fix: Remove JavaScript from HTML reader output

CHANGELOG.md

@@ -1,11 +1,12 @@
-## 0.5.1-dev1
+## 0.5.1-dev2
 
 ### Enhancements
 
 ### Features
 
 ### Fixes
 
+* Fixes an error causing JavaScript to appear in the output of `partition_html` sometimes.
 * Fix several issues with the `requires_dependencies` decorator, including the error message
   and how it was used.
 

unstructured/__version__.py

@@ -1 +1 @@
-__version__ = "0.5.1-dev1"  # pragma: no cover
+__version__ = "0.5.1-dev2"  # pragma: no cover

unstructured/documents/html.py

@@ -265,12 +265,11 @@ def is_narrative_tag(text: str, tag: str) -> bool:
 def _construct_text(tag_elem: etree.Element) -> str:
     """Extracts text from a text tag element."""
     text = ""
-    for item in tag_elem.itertext():
-        if item:
-            text += item
-
-    if tag_elem.tail:
-        text = text + tag_elem.tail
+    for item in tag_elem.iter():
+        if item.text and item.tag != "script":
+            text += item.text
+            if item.tail:
+                text += item.tail
 
     text = replace_unicode_quotes(text)
     return text.strip()