scrutineer

Hive Post performance and quality analytics.

Installation

$ pip install hive-scrutineer

Supported Python versions 3.10+.

Blockchain Support

Limited support to Hive Blockchain HF27.

Basic usage

from scrutineer import Scrutineer

analyzer = Scrutineer()
analyzer.set_weights()
analysis = analyzer.analyze("author", "post-permlink")

Customizations

import json
from nektar import Waggle
from scrutineer import Scrutineer

hive = Waggle("username")

analyzer = Scrutineer(minimum_score=10, max_emojis=0, deep=True, full=False)
analyzer.set_weights(title=5, body=6, emojis=4, images=2, tagging=3, tags=1)

for blog in hive.blogs(limit=5)
    analysis = analyzer.analyze(blog)
    print(json.dumps(analysis, indent=2))

Keywords

import json
from nektar import Waggle
from scrutineer import get_keywords, get_bigrams

hive = Waggle("username")
for blog in hive.blogs(limit=5)
    keywords = get_keywords(blog["body"])
    print("\nget_keywords" + json.dumps(keywords))
    
    keywords = get_bigrams(blog["body"])
    print("\nget_bigrams:" + json.dumps(keywords))

Performance

In version 1.3.0, we've migrated to langdetect to speed up Scrutineer.analyze() by more than 300x versus version 1.2.*!

MIN: 0.02782490011304617
AVG: 0.13722777900053187
MAX: 4.033556599984877

We've also seen a 70% import speed increase, in the said profiling.