Skip to content

Catalog import: 17 SaM listings, SaR 186 and Abl 51 are attached to the wrong objects #668

Closed

Description

@brickbots

The shipped astro_data/pifinder_objects.db (on main and release since #564) attaches 19 catalog listings to the wrong objects. Selecting SaM 1501 (20 Peg) opens M 20, the Trifid. Selecting SaM 463 (61 Cet) opens NGC 5195, and push-to then goes there. The nixos line's file gets all 19 right, and both fixes are already on that line; main doesn't have them.

I found this while building the chart tablet's catalog layer, by comparing the shipped file with the nixos line's file listing by listing, by position.

Cause 1: Herschel double-star aliases match the Herschel 400 (18 listings)

ObjectFinder.get_object_id in catalog_imports/catalog_import_utils.py falls back to normalize(object_name), which lowercases and strips spaces. A double star's alias "H 24" (William Herschel's double-star number) or "h 99" (John Herschel's) becomes h24 or h99. That is the key of Herschel 400 listing H 24 or H 99. find_object_id takes the first hit, so the star's listing joins the H400 galaxy or cluster and takes its position, type and magnitude.

On the nixos line, get_object_id falls back to parse_designation, which accepts only whitelisted prefixes ("herschel" maps to H, a bare "h" doesn't), so these aliases no longer match.

Listing Alias Shipped file: joined to Should be (nixos file)
SaM 137 (107 Aqr) H 24 NGC 5676, galaxy in Boo double star, Aqr 23h46m00s -18°41'
SaM 168 (Nu Aur) H 90 NGC 4245, galaxy in Com double star, Aur 05h51m30s +39°08'
SaM 463 (61 Cet) H 102 NGC 5195, galaxy in CVn double star, Cet 02h03m48s +00°20'
SaM 649 H 38 NGC 637, cluster in Cas double star, CrB 16h23m12s +32°19'
SaM 953 (60 Her) H 133 NGC 5982, galaxy in Dra double star, Her 17h05m18s +12°44'
SaM 999 h 99 NGC 4027, galaxy in Crv double star, Hya 08h37m48s -06°48'
SaM 1042 (17 Crt) H 96 NGC 3962, galaxy in Crt double star, Hya 11h32m18s -29°16'
SaM 1257, SaM 1258 (Nu1 Lyr) H 40 NGC 381, cluster in Cas double star, Lyr 18h49m42s +32°48'
SaM 1501 (20 Peg) h 289 NGC 6514 / M 20 double star, Peg 22h01m06s +13°06'
SaM 1521, SaM 1522 (Xi Peg) h 301 NGC 3938, galaxy in UMa double star, Peg 22h46m42s +12°10'
SaM 1629 H 119 NGC 7086, cluster in Cyg double star, PsA 22h39m42s -28°20'
SaM 1726 h 2 NGC 752 / C 28, cluster in And double star, Pup 08h14m00s -40°21'
SaM 1771 (Nu Sco) H 6 NGC 205 / M 110 double star, Sco 16h12m00s -19°28'
SaM 1887 H 129 NGC 7006 / C 42, globular in Del double star, Sgr 19h04m18s -22°54'
SaM 1940 H 98 NGC 4038 / C 60, the Antennae double star, Tau 04h15m30s +06°11'
SaR 186 (Gamma Lep) H 40 NGC 381, cluster in Cas double star, Lep 05h44m30s -22°27'

The H400 objects also pick up the stars' names, so NGC 381 is called "Nu1 Lyr" as well, and the star's description is lost.

Cause 2: the Abell 50 alias is on row 51 of abell.tsv (1 listing)

On main, astro_data/abell.tsv gives Abell 51 the alias ngc6742, but NGC 6742 is Abell 50, in Draco. The import matched the alias, so Abl 51 joined NGC 6742 at 18h59m +48°27'. Abell 51 is PN G017.6-10.2 in Sagittarius, at the row's own position (285.256, -18.204). Abl 50 became a second object at NGC 6742's position, so that nebula appears twice. The nixos line moved the alias to row 50.

Suggested fix on main

  • Bring over parse_designation and CATALOG_CODE_ALIASES from the nixos line, or at least stop the fallback matching a bare one-letter prefix.
  • Move ngc6742 from row 51 to row 50 of abell.tsv.
  • Rebuild pifinder_objects.db.

Two more differences, less clear-cut

  • Arp 65. The shipped file has it as a galaxy named NGC 0091, at 00h21m51s +22°24'. The nixos file merges it into NGC 91, typed as a star, 2' south.
  • Ta2 52. The shipped file joins it to NGC 2244, the cluster. The nixos file joins it to NGC 2237, the nebula.

Neither is plainly wrong, but someone should look at both.

Why the tablet cares

The chart tablet reads this file and pins one version of it: 8d357eb, the #564 file. Its contract test will show the 19 listings changing when the pin moves to a fixed file.

Activity

  1. added
    bugSomething isn't working
    catalogsIssue with catalog data
    on Oct 12, 2026
  2. brickbots commented on Nov 19, 2026

    @brickbots
    OwnerAuthor

    Fixed by the database rebuild in #697 (4d6e6d6). All 17 SaM listings and SaR 186 are now double stars at their own positions, each with its WDS listing. Abl 51 is its own planetary nebula in Sagittarius, and NGC 381 no longer carries "Nu1 Lyr". The evidence is in #697's "Related issues" section.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't workingcatalogsIssue with catalog data

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions