Support underlines in style map

8 files changed, 107 insertions, 28 deletions

diff --git a/mammoth/conversion.py b/mammoth/conversion.py
index 1569849..fc35efe 100644
--- a/mammoth/conversion.py
+++ b/mammoth/conversion.py
@@ -6,7 +6,7 @@ import base64
 import random
 
 from . import documents, results, html_paths, images, writers
-from .html_generation import HtmlGenerator, satisfy_html_path
+from .html_generation import HtmlGenerator, satisfy_html_path, append_html_path
 
 
 def convert_document_element_to_html(element,
@@ -57,7 +57,7 @@ class DocumentConverter(object):
             documents.NoteReference: self._convert_note_reference,
             documents.Note: self._convert_note,
         }
-        self._convert_underline = convert_underline
+        self._convert_underline = convert_underline or self._default_convert_underline
 
 
     def convert_element_to_html(self, element, html_generator):
@@ -97,12 +97,18 @@ class DocumentConverter(object):
             run_generator.start("sup")
         if run.vertical_alignment == documents.VerticalAlignment.subscript:
             run_generator.start("sub")
-        if run.is_underline and self._convert_underline is not None:
+        if run.is_underline:
             self._convert_underline(run_generator)
         self._convert_elements_to_html(run.children, run_generator)
         run_generator.end_all()
         html_generator.append(run_generator)
-
+    
+    
+    def _default_convert_underline(self, run_generator):
+        style = self._find_style(None, "underline")
+        if style is not None:
+            append_html_path(run_generator, style.html_path)
+    
 
     def _convert_text(self, text, html_generator):
         html_generator.text(text.value)
@@ -201,10 +207,9 @@ class DocumentConverter(object):
         
     
     def _find_html_path(self, element, element_type, default):
-        for style in self._style_map:
-            document_matcher = style.document_matcher
-            if _document_matcher_matches(document_matcher, element, element_type):
-                return style.html_path
+        style = self._find_style(element, element_type)
+        if style is not None:
+            return style.html_path
         
         if element.style_id is not None:
             self.messages.append(results.warning(
@@ -213,7 +218,12 @@ class DocumentConverter(object):
             ))
         
         return default
-        
+    
+    def _find_style(self, element, element_type):
+        for style in self._style_map:
+            document_matcher = style.document_matcher
+            if _document_matcher_matches(document_matcher, element, element_type):
+                return style
 
     def _note_html_id(self, note):
         return self._html_id("{0}-{1}".format(note.note_type, note.note_id))
@@ -226,18 +236,21 @@ class DocumentConverter(object):
         
 
 def _document_matcher_matches(matcher, element, element_type):
-    return (
-        matcher.element_type == element_type and (
-            matcher.style_id is None or
-            matcher.style_id == element.style_id
-        ) and (
-            matcher.style_name is None or
-            element.style_name is not None and (matcher.style_name.upper() == element.style_name.upper())
-        ) and (
-            element_type != "paragraph" or
-            matcher.numbering is None or
-            matcher.numbering == element.numbering
+    if matcher.element_type == "underline":
+        return element_type == "underline"
+    else:
+        return (
+            matcher.element_type == element_type and (
+                matcher.style_id is None or
+                matcher.style_id == element.style_id
+            ) and (
+                matcher.style_name is None or
+                element.style_name is not None and (matcher.style_name.upper() == element.style_name.upper())
+            ) and (
+                element_type != "paragraph" or
+                matcher.numbering is None or
+                matcher.numbering == element.numbering
+            )
         )
-    )
 
 _up_arrow = "↑"
diff --git a/mammoth/document_matchers.py b/mammoth/document_matchers.py
index 222ee73..07d5be2 100644
--- a/mammoth/document_matchers.py
+++ b/mammoth/document_matchers.py
@@ -15,3 +15,7 @@ def run(style_id=None, style_name=None):
 
 RunMatcher = collections.namedtuple("RunMatcher", ["style_id", "style_name"])
 RunMatcher.element_type = "run"
+
+
+class underline(object):
+    element_type = "underline"
diff --git a/mammoth/html_generation.py b/mammoth/html_generation.py
index c7f4f38..e538f6e 100644
--- a/mammoth/html_generation.py
+++ b/mammoth/html_generation.py
@@ -68,12 +68,22 @@ def satisfy_html_path(generator, path):
         generator.end()
     
     for element in path.elements[first_unsatisfied_index:]:
-        attributes = {}
-        if element.class_names:
-            attributes["class"] = _generate_class_attribute(element)
-        generator.start(element.names[0], attributes=attributes)
+        _append_element(generator, element)
+
+
+def append_html_path(generator, path):
+    for element in path.elements:
+        _append_element(generator, element)
+
+
+def _append_element(generator, element):
+    attributes = {}
+    if element.class_names:
+        attributes["class"] = _generate_class_attribute(element)
+    generator.start(element.names[0], attributes=attributes)
     
