def linkreport(config=None, **kwargs):
#Any configuration variables passed in
if config is None: config = {}
try:
#Initialize the output
outstr = ''
while True:
params = yield
model = params['model']
items = {}
#Get the title
#First get the work ID
workid = params['workid']
#simple_lookup() is a little helper for getting a property from a resource
title = simple_lookup(model, workid, TITLE_REL)
#Get the ISBN, just pick the first one
isbn = ''
if params['instanceids']:
inst1 = params['instanceids'][0]
isbn = simple_lookup(model, inst1, ISBN_REL)
envelope = '<div id="{0} isbn="{1}"><title>{2}</title>\n'.format(workid, isbn, title)
#iterate over all the relationship targets to see which is a link
for stmt in model.match():
if iri.matches_uri_syntax(stmt[TARGET]) and iri.split_uri_ref(stmt[TARGET])[1] != BFHOST:
envelope += '<a href="{0}">{0}</a>\n'.format(stmt[TARGET], stmt[TARGET])
envelope += '</div>\n'
outstr += envelope
except GeneratorExit:
#Reached when close() is called on this coroutine
with open(config['output-file'], "w") as outf:
outf.write(outstr)
def handle_record_links(self, loop, model, params):
'''
Task coroutine of the main event loop for MARC conversion, called with
In this case update a report of links encountered in the MARC/XML
model -- raw Versa model with converted resource information from the MARC details from each MARC/XML record processed
params -- parameters passed in from processing:
params['workid']: ID of the work constructed from the MARC record
params['instanceid']: list of IDs of instances constructed from the MARC record
'''
#print ('BF_MARCREC_TASK', linkreport.PLUGIN_ID)
items = {}
#Get the title
#First get the work ID
workid = params['workid']
#simple_lookup() is a little helper for getting a property from a resource
title = simple_lookup(model, workid, TITLE_REL)
#Get the ISBN, just pick the first one
isbn = ''
if params['instanceids']:
inst1 = params['instanceids'][0]
isbn = simple_lookup(model, inst1, ISBN_REL)
envelope = '<div id="{0}" isbn="{1}"><title>{2}</title>\n'.format(workid, isbn, title)
#iterate over all the relationship targets to see which is a link
for stmt in model.match():
if iri.matches_uri_syntax(stmt[TARGET]) and iri.split_uri_ref(stmt[TARGET])[1] != BFHOST:
self._links_found.add(stmt[TARGET])
envelope += '<a href="{0}">{0}</a>\n'.format(stmt[TARGET], stmt[TARGET])
envelope += '</div>\n'
self._outstr += envelope
#print ('DONE BF_MARCREC_TASK', linkreport.PLUGIN_ID)
return
def _abbreviate(self, rid):
'''
Abbreviate a relationship or resource ID target for efficient storage
in the DB. Works only with a prefix/suffix split of hierarchical HTTP-like IRIs,
e.g. 'http://example.org/spam/eggs' becomes something like '{a23}eggs'
and afterward there will be an entry in the prefix map from 'a23' to 'http://example.org/spam/'
The map can then easily be used with str.format
'''
if not isinstance(rid, str) or '/' not in rid or not iri.matches_uri_syntax(rid):
return rid
head, tail = rid.rsplit('/', 1)
head += '/'
pmap = self._db['@_abbreviations']
assert pmap is not None
#FIXME: probably called too often to do this every time
inv_pmap = {v: k for k, v in pmap.items()}
if head in inv_pmap:
prefix = inv_pmap[head]
else:
prefix = f'a{self._abbr_index}'
pmap[prefix] = head
self._abbr_index += 1
self._db['@_abbreviations'] = pmap
post_rid = '{' + prefix + '}' + tail
return post_rid
def _abbreviate(self, rid, txn):
'''
Abbreviate a relationship or resource ID target for efficient storage
in the DB. Works only with a prefix/suffix split of hierarchical HTTP-like IRIs,
