v0.2.0

Highlights

• python bindings:

  • GlareDB can now be used in Python with Pandas, Polars and external data sources. Install the module from PyPI with

  pip install glaredb

  The instance of glaredb is currently limited to in-memory. We're working on connecting to remote instances of glaredb for a future release. For more information, we have examples at https://github.com/GlareDB/glaredb/tree/main/py-glaredb/examples.

• COPY TO:

  • With COPY TO, a local copy of a query can be persisted. We're adding support for remote persistence in the future. For example, you can save a local copy in test.csv of the query select 1 with: copy (select 1) to local (location 'test.csv');
  • Currently, json, parquet and csv formats are supported. The default format is CSV, but you can specify a different format as follows:

  copy (select 1) to local format parquet (location 'test.parquet');

• ndjson_scan:

  • You can now read newline delimited JSON files, ex: select * from ndjson_scan('./your/file.json');

• misc:

  • Added timestamp support for native tables

What's Changed

• chore: Prevent python linking for macos by @scsmithr in #1198
• remove python dep from docker by @universalmind303 in #1199
• feat: Allow reading from the python environment by @scsmithr in #1195
• feat: Add pandas support by @scsmithr in #1201
• feat: Use multi-threaded runtime in python by @scsmithr in #1203
• chore: Break up slt tests into batches by @scsmithr in #1205
• chore: Add slt tests for parquet_scan and csv_scan by @scsmithr in #1206
• chore: add link to protoc installation by @greyscaled in #1209
• feat: Make data dir optional in python bindings by @scsmithr in #1211
• chore: Use raw github link in csv_scan test by @scsmithr in #1212
• feat: add support for scanning ndjson by @universalmind303 in #1213
• ci: setup python workflow by @universalmind303 in #1210
• remove py-glaredb from default members by @universalmind303 in #1224
• build(deps): Bump deps by @scsmithr in #1241
• check permissions on tmp dir by @universalmind303 in #1239
• chore: Bump rust version in docker image by @scsmithr in #1250
• chore: Remove starting metastore from pgproto test script by @scsmithr in #1252
• fix: Downgrade datadriven crate by @scsmithr in #1253
• fix: Fix explain queries by @scsmithr in #1251
• feat: COPY TO support by @vrongmeal in #1159
• refactor: xtask --> justfile by @universalmind303 in #1242
• ci: Keep py-glaredb and glaredb versions in sync by @universalmind303 in #1255
• fix: Allow create tables with timestamps by @vrongmeal in #1257
• chore: Add pytest by @scsmithr in #1258
• feat: support compute_engine authentication parameter by @f0ssel in #1256
• chore: Add more info for pypi by @scsmithr in #1263
• build(deps): Bump deps by @scsmithr in #1277
• refactor: use native execs instead of custom execs by @universalmind303 in #1262
• chore: Add test cases for parsing compute engine from conn string by @scsmithr in #1285
• fix: don't fetch filetype eagerly by @adhish20 in #1286
• feat: Add s3 to object store registry create by @scsmithr in #1291
• feat(py): add lazy execution to python by @universalmind303 in #1292
• chore: Add tests asserting logical plan reuse in python by @scsmithr in #1293
• build: Bump version to 0.2.0 by @scsmithr in #1295

New Contributors

• @adhish20 made their first contribution in #1286

Full Changelog: v0.0.16...v0.2.0