Support batch deletion

[zcutlip]

A user on the 1Password community forums asked if there was a faster way of using the op CLI tool to delete everything in a vault:
https://1password.community/discussion/comment/664638

I think this is something that pyonepassword could be helpful for without too much work.

The op command can take a list of JSON objects via stdin, which should work with the output (or some modified form) of op item list. We already have item_list(self, categories=[], include_archive=False, tags=[], vault=None), so something like:

def item_delete_batch(self, categories=[], include_archive=False, tags=[], vault=None, batch_size=25):
    ...

... shouldn't be too much trouble

[zcutlip]

I've kinda/sorta prototyped this out

basically, I've added _OPCommandInterface._item_delete_batch(), and tried it out by doing the following:

• get a list of item descriptors from "Test Data 3"
• break the list up into the first 25 objects and the rest
• serialize the first batch to json
• call _item_delete_batch()
• serialize the second batch to json
• call _item_delete_batch()

It seems to work, but it remains to be seen if it's faster than deleting the items one at a time.

In [5]: vault = "Test Data 3"

In [6]: for i in range(0, 30):
   ...:     title=f"Example Login Item {i:02d}"
   ...:     username=f"user_{i:02d}"
   ...:     password = LETTERS_DIGITS_SYMBOLS_20
   ...:     op.login_item_create(title, username, password=password, vault=vault)
   ...:

In [7]: items = op.item_list(vault=vault)

In [13]: batch_1=items[0:25]

In [14]: batch_2=items[25:]

In [16]: batch_str=json.dumps(batch_1)

In [18]: op._item_delete_batch(batch_str, vault)

In [19]: batch_str=json.dumps(batch_2)

In [20]: op._item_delete_batch(batch_str, vault)

[zcutlip]

addressed by #92