tue, Nov 26, 10:53 PM (1 d, 14 h)
Storage Engine
blob
Storage Format
Raw Data
Storage Handle
40504
Default Alt Text
(15 KB)

Event Timeline