def list_records(
client,
request,
query,
identifier,
page_number,
sort_by,
search_params,
list_redirect_target,
reset_url,
uuid,
):
resources = LazyPagedSequence(
lambda page, page_size: client.find_collections(
search_pattern=query,
identifier=identifier,
page=page,
page_size=page_size,
sort_by=sort_by,
),
PAGE_SIZE,
client.count_collections(query, identifier),
)
page = helpers.pager(resources, PAGE_SIZE, page_number)
sort_direction = _determine_reverse_sort_direction(sort_by)
return render(request, list_redirect_target, locals())
def search(request):
# FIXME there has to be a better way of handling checkboxes than parsing
# them by hand here, and displaying 'checked' in
# _archival_storage_search_form.html
# Parse checkbox for file mode
yes_options = ('checked', 'yes', 'true', 'on')
if request.GET.get('filemode', '') in yes_options:
file_mode = True
checked_if_in_file_mode = 'checked'
items_per_page = 20
else: # AIP list
file_mode = False
checked_if_in_file_mode = ''
items_per_page = 10
# Parse checkbox for show AICs
show_aics = ''
if request.GET.get('show_aics', '') in yes_options:
show_aics = 'checked'
# get search parameters from request
queries, ops, fields, types = advanced_search.search_parameter_prep(request)
logger.debug('Queries: %s, Ops: %s, Fields: %s, Types: %s', queries, ops, fields, types)
# redirect if no search params have been set
if 'query' not in request.GET:
return helpers.redirect_with_get_params(
'components.archival_storage.views.search',
query='',
field='',
type=''
)
# get string of URL parameters that should be passed along when paging
search_params = advanced_search.extract_url_search_params_from_request(request)
current_page_number = int(request.GET.get('page', 1))
# perform search
es_client = elasticSearchFunctions.get_client()
results = None
query = advanced_search.assemble_query(es_client, queries, ops, fields, types, search_index='aips', doc_type='aipfile')
try:
# use all results to pull transfer facets if not in file mode
# pulling only one field (we don't need field data as we augment
# the results using separate queries)
if not file_mode:
# Fetch all unique AIP UUIDs in the returned set of files
query['aggs'] = {'aip_uuids': {'terms': {'field': 'AIPUUID', 'size': 0}}}
# Don't return results, just the aggregation
query['size'] = 0
# Searching for AIPs still actually searches type 'aipfile', and
# returns the UUID of the AIP the files are a part of. To search
# for an attribute of an AIP, the aipfile must index that
# information about their AIP in
# elasticSearchFunctions.index_mets_file_metadata
results = es_client.search(
body=query,
index='aips',
doc_type='aipfile',
sort='sipName:desc',
)
# Given these AIP UUIDs, now fetch the actual information we want from aips/aip
buckets = results['aggregations']['aip_uuids']['buckets']
uuids = [bucket['key'] for bucket in buckets]
uuid_file_counts = {bucket['key']: bucket['doc_count'] for bucket in buckets}
query = {
'query': {
'terms': {
'uuid': uuids,
},
},
}
index = 'aips'
doc_type = 'aip'
fields = 'name,uuid,size,created,status,AICID,isPartOf,countAIPsinAIC,encrypted'
sort = 'name:desc'
else:
index = 'aips'
doc_type = 'aipfile'
fields = 'AIPUUID,filePath,FILEUUID,encrypted'
sort = 'sipName:desc'
# To reduce amount of data fetched from ES, use LazyPagedSequence
def es_pager(page, page_size):
"""
Fetch one page of normalized aipfile entries from Elasticsearch.
:param page: 1-indexed page to fetch
:param page_size: Number of entries on a page
:return: List of dicts for each entry with additional information
"""
start = (page - 1) * page_size
results = es_client.search(
body=query,
from_=start,
size=page_size,
index=index,
doc_type=doc_type,
fields=fields,
sort=sort,
)
if file_mode:
return search_augment_file_results(es_client, results)
else:
return search_augment_aip_results(results, uuid_file_counts)
count = es_client.count(index=index, doc_type=doc_type, body={'query': query['query']})['count']
results = LazyPagedSequence(es_pager, items_per_page, count)
except ElasticsearchException:
logger.exception('Error accessing index.')
return HttpResponse('Error accessing index.')
