Python - Expose Whitespace PreTokenizer

This commit is contained in:
Anthony MOI
2019-12-30 13:10:33 -05:00
parent 4677a09626
commit 225a886382
2 changed files with 13 additions and 0 deletions

View File

@@ -35,6 +35,7 @@ fn models(_py: Python, m: &PyModule) -> PyResult<()> {
fn pre_tokenizers(_py: Python, m: &PyModule) -> PyResult<()> {
m.add_class::<pre_tokenizers::PreTokenizer>()?;
m.add_class::<pre_tokenizers::ByteLevel>()?;
m.add_class::<pre_tokenizers::Whitespace>()?;
m.add_class::<pre_tokenizers::BertPreTokenizer>()?;
Ok(())
}

View File

@@ -52,6 +52,18 @@ impl ByteLevel {
}
}
#[pyclass]
pub struct Whitespace {}
#[pymethods]
impl Whitespace {
#[staticmethod]
fn new() -> PyResult<PreTokenizer> {
Ok(PreTokenizer {
pretok: Container::Owned(Box::new(tk::pre_tokenizers::whitespace::Whitespace)),
})
}
}
#[pyclass]
pub struct BertPreTokenizer {}
#[pymethods]