Hexa 1441 pipelines list only returns first page #350
Conversation
DimitriKwihangana
requested review from
bramj and
yolanfery
and removed request for
bramj
DimitriKwihangana
requested review from
bramj and
yolanfery
and removed request for
bramj
yolanfery
left a comment
yolanfery
left a comment
There was a problem hiding this comment.
Looking good, minor comments :-)
| current_version = "Jupyter notebook" | ||
| click.echo(f"* {pipeline.code} - {pipeline.name} ({current_version})") | ||
| page = 1 | ||
| click.echo("Pipelines:") |
There was a problem hiding this comment.
[Minor] I am a bit bothered because this is printed before the API call (which takes times). Maybe we should start printing after the API call (even if that means adding a if check)
| for pipeline in response.items: | ||
| if pipeline.type == PipelineType.zipFile: | ||
| version = f"v{pipeline.current_version.version_number}" if pipeline.current_version else "N/A" | ||
| else: # notebook type |
There was a problem hiding this comment.
[Minor] instead of this comment, I would elif pipeline.type == PipelineType.notebook
This avoid the need for comment and helps if ever extend the PipelineType type one day with one more type. Drawback is that you might need to add version = "" to use it ou t of the if-else scope
|
|
||
| for pipeline in response.items: | ||
| if pipeline.type == PipelineType.zipFile: | ||
| version = f"v{pipeline.current_version.version_number}" if pipeline.current_version else "N/A" |
There was a problem hiding this comment.
Can you use a diffrent naming than version ? It conflicts with from importlib.metadata import version
| click.echo(f"* {pipeline.code} - {pipeline.name} ({version})") | ||
|
|
||
| if page >= response.total_pages or not click.confirm( | ||
| f"\nLoad more? (page {page}/{response.total_pages})", default=True |
There was a problem hiding this comment.
[Very-Minor] Show more seems more user friendly than Load more
| click.echo(f"* {pipeline.code} - {pipeline.name} ({current_version})") | ||
| page = 1 | ||
| click.echo("Pipelines:") | ||
|
|
There was a problem hiding this comment.
Would be cool also somewhere to show the total number of pipelines
Like Pipelines (56): or something else
Description
The
openhexa pipelines listcommand was only returning the first page of pipelines (max 10 items), even when workspaces contained 50+ pipelines. This PR adds pagination support to fetch and display all pipelines across multiple pages.Core idea: Instead of fetching only the first page, the CLI now iteratively fetches pages and prompts users to load more, ensuring all pipelines are accessible.
Another fix: Corrected notebook pipeline type detection by using the
PipelineTypeenum instead of string comparison.Changes
Pagination Implementation
page = 1and increment after each fetch.pageandper_page=10parameters toclient.pipelines().while Trueloop to fetch multiple pages..itemsto full response object to accesstotal_pagesmetadata.click.confirm()prompt asking users to load next page.page == 1).Type Detection Fix
from openhexa.graphql.graphql_client.enums import PipelineTypepipeline.type == "zipFile"topipeline.type == PipelineType.zipFile.How/what to test
openhexa pipelines list. Confirm all pipelines display immediately with no "Load more?" prompt.openhexa pipelines list. Confirm the first 10 pipelines display with aLoad more? (page 1/N) [Y/n]:prompt. Press Enter to ensure subsequent pages load correctly.openhexa pipelines listand verify it shows as(Jupyter notebook)while standard pipelines show their version (e.g.,v1).Screenshots / screencast