fasttext 0.2.1 → 0.2.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 597a9c73a6720397b11fb9bd0492c33ed458783747d904deef74618a6f319133
-  data.tar.gz: b03ab1e051cda71fafc499dbf0847d3da1c27050646e949eac148a5548460054
+  metadata.gz: 9aae7e20933f51ebebd802276d7e006ce792fcfc6dd94ea8ceda887ab0f4eca8
+  data.tar.gz: a90a7bbfffe424829052afc3519bc0473a7fc1b22a2d1ea0c8786d9372a0cdb5
 SHA512:
-  metadata.gz: 844ae824ad06aa270a5c58b6dc109701660b2bb0e5bfe63e285df57bc5db854028703c1c0476a845d5b0e8c32dd02ac42e1c8b1c5c0684713928c895c776b1da
-  data.tar.gz: 7f7a07f0041981eb046e84f6b44f4207192eb9e8a786edf41b53ae15b370eb875f5b4d578248c90a371805697ad3f0fa076532d1f6789c44d336d72402a54333
+  metadata.gz: 3042a798560e5960d18d8bcefb66833bf621bc1dcf26f77742602d0d4e7b1e4cc33b078f9225c209ac9a0d2f89c9b3c98ad2446582af4a3ac4effc761192c58d
+  data.tar.gz: d70ce1005916a809a78b02f23232b1efd7cc59dc9d3ed2df7929f946db2f326c0df40729b2ea953846e906c2774d6b6e971650f3d585da65ae6562bdcb9a9b5e

data/CHANGELOG.md

@@ -1,3 +1,7 @@
+## 0.2.2 (2021-10-16)
+
+- Fixed `file cannot be opened` errors
+
 ## 0.2.1 (2021-05-23)
 
 - Improved performance

data/ext/fasttext/ext.cpp

@@ -16,13 +16,12 @@
 #include <rice/rice.hpp>
 #include <rice/stl.hpp>
 
+using fasttext::Args;
 using fasttext::FastText;
 
 using Rice::Array;
 using Rice::Constructor;
-using Rice::Hash;
 using Rice::Module;
-using Rice::Object;
 using Rice::define_class_under;
 using Rice::define_module;
 using Rice::define_module_under;
@@ -47,103 +46,69 @@ namespace Rice::detail
   };
 }
 
-fasttext::Args buildArgs(Hash h) {
-  fasttext::Args a;
-
-  for (const auto& it : h)
-  {
-    auto name = it.key.to_s().str();
-    auto value = (it.value).value();
-
-    if (name == "input") {
-      a.input = Rice::detail::From_Ruby<std::string>().convert(value);
-    } else if (name == "output") {
-      a.output = Rice::detail::From_Ruby<std::string>().convert(value);
-    } else if (name == "lr") {
-      a.lr = Rice::detail::From_Ruby<double>().convert(value);
-    } else if (name == "lr_update_rate") {
-      a.lrUpdateRate = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "dim") {
-      a.dim = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "ws") {
-      a.ws = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "epoch") {
-      a.epoch = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "min_count") {
-      a.minCount = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "min_count_label") {
-      a.minCountLabel = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "neg") {
-      a.neg = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "word_ngrams") {
-      a.wordNgrams = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "loss") {
-      std::string str = Rice::detail::From_Ruby<std::string>().convert(value);
-      if (str == "softmax") {
-        a.loss = fasttext::loss_name::softmax;
-      } else if (str == "ns") {
-        a.loss = fasttext::loss_name::ns;
-      } else if (str == "hs") {
-        a.loss = fasttext::loss_name::hs;
-      } else if (str == "ova") {
-        a.loss = fasttext::loss_name::ova;
-      } else {
-        throw std::invalid_argument("Unknown loss: " + str);
-      }
-    } else if (name == "model") {
-      std::string str = Rice::detail::From_Ruby<std::string>().convert(value);
-      if (str == "supervised") {
-        a.model = fasttext::model_name::sup;
-      } else if (str == "skipgram") {
-        a.model = fasttext::model_name::sg;
-      } else if (str == "cbow") {
-        a.model = fasttext::model_name::cbow;
-      } else {
-        throw std::invalid_argument("Unknown model: " + str);
-      }
-    } else if (name == "bucket") {
-      a.bucket = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "minn") {
-      a.minn = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "maxn") {
-      a.maxn = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "thread") {
-      a.thread = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "t") {
-      a.t = Rice::detail::From_Ruby<double>().convert(value);
-    } else if (name == "label_prefix") {
-      a.label = Rice::detail::From_Ruby<std::string>().convert(value);
-    } else if (name == "verbose") {
-      a.verbose = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "pretrained_vectors") {
-      a.pretrainedVectors = Rice::detail::From_Ruby<std::string>().convert(value);
-    } else if (name == "save_output") {
-      a.saveOutput = Rice::detail::From_Ruby<bool>().convert(value);
-    } else if (name == "seed") {
-      a.seed = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "autotune_validation_file") {
-      a.autotuneValidationFile = Rice::detail::From_Ruby<std::string>().convert(value);
-    } else if (name == "autotune_metric") {
-      a.autotuneMetric = Rice::detail::From_Ruby<std::string>().convert(value);
-    } else if (name == "autotune_predictions") {
-      a.autotunePredictions = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "autotune_duration") {
-      a.autotuneDuration = Rice::detail::From_Ruby<int>().convert(value);
-    } else if (name == "autotune_model_size") {
-      a.autotuneModelSize = Rice::detail::From_Ruby<std::string>().convert(value);
-    } else {
-      throw std::invalid_argument("Unknown argument: " + name);
-    }
-  }
-  return a;
-}
-
 extern "C"
 void Init_ext()
 {
   Module rb_mFastText = define_module("FastText");
   Module rb_mExt = define_module_under(rb_mFastText, "Ext");
 
