def run(self):
_, temp = tempfile.mkstemp(prefix='byoi-')
with self.input().open() as handle:
for path in map(str.strip, handle):
print('processing: %s' % path)
shellout("jq -r -c '.message.items[]' <(unpigz -c {input}) | pigz -c >> {output}", input=path, output=temp)
luigi.File(temp).move(self.output().path)
def run(self):
_, combined = tempfile.mkstemp(prefix='tasktree-')
for target in self.input():
shellout("cat {input} >> {output}", input=target.path,
output=combined)
output = shellout("LANG=C sort -k1,1 -k3,3 {input} > {output}", input=combined)
luigi.File(output).move(self.output().fn)
def run(self):
_, combined = tempfile.mkstemp(prefix='tasktree-')
with self.input().open() as handle:
for row in handle.iter_tsv(cols=('date', 'path')):
shellout("unzip -p {input} >> {output}", input=row.path,
output=combined)
luigi.File(combined).move(self.output().path)
def run(self):
diskdir = self.config.get('ceeol', 'disk-dir')
_, stopover = tempfile.mkstemp(prefix='siskin-')
for path in glob.glob(os.path.join(diskdir, 'articles', 'articles_*xml')):
shellout("span-import -i ceeol {input} | pigz -c >> {output}", input=path, output=stopover)
shellout("cat {update} | pigz -c >> {output}", update=self.input().path, output=stopover)
luigi.LocalTarget(stopover).move(self.output().path)
def run(self):
_, stopover = tempfile.mkstemp(prefix='siskin-')
with self.input().open() as handle:
groups = itertools.groupby(
handle.iter_tsv(cols=('archive', 'member')),
lambda row: row.archive)
for archive, items in groups:
# Write members to extract to temporary file.
_, memberfile = tempfile.mkstemp(prefix='siskin-')
with open(memberfile, 'w') as output:
for item in items:
output.write("%s\n" % item.member)
self.logger.debug("for archive %s extract via: %s", archive,
memberfile)
if not isinstance(archive, six.string_types):
archive = archive.decode(encoding='utf-8')
# The unzippa will not exhaust ARG_MAX.
shellout("""unzippa -v -m {memberfile} {archive} |
sed -e 's@<?xml version="1.0" encoding="UTF-8"?>@@g' | pigz -c >> {output}""",
archive=archive,
memberfile=memberfile,
output=stopover)
try:
os.remove(output.name)
except OSError as err:
self.logger.warn(err)
luigi.LocalTarget(stopover).move(self.output().path)
def complete(self):
path = "{date}-kxp".format(date=self.yesterday.strftime("%y%m%d"))
ids = set()
if not os.path.exists(path):
return False
for index in os.listdir(path):
for f in os.listdir(path + "/" + index):
with gzip.open("{fd}".format(fd=path + "/" + index + "/" + f),
"rt") as inp:
for line in inp:
ids.add(json.loads(line).get("identifier"))
cmd = "zcat {fd} | jq -rc .identifier >> schemaorg-ids-{date}.txt".format(
fd=path + "/" + index + "/" + f,
date=self.yesterday.strftime("%y%m%d"))
shellout(cmd)
es_ids = set()
for record in esidfilegenerator(
host="{host}".format(
**self.config).rsplit("/")[-1].rsplit(":")[0],
port="{host}".format(
**self.config).rsplit("/")[-1].rsplit(":")[1],
index="slub-resources",
type="schemaorg",
idfile="schemaorg-ids-{date}.txt".format(
date=self.yesterday.strftime("%y%m%d")),
source=False):
es_ids.add(record.get("_id"))
if len(es_ids) == len(ids) and len(es_ids) > 0:
return True
return False
def run(self):
_, stopover = tempfile.mkstemp(prefix='siskin-')
temp = shellout("unpigz -c {input} > {output}", input=self.input().get('input').path)
output = shellout("""jq -r '[.doi?, .["rft.issn"][]?, .["rft.eissn"][]?] | @csv' {input} | LC_ALL=C sort -S50% > {output} """,
input=temp, output=stopover)
os.remove(temp)
luigi.File(output).move(self.output().path)
def run(self):
_, combined = tempfile.mkstemp(prefix='siskin-')
for target in self.input():
shellout("cat {input} >> {output}", input=target.path,
output=combined)
output = shellout("marcuniq {input} > {output}", input=combined)
luigi.File(output).move(self.output().path)
def run(self):
_, stopover = tempfile.mkstemp(prefix='siskin-')
shellout("""jq -r '.doi' <(unpigz -c {input}) | grep -v "null" | grep -o "10.*" 2> /dev/null > {output} """,
input=self.input().get('input').path,
output=stopover)
output = shellout("""sort -u {input} > {output} """, input=stopover)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
output = shellout("xsltproc {stylesheet} {input} > {output}",
input=self.input().path,
stylesheet=self.assets('OAIDCtoMARCXML.xsl'))
output = shellout("yaz-marcdump -i marcxml -o marc {input} > {output}",
input=output)
luigi.File(output).move(self.output().path)
def run(self):
_, stopover = tempfile.mkstemp(prefix='siskin-')
with self.input().open() as handle:
for row in handle.iter_tsv(cols=('path',)):
shellout(""" unzip -l {input} | grep "xml$" | awk '{{print "{input}\t"$4}}' >> {output} """,
preserve_whitespace=True, input=row.path, output=stopover)
luigi.File(stopover).move(self.output().path)
def run(self):
"""
Download, maybe unzip, combine with Gold List.
