Added list of abbreviations for addresses in English in Pakistan

[bgo-eiu]

You can take a look at some business directories like these to see how they are used in addresses:
https://www.dibpak.com/wp-content/uploads/2020/01/DubaiIslamicBank-Branch-Contact-Directory-Jan-2020_.pdf
https://psopk.com/files/station_finder/list_of_stations.pdf

I tried to make sure the terms in the YAML are all forms it makes sense for a geocoder to expand, some of the terms you see in the address listing above may look abbreviation-like but are not abbreviations or refer to places more specific than would be common in addresses. Let me know if you have any questions

[lonvia]

Thank you for compiling the list. Just some observations from my side:

There are a couple of very generic terms in the list. They have limit use in the name search as currently implemented by Nominatim but they will increase the database quite a bit because they are so frequent. So I'd prefer to leave them out. These are: particles (between, opposite) and units (number, feet, mile, kilometer, etc.).

Also it is not necessary to repeat abbreviations that already appear in the generic English part (boulevard, street, road, etc.). Better to have a shorter list of abbreviations that are truly specific to Pakistan.

[lonvia]

Taluka and Tehsil appear to be synonyms rather than abbreviation forms. Nominatim doesn't have a proper handling for that yet, but the abbreviation list is not the best place.

[bgo-eiu]

Thank you for your feedback - good to know about the common words and overlapping abbreviations, I will revise to be more specific. Are some of the common ones that are abbreviated in a very short form often in Pakistan still acceptable? For example, Block -> B I don't think is as common in other countries which use English in addresses.

Taluka and Tehsil are synonyms yes, I was on the fence about including it. I do try to add both to the Wikidata "other forms" fields where applicable, or to alt_name at least, though the administrative subdivisions are not complete in OSM for Pakistan yet.

[lonvia]

block, floor, house and sector are indeed a little bit problematic because of their different uses world-wide. Thing is: the abbreviations here are currently applied only to names in the OSM database. So it only makes sense to add an abbreviation if there is really a name tag that uses it.

Looking at the current Pakistan data, there is a clear use for 'block' and 'sector'. It make sense to leave that in.

The few usages of 'floor' are rather questionable and will be filtered out by Nominatim (Example: the housenumber in this node). In all examples I saw, the tag addr:floor should rather be used in stead of the actual word. So let's remove 'floor' abbreviations. (There is an open issue to handle unit and floor numbers, see #587.)

'house' is a bit of a mixed case. There are quite a few uses but I suspect that in many cases the abbreviation shouldn't be used, e.g. for White House Grammar School. I'd rather be on the conservative side for the moment and leave it out, too. We can revisit the decision later.

Regarding Taluka and Tehsil, adding alt_name is indeed the best option, although I see that there are quite a few of those. We will need to sort out synonyms eventually in Nominatim.