v0.2.1

Highlights

• Python bindings:

  • You can now query intermediate results, simply referencing a variable:

    intermediate = con.sql("select * from tbl where a > 2;")
    con.sql("select * from intermediate where b > 4;")

  • show() was added to print results without other libraries

• SQL commands:

  • COPY TO:
    • Add support for remote URLs, including GCS and S3. For more information, see #1280
    • Tables can now be copied with COPY TO (instead of only queries). For more information see #1280

• SQL functions:

  • All functions received support for passing in identifiers and named arguments. For more information, see #1314
  • csv_scan:
    • Add support for GCS and S3 with credentials
  • ndjson_scan:
    • Add support for GCS and S3 with credentials
  • parquet_scan:
    • Add support for GCS and S3 with credentials
  • generate_series:
    • Added support for floats

• Recipes:

  • We've added a new examples directory. Our first example demonstrates how to connect to glaredb with R and query a Snowflake database. Check it out here: https://github.com/GlareDB/glaredb/blob/main/examples/R/out-of-process.R
  • We added an example of using our Python bindings with OpenAI! See it here: https://github.com/GlareDB/glaredb/blob/main/py-glaredb/examples/openai_ex.py

What's Changed

• chore: add semicolon to sql snippet by @greyscaled in #1299
• chore: Add example R script for connecting to glaredb by @scsmithr in #1296
• feat: composable plans by @universalmind303 in #1300
• feat: Support URLs for COPY .. TO ... by @vrongmeal in #1280
• feat: Add support for float in generate_series by @adhish20 in #1298
• chore: move builtins.rs to sqlbuiltins by @adhish20 in #1302
• chore: Add Dependabot configuration for Cargo by @0xf333 in #1301
• fix csv_scan for object_stores by @universalmind303 in #1303
• use pr-target instead of pr by @universalmind303 in #1305
• chore: Add example for running openai on data from postgres by @scsmithr in #1309
• feat: Support object store in *_scan functions. by @vrongmeal in #1306
• fix: stop parsing datasource and tunnel limits from cloud response by @f0ssel in #1313
• chore: move "client" stuff to metastoreproto by @adhish20 in #1311
• fix: Fix not being able to read huggingface parquets by @scsmithr in #1319
• feat: Support named arguments and idents in table functions. by @vrongmeal in #1314
• fix: Explicitly use Ident in table function parameters. by @vrongmeal in #1322
• fix: Add xtask alias back (for zip) by @vrongmeal in #1325
• chore: Update deltalake to da1012. by @vrongmeal in #1326
• feat: adding show() method to python execution results by @adhish20 in #1315
• chore: Use SchemaReference from datafusion instead of our own. by @vrongmeal in #1327
• chore: simplify feat template by @greyscaled in #1328
• build: Bump version to 0.2.1 by @greyscaled in #1332

New Contributors

• @0xf333 made their first contribution in #1301

Full Changelog: v0.2.0...v0.2.1