Merge pull request #166 from gibfahn/allow_overwriting_link_schemes

Add `accepted_url_schemes` option to allow toplevel refs

README.md

@@ -111,6 +111,9 @@ AutoStructify comes with the following options. See [http://recommonmark.readthe
 * __enable_inline_math__: enable the Inline Math feature.
 * __enable_eval_rst__: enable the evaluate embedded reStructuredText feature.
 * __url_resolver__: a function that maps a existing relative position in the document to a http link
+* __known_url_schemes__: a list of url schemes to treat as URLs, schemes not in this list will be assumed to be Sphinx cross-references.
+    Defaults to `None`, which means treat all URL schemes as URLs.
+    Example: `['http', 'https', 'mailto']`
 
 ## Development
 

recommonmark/parser.py

@@ -25,12 +25,22 @@ class CommonMarkParser(parsers.Parser):
     supported = ('md', 'markdown')
     translate_section_name = None
 
+    default_config = {
+        'known_url_schemes': None,
+    }
+
     def __init__(self):
         self._level_to_elem = {}
 
     def parse(self, inputstring, document):
         self.document = document
         self.current_node = document
+        self.config = self.default_config.copy()
+        try:
+            new_cfg = self.document.settings.env.config.recommonmark_config
+            self.config.update(new_cfg)
+        except AttributeError:
+            pass
         self.setup_parse(inputstring, document)
         self.setup_sections()
         parser = Parser()
@@ -153,7 +163,16 @@ class CommonMarkParser(parsers.Parser):
         next_node = ref_node
 
         url_check = urlparse(destination)
-        if not url_check.scheme and not url_check.fragment:
+        # If there's not a url scheme (e.g. 'https' for 'https:...' links),
+        # or there is a scheme but it's not in the list of known_url_schemes,
+        # then assume it's a cross-reference and pass it to Sphinx as an `:any:` ref.
+        known_url_schemes = self.config.get('known_url_schemes')
+        if known_url_schemes:
+            scheme_known = url_check.scheme in known_url_schemes
+        else:
+            scheme_known = bool(url_check.scheme)
+
+        if not url_check.fragment and not scheme_known:
             wrap_node = addnodes.pending_xref(
                 reftarget=unquote(destination),
                 reftype='any',

recommonmark/transform.py

@@ -48,6 +48,7 @@ class AutoStructify(transforms.Transform):
         'enable_inline_math': True,
         'commonmark_suffixes': ['.md'],
         'url_resolver': lambda x: x,
+        'known_url_schemes': None,
     }
 
     def parse_ref(self, ref):

tests/test_basic.py

@@ -17,6 +17,7 @@ class TestParsing(unittest.TestCase):
     def assertParses(self, source, expected, alt=False):  # noqa
         parser = CommonMarkParser()
         parser.parse(dedent(source), new_document('<string>'))
+        self.maxDiff = None
         self.assertMultiLineEqual(
             dedent(expected).lstrip(),
             dedent(parser.document.asdom().toprettyxml(indent='  ')),
@@ -223,6 +224,46 @@ class TestParsing(unittest.TestCase):
             """
         )
 
+    def test_known_schemes(self):
+        self.assertParses(
+            """
+            [https link](https://example.com)
+            [http link](http://example.com)
+            [mailto link](mailto:admin@example.com)
+            [custom scheme](custom:example.com)
+            [ref link](path/to/file:heading)
+            [ref link with spaces](<path/to/file:heading with spaces>)
+            """,
+            """
+            <?xml version="1.0" ?>
+            <document source="&lt;string&gt;">
+              <paragraph>
+                <reference refuri="https://example.com">https link</reference>
+
+
+                <reference refuri="http://example.com">http link</reference>
+
+
+                <reference refuri="mailto:admin@example.com">mailto link</reference>
+
+
+                <reference refuri="custom:example.com">custom scheme</reference>
+
+
+                <pending_xref refdomain="None" refexplicit="True" reftarget="path/to/file:heading" reftype="any" refwarn="True">
+                  <reference refuri="path/to/file:heading">ref link</reference>
+                </pending_xref>
+
+
+                <pending_xref refdomain="None" refexplicit="True" reftarget="path/to/file:heading with spaces" reftype="any" refwarn="True">
+                  <reference refuri="path/to/file:heading%20with%20spaces">ref link with spaces</reference>
+                </pending_xref>
+              </paragraph>
+            </document>
+            """
+        )
+        pass
+
     def test_image(self):
         self.assertParses(
             """