phzhik/gault-millau
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

fix regexp for find images

Browse Source
This commit is contained in:
a.gorbunov 2020-01-20 10:58:00 +00:00
parent df601e3234
commit a31763c550
2 changed files with 1 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
apps/news/management/commands/news_optimize_images.py
View File

@ -1,7 +1,6 @@
# coding=utf-8 # coding=utf-8
from django.core.management.base import BaseCommand from django.core.management.base import BaseCommand
from project.settings import SCHEMA_URI, SITE_DOMAIN_URI
from utils.methods import get_url_images_in_text, get_image_meta_by_url from utils.methods import get_url_images_in_text, get_image_meta_by_url
from news.models import News from news.models import News
from sorl.thumbnail import get_thumbnail from sorl.thumbnail import get_thumbnail
@ -30,11 +29,6 @@ class Command(BaseCommand):
def optimize(self, text, max_size, max_quality): def optimize(self, text, max_size, max_quality):
"""optimize news images""" """optimize news images"""
for image in get_url_images_in_text(text): for image in get_url_images_in_text(text):
if not image.startswith('http'):
image = f'{SCHEMA_URI}://{SITE_DOMAIN_URI}{image}'
self.stdout.write(self.style.SUCCESS(f'{image} {text}'))
try: try:
size, width, height = get_image_meta_by_url(image) size, width, height = get_image_meta_by_url(image)
except IOError as ie: except IOError as ie:

2
apps/utils/methods.py
View File

@ -207,7 +207,7 @@ def section_name_into_index_name(section_name: str):
def get_url_images_in_text(text): def get_url_images_in_text(text):
"""Find images urls in text""" """Find images urls in text"""
return re.findall(r'[^\"\'=\s]+\.jpe?g|png|gif|svg', text) return re.findall(r'\<img.+src=(?!https?:\/\/)([^\/].+?)[\"|\']>', text)
def get_image_meta_by_url(url) -> (int, int, int): def get_image_meta_by_url(url) -> (int, int, int):