+  define_class_under<Args>(rb_mExt, "Args")
+    .define_constructor(Constructor<Args>())
+    .define_attr("input", &Args::input)
+    .define_attr("output", &Args::output)
+    .define_attr("lr", &Args::lr)
+    .define_attr("lr_update_rate", &Args::lrUpdateRate)
+    .define_attr("dim", &Args::dim)
+    .define_attr("ws", &Args::ws)
+    .define_attr("epoch", &Args::epoch)
+    .define_attr("min_count", &Args::minCount)
+    .define_attr("min_count_label", &Args::minCountLabel)
+    .define_attr("neg", &Args::neg)
+    .define_attr("word_ngrams", &Args::wordNgrams)
+    .define_method(
+      "loss=",
+      [](Args& a, const std::string& str) {
+        if (str == "softmax") {
+          a.loss = fasttext::loss_name::softmax;
+        } else if (str == "ns") {
+          a.loss = fasttext::loss_name::ns;
+        } else if (str == "hs") {
+          a.loss = fasttext::loss_name::hs;
+        } else if (str == "ova") {
+          a.loss = fasttext::loss_name::ova;
+        } else {
+          throw std::invalid_argument("Unknown loss: " + str);
+        }
+      })
+    .define_method(
+      "model=",
+      [](Args& a, const std::string& str) {
+        if (str == "supervised") {
+          a.model = fasttext::model_name::sup;
+        } else if (str == "skipgram") {
+          a.model = fasttext::model_name::sg;
+        } else if (str == "cbow") {
+          a.model = fasttext::model_name::cbow;
+        } else {
+          throw std::invalid_argument("Unknown model: " + str);
+        }
+      })
+    .define_attr("bucket", &Args::bucket)
+    .define_attr("minn", &Args::minn)
+    .define_attr("maxn", &Args::maxn)
+    .define_attr("thread", &Args::thread)
+    .define_attr("t", &Args::t)
+    .define_attr("label_prefix", &Args::label)
+    .define_attr("verbose", &Args::verbose)
+    .define_attr("pretrained_vectors", &Args::pretrainedVectors)
+    .define_attr("save_output", &Args::saveOutput)
+    .define_attr("seed", &Args::seed)
+    .define_attr("autotune_validation_file", &Args::autotuneValidationFile)
+    .define_attr("autotune_metric", &Args::autotuneMetric)
+    .define_attr("autotune_predictions", &Args::autotunePredictions)
+    .define_attr("autotune_duration", &Args::autotuneDuration)
+    .define_attr("autotune_model_size", &Args::autotuneModelSize);
+
   define_class_under<FastText>(rb_mExt, "Model")
     .define_constructor(Constructor<FastText>())
     .define_method(
@@ -231,13 +196,12 @@ void Init_ext()
     .define_method(
       "word_vector",
       [](FastText& m, const std::string& word) {
-        int dimension = m.getDimension();
+        auto dimension = m.getDimension();
         fasttext::Vector vec = fasttext::Vector(dimension);
         m.getWordVector(vec, word);
-        float* data = vec.data();
         Array ret;
-        for (int i = 0; i < dimension; i++) {
-          ret.push(data[i]);
+        for (size_t i = 0; i < vec.size(); i++) {
+          ret.push(vec[i]);
         }
         return ret;
       })
@@ -259,20 +223,18 @@ void Init_ext()
       "sentence_vector",
       [](FastText& m, const std::string& text) {
         std::istringstream in(text);
-        int dimension = m.getDimension();
+        auto dimension = m.getDimension();
         fasttext::Vector vec = fasttext::Vector(dimension);
         m.getSentenceVector(in, vec);
-        float* data = vec.data();
         Array ret;
-        for (int i = 0; i < dimension; i++) {
-          ret.push(data[i]);
+        for (size_t i = 0; i < vec.size(); i++) {
+          ret.push(vec[i]);
         }
         return ret;
       })
     .define_method(
       "train",
-      [](FastText& m, Hash h) {
-        auto a = buildArgs(h);
+      [](FastText& m, Args& a) {
         if (a.hasAutotune()) {
           fasttext::Autotune autotune(std::shared_ptr<fasttext::FastText>(&m, [](fasttext::FastText*) {}));
           autotune.train(a);
@@ -282,8 +244,8 @@ void Init_ext()
       })
     .define_method(
       "quantize",
-      [](FastText& m, Hash h) {
-        m.quantize(buildArgs(h));
+      [](FastText& m, Args& a) {
+        m.quantize(a);
       })
     .define_method(
       "supervised?",