if not file_mode:
aic_creation_form = forms.CreateAICForm(initial={'results': uuids})
else: # if file_mode
aic_creation_form = None
page_data = helpers.pager(results, items_per_page, current_page_number)
return render(request, 'archival_storage/search.html',
{
'file_mode': file_mode,
'show_aics': show_aics,
'checked_if_in_file_mode': checked_if_in_file_mode,
'aic_creation_form': aic_creation_form,
'results': page_data.object_list,
'search_params': search_params,
'page': page_data,
}
)
def list_display(request):
if 'aips' not in settings.SEARCH_ENABLED:
return render(request, 'archival_storage/list.html')
current_page_number = int(request.GET.get('page', 1))
logger.debug('Current page: %s', current_page_number)
# get count of AIP files
es_client = elasticSearchFunctions.get_client()
aip_indexed_file_count = aip_file_count(es_client)
# get AIPs
order_by = request.GET.get('order_by', 'name_unanalyzed')
sort_by = request.GET.get('sort_by', 'up')
if sort_by == 'down':
sort_direction = 'desc'
else:
sort_direction = 'asc'
sort_specification = order_by + ':' + sort_direction
sort_params = 'order_by=' + order_by + '&sort_by=' + sort_by
# get list of UUIDs of AIPs that are deleted or pending deletion
aips_deleted_or_pending_deletion = []
should_haves = [
{'match': {'status': 'DEL_REQ'}},
{'match': {'status': 'DELETED'}},
]
query = {
"query": {
"bool": {
"should": should_haves
}
}
}
deleted_aip_results = es_client.search(
body=query,
index='aips',
doc_type='aip',
fields='uuid,status'
)
for deleted_aip in deleted_aip_results['hits']['hits']:
aips_deleted_or_pending_deletion.append(deleted_aip['fields']['uuid'][0])
# Fetch results and paginate
def es_pager(page, page_size):
"""
Fetch one page of normalized entries from Elasticsearch.
:param page: 1-indexed page to fetch
:param page_size: Number of entries on a page
:return: List of dicts for each entry, where keys and values have been cleaned up
"""
start = (page - 1) * page_size
results = es_client.search(
index='aips',
doc_type='aip',
body=elasticSearchFunctions.MATCH_ALL_QUERY,
fields='origin,uuid,filePath,created,name,size,encrypted',
sort=sort_specification,
size=page_size,
from_=start,
)
# normalize results - each of the fields contains a single value,
# but is returned from the ES API as a single-length array
# e.g. {"fields": {"uuid": ["abcd"], "name": ["aip"] ...}}
return [elasticSearchFunctions.normalize_results_dict(d) for d in results['hits']['hits']]
items_per_page = 10
count = es_client.count(index='aips', doc_type='aip', body=elasticSearchFunctions.MATCH_ALL_QUERY)['count']
results = LazyPagedSequence(es_pager, page_size=items_per_page, length=count)
# Paginate
page = helpers.pager(
results,
items_per_page,
current_page_number
)
# process deletion, etc., and format results
aips = []
for aip in page.object_list:
# If an AIP was deleted or is pending deletion, react if status changed
if aip['uuid'] in aips_deleted_or_pending_deletion:
# check with storage server to see current status
api_results = storage_service.get_file_info(uuid=aip['uuid'])
try:
aip_status = api_results[0]['status']
except IndexError:
