Merge pull request #2859 from techAPJ/patch-1

FIX: handle multiple paragraphs in email reply when parsing html
2024-11-27 17:46:05 -05:00 · 2014-10-08 13:44:24 -04:00 · 2014-10-08 13:44:24 -04:00 · a059da5d39
commit a059da5d39
parent 32441ebcac b457005b07
4 changed files with 218 additions and 2 deletions
--- a/lib/email/html_cleaner.rb
+++ b/lib/email/html_cleaner.rb
@ -76,7 +76,7 @@ module Email
    def add_newlines(doc)
      # Replace <br> tags with a markdown \n
      doc.xpath('//br').each do |br|
-        br.replace(new_linebreak_node doc)
+        br.replace(new_linebreak_node doc, 2)
      end
      # Surround <p> tags with newlines, to help with line-wise postprocessing
      # and ensure markdown paragraphs
--- a/spec/components/email/receiver_spec.rb
+++ b/spec/components/email/receiver_spec.rb
@ -82,6 +82,17 @@ Thanks for listening."
      )
    end

+    it "handles multiple paragraphs when parsing html" do
+      test_parse_body(fixture_file("emails/html_paragraphs.eml")).
+          should == (
+"Awesome!
+
+Pleasure to have you here!
+
+:boom:"
+      )
+    end
+
    it "converts back to UTF-8 at the end" do
      result = test_parse_body(fixture_file("emails/big5.eml"))
      result.encoding.should == Encoding::UTF_8
--- a/spec/fixtures/emails/html_paragraphs.eml
+++ b/spec/fixtures/emails/html_paragraphs.eml
@ -0,0 +1,205 @@
+
+MIME-Version: 1.0
+Received: by 10.25.161.144 with HTTP; Tue, 7 Oct 2014 22:17:17 -0700 (PDT)
+X-Originating-IP: [117.207.85.84]
+In-Reply-To: <5434c8b52bb3a_623ff09fec70f049749@discourse-app.mail>
+References: <topic/35@discourse.techapj.com>
+  <5434c8b52bb3a_623ff09fec70f049749@discourse-app.mail>
+Date: Wed, 8 Oct 2014 10:47:17 +0530
+Delivered-To: arpit@techapj.com
+Message-ID: <CAOJeqne=SJ_LwN4sb-0Y95ejc2OpreVhdmcPn0TnmwSvTCYzzQ@mail.gmail.com>
+Subject: Re: [Discourse] [Meta] Welcome to techAPJ's Discourse!
+From: Arpit Jalan <arpit@techapj.com>
+To: Discourse <mail+e1c7f2a380e33840aeb654f075490bad@arpitjalan.com>
+Content-Type: multipart/alternative; boundary=001a114119d8f4e46e0504e26d5b
+
+--001a114119d8f4e46e0504e26d5b
+Content-Type: text/plain; charset=UTF-8
+Content-Transfer-Encoding: quoted-printable
+
+Awesome!
+
+Pleasure to have you here!
+
+:boom:
+
+On Wed, Oct 8, 2014 at 10:46 AM, ajalan <info@unconfigured.discourse.org>
+wrote:
+
+>     ajalan
+> <http://mandrillapp.com/track/click/30081177/discourse.techapj.com?p=3Dey=
+JzIjoiVXgxTTZ3eHpuRWF2QXVoZGRJZVN5MWI0WnhrIiwidiI6MSwicCI6IntcInVcIjozMDA4M=
+TE3NyxcInZcIjoxLFwidXJsXCI6XCJodHRwOlxcXC9cXFwvZGlzY291cnNlLnRlY2hhcGouY29t=
+XFxcL3VzZXJzXFxcL2FqYWxhblwiLFwiaWRcIjpcIjgyNWI5MDYzZWNmMDRkMjk5OTE4Nzk1MmU=
+5YjY2YjE3XCIsXCJ1cmxfaWRzXCI6W1wiNzA3MTNjNTg4MDI3YWQyM2RiM2QwOTVhOGQwYmY4ZT=
+YyMzNjYThiMFwiXX0ifQ>
+> October 8
+>
+> Nice to be here! Thanks! [image: smile]
+>
+> To respond, reply to this email or visit
+> http://discourse.techapj.com/t/welcome-to-techapjs-discourse/35/2
+> <http://mandrillapp.com/track/click/30081177/discourse.techapj.com?p=3Dey=
+JzIjoid1IyWnVqVGRPU2RwLUlFR0Q5QnI1a203eVNjIiwidiI6MSwicCI6IntcInVcIjozMDA4M=
+TE3NyxcInZcIjoxLFwidXJsXCI6XCJodHRwOlxcXC9cXFwvZGlzY291cnNlLnRlY2hhcGouY29t=
+XFxcL3RcXFwvd2VsY29tZS10by10ZWNoYXBqcy1kaXNjb3Vyc2VcXFwvMzVcXFwvMlwiLFwiaWR=
+cIjpcIjgyNWI5MDYzZWNmMDRkMjk5OTE4Nzk1MmU5YjY2YjE3XCIsXCJ1cmxfaWRzXCI6W1wiY2=
+RkYzFlZjc5OThhNzE1ODA4Yjg0MGFlNzVlZmNiYmYzYmViODk4Y1wiXX0ifQ>
+> in your browser.
