1.7.0

What's Changed

• Add check of hashed data when writing new data by @peytondmurray in #296
• Add diff API by @peytondmurray in #293
• Optimize _recreate_raw_dataset to use less memory; add versions method by @peytondmurray in #300
• Prevent tests from generating unwanted test artifacts by @peytondmurray in #304
• Check dataset names against blocklist before committing by @peytondmurray in #305
• Update the build system to use meson-python; add 3.12 compatibility by @peytondmurray in #302
• Add pre-commit; apply formatting across the repository by @peytondmurray in #307
• Resolve pygrep-hooks issues in documentation by @peytondmurray in #311
• Remove unused build/release tooling by @peytondmurray in #312
• Make create_virtual_dataset handle zero-length raw_data correctly by @peytondmurray in #315
• Add tests to the meson.build file so that editable installs can test by @peytondmurray in #317
• Remove APIs deprecated in numpy 2.0; add testing in CI by @peytondmurray in #306
• Make InMemoryGroup provide more informative reprs by @peytondmurray in #319
• Faster and correct chunk reuse logging by @ArvidJB in #324
• Optimize InMemoryDataset.resize to avoid quadratic performance by @peytondmurray in #326
• Improve speed of reused chunk verification by @peytondmurray in #330
• resize: read the entire Dataset by @ArvidJB in #332
• Fix chunk reuse check for multi-dimensional Datasets by @ArvidJB in #334
• Add support for docs version switcher by @peytondmurray in #335
• Add option to disable verify_chunk_reuse by @ArvidJB in #337
• Add 1.7.0 release notes by @peytondmurray in #340

Full Changelog: 1.6.0...1.7.0