"""
key = "http://amsl.technology/discovery/metadata-usage/Dokument/KBART_FREEJOURNALS"
link = "%s%s" % (self.config.get('amsl', 'uri-download-prefix'), key)
downloaded = shellout("curl --fail {link} > {output} ", link=link)
_, stopover = tempfile.mkstemp(prefix='siskin-')
try:
_ = zipfile.ZipFile(downloaded)
output = shellout("unzip -p {input} >> {output}", input=downloaded)
except zipfile.BadZipfile:
# At least the file is not a zip.
output = shellout("cat {input} >> {output}", input=downloaded)
# Include OA list, refs #11579.
shellout("""curl -s https://pub.uni-bielefeld.de/download/2913654/2913655 | cut -d, -f1,2 | tr -d '"' |
grep -E '[0-9]{{4,4}}-[0-9]{{3,3}}[0-9xX]' | tr ',' '\n' |
awk '{{ print "\t\t"$0 }}' >> {output}""",
output=output,
preserve_whitespace=True)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
archive = None
with self.input().get('dblist-fulltext').open() as handle:
for row in handle.iter_tsv(cols=('kind', 'path', 'db')):
if row.db == self.db:
archive = row.path
with self.input().get('dblist-references').open() as handle:
for row in handle.iter_tsv(cols=('kind', 'path', 'db')):
if row.db == self.db:
archive = row.path
if archive is None:
# an non-existent database name amount to an empty dump file
self.logger.debug('no such db: %s' % self.db)
with self.output().open('w') as output:
pass
return
dbzip = shellout("7z x -so {archive} {db}.zip 2> /dev/null > {output}", archive=archive, db=self.db)
output = shellout("""unzip -p {dbzip} \*.xml 2> /dev/null |
iconv -f iso-8859-1 -t utf-8 |
LC_ALL=C grep -v "^<\!DOCTYPE GENIOS PUBLIC" |
LC_ALL=C sed -e 's@<?xml version="1.0" encoding="ISO-8859-1" ?>@@g' |
LC_ALL=C sed -e 's@</Document>@<x-origin>{origin}</x-origin><x-issue>{issue}</x-issue></Document>@' |
pigz -c >> {output} """, dbzip=dbzip, origin=archive, issue=self.issue)
luigi.File(output).move(self.output().path)
def complete(self):
if not self.output().exists():
return False
with chdir(str(self.output().path)):
output = shellout("""git fetch origin {branch} > {output} 2>&1""",
branch=self.branch)
result = True
with open(output, 'rb') as fh:
content = fh.readlines()
if len(content) >= 3:
result = False
revparseoutput = shellout(
"""git rev-parse {branch} > {output} 2>&1""",
branch=self.branch)
originrevparseoutput = shellout(
"""git rev-parse origin/{branch} > {output} 2>&1""",
branch=self.branch)
revparse = getfirstline(revparseoutput, "0")
originrevparse = getfirstline(originrevparseoutput, "1")
if revparse != originrevparse:
result = False
return result
def run(self):
"""
@IHBhY2thZ2UgbWFpbgoK@IGltcG9ydCAoCiAg@@ImVuY29kaW5nL2pzb24iCiAg@@ImVuY29kaW5nL3htbCIK@@IC
AibG9nIgog@@ICJvcyIKCiAg@@InN0cmluZ3MiCgog@@ICJnaXRodWIuY29tL21pa3UveG1sc3RyZWFtIgog@KQoK@
