From 0e48ae679abc0937b4aad583b1b29ee0b3e3eb15 Mon Sep 17 00:00:00 2001
From: Numerlor <25886452+Numerlor@users.noreply.github.com>
Date: Wed, 9 Dec 2020 13:40:01 +0100
Subject: Improve handling of strings

Previously the code assumed ' and " can be used interchangeably,
and strings that were inside of brackets were ignored for depth but
their contents weren't causing strings like "ab[cd" to increase the
depth
---
 bot/exts/info/doc/_parsing.py | 35 ++++++++++++++++++++++++++++-------
 1 file changed, 28 insertions(+), 7 deletions(-)

diff --git a/bot/exts/info/doc/_parsing.py b/bot/exts/info/doc/_parsing.py
index 418405ca9..e6103dde2 100644
--- a/bot/exts/info/doc/_parsing.py
+++ b/bot/exts/info/doc/_parsing.py
@@ -56,6 +56,15 @@ _BRACKET_PAIRS = {
 }
 
 
+def _is_closing_quote(search_string: str, index: int) -> bool:
+    """Check whether the quote at `index` inside `search_string` can be a closing quote."""
+    if search_string[index - 1] != "\\":
+        return True
+    elif search_string[index - 2] == "\\":
+        return True
+    return False
+
+
 def _split_parameters(parameters_string: str) -> List[str]:
     """
     Split parameters of a signature into individual parameter strings on commas.
@@ -67,9 +76,11 @@ def _split_parameters(parameters_string: str) -> List[str]:
     depth = 0
     expected_end = None
     current_search = None
+    quote_character = None
 
-    for index, character in enumerate(parameters_string):
-        if character in _BRACKET_PAIRS:
+    enumerated_string = enumerate(parameters_string)
+    for index, character in enumerated_string:
+        if quote_character is None and character in _BRACKET_PAIRS:
             if current_search is None:
                 current_search = character
                 expected_end = _BRACKET_PAIRS[character]
@@ -77,12 +88,22 @@ def _split_parameters(parameters_string: str) -> List[str]:
                 depth += 1
 
         elif character in {"'", '"'}:
-            if depth == 0:
+            if current_search is not None:
+                # We're currently searching for a bracket, skip all characters that belong to the string
+                # to avoid false positives of closing brackets
+                quote_character = character
+                for index, character in enumerated_string:
+                    if character == quote_character and _is_closing_quote(parameters_string, index):
+                        break
+
+            elif depth == 0:
                 depth += 1
-            elif parameters_string[index-1] != "\\":
-                depth -= 1
-            elif parameters_string[index-2] == "\\":
-                depth -= 1
+                quote_character = character
+            elif character == quote_character:
+                if _is_closing_quote(parameters_string, index):
+                    depth -= 1
+                if depth == 0:
+                    quote_character = None
 
         elif character == expected_end:
             depth -= 1
-- 
cgit v1.2.3