summaryrefslogtreecommitdiffstats
path: root/python/fatcat_tools
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@robocracy.org>2018-11-19 23:04:18 -0800
committerBryan Newbold <bnewbold@robocracy.org>2018-11-19 23:04:18 -0800
commite590eec544ab6f2e54e8770f01e64eef3158fdaa (patch)
tree5f1fe36a489e7e42642d96a3a719dcbd74d60901 /python/fatcat_tools
parent65bdebea35f2ab3c9c8b0f8a8b0a9a577a36bee2 (diff)
downloadfatcat-e590eec544ab6f2e54e8770f01e64eef3158fdaa.tar.gz
fatcat-e590eec544ab6f2e54e8770f01e64eef3158fdaa.zip
initial OAI-PMH harvesters
Diffstat (limited to 'python/fatcat_tools')
-rw-r--r--python/fatcat_tools/harvest/__init__.py2
-rw-r--r--python/fatcat_tools/harvest/doi_registrars.py13
-rw-r--r--python/fatcat_tools/harvest/oaipmh.py157
3 files changed, 167 insertions, 5 deletions
diff --git a/python/fatcat_tools/harvest/__init__.py b/python/fatcat_tools/harvest/__init__.py
index 4de2cbde..7d814696 100644
--- a/python/fatcat_tools/harvest/__init__.py
+++ b/python/fatcat_tools/harvest/__init__.py
@@ -1,3 +1,5 @@
from .harvest_common import HarvestState
from .doi_registrars import HarvestCrossrefWorker, HarvestDataciteWorker
+from .oaipmh import HarvestArxivWorker, HarvestPubmedWorker,\
+ HarvestDoajArticleWorker, HarvestDoajJournalWorker
diff --git a/python/fatcat_tools/harvest/doi_registrars.py b/python/fatcat_tools/harvest/doi_registrars.py
index d5e4b7ec..10492c17 100644
--- a/python/fatcat_tools/harvest/doi_registrars.py
+++ b/python/fatcat_tools/harvest/doi_registrars.py
@@ -10,15 +10,13 @@ import datetime
from pykafka import KafkaClient
from fatcat_tools.workers import most_recent_message
-from .harvest_common import HarvestState
+from .harvest_common import HarvestState, DATE_FMT
# Skip pylint due to:
# AttributeError: 'NoneType' object has no attribute 'scope'
# in 'astroid/node_classes.py'
# pylint: skip-file
-DATE_FMT = "%Y-%m-%d"
-
class HarvestCrossrefWorker:
"""
@@ -68,7 +66,6 @@ class HarvestCrossrefWorker:
self.loop_sleep = 60*60 # how long to wait, in seconds, between date checks
self.api_batch_size = 50
- # for crossref, it's "from-index-date"
self.name = "Crossref"
def params(self, date_str):
@@ -86,6 +83,9 @@ class HarvestCrossrefWorker:
params['cursor'] = resp['message']['next-cursor']
return params
+ def extract_key(self, obj):
+ return obj['DOI'].encode('utf-8')
+
def fetch_date(self, date):
produce_topic = self.kafka.topics[self.produce_topic]
@@ -112,7 +112,7 @@ class HarvestCrossrefWorker:
self.extract_total(resp), http_resp.elapsed))
#print(json.dumps(resp))
for work in items:
- producer.produce(json.dumps(work).encode('utf-8'))
+ producer.produce(json.dumps(work).encode('utf-8'), partition_key=self.extract_key(work))
if len(items) < self.api_batch_size:
break
params = self.update_params(params, resp)
@@ -181,6 +181,9 @@ class HarvestDataciteWorker(HarvestCrossrefWorker):
def extract_total(self, resp):
return resp['meta']['total']
+ def extract_key(self, obj):
+ return obj['doi'].encode('utf-8')
+
def update_params(self, params, resp):
params['page[number]'] = resp['meta']['page'] + 1
return params
diff --git a/python/fatcat_tools/harvest/oaipmh.py b/python/fatcat_tools/harvest/oaipmh.py
new file mode 100644
index 00000000..c3cb90db
--- /dev/null
+++ b/python/fatcat_tools/harvest/oaipmh.py
@@ -0,0 +1,157 @@
+
+"""
+OAI-PMH protocol:
+ https://sickle.readthedocs.io/en/latest/
+
+Pubmed
+ https://www.ncbi.nlm.nih.gov/pmc/tools/oai/
+ https://www.ncbi.nlm.nih.gov/pmc/oai/oai.cgi?verb=GetRecord&identifier=oai:pubmedcentral.nih.gov:152494&metadataPrefix=pmc_fm
+ https://github.com/titipata/pubmed_parser
+
+arxiv
+ some APIs work on a per-version basis, others do not
+
+ http://export.arxiv.org/oai2?verb=GetRecord&identifier=oai:arXiv.org:0804.2273&metadataPrefix=arXiv
+ http://export.arxiv.org/oai2?verb=GetRecord&identifier=oai:arXiv.org:0804.2273&metadataPrefix=arXivRaw
+
+doaj
+ https://github.com/miku/doajfetch
+
+-----
+
+actually, just going to re-use https://github.com/miku/metha for OAI-PMH stuff
+ => shell script from cronjob
+ => call metha-sync daily
+ => metha-cat -since <whenever> | kafkacat output
+ => echo "date" | kafkat state
+ => some shell trick (comm?) to find missing dates; for each, do metha-cat into kafka
+
+or, just skip kafka for this stuff for now? hrm.
