This site is supported by donations to The OEIS Foundation.

User:Georg Fischer/Broken link maintenance

From OeisWiki
Jump to: navigation, search

2008

  • Semi-automated procedure with Perl scripts and wget

2018 - manual repairs

2018-12 Restart of a general analysis

The scripts for this new project can be cloned from the GitHub repository gfis/OEIS-mat.

The following URLs are assumed to be stable, and were filtered out:

domains dx.doi.org|doi.org|oeis.org|en.wikipedia.org|mathworld.wolfram.com|arXiv.org|web.archive.org|lacim.uqam.ca|emis.de
all local files (links starting with "/")

Statistics

all %H lines     389420 (sent by NJAS at 2018-12-10)
total URLs       392253
filtered URLs     75969
protocols:
  ftp://             73
  http://         67087
  https://         8809
distinct domains   4424
distinct URLs     23300

The broken link problem is permanent, but *not hopeless*. From the 400,000 links in the OEIS I count 21872 unique ones, and of them

Currently 2595 URLs do not pass the accessibility test.

There are only 79 broken links which occur in more than 20 sequences, and which may be candidates for a mass-edit procedure, since they affect 2591 sequences. The ones with occurrences <= 20 affect 4884 sequences.