aiohpprinter

Source Code: https://github.com/chemelli74/aiohpprinter

Python library to get information from HP printers

aiohpprinter talks directly to an HP printer's embedded web server (EWS) - the same /DevMgmt/*.xml, /ePrint/*.xml and /IoMgmt/* endpoints the ha-hpprinter Home Assistant integration uses - and exposes the result as typed, async Python instead of raw XML. It has no Home Assistant dependency and can be used from any async Python application.

Installation

Install this via pip (or your favourite package manager):

pip install aiohpprinter

Basic usage

import asyncio

from aiohpprinter import HpPrinter


async def main() -> None:
    async with HpPrinter("192.168.1.100") as printer:
        data = await printer.update()

        print(data.online)
        if data.device:
            print(data.device.make_and_model)
        for consumable in data.consumables.values():
            print(consumable.consumable_id, consumable.percentage_level_remaining)


asyncio.run(main())

update() is the one-call aggregate: it fetches every endpoint the printer answers and returns an HpPrinterData snapshot. Each endpoint is also available individually - printer.device(), printer.status(), printer.consumables(), printer.printer_usage(), printer.scanner_usage(), printer.copy_usage(), printer.fax_usage(), printer.adapters(), printer.eprint(), printer.wifi() - for callers that only need one piece of data, or that want failures raised instead of absorbed (see Exception handling below).

Session handling

HpPrinter accepts an optional session: aiohttp.ClientSession:

• No session given: HpPrinter creates its own aiohttp.ClientSession when entering async with, and closes it automatically on exit (or when close() is called explicitly). Use this for a standalone script.
• Session injected: the caller owns that session and is responsible for closing it - HpPrinter never closes a session it did not create. This is the recommended mode for an application such as Home Assistant that manages one shared session and its connection pooling itself.

import aiohttp

async with aiohttp.ClientSession() as session:
    async with HpPrinter("192.168.1.100", session=session) as printer:
        device = await printer.device()

    # `session` is still open here; `printer`'s own state was released.

Supported HP functionality

Only the status endpoint is treated as required (it is what is_online() and update() use to decide whether the printer is reachable at all). Every other endpoint is optional: not every HP printer supports scanning, copying, faxing, cloud printing or Wireless Direct, and update() leaves the corresponding field None (or an empty list) instead of failing when one is missing.

Exception handling

HpPrinterError
├── HpPrinterConnectionError   # network/transport failure
├── HpPrinterTimeoutError      # request exceeded the client timeout
├── HpPrinterHttpError         # non-2xx HTTP status (`.status` holds it)
├── HpPrinterParseError        # response body was not valid XML/JSON
└── HpPrinterUnsupportedError  # reserved for a required feature genuinely absent

The single-endpoint methods (device(), status(), consumables(), ...) raise these normally, so a caller that wants to distinguish "printer is unreachable" from "printer sent garbage" can catch the specific subclass it cares about, or HpPrinterError to catch anything.

update() and is_online() are the exception to that: they are explicitly designed to tolerate a printer that only partially answers (see Supported HP functionality above), so they catch every HpPrinterError from the optional endpoints internally and never raise for that reason. update() still lets a failure while checking is_online() be represented as online=False rather than an exception, since that is the question it exists to answer.

Adding a New Printer Model

aiohpprinter's test suite is fixture-driven: every directory under tests/fixtures/ is picked up automatically by both the parser suite (tests/test_parsers.py) and an end-to-end update() test that serves the fixture's files from a local HTTP server (test_update_fixture in tests/test_client.py). No new test function is required.

1. Capture the printer with the fixture script, giving its IP address or hostname:

   python scripts/capture_fixture.py 192.168.1.100
   

   The model name, and so the fixture directory, is taken from the printer's ProductConfigDyn.xml: when it only reports its family (MakeAndModel "HP OfficeJet Pro 9020 series"), the variant from SKUIdentifier ("9022e") replaces the series number. Pass --model "..." to override it.

   This creates e.g. tests/fixtures/hp_officejet_pro_9022e/ with:

   • every endpoint aiohpprinter reads (ProductStatusDyn.xml is required; any other endpoint answering 404 is skipped, since that model doesn't support that feature). Adapters is saved as Adapters.json when the printer answers it as JSON;
   • every other resource the printer lists in /DevMgmt/DiscoveryTree.xml (pass --no-discovery to skip these);
   • expected.json, generated from what the current parsers return;
   • a README.md recording the fixture's provenance.

   Serial numbers, IDs, MAC/IP addresses, hostnames, SSIDs, e-mail addresses and credentials are replaced with REDACTED. Use --port / --ssl for a non-default EWS, and --force to recapture over an existing directory.

2. Review the captured files for anything identifying the redaction missed, and check expected.json: it records what the parsers return today, so a wrong or missing value there is a parser bug to fix, not a value to keep.

3. Fix the relevant parser in src/aiohpprinter/parsers/ if a field is missing or wrong, correct expected.json to match, then run the suite: uv run pytest (or uv run pytest -k <model> for just the new fixture).

Known limitations

• Full captures (every endpoint aiohpprinter reads plus every DiscoveryTree.xml resource) exist for two models: HP OfficeJet Pro 9022e and HP ENVY Photo 7830 All-in-One Printer. An HP Color LaserJet M255dw is covered by the endpoints aiohpprinter reads only, and an HP LaserJet 200 color M251nw by ProductUsageDyn.xml only. None of them reports an ePrint PrinterID, so that field is only covered by a unit test. If you have another printer, contributing a capture per Adding a New Printer Model above is very welcome.
• A field is looked up at one fixed XML path per printer generation (tolerant of a PEID attribute appearing or not, see aiohpprinter.parsers.common.text), but not of a field moving to a different element entirely on some model; that would need a fixture demonstrating the variation and a parser update.
• Fax usage has no real-world fixture with non-zero counters: both full captures report a FaxApplicationSubunit with every value at 0.

Manual testing against a real printer

library_test.py connects to a real printer and prints an update() snapshot:

python library_test.py --host 192.168.1.100

Contributors ✨

Thanks goes to these wonderful people (emoji key):

Simone Chemelli

This project follows the all-contributors specification. Contributions of any kind welcome!

Credits

This package was created with Copier and the browniebroke/pypackage-template project template.