def get_url_csl_item_zotero(url):
"""
Use Zotero's translation-server to generate a CSL Item for the specified URL.
"""
from manubot.cite.zotero import export_as_csl, web_query
zotero_data = web_query(url)
csl_data = export_as_csl(zotero_data)
(csl_item,) = csl_data
return csl_item
def get_url_csl_item_zotero(url: str) -> CSLItem:
"""
Use Zotero's translation-server to generate a CSL Item for the specified URL.
"""
from manubot.cite.zotero import export_as_csl, web_query
zotero_data = web_query(url)
csl_data = export_as_csl(zotero_data)
(csl_item, ) = csl_data
if not csl_item.get("URL"):
# some Zotero translators don't set URL. https://github.com/manubot/manubot/issues/244
csl_item["URL"] = url
return csl_item
def test_web_query_returns_single_result_pubmed_url():
"""
See test_web_query_returns_single_result_legacy_manubot_url docstring.
```
curl \
--header "Content-Type: text/plain" \
--data 'https://www.ncbi.nlm.nih.gov/pubmed/?term=sci-hub%5Btitle%5D' \
'https://translate.manubot.org/web?single=1'
```
"""
url = "https://www.ncbi.nlm.nih.gov/pubmed/?term=sci-hub%5Btitle%5D"
zotero_metadata = web_query(url)
assert isinstance(zotero_metadata, list)
assert len(zotero_metadata) == 1
(zotero_metadata,) = zotero_metadata
assert zotero_metadata["title"].startswith("sci-hub[title]")
def test_web_query():
"""
The translation-server web endpoint can be tested via curl:
```
curl \
--header "Content-Type: text/plain" \
--data 'https://bigthink.com/neurobonkers/a-pirate-bay-for-science' \
'https://translate.manubot.org/web'
```
An outdated installation of translation-server caused the web query for this
URL to be extraordinarily slow but has now been fixed. See
https://github.com/zotero/translation-server/issues/63
"""
url = "https://bigthink.com/neurobonkers/a-pirate-bay-for-science"
zotero_data = web_query(url)
assert isinstance(zotero_data, list)
assert len(zotero_data) == 1
assert zotero_data[0]["title"].startswith("Meet the Robin Hood of Science")
def test_web_query_returns_single_result_legacy_manubot_url():
"""
Check that single=1 is specified for web queries. Without this, Zotero
can prefer translators that return multiple choices. This occurs with legacy
Manubot mansucripts, which get assigned the DOI translator as top priority.
https://github.com/zotero/translation-server/issues/65
```
curl \
--header "Content-Type: text/plain" \
--data 'https://greenelab.github.io/scihub-manuscript/v/cfe599e25405d38092bf972b6ea1c9e0dcf3deb9/' \
'https://translate.manubot.org/web?single=1'
```
"""
url = 'https://greenelab.github.io/scihub-manuscript/v/cfe599e25405d38092bf972b6ea1c9e0dcf3deb9/'
zotero_metadata = web_query(url)
assert isinstance(zotero_metadata, list)
assert len(zotero_metadata) == 1
zotero_metadata, = zotero_metadata
assert zotero_metadata[
'title'] == 'Sci-Hub provides access to nearly all scholarly literature'
