refactor(gbif): use testitem

GBIF/Project.toml

@@ -1,13 +1,14 @@
 name = "GBIF"
 uuid = "ee291a33-5a6c-5552-a3c8-0f29a1181037"
 authors = ["Timothée Poisot <[email protected]>"]
-version = "0.4.4"
+version = "0.4.5"
 
 [deps]
 Dates = "ade2ca70-3891-5945-98fb-dc099432e06a"
 HTTP = "cd3eb016-35fb-5094-929b-558a96fad6f3"
 JSON = "682c06a0-de6a-54ab-a142-c8b1cf79cde6"
 Tables = "bd369af6-aec1-5ad0-b16a-f7cc5008161c"
+TestItems = "1c621080-faea-4a02-84b6-bbd5e436b8fe"
 
 [compat]
 HTTP = "0.8, 0.9, 1"

GBIF/src/GBIF.jl

@@ -1,5 +1,7 @@
 module GBIF
 
+using TestItems
+
 using HTTP
 using JSON
 using Dates
@@ -78,4 +80,19 @@ include("paging.jl")
 export occurrence, occurrences
 export occurrences!
 
+@testitem "We can use the Query package" begin
+    using Query
+    using DataFrames
+    t = taxon("Mammalia", strict=false)
+    set = occurrences(t)
+    [occurrences!(set) for i in 1:10]
+
+    tdf = view(set) |>
+        @filter(_.rank == "SPECIES") |>
+        @map({_.key, _.taxon.name, _.country}) |>
+        DataFrame
+
+    @test typeof(tdf) <: DataFrame
+end
+
 end # module

GBIF/src/occurrence.jl

@@ -58,6 +58,20 @@ function occurrence(key::Integer)::GBIFRecord
     return occurrence(string(key))
 end
 
+@testitem "We can get an occurrence by its ID" begin
+    k = 3986160931
+    o = occurrence(k)
+    @test typeof(o) == GBIFRecord
+    @test o.key == k
+end
+
+@testitem "We can get a malformed occurrence" begin
+    k = 1039645472
+    o = occurrence(k)
+    @test typeof(o) == GBIFRecord
+    @test o.key == k
+end
+
 function _internal_occurrences_getter(query::Pair...)
     validate_occurrence_query.(query)
     occ_s_url = gbifurl * "occurrence/search"
@@ -112,6 +126,20 @@ function occurrences(t::GBIFTaxon, query::Pair...)
     return occurrences(taxon_query, query...)
 end
 
+@testitem "Getting occurrences by taxon returns a GBIFRecords" begin
+    iver_id = taxon(5298019)
+    @test typeof(occurrences(iver_id)) <: GBIFRecords
+end
+
+@testitem "Getting occurrences by taxon returns the correct taxa" begin
+    t = taxon(5298019)
+    occ = occurrences(t)
+    @test typeof(occ) <: GBIFRecords
+    for o in occ
+        @test o.taxon.species == Pair("Iris versicolor", 5298019)
+    end
+end
+
 """
     occurrences(t::Vector{GBIFTaxon}, query::Pair...)
 
