Avoid a potential underflow when iterating over invalid inputs

The forward iteration logic already bounds-check for m_it != m_string->end(),
do the same for the backward iteration. The issue with the assert is
that the assert() might not be compiled in, and it is happening after
the dereference, so it was too late.

Author: dirkmueller

src/catch2/internal/catch_textflow.cpp

@@ -148,8 +148,8 @@ namespace Catch {
                 m_it--;
             }
             // Skip back over UTF-8 continuation bytes to the leading byte
-            while ( isUtf8ContinuationByte( *m_it ) ) {
-                assert( m_it != m_string->begin() );
+            while ( m_it != m_string->begin() &&
+                    isUtf8ContinuationByte( *m_it ) ) {
                 m_it--;
             }
         }

tests/SelfTest/IntrospectiveTests/TextFlow.tests.cpp

@@ -437,6 +437,25 @@ TEST_CASE( "TextFlow::AnsiSkippingString iterates UTF-8 codepoints",
     }
 }
 
+TEST_CASE( "TextFlow::AnsiSkippingString handles invalid UTF-8",
+           "[TextFlow][ansiskippingstring][approvals]" ) {
+    SECTION( "Continuation byte at the start" ) {
+        // 0x80 is a continuation byte
+        AnsiSkippingString str( "\x80" );
+        auto it = str.end();
+        --it;
+        CHECK( it == str.begin() );
+        CHECK( *it == static_cast<char>( 0x80 ) );
+    }
+    SECTION( "Multiple continuation bytes at the start" ) {
+        AnsiSkippingString str( "\x80\x80\x80" );
+        auto it = str.end();
+        --it;
+        CHECK( it == str.begin() );
+        CHECK( *it == static_cast<char>( 0x80 ) );
+    }
+}
+
 TEST_CASE( "TextFlow::Column wraps UTF-8 text correctly",
            "[TextFlow][column][approvals]" ) {
     // "äöü äöü äöü" = 11 codepoints, 17 bytes