feat(cli): support datahub ingest mcps
#7871
Conversation
github-actions
bot
added
the
ingestion
| @@ -233,6 +233,31 @@ def parse_restli_response(response): | |||
| return rows | |||
|
|
|||
|
|
|||
| @ingest.command() | |||
| @click.argument("path", type=click.Path(exists=True)) | |||
| def metadata_file(path: str) -> None: | |||
There was a problem hiding this comment.
This works even if you point the source to a directory that contains one or more metadata event files. So the docs should reflect that.
| """ | ||
| Ingest from a metadata json file. | ||
|
|
||
| This requires that you've run `datahub init` to set up your config. |
There was a problem hiding this comment.
It shouldn't require it since sink is auto-inferred.
There was a problem hiding this comment.
The sink is inferred from your env variables or your ~/.datahubenv config - so they do need to run init first
| @@ -233,6 +233,31 @@ def parse_restli_response(response): | |||
| return rows | |||
|
|
|||
|
|
|||
| @ingest.command() | |||
There was a problem hiding this comment.
I'm a bit nervous about taking the top-level default option on this one even though I like the convenience of it.
Why should
datahub ingest <something>
default to ingesting an MCP file or directory with MCP files?
versus:
datahub ingest --source file <path>
There was a problem hiding this comment.
my reading here is that we would have
datahub ingest --path <location>
Is that right? If yes, this feels fairly natural
There was a problem hiding this comment.
You use this as datahub ingest metadata-file my_file.json
$ datahub ingest metadata-file --help
Usage: datahub ingest metadata-file [OPTIONS] PATH
Ingest from a metadata json file or directory of files.
This requires that you've run `datahub init` to set up your config.
Options:
--help Show this message and exit.
hsheth2
changed the title
datahub ingest metadata-filedatahub ingest mcps
Checklist