IHR5cGUgRGVzY3JpcHRpb24gc3RydWN0IHsK@@ICBYTUxOYW1l@IHhtbC5OYW1lIGB4bWw6IkRlc2NyaXB0aW9uImA
K@@ICBJc3Nu@@IFtdc3RyaW5nIGB4bWw6Imlzc24iYAog@@IFNob3J0VGl0bGUgW11zdHJpbmcgYHhtbDoic2hvcnR
UaXRsZSJgCiAg@@VGl0bGUg@ICBbXXN0cmluZyBgeG1sOiJ0aXRsZSJgCiAgICB9Cgog@ZnVuYyBtYWluKCkgewog@
@IHNtIDo9IG1ha2UobWFwW3N0cmluZ11zdHJpbmcpCiAg@@c2Nhbm5lciA6PSB4bWxzdHJlYW0uTmV3U2Nhbm5lcih
vcy5TdGRpbiwgbmV3KERlc2NyaXB0aW9uKSkK@@ICBmb3Igc2Nhbm5lci5TY2FuKCkgewog@@@ICB0YWcgOj0gc2Nh
bm5lci5FbGVtZW50KCkK@@@@aWYgdiwgb2sgOj0gdGFnLigqRGVzY3JpcHRpb24pOyBvayB7CiAg@@@@ICBpZiBsZW
4odi5TaG9ydFRpdGxlKSA9PSAwIHsK@@@@@@ICBjb250aW51ZQog@@@@@fQog@@@@@Zm9yIF8sIHMgOj0gcmFuZ2Ug
di5TaG9ydFRpdGxlIHsK@@@@@@ICBmb3IgXywgdCA6PSByYW5nZSB2LlRpdGxlIHsK@@@@@@@@c21bc10gPSB0CiAg
@@@@@@@IHNtW3N0cmluZ3MuVG9Mb3dlcihzKV0gPSB0CiAg@@@@@@fQog@@@@@fQog@@@ICB9CiAg@@fQog@@IGlmI
GVyciA6PSBzY2FubmVyLkVycigpOyBlcnIgIT0gbmlsIHsK@@@@bG9nLkZhdGFsKGVycikK@@ICB9CiAg@@aWYgZXJ
yIDo9IGpzb24uTmV3RW5jb2Rlcihvcy5TdGRvdXQpLkVuY29kZShzbSk7IGVyciAhPSBuaWwgewog@@@ICBsb2cuRm
F0YWwoZXJyKQog@@IH0K@IH0K
"""
source = self.run.__doc__.replace("\n",
"").replace(" ",
"").replace("@", "ICAg")
tempcode = shellout("""echo '{code}' | base64 -d > {output}.go """,
code=source,
preserve_whitespace=True)
output = shellout(
""" unpigz -c {input} | go run {code}.go > {output} """,
code=tempcode,
input=self.input().path)
os.remove(tempcode)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
_, tmpfile = tempfile.mkstemp(prefix='byoi-')
for target in self.input():
shellout("cat {input} >> {output}",
input=target.path,
output=tmpfile)
luigi.File(tmpfile).move(self.output().path)
def run(self):
_, output = tempfile.mkstemp(prefix='siskin-')
for target in self.input():
shellout("cat {input} >> {output}",
input=target.path,
output=output)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
target = os.path.dirname(self.output().path)
pattern = os.path.join(config.get('lfer', 'glob'))
shellout("rsync -avz {src} {target}", src=pattern, target=target)
with self.output().open('w') as output:
for path in sorted(iterfiles(target)):
output.write_tsv(path)
def run(self):
"""
Download, maybe unzip, combine with Gold List.
"""
key = "http://amsl.technology/discovery/metadata-usage/Dokument/KBART_FREEJOURNALS"
link = "%s%s" % (self.config.get('amsl', 'uri-download-prefix'), key)
downloaded = shellout("curl --fail {link} > {output} ", link=link)
_, stopover = tempfile.mkstemp(prefix='siskin-')
try:
_ = zipfile.ZipFile(downloaded)
output = shellout("unzip -p {input} >> {output}", input=downloaded)
except zipfile.BadZipfile:
# At least the file is not a zip.
output = shellout("cat {input} >> {output}", input=downloaded)
