commands: the stages that rewrite a document

`$addFields`, `$set`, `$unset`, `$replaceRoot` and `$unwind`. The corpus goes
0 pass / 24 fail to 21 / 3, and the three left are `$project`'s computed
fields, renames and nested inclusions, which want the `query.project` that
`find` shares and are their own change.

**The design review was wrong about what this needed, and the corpus is what
settled it.** Tier 2 was scoped as a per-stage iterator on the grounds that
`$unwind` is 1->N and "there is no way to express that in a window over the
input". That was true of the window as it stood, and stopped being true the
moment `$project` was made to rebuild the stream instead of moving bounds over
it -- a stage that rebuilds can emit as many documents as it likes, or none. So
all five share one shape: read the window, build a new list, replace the
stream. No iterator, no rewrite.

What the recording settled, and what a hand-written test would have got wrong:

  - `$addFields` whose expression resolves to nothing leaves the field out
    entirely rather than setting it to null -- so `set_path` is only reached
    when there is a value, and `eval_expr`'s absent/null distinction earns its
    keep a second time.
  - `$addFields: {"n.z": 1}` sets the nested path and keeps its siblings, and
    an existing field is replaced *where it stands*, which is what makes the
    stage "add or overwrite" rather than "append".
  - `$unwind` drops a document whose field is missing or an empty array, keeps
    one whose field is not an array *whole*, and numbers `includeArrayIndex`
    from zero. Three separate behaviours where one guess would have covered
    them all wrongly.
  - `$replaceRoot` of a missing path and of a non-document are the same error,
    40228.

`ReplaceRootNotDocument` joins `EvalError` rather than being reported at the
stage: it is a failure only a document can produce, which is the line that set
already draws.

191/191 unit tests in ReleaseFast and ReleaseSafe, 83/83 fuzz, e2e 49, e2e3 16,
e2e4 17, e2e6 72, e2e7 86, crud corpus unchanged at 201/90/196.

tests/spec/aggregate/README.md

@@ -64,7 +64,7 @@ accumulators in:
 ```
 group-accumulators.json   19 pass    0 fail   0 skip
 expressions.json          27 pass    0 fail   0 skip
-document-stages.json       0 pass   24 fail   0 skip
+document-stages.json      21 pass    3 fail   0 skip
 ```
 
 `group-accumulators` found its first real disagreement on the way to 18:
@@ -73,10 +73,11 @@ that counted documents rather than numbers would have passed every test
 anybody would think to write by hand.
 
 `expressions.json` was recorded before the evaluator was written and read
-1 pass / 26 fail against it; it is green now. `document-stages.json` is the
-next tier's spec, recorded and not implemented -- `$addFields`/`$set`,
-`$unset`, `$replaceRoot`, `$unwind` and `$project`'s computed fields, none of
-which this server has, so it starts at zero.
+1 pass / 26 fail against it; it is green now. `document-stages.json` was recorded at
+0 pass / 24 fail; `$addFields`/`$set`, `$unset`, `$replaceRoot` and `$unwind`
+took it to 21. The three left are `$project`'s computed fields, its renames and
+its nested inclusions, which want the shared `query.project` and are their own
+change.
 
 What recording *that* settled: