Update query_edge_limit config option info and error message.

[danielmai]

• Clarify that the flag --query_edge_limit affects shortest path and recursive queries.
• Reword query edge limit error since it only counts edges and does not measure memory usage.
• Remove "ErrToBig" error variable.
• Add query tests for query edge limit.

---

This change is 

[martinmr]

Reviewed 4 of 4 files at r1.
Reviewable status: complete! all files reviewed, all discussions resolved

[martinmr]

Reviewable status: all files reviewed, 3 unresolved discussions (waiting on @danielmai)

---

query/query3_test.go, line 49 at r1 (raw file):

func TestRecurseEdgeLimitError(t *testing.T) {
	// HACK: Set this flag in the external test cluster when query tests are migrated to

This might not work with the new setup. The current tests are running with the default cluster so it won't be possible to change the edge limit on the fly and setting it to five for all the tests will cause other tests to fail.

One possible solution:

• Move code in common_test.go to a common package and export all the required functions.
• Modify existing calls to those functions to point to the new common package.
• Move these two tests to a different package. TestMain should be identical to the current version.
• Create a custom cluster in that package with the required changes.

It shouldn't be too difficult. Mostly a lot of refactoring but that should be doable with search and replace. Let me know if you have questions.

---

query/query3_test.go, line 62 at r1 (raw file):

	ctx := defaultContext()
	_, err := processToFastJsonCtxVars(t, query, ctx, nil)

change this to processQuery

---

query/query3_test.go, line 294 at r1 (raw file):

	ctx := defaultContext()
	_, err := processToFastJsonCtxVars(t, query, ctx, nil)

change this to processQuery

[manishrjain]

Fix the test and merge.

Reviewable status: all files reviewed, 3 unresolved discussions (waiting on @danielmai)

[manishrjain]

Feel free to transfer to someone if needed, but let's aim to get this in soon.

Reviewable status: all files reviewed, 3 unresolved discussions (waiting on @danielmai)

[manishrjain]

Reviewable status: all files reviewed, 3 unresolved discussions (waiting on @danielmai)

[manishrjain]

Not worth writing test for. Also, discard all unnecessary changes.

@martinmr can give the final LGTM.

Reviewable status: 2 of 11 files reviewed, 3 unresolved discussions (waiting on @danielmai and @martinmr)

[martinmr]

Reviewed 1 of 9 files at r2, 1 of 8 files at r4.
Reviewable status: all files reviewed, 3 unresolved discussions (waiting on @danielmai)