Fix and test case for Mantis bug 1978

inform7/Tests/Test Problems/PM_AmbiguitiesTooDeep.txt

@@ -0,0 +1,3 @@
+Foo is a room.
+
+When play begins: if a room encloses a room that encloses a room that encloses a room that encloses a room that encloses a room that encloses a room that encloses a room that encloses a room that encloses a room that encloses a room that encloses a room, do nothing.

inform7/Tests/Test Problems/_Results_Ideal/PM_AmbiguitiesTooDeep.txt

@@ -0,0 +1,23 @@
+Inform 7 v10.1.0 has started.
+I've now read your source text, which is 55 words long.
+I've also read Basic Inform by Graham Nelson, which is 7691 words long.
+I've also read English Language by Graham Nelson, which is 2328 words long.
+I've also read Standard Rules by Graham Nelson, which is 32092 words long.
+Problem__ PM_AmbiguitiesTooDeep
+  >--> The phrase a room encloses a room that encloses a room that encloses a
+    room that encloses a room that encloses a room that encloses a room that
+    encloses a room that encloses a room that encloses a room that encloses a
+    room that encloses a room is too complicated for me to disentangle without
+    running very, very slowly as I check many ambiguities in it. There ought to
+    be some way to simplify things for me?
+    I was trying to match this phrase:
+      if (a room encloses a room that encloses a room that encloses a room that
+    encloses a room that encloses a room that encloses a room that encloses a
+    room that encloses a room that encloses a room that encloses a room that
+    encloses a room - a condition): 
+    I recognised:
+    a room encloses a room that encloses a room that encloses a room that
+    encloses a room that encloses a room that encloses a room that encloses a
+    room that encloses a room that encloses a room that encloses a room that
+    encloses a room = a condition
+Inform 7 has finished.

inform7/values-module/Chapter 5/Dash.w

@@ -828,8 +828,26 @@ later rules can override earlier ones but still make use of them.
 	for (int i=0; i<no_of_possible_readings; i++)
 		LOGIF(MATCHING, "Possibility (P%d) $e\n", i, list_of_possible_readings[i]);
 
+@ In general, it's not great for typecheckers in compilers to put an upper bound
+on complexity, because although human-written code seldom hits such maxima, there's
+always the possibility of mechanically-generated code which does. On the other hand,
+the result of that doctrine is that a lot of modern compilers (Swift, for example)
+slow to a painful crawl and allocate gigabytes of memory trying to understand
+strange type constraints in two or three lines of code. So, for now at least,
+let's be pragmatic.
+
 @<Add this reading to the list of test cases@> =
-	if (no_of_possible_readings >= MAX_INVOCATIONS_PER_PHRASE) internal_error("overrun");
+	if (no_of_possible_readings >= MAX_INVOCATIONS_PER_PHRASE) {
+		THIS_IS_AN_ORDINARY_PROBLEM;
+		Problems::quote_wording(1, Node::get_text(p));
+		StandardProblems::handmade_problem(Task::syntax_tree(), _p_(PM_AmbiguitiesTooDeep));
+		Problems::issue_problem_segment(
+			"The phrase %1 is too complicated for me to disentangle without "
+			"running very, very slowly as I check many ambiguities in it. There "
+			"ought to be some way to simplify things for me?");
+		Problems::issue_problem_end();
+		return NEVER_MATCH;
+	}
 	list_of_possible_readings[no_of_possible_readings++] = alt;
 	Dash::clear_flags(alt);