+>  ------------------------------
+> Previous Replies    techAPJ
+> <http://mandrillapp.com/track/click/30081177/discourse.techapj.com?p=3Dey=
+JzIjoia2x3LUxac2RSX25uWEFYYWcwVDVha3pIY3RjIiwidiI6MSwicCI6IntcInVcIjozMDA4M=
+TE3NyxcInZcIjoxLFwidXJsXCI6XCJodHRwOlxcXC9cXFwvZGlzY291cnNlLnRlY2hhcGouY29t=
+XFxcL3VzZXJzXFxcL3RlY2hhcGpcIixcImlkXCI6XCI4MjViOTA2M2VjZjA0ZDI5OTkxODc5NTJ=
+lOWI2NmIxN1wiLFwidXJsX2lkc1wiOltcIjk2ZjAyMzVhNmM2NzIyNmU1NjhhMzU1NDE1OTAxNz=
+AyYTkxNjM1NzJcIl19In0>
+> October 8
+>
+> Welcome to techAPJ's Discourse!
+>   ------------------------------
+>
+> To respond, reply to this email or visit
+> http://discourse.techapj.com/t/welcome-to-techapjs-discourse/35/2
+> <http://mandrillapp.com/track/click/30081177/discourse.techapj.com?p=3Dey=
+JzIjoid1IyWnVqVGRPU2RwLUlFR0Q5QnI1a203eVNjIiwidiI6MSwicCI6IntcInVcIjozMDA4M=
+TE3NyxcInZcIjoxLFwidXJsXCI6XCJodHRwOlxcXC9cXFwvZGlzY291cnNlLnRlY2hhcGouY29t=
+XFxcL3RcXFwvd2VsY29tZS10by10ZWNoYXBqcy1kaXNjb3Vyc2VcXFwvMzVcXFwvMlwiLFwiaWR=
+cIjpcIjgyNWI5MDYzZWNmMDRkMjk5OTE4Nzk1MmU5YjY2YjE3XCIsXCJ1cmxfaWRzXCI6W1wiY2=
+RkYzFlZjc5OThhNzE1ODA4Yjg0MGFlNzVlZmNiYmYzYmViODk4Y1wiXX0ifQ>
+> in your browser.
+>
+> To unsubscribe from these emails, visit your user preferences
+> <http://mandrillapp.com/track/click/30081177/discourse.techapj.com?p=3Dey=
+JzIjoiVTNudkpobl9lUUl0cmdsVVRrcm5iaHpyN0JZIiwidiI6MSwicCI6IntcInVcIjozMDA4M=
+TE3NyxcInZcIjoxLFwidXJsXCI6XCJodHRwOlxcXC9cXFwvZGlzY291cnNlLnRlY2hhcGouY29t=
+XFxcL215XFxcL3ByZWZlcmVuY2VzXCIsXCJpZFwiOlwiODI1YjkwNjNlY2YwNGQyOTk5MTg3OTU=
+yZTliNjZiMTdcIixcInVybF9pZHNcIjpbXCI0OTIyMmMyZDgyNzUwMmQyMGZjYzU4MTZkNjhmYT=
+k3NzFkY2YzZDllXCJdfSJ9>
+> .
