Starts an asynchronous bulk download of the PDFs of the selected payables
("cuentas por pagar"). The response is 202 Accepted with a PdfExport whose status
starts as fetching; poll GET /v1/erp/payouts/payables/pdf_exports/{id} until it is
completed (or failed/expired) and then download the ZIP files in files[].url.

Selecting payables

• The payable filters go in the querystring, exactly as in
  GET /v1/erp/payouts/payables (List all payables) (e.g. amount={"$lt": 11000}). See
  How to filter results.
• The control parameters (company_id, scope, ids, excluded_ids,
  include_attachments) can go in the JSON body or in the querystring. company_id and
  scope are required in one of the two places (422, payables/pdf_export_scope, if
  missing). Exports belong to the user, not to a company: these endpoints do not use the
  company-id header, and it does not replace company_id.
• scope=all exports every payable matching the filters minus excluded_ids;
  scope=partial exports only ids (ids from another company or not visible to the caller
  are ignored).
• Only payables of companies where the caller can list payables are included.

Limits

• At most 2.000 payables per export (422, payables/pdf_export_limit). Narrow the
  filters or split the export.
• One export in flight per user (fetching or zipping). A second request returns 409
  with the pdf_export_id of the running export, so you can resume polling it.
• The ZIP files are kept for 7 days after completion; then the export becomes expired.

Result

One ZIP per company (<company name>_<date_from>_<date_to>.zip) with one folder per document
type that has documents (folders such as Facturas/, Boletas de honorarios/, Reembolsos/,
Anticipos/ or Otros/), and an indice.csv (UTF-8 with BOM, ; separator) with the columns
Empresa;Fecha emisión;Razón social emisor;RUT emisor;Tipo;Folio;Monto;Estado;Motivo;Archivo.
Estado is incluido or falló; failed documents are listed with their Motivo, and are
summarized in failures_by_reason. When the export finishes, the user is also notified in-app
and by email, according to their notification preferences.

Polling example

# 1. Create the export (filters in the querystring, control params in the body)
created=$(curl -s -X POST 'https://api.cardda.com/v1/erp/payouts/payables/pdf_exports?amount=%7B%22%24lt%22%3A11000%7D' \
  -H 'Authorization: Bearer <api_key>' \
  -H 'Content-Type: application/json' \
  -d '{"company_id": "550e8400-e29b-41d4-a716-446655440000", "scope": "all"}')
# 202 {"id": "0d6f8a52-...", "status": "fetching", ...}: keep its id.
# 409 {"error_code": "payables/pdf_export_in_progress", "pdf_export_id": "..."}: resume that one.
export_id=$(echo "$created" | jq -r '.id // .pdf_export_id // empty')
if [ -z "$export_id" ]; then echo "$created" >&2; exit 1; fi  # 422: see error_code

# 2. Poll every few seconds until status is "completed", "failed" or "expired"
while true; do
  export=$(curl -s "https://api.cardda.com/v1/erp/payouts/payables/pdf_exports/$export_id" \
    -H 'Authorization: Bearer <api_key>')
  status=$(echo "$export" | jq -r .status)
  echo "$status $(echo "$export" | jq -c .progress)"
  case "$status" in completed|failed|expired) break ;; esac
  sleep 5
done

# 3. Download each ZIP. The URLs are signed for 1 hour and re-signed on every GET:
#    if a link expired, GET the export again to get a fresh one.
echo "$export" | jq -r '.files[] | "\(.url)\t\(.filename)"' |
  while IFS=$'\t' read -r url name; do curl -s -o "$name" "$url"; done