Unescape the XmlEscape method from XUnit (#7323)

XUnit applies some custom escaping to this particular field.
This code is the closest we can come to correctly reversing it so
that it can render normally.

It will produce incorrect results if a control character is followed
by valid hex digits, but there is no way to avoid that, and it should
be relatively rare for raw control characters to appear in error messages.

src/Microsoft.DotNet.Helix/Sdk/tools/azure-pipelines/reporter/formats/xunit.py

@@ -1,7 +1,31 @@
+import re
 import xml.etree.ElementTree
+
 from .result_format import ResultFormat
 from defs import TestResult, TestResultAttachment
 
+_unescape_char_map = {
+    'r': '\r',
+    'n': '\n',
+    't': '\t',
+    '0': '\0',
+    'a': '\a',
+    'b': '\b',
+    'v': '\v',
+    'f': '\f',
+}
+
+def _unescape_xunit_message(value):
+    # xunit does some escaping on the error message we need to do our
+    # best to turn back into something resembling the original message
+    # It only uses \x**, \x**** (indistinguishably), and then the items from __unescape_char_map
+    def bs(match):
+        grp = match.group(0)
+        sym = grp[1]
+        if sym == 'x':
+            return chr(int(grp[2:], 16))
+        return _unescape_char_map.get(match[0][1]) or sym
+    return re.sub(r'\\x[0-9a-fA-F][0-9a-fA-F][0-9a-fA-F]?[0-9a-fA-F]?|\\[^x]', bs, value)
 
 class XUnitFormat(ResultFormat):
 
@@ -38,7 +62,7 @@ def read_results(self, path):
                     exception_type = failure_element.get("exception-type")
                     message_element = failure_element.find("message")
                     if message_element is not None:
-                        failure_message = message_element.text
+                        failure_message = _unescape_xunit_message(message_element.text)
                     stack_trace_element = failure_element.find("stack-trace")
                     if stack_trace_element is not None:
                         stack_trace = stack_trace_element.text