# Include OA list, refs #11579.
shellout(
"""curl -s https://pub.uni-bielefeld.de/download/2913654/2913655 | cut -d, -f1,2 | tr -d '"' |
grep -E '[0-9]{{4,4}}-[0-9]{{3,3}}[0-9xX]' | tr ',' '\n' |
awk '{{ print "\t\t"$0 }}' >> {output}""",
output=output,
preserve_whitespace=True)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
""" The indicator is always recreated, while the subdir
for a given (host, username, base, pattern) is just synced. """
base = os.path.dirname(self.output().path)
subdir = hashlib.sha1('{host}:{username}:{base}:{pattern}'.format(
host=self.host,
username=self.username,
base=self.base,
pattern=self.pattern)).hexdigest()
# target is the root of the mirror
target = os.path.join(base, subdir)
if not os.path.exists(target):
os.makedirs(target)
command = """lftp -u {username},{password}
-e "set net:max-retries 5; set net:timeout 10; mirror --verbose=0
--only-newer -I {pattern} {base} {target}; exit" {host}"""
shellout(command,
host=self.host,
username=pipes.quote(self.username),
password=pipes.quote(self.password),
pattern=pipes.quote(self.pattern),
target=pipes.quote(target),
base=pipes.quote(self.base))
with self.output().open('w') as output:
for path in iterfiles(target):
logger.debug("Mirrored: %s" % path)
output.write_tsv(path)
def run(self):
source = os.path.join(config.get('core', 'swb-mirror'), 'nationallizenzen')
target = os.path.dirname(self.output().path)
shellout("rsync -avz {source} {target}", source=source, target=target)
with self.output().open('w') as output:
for path in iterfiles(target):
output.write_tsv(path)
def run(self):
paths = [p.strip() for p in self.config.get("ceeol", "updates").split(",") if p.strip()]
_, stopover = tempfile.mkstemp(prefix="siskin-")
self.logger.debug("found %d updates", len(paths))
for p in paths:
shellout("span-import -i ceeol-marcxml {input} >> {output}", input=p, output=stopover)
luigi.LocalTarget(stopover).move(self.output().path)
def run(self):
with self.input().open() as handle:
holdings = json.load(handle)
_, stopover = tempfile.mkstemp(prefix='siskin-')
# The property which contains the URI of the holding file. Might change.
urikey = 'DokumentURI'
for holding in holdings:
if holding["ISIL"] == self.isil:
if urikey not in holding:
raise RuntimeError('possible AMSL API change, expected: %s, available keys: %s' % (urikey, holding.keys()))
# refs. #7142
if 'kbart' not in holding[urikey].lower():
self.logger.debug("skipping non-KBART holding URI: %s" % holding[urikey])
continue
link = "%s%s" % (self.config.get('amsl', 'uri-download-prefix'), holding[urikey])
downloaded = shellout("curl --fail {link} > {output} ", link=link)
try:
_ = zipfile.ZipFile(downloaded)
output = shellout("unzip -p {input} >> {output}", input=downloaded, output=stopover)
except zipfile.BadZipfile:
# at least the file is not a zip.
output = shellout("cat {input} >> {output}", input=downloaded, output=stopover)
luigi.File(stopover).move(self.output().path)
def run(self):
paths = [
p.strip() for p in self.config.get("ceeol", "updates").split(",")
if p.strip()
]
_, stopover = tempfile.mkstemp(prefix="siskin-")
self.logger.debug("found %d updates", len(paths))
for p in paths:
shellout("span-import -i ceeol-marcxml {input} >> {output}",
input=p,
output=stopover)
# Append MARC updates.
paths = [
p.strip()
for p in self.config.get("ceeol", "updates-marc").split(",")
if p.strip()
]
_, stopover = tempfile.mkstemp(prefix="siskin-")
self.logger.debug("found %d updates (MARC)", len(paths))
with open(stopover, 'a') as output:
for p in paths:
self.logger.debug("converting: %s", p)
for doc in convert_ceeol_to_intermediate_schema(p):
output.write(json.dumps(doc) + "\n")
luigi.LocalTarget(stopover).move(self.output().path)
def run(self):
output = shellout("""python {script} {input} {output}""",
script=self.assets('30/30_marcbinary.py'),
input=self.input().path)
output = shellout("""yaz-marcdump -o marcxml {input} > {output}""",
input=output)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
"""
itereates over the in LODProcessFromRdi generated JSON-Linked-Data
enriches them with identifier from entityfacts
enriches them with subjects from the GND
enriches them with identifier from wikidata
enriches them with identifier from geonames, if its a geographic Place
ingests them into a elasticsearch node
"""
path = "{date}-aut-data".format(date=self.yesterday.strftime("%y%m%d"))
for index in os.listdir(path):
# doing several enrichment things before indexing the data
for f in os.listdir(path + "/" + index):
cmd = ". ~/git/efre-lod-elasticsearch-tools/init_environment.sh && zcat {fd} | ".format(
fd=path + "/" + index + "/" + f
) # with -pipeline, all the data get's thru, not only enriched docs
# cmd+="~/git/efre-lod-elasticsearch-tools/enrichment/sameAs2id.py -pipeline -stdin -searchserver {host} | ".format(**self.config)
cmd += "~/git/efre-lod-elasticsearch-tools/enrichment/entityfacts-bot.py -pipeline -stdin -searchserver {host} | ".format(
**self.config)
cmd += "~/git/efre-lod-elasticsearch-tools/enrichment/gnd-sachgruppen.py -pipeline -stdin -searchserver {host} | ".format(
**self.config)
cmd += "~/git/efre-lod-elasticsearch-tools/enrichment/wikidata.py -pipeline -stdin | "
if index == "geo":
cmd += "~/git/efre-lod-elasticsearch-tools/enrichment/geonames.py -pipeline -stdin -searchserver {geonames_host} | ".format(
**self.config)
cmd += "esbulk -verbose -server {host} -w 1 -size 20 -index {index} -type schemaorg -id identifier".format(
**self.config, index=index)
shellout(cmd)
put_dict("{host}/date/actual/1".format(**self.config),
{"date": str(self.yesterday.strftime("%Y-%m-%d"))})
def run(self):
"""
TODO(miku): This contains things, that would better be factored out in separate tasks.
