.. comment:: bumpversion marker
- Fixed automated release due to not matching registered email.
- Fixes bad formatting in logging message.
- Fixes wrong message on dupefilter duplicates.
- Fixed typo in default settings.
- Fixed data decoding in Python 3.x.
- Added
REDIS_ENCODING
setting (default utf-8
).
- Default to
CONCURRENT_REQUESTS
value for REDIS_START_URLS_BATCH_SIZE
.
- Renamed queue classes to a proper naming conventiong (backwards compatible).
- Added
REDIS_START_URLS_KEY
setting.
- Fixed spider method
from_crawler
signature.
- Support
redis_cls
parameter in REDIS_PARAMS
setting.
- Python 3.x compatibility fixed.
- Added
SCHEDULER_SERIALIZER
setting.
- Backwards incompatible change: Require explicit
DUPEFILTER_CLASS
setting.
- Added
SCHEDULER_FLUSH_ON_START
setting.
- Added
REDIS_START_URLS_AS_SET
setting.
- Added
REDIS_ITEMS_KEY
setting.
- Added
REDIS_ITEMS_SERIALIZER
setting.
- Added
REDIS_PARAMS
setting.
- Added
REDIS_START_URLS_BATCH_SIZE
spider attribute to read start urls
in batches.
- Added
RedisCrawlSpider
.
- Updated code to be compatible with Scrapy 1.0.
- Added -a domain=... option for example spiders.
- Added REDIS_URL setting to support Redis connection string.
- Added SCHEDULER_IDLE_BEFORE_CLOSE setting to prevent the spider closing too
quickly when the queue is empty. Default value is zero keeping the previous
behavior.
- Schedule preemptively requests on item scraped.
- This version is the latest release compatible with Scrapy 0.24.x.
- Added RedisSpider and RedisMixin classes as building blocks for spiders
to be fed through a redis queue.
- Added redis queue stats.
- Let the encoder handle the item as it comes instead converting it to a dict.
- Added support for different queue classes.
- Changed requests serialization from marshal to cPickle.
- Improved backward compatibility.
- Added example project.