| name | archive-nlmnlp |
| description | Use when maintaining a local NLM NLP concept archive over PubMed for offline chemical, disease, gene, or GeneRIF lookups. |
| disable-model-invocation | true |
| user-invocable | true |
archive-nlmnlp
Quick Start
- Command:
archive-nlmnlp [daily|-index]
- Local executable:
/home/vimalinx/miniforge3/envs/bio/bin/archive-nlmnlp
- Full reference: See references/help.md for complete documentation
When To Use This Tool
- Build a local PubMed concept archive backed by PubTator Central, GeneRIF, and gene metadata files.
- Support offline lookup of chemical, disease, gene, synonym, and GeneRIF-derived postings in EDirect workflows.
- Refresh the archive after upstream concept files change, then rebuild postings in the usual EDirect layout.
- Use this instead of stitching together the PubTator and GeneRIF files manually.
Common Patterns
export EDIRECT_LOCAL_ARCHIVE=/data/edirect
archive-nlmnlp -https
archive-nlmnlp daily
archive-nlmnlp -index
Recommended Workflow
- Set
EDIRECT_LOCAL_ARCHIVE and confirm the archive root exists and is writable.
- Make sure
pm-setup, pm-prepare, and the local Go toolchain are available.
- Run
archive-nlmnlp or archive-nlmnlp -https to stage the latest PubTator / GeneRIF support files.
- Follow with
daily or -index as a separate command when you need refreshed postings.
- Check the
DWN, IDX, INV, MRG, and PST timing blocks to confirm each stage actually ran.
Guardrails
--help and --version are not safe inspection switches; they still trigger archive setup and prerequisite checks.- The script depends on
EDIRECT_LOCAL_ARCHIVE, pm-setup, pm-prepare, and a local go compiler.
- Default transport may try the FTP / Aspera path; use
-https or -ftp explicitly if you need predictable transfer behavior.
- Cleaning and indexing must be run separately.
- The cleanup aliases remove scratch data, so do not use them casually on a populated archive.
