Milli v0.26.2

⚠️ Release done on release-v0.26.2 branch, not on main.

Fixes

• min-word-len-for-typo not reset properly (#499) @MarinPostma

Milli v0.26.1

Changes

• Enforce labelling for the PRs (#490) @curquiza
• Add the new Specify breaking check to bors.toml (#492) @curquiza
• Fix distinct count bug (#489) @MarinPostma
• Improve the performances of the flattening subcrate (#496) @Kerollmops
• Use smartstring to store the external id in our hashmap (#493) @irevoire
• Only flatten the required objects (#494) @irevoire

Thanks again to @Kerollmops, @MarinPostma, @curquiza and @irevoire! 🎉

Milli v0.26.0

Breaking changes ⚠️

• Nested fields (#458) @irevoire

Thanks again to @irevoire! 🎉

Milli v0.25.0

Changes

• Increase benchmarks (push) CI timeout (#480) @Kerollmops
• add authorize typo setting (#469) @MarinPostma
• fix bug on 2 typos derivation (#485) @MarinPostma
• set minimum word len for typos (#473) @MarinPostma
• Disable typos on exact word (#474) @MarinPostma

Breaking changes ⚠️

• Disable typo on attribute (#478) @MarinPostma

Thanks again to @Kerollmops, @MarinPostma, and @curquiza! 🎉

Milli v0.24.1

Changes

• Remove useless variables in proximity (#472) @ManyTheFish

Misc

• Rollback meilisearch-tokenizer version (#476) and then re-bump it (#475) @irevoire

Thanks again to @ManyTheFish, and @irevoire! 🎉

Milli v0.24.0

Improvements

• Speed up the word prefix databases computation time (#436) @Kerollmops
• Reintroduce appending sorted entries when possible (#454) @Kerollmops
• Remove useless grenad merging (#456) @Kerollmops
• Avoid iterating on big databases when useless (#457) @Kerollmops
• Add a new error message when the valid_fields is empty (#461) @brunoocasali
• cli improvements (#462) @MarinPostma
• Allow setting the primary-key in the cli (#463) @irevoire
• optimize prefix database (#467) @MarinPostma
• Optimize typo criterion (#439) @MarinPostma
• Add a new error message when the filterableAttributes are empty (#468) @brunoocasali

Misc

• Benchmark multi batch indexing (#453) @Kerollmops
• Raise the GitHub CI timeout limit to 72h (#455) @Kerollmops
• Update heed link in cargo toml (#459) @curquiza
• Set the cargo crate resolver to v2 (#470) @MarinPostma
• Update dependencies (#465) @Kerollmops

Breaking changes ⚠️

• exporting heed to avoid having different versions of Heed in Meilisearch (#464) @psvnlsaikumar

Thanks again to @Kerollmops, @MarinPostma, @brunoocasali, @curquiza, @irevoire and @psvnlsaikumar! 🎉

Milli v0.23.0

Changes

• Get rid of chrono in favor of time (#450) @irevoire

Thanks again to @irevoire! 🎉

Milli v0.22.1

Changes

• fix(fuzzer): fix the fuzzer after #430 (#440) @irevoire
• CLI improvements (#438) @MarinPostma
• Fix and improve word prefix pair proximity (#431) @Kerollmops
• Fix the parsing of ndjson requests to index more than the first line (#444) @Kerollmops
• Change chunk size to 4MiB to fit more the end user usage (#417) @ManyTheFish
• allow null values in csv (#445) @MarinPostma
• fix phrase search (#442) @MarinPostma

Thanks again to @Kerollmops, @ManyTheFish, @MarinPostma, @irevoire ! 🎉

v0.22.0

Breaking changes

• Store the geopoint in three dimensions (#424) @irevoire
• Prefer returning None instead of using an FilterCondition::Empty state (#422) @Kerollmops

Enhancement

• Introduce the depth method on FilterCondition (#421) @Kerollmops
• Handle escaped characters in filters (#427) @irevoire
• Document batch support (#430) @MarinPostma
• Add fuzzing (#432) @irevoire

Fixes

• Fix search highlight for non-unicode chars (#426) @Samyak2
• Fix two filtering bugs (#433) @irevoire
• Ensure we get no documents and no error when filtering on an empty db (#435) @irevoire

Misc

• Remove an unused file (#423) @irevoire
• Push the result of the benchmarks to influxdb (#425) @irevoire
• Reintroduce the gitignore for the fuzzer (#428) @irevoire
• Benchmark CIs: not use a default label to call the GH runner (#429) @curquiza

Milli v0.21.1

Changes

• Fix filter issue for meilisearch/meilisearch#2082 (#433) @irevoire