Diario Oficial Chile Scraper & Monitor — Apify Store Overview

Actor: stefano_seggio/diario-oficial-cl-monitor (ID: qfBeEKuLfYUw9UOuW) Store URL: https://apify.com/stefano_seggio/diario-oficial-cl-monitor Version: 2.0

---

Executive Summary & Business Use Case

The Diario Oficial Chile Scraper & Monitor extracts today's edition of Chile's Diario Oficial — the country's official government gazette, published at diariooficial.interior.gob.cl — pulling every law, decree and resolution listed that day with its full institutional hierarchy (branch of government, ministry, sub-agency) and a direct link to the source PDF. The actor was built after a technical audit (2026-09-04) confirmed no existing Apify actor covered this source: Mexico's DOF and the Colombia/Paraguay procurement systems already had coverage from other developers, but Chile's gazette did not. It ships with an optional delta mode that returns only publications that are genuinely new or corrected since an earlier run the same day, so it can be scheduled intraday without re-delivering the same entries.

Because the source is a legal-publication registry rather than a commercial listing or tender board, the realistic use cases sit squarely in regulatory monitoring rather than pricing or inventory analysis — there is no monetary-value field anywhere in this dataset, so use cases are framed around what changed and who published it, not what it's worth. Three concrete examples: a legal or compliance team at a company operating in Chile watches ministerio and organismo for its sector (e.g. fisheries, mining, telecom) to catch new resolutions affecting its operations and route them to the right internal owner using event_type=NEW_LISTING; a law firm or regulatory consultancy tracks specific publications over time and uses event_type=UPDATED to know the moment a fe de erratas (correction) is re-published under the same CVE, so it can re-check the amended text before advising a client; and a market-research or policy-analytics team aggregates rama, ministerio and seccion across runs to build trend views of regulatory activity by government branch and agency.

What the actor deliberately does not claim is also part of its value proposition: a published legal notice in this domain is a permanent public record — it is never "adjudicated," "closed," or withdrawn the way a tender or licence might be — so the dataset has no status-change or closure event, and the documentation says so explicitly rather than forcing a generic lifecycle model onto a domain that doesn't have one.

Technical Features & V2 Architecture Highlights

This actor ships with named key-value-store cross-run delta persistence: per its own input_schema.json, the onlyNew flag returns "only publications that are new or amended (a correction re-published under the same CVE) since a prior run today, tracked in a named key-value store scoped to this actor." In practice this means the actor remembers what it has already seen (keyed by record_id) across scheduled runs on the same day, so a run at 9am and a run at 3pm against the same edition won't re-deliver identical entries — only genuinely new listings or corrected ones.

The real event_type values this actor emits are NEW_LISTING, UPDATED, and UNCHANGED (the last one only surfaces when onlyNew is off, per dataset_schema.json's own description: "NEW_LISTING (never seen before), UPDATED (a correction to a known entry) or UNCHANGED (only when Only new records is off)."). This is a deliberately narrower set than the generic NEW_LISTING/UPDATED/CLOSED pattern used elsewhere in the fleet: there is no CLOSED or STATUS_CHANGE event in this actor at all. Per its own AGENTS.md, this was a considered decision, not an oversight — "this domain has no status or lifecycle concept at all — the fleet's usual STATUS_CHANGE/CLOSED categories were considered and explicitly NOT built here... A published law, decree or resolution is a permanent public record; it does not get 'adjudicated' or 'closed' the way a tender does. The one real thing that CAN happen to a published entry is a correction (a fe de erratas) re-published under the same CVE with amended text — that is UPDATED."

The onlyNew boolean (default false) is the actor's delta-mode switch. When true, only records flagged NEW_LISTING or UPDATED are delivered (narrowed further by the eventTypes array, which defaults to both values); when false, every entry from the resolved edition is pushed, including ones already seen (UNCHANGED). The companion is_new boolean field records, independent of the event-type gate, whether that specific record_id was already present in the persisted seen-set when the run started.

On the dataset schema itself: dataset_schema.json for this actor documents 15 fields (rama, ministerio, organismo, descripcion, pdfUrl, cve, seccion, edicion, fecha, scrapedAt, record_id, event_type, is_new, source_url, contentHash) — this actor's own schema does not confirm an 18-field structure, so this overview reports the actual count rather than the fleet's usual figure.

Two more real, source-specific engineering facts worth surfacing to a technical buyer: this actor does not use Crawlee's CheerioCrawler — a live audit found the site serves a real ~22KB page with the full data table to a plain fetch() call, but a ~6KB JS bot-check stub (no table) to got-scraping (CheerioCrawler's default HTTP client) at the same URL, same machine, same moment. The actor is built on native fetch() with its own exponential-backoff retry logic instead. And record_id uses the source's own CVE (Código de Verificación Electrónica) when the link text matches the expected pattern, falling back to a sha1 content hash (seccion|edicion|descripcion|pdfUrl) on the disclosed edge case where it doesn't.

Input Schema & JSON Configuration Example

Field	Type	Default	Description
sections	array (enum, select)	["normas_generales"]	Which sections of the edition to fetch. Not every section publishes content every day - a section with nothing that day returns 404 from the source and is skipped, not treated as an error. Allowed values: normas_generales, avisos_destacados, marcas_patentes, normas_particulares, publicaciones_judiciales, empresas_cooperativas, bom.
maxItems	integer	300	Hard cap on the number of publications returned this run, across all selected sections.
onlyNew	boolean	false	Return only publications that are new or amended (a correction re-published under the same CVE) since a prior run today, tracked in a named key-value store scoped to this actor. Most useful re-running the same day's edition on a schedule to catch late additions/corrections. There is no CLOSED or status-change event in this domain - a published legal notice is never withdrawn or 'closed', it is a permanent public record.
eventTypes	array (enum, select)	["NEW_LISTING", "UPDATED"]	Which kinds of change to deliver when 'Only new records' is on (ignored, everything delivered, when it is off). NEW_LISTING = never seen before. UPDATED = same CVE (or identity), the content changed - a correction. Allowed values: NEW_LISTING, UPDATED.

