Rich text benchmarking - management command

results.txt

On a real-world site with 3058 HTML fragments

with richtext-benchmark/original:

1. 13.098440
2. 14.320241
3. 13.525923

with richtext-benchmark/loic:

1. 32.462429
2. 34.154890
3. 34.121001

with richtext-benchmark/matthew:

1. 13.270576
2. 13.325513
3. 15.039238

rich_text_benchmark.py

import time

from django.core.management.base import BaseCommand

from wagtail.core.blocks import RichTextBlock
from wagtail.core.models import get_page_models
from wagtail.core.fields import RichTextField, StreamField
from wagtail.core.rich_text import expand_db_html


class Command(BaseCommand):
    def handle(self, **options):
        print("Building test corpus...")
        corpus = []

        for page_model in get_page_models():
            rich_text_fields = []
            stream_fields = []

            for field in page_model._meta.get_fields():
                if isinstance(field, RichTextField):
                    rich_text_fields.append(field.name)
                elif isinstance(field, StreamField):
                    stream_fields.append(field.name)

            for page in page_model.objects.all():
                for field in rich_text_fields:
                    corpus.append(getattr(page, field))
                for field in stream_fields:
                    for block in getattr(page, field):
                        if isinstance(block.block, RichTextBlock):
                            corpus.append(block.value.source)

        print("Built test corpus of %d HTML fragments." % len(corpus))

        start = time.time()
        for i, html in enumerate(corpus):
            if i % 1000 == 0:
                print("processed %d fragments" % i)
            expand_db_html(html)
        end = time.time()

        print("Completed in %fs" % (end - start))