| 1 |
.. title:: clang-tidy - bugprone-signed-char-misuse |
| 2 |
|
| 3 |
bugprone-signed-char-misuse |
| 4 |
=========================== |
| 5 |
|
| 6 |
`cert-str34-c` redirects here as an alias for this check. For the CERT alias, |
| 7 |
the `DiagnoseSignedUnsignedCharComparisons` option is set to `false`. |
| 8 |
|
| 9 |
Finds those ``signed char`` -> integer conversions which might indicate a |
| 10 |
programming error. The basic problem with the ``signed char``, that it might |
| 11 |
store the non-ASCII characters as negative values. This behavior can cause a |
| 12 |
misunderstanding of the written code both when an explicit and when an |
| 13 |
implicit conversion happens. |
| 14 |
|
| 15 |
When the code contains an explicit ``signed char`` -> integer conversion, the |
| 16 |
human programmer probably expects that the converted value matches with the |
| 17 |
character code (a value from [0..255]), however, the actual value is in |
| 18 |
[-128..127] interval. To avoid this kind of misinterpretation, the desired way |
| 19 |
of converting from a ``signed char`` to an integer value is converting to |
| 20 |
``unsigned char`` first, which stores all the characters in the positive [0..255] |
| 21 |
interval which matches the known character codes. |
| 22 |
|
| 23 |
In case of implicit conversion, the programmer might not actually be aware |
| 24 |
that a conversion happened and char value is used as an integer. There are |
| 25 |
some use cases when this unawareness might lead to a functionally imperfect code. |
| 26 |
For example, checking the equality of a ``signed char`` and an ``unsigned char`` |
| 27 |
variable is something we should avoid in C++ code. During this comparison, |
| 28 |
the two variables are converted to integers which have different value ranges. |
| 29 |
For ``signed char``, the non-ASCII characters are stored as a value in [-128..-1] |
| 30 |
interval, while the same characters are stored in the [128..255] interval for |
| 31 |
an ``unsigned char``. |
| 32 |
|
| 33 |
It depends on the actual platform whether plain ``char`` is handled as ``signed char`` |
| 34 |
by default and so it is caught by this check or not. To change the default behavior |
| 35 |
you can use ``-funsigned-char`` and ``-fsigned-char`` compilation options. |
| 36 |
|
| 37 |
Currently, this check warns in the following cases: |
| 38 |
- ``signed char`` is assigned to an integer variable |
| 39 |
- ``signed char`` and ``unsigned char`` are compared with equality/inequality operator |
| 40 |
- ``signed char`` is converted to an integer in the array subscript |
| 41 |
|
| 42 |
See also: |
| 43 |
`STR34-C. Cast characters to unsigned char before converting to larger integer sizes |
| 44 |
<https://wiki.sei.cmu.edu/confluence/display/c/STR34-C.+Cast+characters+to+unsigned+char+before+converting+to+larger+integer+sizes>`_ |
| 45 |
|
| 46 |
A good example from the CERT description when a ``char`` variable is used to |
| 47 |
read from a file that might contain non-ASCII characters. The problem comes |
| 48 |
up when the code uses the ``-1`` integer value as EOF, while the 255 character |
| 49 |
code is also stored as ``-1`` in two's complement form of char type. |
| 50 |
See a simple example of this below. This code stops not only when it reaches |
| 51 |
the end of the file, but also when it gets a character with the 255 code. |
| 52 |
|
| 53 |
.. code-block:: c++ |
| 54 |
|
| 55 |
#define EOF (-1) |
| 56 |
|
| 57 |
int read(void) { |
| 58 |
char CChar; |
| 59 |
int IChar = EOF; |
| 60 |
|
| 61 |
if (readChar(CChar)) { |
| 62 |
IChar = CChar; |
| 63 |
} |
| 64 |
return IChar; |
| 65 |
} |
| 66 |
|
| 67 |
A proper way to fix the code above is converting the ``char`` variable to |
| 68 |
an ``unsigned char`` value first. |
| 69 |
|
| 70 |
.. code-block:: c++ |
| 71 |
|
| 72 |
#define EOF (-1) |
| 73 |
|
| 74 |
int read(void) { |
| 75 |
char CChar; |
| 76 |
int IChar = EOF; |
| 77 |
|
| 78 |
if (readChar(CChar)) { |
| 79 |
IChar = static_cast<unsigned char>(CChar); |
| 80 |
} |
| 81 |
return IChar; |
| 82 |
} |
| 83 |
|
| 84 |
Another use case is checking the equality of two ``char`` variables with |
| 85 |
different signedness. Inside the non-ASCII value range this comparison between |
| 86 |
a ``signed char`` and an ``unsigned char`` always returns ``false``. |
| 87 |
|
| 88 |
.. code-block:: c++ |
| 89 |
|
| 90 |
bool compare(signed char SChar, unsigned char USChar) { |
| 91 |
if (SChar == USChar) |
| 92 |
return true; |
| 93 |
return false; |
| 94 |
} |
| 95 |
|
| 96 |
The easiest way to fix this kind of comparison is casting one of the arguments, |
| 97 |
so both arguments will have the same type. |
| 98 |
|
| 99 |
.. code-block:: c++ |
| 100 |
|
| 101 |
bool compare(signed char SChar, unsigned char USChar) { |
| 102 |
if (static_cast<unsigned char>(SChar) == USChar) |
| 103 |
return true; |
| 104 |
return false; |
| 105 |
} |
| 106 |
|
| 107 |
.. option:: CharTypdefsToIgnore |
| 108 |
|
| 109 |
A semicolon-separated list of typedef names. In this list, we can list |
| 110 |
typedefs for ``char`` or ``signed char``, which will be ignored by the |
| 111 |
check. This is useful when a typedef introduces an integer alias like |
| 112 |
``sal_Int8`` or ``int8_t``. In this case, human misinterpretation is not |
| 113 |
an issue. |
| 114 |
|
| 115 |
.. option:: DiagnoseSignedUnsignedCharComparisons |
| 116 |
|
| 117 |
When `true`, the check will warn on ``signed char``/``unsigned char`` comparisons, |
| 118 |
otherwise these comparisons are ignored. By default, this option is set to `true`. |