adjustments to two tests (#18)

The previous CIDs being matched in the test do not exist in the PubChem database. Consequently, when testing on the PubChem website directly instead of using the BrokenRecord files, the tests were failing. With this change tests do not fail anymore, as I replaced the erroneous CID with one that exists in PubChem.

test/runtests.jl

@@ -46,11 +46,11 @@ BrokenRecord.configure!(; path="http_record")
     @test 27125 ∈ df.CID   # check that estetrol has estriol as a substructure
     sleep(2.0 * get_recordings)
     df13 = CSV.File(playback(() -> query_substructure(;smarts="[r13]Br", output="CSV"), "smarts.bson")) |> DataFrame  # brominated 13-atom ring structures
-    @test 153064026 ∈ df13.CID
+    @test 118303825 ∈ df13.CID
     # The recommended approach for substructure searches is `query_substructure_pug`
     sleep(5.0 * get_recordings)
     cids13 = playback(() -> query_substructure_pug(;smarts="[r13]Br", poll_interval=10*get_recordings), "smarts_pug.bson")  # brominated 13-atom ring structures, via PUG interface
-    @test 153064026 ∈ cids13
+    @test 118303825 ∈ cids13
 
     # properties
     sleep(5.0 * get_recordings)