v0.4.1

[0.4.1] - 2018-08-16

Added

Changed

Deprecated

Removed

Fixed

- [Documentation] Fixed details on RELEASING.md and updated README.md files

- [Deployment] Updated to the correct version number

Security

v0.4.0

[0.4.0] - 2018-08-15

Added

• [Resources] Add ability to define resource files that are version-independent. [#223]

• [Versions] Add a class to represent Standard Versions. Handles current and proposed formats. [#273]

• [Versions] STANDARD_VERSIONS_SUPPORTED lists all versions of the Standard that are fully supported by pyIATI. [#223]

• [Versions] STANDARD_VERSIONS_MINOR lists all Minor versions of the IATI Standard. [#264]

• [Tests] Separate testing of bytes and str type values for Python 3+. [#286]

• [Tests] Mark tests requiring updates when adding a new version. [#288]

• [Tests] Add another build stage to check that docs build when merging to dev or master. [#292]

• [Tests] Significantly improve test coverage of the resources module. [#297]

Changed

• [Defaults] Make version argument mandatory when accessing Standard content that may differ between versions. [#243]

• [Constants] Move lists of versions to version module. [#280]

• [General] Change default behavior of version argument as per proposal in #218. [#223]

• [Resources] Change the folder containing version-independent data from version-independent to version_independent for consistency. [#297]

• [Resources] create_*_path() functions will raise errors when given an unknown version. [#297]

• [Resources] get_*_paths() functions now verify that paths point to actual files before returning them. [#297]

• [Resources] get_*_paths() functions raise errors when given values that cannot represent a version. [#297]

• [Utility] Move function acting on versions to version module. [#280]

• [Utility] Raise a TypeError rather than a ValueError when convert_xml_to_tree() is given a value of incorrect type. [#286]

• [Versions] The value to represent 'version independent' has been changed to be a value that is not None, and a constant added to specify the exact value. [#281]

Deprecated

Removed

• [Defaults] Remove get_default_version_if_none(). [#243]

• [Defaults] Remove version argument for ruleset_schema() since this is version-independent. [#243]

• [Resources] Functions in the resources module no longer have default values for the version argument. [#297]

• [Compatibility] Drop support for Python 2.7. [#311]

• [Compatibility] Drop support for Python 3.4. [#309]

Fixed

• [Resources] Re-implement support for 2.03 from the ground up to improve correctness and reduce number of undetectable bugs. [#307]

• [Schemas] A tree is now returned from _change_include_to_xinclude() when there are no includes to convert. [#244]

• [Validation] The _check_codes() implementation is no longer fixed to version 2.02. [#291]

• [Validation] Separate two similar but distinct ValueErrors that lxml may raise so that pyIATI treats them differently. [#287]

• [Versions] STANDARD_VERSIONS now lists all versions of the Standard, not just those that are fully supported by pyIATI. [#223]

• [Tests] Correct some very old test fixtures and documentation. [#286]

Security

v0.3.0

[0.3.0] - 2018-02-16

Added

• [Datasets] A Dataset xml_tree may be set with an ElementTree. [#235]

• [Resources] Updated SSOT to latest content as of 2017-11-14. [#237]

• [Resources] Remove SSOT organisation test files that are not valid XML. [IATI/IATI-Schemas#376, #242]

• [Resources] Add get_ruleset_paths() and get_codelist_mapping_paths() to improve consistency of resources module. [#260]

• [Resources] Add Schemas, Codelists and Rulesets for v2.03. [#272]

• [Schemas] Test that multiple Rulesets can be added to a Schema. [#254]

• [Utility] Non-resource files may be loaded using utility functions. [#235]

• [Validation] full_validation() now checks whether a Dataset is IATI XML. [#239]

• [Validation] Test that SSOT organisation test files are valid IATI XML. [#242]

Changed

• [Codelists] complete attribute included in equality comparison and hash calculations. [#247]

• [Codelists] Codes must have a value to instantiate. [#247]

• [Codelists] A number of Codelists have changed from Embedded to Non-Embedded. [IATI/IATI-Codelists-NonEmbedded#220, #272]

• [Defaults] When a specific version is not requested, it will now assume v2.03 rather than v2.02. [#272]

• [Resources] Move load_as_x functions to iati.utilities. [#235]

• [Resources] Rename version-specific resource folders to reduce ambiguity. [#217]

• [Resources] Rename various resource functions to improve clarity. [#259]

• [Rulesets] validate_ruleset() changed from public to private function. [#246]

• [Rulesets] case attribute on a Rule changed from public to private. [#252]

• [Rulesets] context attribute on a Rule changed to read-only property. [#253]

• [Validation] _check_is_iati_xml() will raise a TypeError when given a non-dataset. This replaces an undocumented AttributeError. [#239]

Removed

• [Documentation] Stop tracking auto-generated docs templates. [#236]

• [Rulesets] ruleset attribute removed from Rulesets. [#246]

