fix: add additional URL patterns (#148)

* fix: add additional URL patterns * test: add test class and tests for URL patterns * style: fix rstcheck error
PyCQA · Jan 8, 2023 · 3c9c239 · 3c9c239
1 parent 518018f
commit 3c9c239
Show file tree

Hide file tree

Showing 3 changed files with 508 additions and 23 deletions.
diff --git a/README.rst b/README.rst
@@ -15,8 +15,12 @@ docformatter
     :target: https://pycqa.github.io/isort/
 .. |SELF| image:: https://img.shields.io/badge/%20formatter-docformatter-fedcba.svg
     :target: https://github.com/PyCQA/docformatter
-.. |DOCSTYLE| image:: https://img.shields.io/badge/%20style-numpy-459db9.svg
+.. |SPHINXSTYLE| image:: https://img.shields.io/badge/%20style-sphinx-0a507a.svg
+    :target: https://www.sphinx-doc.org/en/master/usage/index.html
+.. |NUMPSTYLE| image:: https://img.shields.io/badge/%20style-numpy-459db9.svg
     :target: https://numpydoc.readthedocs.io/en/latest/format.html
+.. |GOOGSTYLE| image:: https://img.shields.io/badge/%20style-google-3666d6.svg
+    :target: https://google.github.io/styleguide/pyguide.html#s3.8-comments-and-docstrings
 
 .. |VERSION| image:: https://img.shields.io/pypi/v/docformatter
 .. |LICENSE| image:: https://img.shields.io/pypi/l/docformatter
@@ -27,7 +31,7 @@ docformatter
 +----------------+----------------------------------------------------------+
 | **Code**       + |BLACK| |ISORT|                                          +
 +----------------+----------------------------------------------------------+
-| **Docstrings** + |SELF| |DOCSTYLE|                                        +
+| **Docstrings** + |SELF| |NUMPSTYLE|                                       +
 +----------------+----------------------------------------------------------+
 | **GitHub**     + |CI| |CONTRIBUTORS| |COMMIT|                             +
 +----------------+----------------------------------------------------------+
@@ -171,28 +175,26 @@ Do you use *docformatter*?  What style docstrings do you use?  Add some badges t
     .. image:: https://img.shields.io/badge/%20formatter-docformatter-fedcba.svg
         :target: https://github.com/PyCQA/docformatter
 
-    .. image:: https://img.shields.io/badge/%20style-google-3666d6.svg
-        :target: https://google.github.io/styleguide/pyguide.html#s3.8-comments-and-docstrings
+|SPHINXSTYLE|
 
 .. code-block::
 
-    .. image:: https://img.shields.io/badge/%20style-google-3666d6.svg
-        :target: https://google.github.io/styleguide/pyguide.html#s3.8-comments-and-docstrings
+    .. image:: https://img.shields.io/badge/%20style-sphinx-0a507a.svg
+        :target: https://www.sphinx-doc.org/en/master/usage/index.html
 
-|DOCSTYLE|
+|NUMPSTYLE|
 
 .. code-block::
 
     .. image:: https://img.shields.io/badge/%20style-numpy-459db9.svg
         :target: https://numpydoc.readthedocs.io/en/latest/format.html
 
-    .. image:: https://img.shields.io/badge/%20style-sphinx-0a507a.svg
-        :target: https://www.sphinx-doc.org/en/master/usage/index.html
+|GOOGSTYLE|
 
 .. code-block::
 
-    .. image:: https://img.shields.io/badge/%20style-sphinx-0a507a.svg
-        :target: https://www.sphinx-doc.org/en/master/usage/index.html
+    .. image:: https://img.shields.io/badge/%20style-google-3666d6.svg
+        :target: https://google.github.io/styleguide/pyguide.html#s3.8-comments-and-docstrings
 
 Issues
 ======

diff --git a/src/docformatter/syntax.py b/src/docformatter/syntax.py
@@ -107,14 +107,7 @@ def do_preserve_links(
     if len(lines) < 2:
         return lines
 
-    url = next(
-        (
-            line
-            for line in lines
-            if re.search(r"<?(http://|https://|ftp://|sftp://)", line)
-        ),
-        "",
-    )
+    url = is_some_sort_of_link(lines)
 
     if url != "":
         url_idx = lines.index(url)
@@ -165,6 +158,77 @@ def do_preserve_links(
     return lines
 
 
+def is_some_sort_of_link(lines: List[str]) -> str:
+    """Determine if docstring line contains a link.
+
+    URL patterns based on table at
+    <https://en.wikipedia.org/wiki/List_of_URI_schemes>
+
+    Parameters
+    ----------
+    lines: str
+        the list of docstring lines to check for a link pattern.
+
+    Returns
+    -------
+    url: str
+        the line with the url pattern.
+    """
+    url_patterns = (
+        "("
+        "afp://|"
+        "apt:|"
+        "bitcoin:|"
+        "chrome://|"
+        "cvs://|"
+        "dav://|"
+        "dns:|"
+        "file://|"
+        "finger://|"
+        "fish://|"
+        "ftp://|"
+        "ftps://|"
+        "git://|"
+        "http://|"
+        "https://|"
+        "imap://|"
+        "ipp://|"
+        "ipps://|"
+        "irc://|"
+        "irc6://|"
+        "ircs://|"
+        "jar:|"
+        "ldap://|"
+        "ldaps://|"
+        "mailto:|"
+        "news:|"
+        "nfs://|"
+        "nntp://|"
+        "pop://|"
+        "rsync://|"
+        "s3://|"
+        "sftp://|"
+        "shttp://|"
+        "sip:|"
+        "sips:|"
+        "smb://|"
+        "sms:|"
+        "snmp://|"
+        "ssh://|"
+        "svn://|"
+        "telnet://|"
+        "vnc://|"
+        "xmpp:|"
+        "xri://"
+        ")"
+    )
+
+    return next(
+        (line for line in lines if re.search(rf"<?{url_patterns}", line)),
+        "",
+    )
+
+
 # pylint: disable=line-too-long
 def is_some_sort_of_list(text, strict) -> bool:
     """Determine if docstring is a reST list.