e.g. 'http://example.org/spam/eggs' becomes something like '{a23}eggs'
and afterward there will be an entry in the prefix map from 'a23' to 'http://example.org/spam/'
The map can then easily be used with str.format
'''
if not isinstance(
rid, str) or '/' not in rid or not iri.matches_uri_syntax(rid):
return rid
head, tail = rid.rsplit('/', 1)
head += '/'
prefix_map = txn.get(b'@_abbreviations')
assert prefix_map is not None
prefix_map = msgpack.loads(prefix_map, raw=False)
#FIXME: probably called too often to do this every time
inv_prefix_map = {v: k for k, v in prefix_map.items()}
if head in inv_prefix_map:
prefix = inv_prefix_map[head]
else:
prefix = f'a{self._abbr_index}'
prefix_map[prefix] = head
self._abbr_index += 1
txn.put(b'@_abbreviations',
msgpack.dumps(prefix_map, use_bin_type=True))
post_rid = '{' + prefix + '}' + tail.replace('{', '{{').replace(
'}', '}}')
return post_rid
def _abbreviate(self, rid):
'''
Abbreviate a relationship or resource ID target for efficient storage
in the DB. Works only with a prefix/suffix split of hierarchical HTTP-like IRIs,
e.g. 'http://example.org/spam/eggs' becomes something like '{a23}eggs'
and afterward there will be an entry in the prefix map from 'a23' to 'http://example.org/spam/'
The map can then easily be used with str.format
'''
if not isinstance(rid, str) or not iri.matches_uri_syntax(rid):
return rid
head, tail = rid.rsplit('/', 1)
head += '/'
abbrev_obj = self._db_coll.find_one({'origin': '@_abbreviations'})
assert abbrev_obj is not None
pmap = abbrev_obj['map']
#FIXME: probably called too often to do this every time
inv_pmap = {v: k for k, v in pmap.items()}
if head in inv_pmap:
prefix = inv_pmap[head]
else:
prefix = f'a{self._abbr_index}'
pmap[prefix] = head
self._abbr_index += 1
self._db_coll.replace_one({'origin': '@_abbreviations'}, {
'origin': '@_abbreviations',
'map': pmap
})
post_rid = '{' + prefix + '}' + tail
return post_rid
def handle_record_links(self, loop, model, params):
'''
Task coroutine of the main event loop for MARC conversion, called with
In this case update a report of links encountered in the MARC/XML
model -- raw Versa model with converted resource information from the MARC details from each MARC/XML record processed
params -- parameters passed in from processing:
params['workid']: ID of the work constructed from the MARC record
params['instanceid']: list of IDs of instances constructed from the MARC record
'''
#print ('BF_MARCREC_TASK', linkreport.PLUGIN_ID)
items = {}
#Get the title
#First get the work ID
workid = params['workid']
#simple_lookup() is a little helper for getting a property from a resource
title = simple_lookup(model, workid, TITLE_REL)
#Get the ISBN, just pick the first one
isbn = ''
if params['instanceids']:
inst1 = params['instanceids'][0]
isbn = simple_lookup(model, inst1, ISBN_REL)
envelope = '<div id="{0}" isbn="{1}"><title>{2}</title>\n'.format(
workid, isbn, title)
#iterate over all the relationship targets to see which is a link
for stmt in model.match():
if iri.matches_uri_syntax(stmt[TARGET]) and iri.split_uri_ref(
stmt[TARGET])[1] != BFHOST:
self._links_found.add(stmt[TARGET])
envelope += '<a href="{0}">{0}</a>\n'.format(
stmt[TARGET], stmt[TARGET])
envelope += '</div>\n'
self._outstr += envelope
#print ('DONE BF_MARCREC_TASK', linkreport.PLUGIN_ID)
return
def __init__(self,
obj,
siri=None,
encoding=None,
streamopenmode='rb',
sourcetype=inputsourcetype.unknown):
'''
obj - byte string, proper string (only if you really know what you're doing),
file-like object (stream), file path or URI.
