Handle attached leading and trailing characters of links (fixes #55) (#…

…63)

markflow/_utils/textwrap.py

@@ -4,9 +4,9 @@
 from markflow.typing import Number
 
 INLINE_CODE_MARKER_REGEX = re.compile(r"(((?!<\\)`)+)")
-FOOTNOTE_REGEX = re.compile(r"\[[^\[]+\]\[[^\]]+\]")
+FOOTNOTE_REGEX = re.compile(r"[^\s\]\)]*\[[^\[]+\]\[[^\]]+\][^\s\[\(]*")
 HTML_NEWLINE_REGEX = re.compile(r"<br ?/?>")
-URL_REGEX = re.compile(r"\[[^\[]+\]\([^\)]+\)")
+URL_REGEX = re.compile(r"[^\s\]\)]*\[[^\[]+\]\([^\)]+\)[^\s\[\(]*")
 
 
 def join(split_text: List[str], leading_spaces: List[bool], width: Number) -> str:
@@ -134,7 +134,7 @@ def link_split(
     leading_spaces: List[bool] = []
     evaluates: List[bool] = []
     last_end = 0
-    # Each iteration of this for loop operates operates on non-link text followed by
+    # Each iteration of this for loop operates on non-link text followed by
     # link text.
     for match in matches:
         non_link_text = text[last_end : match.start()]

tests/files/0018_in_urls_with_trailing_characters.md

@@ -0,0 +1,9 @@
+Ever since I sold the first MacBook, it felt like I was fighting Windows audio drivers
+constantly. Between latency issues, BSODs due to Focusrite's unstable audio interface
+drivers, more BSODs trying to get around [multiple audio interface setups in software](
+https://help.ableton.com/hc/en-us/articles/209071609-Using-Aggregate-Devices-and-multiple-audio-interfaces),
+I started steering away from in-the-box production and started looking into hardware.
+
+[This is a really long URL](http://www.example.com/ssssssssssssssssssssssssssssssssssss),
+
+[This is a really long footnote][fffffffffffffffffffffffffffffffffffffffffffffffffffffffff],

tests/files/0018_out_urls_with_trailing_characters.md

@@ -0,0 +1,11 @@
+Ever since I sold the first MacBook, it felt like I was fighting Windows audio drivers
+constantly. Between latency issues, BSODs due to Focusrite's unstable audio interface
+drivers, more BSODs trying to get around [multiple audio interface setups in software](
+https://help.ableton.com/hc/en-us/articles/209071609-Using-Aggregate-Devices-and-multiple-audio-interfaces),
+I started steering away from in-the-box production and started looking into hardware.
+
+[This is a really long URL](
+http://www.example.com/ssssssssssssssssssssssssssssssssssss),
+
+[This is a really long footnote][
+fffffffffffffffffffffffffffffffffffffffffffffffffffffffff],

tests/test_utils.py

@@ -43,7 +43,7 @@ def test_all_splits(self) -> None:
             abc abc abc abc abc abc abc abc abc
             ``abc ``` abc[0][0] ``abc abc abc abc <br />
             <br />
-            abc abc [url](http://example.com) abc[url][
+            abc abc [url](http://example.com)abc[url][
             http://example.com]abc[url][URL][url][URL] <br/>
             abc<br/>"""
         )
@@ -96,16 +96,14 @@ def test_code_split_solo_tilda(self) -> None:
     def test_link_split(self) -> None:
         input_ = "a[URL][url] b [URL](http://example.com)c"
         expected_split_text = [
-            "a",
-            "[URL][",
+            "a[URL][",
             "url]",
             "b",
             "[URL](",
-            "http://example.com)",
-            "c",
+            "http://example.com)c",
         ]
-        expected_leading_spaces = [False, False, False, True, True, False, False]
-        expected_evaluates = [True, True, False, True, True, False, True]
+        expected_leading_spaces = [False, False, True, True, False]
+        expected_evaluates = [True, False, True, True, False]
         split_text, leading_spaces, evaluates = link_split(input_, False)
         assert len(split_text) == len(leading_spaces) == len(evaluates)
         assert split_text == expected_split_text
@@ -115,16 +113,14 @@ def test_link_split(self) -> None:
     def test_link_split_sentence(self) -> None:
         input_ = "a[URL][url]. b [URL](http://example.com).c"
         expected_split_text = [
-            "a",
-            "[URL][",
+            "a[URL][",
             "url].",
             "b",
             "[URL](",
-            "http://example.com).",
-            "c",
+            "http://example.com).c",
         ]
-        expected_leading_spaces = [False, False, False, True, True, False, False]
-        expected_evaluates = [True, True, False, True, True, False, True]
+        expected_leading_spaces = [False, False, True, True, False]
+        expected_evaluates = [True, False, True, True, False]
         split_text, leading_spaces, evaluates = link_split(input_, False)
         assert len(split_text) == len(leading_spaces) == len(evaluates)
         assert split_text == expected_split_text