From aa3b588cf40d350710d345e13df8f04656577e1f Mon Sep 17 00:00:00 2001
From: Matt Westcott <matt@west.co.tt>
Date: Tue, 5 Dec 2017 16:40:52 +0000
Subject: [PATCH] Gracefully handle bare content at the top level

---
 .../converters/html_to_contentstate.py        | 14 +++++++----
 wagtail/admin/tests/test_contentstate.py      | 24 +++++++++++++++++++
 2 files changed, 34 insertions(+), 4 deletions(-)

diff --git a/wagtail/admin/rich_text/converters/html_to_contentstate.py b/wagtail/admin/rich_text/converters/html_to_contentstate.py
index c53ffc0cb8..cd3627a1eb 100644
--- a/wagtail/admin/rich_text/converters/html_to_contentstate.py
+++ b/wagtail/admin/rich_text/converters/html_to_contentstate.py
@@ -58,7 +58,6 @@ class BlockElementHandler(object):
         self.block_type = block_type
 
     def create_block(self, name, attrs, state, contentstate):
-        assert state.current_block is None, "%s element found nested inside another block" % name
         return Block(self.block_type, depth=state.list_depth)
 
     def handle_starttag(self, name, attrs, state, contentstate):
@@ -248,7 +247,14 @@ class HtmlToContentStateHandler(HTMLParser):
 
     def handle_data(self, content):
         if self.state.current_block is None:
-            assert not content.strip(), "Bare text content found at the top level: %r" % content
+            content = content.strip()
+            if content:
+                # create a new paragraph block for this content
+                block = Block('unstyled', depth=self.state.list_depth)
+                self.contentstate.blocks.append(block)
+                self.state.current_block = block
+            else:
+                # ignore top-level whitespace
+                return
 
-        else:
-            self.state.current_block.text += content
+        self.state.current_block.text += content
diff --git a/wagtail/admin/tests/test_contentstate.py b/wagtail/admin/tests/test_contentstate.py
index e0c0adf16a..88d2301821 100644
--- a/wagtail/admin/tests/test_contentstate.py
+++ b/wagtail/admin/tests/test_contentstate.py
@@ -53,6 +53,7 @@ class TestHtmlToContentState(TestCase):
         result = json.loads(converter.from_database_format(
             '''
             <foo>Hello world!</foo>
+            <foo>I said hello world!</foo>
             <p>Goodbye world!</p>
             '''
         ))
@@ -60,6 +61,29 @@ class TestHtmlToContentState(TestCase):
             'entityMap': {},
             'blocks': [
                 {'inlineStyleRanges': [], 'text': 'Hello world!', 'depth': 0, 'type': 'unstyled', 'key': '00000', 'entityRanges': []},
+                {'inlineStyleRanges': [], 'text': 'I said hello world!', 'depth': 0, 'type': 'unstyled', 'key': '00000', 'entityRanges': []},
                 {'inlineStyleRanges': [], 'text': 'Goodbye world!', 'depth': 0, 'type': 'unstyled', 'key': '00000', 'entityRanges': []},
             ]
         })
+
+    def test_bare_text_becomes_paragraph(self):
+        converter = ContentstateConverter(features=[])
+        result = json.loads(converter.from_database_format(
+            '''
+            before
+            <p>paragraph</p>
+            between
+            <p>paragraph</p>
+            after
+            '''
+        ))
+        self.assertContentStateEqual(result, {
+            'entityMap': {},
+            'blocks': [
+                {'inlineStyleRanges': [], 'text': 'before', 'depth': 0, 'type': 'unstyled', 'key': '00000', 'entityRanges': []},
+                {'inlineStyleRanges': [], 'text': 'paragraph', 'depth': 0, 'type': 'unstyled', 'key': '00000', 'entityRanges': []},
+                {'inlineStyleRanges': [], 'text': 'between', 'depth': 0, 'type': 'unstyled', 'key': '00000', 'entityRanges': []},
+                {'inlineStyleRanges': [], 'text': 'paragraph', 'depth': 0, 'type': 'unstyled', 'key': '00000', 'entityRanges': []},
+                {'inlineStyleRanges': [], 'text': 'after', 'depth': 0, 'type': 'unstyled', 'key': '00000', 'entityRanges': []},
+            ]
+        })