"""
titles = {}
output = shellout("span-gh-dump {hfile} > {output}", hfile=self.hfile)
with luigi.File(output, format=TSV).open() as handle:
for row in handle.iter_tsv(cols=('issn', 'title')):
titles[row.issn] = row.title
issns_held = set()
output = shellout("xmlstarlet sel -t -v '//issn' {hfile} | sort | uniq > {output}", hfile=self.hfile)
with luigi.File(output, format=TSV).open() as handle:
for row in handle.iter_tsv(cols=('issn',)):
issns_held.add(row.issn)
issns_crossref = set()
with self.input().open() as handle:
for row in handle.iter_tsv(cols=('issn',)):
issns_crossref.add(row.issn)
covered = issns_held.intersection(issns_crossref)
with self.output().open('w') as output:
for issn in issns_held:
if issn in covered:
output.write_tsv("COVERED", issn, titles[issn])
else:
output.write_tsv("NOT_COVERED", issn, titles[issn])
def run(self):
""" The indicator is always recreated, while the subdir
for a given (host, username, base, pattern) is just synced. """
base = os.path.dirname(self.output().path)
subdir = hashlib.sha1('{host}:{username}:{base}:{pattern}'.format(
host=self.host, username=self.username, base=self.base,
pattern=self.pattern)).hexdigest()
# target is the root of the mirror
target = os.path.join(base, subdir)
if not os.path.exists(target):
os.makedirs(target)
command = """lftp -u {username},{password}
-e "set net:max-retries 5; set net:timeout 10; mirror --verbose=0
--only-newer -I {pattern} {base} {target}; exit" {host}"""
shellout(command, host=self.host, username=pipes.quote(self.username),
password=pipes.quote(self.password),
pattern=pipes.quote(self.pattern),
target=pipes.quote(target),
base=pipes.quote(self.base))
with self.output().open('w') as output:
for path in iterfiles(target):
logger.debug("Mirrored: %s" % path)
output.write_tsv(path)
def run(self):
_, stopover = tempfile.mkstemp(prefix='siskin-')
shellout(r"""cat {input} | awk '{{ print $0"\treferences"}}' >> {output}""",
input=self.input().get('references').path, output=stopover)
shellout(r"""cat {input} | awk '{{ print $0"\tfulltext"}}' >> {output}""",
input=self.input().get('fulltext').path, output=stopover)
luigi.File(stopover).move(self.output().path)
def pypi_build(name, target='deb'):
""" Take a package name and return the filename of the target. """
cache_key = hashlib.sha1('%s:%s' % (name, target)).hexdigest()
cache = shelve.open(CACHE)
if not cache_key in cache:
logger.debug('Building %s for %s...' % (target, name))
stopover = tempfile.mkdtemp(prefix='pkpy-')
try:
shellout("""cd {stopover} &&
fpm --verbose -s python -t {target} {name}""",
stopover=stopover, name=name, target=target)
src = iterfiles(stopover).next()
basename = os.path.basename(src)
dst = os.path.join(PACKAGE_CACHE, basename)
shutil.copyfile(src, dst)
shutil.rmtree(stopover)
cache[cache_key] = basename
except RuntimeError as err:
logger.error(err)
return abort(404)
else:
logger.debug('Cache hit...')
filename = cache[cache_key]
cache.close()
return filename
def github_clone_and_build(username, repo, target='deb'):
""" Clone a repo (username, repo) and build the `target` package.
Returns the filename, that is placed directly under `static`. """
repo_url = '[email protected]:%s/%s.git' % (username, repo)
cache_key = hashlib.sha1('%s:%s' % (repo_url, target)).hexdigest()
cache = shelve.open(CACHE)
if not cache_key in cache:
logger.debug('Building (%s, %s) ...' % (repo_url, target))
stopover = tempfile.mkdtemp(prefix='pkpy-')
shellout("""
cd {stopover} && git clone {repo_url} &&
cd {repo} && fpm --verbose -s python -t {target} .""",
stopover=stopover, repo_url=repo_url, repo=repo, target=target)
src = iterfiles(stopover, fun=lambda fn: fn.endswith(target)).next()
basename = os.path.basename(src)
dst = os.path.join(PACKAGE_CACHE, basename)
shutil.copyfile(src, dst)
shutil.rmtree(stopover)
cache[cache_key] = basename
else:
logger.debug('Cache hit...')
