Handle unnecessarily escaped strings (#128)

diff --git a/README.md b/README.md
index 65eda2accff3fd7ac27cb8b0a5d4e303898e381d..cff4bdf59dbae0f0080890cc8a42ff2c48231bd3 100644 (file)
--- a/README.md
+++ b/README.md
@@ -499,6 +499,9 @@ More details can be found in [CONTRIBUTING](CONTRIBUTING.md).
 
 * Vim plugin now works on Windows, too
 
 
 * Vim plugin now works on Windows, too
 

+* fixed unstable formatting when encountering unnecessarily escaped quotes
+  in a string (#120)
+

 
 ### 18.4a1
 
 
 ### 18.4a1
 

diff --git a/black.py b/black.py
index 587d9b35acc48642a9981423e6a81a10bddc3c37..ccc1e942cc8ae40071ed89c3533d8269c4c4436b 100644 (file)
--- a/black.py
+++ b/black.py
@@ -24,6 +24,7 @@ from typing import (
     Iterator,
     List,
     Optional,
     Iterator,
     List,
     Optional,

+    Pattern,

     Set,
     Tuple,
     Type,
     Set,
     Tuple,
     Type,


@@ -1984,9 +1985,10 @@ def normalize_string_quotes(leaf: Leaf) -> None:
         return  # There's an internal error
 
     prefix = leaf.value[:first_quote_pos]
         return  # There's an internal error
 
     prefix = leaf.value[:first_quote_pos]

-    body = leaf.value[first_quote_pos + len(orig_quote):-len(orig_quote)]

     unescaped_new_quote = re.compile(rf"(([^\\]|^)(\\\\)*){new_quote}")
     unescaped_new_quote = re.compile(rf"(([^\\]|^)(\\\\)*){new_quote}")

-    escaped_orig_quote = re.compile(rf"\\(\\\\)*{orig_quote}")
+    escaped_new_quote = re.compile(rf"([^\\]|^)\\(\\\\)*{new_quote}")
+    escaped_orig_quote = re.compile(rf"([^\\]|^)\\(\\\\)*{orig_quote}")
+    body = leaf.value[first_quote_pos + len(orig_quote):-len(orig_quote)]

     if "r" in prefix.casefold():
         if unescaped_new_quote.search(body):
             # There's at least one unescaped new_quote in this raw string
     if "r" in prefix.casefold():
         if unescaped_new_quote.search(body):
             # There's at least one unescaped new_quote in this raw string


@@ -1996,11 +1998,14 @@ def normalize_string_quotes(leaf: Leaf) -> None:
         # Do not introduce or remove backslashes in raw strings
         new_body = body
     else:
         # Do not introduce or remove backslashes in raw strings
         new_body = body
     else:

-        new_body = escaped_orig_quote.sub(rf"\1{orig_quote}", body)
-        new_body = unescaped_new_quote.sub(rf"\1\\{new_quote}", new_body)
-        # Add escapes again for consecutive occurences of new_quote (sub
-        # doesn't match overlapping substrings).
-        new_body = unescaped_new_quote.sub(rf"\1\\{new_quote}", new_body)
+        # remove unnecessary quotes
+        new_body = sub_twice(escaped_new_quote, rf"\1\2{new_quote}", body)
+        if body != new_body:
+            # Consider the string without unnecessary quotes as the original
+            body = new_body
+            leaf.value = f"{prefix}{orig_quote}{body}{orig_quote}"
+        new_body = sub_twice(escaped_orig_quote, rf"\1\2{orig_quote}", new_body)
+        new_body = sub_twice(unescaped_new_quote, rf"\1\\{new_quote}", new_body)

     if new_quote == '"""' and new_body[-1] == '"':
         # edge case:
         new_body = new_body[:-1] + '\\"'
     if new_quote == '"""' and new_body[-1] == '"':
         # edge case:
         new_body = new_body[:-1] + '\\"'


@@ -2374,5 +2379,14 @@ def shutdown(loop: BaseEventLoop) -> None:
         loop.close()
 
 
         loop.close()
 
 

+def sub_twice(regex: Pattern[str], replacement: str, original: str) -> str:
+    """Replace `regex` with `replacement` twice on `original`.
+
+    This is used by string normalization to perform replaces on
+    overlapping matches.
+    """
+    return regex.sub(replacement, regex.sub(replacement, original))
+
+

 if __name__ == "__main__":
     main()
 if __name__ == "__main__":
     main()

diff --git a/docs/reference/reference_functions.rst b/docs/reference/reference_functions.rst
index 2b8c08e0042a94505b9822a2b9314fef7bcc10a5..19128ba7098bb2226d81a4e7026b41318a6150cf 100644 (file)
--- a/docs/reference/reference_functions.rst
+++ b/docs/reference/reference_functions.rst
@@ -91,4 +91,6 @@ Utilities
 
 .. autofunction:: black.preceding_leaf
 
 
 .. autofunction:: black.preceding_leaf
 

+.. autofunction:: black.sub_twice
+

 .. autofunction:: black.whitespace
 .. autofunction:: black.whitespace

diff --git a/tests/string_quotes.py b/tests/string_quotes.py
index 8ccd041b1aac5ff9f569d5d5dcdc1177bc7dc9ec..1ac6b06a56e4249b79c00c98a7e060a6f08bafb9 100644 (file)
--- a/tests/string_quotes.py
+++ b/tests/string_quotes.py
@@ -38,6 +38,10 @@ re.compile(r'[\\"]')
 "x = '''; y = \"\"\"\""
 "x = ''''; y = \"\"\"\"\""
 "x = '' ''; y = \"\"\"\"\""
 "x = '''; y = \"\"\"\""
 "x = ''''; y = \"\"\"\"\""
 "x = '' ''; y = \"\"\"\"\""

+'unnecessary \"\"escaping'
+"unnecessary \'\'escaping"
+'\\""'
+"\\''"

 
 # output
 
 
 # output
 


@@ -81,3 +85,7 @@ re.compile(r'[\\"]')
 'x = \'\'\'; y = """"'
 'x = \'\'\'\'; y = """""'
 'x = \'\' \'\'; y = """""'
 'x = \'\'\'; y = """"'
 'x = \'\'\'\'; y = """""'
 'x = \'\' \'\'; y = """""'

+'unnecessary ""escaping'
+"unnecessary ''escaping"
+'\\""'
+"\\''"