summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorMartin Czygan <martin.czygan@gmail.com>2020-08-10 19:55:14 +0200
committerMartin Czygan <martin.czygan@gmail.com>2020-08-10 19:58:12 +0200
commite18d48642cecb55d9f2270f9048953a7b543472e (patch)
treefa38a037143ae65d86bc91ffc3aba8d4ff97272e
parentde0fb59f0e36d8079649feefb7592189d8f7c6ed (diff)
downloadfatcat-e18d48642cecb55d9f2270f9048953a7b543472e.tar.gz
fatcat-e18d48642cecb55d9f2270f9048953a7b543472e.zip
harvest: datacite API yields HTTP 200 with broken JSON
As a first step: log response body for debugging.
-rw-r--r--python/fatcat_tools/harvest/doi_registrars.py9
1 files changed, 8 insertions, 1 deletions
diff --git a/python/fatcat_tools/harvest/doi_registrars.py b/python/fatcat_tools/harvest/doi_registrars.py
index 2554fe96..03773c15 100644
--- a/python/fatcat_tools/harvest/doi_registrars.py
+++ b/python/fatcat_tools/harvest/doi_registrars.py
@@ -136,7 +136,14 @@ class HarvestCrossrefWorker:
time.sleep(30.0)
continue
http_resp.raise_for_status()
- resp = http_resp.json()
+ try:
+ resp_body = http_resp.text
+ resp = json.loads(resp_body)
+ except json.JSONDecodeError as exc:
+ # Datacite API returned HTTP 200, but JSON seemed unparseable.
+ # It might be a glitch, so we retry.
+ print("failed to decode body from {}: {}".format(http_resp.url, resp_body), file=sys.stderr)
+ raise
items = self.extract_items(resp)
count += len(items)
print("... got {} ({} of {}), HTTP fetch took {}".format(len(items), count,