uri - optional override URI. Base URI for the input source will be set to
this value
>>> from amara3 import inputsource
>>> inp = inputsource('abc')
>>> inp.stream
<_io.StringIO object at 0x1056fbf78>
>>> inp.iri
>>> print(inp.iri)
None
>>> inp = inputsource(['abc', 'def']) #Now multiple streams in one source
>>> inp.stream
<_io.StringIO object at 0x1011aff78>
>>> print(inp.iri)
None
>>> inp = next(inp)
>>> inp.stream
<_io.StringIO object at 0x1011af5e8>
>>> print(inp.iri)
None
>>>
'''
# from amara3 import inputsource; inp = inputsource('foo.zip')
# from amara3 import inputsource; inp = inputsource('test/resource/std-examples.zip')
# s = inp.stream.read(100)
# s
# b'<?xml version="1.0" encoding="UTF-8"?>\r\n<!-- edited with XML Spy v4.3 U (http://www.xmlspy.com) by M'
# s
# b'<?xml version="1.0" encoding="UTF-8"?>\r\n<collection xmlns="http://www.loc.gov/MARC21/slim">\r\n <reco'
self.stream = None
self.iri = siri
self.sourcetype = sourcetype
if obj in ('', b''):
raise ValueError("Cannot parse an empty string as XML")
if hasattr(obj, 'read'):
#Create dummy Uri to use as base
#uri = uri or uuid4().urn
self.stream = obj
#elif sourcetype == inputsourcetype.xmlstring:
#See this article about XML detection heuristics
#http://www.xml.com/pub/a/2007/02/28/what-does-xml-smell-like.html
#uri = uri or uuid4().urn
elif self.sourcetype == inputsourcetype.iri or (
siri and iri.matches_uri_syntax(obj)):
self.iri = siri or obj
self.stream = urlopen(iri)
elif self.sourcetype == inputsourcetype.filename or (
siri and iri.is_absolute(obj) and not os.path.isfile(obj)):
#FIXME: convert path to URI
self.iri = siri or iri.os_path_to_uri(obj)
self.stream = open(obj, streamopenmode)
elif self.sourcetype == inputsourcetype.string or isinstance(
obj, str) or isinstance(obj, bytes):
self.stream = StringIO(obj)
#If obj is beyond a certain length, don't even try it as a URI
#if len(obj) < MAX_URI_LENGTH_FOR_HEURISTIC:
# self.iri = iri.os_path_to_uri(obj)
# self.stream = urlopen(siri)
else:
raise ValueError("Unable to recognize as an inputsource")
return
def __init__(self, obj, siri=None, encoding=None, streamopenmode='rb',
sourcetype=inputsourcetype.unknown):
'''
obj - byte string, proper string (only if you really know what you're doing),
file-like object (stream), file path or URI.
uri - optional override URI. Base URI for the input source will be set to
this value
>>> from amara3 import inputsource
>>> inp = inputsource('abc')
>>> inp.stream
<_io.StringIO object at 0x1056fbf78>
>>> inp.iri
>>> print(inp.iri)
None
>>> inp = inputsource(['abc', 'def']) #Now multiple streams in one source
>>> inp.stream
<_io.StringIO object at 0x1011aff78>
>>> print(inp.iri)
None
>>> inp = next(inp)
>>> inp.stream
<_io.StringIO object at 0x1011af5e8>
>>> print(inp.iri)
None
>>>
'''
# from amara3 import inputsource; inp = inputsource('foo.zip')
# from amara3 import inputsource; inp = inputsource('test/resource/std-examples.zip')
# s = inp.stream.read(100)
# s
# b'<?xml version="1.0" encoding="UTF-8"?>\r\n<!-- edited with XML Spy v4.3 U (http://www.xmlspy.com) by M'
# s
# b'<?xml version="1.0" encoding="UTF-8"?>\r\n<collection xmlns="http://www.loc.gov/MARC21/slim">\r\n <reco'
self.stream = None
self.iri = siri
self.sourcetype = sourcetype
if obj in ('', b''):
raise ValueError("Cannot parse an empty string as XML")
if hasattr(obj, 'read'):
#Create dummy Uri to use as base
#uri = uri or uuid4().urn
self.stream = obj
#elif sourcetype == inputsourcetype.xmlstring:
#See this article about XML detection heuristics
#http://www.xml.com/pub/a/2007/02/28/what-does-xml-smell-like.html
#uri = uri or uuid4().urn
elif self.sourcetype == inputsourcetype.iri or (siri and iri.matches_uri_syntax(obj)):
self.iri = siri or obj
self.stream = urlopen(iri)
elif self.sourcetype == inputsourcetype.filename or (siri and iri.is_absolute(obj) and not os.path.isfile(obj)):
#FIXME: convert path to URI
self.iri = siri or iri.os_path_to_uri(obj)
self.stream = open(obj, streamopenmode)
elif self.sourcetype == inputsourcetype.string or isinstance(obj, str) or isinstance(obj, bytes):
self.stream = StringIO(obj)
#If obj is beyond a certain length, don't even try it as a URI
#if len(obj) < MAX_URI_LENGTH_FOR_HEURISTIC:
# self.iri = iri.os_path_to_uri(obj)
# self.stream = urlopen(siri)
else:
raise ValueError("Unable to recognize as an inputsource")
return