 
+
 def _find_first_unsatisfied_index(generator, path):
     for index, (generated_element, path_element) in enumerate(zip(generator._stack, path.elements)):
         if not _is_element_match(generated_element, path_element):
diff --git a/mammoth/style_reader/document_matcher_reader.py b/mammoth/style_reader/document_matcher_reader.py
index 3dc9892..56c4626 100644
--- a/mammoth/style_reader/document_matcher_reader.py
+++ b/mammoth/style_reader/document_matcher_reader.py
@@ -14,6 +14,8 @@ def read_document_matcher_node(root_node):
         return _read_paragraph_node(element_node)
     elif element_node.expr_name == "run":
         return _read_run_node(element_node)
+    elif element_node.expr_name == "underline":
+        return document_matchers.underline
     
 
 def _read_paragraph_node(paragraph_node):
@@ -53,7 +55,9 @@ def _read_list_node(list_node):
         return None
 
 grammar_text = r"""
-document_matcher = paragraph / run
+document_matcher = paragraph / run / underline
+
+underline = "u"
 
 paragraph = "p" style_id? style_name_specifier? list?
 
diff --git a/tests/conversion_tests.py b/tests/conversion_tests.py
index 4b5b461..19df77f 100644
--- a/tests/conversion_tests.py
+++ b/tests/conversion_tests.py
@@ -172,7 +172,7 @@ def underline_runs_are_ignored_by_default():
     
 
 @istest
-def underline_runs_can_be_wrapped_in_tags():
+def underline_runs_can_be_wrapped_in_tags_using_convert_underline_argument():
     result = convert_document_element_to_html(
         documents.run(children=[documents.text("Hello")], is_underline=True),
         convert_underline=mammoth.underline.element("u")
@@ -181,6 +181,28 @@ def underline_runs_can_be_wrapped_in_tags():
     
 
 @istest
+def underline_runs_can_be_mapped_using_style_mapping():
+    result = convert_document_element_to_html(
+        documents.run(children=[documents.text("Hello")], is_underline=True),
+        style_map=[
+            style_reader.read_style("u => em")
+        ]
+    )
+    assert_equal("<em>Hello</em>", result.value)
+
+
+@istest
+def style_mapping_for_underline_runs_does_not_close_parent_elements():
+    result = convert_document_element_to_html(
+        documents.run(children=[documents.text("Hello")], is_underline=True, is_bold=True),
+        style_map=[
+            style_reader.read_style("u => em")
+        ]
+    )
+    assert_equal("<strong><em>Hello</em></strong>", result.value)
+
+
+@istest
 def superscript_runs_are_wrapped_in_sup_tags():
     result = convert_document_element_to_html(
         documents.run(
diff --git a/tests/html_generation_tests.py b/tests/html_generation_tests.py
index ccb0e7e..2375564 100644
--- a/tests/html_generation_tests.py
+++ b/tests/html_generation_tests.py
@@ -1,6 +1,6 @@
 from nose.tools import istest, assert_equal
 
-from mammoth.html_generation import HtmlGenerator, satisfy_html_path
+from mammoth.html_generation import HtmlGenerator, satisfy_html_path, append_html_path
 from mammoth import html_paths, writers
 
 
@@ -171,5 +171,17 @@ class SatisfyPathTests(object):
         assert_equal('<p class="tip">HelpTip', generator.as_string())
 
 
+@istest
+class AppendPathTests(object):
+    @istest
+    def already_opened_elements_are_not_closed(self):
+        generator = _create_html_generator()
+        generator.start("strong")
+        path = html_paths.path([html_paths.element(["em"])])
+        append_html_path(generator, path)
+        generator.text("Hello!")
+        assert_equal('<strong><em>Hello!', generator.as_string())
+
+
 def _create_html_generator():
     return HtmlGenerator(writers.writer)
diff --git a/tests/mammoth_tests.py b/tests/mammoth_tests.py
index a1b396c..e7e7e64 100644
--- a/tests/mammoth_tests.py
+++ b/tests/mammoth_tests.py
@@ -124,6 +124,13 @@ def underline_can_be_configured_with_convert_underline_option():
 
 
 @istest
+def underline_can_be_configured_with_style_mapping():
+    with open(test_path("underline.docx"), "rb") as fileobj:
+        result = mammoth.convert_to_html(fileobj=fileobj, style_map="u => em")
+        assert_equal('<p><strong>The </strong><strong><em>Sunset</em></strong><strong> Tree</strong></p>', result.value)
+
+
+@istest
 def transform_document_is_applied_to_document_before_conversion():
     def transform_document(document):
         document.children[0].style_id = "Heading1"
diff --git a/tests/style_reader_tests.py b/tests/style_reader_tests.py
index 464cb4a..501629a 100644
--- a/tests/style_reader_tests.py
+++ b/tests/style_reader_tests.py
@@ -134,6 +134,13 @@ class ReadDocumentMatcherTests(object):
             read_document_matcher("r[style-name='Emphasis']")
         )
 
+    @istest
+    def reads_underline(self):
+        assert_equal(
+            document_matchers.underline,
+            read_document_matcher("u")
+        )
+
 
 @istest
 def document_matcher_is_mapped_to_html_path_using_fat_arrow():