Skip to content

Releases: aschleg/petpy

petpy 2.1.0 Release

16 Aug 13:02
Compare
Choose a tag to compare

The 2.1.0 release of petpy implements several user-defined exceptions to help users debug any
errors that may occur. Although the petpy library attempts to find any invalid passed parameter values and raise the appropriate error before the call to the Petfinder API is made to reduce the number of calls made to the API; however, some errors cannot be caught until after the API call is made. This update introduces these custom, user-defined exceptions for debugging error responses from the Petfinder API. For more information on the Petfinder API error definitions, please see the [Petfinder API documentation (https://www.petfinder.com/developers/v2/docs/#errors).

The following is a list of the new user-defined exceptions.

  • PetfinderInvalidCredentials
    • Raised when a user enters invalid API key and secret key credentials.
  • PetfinderInsufficientAccess
    • Raised when a status code 403 occurs. This error would typically be
      raised when the current authentication token to the Petfinder API has expired and requires the connnection to be re-authenticated.
  • PetfinderResourceNotFound
  • PetfinderUnexpectedError
  • PetfinderInvalidParameters
    • Raised when a 400 status code occurs. The exception will include the invalid parameters detected by the Petfinder API and include those parameters as part of the error output as a JSON object. For more information on this error, please see the Petfinder API documentation.
      • Please note the petpy library will attempt to catch any invalid parameters before the API call is made to avoid extraneous issuage of the API, but if an invalid parameter does get through then this error should help provide the necessary information for users to debug any errors related to their chosen parameters.

The following other changes have been made in the 2.1.0 release.

  • The host and auth attributes of the Petfinder class are now private (to the extent that Python allows private attributes, denoted with an underscore in front of the attribute).

2.0.2 Relase

08 Aug 19:17
Compare
Choose a tag to compare

Minor bug fix release with following:

  • breeds() method now correctly returns data when a single animal type is supplied.
  • animals() method now properly displays the correct error message when the distance parameter is
    0 <= distance <= 500.

petpy 2.0.1 released

27 Jul 17:36
Compare
Choose a tag to compare

The 2.0.1 release of petpy introduces some small bug fixes and a few other minor changes to shore up changes made for the 2.0.0 release.

Full Changelog, which can also be found in the repository's root directory.

  • Fixes the animals() and organizations() method to return all matching search results when the pages parameter
    is set to None.
  • The resulting JSON (dictionary) from the breeds() method when the parameter return_df=False will now consistently
    have types as the key. Prior to this change, if the breeds() method was called with a single animal type, the
    resulting key name in the returned object would be named type, whereas if more than one animal type is specified
    the key name would be types.
  • The distance parameter for the animals() and organizations() parameters will now raise a ValueError if it is
    less than 0.

petpy 2.0.0 Released!

20 Jul 14:44
Compare
Choose a tag to compare

New major release coinciding with the release of v2.0 of the Petfinder API!
The legacy version of the Petfinder API, v1.0, will be retired in January 2020, therefore, the petpy library has
been updated almost from the ground up to be compatible as possible with the new version of the Petfinder API! The
new version of the Petfinder API is a huge improvement over the legacy version, with many changes and additions to
the design of the API itself.

Below is a summary of all the changes made in the release of petpy 2.0.

  • petpy now supports the latest release of Python 3.7
  • Support for Python 2.7 is discontinued as Python 2.7 will be officially discontinued in January 2020.
  • The following methods have been added to petpy to make it compatible with v2.0 of the Petfinder API.
    • animal_types() is used to getting animal types (or type) available from the Petfinder API. The release
      of v2.0 of the Petfinder API added several endpoints for accessing animal types in the Petfinder database.
      This method wraps both Petfinder API endpoints for getting animal types. More information on the animal type
      endpoints in the Petfinder API can be found in its documentation:
    • breeds() is the new method for getting available animal breeds from the Petfinder database. The API
      endpoint documentation is available on the Petfinder API documentation page.
    • animals() is the method for extracting animal data available on the Petfinder API and deprecates the
      pets() related methods. The method wraps both the animals and animal/{id} endpoints of the Petfinder API. The documentation for these endpoints can be
      found in the Petfinder API documentation:
    • organizations() is now the method for extracting animal welfare organization data available on Petfinder
      and deprecates previous shelter() related methods and endpoints. The organizations() method wraps both
      the Petfinder API organizations and organizations/{id} endpoints. The Petfinder API documentation for
      these two endpoints can be found below:
  • The following methods have been removed as they are no longer valid endpoints with the release of v2.0 of the
    PetFinder API.
    • pet_get_random()
    • shelter_list_by_breed()
    • shelter_get_pets()
  • General refactoring and code clean-up.

petpy 1.8.2 release

19 Sep 13:14
Compare
Choose a tag to compare

This release fixes some import issues with pandas (issue #8) as well as implementing some additional checks around parameter inputs. The latter was officially released in version 1.8.1.