+>
+
+--001a114119d8f4e46e0504e26d5b
+Content-Type: text/html; charset=UTF-8
+Content-Transfer-Encoding: quoted-printable
+
+<div dir=3D"ltr">Awesome!<div><br></div><div>Pleasure to have you here!</di=
+v><div><br></div><div>:boom:</div></div><div class=3D"gmail_extra"><br><div=
+ class=3D"gmail_quote">On Wed, Oct 8, 2014 at 10:46 AM, ajalan <span dir=3D=
+"ltr">&lt;<a href=3D"mailto:info@unconfigured.discourse.org" target=3D"_bla=
+nk">info@unconfigured.discourse.org</a>&gt;</span> wrote:<br><blockquote cl=
+ass=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;p=
+adding-left:1ex"><div>
+
+<table style=3D"margin-bottom:25px;max-width:761px" cellspacing=3D"0" cellp=
+adding=3D"0" border=3D"0">
+  <tbody>
+    <tr>
+      <td style=3D"vertical-align:top;width:55px">
+        <img src=3D"http://discourse.techapj.com/user_avatar/discourse.tech=
+apj.com/ajalan/45/35.png" title=3D"ajalan" style=3D"max-width:694px" width=
+=3D"45" height=3D"45">
+      </td>
+      <td>
+        <a href=3D"http://mandrillapp.com/track/click/30081177/discourse.te=
+chapj.com?p=3DeyJzIjoiVXgxTTZ3eHpuRWF2QXVoZGRJZVN5MWI0WnhrIiwidiI6MSwicCI6I=
+ntcInVcIjozMDA4MTE3NyxcInZcIjoxLFwidXJsXCI6XCJodHRwOlxcXC9cXFwvZGlzY291cnNl=
+LnRlY2hhcGouY29tXFxcL3VzZXJzXFxcL2FqYWxhblwiLFwiaWRcIjpcIjgyNWI5MDYzZWNmMDR=
+kMjk5OTE4Nzk1MmU5YjY2YjE3XCIsXCJ1cmxfaWRzXCI6W1wiNzA3MTNjNTg4MDI3YWQyM2RiM2=
+QwOTVhOGQwYmY4ZTYyMzNjYThiMFwiXX0ifQ" style=3D"font-size:13px;font-family:&=
+#39;lucida grande&#39;,tahoma,verdana,arial,sans-serif;color:#3b5998;text-d=
+ecoration:none;font-weight:bold;text-decoration:none;font-weight:bold;color=
+:#006699" target=3D"_blank">ajalan</a><br>
+        <span style=3D"text-align:right;color:#999999;padding-right:5px;fon=
+t-family:&#39;lucida grande&#39;,tahoma,verdana,arial,sans-serif;font-size:=
+11px">October 8</span>
+      </td>
+    </tr>
+    <tr>
+      <td style=3D"padding-top:5px" colspan=3D"2"><p style=3D"margin-top:0;=
+border:0">Nice to be here! Thanks! <img src=3D"http://discourse.techapj.com=
+/plugins/emoji/images/smile.png" title=3D":smile:" alt=3D"smile" width=3D"2=
+0" height=3D"20"></p></td>
+    </tr>
+  </tbody>
+</table>
+
+
+  <div style=3D"color:#666">
+    <p>To respond, reply to this email or visit <a href=3D"http://mandrilla=
+pp.com/track/click/30081177/discourse.techapj.com?p=3DeyJzIjoid1IyWnVqVGRPU=
+2RwLUlFR0Q5QnI1a203eVNjIiwidiI6MSwicCI6IntcInVcIjozMDA4MTE3NyxcInZcIjoxLFwi=
+dXJsXCI6XCJodHRwOlxcXC9cXFwvZGlzY291cnNlLnRlY2hhcGouY29tXFxcL3RcXFwvd2VsY29=
+tZS10by10ZWNoYXBqcy1kaXNjb3Vyc2VcXFwvMzVcXFwvMlwiLFwiaWRcIjpcIjgyNWI5MDYzZW=
+NmMDRkMjk5OTE4Nzk1MmU5YjY2YjE3XCIsXCJ1cmxfaWRzXCI6W1wiY2RkYzFlZjc5OThhNzE1O=
+DA4Yjg0MGFlNzVlZmNiYmYzYmViODk4Y1wiXX0ifQ" style=3D"color:#666;text-decorat=
+ion:none;font-weight:bold;color:#006699" target=3D"_blank">http://discourse=
+.techapj.com/t/welcome-to-techapjs-discourse/35/2</a> in your browser.</p>
+  </div>
+  <hr style=3D"background-color:#ddd;min-height:1px;border:1px;background-c=
+olor:#ddd;min-height:1px;border:1px">
+  <h4>Previous Replies</h4>
+
+  <table style=3D"margin-bottom:25px;max-width:761px" cellspacing=3D"0" cel=
+lpadding=3D"0" border=3D"0">
+  <tbody>
+    <tr>
+      <td style=3D"vertical-align:top;width:55px">
+        <img src=3D"http://discourse.techapj.com/user_avatar/discourse.tech=
