utils/html: Remove the hard-coded list of allowed elements and attrib…

…utes

These changes provide full control over the management of "allowed-elements" and "allowed-attributes" through the configuration file.

Fixes #751

CHANGES.rst

@@ -16,6 +16,11 @@ New Features
 Breaking Changes
 ^^^^^^^^^^^^^^^^
 
+- Provide full control of allowed-elements and allowed-attributes via the configuration
+  file (`#1007`_, pkvach)
+  Now the configuration options ``allowed-elements`` and ``allowed-attributes`` are not additional.
+  That is, they specify the full list of allowed elements and attributes.
+
 - TBD
 
 Bugfixes & Improvements
@@ -27,13 +32,16 @@ Bugfixes & Improvements
 - Prevent auto creation of invalid links in comments (`#995`_, pkvach)
 - Fix W3C Validation issues (`#999`_, pkvach)
 - Handle deleted comments in Disqus migration (`#994`_, pkvach)
+- Provide full control of allowed-elements and allowed-attributes via the configuration
+  file (`#1007`_, pkvach)
 
 .. _#951: https://github.com/posativ/isso/pull/951
 .. _#967: https://github.com/posativ/isso/pull/967
 .. _#983: https://github.com/posativ/isso/pull/983
 .. _#995: https://github.com/isso-comments/isso/pull/995
 .. _#999: https://github.com/isso-comments/isso/pull/999
 .. _#994: https://github.com/isso-comments/isso/pull/994
+.. _#1007: https://github.com/isso-comments/isso/pull/1007
 
 0.13.1.dev0 (2023-02-05)
 ------------------------

contrib/isso-dev.cfg

@@ -37,8 +37,9 @@ reply-to-self = true
 [markup]
 options = autolink, fenced-code, no-intra-emphasis, strikethrough, superscript
 flags =
-allowed-elements =
-allowed-attributes =
+allowed-elements = a, p, hr, br, ol, ul, li, pre, code, blockquote, del, ins,
+   strong, em, h1, h2, h3, h4, h5, h6, sub, sup, table, thead, tbody, th, td
+allowed-attributes = align, href
 
 [hash]
 salt = Eech7co8Ohloopo9Ol6baimi

docs/docs/reference/server-config.rst

@@ -428,37 +428,19 @@ flags
     .. versionadded:: 0.12.4
 
 allowed-elements
-    **Additional** HTML tags to allow in the generated output, comma-separated.
-
-    By default, only ``a``, ``blockquote``, ``br``, ``code``, ``del``, ``em``,
-    ``h1``, ``h2``, ``h3``, ``h4``, ``h5``, ``h6``, ``hr``, ``ins``, ``li``,
-    ``ol``, ``p``, ``pre``, ``strong``, ``table``, ``tbody``, ``td``, ``th``,
-    ``thead`` and ``ul`` are allowed.
+    HTML tags to allow in the generated output, comma-separated.
 
     For a more detailed explanation, see :doc:`/docs/reference/markdown-config`.
 
-    .. warning::
-
-       This option (together with ``allowed-attributes``) is frequently
-       misunderstood. Setting e.g. this list to only ``a, blockquote`` will
-       mean that ``br, code, del, ...`` and all other default allowed tags are
-       still allowed. You can only add *additional* elements here.
-
-       It is planned to change this behavior, see
-       `this issue <https://github.com/isso-comments/isso/issues/751>`_.
-
-    Default: (empty)
+    Default: ``a, p, hr, br, ol, ul, li, pre, code, blockquote, del, ins, strong, em, h1, h2, h3, h4, h5, h6, sub, sup, table, thead, tbody, th, td``
 
 allowed-attributes
-    **Additional** HTML attributes (independent from elements) to allow in the
+    HTML attributes (independent from elements) to allow in the
     generated output, comma-separated.
 
