Handle arbitrary number of backslashes during string normalization (#110)

diff --git a/black.py b/black.py
index 2ee7634c80b151e0e832a883ba7b55fa5c5b0643..9b144edb579d923d7479bdae8565f5df7507fd81 100644 (file)
--- a/black.py
+++ b/black.py
@@ -10,6 +10,7 @@ import logging
 from multiprocessing import Manager
 import os
 from pathlib import Path
 from multiprocessing import Manager
 import os
 from pathlib import Path

+import re

 import tokenize
 import signal
 import sys
 import tokenize
 import signal
 import sys


@@ -1922,8 +1923,10 @@ def normalize_string_quotes(leaf: Leaf) -> None:
 
     prefix = leaf.value[:first_quote_pos]
     body = leaf.value[first_quote_pos + len(orig_quote):-len(orig_quote)]
 
     prefix = leaf.value[:first_quote_pos]
     body = leaf.value[first_quote_pos + len(orig_quote):-len(orig_quote)]

+    unescaped_new_quote = re.compile(r"(([^\\]|^)(\\\\)*)" + new_quote)
+    escaped_orig_quote = re.compile(r"\\(\\\\)*" + orig_quote)

     if "r" in prefix.casefold():
     if "r" in prefix.casefold():

-        if body.count(new_quote) != body.count(f"\\{new_quote}"):
+        if unescaped_new_quote.search(body):

             # There's at least one unescaped new_quote in this raw string
             # so converting is impossible
             return
             # There's at least one unescaped new_quote in this raw string
             # so converting is impossible
             return


@@ -1931,9 +1934,8 @@ 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 = body.replace(f"\\{orig_quote}", orig_quote).replace(
-            new_quote, f"\\{new_quote}"
-        )
+        new_body = escaped_orig_quote.sub(f"\\1{orig_quote}", body)
+        new_body = unescaped_new_quote.sub(f"\\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] + '\\"'

diff --git a/tests/string_quotes.py b/tests/string_quotes.py
index f2d720bbdc6520bd889d92c47aed3a7f2d3a0975..c66b6ee8fbd17bdf08b608f21b1d6718832abfd9 100644 (file)
--- a/tests/string_quotes.py
+++ b/tests/string_quotes.py
@@ -1,3 +1,7 @@
+'\''
+'"'
+"'"
+"\""

 "Hello"
 "Don't do that"
 'Here is a "'
 "Hello"
 "Don't do that"
 'Here is a "'


@@ -18,9 +22,20 @@ r"raw string ftw"
 r'Date d\'expiration:(.*)'
 r'Tricky "quote'
 r'Not-so-tricky \"quote'
 r'Date d\'expiration:(.*)'
 r'Tricky "quote'
 r'Not-so-tricky \"quote'

+'\n\
+The \"quick\"\n\
+brown fox\n\
+jumps over\n\
+the \'lazy\' dog.\n\
+'
+re.compile(r'[\\"]')

 
 # output
 
 
 # output
 

+"'"
+'"'
+"'"
+'"'

 "Hello"
 "Don't do that"
 'Here is a "'
 "Hello"
 "Don't do that"
 'Here is a "'


@@ -41,3 +56,10 @@ r"raw string ftw"
 r"Date d\'expiration:(.*)"
 r'Tricky "quote'
 r"Not-so-tricky \"quote"
 r"Date d\'expiration:(.*)"
 r'Tricky "quote'
 r"Not-so-tricky \"quote"

+"\n\
+The \"quick\"\n\
+brown fox\n\
+jumps over\n\
+the 'lazy' dog.\n\
+"
+re.compile(r'[\\"]')