# Storage service does not know about this AIP
# TODO what should happen here?
logger.info("AIP not found in storage service: {}".format(aip))
continue
# delete AIP metadata in ElasticSearch if AIP has been deleted from the
# storage server
# TODO: handle this asynchronously
if aip_status == 'DELETED':
elasticSearchFunctions.delete_aip(es_client, aip['uuid'])
elasticSearchFunctions.delete_aip_files(es_client, aip['uuid'])
elif aip_status != 'DEL_REQ':
# update the status in ElasticSearch for this AIP
elasticSearchFunctions.mark_aip_stored(es_client, aip['uuid'])
else:
aip_status = 'UPLOADED'
# Tweak AIP presentation and add to display array
if aip_status != 'DELETED':
aip['status'] = AIP_STATUS_DESCRIPTIONS[aip_status]
try:
size = '{0:.2f} MB'.format(float(aip['size']))
except (TypeError, ValueError):
size = 'Removed'
aip['size'] = size
aip['href'] = aip['filePath'].replace(AIPSTOREPATH + '/', "AIPsStore/")
aip['date'] = aip['created']
aips.append(aip)
total_size = total_size_of_aips(es_client)
# Find out which AIPs are encrypted
return render(request, 'archival_storage/list.html',
{
'total_size': total_size,
'aip_indexed_file_count': aip_indexed_file_count,
'aips': aips,
'page': page,
'search_params': sort_params,
}
)
def search(request):
# FIXME there has to be a better way of handling checkboxes than parsing
# them by hand here, and displaying 'checked' in
# _archival_storage_search_form.html
# Parse checkbox for file mode
yes_options = ("checked", "yes", "true", "on")
if request.GET.get("filemode", "") in yes_options:
file_mode = True
checked_if_in_file_mode = "checked"
items_per_page = 20
else: # AIP list
file_mode = False
checked_if_in_file_mode = ""
items_per_page = 10
# Parse checkbox for show AICs
show_aics = ""
if request.GET.get("show_aics", "") in yes_options:
show_aics = "checked"
# get search parameters from request
queries, ops, fields, types = advanced_search.search_parameter_prep(
request)
logger.debug("Queries: %s, Ops: %s, Fields: %s, Types: %s", queries, ops,
fields, types)
# redirect if no search params have been set
if "query" not in request.GET:
return helpers.redirect_with_get_params(
"components.archival_storage.views.search",
query="",
field="",
type="")
# get string of URL parameters that should be passed along when paging
search_params = advanced_search.extract_url_search_params_from_request(
request)
current_page_number = int(request.GET.get("page", 1))
# perform search
es_client = elasticSearchFunctions.get_client()
results = None
query = advanced_search.assemble_query(queries, ops, fields, types)
try:
# Use all results to pull transfer facets if not in file mode
# pulling only one field (we don't need field data as we augment
# the results using separate queries).
if not file_mode:
# Fetch all unique AIP UUIDs in the returned set of files
# ES query will limit to 10 aggregation results by default,
# add size parameter in terms to override.
# TODO: Use composite aggregation when it gets out of beta.
query["aggs"] = {
"aip_uuids": {
"terms": {
"field": "AIPUUID",
"size": "10000"
}
}
}
# Don't return results, just the aggregation
query["size"] = 0
# Searching for AIPs still actually searches type 'aipfile', and
# returns the UUID of the AIP the files are a part of. To search
# for an attribute of an AIP, the aipfile must index that
# information about their AIP.
results = es_client.search(body=query, index="aipfiles")
# Given these AIP UUIDs, now fetch the actual information we want from aips/aip
buckets = results["aggregations"]["aip_uuids"]["buckets"]
uuids = [bucket["key"] for bucket in buckets]
uuid_file_counts = {
bucket["key"]: bucket["doc_count"]
for bucket in buckets
}
query = {"query": {"terms": {"uuid": uuids}}}
index = "aips"
fields = (
"name,uuid,size,created,status,AICID,isPartOf,countAIPsinAIC,encrypted"
)
sort = "name.raw:desc"
else:
index = "aipfiles"
fields = "AIPUUID,filePath,FILEUUID,encrypted"
sort = "sipName.raw:desc"
# To reduce amount of data fetched from ES, use LazyPagedSequence
def es_pager(page, page_size):
"""
Fetch one page of normalized aipfile entries from Elasticsearch.
:param page: 1-indexed page to fetch
:param page_size: Number of entries on a page
:return: List of dicts for each entry with additional information
"""
start = (page - 1) * page_size
results = es_client.search(
body=query,
from_=start,
size=page_size,
index=index,
_source=fields,
sort=sort,
)
if file_mode:
return search_augment_file_results(es_client, results)
else:
return search_augment_aip_results(results, uuid_file_counts)
count = es_client.count(index=index, body={"query":
query["query"]})["count"]
results = LazyPagedSequence(es_pager, items_per_page, count)
except ElasticsearchException:
logger.exception("Error accessing index.")
return HttpResponse("Error accessing index.")
