Fix handling &lrm; and &rlm; marks

- mid-text within stressed tags - right after stressed tags Fixes Alir3z4#201.
Unit03 · Jul 15, 2018 · 802fad1 · 802fad1
1 parent bef7ec0
commit 802fad1
Show file tree

Hide file tree

Showing 8 changed files with 19 additions and 1 deletion.
diff --git a/ChangeLog.rst b/ChangeLog.rst
@@ -1,3 +1,10 @@
+0000.00.00
+==========
+----
+
+* Fix #201: handle &lrm;/&rlm; marks mid-text within stressed tags or right after stressed tags
+
+
 2018.9.1
 ========
 ----

diff --git a/html2text/__init__.py b/html2text/__init__.py
@@ -186,7 +186,9 @@ def handle_charref(self, c):
         self.handle_data(self.charref(c), True)
 
     def handle_entityref(self, c):
-        self.handle_data(self.entityref(c), True)
+        ref = self.entityref(c)
+        if ref:
+            self.handle_data(ref, True)
 
     def handle_starttag(self, tag, attrs):
         self.handle_tag(tag, attrs, 1)

diff --git a/test/lrm_after_i.html b/test/lrm_after_i.html
@@ -0,0 +1 @@
+<i>Foo</i>&lrm;
diff --git a/test/lrm_after_i.md b/test/lrm_after_i.md
@@ -0,0 +1,2 @@
+_Foo_
+
diff --git a/test/lrm_inside_i.html b/test/lrm_inside_i.html
@@ -0,0 +1 @@
+<i>Foo&lrm;bar</i>
diff --git a/test/lrm_inside_i.md b/test/lrm_inside_i.md
@@ -0,0 +1,2 @@
+_Foo bar_
+
diff --git a/test/rlm_inside_strong.html b/test/rlm_inside_strong.html
@@ -0,0 +1 @@
+<strong>Foo&rlm;bar</strong>
diff --git a/test/rlm_inside_strong.md b/test/rlm_inside_strong.md
@@ -0,0 +1,2 @@
+**Foo bar**
+