filename = cache[cache_key]
cache.close()
return filename
def run(self):
_, dbpath = tempfile.mkstemp(prefix='siskin-')
shellout("cayley init -alsologtostderr -config {config} -dbpath={dbpath}",
config=self.assets('cayley.bolt.conf'), dbpath=dbpath)
shellout("GOMAXPROCS={gomaxprocs} cayley load -config {config} -alsologtostderr -dbpath={dbpath} --triples {input}",
gomaxprocs=self.gomaxprocs, config=self.assets('cayley.bolt.conf'), dbpath=dbpath, input=self.input().get('ntriples').path)
shutil.move(dbpath, self.output().path)
def run(self):
grobid_dir = "work/{}/{}/grobid_tei".format(self.crawl, self.item)
json_dir = "work/{}/{}/grobid_json".format(self.crawl, self.item)
shellout("mkdir -p {json_dir}", json_dir=json_dir)
# Generate fulltext json (one file per PDF)
for tei_path in glob.glob(grobid_dir + "/*.tei.xml"):
json_path = tei_path.replace('/grobid_tei/', '/grobid_json/')\
.replace('.tei.xml', '.json')
# jq in the pipeline validates JSON
shellout("""
bin/grobid2json.py {tei_path}
| jq -c .
> {json_path}""",
tei_path=tei_path,
json_path=json_path)
# Just the header info (one json file for the whole item)
# jq in the pipeline validates JSON
output = shellout("""
cd {grobid_dir}
&& find . -name "*.tei.xml"
| parallel -j 4 ../../../../bin/grobid2json.py --no-encumbered {{}}
| jq -c .
> {output}""",
grobid_dir=grobid_dir)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
"""
transforms the geonames TSV Dump to line-delimited JSON
"""
cmd = ". ~/git/efre-lod-elasticsearch-tools/init_environment.sh && ~/git/efre-lod-elasticsearch-tools/helperscripts/tsv2json.py {file}.txt | gzip > {file}.ldj.gz".format(
**self.config)
shellout(cmd)
def run(self):
url = self.config.get('thieme', 'oai')
shellout("METHA_DIR={dir} metha-sync -set {set} -format {prefix} {url}",
set=self.set, prefix=self.prefix, url=url, dir=self.config.get('core', 'metha-dir'))
output = shellout("METHA_DIR={dir} metha-cat -format {prefix} {url} | pigz -c > {output}",
prefix=self.prefix, url=url, dir=self.config.get('core', 'metha-dir'))
luigi.File(output).move(self.output().path)
def run(self):
pdf_dir = "work/{}/{}/pdfs".format(self.crawl, self.item)
grobid_dir = "work/{}/{}/grobid_tei".format(self.crawl, self.item)
shellout("mkdir -p {grobid_dir}", grobid_dir=grobid_dir)
output = shellout("""
/usr/bin/time -v -o {output}
java
-Xmx6G
-jar {GROBID_JAR}
-gH {GROBID_HOME}
-dIn {pdf_dir}
-r
-dOut {grobid_dir}
-exe processFullText""",
pdf_dir=pdf_dir,
grobid_dir=grobid_dir,
GROBID_JAR=GROBID_JAR,
GROBID_HOME=GROBID_HOME)
# java/GROBID doesn't error when it runs. One way to check that it ran
# is that there should be a TEI file for every file in PDF dir.
#assert(glob_count(pdf_dir + "/*.pdf") == glob_count(grobid_dir + "/*.tei.xml"))
luigi.LocalTarget(output).move(self.output().path)
def run(self):
_, stopover = tempfile.mkstemp(prefix='siskin-')
for i in range(27):
url = self.template.format(part=i)
shellout("""curl --retry 1 --compress "{url}" >> {output}""",
url=url, output=stopover)
luigi.File(stopover).move(self.output().path)
def run(self):
_, stopover = tempfile.mkstemp(prefix='siskin-')
with self.input().open() as handle:
groups = itertools.groupby(handle.iter_tsv(cols=('archive', 'member')), lambda row: row.archive)
for archive, items in groups:
