auto_linker

commit: 479dd343f4e563ff91215c8275f3b5c67e032850
parent: 5e2661980a55a93bd34859abd094cf2185ed1dc9
Author: Egor <egor@kislitsyn.com>
Date:   Wed,  3 Apr 2019 12:02:31 +0000

Merge branch 'improve-parsing' into 'master'

Improve detection of URLs and Asian scripts

Closes #2 and #3

See merge request pleroma/auto_linker!9
Diffstat:
M
lib/auto_linker/parser.ex
23
++++++++++++++++-------
M
test/auto_linker_test.exs
23
+++++++++++++++++++++++

2 files changed, 39 insertions(+), 7 deletions(-)
diff --git a/lib/auto_linker/parser.ex b/lib/auto_linker/parser.ex
@@ -29,24 +29,22 @@ defmodule AutoLinker.Parser do
 
   @match_url ~r{^[\w\.-]+(?:\.[\w\.-]+)+[\w\-\._~%:/?#[\]@!\$&'\(\)\*\+,;=.]+$}
 
-  @match_scheme ~r{^(?:http(s)?:\/\/)?[\w.-]+(?:\.[\w\.-]+)+[\w\-\._~%:/?#[\]@!\$&'\(\)\*\+,;=.]+$}
+  @match_scheme ~r{^(?:\W*)?(?<url>(?:\W*https?:\/\/)?[\w.-]+(?:\.[\w\.-]+)+[\w\-\._~%:\/?#[\]@!\$&'\(\)\*\+,;=.]+$)}u
 
   @match_phone ~r"((?:x\d{2,7})|(?:(?:\+?1\s?(?:[.-]\s?)?)?(?:\(\s?(?:[2-9]1[02-9]|[2-9][02-8]1|[2-9][02-8][02-9])\s?\)|(?:[2-9]1[02-9]|[2-9][02-8]1|[2-9][02-8][02-9]))\s?(?:[.-]\s?)?)(?:[2-9]1[02-9]|[2-9][02-9]1|[2-9][02-9]{2})\s?(?:[.-]\s?)?(?:[0-9]{4}))"
 
-  @match_hostname ~r{^(?:https?:\/\/)?(?:[^@\n]+\\w@)?(?<host>[^:#~\/\n?]+)}
+  @match_hostname ~r{^(?:\W*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/^@[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])?)*|@[a-zA-Z\d_-]+/u
+  @match_mention ~r"^@[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])?)*|@[a-zA-Z\d_-]+"u
 
   # https://www.w3.org/TR/html5/forms.html#valid-e-mail-address
-  # credo:disable-for-next-line
-  @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_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:]_·]*)/u
+  @match_hashtag ~r/^(?<tag>\#[[:word:]_]*[[:alpha:]_·][[:word:]_·\p{M}]*)/u
 
   @prefix_extra [
     "magnet:?",
@@ -261,6 +259,17 @@ defmodule AutoLinker.Parser do
   defp do_parse({<<ch::8>> <> text, user_acc}, opts, {buffer, acc, state}, handler),
     do: do_parse({text, user_acc}, opts, {buffer <> <<ch::8>>, acc, state}, handler)
 
+  def check_and_link(buffer, %{scheme: true} = opts, _user_acc) do
+    if is_url?(buffer, opts[:scheme]) do
+      case Regex.run(@match_scheme, buffer, capture: [:url]) do
+        [^buffer] -> link_url(true, buffer, opts)
+        [url] -> String.replace(buffer, url, link_url(true, url, opts))
+      end
+    else
+      buffer
+    end
+  end
+
   def check_and_link(buffer, opts, _user_acc) do
     buffer
     |> is_url?(opts[:scheme])
diff --git a/test/auto_linker_test.exs b/test/auto_linker_test.exs
@@ -195,6 +195,21 @@ defmodule AutoLinkerTest do
              ) == expected
     end
 
+    test "support Telugu" do
+      text = "#చక్రం #కకకకక్ #కకకకాక #కకకక్రకకకక"
+
+      expected =
+        "<a href=\"/t/చక్రం\">#చక్రం</a> <a href=\"/t/కకకకక్\">#కకకకక్</a> <a href=\"/t/కకకకాక\">#కకకకాక</a> <a href=\"/t/కకకక్రకకకక\">#కకకక్రకకకక</a>"
+
+      assert AutoLinker.link(text,
+               hashtag: true,
+               hashtag_prefix: "/t/",
+               class: false,
+               rel: false,
+               new_window: false
+             ) == expected
+    end
+
     test "do not turn urls with hashes into hashtags" do
       text = "google.com#test #test google.com/#test #tag"
 
@@ -242,6 +257,14 @@ defmodule AutoLinkerTest do
       assert AutoLinker.link(text, scheme: true) == expected
     end
 
+    test "turn urls with schema into urls" do
+      text = "📌https://google.com"
+      expected = "📌<a href=\"https://google.com\">google.com</a>"
+
+      assert AutoLinker.link(text, scheme: true, class: false, new_window: false, rel: false) ==
+               expected
+    end
+
     test "hostname/@user" do
       text = "https://example.com/@user"