mirrors/bookwyrm
1
0
Fork 1
mirror of https://github.com/bookwyrm-social/bookwyrm.git synced 2024-12-16 13:16:33 +00:00
Code Issues Projects Releases Wiki Activity

hide isni authors if isni in local db

Browse source 
Check the authors suggested from the local DB for a matching ISNI when pulling authors from ISNI.
i.e. we do not want to suggest the same author twice when we *know* it is the same author.
This commit is contained in:
Hugh Rundle 2021-10-31 20:48:47 +11:00
parent bce5f3f3b5
commit 3cfd31f1fe
2 changed files with 43 additions and 29 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

37
bookwyrm/utils/isni.py
View file

@ -1,29 +1,30 @@
"""ISNI author checking utilities"""
import xml.etree.ElementTree as ET
import requests
from bookwyrm.settings import BASE_DIR
def url_stringify(string):
"""replace spaces for url encoding"""
# TODO: this is very lazy and incomplete
return string.replace(" ", "+")
def request_isni_data(search_index, search_term, maxRecords=10):
def request_isni_data(search_index, search_term, maxRecords=5):
"""Request data from the ISNI API"""
search_string = url_stringify(search_term)
query_parts = [
"http://isni.oclc.org/sru/?query=",
"http://isni.oclc.org/sru/?query=",
search_index,
"+%3D+%22",
search_string,
"%22&version=1.1&operation=searchRetrieve&recordSchema=isni-b",
"&maximumRecords=",
str(maxRecords),
"&startRecord=1&recordPacking=xml&sortKeys=RLV%2Cpica%2C0%2C%2C"
]
"&startRecord=1&recordPacking=xml&sortKeys=RLV%2Cpica%2C0%2C%2C",
]
query_url = "".join(query_parts)
result = requests.get(query_url)
# the OCLC ISNI server asserts the payload is encoded
@ -31,29 +32,32 @@ def request_isni_data(search_index, search_term, maxRecords=10):
result.encoding = "utf-8"
return result.text
def make_name_string(element):
"""create a string of form 'personal_name surname'"""
# NOTE: this will often be incorrect, many naming systems
# NOTE: this will often be incorrect, many naming systems
# list "surname" before personal name
forename = element.find(".//forename")
surname = element.find(".//surname")
if forename is not None:
return "".join([forename.text," ",surname.text])
return "".join([forename.text, " ", surname.text])
return surname.text
def get_other_identifier(element, code):
"""Get other identifiers associated with an author from their ISNI record"""
identifiers = element.findall(".//otherIdentifierOfIdentity")
for section_head in identifiers:
if (
section_head.find(".//type") is not None
section_head.find(".//type") is not None
and section_head.find(".//type").text == code
and section_head.find(".//identifier") is not None
and section_head.find(".//identifier") is not None
):
return section_head.find(".//identifier").text
return
return
def get_external_information_uri(element, match_string):
"""Get URLs associated with an author from their ISNI record"""
@ -61,13 +65,11 @@ def get_external_information_uri(element, match_string):
sources = element.findall(".//externalInformation")
for source in sources:
uri = source.find(".//URI")
if (
uri is not None
and uri.text.find(match_string) is not None
):
return uri.text
if uri is not None and uri.text.find(match_string) is not None:
return uri.text
return
def find_authors_by_name(name_string):
"""Query the ISNI database for possible author matches by name"""
@ -94,6 +96,7 @@ def find_authors_by_name(name_string):
return possible_authors
def get_author_isni_data(isni):
payload = request_isni_data("pica.isn", isni)
@ -110,9 +113,9 @@ def get_author_isni_data(isni):
author["viaf_id"] = get_other_identifier(element, "viaf")
author["wikipedia_link"] = get_external_information_uri(element, "Wikipedia")
author["bio"] = bio.text if bio is not None else ""
author["aliases"] = [] # CHECK can we send a list for this?
author["aliases"] = []
aliases = element.findall(".//personalNameVariant")
for entry in aliases:
author["aliases"].append( make_name_string(entry) )
author["aliases"].append(make_name_string(entry))
return author

35
bookwyrm/views/books/edit_book.py
View file

@ -57,22 +57,32 @@ class EditBook(View):
"aliases", weight="B"
)
author_matches = (
models.Author.objects.annotate(search=vector)
.annotate(rank=SearchRank(vector, author))
.filter(rank__gt=0.4)
.order_by("-rank")[:5]
)
isni_authors = find_authors_by_name(
author
) # find matches from ISNI API
# do not show isni results for authors we already have in the DB
exists = [
i
for i in isni_authors
for a in author_matches
if i["isni"] == a.isni
]
isni_matches = list(filter(lambda x: x not in exists, isni_authors))
data["author_matches"].append(
{
"name": author.strip(),
"matches": (
models.Author.objects.annotate(search=vector)
.annotate(rank=SearchRank(vector, author))
.filter(rank__gt=0.4)
.order_by("-rank")[:5]
),
"isni_matches": find_authors_by_name(
author
), # find matches from ISNI API
"matches": author_matches,
"isni_matches": isni_matches,
}
)
# TODO: check if an isni record matches an existing record
# to bring these two records together
# we're creating a new book
if not book:
@ -157,7 +167,8 @@ class ConfirmEditBook(View):
# otherwise it's a name with or without isni id
isni = request.POST.get(f"isni_match-{i}")
author_data = (
get_author_isni_data(isni) if isni is not None
get_author_isni_data(isni)
if isni is not None
else {"name": match}
)
author = models.Author.objects.create(**author_data)