+apj.com/techapj/45/34.png" title=3D"techAPJ" style=3D"max-width:694px" widt=
+h=3D"45" height=3D"45">
+      </td>
+      <td>
+        <a href=3D"http://mandrillapp.com/track/click/30081177/discourse.te=
+chapj.com?p=3DeyJzIjoia2x3LUxac2RSX25uWEFYYWcwVDVha3pIY3RjIiwidiI6MSwicCI6I=
+ntcInVcIjozMDA4MTE3NyxcInZcIjoxLFwidXJsXCI6XCJodHRwOlxcXC9cXFwvZGlzY291cnNl=
+LnRlY2hhcGouY29tXFxcL3VzZXJzXFxcL3RlY2hhcGpcIixcImlkXCI6XCI4MjViOTA2M2VjZjA=
+0ZDI5OTkxODc5NTJlOWI2NmIxN1wiLFwidXJsX2lkc1wiOltcIjk2ZjAyMzVhNmM2NzIyNmU1Nj=
+hhMzU1NDE1OTAxNzAyYTkxNjM1NzJcIl19In0" style=3D"font-size:13px;font-family:=
+&#39;lucida grande&#39;,tahoma,verdana,arial,sans-serif;color:#3b5998;text-=
+decoration:none;font-weight:bold;text-decoration:none;font-weight:bold;colo=
+r:#006699" target=3D"_blank">techAPJ</a><br>
+        <span style=3D"text-align:right;color:#999999;padding-right:5px;fon=
+t-family:&#39;lucida grande&#39;,tahoma,verdana,arial,sans-serif;font-size:=
+11px">October 8</span>
+      </td>
+    </tr>
+    <tr>
+      <td style=3D"padding-top:5px" colspan=3D"2"><p style=3D"margin-top:0;=
+border:0">Welcome to techAPJ&#39;s Discourse!</p></td>
+    </tr>
+  </tbody>
+</table>
+
+
+<hr style=3D"background-color:#ddd;min-height:1px;border:1px;background-col=
+or:#ddd;min-height:1px;border:1px">
+
+<div style=3D"color:#666">
+<p>To respond, reply to this email or visit <a href=3D"http://mandrillapp.c=
+om/track/click/30081177/discourse.techapj.com?p=3DeyJzIjoid1IyWnVqVGRPU2RwL=
+UlFR0Q5QnI1a203eVNjIiwidiI6MSwicCI6IntcInVcIjozMDA4MTE3NyxcInZcIjoxLFwidXJs=
+XCI6XCJodHRwOlxcXC9cXFwvZGlzY291cnNlLnRlY2hhcGouY29tXFxcL3RcXFwvd2VsY29tZS1=
+0by10ZWNoYXBqcy1kaXNjb3Vyc2VcXFwvMzVcXFwvMlwiLFwiaWRcIjpcIjgyNWI5MDYzZWNmMD=
+RkMjk5OTE4Nzk1MmU5YjY2YjE3XCIsXCJ1cmxfaWRzXCI6W1wiY2RkYzFlZjc5OThhNzE1ODA4Y=
+jg0MGFlNzVlZmNiYmYzYmViODk4Y1wiXX0ifQ" style=3D"color:#666;text-decoration:=
+none;font-weight:bold;color:#006699" target=3D"_blank">http://discourse.tec=
+hapj.com/t/welcome-to-techapjs-discourse/35/2</a> in your browser.</p>
+</div><span class=3D"">
+<div style=3D"color:#666">
+<p>To unsubscribe from these emails, visit your <a href=3D"http://mandrilla=
+pp.com/track/click/30081177/discourse.techapj.com?p=3DeyJzIjoiVTNudkpobl9lU=
+Ul0cmdsVVRrcm5iaHpyN0JZIiwidiI6MSwicCI6IntcInVcIjozMDA4MTE3NyxcInZcIjoxLFwi=
+dXJsXCI6XCJodHRwOlxcXC9cXFwvZGlzY291cnNlLnRlY2hhcGouY29tXFxcL215XFxcL3ByZWZ=
+lcmVuY2VzXCIsXCJpZFwiOlwiODI1YjkwNjNlY2YwNGQyOTk5MTg3OTUyZTliNjZiMTdcIixcIn=
+VybF9pZHNcIjpbXCI0OTIyMmMyZDgyNzUwMmQyMGZjYzU4MTZkNjhmYTk3NzFkY2YzZDllXCJdf=
+SJ9" style=3D"color:#666;text-decoration:none;font-weight:bold;color:#00669=
+9" target=3D"_blank">user preferences</a>.</p>
+</div>
+</span></div>
+
+<img src=3D"http://mandrillapp.com/track/open.php?u=3D30081177&amp;id=3D825=
+b9063ecf04d2999187952e9b66b17" height=3D"1" width=3D"1"></blockquote></div>=
+<br></div>
+
+--001a114119d8f4e46e0504e26d5b--
--- a/spec/fixtures/emails/paragraphs.eml
+++ b/spec/fixtures/emails/paragraphs.eml
@ -39,4 +39,4 @@ On Sun, Jun 9, 2013 at 1:39 PM, eviltrout via Discourse Meta
 > Please visit this link to respond: http://localhost:3000/t/adventure-time-sux/1234/3
 >
 > To unsubscribe from these emails, visit your [user preferences](http://localhost:3000/user_preferences).
->
+>