From e2d510d2539f034265fae61cd27d2b464bb5792d Mon Sep 17 00:00:00 2001
From: dzhuang <dzhuang.scut@gmail.com>
Date: Fri, 23 Feb 2018 23:08:12 +0800
Subject: [PATCH] Removed lxml dependency for ipynb utility.
---
course/content.py | 17 ++---------------
.../course/jinja2/nbconvert_template.tpl | 2 +-
course/utils.py | 17 +++++++++++++++++
requirements.txt | 3 ---
setup.py | 1 -
tests/test_content.py | 2 ++
6 files changed, 22 insertions(+), 20 deletions(-)
diff --git a/course/content.py b/course/content.py
index 2a31b47f..f4b51531 100644
--- a/course/content.py
+++ b/course/content.py
@@ -912,18 +912,6 @@ def expand_markup(
return text
-def unwrap_relate_tmp_pre_tag(html_string):
- # type: (Text) -> (Text)
-
- from lxml.html import fromstring, tostring
- tree = fromstring(html_string)
-
- for node in tree.iterdescendants("pre"):
- if "relate_tmp_pre" in node.attrib.get("class", ""):
- node.drop_tag()
- return tostring(tree, encoding="unicode")
-
-
def markup_to_html(
course, # type: Optional[Course]
repo, # type: Repo_ish
@@ -976,11 +964,13 @@ def markup_to_html(
return ""
from course.mdx_mathjax import MathJaxExtension
+ from course.utils import NBConvertExtension
import markdown
extensions = [
LinkFixerExtension(course, commit_sha, reverse_func=reverse_func),
MathJaxExtension(),
+ NBConvertExtension(),
"markdown.extensions.extra",
]
@@ -998,9 +988,6 @@ def markup_to_html(
extensions=extensions,
output_format="html5")
- if result.strip():
- result = unwrap_relate_tmp_pre_tag(result)
-
assert isinstance(result, six.text_type)
if cache_key is not None:
def_cache.add(cache_key, result, None)
diff --git a/course/templates/course/jinja2/nbconvert_template.tpl b/course/templates/course/jinja2/nbconvert_template.tpl
index f040623a..c59f9a8b 100644
--- a/course/templates/course/jinja2/nbconvert_template.tpl
+++ b/course/templates/course/jinja2/nbconvert_template.tpl
@@ -2,7 +2,7 @@
{# This is to prevent code_cell being process by markdown_to_html #}
-{% block input %}<pre class="relate_tmp_pre">{{ super() }}</pre>
+{% block input %}<pre><relate_ipynb>{{ super() }}</relate_ipynb></pre>
{%- endblock input %}
{# This is to remove the empty cells ahead of markdown_cells #}
diff --git a/course/utils.py b/course/utils.py
index 428576ba..b4ca3ea9 100644
--- a/course/utils.py
+++ b/course/utils.py
@@ -28,6 +28,7 @@ from typing import cast
import six
import datetime # noqa
+import markdown
from django.shortcuts import ( # noqa
render, get_object_or_404)
@@ -1322,6 +1323,22 @@ class IpynbJinjaMacro(RelateJinjaMacroBase):
return body
+
+NBCONVERT_PRE_OPEN_RE = re.compile(r"<pre\s*>\s*<relate_ipynb\s*>")
+NBCONVERT_PRE_CLOSE_RE = re.compile(r"</relate_ipynb\s*>\s*</pre\s*>")
+
+
+class NBConvertHTMLPostprocessor(markdown.postprocessors.Postprocessor):
+ def run(self, text):
+ text = NBCONVERT_PRE_OPEN_RE.sub("", text)
+ text = NBCONVERT_PRE_CLOSE_RE.sub("", text)
+ return text
+
+
+class NBConvertExtension(markdown.Extension):
+ def extendMarkdown(self, md, md_globals): # noqa
+ md.postprocessors['relate_nbconvert'] = NBConvertHTMLPostprocessor(md)
+
# }}}
# vim: foldmethod=marker
diff --git a/requirements.txt b/requirements.txt
index 3dc2f8e9..62959600 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -117,7 +117,4 @@ typing
nbconvert>=5.2.1
IPython
-# For parsing and edit html
-lxml
-
# vim: foldmethod=marker
diff --git a/setup.py b/setup.py
index 286d45ec..64eee057 100644
--- a/setup.py
+++ b/setup.py
@@ -27,7 +27,6 @@ setup(name="relate-courseware",
"markdown",
"dulwich",
"pyyaml",
- "lxml",
"nbconvert>=5.2.1",
"pymbolic",
"sympy",
diff --git a/tests/test_content.py b/tests/test_content.py
index 589e9f3c..e8f37b04 100644
--- a/tests/test_content.py
+++ b/tests/test_content.py
@@ -240,6 +240,7 @@ CODE_CELL_HTML_CLASS = "code_cell"
CODE_CELL_IN_STR_PATTERN = '<div class="prompt input_prompt">In[%s]:</div>'
CODE_CELL_PRINT_STR1 = "This is function1"
CODE_CELL_PRINT_STR2 = "This is function2"
+RELATE_IPYNB_CONVERT_PRE_WRAPPER_TAG_NAME = "relate_ipynb"
def strip_nbsp(s):
@@ -261,6 +262,7 @@ class NbconvertRenderTestMixin(SingleCoursePageSandboxTestBaseMixin):
self.assertNotContains(response, "```")
self.assertNotContains(response, "# First Title of Test NoteBook")
self.assertNotContains(response, "# Second Title of Test NoteBook")
+ self.assertNotContains(response, RELATE_IPYNB_CONVERT_PRE_WRAPPER_TAG_NAME)
def setUp(self):
super(NbconvertRenderTestMixin, self).setUp()
--
GitLab