• [Tests] iati.tests.resources.get_test_ruleset_path() removed due to no calls to function. [#256]

Fixed

• [Codelists] Fixed impossible XPath in Codelist Mapping File. [IATI/IATI-Codelists#119, #229]

• [Codelists] Sort Codes in a Codelist before hashing so that Codelists with the same Codes always have the same hash. [#247]

• [Defaults] Test and document ValueErrors that can be raised by functions in iati.default. [#241]

• [Documentation] Minor fixes to README. [#266, #267]

• [Rulesets] name attribute on a Rule changed to read-only property. [#251]

• [Rulesets] Equal Rulesets are now deemed to be equal. [#249]

• [Validation] Prevent XPathEvalErrors occurring when given a Codelist Mapping XPath that identifies something other than an attribute. [#229]

• [Validation] Datasets with an xml:lang attribute no longer raise a KeyError upon performing Codelist validation against a Schema populated with the Language Codelist. [#226]

• [Tools] Updated dependencies to latest version as of 2018-01-25 and fixed newly identified linting issue. [#269]

v0.2.0

[0.2.0] - 2017-11-07

Added

• [Codelists] Implement the complete attribute. [#45]

• [Codelists] Codes may have equality compared with strings - the value of a Code is compared. [#45]

• [Codelists] Add v2.02 Codelist mapping file. [#45]

• [Documentation] Clarify version support in README. [#216]

• [Exceptions] Add an error_log attribute to ValidationErrors. [#45]

• [Resources] Add method to load data files relating to pyIATI (rather than the IATI Standard, or tests). [#45]

• [Resources] Allow test files to be located within sub-folders by including slashes (/) in the name. [#45]

• [Resources] Detect encoding of files that are not UTF-8. [#45]

• [Validation] Change from validate.py to validator.py to improve readability of code using this module. [#86]

• [Validation] Add a YAML file containing error information. [#117]

• [Validation] Check whether a string is valid XML - truthy. [#45]

• [Validation] Check whether a string is valid XML - detailed error information. [#45]

• [Validation] Provide custom error messages when lxml returns errors for a string that is not XML. [#90]

• [Validation] Check whether a Dataset is valid against an IATI Schema - truthy. [#45]

• [Validation] Check whether a Dataset is valid against an IATI Schema - detailed error information. [#45]

• [Validation] Provide custom error messages when lxml returns errors for Datasets that do not contain valid IATI XML. [#92]

• [Validation] Check whether attributes in a Dataset have values from Codelists where required - truthy. [#45]

• [Validation] Check whether attributes in a Dataset have values from Codelists where required - detailed error information. [#45]

• [Validation] Check whether a Dataset conforms with Rules in a Ruleset - truthy. [#58]

• [Validation] Check whether a Dataset conforms with Rules in a Ruleset - basic information about which Rules fail. [#58]

• [Validation] Add a ValidationErrorLog containing ValidationErrors to track validation errors. [#87]

• [Validation] Rudimentary differentiation of errors and warnings. [#45]

• [Tests] Add a range of test XML files. [#45]

• [Tests] Add some missing tests. [#45]

• [Tests] Add a somewhat normal-looking string to use for fuzzing. [#113]

Changed

• [Codelists] The default name for a Code is now an empty string. [#45]

• [Codelists] The name of a Code is no longer included when computing the hash. [#45]

• [Datasets] pyIATI validation functionality used to determine whether a string is XML. This changes the types of Error that may be raised when updating the XML that a Dataset represents. [#95]

• [Tests] Re-organise test data to use folders to separate logical groups. [#58]

Fixed

• [Codelists] The names of Codes are detected in CLv2 XML Codelists (when there are no <narrative> elements).

• [Documentation] Corrected some out-of-date documentation. [#45]

• [Rulesets] Update date_order XPaths in Standard Ruleset. [IATI/IATI-Rulesets#31]

v0.1.1

[0.1.1] - 2017-10-25

Fixed

• [Deployment] Include resource data in package distribution [#195].

• [Docs] The README is up-to-date, and so better matches the code. [#198].

Removed

• [Deployment] Remove attempt at excluding test files from deployed package [#202].

v0.1.0

Initial release for PyPi.

[0.1.0] - 2017-10-19

Added

• [Codelists] Blank Codelists may be created and given a name.

• [Codelists] Codelists may be created from a string conforming to the Codelist Schema.

• [Codelists] Codelists contain a set of Codes.

• [Codelists] Codelists may be output as a XSD simpleType restriction containing enumeration elements.

• [Codelists] Codelist equality is based on whether the names and contained Codes are the same.

• [Codelists] Codes may be created from an optional value and name.

• [Codelists] Codes may be output as a XSD enumeration to be contained within a simpleType restriction.

• [Codelists] Code equality is determined based on whether name and value are the same.

• [Constants] A number of constants are defined, though should be deemed private to the library.

• [Datasets] Datasets can represent each Activity and Organisation files (though not at the same time!).

