Implement __new__ on Decoders

Allow decoders to be initialized from python using the class
constructor.

bindings/python/src/decoders.rs

@@ -26,25 +26,25 @@ impl Decoder {
     }
 }
 
-#[pyclass]
+#[pyclass(extends=Decoder)]
 pub struct ByteLevel {}
 #[pymethods]
 impl ByteLevel {
-    #[staticmethod]
-    fn new() -> PyResult<Decoder> {
-        Ok(Decoder {
+    #[new]
+    fn new(obj: &PyRawObject) -> PyResult<()> {
+        Ok(obj.init(Decoder {
             decoder: Container::Owned(Box::new(tk::decoders::byte_level::ByteLevel::new(false))),
-        })
+        }))
     }
 }
 
-#[pyclass]
+#[pyclass(extends=Decoder)]
 pub struct WordPiece {}
 #[pymethods]
 impl WordPiece {
-    #[staticmethod]
+    #[new]
     #[args(kwargs="**")]
-    fn new(kwargs: Option<&PyDict>) -> PyResult<Decoder> {
+    fn new(obj: &PyRawObject, kwargs: Option<&PyDict>) -> PyResult<()> {
         let mut prefix = String::from("##");
 
         if let Some(kwargs) = kwargs {
@@ -53,19 +53,19 @@ impl WordPiece {
             }
         }
 
-        Ok(Decoder {
+        Ok(obj.init(Decoder {
             decoder: Container::Owned(Box::new(tk::decoders::wordpiece::WordPiece::new(prefix))),
-        })
+        }))
     }
 }
 
-#[pyclass]
+#[pyclass(extends=Decoder)]
 pub struct Metaspace {}
 #[pymethods]
 impl Metaspace {
-    #[staticmethod]
+    #[new]
     #[args(kwargs = "**")]
-    fn new(kwargs: Option<&PyDict>) -> PyResult<Decoder> {
+    fn new(obj: &PyRawObject, kwargs: Option<&PyDict>) -> PyResult<()> {
         let mut replacement = '▁';
         let mut add_prefix_space = true;
 
@@ -85,22 +85,22 @@ impl Metaspace {
             }
         }
 
-        Ok(Decoder {
+        Ok(obj.init(Decoder {
             decoder: Container::Owned(Box::new(tk::decoders::metaspace::Metaspace::new(
                 replacement,
                 add_prefix_space,
             ))),
-        })
+        }))
     }
 }
 
-#[pyclass]
+#[pyclass(extends=Decoder)]
 pub struct BPEDecoder {}
 #[pymethods]
 impl BPEDecoder {
-    #[staticmethod]
+    #[new]
     #[args(kwargs = "**")]
-    fn new(kwargs: Option<&PyDict>) -> PyResult<Decoder> {
+    fn new(obj: &PyRawObject, kwargs: Option<&PyDict>) -> PyResult<()> {
         let mut suffix = String::from("</w");
 
         if let Some(kwargs) = kwargs {
@@ -113,9 +113,9 @@ impl BPEDecoder {
             }
         }
 
-        Ok(Decoder {
+        Ok(obj.init(Decoder {
             decoder: Container::Owned(Box::new(tk::decoders::bpe::BPEDecoder::new(suffix))),
-        })
+        }))
     }
 }
 

bindings/python/tokenizers/decoders/__init__.pyi

@@ -14,8 +14,7 @@ class Decoder:
 class ByteLevel:
     """ ByteLevel Decoder """
 
-    @staticmethod
-    def new() -> Decoder:
+    def __init__(self) -> None:
         """ Instantiate a new ByteLevel Decoder """
         pass
 
@@ -23,7 +22,7 @@ class WordPiece:
     """ WordPiece Decoder """
 
     @staticmethod
-    def new(prefix: str="##") -> Decoder:
+    def __init__(self, prefix: str = "##") -> Decoder:
         """ Instantiate a new WordPiece Decoder
 
         Args:
@@ -35,9 +34,7 @@ class WordPiece:
 class Metaspace:
     """ Metaspace decoder """
 
-    @staticmethod
-    def new(replacement: str="▁",
-            add_prefix_space: bool=True) -> Decoder:
+    def __init__(self, replacement: str = "▁", add_prefix_space: bool = True) -> None:
         """ Instantiate a new Metaspace
 
         Args:
@@ -54,8 +51,7 @@ class Metaspace:
 class BPEDecoder:
     """ BPEDecoder """
 
-    @staticmethod
-    def new(suffix: str="</w>") -> Decoder:
+    def __init__(self, suffix: str = "</w>") -> None:
         """ Instantiate a new BPEDecoder
 
         Args:

bindings/python/tokenizers/implementations/bert_wordpiece.py

@@ -48,7 +48,7 @@ class BertWordPieceTokenizer(BaseTokenizer):
                 (sep_token, sep_token_id),
                 (cls_token, cls_token_id)
             )
-        tokenizer.decoders = decoders.WordPiece.new(prefix=wordpieces_prefix)
+        tokenizer.decoders = decoders.WordPiece(prefix=wordpieces_prefix)
 
         parameters = {
             "model": "BertWordPiece",

bindings/python/tokenizers/implementations/byte_level_bpe.py

@@ -46,7 +46,7 @@ class ByteLevelBPETokenizer(BaseTokenizer):
                 tokenizer.normalizer = normalizers[0]
 
         tokenizer.pre_tokenizer = pre_tokenizers.ByteLevel.new(add_prefix_space=add_prefix_space)
-        tokenizer.decoder = decoders.ByteLevel.new()
+        tokenizer.decoder = decoders.ByteLevel()
 
         parameters = {
             "model": "ByteLevelBPE",

bindings/python/tokenizers/implementations/char_level_bpe.py

@@ -52,7 +52,7 @@ class CharBPETokenizer(BaseTokenizer):
                 tokenizer.normalizer = normalizers[0]
 
         tokenizer.pre_tokenizer = pre_tokenizers.WhitespaceSplit.new()
-        tokenizer.decoder = decoders.BPEDecoder.new(suffix=suffix)
+        tokenizer.decoder = decoders.BPEDecoder(suffix=suffix)
 
         parameters = {
             "model": "BPE",

bindings/python/tokenizers/implementations/sentencepiece_bpe.py

@@ -31,8 +31,9 @@ class SentencePieceBPETokenizer(BaseTokenizer):
         tokenizer.normalizer = NFKC()
         tokenizer.pre_tokenizer = pre_tokenizers.Metaspace.new(replacement=replacement,
                                                                add_prefix_space=add_prefix_space)
-        tokenizer.decoder = decoders.Metaspace.new(replacement=replacement,
-                                                   add_prefix_space=add_prefix_space)
+        tokenizer.decoder = decoders.Metaspace(
+            replacement=replacement, add_prefix_space=add_prefix_space
+        )
 
         parameters = {
             "model": "SentencePieceBPE",