-    By default, only ``align`` and ``href`` are allowed (same caveats as for
-    ``allowed-elements`` above apply)
-
     For a more detailed explanation, see :doc:`/docs/reference/markdown-config`.
 
-    Default: (empty)
+    Default: ``align, href``
 
 .. note:: To allow images in comments, you need to add
    ``allowed-elements = img`` and *also* ``allowed-attributes = src``.

isso/isso.cfg

@@ -211,15 +211,13 @@ options = autolink, fenced-code, no-intra-emphasis, strikethrough, superscript
 # Per Misaka's defaults, no flags are set.
 flags =
 
-# Additional HTML tags to allow in the generated output, comma-separated. By
-# default, only a, blockquote, br, code, del, em, h1, h2, h3, h4, h5, h6, hr,
-# ins, li, ol, p, pre, strong, table, tbody, td, th, thead and ul are allowed.
-allowed-elements =
-
-# Additional HTML attributes (independent from elements) to allow in the
-# generated output, comma-separated. By default, only align and href are
-# allowed.
-allowed-attributes =
+# HTML tags to allow in the generated output, comma-separated.
+allowed-elements = a, p, hr, br, ol, ul, li, pre, code, blockquote, del, ins,
+    strong, em, h1, h2, h3, h4, h5, h6, sub, sup, table, thead, tbody, th, td
+
+# HTML attributes (independent from elements) to allow in the generated output,
+# comma-separated.
+allowed-attributes = align, href
 
 
 [hash]

isso/tests/test_html.py

@@ -60,7 +60,7 @@ def test_github_flavoured_markdown(self):
             </code></pre>""")
 
     def test_sanitizer(self):
-        sanitizer = html.Sanitizer(elements=[], attributes=[])
+        sanitizer = html.Sanitizer(elements=["p", "a", "code"], attributes=["href"])
         examples = [
             ('Look: <img src="..." />', 'Look: '),
             ('<a href="http://example.org/">Ha</a>',
@@ -94,8 +94,8 @@ def test_render(self):
             "markup": {
                 "options": "autolink",
                 "flags": "",
-                "allowed-elements": "",
-                "allowed-attributes": ""
+                "allowed-elements": "a, p",
+                "allowed-attributes": "href"
             }
         })
         renderer = html.Markup(conf.section("markup")).render
@@ -109,7 +109,7 @@ def test_sanitized_render_extensions(self):
             "markup": {
                 "options": "no_intra_emphasis",  # Deliberately snake_case
                 "flags": "",
-                "allowed-elements": "",
+                "allowed-elements": "p",
                 "allowed-attributes": ""
             }
         })

isso/utils/html.py

@@ -17,25 +17,17 @@ def allow_attribute_class(tag, name, value):
         return name == "class" and bool(Sanitizer.code_language_pattern.match(value))
 
     def __init__(self, elements, attributes):
-        # attributes found in Sundown's HTML serializer [1]
-        # - except for <img> tag, because images are not generated anyways.
-        # - sub and sup added
-        #
-        # [1] https://github.com/vmg/sundown/blob/master/html/html.c
-        self.elements = ["a", "p", "hr", "br", "ol", "ul", "li",
-                         "pre", "code", "blockquote",
-                         "del", "ins", "strong", "em",
-                         "h1", "h2", "h3", "h4", "h5", "h6", "sub", "sup",
-                         "table", "thead", "tbody", "th", "td"] + elements
+        self.elements = elements
 
         # allowed attributes for tags
         self.attributes = {
-            "table": ["align"],
-            "a": ["href"],
-            "code": Sanitizer.allow_attribute_class,
             "*": attributes
         }
 
+        # If "code" elements are allowed, allow "language-*" CSS classes for syntax highlighting
+        if "code" in self.elements:
+            self.attributes["code"] = Sanitizer.allow_attribute_class
+
     def sanitize(self, text):
         clean_html = bleach.clean(text, tags=self.elements, attributes=self.attributes, strip=True)