• [Datasets] A Dataset may be created either from a string or a lxml ElementTree.

• [Datasets] The contents of a Dataset can be accessed as each a string and an ElementTree. These values remain in sync with each other.

• [Datasets] The version of the IATI Standard at which a Dataset is defined can be obtained through the version property.

• [Datasets] Subsections of a Dataset (indexed by line number) may be obtained in string format.

• [Defaults] The defaults module provides access to the contents of the SSOT.

• [Defaults] To obtain SSOT contents, the desired version of the Standard must be specified. Should a version not be specified, the latest version will be assumed.

• [Defaults] Specific Codelists from the SSOT may be accessed by name.

• [Defaults] All Codelists from the SSOT may be accessed at once. No differentiation is made between Embedded and Non-Embedded Codelists.

• [Defaults] The Standard Ruleset may be accessed.

• [Defaults] The Ruleset Schema may be accessed. NOTE: This is a JSON (dict) object, not an iati.Schema.

• [Defaults] The Activity Schema may be accessed.

• [Defaults] The Organisation Schema may be accessed.

• [Defaults] By default, the Activity and Organisation Schemas will be populated with all the Codelists and Rulesets at the desired version of the Standard.

• [Defaults] Unpopulated Activity and Organisation Schemas may be accessed by overriding a default argument.

• [Exceptions] There is very rudimentary exception handling.

• [Exceptions] Some exceptions are logged.

• [Exceptions] Some raised exceptions have descriptions to make it possible to consistently determine the cause.

• [Resources] Resources are static files that come with the library, split into 'The Standard' and 'Test Data'.

• [Resources] Resources are stored and accessed based on a particular Decimal Version.

• [Resources] Schemas, Codelists and Rulesets are available for V1.04-2.02.

• [Resources] Schemas are available for V1.03 and prior.

• [Resources] Non-embedded Codelists are deemed to have a single form valid for V1.04-2.02.

• [Resources] Schema test data from the SSOT is present for V1.01-2.02.

• [Resources] Custom test data has been created for V2.02.

• [Resources] Resources may be loaded from disk as bytes, a Dataset, a string or an ElementTree.

• [Rulesets] Blank Rulesets may be created by providing no arguments.

• [Rulesets] Rulesets may be created from a string conforming to the Ruleset Schema.

• [Rulesets] Rulesets contain a set of Rules.

• [Rulesets] A Dataset may have truthy conformance checked against a Ruleset.

• [Rulesets] Rules may be created from a context and a case.

• [Rulesets] Each of the Rules in the Ruleset Schema have been implemented for instantiation and truthy conformance. These are: atleast_one, date_order, dependent, no_more_than_one, regex_matches, regex_no_matches, startswith, sum, unique

• [Rulesets] A Rule base class exists to allow the creation of custom types of Rule.

• [Rulesets] The string representation of a Rule details what must occur for a Dataset to be conformant.

• [Rulesets] A Dataset may have truthy conformance checked against a Rule. NOTE: A Rule may skip or raise an error.

• [Schemas] Schemas may be created by providing a path to a file containing a valid XSD.

• [Schemas] Schemas contain a set of Codelists.

• [Schemas] Schemas contain a set of Rulesets.

• [Schemas] Should a XSD be defined across multiple files, the resulting Schema shall be flattened to allow programmatic access to all of its contents.

• [Schemas] Activity and Organisation Schemas have been implemented.

• [Utility] A new namespace may be added to a Schema.

• [Utility] There is a function to convert from ElementTree to XMLSchema.

• [Utility] There is a function to convert from a string to an ElementTree.

• [Utility] There is a function to work around RFC4627 by preventing duplicate keys in a JSON file.

• [Utility] There is very rudimentary logging functionality.

• [Utility] Various functionality to access Version Numbers is present.

• [Compatibility] Compatible with Python 2.7, 3.4, 3.5 and 3.6.

• [Tests] All code is comprehensively unit tested.

• [Tests] Datasets, Rulesets, and other components permitting user-defined input tested with range of valid and invalid data.

• [Tests] Variety of never-before-seen edge-cases tested for correct behavior.

• [Tests] Rudimentary custom fuzzing used to ensure expected errors are raised.

• [Tools] Documentation is generated with Sphinx, based on Google-style docstrings.

• [Tools] Documentation is presented using the Sphinx Read the Docs theme.

• [Tools] Linting performed with pylint, flake8 and pydocstyle.

• [Tools] Linters are configured to generally follow PEP8 and PEP257, with a few custom modifications such as line length.

• [Tools] Code complexity is assessed using radon.

• [Tools] Unit testing is performed with pytest.

• [Tools] Full testing is performed by Travis.

• [Tools] Python 2.7 compatibility is provided by both future and six. Where a custom workaround is used rather than one of these, python2/3 is present in a comment.

• [XML] lxml utilised for XML functionality.

• [XML] Support XML 1.0.

• [XML] Support XMLSchema 1.0.

• [XML] Support XPath 1.0.