HarperDB-SDK 0.2.0

Simple SDK for HarperDB.

harperdb

Python3 implementations of HarperDB API functions. Also provides wrappers for an object-oriented interface.

Installation

pip3 install harperdb

Requirements

• Python>=3.4
• requests~=2.0
• responses~=0.10 (required for testing only)

---

harperdb.HarperDB

Each instance of HarperDB represents a running HarperDB instance at a URL, passed to the constructor. Optionally implement Basic Auth as keyword arguments. HarperDB API functions are exposed as instance methods, which produce and consume JSON following the API documentation.

import harperdb
db = harperdb.HarperDB(
    url=HARPERDB_URL,
    username=HARPERDB_USERNAME,
    password=HARPERDB_PASSWORD)

Instance Parameters:

• url (string): Full URL of HarperDB instance
• username (string): (optional) Basic Auth username
• password (string): (optional) Basic Auth password
• timeout (float): Seconds to wait for a server response, default 10

Instance Attributes:

• token (string): Value used in Authorization header, or None. The value is generated automatically when instantiated with both username and password
• timeout (float): Seconds to wait for a server response
• url (string): Full URL of HarperDB instance

Instance Methods:

These methods expose the HarperDB API functions, and return JSON from the target database instance at HarperDB.url

Schemas and Tables:

• describe_all()
• create_schema(schema)
• describe_schema(schema)
• drop_schema(schema)
• create_table(schema, table, hash_attribute)
• describe_table(schema, table)
• drop_table(schema, table)
• drop_attribute(schema, table, attribute)

NoSQL Operations:

• insert(schema, table, [records])
• update(schema, table, [records])
• delete(schema, table, [hashes])
• search_by_hash(schema, table, [hashes], get_attributes=['*'])
• search_by_value(schema, table, search_attribute, search_value, get_attributes=['*'])

SQL Operations:

• sql(SQL)

CSV Operations:

• csv_data_load(schema, table, path, action="insert")

Jobs:

• get_job(id)

---

harperdb.wrappers.HarperDBWrapper

HarperDBWrapper provides a high-level, object-oriented interface for HarperDB. From this top-level object an application programmer can make references to schemas, tables, and records, while making minimal transactions with the server when values are used or modified. Each instance of HarperDBWrapper represents a running HarperDB instance at a URL, passed to the constructor. Optionally implement Basic Auth as keyword arguments.

Schemas are subscriptable by name, and iterating yields instances of HarperDBSchema. The length of a HarperDBWrapper instance returns the number of schemas in the target database. HarperDB API functions are implemented as low-level instance methods, which produce and consume JSON following the API documentation.

import harperdb
db = harperdb.wrappers.HarperDBWrapper(
    url=HARPERDB_URL,
    username=HARPERDB_USERNAME,
    password=HARPERDB_PASSWORD)
dev_schema = db.create_schema('dev')
len(db)  # returns 1
db['dev']  # returns dev_schema
for schema in db:
    schema.name  # returns "dev"

Schemas can be dropped using HarperDB.drop_schema, or using the del keyword and HarperDBSchema.name value like a dictionary:

db.drop_schema('dev')
# same as
del db['dev']

Instance Parameters:

• url (string): Full URL of HarperDB instance
• username (string): (optional) Basic Auth username
• password (string): (optional) Basic Auth password
• timeout (float): Seconds to wait for a server response, default 10

Instance Attributes:

• token (string): Value used in Authorization header, or None. The value is generated automatically when instantiated with both username and password
• timeout (float): Seconds to wait for a server response
• url (string): Full URL of HarperDB instance

High-Level Instance Methods:

• create_schema(name): Create a schema, returns HarperDBSchema
• drop_schema(name): Drop a schema

Low-Level Instance Methods:

These methods expose the HarperDB API functions, and return JSON from the target database instance at HarperDBWrapper.url

Schemas and Tables:

• _describe_all()
• _create_schema(schema)
• _describe_schema(schema)
• _drop_schema(schema)
• _create_table(schema, table, hash_attribute)
• _describe_table(schema, table)
• _drop_table(schema, table)
• _drop_attribute(schema, table, attribute)

NoSQL Operations:

• _insert(schema, table, [records])
• _update(schema, table, [records])
• _delete(schema, table, [hashes])
• _search_by_hash(schema, table, [hashes], get_attributes=['*'])
• _search_by_value(schema, table, search_attribute, search_value, get_attributes=['*'])

SQL Operations:

• _sql(SQL)

CSV Operations:

• _csv_data_load(schema, table, path, action="insert")

Jobs:

• _get_job(id)

---

harperdb.wrappers.HarperDBSchema

Tables are subscriptable by name, and iterating yields instances of HarperDBTable. The length of a HarperDBSchema instance returns the number of tables in the schema. Schema metadata is contained in instance attributes.

