One source-document instance
prodigal.json is the parsed representation of the unchanged
nmdc-prodigal corpus file.
Its origin is NMDC DataObject nmdc:dobj-11-2jq65c25, licensed CC BY 4.0 by NMDC;
the source URL is recorded in the JSON and in corpus/index.yaml.
The original is 2,033 bytes with MD5 92d78177f3c3d0c410d0b57c94ebf22c;
the JSON also records its SHA-256 digest.
Reproduce it from the repository root into a new scratch path:
just source-parse \
corpus/sources/nmdc/nmdc_wfmgan-11-9ya9xh30.1_prodigal.gff gff3 \
local/source-documents/prodigal-reproduced.json --profile prodigal \
--source-uri 'https://data.microbiomedata.org/data/nmdc:omprc-11-cfaemn69/nmdc:wfmgan-11-9ya9xh30.1/nmdc_wfmgan-11-9ya9xh30.1_prodigal.gff'
cmp model/examples/source-documents/prodigal.json local/source-documents/prodigal-reproduced.json
just validate-source-example
The validation recipe also compares the replayed bytes against the independently
retained corpus source. For a different example, supply both paths with
just validate-source-example your-document.json your-original.gff.
There are five Sequence Data / Model Data blocks around six feature rows. The
sequence lengths are 359, 303, 298, 298, and 295 bases. The first block uses translation
table 4; the others use table 11. These are sequence-scoped settings, not one file-wide
value. seqhdr quotes and ordered/repeated metadata survive; training model names
are not interpreted as sample taxonomy.
This is a source-preservation example, separate from the harmonized Dataset examples. See the document model and parser contract for scope, generic versus producer profiles, exact replay, and validation limits.