Revert "quieten tests"

src/Embeddings.jl

@@ -1,6 +1,6 @@
 module Embeddings
 
-using LinearAlgebra: norm
+using Statistics: norm
 using DataDeps
 using AutoHashEquals
 

src/word2vec.jl

@@ -34,18 +34,18 @@ function _load_embeddings(::Type{<:Word2Vec}, embedding_file, max_vocab_size, ke
 
         index = 1
         @inbounds for _ in 1:vocab_size
-            word = readuntil(fh, ' '; keep=false)
+            word = readuntil(fh, ' ', keep=false)
             vector = Vector{Float32}(undef, vector_size)
             @inbounds for i = 1:vector_size
                 vector[i] = read(fh, Float32)
             end
 
-            if isempty(keep_words) || word ∈ keep_words
-                LL[:, index] = vector ./ norm(vector)
+            if !occursin("_", word) && (length(keep_words)==0 || word in keep_words ) #If it isn't a phrase
+                LL[:,index]=vector./norm(vector)
                 indexed_words[index] = word
 
-                index += 1
-                if index > max_stored_vocab_size
+                index+=1
+                if index>max_stored_vocab_size
                     break
                 end
             end

test/runtests.jl

@@ -1,8 +1,7 @@
 using Embeddings
 using Test
-using Suppressor
-using DataDeps
 
+using DataDeps
 
 """
     tempdatadeps(fun)
@@ -36,7 +35,7 @@ end
 @testset_nokeep_data
 
 Use just like @testset,
-but know that it deletes any downquiet_loaded data dependencies when it is done.
+but know that it deletes any downloaded data dependencies when it is done.
 """
 macro testset_nokeep_data(name, expr)
     quote
@@ -47,29 +46,20 @@ macro testset_nokeep_data(name, expr)
 end
 
 
-"""
-Loads embeddings with suppressed output.
-For purposes of not flooding the logs.
-During downloads.
-"""
-function quiet_load_embeddings(args...; kwargs...)
-     @suppress load_embeddings(args...; kwargs...)
-end
-
 @testset_nokeep_data "Word2Vec" begin
-    embs_full =  quiet_load_embeddings(Word2Vec)
+    embs_full = load_embeddings(Word2Vec)
 
     @test size(embs_full.embeddings) == (300, length(embs_full.vocab))
 
-     embs_mini = quiet_load_embeddings(Word2Vec; max_vocab_size=100)
+    embs_mini = load_embeddings(Word2Vec; max_vocab_size=100)
     @test length(embs_mini.vocab)==100
 
     @test embs_mini.embeddings == embs_full.embeddings[:, 1:100]
     @test embs_mini.vocab == embs_full.vocab[1:100]
 
     @test "for" ∈ embs_mini.vocab
 
-     embs_specific =  quiet_load_embeddings(Word2Vec; keep_words=Set(["red", "green", "blue"]))
+    embs_specific =  load_embeddings(Word2Vec; keep_words=Set(["red", "green", "blue"]))
 
     @test size(embs_specific.embeddings) == (300, 3)
     @test Set(embs_specific.vocab) == Set(["red", "green", "blue"])
@@ -78,7 +68,7 @@ end
 @testset "GloVe" begin
     # just test one file from each of provided sets
     tests = ["glove.6B/glove.6B.50d.txt",
-             #"glove.42B.300d/glove.42B.300d.txt",     # These files are too slow to downquiet_load
+             #"glove.42B.300d/glove.42B.300d.txt",     # These files are too slow to download
              #"glove.840B.300d/glove.840B.300d.txt",   # They are not that big bt are on a slow server
              "glove.twitter.27B/glove.twitter.27B.25d.txt"]
 
@@ -90,15 +80,15 @@ end
 
         @testset_nokeep_data "$filename" begin
             @testset "Basic" begin
-                 glove = quiet_load_embeddings(GloVe{:en}, @datadep_str(file), max_vocab_size=1000)
+                glove = load_embeddings(GloVe{:en}, @datadep_str(file), max_vocab_size=1000)
                 @test length(glove.vocab) == 1000
                 @test size(glove.embeddings) == (dim(file), 1000)
                 @test "for" ∈ glove.vocab
             end
 
             @testset "Specific" begin
                 colors = ["red", "green", "blue"]
-                 glove_colors = quiet_load_embeddings(GloVe, @datadep_str(file), keep_words=colors)
+                glove_colors = load_embeddings(GloVe, @datadep_str(file), keep_words=colors)
                 @test length(glove_colors.vocab) == 3
                 @test size(glove_colors.embeddings) == (dim(file), 3)
                 @test Set(glove_colors.vocab) == Set(colors)
@@ -110,14 +100,14 @@ end
         # first 100 lines of official glove.6B.50d.txt
         custom_glove_file = joinpath(@__DIR__, "data", "custom.glove.txt")
         @testset "Basic" begin
-            glove = quiet_load_embeddings(GloVe, custom_glove_file)
+            glove = load_embeddings(GloVe, custom_glove_file)
             @test length(glove.vocab) == 100
             @test size(glove.embeddings) == (50, 100)
             @test "the" ∈ glove.vocab
         end
         @testset "Specific" begin
             punct = [".", ","]
-            glove_punct = quiet_load_embeddings(GloVe, custom_glove_file, keep_words=punct)
+            glove_punct = load_embeddings(GloVe, custom_glove_file, keep_words=punct)
             @test length(glove_punct.vocab) == 2
             @test size(glove_punct.embeddings) == (50, 2)
             @test Set(glove_punct.vocab) == Set(punct)
@@ -129,29 +119,29 @@ end
 @testset "FastText" begin
     @testset_nokeep_data "English 1" begin
         @testset "Basic" begin
-             embs1 = quiet_load_embeddings(FastText_Text; max_vocab_size=100)
+            embs1 = load_embeddings(FastText_Text; max_vocab_size=100)
             @test length(embs1.vocab)==100
             @test size(embs1.embeddings) == (300, 100)
         end
 
         @testset "Specific" begin           
-             embs_specific =  quiet_load_embeddings(FastText_Text; keep_words=Set(["red", "green", "blue"]))
+            embs_specific =  load_embeddings(FastText_Text; keep_words=Set(["red", "green", "blue"]))
             @test size(embs_specific.embeddings) == (300, 3)
             @test Set(embs_specific.vocab) == Set(["red", "green", "blue"])
         end
     end
 
 
     @testset_nokeep_data "French" begin
-         embs_fr = quiet_load_embeddings(FastText_Text{:fr}; max_vocab_size=100)
+        embs_fr = load_embeddings(FastText_Text{:fr}; max_vocab_size=100)
         @test length(embs_fr.vocab)==100
         @test size(embs_fr.embeddings) == (300, 100)
     end
 
 
 
     @testset_nokeep_data "English file number 2" begin
-        embs_specific =  quiet_load_embeddings(FastText_Text, 2; keep_words=Set(["red", "green", "blue"]))
+        embs_specific =  load_embeddings(FastText_Text, 2; keep_words=Set(["red", "green", "blue"]))
         @test size(embs_specific.embeddings) == (300, 3)
         @test Set(embs_specific.vocab) == Set(["red", "green", "blue"])
     end