Version 1.8.2

  • Add pandas back as an installation requirement as it is included in top-level imports. pandas is
    still not necessary to use the primary functionality of petpy.

Version 1.8.1

  • Implement check to make sure count parameter is not larger than 1,000 records (per the Petfinder API
    limits). If count exceeds 1,000 a ValueError is raised.
  • Add check for ensuring the number of total records to return does not exceed 2,000 when paging results with
    the pages and count parameters. If the desired amount of records is higher than 2,000, a
    ValueError is raised.
  • Remove Python 3.3 support. Although petpy should work fine for those still using Python 3.3, testing for 3.3
    has been discontinued.

petpy 1.8.0 Release

16 May 02:45
Compare
Choose a tag to compare

General refactoring of the petpy library to remove unneeded methods from being exposed when importing the library. The best way to import and begin using petpy is from petpy import Petfinder or, less optimally, import petpy, then calling the Petfinder class by petpy.Petfinder.

petpy 1.7.2 Release

20 Apr 02:01
Compare
Choose a tag to compare

The 1.7.2 release of petpy fixes a variety of bugs and changes to requirements. The specific changes notes are:

  • There is now a proper message when the daily API call limit is exceeded. Prior to the change, methods would
    return a JSONDecodeError.
  • The Python 2 to 3 compatibility package six, has been added as a requirement for package installation.
    The six library is lightweight and ensures the petpy package works properly for Python 2 and 3.
  • Numpy is no longer required for installing the package. Numpy's nan was originally used to denoted shelters
    animals that were not found in the Petfinder database. The value returned when a shelter or animal is not found
    is now 'na'.

The full change history can be found in the documentation.

petpy 1.7.1 Release

04 Apr 13:08
Compare
Choose a tag to compare

The 1.7.1 release of petpy is intended to fix a couple of bugs when using the shelter methods and transforming the results into pandas DataFrames. This version and other version changes can also be found in the complete petpy documentation.

Release Notes

  • Fix to the shelter_get() method for handling empty responses when no shelters returned for when
    the parameter return_df = True.
  • Fix to getting pets available at a shelter with shelter_get_pets() when the parameter
    return_df = True when the given shelter does not return any available animals.

petpy 1.7.0 Release

15 Mar 23:19
Compare
Choose a tag to compare

The 1.7.0 release introduces a variety of fixes and changes to the package. The full release notes can be found below, as well as in the petpy version documentation.

Release Notes

  • Refactoring of library to clean up code files.
  • Fixed a bug with the shelter_get_pets() method that caused an error to be thrown when there is only
    one pet record and the parameter return_df = True.
  • Many changes to simplify expressions and internal code within methods.
  • The Petfinder class method names and parameters have been renamed to be PEP8 compatible. I apologize as this
    will cause backward compatibility issues upon updating for anyone using previous versions. The original
    intention of the naming was to reflect the Petfinder API's method names as closely as possible, but after
    further consideration and given the relative short life of the library, I think the change is necessary for the
    future growth and maturity of the package.
    • How the methods interact with the Petfinder API is unchanged, thus there is no immediate need to update your
      version of petpy if it will break any current code.
  • The Github repo README has been updated with the new API methods.
  • Below is a table detailing the changed method names:
Previous Method Name New Method Name
pet_getRandom() pet_get_random()
shelter_getPets() shelter_get_pets()
shelter_listByBreed() shelter_list_by_breed()
  • The following lists the method parameter names that have changed with the release:
Previous Parameter Name New Parameter Name
petId pet_id
shelterId shelter_id

petpy 1.6 released

18 Feb 23:26
Compare
Choose a tag to compare

This release removes pandas as an installation requirement for the package. Although pandas is required to convert the API results into a DataFrame, this is optional and not necessary to the building or use of the package itself.