diff options
| author | Bryan Newbold <bnewbold@archive.org> | 2020-01-15 22:49:23 -0800 |
|---|---|---|
| committer | Bryan Newbold <bnewbold@archive.org> | 2020-01-15 22:49:34 -0800 |
| commit | a8ada6d50e6522d1d8213b5aeb5662db9c96ed79 (patch) | |
| tree | a09b5b69f35498e74deb2d244e0b0a09708cdf16 | |
| parent | 9acda8f48c0a2bb117600635df3b61c7f6272698 (diff) | |
| download | sandcrawler-a8ada6d50e6522d1d8213b5aeb5662db9c96ed79.tar.gz sandcrawler-a8ada6d50e6522d1d8213b5aeb5662db9c96ed79.zip | |
make failed replay fetch an error, not assert error
| -rw-r--r-- | python/sandcrawler/ia.py | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/python/sandcrawler/ia.py b/python/sandcrawler/ia.py index 058f4ca..8895500 100644 --- a/python/sandcrawler/ia.py +++ b/python/sandcrawler/ia.py @@ -406,7 +406,8 @@ class WaybackClient: #print(resp.url, file=sys.stderr) # defensively check that this is actually correct replay based on headers - assert "X-Archive-Src" in resp.headers + if not "X-Archive-Src" in resp.headers: + raise WaybackError("replay fetch didn't return X-Archive-Src in headers") if not datetime in resp.url: raise WaybackError("didn't get exact reply (redirect?) datetime:{} got:{}".format(datetime, resp.url)) |