Skip to content

Processing revisions and errata

Matt Post edited this page Jul 8, 2020 · 1 revision

This page describes how to process revisions and errata to papers int he Anthology.

  1. Per the instructions, there should be a Github issue with (a) the Anthology ID and (b) the revised file attached, and (c) a description of the changes.

  2. Look at the original file, the changed file, and the submitter. Make at least a cursory verification that the changes are well-described, and that the Github submitter is one of the authors of the paper.

  3. Ensure that the comment describing the changes is written from a neutral, third-person, scientific perspective. The corrections page listed above contains a few good examples.

  4. Copy the path to the file (either remote or local works) and run the add_revision.py script:

    ./bin/add_revision.py {anth_id} {path} "Revision comment"
    

    This will put the file in place locally under ~/anthology-files/pdf/{local_path}, where {local_path} varies according to whether the paper has a new- or old-style ID.

    For errata, you need to add the -e flag:

    ./bin/add_revision.py -e {anth_id} {path} "Errata comment"
    
  5. Verify the XML change with git diff. Commit the change. I typically use a git comment of "Revision to {anth_id} (closes #{github_id})", where {github_id} is the issue ID on Github.

  6. Push the change up and create a PR for verification. You will probably be processing corrections in batches, so you can wait to create this PR until you've processed all the corrections in your batch.

  7. Sync the files to the server. If you have an ssh alias aclweb for the ACL server, you can type:

    rsync -ave ssh ~/anthology-files/pdf aclweb:anthology-files/pdf