Skip to content

CHANGELOG.md

Latest commit

 

History

History
111 lines (88 loc) · 8.25 KB

CHANGELOG.md

File metadata and controls

111 lines (88 loc) · 8.25 KB

Notation Changelog

All notable changes to this project will be documented in this file. The format is based on Keep a Changelog and this project adheres to Semantic Versioning.

2.0.0 (2020-09-04)

This is a big major release with lots of improvements and some breaking changes. Please read the changes below and re-run your application tests after you upgrade to v2.

Added

  • Array index notation and object bracket notation support. This brings ability to manipulate arrays with indexes and objects with non-standard property names (keys). See documentation for more info.
  • strict: boolean class option that specifies whether to throw when a notation path does not exist on the source. (Note that #inspectGet() and #inspectRemove() methods are exceptions). It's recommended to set this to true and prevent silent failures if you're working with sensitive data. Regardless of strict option, it will always throw on invalid notation syntax or other crucial failures.
  • restrictive: boolean option for Notation#filter() instance method that specifies whether negated items strictly remove every match. Note that, regardless of this option; if any item has an exact negated version; non-negated is always removed. Default is false. This option is also available for .normalize() and .union() static methods of Notation.Glob class. For example; ['*', 'name', 'car.model', '!car.*'] will normalize to ['*', '!car.*', 'car.model']. In restrictive mode, it will normalize to ['*', '!car'].
  • Notation.join() static method for joining given notes into a notation string.
  • Notation.split() static method for splitting given regular notation into a notes array.
  • Notation.Glob.split() static method for splitting given glob notation into a notes array.
  • Notation.Glob#intersect() instance method for checking whether current glob has any intersection with the given glob/notation.
  • Notation.Glob#covers() instance method for checking whether given glob covers another specific glob.
  • Ability to "insert" array item at given index via Notation#set() method; instead of overwriting only.
  • 100% full test coverage.

Changed

  • Breaking » (For browser) dropped support for IE 10 and below.
  • Breaking » (For Node) dropped support for Node v8 and below (might still work though).
  • Breaking » Notation is a named export now. (Meaning require or import within curly brackets. e.g. const { Notation } = require('notation'))
  • Breaking » Now that bracket-notation support is added, there will be some changed behaviour. Notation that has an array value is now also notated with a bracket-index for each item, instead of only the key (array name) itself. e.g. prop1.prop2[3]
  • Breaking » Improved notation and glob validation. Now we strictly validate each note of a notation against EcmaScript variable syntax, array index notation and object bracket notation. For example:
    • x[y], x.1, x.y-z, x.@ are incorrect and will never match.
    • x["y"], x['1'], x["y-z"], x["@"] are correct.
    • x.* is valid (wildcard) for glob notation but invalid (object property) as regular notation.
    • x[*] is valid (wildcard) for glob notation but invalid (array index) as regular notation.
    • x['*'] just indicates a property/key (star), not a wildcard. Valid regular notation.
  • When an element is removed from an array; that item will be emptied and indices will be preserved (and will NOT shift). e.g. 
    Notation.create([0, 1, 2]).remove('[1]').value
// will return [0, (empty), 2]
    The empty item can be treated as undefined in the sparse array result above. Set preserveIndices option to false to change this behavior. 
    const options = { preserveIndices: false };
Notation.create([0, 1, 2], options).remove('[1]').value
// will return [0, 2]
  • Breaking » Changed/improved normalization and union logic. Also now, introducing (linear) intersections within normalization and (cross) intersections within union. An intersection glob is only produced when needed. For example; previously, ['!*.y', 'x'] would normalize as is but this had side-effects when union'ed with another glob list. Now it normalizes to ['x', '!x.y']. Notice that in this example, intersection glob '!x.y' is added and '!*.y' is removed. See documentation for more.
  • Breaking » All levels of negated globs are now respected. For example; when an object is filtered with ['*', '!x.*'], the x property would be completely removed. Now, x will be emptied instead (i.e. {}) due to explicit trailing wildcard star. To completely remove the x property; filtering globs should be ['*', '!x'].
  • Passing undefined as the source object will now throw. This prevents accidental empty initialization. To initialize a Notation instance with a new empty object, just omit the argument or explicitly define it. e.g. new Notation() or new Notation({}).
  • Breaking » #inspectGet() method of Notation class is renamed to #inspectGet() for compatibility with Node.js version 10+. See DEP0079.
  • Improved collection deep cloning.
  • Updated globs comparison/sort logic.

Removed

  • Breaking » Instance method Notation#eachKey() (alias of #each()) is removed. Now that bracket (and array) notation support is added, this name is misleading. (Now, "each" indicates each key and/or index.)

Fixed

  • In some cases, when an empty string or invalid notation is passed, it would silently fail.
  • An issue with Notation.Glob.normalize() where it would throw RangeError: Maximum call stack size exceeded when a glob list had both object and array notations for root level.
  • An issue with Notatin.Glob.union() where less restrictive globA would be removed incorrectly when globB had multiple trailing wildcards and both globs were negated.
  • An issue with Notation.Glob.normalize() where some redundant non-negated globs were not removed.
  • When Notation throws an error, error instanceof Notation.Error would return false. Fixed.

Thanks to @marcinkumorek and @BenoitRanque for their help.

1.3.6 (2018-02-24)

Fixed

  • An issue with Notation.Glob.toRegExp() method that would cause some globs to be cleared out incorrectly when .normalize()d. e.g. "!password" would match "!password_reset" and get removed the later. Fixes #7.

1.3.5 (2017-10-04)

Changed

  • Redundant, negated globs are also removed when normalized. Fixes #5.
  • Minor revisions.

Fixed

  • Shifted index issue with Notation.Glob.normalize(array).
  • countNotes() method.

1.3.0 (2017-09-30)

Added

  • Notation.Glob.normalize(array) static method.
  • Notation.Glob.toRegExp(glob) static method.
  • Notation.countNotes(notation) convenience method.

Changed

  • Completely re-wrote Notation.Glob.union() static method. sort (boolean) argument is removed (the output is now always sorted.)
  • Union output is now properly normalized, duplicates and redundant globs are removed, etc...
  • Improved glob validation.

Fixed

  • Array mutation issue. Fixes #2.
  • An issue where a glob with wildcard is not properly union'ed. Fixes [#3(onury#3).
  • An issue where negated wildcards would be filtered incorrectly in some edge cases (e.g. !*.*.*).
  • Import typo that prevents Travis builds succeed.

Removed

  • (dev) Removed dev-dependencies (Grunt and plugins) in favor of NPM scripts. Updated other dev-dependencies. Added more, comprehensive tests.

1.1.0 (2016-09-27)

Added

  • Notation#expand() method (alias Notation#aggregate()).

Changed

  • Refactored Notation#getFlat() to Notation#flatten(). Returns instance (chainable) instead of source.
  • Notation#separate() returns instance (chainable) instead of source.
  • Minor revisions.

1.0.0 (2016-04-10)

Added

  • initial release.