{
  "sections": ["normas_generales", "marcas_patentes"],
  "maxItems": 150,
  "onlyNew": true,
  "eventTypes": ["NEW_LISTING", "UPDATED"]
}

Output Dataset Sample & Data Dictionary

Field	Type	Description
rama	string | null	Branch of government, e.g. PODER EJECUTIVO
ministerio	string | null	Ministry or top-level agency
organismo	string | null	Sub-agency, e.g. a Subsecretaria
descripcion	string	The publication entry text
pdfUrl	string | null	Direct link to the official PDF
cve	string | null	Codigo de Verificacion Electronica
seccion	string	Which section this entry came from
edicion	string	Edition number
fecha	string	Edition date, DD-MM-YYYY
scrapedAt	string	ISO timestamp of extraction
record_id	string	The CVE when present, otherwise a content hash - see AGENTS.md.
event_type	string	NEW_LISTING (never seen before), UPDATED (a correction to a known entry) or UNCHANGED (only when Only new records is off).
is_new	boolean	True if this id was not in the persisted seen-set when this run started.
source_url	string	The publication's own PDF link when present, otherwise the shared section page.
contentHash	string	sha1 fingerprint used to detect UPDATED (a correction) between runs.

{
  "rama": "PODER EJECUTIVO",
  "ministerio": "MINISTERIO DE ECONOMIA, FOMENTO Y TURISMO",
  "organismo": "Subsecretaria de Pesca y Acuicultura",
  "descripcion": "Extracto de resolucion exenta numero 2.092, de 2026.- Modifica resolucion N 159 exenta, de 2026",
  "pdfUrl": "https://www.diariooficial.interior.gob.cl/publicaciones/2026/09/04/44542/01/2865015.pdf",
  "cve": "2865015",
  "seccion": "normas_generales",
  "edicion": "44542",
  "fecha": "04-09-2026",
  "scrapedAt": "2026-09-04T16:16:42.538Z",
  "record_id": "2865015",
  "event_type": "NEW_LISTING",
  "is_new": true,
  "source_url": "https://www.diariooficial.interior.gob.cl/publicaciones/2026/09/04/44542/01/2865015.pdf",
  "contentHash": "a1b2c3d4e5f6789012345678901234567890abcd"
}

Multi-language Integration Snippets

cURL

curl "https://api.apify.com/v2/acts/stefano_seggio~diario-oficial-cl-monitor/run-sync-get-dataset-items?token=$APIFY_TOKEN" \
  -X POST \
  -H "Content-Type: application/json" \
  -d '{
    "sections": ["normas_generales"],
    "maxItems": 150,
    "onlyNew": true,
    "eventTypes": ["NEW_LISTING", "UPDATED"]
  }'

Python (apify-client)

from apify_client import ApifyClient

client = ApifyClient("YOUR_TOKEN")

run_input = {
    "sections": ["normas_generales"],
    "maxItems": 150,
    "onlyNew": True,
    "eventTypes": ["NEW_LISTING", "UPDATED"],
}

run = client.actor("stefano_seggio/diario-oficial-cl-monitor").call(run_input=run_input)

for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(f"[{item['event_type']}] {item['ministerio']} - {item['descripcion']}")

Node.js (apify-client)

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });

const runInput = {
    sections: ['normas_generales'],
    maxItems: 150,
    onlyNew: true,
    eventTypes: ['NEW_LISTING', 'UPDATED'],
};

const run = await client.actor('stefano_seggio/diario-oficial-cl-monitor').call(runInput);

const { items } = await client.dataset(run.defaultDatasetId).listItems();

for (const item of items) {
    console.log(`[${item.event_type}] ${item.ministerio} - ${item.descripcion}`);
}

Pricing Model Explanation

This actor is pay-per-event, with platform usage included:

Event	Price	What it means here
result	$0.003 / record	Charged once per dataset item pushed. There is no second, cheaper tier for this actor - a gazette entry has no detail/summary distinction, because every publication row is already fully parsed inline from the edition's hierarchical table (branch, ministry, agency, description and PDF link all come from the same pass), so there's no "listing-only" vs "full detail" split to price differently. Every record that reaches the dataset costs the same.
apify-actor-start	$0.00005 / run	Charged once per run, regardless of how many sections are selected or how many records are returned.

Delta mode and billing: when onlyNew is set to true, the actor only pushes dataset items for entries that are NEW_LISTING or UPDATED (further narrowed by eventTypes). An entry that is already known and unchanged is not delivered at all — it is not pushed as an UNCHANGED record and then charged at $0, it is simply never created as a dataset item in the first place, so it is never a chargeable result event. This is what makes scheduling the actor intraday economical: re-running the same day's edition every few hours to catch late additions or corrections only bills for what actually changed, not for re-confirming everything that didn't.

---

Sources for this overview: .actor/actor.json, .actor/input_schema.json, .actor/dataset_schema.json, .actor/output_schema.json, README.md, and AGENTS.md from the actor's real repository at C:\Users\Stef\apify-portfolio\diario-oficial-cl-monitor. No CHANGELOG.md exists in this repo; V2 architecture detail came from AGENTS.md's "Delta engine v2 (2026-09-08)" section instead.