https://github.com/sloev/spacy-syllables
Multilingual syllable annotation pipeline component for spacy
https://github.com/sloev/spacy-syllables
Last synced: over 1 year ago
JSON representation
Multilingual syllable annotation pipeline component for spacy
- Host: GitHub
- URL: https://github.com/sloev/spacy-syllables
- Owner: sloev
- License: mit
- Created: 2020-03-13T16:29:12.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2023-03-08T11:15:47.000Z (over 3 years ago)
- Last Synced: 2025-04-12T06:03:38.969Z (over 1 year ago)
- Language: Python
- Size: 160 KB
- Stars: 39
- Watchers: 5
- Forks: 2
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- Funding: .github/FUNDING.yml
- License: LICENSE
Awesome Lists containing this project
README
# Spacy Syllables
 [](https://pypi.python.org/pypi/spacy-syllables) [](https://pypi.python.org/pypi/spacy-syllables)
A [spacy 2+ pipeline component](https://spacy.io/universe/category/pipeline) for adding multilingual syllable annotation to tokens.
* Uses well established [pyphen](https://github.com/Kozea/Pyphen) for the syllables.
* Supports [a ton of languages](https://github.com/Kozea/Pyphen/tree/master/pyphen/dictionaries)
* Ease of use thx to the awesome pipeline framework in spacy
## Install
```bash
$ pip install spacy_syllables
```
which also installs the following dependencies:
* spacy = "^2.2.3"
* pyphen = "^0.9.5"
## Usage
The [`SpacySyllables`](spacy_syllables/__init__.py) class autodetects language from the given spacy nlp instance, but you can also override the detected language by specifying the `lang` parameter during instantiation, see how [here](tests/test_all.py).
### Normal usecase
```python
import spacy
from spacy_syllables import SpacySyllables
nlp = spacy.load("en_core_web_sm")
nlp.add_pipe("syllables", after="tagger")
assert nlp.pipe_names == ["tok2vec", "tagger", "syllables", "parser", "ner", "attribute_ruler", "lemmatizer"]
doc = nlp("terribly long")
data = [(token.text, token._.syllables, token._.syllables_count) for token in doc]
assert data == [("terribly", ["ter", "ri", "bly"], 3), ("long", ["long"], 1)]
```
more examples in [tests](tests/test_all.py)
## Migrating from spacy 2.x to 3.0
In spacy 2.x, spacy_syllables was originally added to the pipeline by instantiating a [`SpacySyllables`](spacy_syllables/__init__.py) object with the desired options and adding it to the pipeline:
```python
from spacy_syllables import SpacySyllables
syllables = SpacySyllables(nlp, "en_US")
nlp.add_pipe(syllables, after="tagger")
```
In spacy 3.0, you now add the component to the pipeline simply by adding it by name, setting custom configuration information in the `add_pipe()` parameters:
```python
from spacy_syllables import SpacySyllables
nlp.add_pipe("syllables", after="tagger", config={"lang": "en_US"})
```
In addition, the default pipeline components have changed between 2.x and 3.0; please make sure to update any asserts you have that check for these.
e.g.:
spacy 2.x:
```python
assert nlp.pipe_names == ["tagger", "syllables", "parser", "ner"]
```
spacy 3.0:
```python
assert nlp.pipe_names == ["tok2vec", "tagger", "syllables", "parser", "ner", "attribute_ruler", "lemmatizer"]
```
## Dev setup / testing
### install
install the dev package and pyenv versions
```bash
$ pip install -e ".[dev]"
$ python -m spacy download en_core_web_sm
```
### run tests
```bash
$ black .
$ pytest
```