Skip to content

3.13.1 — Parse-tree node flags are read by name

Choose a tag to compare

View all tags
@github-actions github-actions released this 10 Sep 04:41
· 50 commits to master since this release
v3.13.1
86dbd1e
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Parse-tree node flags are read by name.

  • "Display Built Only" no longer hides nodes that were built. The engine writes a node's flags only when they are set -- b, un, sem, fired, blt, in that order -- so which field holds blt depends on how many flags precede it. The reader took a fixed field, which is right for a node written node,fired,blt and wrong for one written node,un,fired,blt, where the flag has shifted along by one. Those nodes came back as not built and were dropped from the tree in BUILT mode. In the sample analyzers 63 nodes are written that way, _sentence and _year among them.
  • The same fixed-field reading in the other tree reader marked any unsealed node as fired: node,un has un where the reader expected fired. That field is not consulted by anything today, so nothing displayed wrongly, but it was wrong on 23,621 of the 37,863 node lines in the sample analyzers -- and it is exactly the field a future reader would reach for.
  • Both readers now share one flag parser with the .tree reader added in 3.13.0, so the vocabulary lives in one place and the two cannot drift apart again.
Assets 3
Loading