fixes HTML issue for h1 containing p (#243)

Introduces a helper to strip p tags where needed

Author: jaimeiniesta

lib/school_house_web/templates/post/post.html.heex

@@ -15,7 +15,7 @@
     </div>
 
     <div class="pt-6 text-center">
-      <h1 class="text-2xl leading-8 font-extrabold tracking-tight sm:text-2xl prose dark:prose-dark"><%= raw @post.title %></h1>
+      <h1 class="text-2xl leading-8 font-extrabold tracking-tight sm:text-2xl prose dark:prose-dark"><%= @post.title |> strip_p_tags() |> raw() %></h1>
     </div>
 
     <div class="py-2 text-center">

lib/school_house_web/views/html_helpers.ex

@@ -99,4 +99,19 @@ defmodule SchoolHouseWeb.HtmlHelpers do
   def translation_status_css_class(%{status: :minor}), do: "bg-yellow-400"
   def translation_status_css_class(%{status: :patch}), do: "bg-yellow-200"
   def translation_status_css_class(_line), do: "bg-white"
+
+  @doc """
+  Removes all <p> and </p> tags from the given string, inserting newline if needed.
+
+  This is useful to avoid nesting <p> tags inside elements that don't allow them like headings or span,
+  for example in the blog posts titles.
+  """
+  def strip_p_tags(str) when is_binary(str) do
+    str
+    |> String.replace("<p>", "")
+    |> String.replace("</p>", "\n")
+    |> String.trim()
+  end
+
+  def strip_p_tags(str), do: str
 end

test/school_house_web/views/html_helpers_test.exs

@@ -69,4 +69,14 @@ defmodule SchoolHouseWeb.HtmlHelpersTest do
       assert 1 < length(locales)
     end
   end
+
+  describe "strip_p_tags" do
+    test "removes <p> and </p> tags from string" do
+      assert HtmlHelpers.strip_p_tags("<p>some content</p>") == "some content"
+    end
+
+    test "converts </p> tags to newline" do
+      assert HtmlHelpers.strip_p_tags("<p>one</p><p>two</p>") == "one\ntwo"
+    end
+  end
 end