postfinder

Post offices, parcel lockers and post boxes, as an API. Ask what is nearest a coordinate, look up what a postcode covers, or read a suburb's locations.

• Free and keyless. No account, no quota to buy, nothing to configure
• Answers are cached at the edge, so a repeated question is fast and costs nobody anything
• Every row knows the page it belongs to on postfinder.io, so linking out takes no second request
• No dependencies

pip install postfinder

Quick start

from postfinder import PostFinder

pf = PostFinder()

for p in pf.nearby(-37.7404, 144.9633, category="post-offices", country="australia"):
    print(p.name, p.address, f"{p.distance_m} m")

Coburg Post Office 484 Sydney Rd 420 m
Coburg North LPO 12 Elizabeth St 1800 m

Nearest first, within 50km, at most 30 rows. That is the question "where do I post this", which is a different question from "list every post box in Victoria".

The six categories

from postfinder import CATEGORIES   # what nearby() accepts

post-offices, post-boxes, express-post-boxes, parcel-lockers, drop-off-points, collection-points.

A name that is not one of those raises ValueError before a request goes out, rather than spending a round trip to be told 400.

Typeahead

for hit in pf.search("coburg"):
    print(hit.kind, hit.name, hit.postcode, hit.path)

locality Coburg 3058 /en/australia/victoria/coburg/
place Coburg Post Office 3058 /en/australia/victoria/coburg/

Two characters minimum: below that the client returns an empty list without asking, which is what the service answers anyway. Debounce by at least 150ms. Firing on every keystroke spends bandwidth for no better answer.

Postcodes

A postcode is not a suburb. 3058 is Coburg, Coburg North and Merlynston, and an address in any of them is written with the same four digits.

detail = pf.postcode("australia", "3058")
for suburb in detail.localities:
    print(suburb.name, suburb.place_count, suburb.path)

The whole country comes in one response, and it is meant to be kept:

index = pf.postcodes("australia")        # one request
index.suburbs_in("3058")                 # no request, and no rescan
index.suburbs_in("2044")

suburbs_in builds its map on first use and keeps it, so resolving a column of ten thousand postcodes is one pass over the index rather than ten thousand walks through every state.

A location, and a suburb

place = pf.place("k7m2p9x4")
print(place.place.name, place.locality.name, place.path)
print(place.reviews.summary.count, place.reviews.summary.average)

suburb = pf.locality("australia", "victoria", "coburg")
print(suburb.locality.place_count, suburb.count_of("parcel-lockers"))
for p in suburb.places:
    print(p.name, suburb.path_of(p))

A place's public id is permanent. It is minted once and never derived from a source record, so a feed that renumbers its rows does not change it. Store the id, not the name or the path.

Browsing

pf.countries()                                   # every country with pages
pf.country("australia")                          # its states
pf.region("australia", "victoria", limit=100)    # its localities, a page at a time
pf.category("australia", "parcel-lockers")       # counts per state, busiest suburbs

Errors

from postfinder import NotFound, BadRequest, RateLimited, PostFinderError

try:
    pf.place(stored_id)
except NotFound:
    ...     # retired, or never there

NotFound is ordinary rather than a failure: a suburb with no locations has no page, and a location that closed is retired. Every error carries the status, title and detail the service sent, because that is the part that says what to do about it.

Being a good citizen

The API is free and asks for care in return: around a thousand requests a month from one address, results kept rather than re-fetched, typing debounced. If you need more than that, say what you are building at postfinder.io/en/contact/.

Introduce yourself and it is easier to help you before a rate limit does:

pf = PostFinder(contact="https://example.com/about-our-bot")

Attribution

Locations come from OpenStreetMap (ODbL), localities and postcodes from GeoNames (CC BY 4.0). If you publish what you get back, you carry those credits with it. The sources page names each one.

Also available

Looking for Australian street addresses rather than locations? That is Locio: G-NAF address autocomplete, validation and geocoding.

Licence

MIT.