edward/depicts
1
0
Fork 0
Code Issues 9 Pull requests Packages Projects Releases Wiki Activity

Bug fix saam parser

Browse source
This commit is contained in:
Edward Betts 2020-06-30 09:05:22 +01:00
parent dc9989157c
commit 3191ae51b0
1 changed files with 7 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
depicts/saam.py
View file

@ -18,17 +18,22 @@ def get_html(saam_id):
def parse_html(html): def parse_html(html):
root = lxml.html.fromstring(html) root = lxml.html.fromstring(html)
ld = json.loads(root.findtext('.//script[@type="application/ld+json"]')) ld_json = root.findtext('.//script[@type="application/ld+json"]')
if ld_json is None:
return {'ld': {}, 'keywords': []}
ld = json.loads(ld_json)
ul = root.find('.//ul[@class="ontology-list"]') ul = root.find('.//ul[@class="ontology-list"]')
if ul is None: if ul is None:
return {'ld': ld, 'keywords': []} return
assert ul.tag == 'ul' assert ul.tag == 'ul'
keywords = [li.text for li in ul] keywords = [li.text for li in ul]
return {'ld': ld, 'keywords': keywords} return {'ld': ld, 'keywords': keywords}
def get_catalog(saam_id): def get_catalog(saam_id):
data = parse_html(get_html(saam_id)) data = parse_html(get_html(saam_id))
if not data:
return {}
ret = { ret = {
'institution': 'Smithsonian American Art Museum', 'institution': 'Smithsonian American Art Museum',
} }