• Files Changed
• Browse Source

- switch to modern python for sle15

- Restrict numpy to < 2.

• Files Changed
• Browse Source

- Restrict numpy to < 2.

• Files Changed
• Browse Source

- update to 2.0.4:
  * readd entrypoints for old style sources
  * Fix "plugins" in old catalogs

  * allow catalog TTL as None
  * add intake-nested-yaml-catalog plugin
  * Don't warn twice
  * Pass filesystem stuff through
  * add to API docs

• Files Changed
• Browse Source

- Don't try to build with Python 3.9
- Update to 2.0.3
  * Complete rewrite of the package
  * We are making Take2 as a full release. It is still “beta” in
    the sense that we will be adding many data types, readers and
    transformers, and are prepared to revisit the API in general.
    The reason not to use a pre-release or RC, is that users never
    see these.
  ## Relationship to V1
  * We aim to be largely backward compatible with pre-V2 Intake
    sources and catalogs. Many data sources have been rewritten to
    use the new framework, and many rarely-used features have been
    removed. In particular, the following features are no longer
    supported for V1 sources:
    - the intake server (use tiled instead)
    - caching (use fsspec caching instead or custom caching
      pipelines)
    - “persist” and “export” (use the new converters and output
      classes)
    - automatic hvplot (this is now an “output” converter for
      pandas and xarray types)
    - some niche source features such as CSV file pattern matching
  * In addition, not all existing intake_* packages have
    corresponding readers in Take2, but we are making progress.
- Drop fix-tests.patch

• Files Changed
• Browse Source

- Add fix-tests.patch to make the current version (0.7.0) compatible
  with latest version of panel, pandas and fsspec.

• Files Changed
• Browse Source

• Files Changed
• Browse Source

- Update to 0.6.8:
  * user parameter parsed as string before conversion to given type
  * numpy source becomes first to have read() path avoid dask
  * when registering drivers dynamically, corresponding open_* functions will
    be created automatically (plus refactor/cleanup of the discovery code)
  * catalog .gui attribute will make top-level GUI instance instead of cut
    down one-catalog version
  * server fix for upstream dask change giving newlined in report
  * editable plots, based on hvPlot's "explorer"
  * remove "text" input to YAMLFileCatalog
  * GUI bug fixes
  * allow catalog TTL as None 
- Add patch use-sys-executable.patch:
  * Use sys.executable, rather than hardcoding python.

• Files Changed
• Browse Source

• Files Changed
• Browse Source

- Remove skip_python310, it is not needed anymore.

• Files Changed
• Browse Source

• Files Changed
• Browse Source

• Files Changed
• Browse Source

• Files Changed
• Browse Source

• Files Changed
• Browse Source

• Files Changed
• Browse Source

• Files Changed
• Browse Source

• Files Changed
• Browse Source

Data loading and cataloging system

• Browse Source