nlpodyssey/gotokenizers

// Copyright (c) 2020, NLP Odyssey Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.

package sequencenormalizer

import (
    "github.com/nlpodyssey/gotokenizers/normalizedstring"
    "github.com/nlpodyssey/gotokenizers/normalizers"
)

// SequenceNormalizer allows concatenating multiple other Normalizers as a
// Sequence.
type SequenceNormalizer struct{ normalizers []normalizers.Normalizer }

var _ normalizers.Normalizer = &SequenceNormalizer{}

// NewSequenceNormalizer returns a new SequenceNormalizer, initializing it
// with the ordered sequence of Normalizers.
func NewSequenceNormalizer(normalizers []normalizers.Normalizer) *SequenceNormalizer {
    return &SequenceNormalizer{normalizers: normalizers}
}

// Normalize transform the NormalizedString running the ordered sequence of
// normalizers (against the same NormalizedString).
//
// If one Normalizer returns an error, the same error is returned and
// the subsequent Normalizers (if any) are ignored.
func (sn *SequenceNormalizer) Normalize(ns *normalizedstring.NormalizedString) error {
    for _, normalizer := range sn.normalizers {
        err := normalizer.Normalize(ns)
        if err != nil {
            return err
        }
    }
    return nil
}