Add wxWhateverWorksConv and use it for file names under Unix

This ensures that we can create output files with Unicode names even when
they're not representable in the current locale encoding, notably when the
current locale has never been changed and is still the default "C" one, not
supporting anything else other than 7 bit ASCII.

Credits for the new class name go to Woody Allen.

interface/wx/strconv.h

@@ -483,6 +483,30 @@ public:
     bool IsOk() const;
 };
 
+/**
+    Conversion object always producing non-empty output for non-empty input.
+
+    Conversions done using this object never lose data, at the cost of possibly
+    producing the output in an unwanted encoding or misinterpreting input
+    encoding.
+
+    To be precise, converting Unicode to multibyte strings using this object
+    tries to use the current locale encoding first but if this doesn't work, it
+    falls back to using UTF-8. In the other direction, UTF-8 is tried first,
+    then the current locale encoding and if this fails too, input is
+    interpreted as using ISO 8859-1, which never fails.
+
+    It is almost always @e wrong to use this converter for multibyte-to-Unicode
+    direction as the program should know which encoding the input data is
+    supposed to use and use the appropriate converter instead. However it may
+    be useful in the Unicode-to-multibyte direction if the goal is to produce
+    the output in the current locale encoding if possible, but still output
+    something, instead of nothing at all, even if the Unicode string is not
+    representable in this encoding.
+
+    @since 3.1.0
+ */
+extern wxMBConv& wxConvWhateverWorks;
 
 
 /**