Skip to content

TextBlob Extensions

Usage

TextBlob supports adding custom models and new languages through “extensions”. Check out the TextBlob docs (https://textblob.readthedocs.io/en/dev/extensions.html#extensions) for more details.

spacytextblob also supports the use of TextBlob extensions. To use a TextBlob extension you need to pass some additional information to the config parameter when initializing the spacytextblob pipeline component.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
import spacy
from spacytextblob.spacytextblob import SpacyTextBlob
from textblob import Blobber
from textblob_fr import PatternTagger, PatternAnalyzer # 

text = u"Quelle belle matinée"

@spacy.registry.misc("spacytextblob.fr_blob") # 
def create_fr_blob():
    tb = Blobber(pos_tagger=PatternTagger(), analyzer=PatternAnalyzer())
    return tb # 

nlp_fr = spacy.load("fr_core_news_sm")

nlp_fr.add_pipe(
    "spacytextblob", # 
    config={ # 
        "custom_blob": {"@misc": "spacytextblob.fr_blob"} # 
    }
)

doc = nlp_fr(text)

print(doc)
# Quelle belle matinée
print(doc._.blob)
# Quelle belle matinée
print(doc._.blob.sentiment)
# (0.8, 0.8)

Extensions

The following extensions have been tested and are supported. Other extensions may work, but have not been tested.

textblob-fr

textblob-fr is a TextBlob extension that enables French language support for TextBlob (https://github.com/sloria/textblob-fr).

pip install textblob-fr

To use it with spacytextblob First install a spaCy model that supports French (https://spacy.io/models/fr):

python -m spacy download fr_core_news_sm

The code below demonstrates how you can then use and access textblob-fr within spacytextblob.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
import spacy
from textblob import Blobber
from textblob_fr import PatternAnalyzer, PatternTagger

from spacytextblob.spacytextblob import SpacyTextBlob  # noqa: F401


@spacy.registry.misc("spacytextblob.fr_blob")
def create_fr_blob():
    tb = Blobber(pos_tagger=PatternTagger(), analyzer=PatternAnalyzer())
    return tb


config = {"custom_blob": {"@misc": "spacytextblob.fr_blob"}}

nlp_fr = spacy.load("fr_core_news_sm")
nlp_fr.add_pipe("spacytextblob", config=config)
text = "Quelle belle matinée"
doc = nlp_fr(text)

print(doc)
# Quelle belle matinée
print(doc._.blob)
# Quelle belle matinée
print(doc._.blob.sentiment)
# (0.8, 0.8)

textblob-de

Warning

textblob-de is not supported. As of spacytextblob 4.1.0. The textblob-de library depends on a Google Translate feature that no longer works. More details can be found in this issue https://github.com/markuskiller/textblob-de/issues/24.

textblob-aptagger

Warning

textblob-aptagger is not supported. As of TextBlob 0.11.0, TextBlob uses NLTK's averaged perceptron tagger by default. This package is no longer necessary (https://github.com/sloria/textblob-aptagger).