+
+crossref-like stuff is far enough along to keep
+
+## More Miku Magic!
+
+wowa, JSTOR KBART files!
+ http://www.jstor.org/kbart/collections/all-archive-titles
+
+https://github.com/miku/ldjtab: faster than jq for just grabbing
+
+sort can be told how much memory to use; eg: `sort -S50%`, and threads to use
+
+"""
+
+import re
+import sys
+import csv
+import json
+import time
+import requests
+import itertools
+import datetime
+from pykafka import KafkaClient
+import sickle
+
+from fatcat_tools.workers import most_recent_message
+from .harvest_common import HarvestState, DATE_FMT
+
+
+class HarvestOaiPmhWorker:
+ """
+ Base class for OAI-PMH harvesters.
+
+ Based on Crossref importer
+ """
+
+
+ def __init__(self, kafka_hosts, produce_topic, state_topic,
+ start_date=None, end_date=None):
+
+ self.produce_topic = produce_topic
+ self.state_topic = state_topic
+ self.kafka = KafkaClient(hosts=kafka_hosts, broker_version="1.0.0")
+
+ self.loop_sleep = 60*60 # how long to wait, in seconds, between date checks
+
+ self.endpoint_url = None # needs override
+ self.metadata_prefix = None # needs override
+ self.state = HarvestState(start_date, end_date)
+ self.state.initialize_from_kafka(self.kafka.topics[self.state_topic])
+
+
+ def fetch_date(self, date):
+
+ api = sickle.Sickle(self.endpoint_url)
+ date_str = date.strftime(DATE_FMT)
+ produce_topic = self.kafka.topics[self.produce_topic]
+ # this dict kwargs hack is to work around 'from' as a reserved python keyword
+ # recommended by sickle docs
+ records = api.ListRecords(**{
+ 'metadataPrefix': self.metadata_prefix,
+ 'from': date_str,
+ 'until': date_str,
+ })
+
+ count = 0
+ with produce_topic.get_producer() as producer:
+ for item in records:
+ count += 1
+ if count % 50 == 0:
+ print("... up to {}".format(count))
+ producer.produce(item.raw.encode('utf-8'), partition_key=item.header.identifier.encode('utf-8'))
+
+ def run(self, continuous=False):
+
+ while True:
+ current = self.state.next(continuous)
+ if current:
+ print("Fetching DOIs updated on {} (UTC)".format(current))
+ self.fetch_date(current)
+ self.state.complete(current, kafka_topic=self.kafka.topics[self.state_topic])
+ continue
+
+ if continuous:
+ print("Sleeping {} seconds...".format(self.loop_sleep))
+ time.sleep(self.loop_sleep())
+ else:
+ break
+ print("{} DOI ingest caught up".format(self.name))
+
+
+class HarvestArxivWorker(HarvestOaiPmhWorker):
+
+ def __init__(self, **kwargs):
+ super().__init__(**kwargs)
+ self.endpoint_url = "https://export.arxiv.org/oai2"
+ self.metadata_prefix = "arXiv"
+
+
+class HarvestPubmedWorker(HarvestOaiPmhWorker):
+
+ def __init__(self, **kwargs):
+ super().__init__(**kwargs)
+ self.endpoint_url = "https://www.ncbi.nlm.nih.gov/pmc/oai/oai.cgi"
+ self.metadata_prefix = "pmc_fm"
+
+
+class HarvestDoajJournalWorker(HarvestOaiPmhWorker):
+ """
+ WARNING: DOAJ OAI-PMH doesn't seem to respect 'from' and 'until' params
+ """
+
+ def __init__(self, **kwargs):
+ super().__init__(**kwargs)
+ self.endpoint_url = "https://www.doaj.org/oai"
+ self.metadata_prefix = "oai_dc"
+
+
+class HarvestDoajArticleWorker(HarvestOaiPmhWorker):
+ """
+ WARNING: DOAJ OAI-PMH doesn't seem to respect 'from' and 'until' params
+ """
+
+ def __init__(self, **kwargs):
+ super().__init__(**kwargs)
+ self.endpoint_url = "https://www.doaj.org/oai.article"
+ self.metadata_prefix = "oai_doaj"
+