# Write members to extract to temporary file.
_, memberfile = tempfile.mkstemp(prefix='siskin-')
with open(memberfile, 'w') as output:
for item in items:
output.write("%s\n" % item.member)
self.logger.debug("for archive %s extract via: %s", archive, memberfile)
# The unzippa will not exhaust ARG_MAX.
shellout("""unzippa -v -m {memberfile} {archive} |
sed -e 's@<?xml version="1.0" encoding="UTF-8"?>@@g' | pigz -c >> {output}""",
archive=archive.decode(encoding='utf-8'),
memberfile=memberfile,
output=stopover)
try:
os.remove(output.name)
except OSError as err:
self.logger.warn(err)
luigi.LocalTarget(stopover).move(self.output().path)
def run(self):
shellout(
"""metha-sync -rm -format cmdi -set dta https://clarin.bbaw.de/oai-dta/"""
)
output = shellout(
"""metha-cat -format cmdi -set dta https://clarin.bbaw.de/oai-dta/ > {output}"""
)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
_, combined = tempfile.mkstemp(prefix='siskin-')
for target in self.input():
tmp = shellout("""yaz-marcdump -f utf-8 -t utf-8 -i
marcxml -o marc {input} > {output}""",
input=target.fn, ignoremap={5: 'TODO: fix this'})
shellout("cat {input} >> {output}", input=tmp, output=combined)
luigi.File(combined).move(self.output().path)
def run(self):
_, stopover = tempfile.mkstemp(prefix='siskin-')
with self.input().open() as handle:
for row in handle.iter_tsv(cols=('path',)):
shellout(""" unzip -p {input} | grep -o 'DB="[^"]*' | sed -e 's/DB="//g' >> {output} """,
input=row.path, output=stopover)
output = shellout("sort -u {input} > {output} && rm {input}", input=stopover)
luigi.File(output).move(self.output().path)
def merge(targets):
"""
Helper function to concatenate the outputs for a number of targets.
"""
_, tf = tempfile.mkstemp(prefix='lab-')
for target in targets:
shellout("cat {input} >> {output}", input=target.path, output=tf)
return tf
def run(self):
_, output = tempfile.mkstemp(prefix='siskin-')
with self.input().open() as handle:
for row in sorted(handle.iter_tsv(cols=('path', ))):
if not str(row.path).endswith('.tar'):
continue
shellout("span-import -i elsevier-tar {input} | pigz -c >> {output}", input=row.path, output=output)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
_, stopover = tempfile.mkstemp(prefix='siskin-')
shellout("""LC_ALL=C zgrep -E "http(s)?://.*.crossref.org" {input} >> {output}""",
input=self.input().path, output=stopover)
shellout("""LC_ALL=C zgrep -v "^200" {input} >> {output}""",
input=self.input().path, output=stopover)
output = shellout("sort -S50% -u {input} | cut -f4 | sed s@http://doi.org/api/handles/@@g > {output}", input=stopover)
luigi.File(output).move(self.output().path)
def run(self):
# XXX: Does this really work?
output = shellout("""iconv -f utf-8 -t utf-8 -c {input} > {output}""",
input=self.input().path)
output = shellout("""flux.sh {flux} in={input} > {output}""",
flux=self.assets("127/127.flux"),
input=output)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
output = shellout(""" span-tag -c {config} {input} > {output} """,
config=self.input().get('config').path,
input=self.input().get('file').path)
output = shellout(""" span-export -o {format} {input} > {output} """,
input=output,
format=self.format)
luigi.LocalTarget(output).move(self.output().path)
def merge(targets):
"""
Helper function to concatenate the outputs for a number of targets.
"""
_, tf = tempfile.mkstemp(prefix='lab-')
for target in targets:
shellout("cat {input} >> {output}", input=target.path, output=tf)
return tf
def run(self):
output = shellout("span-tag -c {config} <(unpigz -c {input}) | pigz -c > {output}",
config=self.input().get('config').path,
input=self.input().get('file').path)
output = shellout("span-export -o {format} <(unpigz -c {input}) | pigz -c > {output}",
format=self.format,
input=output)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
"""
iterates over the id-file from LODKXPTransform2ldj, searches for the right titledata (de-14) and merges them with the merge_lok_with_tit script. finally, the data gets loaded into the kxp-de14 index
"""
shellout(
""". ~/git/efre-lod-elasticsearch-tools/init_environment.sh && ~/git/efre-lod-elasticsearch-tools/helperscripts/merge_lok_with_tit.py -selectbody \'{{\"query\": {{\"match\": {{\"852.__.a.keyword\": \"DE-14\"}}}}}}\' -title_server {rawdata_host}/kxp-tit/mrc -local_server {rawdata_host}/kxp-lok/mrc -idfile {date}-lok-ppns.txt | tee data.ldj | esbulk -server {rawdata_host} -index kxp-de14 -type mrc -id 001 -w 1 -verbose && jq -rc \'.\"001\"' data.ldj > ids.txt && rm data.ldj""",
rawdata_host=self.config.get("rawdata_host"),
date=self.yesterday.strftime("%y%m%d"))
def run(self):
_, stopover = tempfile.mkstemp(prefix='siskin-')
shellout(
"""jq -r '.doi' <(unpigz -c {input}) | grep -v "null" | grep -o "10.*" 2> /dev/null > {output} """,
input=self.input().get('input').path,
output=stopover)
output = shellout("""sort -u {input} > {output} """, input=stopover)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
output = shellout(
"""sed -e 's/'$(echo "\o001")'/ /g' < {input} > {output}""",
input=self.config.get('khm', 'dump'))