data/lib/fasttext/classifier.rb

@@ -30,14 +30,16 @@ module FastText
     }
 
     def fit(x, y = nil, autotune_set: nil)
-      input = input_path(x, y)
+      input, _ref = input_path(x, y)
       @m ||= Ext::Model.new
-      opts = DEFAULT_OPTIONS.merge(@options).merge(input: input, model: "supervised")
+      a = build_args(DEFAULT_OPTIONS)
+      a.input = input
+      a.model = "supervised"
       if autotune_set
         x, y = autotune_set
-        opts.merge!(autotune_validation_file: input_path(x, y))
+        a.autotune_validation_file, _autotune_ref = input_path(x, y)
       end
-      m.train(opts)
+      m.train(a)
     end
 
     def predict(text, k: 1, threshold: 0.0)
@@ -47,16 +49,16 @@ module FastText
       # TODO predict multiple in C++ for performance
       result =
         text.map do |t|
-          m.predict(prep_text(t), k, threshold).map do |v|
+          m.predict(prep_text(t), k, threshold).to_h do |v|
             [remove_prefix(v[1]), v[0]]
-          end.to_h
+          end
         end
 
       multiple ? result : result.first
     end
 
     def test(x, y = nil, k: 1)
-      input = input_path(x, y)
+      input, _ref = input_path(x, y)
       res = m.test(input, k)
       {
         examples: res[0],
@@ -67,7 +69,8 @@ module FastText
 
     # TODO support options
     def quantize
-      m.quantize({})
+      a = Ext::Args.new
+      m.quantize(a)
     end
 
     def labels(include_freq: false)
@@ -85,7 +88,7 @@ module FastText
     def input_path(x, y)
       if x.is_a?(String)
         raise ArgumentError, "Cannot pass y with file" if y
-        x
+        [x, nil]
       else
         tempfile = Tempfile.new("fasttext")
         x.zip(y) do |xi, yi|
@@ -95,7 +98,7 @@ module FastText
           tempfile.write("\n")
         end
         tempfile.close
-        tempfile.path
+        [tempfile.path, tempfile]
       end
     end
 

data/lib/fasttext/model.rb

@@ -56,5 +56,15 @@ module FastText
     def m
       @m || (raise Error, "Not fit")
     end
+
+    def build_args(default_options)
+      a = Ext::Args.new
+      opts = @options.dup
+      default_options.each do |k, v|
+        a.send("#{k}=", opts.delete(k) || v)
+      end
+      raise ArgumentError, "Unknown argument: #{opts.keys.first}" if opts.any?
+      a
+    end
   end
 end

data/lib/fasttext/vectorizer.rb

@@ -29,9 +29,10 @@ module FastText
     }
 
     def fit(x)
-      input = input_path(x)
       @m ||= Ext::Model.new
-      m.train(DEFAULT_OPTIONS.merge(@options).merge(input: input))
+      a = build_args(DEFAULT_OPTIONS)
+      a.input, _ref = input_path(x)
+      m.train(a)
     end
 
     def nearest_neighbors(word, k: 10)
@@ -48,7 +49,7 @@ module FastText
     # https://github.com/facebookresearch/fastText/issues/518
     def input_path(x)
       if x.is_a?(String)
-        x
+        [x, nil]
       else
         tempfile = Tempfile.new("fasttext")
         x.each do |xi|
@@ -56,7 +57,7 @@ module FastText
           tempfile.write("\n")
         end
         tempfile.close
-        tempfile.path
+        [tempfile.path, tempfile]
       end
     end
   end

data/lib/fasttext/version.rb

@@ -1,3 +1,3 @@
 module FastText
-  VERSION = "0.2.1"
+  VERSION = "0.2.2"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: fasttext
 version: !ruby/object:Gem::Version
-  version: 0.2.1
+  version: 0.2.2
 platform: ruby
 authors:
 - Andrew Kane
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2021-05-23 00:00:00.000000000 Z
+date: 2021-10-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rice
@@ -90,7 +90,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.2.3
+rubygems_version: 3.2.22
 signing_key:
 specification_version: 4
 summary: fastText - efficient text classification and representation learning - for