PMS (Campus Print)

Read printer status and the upload queue, and upload documents to the campus PMS.

Authentication

Install the [pms] extra, which includes the legacy RSA login dependency.

from sustech_survival.sso import PMSAuth

auth = PMSAuth()
ok, reason = auth.ensure()
if not ok:
    raise RuntimeError(reason)

ensure() follows the site's Auth/SSoPage entry, discovers the current CAS service callback, and reuses the same in-memory cookie jar through authcenter, CAS and PMS. It verifies the resulting session with Auth/Check. No browser is required. Interactive authentication challenges stop the flow.

login_via_cas() uses this same flow (headless is retained for compatibility). login_password() is an explicit legacy RSA print-account login requiring the [pms] extra. It is not the default CAS path, and an invalid-session response from it does not establish that the campus password is wrong. No automatic fallback to that endpoint occurs.

CLI

sustech pms check
sustech pms jobs --json
sustech pms stations --json

Authentication/read failures exit nonzero. Authentication diagnostics go to stderr so JSON output remains parseable.

Python API and upload receipts

from sustech_survival.pms import pms

client = pms()  # verifies/reuses the PMS authorizer session
jobs = client.list_print_jobs()
preview = client.upload_print("PMS_TEST.pdf", paper="A4", color="bw",
                              duplex="long", copies=1, dry_run=True)
# After reviewing the file/options and obtaining authorization:
receipt = client.upload_print("PMS_TEST.pdf", paper="A4", color="bw",
                              duplex="long", copies=1)
print(receipt.status, receipt.job_id, receipt.verification_url)

Dry runs perform no network calls. A real upload snapshots queue IDs, sends one multipart POST with the correct queue page as BackURL, parses JSON or the same-site result redirect, and reads the queue once. It does not automatically follow upload redirects or repeat the POST.

• status="confirmed", ok=True, uploaded=True: one new matching queue job was read back; job_id identifies it.
• status="unknown", ok=False, uploaded=None: the result could not be confirmed, including a lost response, delayed/failed queue read or multiple new matches. Inspect observed_job_ids, http_status, response_format and verification_error, then query the queue before considering any retry.
• status="rejected", uploaded=False: an explicit rejection and no new matching queue job. An unavailable initial queue snapshot sends no upload.

ok=False alone is never a reason to resend a file. Uploading queues a document; this API does not trigger physical printing.

Queue verification URL: https://pms.sustech.edu.cn/client/new/cprintPc/printDoc.html.

Duplex values and queue labels

The client follows the PMS queue list's long/short-edge labels:

Option	dwDuplex	Queue flag	Queue label
"single" / DUPLEX_SINGLE	1	single	单面
"long" / "双面长边" / DUPLEX_LONG_EDGE	2	vdup	双面长边
"short" / "双面短边" / DUPLEX_SHORT_EDGE	3	hdup	双面短边

Numeric inputs and numeric strings use the same wire value. The upload page reverses the queue's edge labels; aliases, previews and queue records all use the queue convention. This corrects the previous long=3 / short=2 alias and constant mapping. Callers using raw numeric values keep those values.

Queue records preserve duplex_flag and expose duplex_edge="long" for vdup and "short" for hdup. Missing or conflicting flags remain unknown (is_duplex=None, duplex_edge=None).

Other client methods: list_server_groups(), list_stations(group_sn=None), list_scan_jobs(), history(begin=..., end=...), delete_print_job(job_id) and delete_scan_job(job_id). Deletion changes the account and needs authorization.