@@ -126,3 +154,63 @@ function occurrences(ts::Vector{GBIFTaxon}, query::Pair...)
     end
     return occurrences(taxon_query..., query...)
 end
+
+@testitem "We get the correct country when looking for occurrences by country" begin
+    occ = occurrences("country" => "ES")
+    @test typeof(occ) <: GBIFRecords
+    for o in occ
+        @test o.country == "Spain"
+    end
+end
+
+@testitem "We can pass the taxon data manually" begin
+set1 = occurrences("scientificName" => "Mus musculus", "year" => 1999, "hasCoordinate" => true)
+@test typeof(set1) == GBIFRecords
+@test length(set1) == 20
+end
+
+@testitem "We can get the latest occurrences" begin
+set2 = occurrences()
+@test typeof(set2) == GBIFRecords
+@test length(set2) == 20
+end
+
+@testitem "We can use tuples to show intervals" begin
+set3 = occurrences("scientificName" => "Mus musculus", "year" => 1999, "hasCoordinate" => true, "decimalLatitude" => (0.0, 50.0))
+@test typeof(set3) == GBIFRecords
+@test length(set3) == 20
+end
+
+@testitem "We can complete a query using a while loop" begin
+serval = GBIF.taxon("Leptailurus serval", strict=true)
+obs = occurrences(serval, "hasCoordinate" => "true", "continent" => "AFRICA", "decimalLongitude" => (-30, 40))
+while length(obs) < count(obs)
+    occurrences!(obs)
+end
+@test length(obs) == count(obs)
+end
+
+@testitem "We can query multiple taxa at once" begin
+serval = GBIF.taxon("Leptailurus serval", strict=true)
+leopard = GBIF.taxon("Panthera pardus", strict=true)
+obs = occurrences([leopard, serval], "hasCoordinate" => true, "occurrenceStatus" => "PRESENT")
+@test typeof(obs) == GBIFRecords
+end
+
+@testitem "We can complete a query with a specific page size" begin
+obs = occurrences(serval, "hasCoordinate" => "true", "continent" => "AFRICA", "decimalLongitude" => (-30, 40), "limit" => 45)
+while length(obs) < count(obs)
+    occurrences!(obs)
+end
+@test length(obs) == count(obs)
+end
+
+@testitem "Records of absence are correctly represented" begin
+o = occurrences("occurrenceStatus" => "ABSENT")
+@test o[1].presence == false
+end
+
+@testitem "Records of presence are correctly represented" begin
+o = occurrences("occurrenceStatus" => "PRESENT")
+@test o[1].presence == true
+end

GBIF/src/paging.jl

@@ -47,3 +47,24 @@ function occurrences!(o::GBIFRecords)
         o.occurrences[start:stop] = retrieved
     end
 end
+
+@testitem "We always get one page by default" begin
+    set = occurrences()
+    occurrences!(set)
+    @test length(set) == 40
+end
+
+@testitem "We can get multiple pages on repeated calls" begin
+    set = occurrences("limit" => 40)
+    occurrences!(set)
+    @test length(set) == 80
+    @test length(set) == length(unique([o.key for o in set]))
+end
+
+@testitem "We can page over queries" begin
+    setQ = occurrences(taxon("Iris versicolor", rank=:SPECIES), "limit" => 10)
+    occurrences!(setQ)
+    @test length(setQ) == 20
+    occurrences!(setQ)
+    @test length(setQ) == length(unique([o.key for o in setQ]))
+end

GBIF/src/query.jl

@@ -92,3 +92,26 @@ function validate_occurrence_query(query::Pair)
         end
     end
 end
+
+@testitem "Using a single argument does not change the type of the argument" begin
+    queryset = ["hasCoordinate" => true]
+    @test length(occurrences(queryset...)) > 0
+
+    queryset = ["hasCoordinate" => true]
+    @test length(occurrences(taxon("Alces alces"), queryset...)) > 0
+end
+
+@testitem "We cannot search for a wrong country code" begin
+  qpars = Pair("country", "ABC")
+  @test_warn "country code" GBIF.validate_occurrence_query(qpars)
+end
+
+@testitem "We cannot search with an unsupported keyword" begin
+  qpars = Pair("years", "1234")
+  @test_warn "parameter is not allowed" GBIF.validate_occurrence_query(qpars)
+end
+
+@testitem "We cannot search with an unsupported enumerated value" begin
+  qpars = Pair("establishmentMeans", "just vibes")
+  @test_warn "is not a valid value" GBIF.validate_occurrence_query(qpars)
+end

GBIF/src/taxon.jl

@@ -8,15 +8,15 @@ reference taxonomy.
 
