[FIRParser] Accepting negative numbers where it doesn't make sense

Nobody has claimed this yet.

In the FIRRTL parser we have a function `parseIntLit`. This function happily parses `-1`, and is used in contexts where this doesn't make sense. Here are a couple of examples:

---

Negative integer literal. SFC produces "Exception in thread "main" firrtl.passes.CheckHighFormLike$NegUIntException: : [module Foo] UIntLiteral cannot be negative."

```
circuit Foo:
module Foo:
output z : UInt<8>
z <= UInt<8>(-1)
```
```mlir
firrtl.circuit "Foo" {
firrtl.module @Foo(out %z: !firrtl.uint<8>) {
%c15_ui8 = firrtl.constant 15 : !firrtl.uint<8>
firrtl.strictconnect %z, %c15_ui8 : !firrtl.uint<8>
}
}
```
---

Using a negative integer with `tail` happily compiles:

```
circuit Foo:
module Foo:
input a : UInt<100>
output z : UInt<100>
z <= tail(a, -1)
```
```
firrtl.circuit "Foo" {
firrtl.module @Foo(in %a: !firrtl.uint<100>, out %z: !firrtl.uint<100>) {
%0 = firrtl.tail %a, 15 : (!firrtl.uint<100>) -> !firrtl.uint<85>
%1 = firrtl.pad %0, 100 : (!firrtl.uint<85>) -> !firrtl.uint<100>
firrtl.strictconnect %z, %1 : !firrtl.uint<100>
}
}
```

---

Negative array length. Similar to above the `-1` is translated to `15`:
```
circuit Foo:
module Foo:
output z : UInt<8>[-1]
z is invalid
```
```mlir
firrtl.circuit "Foo" {
firrtl.module @Foo(out %z: !firrtl.vector, 15>) {
%invalid = firrtl.invalidvalue : !firrtl.vector, 15>
firrtl.strictconnect %z, %invalid : !firrtl.vector, 15>
}
}
```

Start at the FIRParser function parseIntLit and trace its uses in integer literals, tail, and array lengths. Reproduce the three examples from the issue and verify that negative values are rejected where they are invalid instead of being converted to unsigned values or accepted silently.

Written by the indexing model from the issue text.