feat(stream): stream to zenodo-rdm deposit form by mairasalazar · Pull Request #13 · inveniosoftware/orcha

mairasalazar · 2026-03-06T15:07:50Z

❤️ Thank you for your review!

This PR allows integrating the Orcha workflow with InvenioRDM by:

Allowing a file URI to be passed instead of a URL.
- The button added in Add trigger workflow button invenio-app-rdm#3355 allows triggering a workflow and passing a file path to it. In this case, the activity must read from a file uri instead of the url.
Adding support for Server-Sent Events, to allow streaming.

OliverGeneser · 2026-03-13T11:05:06Z

-        pdf_bytes = response.content
+    """Read a file and extract its text content using the specified extractor."""
+    if settings.orcha_env in [Environment.LOCAL, Environment.DEV]:
+        with open(request.url, "rb") as f:


Maybe we should rename it to request.source as it can now be a locale file path which is not an URL.

I think we can keep it as it is since the local file path is a temporary thing for local dev

OliverGeneser · 2026-03-13T11:08:01Z

+                    return
+
            except SQLAlchemyError as e:
+                print("Error in fetching from database (stream_workflow)", e)


Maybe we should move to using logging, so we get the stack traces + observability.

ntarocco · 2026-03-19T10:36:52Z

+                if status == WorkflowStatus.ERROR:
+                    yield ServerSentEvent(
+                        # TODO: improve it with a better error message for end users
+                        data="The Temporal Workflow failed",


including translations, if this error will be directly shown in the interface. Otherwise, it is better to return error IDs or similar (numbers?), so that the error msg will be converted by the UI and translated.

mairasalazar · 2026-03-20T10:17:33Z

+                        select(Workflow).where(Workflow.public_id == workflow_id)
+                    ).one()
+                except NoResultFound:
+                    raise WorkflowEventError(error_code="WORKFLOW_NOT_FOUND")


todo: still need to handle error codes on the ui side

mairasalazar · 2026-03-20T10:50:48Z

Failing one test because some changes here depend on #18 (e.g.: Workflow class)

slint

Some minor/nits, but otherwise LGTM

slint · 2026-04-28T08:38:02Z

 from sqlalchemy.exc import SQLAlchemyError
+from sqlalchemy.orm.exc import NoResultFound
 from sqlmodel import Session, select
+from sse_starlette import EventSourceResponse, ServerSentEvent


minor: I see from the FastAPI SSE docs, you can import these from fastapi.sse. Does this maybe imply also that sse-starlette is already bundled with FastAPI (or comes via an extra?)

fastapi.sse was introduced in version 0.135.0 and Orcha is on version 0.129.0, which I guess we could bump without issues :) From what I've seen, the implementation doesn't use sse-starlette for the SSE support

@yashlamba fyi, maybe we can look into using a newer version of fastapi

mairasalazar requested a review from yashlamba March 6, 2026 15:08

mairasalazar force-pushed the allow-reading-file-from-path branch 2 times, most recently from c03fbb6 to 77626f8 Compare March 10, 2026 13:29

mairasalazar changed the title ~~refactor: read from path instead of url~~ integrate with invenio-app-rdm Mar 10, 2026

mairasalazar force-pushed the allow-reading-file-from-path branch from 77626f8 to ceddece Compare March 11, 2026 09:22

yashlamba reviewed Mar 12, 2026

View reviewed changes

Comment thread app/activities/extract_pdf_content.py Outdated

mairasalazar force-pushed the allow-reading-file-from-path branch 2 times, most recently from 497936d to 1195026 Compare March 12, 2026 16:10

mairasalazar requested a review from OliverGeneser March 13, 2026 07:56

OliverGeneser requested changes Mar 13, 2026

View reviewed changes

mairasalazar force-pushed the allow-reading-file-from-path branch 2 times, most recently from c80e161 to b923845 Compare March 16, 2026 16:51

mairasalazar changed the title ~~integrate with invenio-app-rdm~~ feat(stream): stream to zenodo-rdm deposit form Mar 18, 2026

ntarocco reviewed Mar 19, 2026

View reviewed changes

Comment thread app/activities/extract_pdf_content.py Outdated

ntarocco reviewed Mar 19, 2026

View reviewed changes

Comment thread app/routers/workflows.py Outdated

ntarocco reviewed Mar 19, 2026

View reviewed changes

Comment thread app/routers/workflows.py Outdated

ntarocco reviewed Mar 19, 2026

View reviewed changes

Comment thread app/config.py Outdated

mairasalazar force-pushed the allow-reading-file-from-path branch from bebe3ab to 80919a9 Compare March 20, 2026 10:15

mairasalazar commented Mar 20, 2026

View reviewed changes

mairasalazar force-pushed the allow-reading-file-from-path branch from 80919a9 to 20d2e76 Compare March 20, 2026 10:20

mairasalazar force-pushed the allow-reading-file-from-path branch 2 times, most recently from 72dde08 to 7577f4a Compare April 15, 2026 13:22

slint reviewed Apr 28, 2026

View reviewed changes

slint added this to Sprint Q2 2026 ☀️ Apr 28, 2026

github-project-automation Bot moved this from In review 🔍 to In progress in Sprint Q2 2026 ☀️ Apr 28, 2026

mairasalazar added 3 commits April 28, 2026 15:33

feat(stream): add SSE support

4e1af90

fix: accept both url and uri file paths

9399e23

refactor: manage errors in open file

9ad4213

mairasalazar added 4 commits April 28, 2026 15:33

feature(stream): improve streaming of results

6e9a647

refactor: change README header for clarity

21d3e79

fix: better error handling

2a74978

refactor: use match/case

e970fce

mairasalazar force-pushed the allow-reading-file-from-path branch from 7577f4a to e970fce Compare April 28, 2026 13:51

slint removed this from Sprint Q2 2026 ☀️ Jun 2, 2026

Conversation

mairasalazar commented Mar 6, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mairasalazar commented Mar 20, 2026

Uh oh!

slint left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

mairasalazar commented Mar 6, 2026 •

edited

Loading