eclipse-aascw / aas-core-codegen / 35329863572

90%

push

github

Fix the XSD lexical forms in the C# XML (#724)

`ReadContentAsDouble` reads the three named literals
correctly, but it also takes `Infinity` and `-Infinity`, and `nan` and
`NAN` -- none of which `xs:double` admits. The text is now matched
against the lexical space first.

Mind that the pattern ends in `\z` and not in `$`. In .NET, as in
Python, `$` matches not only at the end of the text but also just before
a trailing newline, so `"1.0\n"` would otherwise have passed.

An integer beyond the range of a long leaves `ReadContentAsLong` as
an `OverflowException`, and the filter around the conversion named only
`FormatException` and `XmlException`, so that exception escaped
the error channel and left the de-serialization through a type which
nothing in the API declares.  It is caught now, and reported like any
other bad content.

2 of 2 new or added lines in 1 file covered. (100.0%)

34766 of 40513 relevant lines covered (85.81%)

2.57 hits per line