Skip to content

Latest commit

 

History

History
69 lines (42 loc) · 3.56 KB

CHANGELOG.md

File metadata and controls

69 lines (42 loc) · 3.56 KB
Raw

Changelog

6.2

  • support proper stream parsing using (kotlinx-io's) Source
  • fix FeatureDictionary.getById didn't always exclude country-specific results when the country parameter was null even though it was documented as such

6.1

  • Fix crash when single presets were translated only partially for a dialect (#26), by @logan12358
  • Aliases and terms of presets in the parent language are now merged into any dialects (#26)

6.0

The library is now a Kotlin Multiplatform library. This is a breaking API change. There is also no separate artifact for Android anymore.

5.2

  • Add support placeholders for preset names (breaking change in v5.0.0 of iD presets schema)
  • When searching by term, include results that match with the tag value at the bottom of the result list. (iD has the same behavior)

5.1

  • add property boolean isSuggestion to Feature to be able to tell if a feature is a brand feature or not
  • add method to get a feature by its id to FeatureDictionary

5.0

Added support for aliases of presets. They are treated similarily as alternative names.

4.1

Added support for scripts.

E.g. there can be localization of presets in "bg" (Bulgarian) and also "bg-Cyrl" (Bulgarian in Cryllic).

4.0

Brand features are becoming too big.

So now, brand features are loaded lazily per-country. I.e. additionally to a presets.json, there can be presets-US.json (using ISO 3166-1 alpha2 codes) and even presets-US-NY.json (ISO 3166-2) in the same directory, which will be loaded on demand. This functionality is used in StreetComplete.

3.0

Support for NSI brand names (v6.x). Pass the path to the NSI presets as second parameter to the create function.

Note that the NSI presets are actually not on the root level of https://github.com/osmlab/name-suggestion-index/blob/main/dist/presets/nsi-id-presets.json but in the "presets" object. The library expects a presets.json to have all the presets at the root level.

2.0

  • dictionary now uses indices for both the lookup by tags and the lookup by term, which speeds up individual lookups somewhat.
  • added support for fallback locales. You can now specify several locales in which to search for a term
  • added support for presets that are available everywhere except in certain countries
  • added support for filtering results by term whether they are suggestions (brand names) or not
  • uses new source for presets and translations (iD presets have been outsourced into an own repository) that has a slightly different format
  • now, Feature objects are returned (not Match) objects in lookups which contain most informations of a preset in iD format
  • added possibility to add (and thus merge) presets from several directories

1.2

  • enable to also either exclusively search for presets that are suggestions (=brand names) and to exclusively search for presets that are not suggestions. (QueryByTagBuilder::isSuggestion(Boolean))
  • when searching by tags and the search is not limited by locale, sort those matches further up in the list of search results that are also not limited by locale
  • when searching by tags, sort those matches further up in the list of search results of whose addTags match with more of the given tags
  • internally use LinkedHashMap. This should have no effect on pure Java. On Android API 20+, this should sort presets that were defined further up in the presets.json also further up in the list of results if all other sort criteria are the same for any two matches

1.1

  • correct the gradle script in the README