Render empty paragraph tags

pydocx/export/base.py

@@ -341,8 +341,10 @@ def export_run(self, run):
 
         # TODO squash multiple sequential text nodes into one?
         results = self.yield_nested(run.children, self.export_node)
+
         if run.effective_properties:
             results = self.export_run_apply_properties(run, results)
+
         return results
 
     def get_run_styles_to_apply(self, run):

pydocx/export/html.py

@@ -101,13 +101,15 @@ def __init__(
         allow_self_closing=False,
         closed=False,
         allow_whitespace=False,
+        custom_text=None,
         **attrs
     ):
         self.tag = tag
         self.allow_self_closing = allow_self_closing
         self.attrs = attrs
         self.closed = closed
         self.allow_whitespace = allow_whitespace
+        self.custom_text = custom_text
 
     def apply(self, results, allow_empty=True):
         if not allow_empty:
@@ -116,12 +118,15 @@ def apply(self, results, allow_empty=True):
                 return
 
         sequence = [[self]]
+
+        if self.custom_text:
+            sequence.append([self.custom_text])
+
         if results is not None:
             sequence.append(results)
 
         if not self.allow_self_closing:
             sequence.append([self.close()])
-
         results = chain(*sequence)
 
         for result in results:
@@ -178,6 +183,9 @@ def style(self):
         styles = {
             'body': {
                 'margin': '0px auto',
+            },
+            'p': {
+                'margin': '0'
             }
         }
 
@@ -248,17 +256,15 @@ def export_footnote(self, footnote):
         return tag.apply(results, allow_empty=False)
 
     def get_paragraph_tag(self, paragraph):
+        if paragraph.is_empty:
+            return HtmlTag('p', custom_text=' ')
+
         heading_style = paragraph.heading_style
         if heading_style:
             tag = self.get_heading_tag(paragraph)
             if tag:
                 return tag
-        if self.in_table_cell:
-            return
-        if paragraph.has_structured_document_parent():
-            return
-        if isinstance(paragraph.parent, NumberingItem):
-            return
+
         return HtmlTag('p')
 
     def get_heading_tag(self, paragraph):
@@ -277,10 +283,6 @@ def get_heading_tag(self, paragraph):
     def export_paragraph(self, paragraph):
         results = super(PyDocXHTMLExporter, self).export_paragraph(paragraph)
 
-        results = is_not_empty_and_not_only_whitespace(results)
-        if results is None:
-            return
-
         tag = self.get_paragraph_tag(paragraph)
         if tag:
             results = tag.apply(results)
@@ -585,9 +587,10 @@ def export_table_cell(self, table_cell):
             tag = HtmlTag('td', **attrs)
 
         numbering_spans = self.yield_numbering_spans(table_cell.children)
-        results = self.yield_nested_with_line_breaks_between_paragraphs(
+
+        results = self.yield_nested(
             numbering_spans,
-            self.export_node,
+            self.export_node
         )
         if tag:
             results = tag.apply(results)
@@ -733,10 +736,8 @@ def export_numbering_span(self, numbering_span):
         return tag.apply(results)
 
     def export_numbering_item(self, numbering_item):
-        results = self.yield_nested_with_line_breaks_between_paragraphs(
-            numbering_item.children,
-            self.export_node,
-        )
+        results = super(PyDocXHTMLExporter, self).export_numbering_item(numbering_item)
+
         tag = HtmlTag('li')
         return tag.apply(results)
 

pydocx/openxml/wordprocessing/paragraph.py

@@ -6,17 +6,17 @@
 )
 
 from pydocx.models import XmlModel, XmlCollection, XmlChild
+from pydocx.openxml.wordprocessing.bookmark import Bookmark
+from pydocx.openxml.wordprocessing.deleted_run import DeletedRun
 from pydocx.openxml.wordprocessing.hyperlink import Hyperlink
+from pydocx.openxml.wordprocessing.inserted_run import InsertedRun
 from pydocx.openxml.wordprocessing.paragraph_properties import ParagraphProperties  # noqa
 from pydocx.openxml.wordprocessing.run import Run
-from pydocx.openxml.wordprocessing.tab_char import TabChar
-from pydocx.openxml.wordprocessing.text import Text
-from pydocx.openxml.wordprocessing.smart_tag_run import SmartTagRun
-from pydocx.openxml.wordprocessing.inserted_run import InsertedRun
-from pydocx.openxml.wordprocessing.deleted_run import DeletedRun
 from pydocx.openxml.wordprocessing.sdt_run import SdtRun
 from pydocx.openxml.wordprocessing.simple_field import SimpleField
-from pydocx.openxml.wordprocessing.bookmark import Bookmark
+from pydocx.openxml.wordprocessing.smart_tag_run import SmartTagRun
+from pydocx.openxml.wordprocessing.tab_char import TabChar
+from pydocx.openxml.wordprocessing.text import Text
 
 
 class Paragraph(XmlModel):
@@ -39,6 +39,20 @@ def __init__(self, **kwargs):
         super(Paragraph, self).__init__(**kwargs)
         self._effective_properties = None
 
+    @property
+    def is_empty(self):
+        if not self.children:
+            return True
+
+        # we may have cases when a paragraph has a Bookmark with name '_GoBack'
+        # and we should treat it as empty paragraph
+        if len(self.children) == 1 and \
+                isinstance(self.children[0], Bookmark) and \
+                self.children[0].name in ('_GoBack',):
+            return True
+
+        return False
+
     @property
     def effective_properties(self):
         # TODO need to calculate effective properties like Run

pydocx/test/testcases.py

@@ -50,6 +50,7 @@
     '.pydocx-tab {display:inline-block;width:4em}'
     '.pydocx-underline {text-decoration:underline}'
     'body {margin:0px auto;width:51.00em}'
+    'p {margin:0}'
     '</style>'
 )