You should never need to instantiate this class directly, use HarperDBWrapper.create_schema instead.

dog_table = dev_schema.create_table(
        name='dog',
        hash_attribute='id')
len(dev_schema)  # returns 1
dev_schema.database  # returns db

Tables can be dropped using HarperDBSchema.drop_table, or using the del keyword and HarperDBTable.name value like a dictionary:

dev_schema.drop_table('dog')
# same as
del dev_schema['dog']

Schemas can be dropped using the instance method dev_schema.drop().

Instance Attributes:

• name (string): Name of this schema
• database (HarperDBWrapper): Instance of the parent database

Instance Methods:

• create_table(name): Create a table, returns HarperDBTable
• drop(): Drop this schema
• drop_table(name): Drop a table

---

harperdb.wrappers.HarperDBTable

Records are subscriptable by hash_attribute, but HarperDBTable is not iterable. The length of a HarperDBTable instance returns the number of records in the table. Table metadata is contained in instance attributes.

You should never need to instantiate this class directly, use HarperDBSchema.create_table instead.

HarperDBTable.upsert inserts a record from a dictionary, and updates records if a value is given for the table's hash_attribute and a matching record is found in the table.

penny = dog_table.upsert({
    'id': 1,
    'dog_name': 'Penny',
    'owner_name': 'Kyle',
    'breed_id': 154,
    'age': 5,
    'weight_lbs': 35,
    'adorable': True,
})
dog_table[1]  # returns penny
dog_table.record_count  # same as len(dog_table)
dog_table.__createdtime__  # returns int, Unix time with milliseconds
dog_table.created_time  # returns datetime.datetime
dog_table.hash_attribute  # returns "id"
dog_table.schema  # returns dev_schema

HarperDBTable.upsert accepts either a dictionary of record data, or a list of such dictionaries, returning an instance of HarperDBRecord for each record. Any records skipped by the server are omitted from the return value.

Use HarperDBTable.upsert_from_csv to load record data in bulk from a CSV file. Returns an instance of HarperDBRecord for each record. Any records skipped by the server are omitted from the return value.

Records can be deleted using HarperDBTable.delete, or using the del keyword and HarperDBTable.hash_attribute value like a dictionary:

dog_table.delete(1)
# or
del dog_table[1]

Searching by a record value returns a list of matching HarperDBRecord instances.

dog_table.search_by_value(
    attribute='name',
    value='penny')  # returns a list containing penny

Tables can be dropped using the instance method dog_table.drop().

Instance Attributes:

• attributes (list): All record attributes (string) in this table
• created_time (datetime.datetime): equal to __createdtime__
• hash_attribute (string): Primary key of this table
• id (string): Unique identifier assigned to this table
• name (string): Name of this table
• record_count (int): Number of records in this table
• schema (HarperDBSchema): Instance of the parent schema
• updated_time (datetime.datetime): equal to __updatedtime__
• __createdtime__ (int): Epoch time in milliseconds
• __updatedtime__ (int): Epoch time in milliseconds

Instance Methods:

• delete(hash): Delete a record by hash value
• drop(): Drop this table
• search_by_value(search_attribute, search_value): Return a list of matching HarperDBRecord instances.
• upsert(record): Insert a record from a dictionary, or list of dictionaries. If a value is given for the table's hash_attribute, and this table has a matching record, that record will be updated. Any records skipped by the server will be omitted from the return value. Returns HarperDBRecord, or a list of HarperDBRecord instances.
• upsert_from_csv(path): Insert records from a CSV file, with headers in the first row. Any records which have a value for the table's hash_attribute will be updated. Any records skipped by the server will be omitted from the return value. Returns a list of HarperDBRecord instances.

---

harperdb.wrappers.HarperDBRecord

Record data is subscriptable by record data key, and supports item assignment. Record metadata is stored in instance attributes.

You should never need to instantiate this class directly, use HarperDBTable.upsert instead.

penny['owner_name']  # returns "Kyle"
penny['age'] = 6  # Happy Birthday!
penny.__updatedtime__  # returns int, Unix time with milliseconds
penny.updated_time  # returns datetime.datetime
penny.table  # returns dog_table

Records can be deleted using the instance method penny.delete().

HarperDBRecord.to_dict() returns a dictionary of the record.

Instance Attributes:

• created_time (datetime.datetime): equal to __createdtime__
• table (HarperDBTable): Instance of parent table
• updated_time (datetime.datetime): equal to __updatedtime__
• __createdtime__ (int): Epoch time in milliseconds
• __updatedtime__ (int): Epoch time in milliseconds

Instance Methods:

• delete(): Delete this record
• to_dict(): Returns record data as a dictionary

---

harperdb.exceptions.HarperDBError

Raised when the server returns an error (500), or a hash is not found.

This is the only Exception raised explicitly.