diff --git a/mwmbl/crawler/stats.py b/mwmbl/crawler/stats.py new file mode 100644 index 0000000..c71973c --- /dev/null +++ b/mwmbl/crawler/stats.py @@ -0,0 +1,77 @@ +import gzip +import json +from datetime import datetime +from glob import glob +from itertools import islice +from logging import getLogger +from urllib.parse import urlparse + +from redis import Redis + +from mwmbl.crawler.batch import HashedBatch +from mwmbl.indexer.update_urls import get_datetime_from_timestamp + +logger = getLogger(__name__) + +URL_DATE_COUNT_KEY = "url-count-{date}" +URL_HOUR_COUNT_KEY = "url-count-hour-{hour}" +USER_COUNT_KEY = "user-count-{date}" +HOST_COUNT_KEY = "host-count-{date}" +EXPIRE_SECONDS = 60*60*24 + + +class StatsManager: + def __init__(self, redis: Redis): + self.redis = redis + + def record_batch(self, hashed_batch: HashedBatch): + date_time = get_datetime_from_timestamp(hashed_batch.timestamp) + + num_crawled_urls = sum(1 for item in hashed_batch.items if item.content is not None) + + url_count_key = URL_DATE_COUNT_KEY.format(date=date_time.date) + self.redis.incrby(url_count_key, num_crawled_urls) + self.redis.expire(url_count_key, EXPIRE_SECONDS) + + hour = datetime(date_time.year, date_time.month, date_time.day, date_time.hour) + hour_key = URL_HOUR_COUNT_KEY.format(hour=hour) + self.redis.incrby(hour_key, num_crawled_urls) + self.redis.expire(hour_key, EXPIRE_SECONDS) + + user_count_key = USER_COUNT_KEY.format(date=date_time.date) + self.redis.zincrby(user_count_key, num_crawled_urls, hashed_batch.user_id_hash) + self.redis.expire(user_count_key, EXPIRE_SECONDS) + + host_key = HOST_COUNT_KEY.format(date=date_time.date) + for item in hashed_batch.items: + if item.content is None: + continue + + host = urlparse(item.url).netloc + self.redis.zincrby(host_key, 1, host) + self.redis.expire(host_key, EXPIRE_SECONDS) + + def get_stats(self): + pass + + +def get_test_batches(): + for path in glob("./devdata/batches/**/*.json.gz", recursive=True): + print("Processing path", path) + with gzip.open(path) as gzip_file: + yield HashedBatch.parse_raw(gzip_file.read()) + + +if __name__ == '__main__': + redis = Redis(host='localhost', port=6379, decode_responses=True) + stats = StatsManager(redis) + batches = get_test_batches() + start = datetime.now() + processed = 0 + for batch in islice(batches, 100): + stats.record_batch(batch) + processed += 1 + total_time = (datetime.now() - start).total_seconds() + print("Processed", processed) + print("Total time", total_time) + print("Time per batch", total_time/processed) diff --git a/poetry.lock b/poetry.lock index 6c15500..f2e3102 100644 --- a/poetry.lock +++ b/poetry.lock @@ -40,6 +40,18 @@ typing-extensions = {version = ">=4", markers = "python_version < \"3.11\""} [package.extras] tests = ["mypy (>=0.800)", "pytest", "pytest-asyncio"] +[[package]] +name = "async-timeout" +version = "4.0.3" +description = "Timeout context manager for asyncio programs" +category = "main" +optional = false +python-versions = ">=3.7" +files = [ + {file = "async-timeout-4.0.3.tar.gz", hash = "sha256:4640d96be84d82d02ed59ea2b7105a0f7b33abe8703703cd0ab0bf87c427522f"}, + {file = "async_timeout-4.0.3-py3-none-any.whl", hash = "sha256:7405140ff1230c310e51dc27b3145b9092d659ce68ff733fb0cefe3ee42be028"}, +] + [[package]] name = "attrs" version = "23.1.0" @@ -538,6 +550,105 @@ files = [ {file = "h11-0.14.0.tar.gz", hash = "sha256:8f19fbbe99e72420ff35c00b27a34cb9937e902a8b810e2c88300c6f0a3b699d"}, ] +[[package]] +name = "hiredis" +version = "2.2.3" +description = "Python wrapper for hiredis" +category = "main" +optional = false +python-versions = ">=3.7" +files = [ + {file = "hiredis-2.2.3-cp310-cp310-macosx_10_12_universal2.whl", hash = "sha256:9a1a80a8fa767f2fdc3870316a54b84fe9fc09fa6ab6a2686783de6a228a4604"}, + {file = "hiredis-2.2.3-cp310-cp310-macosx_10_12_x86_64.whl", hash = "sha256:3f006c28c885deb99b670a5a66f367a175ab8955b0374029bad7111f5357dcd4"}, + {file = "hiredis-2.2.3-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:ffaf841546905d90ff189de7397aa56413b1ce5e54547f17a98f0ebf3a3b0a3b"}, + {file = "hiredis-2.2.3-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:1cadb0ac7ba3babfd804e425946bec9717b320564a1390f163a54af9365a720a"}, + {file = "hiredis-2.2.3-cp310-cp310-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:33bc4721632ef9708fa44e5df0066053fccc8e65410a2c48573192517a533b48"}, + {file = "hiredis-2.2.3-cp310-cp310-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:227c5b4bcb60f89008c275d596e4a7b6625a6b3c827b8a66ae582eace7051f71"}, + {file = "hiredis-2.2.3-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:61995eb826009d99ed8590747bc0da683a5f4fbb4faa8788166bf3810845cd5c"}, + {file = "hiredis-2.2.3-cp310-cp310-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:6f969edc851efe23010e0f53a64269f2629a9364135e9ec81c842e8b2277d0c1"}, + {file = "hiredis-2.2.3-cp310-cp310-musllinux_1_1_aarch64.whl", hash = "sha256:d27e560eefb57914d742a837f1da98d3b29cb22eff013c8023b7cf52ae6e051d"}, + {file = "hiredis-2.2.3-cp310-cp310-musllinux_1_1_i686.whl", hash = "sha256:3759f4789ae1913b7df278dfc9e8749205b7a106f888cd2903d19461e24a7697"}, + {file = "hiredis-2.2.3-cp310-cp310-musllinux_1_1_ppc64le.whl", hash = "sha256:c6cb613148422c523945cdb8b6bed617856f2602fd8750e33773ede2616e55d5"}, + {file = "hiredis-2.2.3-cp310-cp310-musllinux_1_1_s390x.whl", hash = "sha256:1d274d5c511dfc03f83f997d3238eaa9b6ee3f982640979f509373cced891e98"}, + {file = "hiredis-2.2.3-cp310-cp310-musllinux_1_1_x86_64.whl", hash = "sha256:3b7fe075e91b9d9cff40eba4fb6a8eff74964d3979a39be9a9ef58b1b4cb3604"}, + {file = "hiredis-2.2.3-cp310-cp310-win32.whl", hash = "sha256:77924b0d32fd1f493d3df15d9609ddf9d94c31a364022a6bf6b525ce9da75bea"}, + {file = "hiredis-2.2.3-cp310-cp310-win_amd64.whl", hash = "sha256:dcb0569dd5bfe6004658cd0f229efa699a3169dcb4f77bd72e188adda302063d"}, + {file = "hiredis-2.2.3-cp311-cp311-macosx_10_12_universal2.whl", hash = "sha256:d115790f18daa99b5c11a506e48923b630ef712e9e4b40482af942c3d40638b8"}, + {file = "hiredis-2.2.3-cp311-cp311-macosx_10_12_x86_64.whl", hash = "sha256:4c3b8be557e08b234774925622e196f0ee36fe4eab66cd19df934d3efd8f3743"}, + {file = "hiredis-2.2.3-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:3f5446068197b35a11ccc697720c41879c8657e2e761aaa8311783aac84cef20"}, + {file = "hiredis-2.2.3-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:aa17a3b22b3726d54d7af20394f65d4a1735a842a4e0f557dc67a90f6965c4bc"}, + {file = "hiredis-2.2.3-cp311-cp311-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:7df645b6b7800e8b748c217fbd6a4ca8361bcb9a1ae6206cc02377833ec8a1aa"}, + {file = "hiredis-2.2.3-cp311-cp311-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:2fb9300959a0048138791f3d68359d61a788574ec9556bddf1fec07f2dbc5320"}, + {file = "hiredis-2.2.3-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:2d7e459fe7313925f395148d36d9b7f4f8dac65be06e45d7af356b187cef65fc"}, + {file = "hiredis-2.2.3-cp311-cp311-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:8eceffca3941775b646cd585cd19b275d382de43cc3327d22f7c75d7b003d481"}, + {file = "hiredis-2.2.3-cp311-cp311-musllinux_1_1_aarch64.whl", hash = "sha256:b17baf702c6e5b4bb66e1281a3efbb1d749c9d06cdb92b665ad81e03118f78fc"}, + {file = "hiredis-2.2.3-cp311-cp311-musllinux_1_1_i686.whl", hash = "sha256:4e43e2b5acaad09cf48c032f7e4926392bb3a3f01854416cf6d82ebff94d5467"}, + {file = "hiredis-2.2.3-cp311-cp311-musllinux_1_1_ppc64le.whl", hash = "sha256:a7205497d7276a81fe92951a29616ef96562ed2f91a02066f72b6f93cb34b40e"}, + {file = "hiredis-2.2.3-cp311-cp311-musllinux_1_1_s390x.whl", hash = "sha256:126623b03c31cb6ac3e0d138feb6fcc36dd43dd34fc7da7b7a0c38b5d75bc896"}, + {file = "hiredis-2.2.3-cp311-cp311-musllinux_1_1_x86_64.whl", hash = "sha256:071c5814b850574036506a8118034f97c3cbf2fe9947ff45a27b07a48da56240"}, + {file = "hiredis-2.2.3-cp311-cp311-win32.whl", hash = "sha256:d1be9e30e675f5bc1cb534633324578f6f0944a1bcffe53242cf632f554f83b6"}, + {file = "hiredis-2.2.3-cp311-cp311-win_amd64.whl", hash = "sha256:b9a7c987e161e3c58f992c63b7e26fea7fe0777f3b975799d23d65bbb8cb5899"}, + {file = "hiredis-2.2.3-cp37-cp37m-macosx_10_12_x86_64.whl", hash = "sha256:f2dcb8389fa3d453927b1299f46bdb38473c293c8269d5c777d33ea0e526b610"}, + {file = "hiredis-2.2.3-cp37-cp37m-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:a2df98f5e071320c7d84e8bd07c0542acdd0a7519307fc31774d60e4b842ec4f"}, + {file = "hiredis-2.2.3-cp37-cp37m-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:61a72e4a523cdfc521762137559c08dfa360a3caef63620be58c699d1717dac1"}, + {file = "hiredis-2.2.3-cp37-cp37m-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:c9b9e5bde7030cae83aa900b5bd660decc65afd2db8c400f3c568c815a47ca2a"}, + {file = "hiredis-2.2.3-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:cd2614f17e261f72efc2f19f5e5ff2ee19e2296570c0dcf33409e22be30710de"}, + {file = "hiredis-2.2.3-cp37-cp37m-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:46525fbd84523cac75af5bf524bc74aaac848beaf31b142d2df8a787d9b4bbc4"}, + {file = "hiredis-2.2.3-cp37-cp37m-musllinux_1_1_aarch64.whl", hash = "sha256:d1a4ce40ba11da9382c14da31f4f9e88c18f7d294f523decd0fadfb81f51ad18"}, + {file = "hiredis-2.2.3-cp37-cp37m-musllinux_1_1_i686.whl", hash = "sha256:5cda592405bbd29d53942e0389dc3fa77b49c362640210d7e94a10c14a677d4d"}, + {file = "hiredis-2.2.3-cp37-cp37m-musllinux_1_1_ppc64le.whl", hash = "sha256:5e6674a017629284ef373b50496d9fb1a89b85a20a7fa100ecd109484ec748e5"}, + {file = "hiredis-2.2.3-cp37-cp37m-musllinux_1_1_s390x.whl", hash = "sha256:e62ec131816c6120eff40dffe43424e140264a15fa4ab88c301bd6a595913af3"}, + {file = "hiredis-2.2.3-cp37-cp37m-musllinux_1_1_x86_64.whl", hash = "sha256:17e938d9d3ee92e1adbff361706f1c36cc60eeb3e3eeca7a3a353eae344f4c91"}, + {file = "hiredis-2.2.3-cp37-cp37m-win32.whl", hash = "sha256:95d2305fd2a7b179cacb48b10f618872fc565c175f9f62b854e8d1acac3e8a9e"}, + {file = "hiredis-2.2.3-cp37-cp37m-win_amd64.whl", hash = "sha256:8f9dbe12f011a9b784f58faecc171d22465bb532c310bd588d769ba79a59ef5a"}, + {file = "hiredis-2.2.3-cp38-cp38-macosx_10_12_universal2.whl", hash = "sha256:5a4bcef114fc071d5f52c386c47f35aae0a5b43673197b9288a15b584da8fa3a"}, + {file = "hiredis-2.2.3-cp38-cp38-macosx_10_12_x86_64.whl", hash = "sha256:232d0a70519865741ba56e1dfefd160a580ae78c30a1517bad47b3cf95a3bc7d"}, + {file = "hiredis-2.2.3-cp38-cp38-macosx_11_0_arm64.whl", hash = "sha256:9076ce8429785c85f824650735791738de7143f61f43ae9ed83e163c0ca0fa44"}, + {file = "hiredis-2.2.3-cp38-cp38-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:ec58fb7c2062f835595c12f0f02dcda76d0eb0831423cc191d1e18c9276648de"}, + {file = "hiredis-2.2.3-cp38-cp38-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:7f2b34a6444b8f9c1e9f84bd2c639388e5d14f128afd14a869dfb3d9af893aa2"}, + {file = "hiredis-2.2.3-cp38-cp38-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:818dfd310aa1020a13cd08ee48e116dd8c3bb2e23b8161f8ac4df587dd5093d7"}, + {file = "hiredis-2.2.3-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:96d9ea6c8d4cbdeee2e0d43379ce2881e4af0454b00570677c59f33f2531cd38"}, + {file = "hiredis-2.2.3-cp38-cp38-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:f1eadbcd3de55ac42310ff82550d3302cb4efcd4e17d76646a17b6e7004bb42b"}, + {file = "hiredis-2.2.3-cp38-cp38-musllinux_1_1_aarch64.whl", hash = "sha256:477c34c4489666dc73cb5e89dafe2617c3e13da1298917f73d55aac4696bd793"}, + {file = "hiredis-2.2.3-cp38-cp38-musllinux_1_1_i686.whl", hash = "sha256:14824e457e4f5cda685c3345d125da13949bcf3bb1c88eb5d248c8d2c3dee08f"}, + {file = "hiredis-2.2.3-cp38-cp38-musllinux_1_1_ppc64le.whl", hash = "sha256:9cd32326dfa6ce87edf754153b0105aca64486bebe93b9600ccff74fa0b224df"}, + {file = "hiredis-2.2.3-cp38-cp38-musllinux_1_1_s390x.whl", hash = "sha256:51341e70b467004dcbec3a6ce8c478d2d6241e0f6b01e4c56764afd5022e1e9d"}, + {file = "hiredis-2.2.3-cp38-cp38-musllinux_1_1_x86_64.whl", hash = "sha256:2443659c76b226267e2a04dbbb21bc2a3f91aa53bdc0c22964632753ae43a247"}, + {file = "hiredis-2.2.3-cp38-cp38-win32.whl", hash = "sha256:4e3e3e31423f888d396b1fc1f936936e52af868ac1ec17dd15e3eeba9dd4de24"}, + {file = "hiredis-2.2.3-cp38-cp38-win_amd64.whl", hash = "sha256:20f509e3a1a20d6e5f5794fc37ceb21f70f409101fcfe7a8bde783894d51b369"}, + {file = "hiredis-2.2.3-cp39-cp39-macosx_10_12_universal2.whl", hash = "sha256:d20891e3f33803b26d54c77fd5745878497091e33f4bbbdd454cf6e71aee8890"}, + {file = "hiredis-2.2.3-cp39-cp39-macosx_10_12_x86_64.whl", hash = "sha256:50171f985e17970f87d5a29e16603d1e5b03bdbf5c2691a37e6c912942a6b657"}, + {file = "hiredis-2.2.3-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:9944a2cac25ffe049a7e89f306e11b900640837d1ef38d9be0eaa4a4e2b73a52"}, + {file = "hiredis-2.2.3-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:5a5c8019ff94988d56eb49b15de76fe83f6b42536d76edeb6565dbf7fe14b973"}, + {file = "hiredis-2.2.3-cp39-cp39-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:a286ded34eb16501002e3713b3130c987366eee2ba0d58c33c72f27778e31676"}, + {file = "hiredis-2.2.3-cp39-cp39-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:4b3e974ad15eb32b1f537730dea70b93a4c3db7b026de3ad2b59da49c6f7454d"}, + {file = "hiredis-2.2.3-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:08415ea74c1c29b9d6a4ca3dd0e810dc1af343c1d1d442e15ba133b11ab5be6a"}, + {file = "hiredis-2.2.3-cp39-cp39-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:7e17d04ea58ab8cf3f2dc52e875db16077c6357846006780086fff3189fb199d"}, + {file = "hiredis-2.2.3-cp39-cp39-musllinux_1_1_aarch64.whl", hash = "sha256:6ccdcb635dae85b006592f78e32d97f4bc7541cb27829d505f9c7fefcef48298"}, + {file = "hiredis-2.2.3-cp39-cp39-musllinux_1_1_i686.whl", hash = "sha256:69536b821dd1bc78058a6e7541743f8d82bf2d981b91280b14c4daa6cdc7faba"}, + {file = "hiredis-2.2.3-cp39-cp39-musllinux_1_1_ppc64le.whl", hash = "sha256:3753df5f873d473f055e1f8837bfad0bd3b277c86f3c9bf058c58f14204cd901"}, + {file = "hiredis-2.2.3-cp39-cp39-musllinux_1_1_s390x.whl", hash = "sha256:6f88cafe46612b6fa68e6dea49e25bebf160598bba00101caa51cc8c1f18d597"}, + {file = "hiredis-2.2.3-cp39-cp39-musllinux_1_1_x86_64.whl", hash = "sha256:33ee3ea5cad3a8cb339352cd230b411eb437a2e75d7736c4899acab32056ccdb"}, + {file = "hiredis-2.2.3-cp39-cp39-win32.whl", hash = "sha256:b4f3d06dc16671b88a13ae85d8ca92534c0b637d59e49f0558d040a691246422"}, + {file = "hiredis-2.2.3-cp39-cp39-win_amd64.whl", hash = "sha256:4f674e309cd055ee7a48304ceb8cf43265d859faf4d7d01d270ce45e976ae9d3"}, + {file = "hiredis-2.2.3-pp37-pypy37_pp73-macosx_10_12_x86_64.whl", hash = "sha256:8f280ab4e043b089777b43b4227bdc2035f88da5072ab36588e0ccf77d45d058"}, + {file = "hiredis-2.2.3-pp37-pypy37_pp73-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:15c2a551f3b8a26f7940d6ee10b837810201754b8d7e6f6b1391655370882c5a"}, + {file = "hiredis-2.2.3-pp37-pypy37_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:60c4e3c258eafaab21b174b17270a0cc093718d61cdbde8c03f85ec4bf835343"}, + {file = "hiredis-2.2.3-pp37-pypy37_pp73-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:cc36a9dded458d4e37492fe3e619c6c83caae794d26ad925adbce61d592f8428"}, + {file = "hiredis-2.2.3-pp37-pypy37_pp73-win_amd64.whl", hash = "sha256:4ed68a3b1ccb4313d2a42546fd7e7439ad4745918a48b6c9bcaa61e1e3e42634"}, + {file = "hiredis-2.2.3-pp38-pypy38_pp73-macosx_10_12_x86_64.whl", hash = "sha256:3bf4b5bae472630c229518e4a814b1b68f10a3d9b00aeaec45f1a330f03a0251"}, + {file = "hiredis-2.2.3-pp38-pypy38_pp73-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:33a94d264e6e12a79d9bb8af333b01dc286b9f39c99072ab5fef94ce1f018e17"}, + {file = "hiredis-2.2.3-pp38-pypy38_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:3fa6811a618653164f918b891a0fa07052bd71a799defa5c44d167cac5557b26"}, + {file = "hiredis-2.2.3-pp38-pypy38_pp73-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:af33f370be90b48bbaf0dab32decbdcc522b1fa95d109020a963282086518a8e"}, + {file = "hiredis-2.2.3-pp38-pypy38_pp73-win_amd64.whl", hash = "sha256:b9953d87418ac228f508d93898ab572775e4d3b0eeb886a1a7734553bcdaf291"}, + {file = "hiredis-2.2.3-pp39-pypy39_pp73-macosx_10_12_x86_64.whl", hash = "sha256:5e7bb4dd524f50b71c20ef5a12bd61da9b463f8894b18a06130942fe31509881"}, + {file = "hiredis-2.2.3-pp39-pypy39_pp73-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:89a258424158eb8b3ed9f65548d68998da334ef155d09488c5637723eb1cd697"}, + {file = "hiredis-2.2.3-pp39-pypy39_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:9f4a65276f6ecdebe75f2a53f578fbc40e8d2860658420d5e0611c56bbf5054c"}, + {file = "hiredis-2.2.3-pp39-pypy39_pp73-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:334f2738700b20faa04a0d813366fb16ed17287430a6b50584161d5ad31ca6d7"}, + {file = "hiredis-2.2.3-pp39-pypy39_pp73-win_amd64.whl", hash = "sha256:d194decd9608f11c777946f596f31d5aacad13972a0a87829ae1e6f2d26c1885"}, + {file = "hiredis-2.2.3.tar.gz", hash = "sha256:e75163773a309e56a9b58165cf5a50e0f84b755f6ff863b2c01a38918fe92daa"}, +] + [[package]] name = "idna" version = "3.3" @@ -1536,6 +1647,26 @@ files = [ [package.extras] full = ["numpy"] +[[package]] +name = "redis" +version = "5.0.1" +description = "Python client for Redis database and key-value store" +category = "main" +optional = false +python-versions = ">=3.7" +files = [ + {file = "redis-5.0.1-py3-none-any.whl", hash = "sha256:ed4802971884ae19d640775ba3b03aa2e7bd5e8fb8dfaed2decce4d0fc48391f"}, + {file = "redis-5.0.1.tar.gz", hash = "sha256:0dab495cd5753069d3bc650a0dde8a8f9edde16fc5691b689a566eda58100d0f"}, +] + +[package.dependencies] +async-timeout = {version = ">=4.0.2", markers = "python_full_version <= \"3.11.2\""} +hiredis = {version = ">=1.0.0", optional = true, markers = "extra == \"hiredis\""} + +[package.extras] +hiredis = ["hiredis (>=1.0.0)"] +ocsp = ["cryptography (>=36.0.1)", "pyopenssl (==20.0.1)", "requests (>=2.26.0)"] + [[package]] name = "requests" version = "2.31.0" @@ -2299,4 +2430,4 @@ indexer = ["ujson", "warcio", "idna", "beautifulsoup4", "lxml", "langdetect", "p [metadata] lock-version = "2.0" python-versions = ">=3.10,<3.11" -content-hash = "0f7cb7135a2dd6ec11e37eb3bbc28071662f80a8ec40398db62c411992bd1b2a" +content-hash = "fe5f238c57ec2d09acb6bdf8f46f33c7bbe499f68a7e34ab7bca1336e0ae881c" diff --git a/pyproject.toml b/pyproject.toml index c52dd5f..4a4a725 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -36,6 +36,7 @@ Levenshtein = {version= "==0.16.0", optional = true} django = "^4.2.4" django-ninja = "^0.22.2" requests-cache = "^1.1.0" +redis = {extras = ["hiredis"], version = "^5.0.1"} [tool.poetry.extras] indexer = [