Use ascii encoding for char strings in string convenience functions

ssbssa · Dec 3, 2023 · 1256c3a · 1256c3a
1 parent 1769cbc
commit 1256c3a
Show file tree

Hide file tree

Showing 3 changed files with 31 additions and 4 deletions.
diff --git a/gdb/python/lib/gdb/function/strfns.py b/gdb/python/lib/gdb/function/strfns.py
@@ -44,6 +44,17 @@ def invoke(self, a, b, length):
         return a_ptr.dereference() == b_ptr.dereference()
 
 
+def char_as_ascii_string(a):
+    t = a.type.strip_typedefs()
+    if (
+        t.code in [gdb.TYPE_CODE_ARRAY, gdb.TYPE_CODE_PTR]
+        and t.target().strip_typedefs().name == "char"
+    ):
+        return a.string(encoding="ascii", errors="surrogateescape")
+    else:
+        return a.string(errors="surrogateescape")
+
+
 class _StrLen(gdb.Function):
     """$_strlen - compute string length.
 
@@ -56,7 +67,7 @@ def __init__(self):
         super(_StrLen, self).__init__("_strlen")
 
     def invoke(self, a):
-        s = a.string()
+        s = char_as_ascii_string(a)
         return len(s)
 
 
@@ -76,7 +87,7 @@ def __init__(self):
         super(_StrEq, self).__init__("_streq")
 
     def invoke(self, a, b):
-        return a.string() == b.string()
+        return char_as_ascii_string(a) == char_as_ascii_string(b)
 
 
 class _RegEx(gdb.Function):
@@ -92,8 +103,8 @@ def __init__(self):
         super(_RegEx, self).__init__("_regex")
 
     def invoke(self, string, regex):
-        s = string.string()
-        r = re.compile(regex.string())
+        s = char_as_ascii_string(string)
+        r = re.compile(char_as_ascii_string(regex))
         return bool(r.match(s))
 
 

diff --git a/gdb/testsuite/gdb.python/py-strfns.c b/gdb/testsuite/gdb.python/py-strfns.c
@@ -18,6 +18,8 @@
 const char str1[] = "Hello.";
 const char str2[] = "Hello.";
 const char str3[] = "Goodbye.";
+const char str4[] = "\xed\x95\x9c";
+const char str5[] = "\xff";
 
 const char buf1[] = { 0, 1, 2, 3 };
 const char buf2[] = { 0, 1, 2, 3 };

diff --git a/gdb/testsuite/gdb.python/py-strfns.exp b/gdb/testsuite/gdb.python/py-strfns.exp
@@ -44,6 +44,20 @@ proc test_all_strfns { } {
     gdb_test {p $_regex (str1, "Help")} " = 0"
     gdb_test {p $_regex (str1, "^Hello")} " = 1"
     gdb_test {p $_regex (str1, "^Hello.$")} " = 1"
+
+    foreach_with_prefix charset {CP1252 UTF-8 ASCII} {
+	gdb_test_no_output "set target-charset $charset"
+
+	gdb_test "p \$_strlen (str4)" " = 3"
+	gdb_test "p \$_strlen (str5)" " = 1"
+
+	gdb_test {p $_streq (str1, str4)} " = 0"
+	gdb_test {p $_streq (str5, "test")} " = 0"
+	gdb_test {p $_streq (str5, "\xff")} " = 1"
+	gdb_test {p $_streq (str1, "\xff")} " = 0"
+
+	gdb_test {p $_regex (str4, ".\x95")} " = 1"
+    }
 }
 
 test_all_strfns