portal.just.ro ⟶ sqlite

Descarcă arhiva via portalquery.just.ro API către un db sqlite

fetchData(dateRange, instanta='all')

instanta = 'all' | [slug_instanta1, slug_instanta2, ...]

if instanta='all' or None, loop all instanțe then fetch dosare & ședințe filtered by dateRange

Commands

• createDb.py <pato/to/db> : creates sqlite db from config, if no path given
• fetchAPI.py <date> <days=1> <direction=back> : fetches date + days in direction [back|fwd] – fetches 24h after date
• updateDb.py : looks into xml folder, writes to db, moves to /parsed folder; also checks for dupes? – if dupe create log

SQL db schema + Prompts

Given the following SQLite Database containing data about lawsuits archive with the following tables schema:

• "trials" (id: TEXT, data: DATE, court: TEXT, category: TEXT, status: TEXT)
• "parties" (trial_id: TEXT, name: TEXT, type: TEXT)
• "appeals" (trial_id: TEXT, date: DATE, appealing_party: TEXT, type: TEXT)
• "courts" (name: TEXT, id: TEXT, county: TEXT, type: TEXT, parent: TEXT)

Where we have the following relationships:

• trial_id in parties and appeals tabbles is linked to the trials.id
• courts.parent is the parent of the court, which is a value from courts.id
• trials.court is one of the values from courts.id

---

Roadmap

Stats / SQL

• SQL stats - via GPT
  • describe db schema

NLP-ish

categorisation, clustering, entity extraction

• n-grams
  • cli script interactive (optional) parameters
• k-means
• manual definition, recipes
• entity detection, semantics

---

UI

Datasette

• Datasette Template
• Dashboards
• Search

Fetching

v0.1

• fetch api ⟶ local xml
• local xml ⟶ sqlite prototype

v0.2

• prototype fetch ⟶ db sequence for 24h (1day)
• add relationships (Dosar)
  • DosarParte
  • DosarSedinta
  • DosarCaleAtac

v0.25

• sedinte
  • Sedinta
  • SedintaDosar

v0.3

• commands w arguments
• logging
• fetch api ⟶ update sqlite

v0.4

• check for fetch errors
• check dupes?
• cron

v0.45

• Datasette template + dashboards

v0.5

• nosql db

---

NLP-ish