diff options
Diffstat (limited to 'googletest/src/gtest-printers.cc')
-rw-r--r-- | googletest/src/gtest-printers.cc | 101 |
1 files changed, 91 insertions, 10 deletions
diff --git a/googletest/src/gtest-printers.cc b/googletest/src/gtest-printers.cc index dd67f64..de4d245 100644 --- a/googletest/src/gtest-printers.cc +++ b/googletest/src/gtest-printers.cc @@ -26,10 +26,9 @@ // THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT // (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE // OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. -// -// Author: wan@google.com (Zhanyong Wan) -// Google Test - The Google C++ Testing Framework + +// Google Test - The Google C++ Testing and Mocking Framework // // This file implements a universal value printer that can print a // value of any type T: @@ -43,12 +42,13 @@ // defines Foo. #include "gtest/gtest-printers.h" -#include <ctype.h> #include <stdio.h> +#include <cctype> #include <cwchar> #include <ostream> // NOLINT #include <string> #include "gtest/internal/gtest-port.h" +#include "src/gtest-internal-inl.h" namespace testing { @@ -89,7 +89,7 @@ void PrintBytesInObjectToImpl(const unsigned char* obj_bytes, size_t count, // If the object size is bigger than kThreshold, we'll have to omit // some details by printing only the first and the last kChunkSize // bytes. - // TODO(wan): let the user control the threshold using a flag. + // FIXME: let the user control the threshold using a flag. if (count < kThreshold) { PrintByteSegmentInObjectTo(obj_bytes, 0, count, os); } else { @@ -123,7 +123,7 @@ namespace internal { // Depending on the value of a char (or wchar_t), we print it in one // of three formats: // - as is if it's a printable ASCII (e.g. 'a', '2', ' '), -// - as a hexidecimal escape sequence (e.g. '\x7F'), or +// - as a hexadecimal escape sequence (e.g. '\x7F'), or // - as a special escape sequence (e.g. '\r', '\n'). enum CharFormat { kAsIs, @@ -230,7 +230,7 @@ void PrintCharAndCodeTo(Char c, ostream* os) { return; *os << " (" << static_cast<int>(c); - // For more convenience, we print c's code again in hexidecimal, + // For more convenience, we print c's code again in hexadecimal, // unless c was already printed in the form '\x##' or the code is in // [1, 9]. if (format == kHexEscape || (1 <= c && c <= 9)) { @@ -262,11 +262,12 @@ template <typename CharType> GTEST_ATTRIBUTE_NO_SANITIZE_MEMORY_ GTEST_ATTRIBUTE_NO_SANITIZE_ADDRESS_ GTEST_ATTRIBUTE_NO_SANITIZE_THREAD_ -static void PrintCharsAsStringTo( +static CharFormat PrintCharsAsStringTo( const CharType* begin, size_t len, ostream* os) { const char* const kQuoteBegin = sizeof(CharType) == 1 ? "\"" : "L\""; *os << kQuoteBegin; bool is_previous_hex = false; + CharFormat print_format = kAsIs; for (size_t index = 0; index < len; ++index) { const CharType cur = begin[index]; if (is_previous_hex && IsXDigit(cur)) { @@ -276,8 +277,13 @@ static void PrintCharsAsStringTo( *os << "\" " << kQuoteBegin; } is_previous_hex = PrintAsStringLiteralTo(cur, os) == kHexEscape; + // Remember if any characters required hex escaping. + if (is_previous_hex) { + print_format = kHexEscape; + } } *os << "\""; + return print_format; } // Prints a (const) char/wchar_t array of 'len' elements, starting at address @@ -347,15 +353,90 @@ void PrintTo(const wchar_t* s, ostream* os) { } #endif // wchar_t is native +namespace { + +bool ContainsUnprintableControlCodes(const char* str, size_t length) { + const unsigned char *s = reinterpret_cast<const unsigned char *>(str); + + for (size_t i = 0; i < length; i++) { + unsigned char ch = *s++; + if (std::iscntrl(ch)) { + switch (ch) { + case '\t': + case '\n': + case '\r': + break; + default: + return true; + } + } + } + return false; +} + +bool IsUTF8TrailByte(unsigned char t) { return 0x80 <= t && t<= 0xbf; } + +bool IsValidUTF8(const char* str, size_t length) { + const unsigned char *s = reinterpret_cast<const unsigned char *>(str); + + for (size_t i = 0; i < length;) { + unsigned char lead = s[i++]; + + if (lead <= 0x7f) { + continue; // single-byte character (ASCII) 0..7F + } + if (lead < 0xc2) { + return false; // trail byte or non-shortest form + } else if (lead <= 0xdf && (i + 1) <= length && IsUTF8TrailByte(s[i])) { + ++i; // 2-byte character + } else if (0xe0 <= lead && lead <= 0xef && (i + 2) <= length && + IsUTF8TrailByte(s[i]) && + IsUTF8TrailByte(s[i + 1]) && + // check for non-shortest form and surrogate + (lead != 0xe0 || s[i] >= 0xa0) && + (lead != 0xed || s[i] < 0xa0)) { + i += 2; // 3-byte character + } else if (0xf0 <= lead && lead <= 0xf4 && (i + 3) <= length && + IsUTF8TrailByte(s[i]) && + IsUTF8TrailByte(s[i + 1]) && + IsUTF8TrailByte(s[i + 2]) && + // check for non-shortest form + (lead != 0xf0 || s[i] >= 0x90) && + (lead != 0xf4 || s[i] < 0x90)) { + i += 3; // 4-byte character + } else { + return false; + } + } + return true; +} + +void ConditionalPrintAsText(const char* str, size_t length, ostream* os) { + if (!ContainsUnprintableControlCodes(str, length) && + IsValidUTF8(str, length)) { + *os << "\n As Text: \"" << str << "\""; + } +} + +} // anonymous namespace + // Prints a ::string object. #if GTEST_HAS_GLOBAL_STRING void PrintStringTo(const ::string& s, ostream* os) { - PrintCharsAsStringTo(s.data(), s.size(), os); + if (PrintCharsAsStringTo(s.data(), s.size(), os) == kHexEscape) { + if (GTEST_FLAG(print_utf8)) { + ConditionalPrintAsText(s.data(), s.size(), os); + } + } } #endif // GTEST_HAS_GLOBAL_STRING void PrintStringTo(const ::std::string& s, ostream* os) { - PrintCharsAsStringTo(s.data(), s.size(), os); + if (PrintCharsAsStringTo(s.data(), s.size(), os) == kHexEscape) { + if (GTEST_FLAG(print_utf8)) { + ConditionalPrintAsText(s.data(), s.size(), os); + } + } } // Prints a ::wstring object. |