Merge pull request #41 from oreillymedia/bug-fix-rm-spans-pre-class

Bug fix for platform surprise class styling

Author: delfanbaum

README.md

@@ -72,6 +72,11 @@ Options:
 
 ## Release Notes
 
+### 1.0.9
+
+Bug fixes:
+- Remove spans inside `<code>` tags that display incorrectly on the ORM learning platform.
+
 ### 1.0.8
 
 Bug fixes:

jupyter_book_to_htmlbook/code_processing.py

@@ -60,6 +60,25 @@ def process_code(chapter, skip_numbering=False):
     return chapter
 
 
+def process_inline_code(chapter):
+    """
+    Because the platform occasionally has unexpected class styling, we want to
+    make sure that inline code in particular is clean. Note that the current
+    target version of Jupyter Book doesn't put <code> inside <pre> tags, so our
+    "dumb" inline searcher should work. There is a test to confirm this.
+
+    NOTE: This is a temporary fix; we should really clean out all classes
+    except those we explicitly want from the chapter files
+    """
+    inline_codes = chapter.find_all("code")
+
+    for code in inline_codes:
+        if code.find("span"):
+            code.span.unwrap()
+
+    return chapter
+
+
 def number_codeblock(pre_block, cell_number):
     """
     Adds numbering markers (`In [##]: ` or `Out[##]: `) to cell blocks

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "jupyter-book-to-htmlbook"
-version = "1.0.8"
+version = "1.0.9"
 description = "A script to convert jupyter book html files to htmlbook for consumption in Atlas"
 authors = ["delfanbaum"]
 

tests/test_code_processing.py

@@ -5,6 +5,7 @@
 from bs4 import BeautifulSoup  # type: ignore
 from jupyter_book_to_htmlbook.code_processing import (
         process_code,
+        process_inline_code,
         number_codeblock,
         process_code_examples
     )
@@ -557,3 +558,29 @@ def test_examples_malformed_r(self, caplog):
         assert not example_div.find("h5", string="A formal R example")
         assert result == malformed_example
         assert "Missing first two line comments for" in log
+
+
+class TestInlineCode:
+    """
+    Smoke tests around the translation of inline code
+    """
+
+    def test_unwrap_inline_spans(self):
+        """
+        We should not allow spans inside inline code
+        """
+
+        html = BeautifulSoup("""<p>Some text with
+<code><span class="pre">code</span></code>.</p>""",
+                             "html.parser")
+        result = process_inline_code(html)
+
+        assert not result.find("span")
+        assert str(result.find("code")) == "<code>code</code>"
+
+    def test_unwrap_inline_spans_does_not_affect_pre(self,
+                                                     code_example_data_type):
+        expected = str(code_example_data_type)
+        result = process_inline_code(code_example_data_type)
+        assert str(result) == expected
+