Skip to content

Improve designation/exact ref parsing

Lars Willighagen requested to merge phoenix/feature/exact-ref-parsing into phoenix/develop

Algorithm:

  • Segment all designations on RegExp "word" boundaries (boundaries between a character in the set A-Za-z0-9_ and a character outside that set).
  • Compute the longest common subsequence (LCS) of those designations.
  • For each publication, split its designation-exact ref string on the LCS and take the last element as its exact reference.

Merge request reports