# TODO(miku): maybe check, if cleanup is still required.
output = shellout("python {script} {input} {output}",
script=self.assets("109/109_marcbinary.py"),
input=output)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
"""
calls esmarc with the idfile created in LODTITTransform2ldj
"""
if os.stat("{date}.mrc.bz2".format(date=self.date)).st_size > 0:
cmd = ". ~/git/efre-lod-elasticsearch-tools/init_environment.sh && ~/git/efre-lod-elasticsearch-tools/processing/esmarc.py -z -server {rawdata_host}/finc-main-k10plus/mrc -idfile {date}-ppns.txt -prefix {date}-data".format(
**self.config, date=self.date)
shellout(cmd)
sleep(5)
def run(self):
endpoint, set = "https://monami.hs-mittweida.de/oai", "institutes:medien"
shellout("""metha-sync -set {set} {endpoint}""",
set=set,
endpoint=endpoint)
output = shellout(
"""metha-cat -set "institutes:medien" {endpoint} > {output} """,
endpoint=endpoint)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
shellout("""metha-sync -rm -format {format} {endpoint} """,
endpoint=self.endpoint,
format=self.format)
output = shellout(
"""metha-cat -format {format} {endpoint} > {output}""",
endpoint=self.endpoint,
format=self.format)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
shellout(
"metha-sync -format {format} http://digi.ub.uni-heidelberg.de/cgi-bin/digioai.cgi",
format=self.format)
output = shellout(
"metha-cat -format {format} -root Records {url} > {output}",
format=self.format,
url=self.url)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
"""
takes the in LODTransform2ldj created TXT file and gets those records from the elasticsearch node
transforms them to JSON-Linked Data with esmarc.py, gzips the files
"""
cmd = ". ~/git/efre-lod-elasticsearch-tools/init_environment.sh && ~/git/efre-lod-elasticsearch-tools/processing/esmarc.py -z -server {host}/swb-aut/mrc -idfile {date}-norm-aut-ppns.txt -prefix {date}-aut-data".format(
**self.config, date=self.yesterday.strftime("%y%m%d"))
shellout(cmd)
sleep(5)
def run(self):
shellout("""METHA_DIR={dir} metha-sync "{endpoint}" """,
dir=self.config.get('core', 'metha-dir'),
endpoint=self.endpoint)
output = shellout(
"""METHA_DIR={dir} metha-cat -root Records "{endpoint}" > {output}""",
dir=self.config.get('core', 'metha-dir'),
endpoint=self.endpoint)
luigi.LocalTarget(output).move(self.output().path)
def run(self):
"""
iterates over the id-file from LODKXPTransform2ldj, gets this set of records from the kxp-de14 index, transforms them to JSON-Linked-Data
"""
# delete("{rawdata_host}/kxp-tit-{date}".format(**self.config,date=self.yesterday.strftime("%y%m%d")))
# delete("{rawdata_host}/kxp-lok-{date}".format(**self.config,date=self.yesterday.strftime("%y%m%d")))
cmd = ". ~/git/efre-lod-elasticsearch-tools/init_environment.sh && ~/git/efre-lod-elasticsearch-tools/processing/esmarc.py -z -server {rawdata_host}/kxp-de14/mrc -idfile ids.txt -prefix {date}-kxp".format(
**self.config, date=self.yesterday.strftime("%y%m%d"))
shellout(cmd)
sleep(5)
def run(self):
"""
fills the local data index and the source title data index with the data transformed in LODKXPTransform2ldj
"""
for typ in ["tit", "lok"]:
# put_dict("{rawdata_host}/kxp-{typ}".format(**self.config,typ=typ,date=self.yesterday.strftime("%y%m%d")),{"mappings":{"mrc":{"date_detection":False}}})
# put_dict("{rawdata_host}/kxp-{typ}/_settings".format(**self.config,typ=typ,date=self.yesterday.strftime("%y%m%d")),{"index.mapping.total_fields.limit":5000})
cmd = "esbulk -z -verbose -server {rawdata_host} -w {workers} -index kxp-{typ} -type mrc -id 001 {date}-{typ}.ldj.gz" "".format(
**self.config, typ=typ, date=self.yesterday.strftime("%y%m%d"))
shellout(cmd)
def run(self):
directory = self.config.get('elsevierjournals', 'backlog-dir')
_, output = tempfile.mkstemp(prefix='siskin-')
for path in sorted(
iterfiles(directory, fun=lambda p: p.endswith('.tar'))):
shellout(
"span-import -i elsevier-tar {input} | pigz -c >> {output}",
input=path,
output=output)
luigi.LocalTarget(output).move(self.output().path)
