def convertAllFilesAndAddToDB(self, ALL_INPUT_FILES, inputdir, import_options):
"""
Loads each XML file, saves it as a SciDoc JSON file, adds its metadata to
the database
"""
progress=ProgressIndicator(True, self.num_files_to_process, dot_every_xitems=20)
tasks=[]
for fn in ALL_INPUT_FILES[FILES_TO_PROCESS_FROM:FILES_TO_PROCESS_TO]:
corpus_id=self.generate_corpus_id(fn)
match=cp.Corpus.getMetadataByField("metadata.filename",os.path.basename(fn))
if not match or import_options.get("reload_xml_if_doc_in_collection",False):
if self.use_celery:
match_id=match["guid"] if match else None
tasks.append(importXMLTask.apply_async(
args=[
os.path.join(inputdir,fn),
corpus_id,
self.import_id,
self.collection_id,
import_options,
match_id
],
queue="import_xml"
))
else:
# main loop over all files
filename=cp.Corpus.paths.inputXML+fn
corpus_id=self.generate_corpus_id(fn)
match=cp.Corpus.getMetadataByField("metadata.filename",os.path.basename(fn))
if not match:
try:
doc=convertXMLAndAddToCorpus(
os.path.join(inputdir,fn),
corpus_id,
self.import_id,
self.collection_id,
import_options
)
except ValueError:
logging.exception("ERROR: Couldn't convert %s" % fn)
continue
progress.showProgressReport("Importing -- latest file %s" % fn)
def reloadSciDocsOnly(self, conditions, inputdir, file_mask):
"""
Iterates through the papers already in the collection given the
condition. Tries to load their scidoc. If KeyError occurs, it loads
the XML again
"""
## filenames=cp.Corpus.SQLQuery("SELECT guid,metadata.filename FROM papers where %s limit 10000" % conditions)
in_collection=[item["_source"] for item in cp.Corpus.unlimitedQuery(
index="papers",
doc_type="paper",
_source=["metadata.corpus_id","metadata.filename","guid"],
q=conditions
)]
print("Fixing broken SciDocs")
print("Listing all loaded papers...")
ALL_INPUT_FILES=self.loadListOrListAllFiles(inputdir,file_mask)
files_to_process=[]
files_hash={}
for input_file in ALL_INPUT_FILES:
corpus_id=self.generate_corpus_id(input_file)
files_hash[corpus_id]=input_file
print("Iterating over all papers trying to load them...")
tasks=[]
import_options={"reload_xml_if_doc_in_collection": True,}
progress=ProgressIndicator(True,len(in_collection))
for item in in_collection:
corpus_id=self.generate_corpus_id(item["metadata"]["filename"])
assert corpus_id==item["metadata"]["corpus_id"]
try:
doc=cp.Corpus.loadSciDoc(item["guid"])
except KeyError:
print("File %s is broken" % item["guid"])
if self.use_celery:
tasks.append(importXMLTask.apply_async(args=[
os.path.join(cp.Corpus.paths.inputXML,files_hash[corpus_id]),
corpus_id,
self.import_id,
self.collection_id,
import_options
],
kwargs={"existing_guid":item["guid"]},
queue="import_xml"
))
else:
files_to_process.append([files_hash[corpus_id],item["guid"]])
progress.showProgressReport("Checking papers")
if self.use_celery:
return
print("Processing all %s broken files..." % len(files_to_process))
progress=ProgressIndicator(True,len(files_to_process))
for fn in files_to_process:
corpus_id=self.generate_corpus_id(fn[0])
try:
doc=convertXMLAndAddToCorpus(
os.path.join(cp.Corpus.paths.inputXML,fn[0]),
corpus_id,
self.import_id,
self.collection_id,
import_options,
existing_guid=fn[1],
)
except ValueError:
logging.exception("ERROR: Couldn't convert %s" % fn)
continue
progress.showProgressReport("Importing -- latest file %s" % fn)