 Optional arguments are
 
-- `rank::Union{Symbol,Nothing}=:SPECIES` -- the rank of the taxon you want. This
-  is part of a controlled vocabulary, and can only be one of `:DOMAIN`,
-  `:CLASS`, `:CULTIVAR`, `:FAMILY`, `:FORM`, `:GENUS`, `:INFORMAL`, `:ORDER`,
-  `:PHYLUM,`, `:SECTION`, `:SUBCLASS`, `:VARIETY`, `:TRIBE`, `:KINGDOM`,
-  `:SUBFAMILY`, `:SUBFORM`, `:SUBGENUS`, `:SUBKINGDOM`, `:SUBORDER`,
-  `:SUBPHYLUM`, `:SUBSECTION`, `:SUBSPECIES`, `:SUBTRIBE`, `:SUBVARIETY`,
-  `:SUPERCLASS`, `:SUPERFAMILY`, `:SUPERORDER`, and `:SPECIES`
+  - `rank::Union{Symbol,Nothing}=:SPECIES` -- the rank of the taxon you want. This
+    is part of a controlled vocabulary, and can only be one of `:DOMAIN`,
+    `:CLASS`, `:CULTIVAR`, `:FAMILY`, `:FORM`, `:GENUS`, `:INFORMAL`, `:ORDER`,
+    `:PHYLUM,`, `:SECTION`, `:SUBCLASS`, `:VARIETY`, `:TRIBE`, `:KINGDOM`,
+    `:SUBFAMILY`, `:SUBFORM`, `:SUBGENUS`, `:SUBKINGDOM`, `:SUBORDER`,
+    `:SUBPHYLUM`, `:SUBSECTION`, `:SUBSPECIES`, `:SUBTRIBE`, `:SUBVARIETY`,
+    `:SUPERCLASS`, `:SUPERFAMILY`, `:SUPERORDER`, and `:SPECIES`
 
-- `strict::Bool=true` -- whether the match should be strict, or fuzzy
+  - `strict::Bool=true` -- whether the match should be strict, or fuzzy
 
 Finally, one can also specify other levels of the taxonomy, using  `kingdom`,
 `phylum`, `class`, `order`, `family`, and `genus`, all of which can either be
@@ -26,17 +26,19 @@ If a match is found, the result will be given as a `GBIFTaxon`. If not, this
 function will return `nothing` and give a warning.
 """
 function taxon(name::String;
-    rank::Union{Symbol,Nothing}=:SPECIES, strict::Bool=true,
-    kingdom::Union{String,Nothing}=nothing, phylum::Union{String,Nothing}=nothing, class::Union{String,Nothing}=nothing,
-    order::Union{String,Nothing}=nothing, family::Union{String,Nothing}=nothing, genus::Union{String,Nothing}=nothing)
+    rank::Union{Symbol, Nothing} = :SPECIES, strict::Bool = true,
+    kingdom::Union{String, Nothing} = nothing, phylum::Union{String, Nothing} = nothing,
+    class::Union{String, Nothing} = nothing,
+    order::Union{String, Nothing} = nothing, family::Union{String, Nothing} = nothing,
+    genus::Union{String, Nothing} = nothing)
     @assert rank ∈ [
         :DOMAIN, :CLASS, :CULTIVAR, :FAMILY, :FORM, :GENUS, :INFORMAL, :ORDER, :PHYLUM,
         :SECTION, :SUBCLASS, :VARIETY, :TRIBE, :KINGDOM, :SUBFAMILY, :SUBFORM,
         :SUBGENUS, :SUBKINGDOM, :SUBORDER, :SUBPHYLUM, :SUBSECTION, :SUBSERIES,
         :SUBSPECIES, :SUBTRIBE, :SUBVARIETY, :SUPERCLASS, :SUPERFAMILY, :SUPERORDER,
-        :SPECIES
+        :SPECIES,
     ]
-    args = Dict{String,Any}("name" => name, "strict" => strict)
+    args = Dict{String, Any}("name" => name, "strict" => strict)
 
     isnothing(rank) || (args["rank"] = String(rank))
     isnothing(kingdom) || (args["kingdom"] = String(kingdom))
@@ -47,7 +49,7 @@ function taxon(name::String;
     isnothing(genus) || (args["genus"] = String(genus))
 