if not file_mode:
aic_creation_form = forms.CreateAICForm(initial={"results": uuids})
else: # if file_mode
aic_creation_form = None
page_data = helpers.pager(results, items_per_page, current_page_number)
return render(
request,
"archival_storage/search.html",
{
"file_mode": file_mode,
"show_aics": show_aics,
"checked_if_in_file_mode": checked_if_in_file_mode,
"aic_creation_form": aic_creation_form,
"results": page_data.object_list,
"search_params": search_params,
"page": page_data,
},
)
def list_display(request):
if "aips" not in settings.SEARCH_ENABLED:
return render(request, "archival_storage/list.html")
current_page_number = int(request.GET.get("page", 1))
logger.debug("Current page: %s", current_page_number)
# get count of AIP files
es_client = elasticSearchFunctions.get_client()
aip_indexed_file_count = aip_file_count(es_client)
# get AIPs
order_by = request.GET.get("order_by", "name")
sort_by = request.GET.get("sort_by", "up")
sort_params = "order_by=" + order_by + "&sort_by=" + sort_by
# use raw subfield to sort by name
if order_by == "name":
order_by = order_by + ".raw"
# change sort_by param to ES sort directions
if sort_by == "down":
sort_by = "desc"
else:
sort_by = "asc"
sort_specification = order_by + ":" + sort_by
# get list of UUIDs of AIPs that are deleted or pending deletion
aips_deleted_or_pending_deletion = []
should_haves = [{
"match": {
"status": "DEL_REQ"
}
}, {
"match": {
"status": "DELETED"
}
}]
query = {"query": {"bool": {"should": should_haves}}}
deleted_aip_results = es_client.search(body=query,
index="aips",
_source="uuid,status")
for deleted_aip in deleted_aip_results["hits"]["hits"]:
aips_deleted_or_pending_deletion.append(deleted_aip["_source"]["uuid"])
# Fetch results and paginate
def es_pager(page, page_size):
"""
Fetch one page of normalized entries from Elasticsearch.
:param page: 1-indexed page to fetch
:param page_size: Number of entries on a page
:return: List of dicts for each entry, where keys and values have been cleaned up
"""
start = (page - 1) * page_size
results = es_client.search(
index="aips",
body={"query": {
"match_all": {}
}},
_source="origin,uuid,filePath,created,name,size,encrypted",
sort=sort_specification,
size=page_size,
from_=start,
)
return [d["_source"] for d in results["hits"]["hits"]]
items_per_page = 10
count = es_client.count(index="aips", body={"query": {
"match_all": {}
}})["count"]
results = LazyPagedSequence(es_pager,
page_size=items_per_page,
length=count)
# Paginate
page = helpers.pager(results, items_per_page, current_page_number)
# process deletion, etc., and format results
aips = []
for aip in page.object_list:
# If an AIP was deleted or is pending deletion, react if status changed
if aip["uuid"] in aips_deleted_or_pending_deletion:
# check with storage server to see current status
api_results = storage_service.get_file_info(uuid=aip["uuid"])
try:
aip_status = api_results[0]["status"]
except IndexError:
# Storage service does not know about this AIP
# TODO what should happen here?
logger.info("AIP not found in storage service: {}".format(aip))
continue
# delete AIP metadata in ElasticSearch if AIP has been deleted from the
# storage server
# TODO: handle this asynchronously
if aip_status == "DELETED":
elasticSearchFunctions.delete_aip(es_client, aip["uuid"])
elasticSearchFunctions.delete_aip_files(es_client, aip["uuid"])
elif aip_status != "DEL_REQ":
# update the status in ElasticSearch for this AIP
elasticSearchFunctions.mark_aip_stored(es_client, aip["uuid"])
else:
aip_status = "UPLOADED"
# Tweak AIP presentation and add to display array
if aip_status != "DELETED":
aip["status"] = AIP_STATUS_DESCRIPTIONS[aip_status]
try:
size = "{0:.2f} MB".format(float(aip["size"]))
except (TypeError, ValueError):
size = "Removed"
aip["size"] = size
aip["href"] = aip["filePath"].replace(AIPSTOREPATH + "/",
"AIPsStore/")
aip["date"] = aip["created"]
aips.append(aip)
total_size = total_size_of_aips(es_client)
# Find out which AIPs are encrypted
return render(
request,
"archival_storage/list.html",
{
"total_size": total_size,
"aip_indexed_file_count": aip_indexed_file_count,
"aips": aips,
"page": page,
"search_params": sort_params,
},
)
