diff options
| author | Bryan Newbold <bnewbold@robocracy.org> | 2019-06-24 14:46:24 -0700 |
|---|---|---|
| committer | Bryan Newbold <bnewbold@robocracy.org> | 2019-06-24 14:46:24 -0700 |
| commit | 135d262b3e92fc14d4e2572d4ea93f9980357c9b (patch) | |
| tree | ca6366b85075254a4c45db0a5aeac6288df21a82 /python/fatcat_tools/importers | |
| parent | b1e26b8f2c53141d0cfed9199a771ff8f07926fd (diff) | |
| download | fatcat-135d262b3e92fc14d4e2572d4ea93f9980357c9b.tar.gz fatcat-135d262b3e92fc14d4e2572d4ea93f9980357c9b.zip | |
fix existing updater
Diffstat (limited to 'python/fatcat_tools/importers')
| -rw-r--r-- | python/fatcat_tools/importers/matched.py | 5 |
1 files changed, 3 insertions, 2 deletions
diff --git a/python/fatcat_tools/importers/matched.py b/python/fatcat_tools/importers/matched.py index 1e154204..4da1d646 100644 --- a/python/fatcat_tools/importers/matched.py +++ b/python/fatcat_tools/importers/matched.py @@ -159,9 +159,10 @@ class MatchedImporter(EntityImporter): # minimum viable "existing" URL cleanup to fix dupes and broken links: # remove 'None' wayback URLs, and set archive.org rel 'archive' existing.urls = [u for u in existing.urls if not ('://web.archive.org/web/None/' in u.url)] - for u in existing.urls: + for i in len(existing.urls): + u = existing.urls[i] if u.rel == 'repository' and '://archive.org/download/' in u.url: - u.rel == 'archive' + existing.urls[i].rel == 'archive' # merge the existing into this one and update existing.urls = list(set([(u.rel, u.url) for u in fe.urls + existing.urls])) |