     sp_s_url = gbifurl * "species/match"
-    sp_s_req = HTTP.get(sp_s_url, query=args)
+    sp_s_req = HTTP.get(sp_s_url; query = args)
     if sp_s_req.status == 200
         body = JSON.parse(String(sp_s_req.body))
         # This will throw warnings for various reasons related to matchtypes
@@ -78,17 +80,35 @@ This function will look for a taxon by its taxonID in the GBIF
 reference taxonomy.
 """
 function taxon(id::Int)
-    args = Dict{String,Any}("id" => id)
+    args = Dict{String, Any}("id" => id)
 
     sp_s_url = gbifurl * "species/$id"
-    sp_s_req = HTTP.get(sp_s_url, query=args)
+    sp_s_req = HTTP.get(sp_s_url; query = args)
     if sp_s_req.status == 200
         body = JSON.parse(String(sp_s_req.body))
         return GBIFTaxon(body)
     else
-        throw(ErrorException("Impossible to retrieve information for taxonID $(id) -- HTML error code $(sp_s_req.status)"))
+        throw(
+            ErrorException(
+                "Impossible to retrieve information for taxonID $(id) -- HTML error code $(sp_s_req.status)",
+            ),
+        )
     end
-
 end
 
 taxon(t::Pair) = taxon(t.second)
+
+@testitem "We can get a taxon by its name" begin
+    iver = taxon("Iris versicolor"; rank = :SPECIES)
+    @test iver.species == Pair("Iris versicolor", 5298019)
+end
+
+@testitem "We can get a taxon by its GBIF ID" begin
+    iver_id = taxon(5298019)
+    @test iver_id.species == ("Iris versicolor" => 5298019)
+end
+
+@testitem "We do not get a species field when looking for a genus" begin
+    i_sp = taxon("Lamellodiscus"; rank = :GENUS, strict = true)
+    @test ismissing(i_sp.species)
+end

GBIF/src/types/iterators.jl

@@ -21,6 +21,12 @@ function getindex(o::GBIFRecords, r::UnitRange{Int64})
     return view(o)[r]
 end
 
+@testitem "Occurrences have a length and a typeof" begin
+    set = occurrences()
+    @test typeof(set[1]) == GBIFRecord
+    @test length(set[1:4]) == 4
+end
+
 function iterate(o::GBIFRecords)
     return iterate(collect(view(o)))
 end
@@ -29,7 +35,30 @@ function iterate(o::GBIFRecords, t::Union{Int64, Nothing})
     return iterate(collect(view(o)), t)
 end
 
+@testitem "We can iterate over records" begin
+    set = occurrences()
+    @test iterate(set) == (set[1], 2)
+    @test iterate(set, 2) == (set[2], 3)
+end
+
+@testitem "We can iterate over records made with a non-empty query" begin
+    plotor = taxon("Procyon lotor")
+    plotor_occ = occurrences(plotor)
+    occurrences!(plotor_occ)
+    for o in plotor_occ
+        @test typeof(o) <: GBIFRecord
+        @test o.taxon.species.second == plotor.species.second
+    end
+end
+
 # Tables.jl interface
 Tables.istable(::Type{GBIFRecords}) = true
 Tables.rowaccess(::Type{GBIFRecords}) = true
-Tables.rows(records::GBIFRecords) = view(records)
+Tables.rows(records::GBIFRecords) = view(records)
+
+@testitem "We can convert records to a data frame" begin
+    using DataFrames
+
+    df = DataFrame(occurrences())
+    @test typeof(df) <: DataFrame
+end

GBIF/test/Project.toml

@@ -2,3 +2,4 @@
 DataFrames = "a93c6f00-e57d-5684-b7b6-d8193f3e46c0"
 Query = "1a8c2f83-1ff3-5112-b086-8aa67b057ba1"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
+TestItemRunner = "f8b46487-2199-4994-9208-9a1283c18c0a"