Per-flow librarian clients and per-workspace response queues (#865)

Replace singleton LibrarianClient with per-flow instances via the new
LibrarianSpec, giving each flow its own librarian tied to the
workspace-scoped request/response queues from the blueprint.

Move all workspace-scoped services (config, flow, librarian, knowledge)
from a single base-queue response producer to per-workspace response
producers created alongside the existing per-workspace request
consumers.  Update the gateway dispatcher and bootstrapper flow client
to subscribe to the matching workspace-scoped response queues.

Fix WorkspaceInit to register workspaces through the IAM
create-workspace API so they appear in __workspaces__ and are visible
to the gateway.  Simplify the bootstrapper gate to only check
config-svc reachability.

Updated tests accordingly.
This commit is contained in:
cybermaggedon 2026-05-06 12:01:01 +01:00 committed by GitHub
parent 01bf1d89d5
commit 03cc5ac80f
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
30 changed files with 405 additions and 735 deletions

View file

@ -177,8 +177,7 @@ class TestRecursiveChunkerSimple(IsolatedAsyncioTestCase):
processor = Processor(**config)
# Mock save_child_document to avoid waiting for librarian response
processor.librarian.save_child_document = AsyncMock(return_value="mock-doc-id")
# Mock save_child_document on flow to avoid waiting for librarian response
# Mock message with TextDocument
mock_message = MagicMock()
@ -204,6 +203,7 @@ class TestRecursiveChunkerSimple(IsolatedAsyncioTestCase):
"output": mock_producer,
"triples": mock_triples_producer,
}.get(key)
mock_flow.librarian.save_child_document = AsyncMock(return_value="mock-doc-id")
# Act
await processor.on_message(mock_message, mock_consumer, mock_flow)

View file

@ -177,8 +177,7 @@ class TestTokenChunkerSimple(IsolatedAsyncioTestCase):
processor = Processor(**config)
# Mock save_child_document to avoid librarian producer interactions
processor.librarian.save_child_document = AsyncMock(return_value="chunk-id")
# Mock save_child_document on flow to avoid librarian producer interactions
# Mock message with TextDocument
mock_message = MagicMock()
@ -204,6 +203,7 @@ class TestTokenChunkerSimple(IsolatedAsyncioTestCase):
"output": mock_producer,
"triples": mock_triples_producer,
}.get(key)
mock_flow.librarian.save_child_document = AsyncMock(return_value="chunk-id")
# Act
await processor.on_message(mock_message, mock_consumer, mock_flow)

View file

@ -156,6 +156,7 @@ class TestMistralOcrProcessor(IsolatedAsyncioTestCase):
"output": mock_output_flow,
"triples": mock_triples_flow,
}.get(name))
mock_flow.librarian.save_child_document = AsyncMock(return_value="mock-doc-id")
config = {
'id': 'test-mistral-ocr',
@ -171,9 +172,6 @@ class TestMistralOcrProcessor(IsolatedAsyncioTestCase):
("# Page 2\nMore content", 2),
]
# Mock save_child_document
processor.librarian.save_child_document = AsyncMock(return_value="mock-doc-id")
with patch.object(processor, 'ocr', return_value=ocr_result):
await processor.on_message(mock_msg, None, mock_flow)
@ -227,8 +225,7 @@ class TestMistralOcrProcessor(IsolatedAsyncioTestCase):
Processor.add_args(mock_parser)
mock_parent_add_args.assert_called_once_with(mock_parser)
assert mock_parser.add_argument.call_count == 3
# Check the API key arg is among them
assert mock_parser.add_argument.call_count == 1
call_args_list = [c[0] for c in mock_parser.add_argument.call_args_list]
assert ('-k', '--api-key') in call_args_list

View file

@ -72,6 +72,7 @@ class TestPdfDecoderProcessor(IsolatedAsyncioTestCase):
"output": mock_output_flow,
"triples": mock_triples_flow,
}.get(name))
mock_flow.librarian.save_child_document = AsyncMock(return_value="mock-doc-id")
config = {
'id': 'test-pdf-decoder',
@ -80,9 +81,6 @@ class TestPdfDecoderProcessor(IsolatedAsyncioTestCase):
processor = Processor(**config)
# Mock save_child_document to avoid waiting for librarian response
processor.librarian.save_child_document = AsyncMock(return_value="mock-doc-id")
await processor.on_message(mock_msg, None, mock_flow)
# Verify output was sent for each page
@ -148,6 +146,7 @@ class TestPdfDecoderProcessor(IsolatedAsyncioTestCase):
"output": mock_output_flow,
"triples": mock_triples_flow,
}.get(name))
mock_flow.librarian.save_child_document = AsyncMock(return_value="mock-doc-id")
config = {
'id': 'test-pdf-decoder',
@ -156,9 +155,6 @@ class TestPdfDecoderProcessor(IsolatedAsyncioTestCase):
processor = Processor(**config)
# Mock save_child_document to avoid waiting for librarian response
processor.librarian.save_child_document = AsyncMock(return_value="mock-doc-id")
await processor.on_message(mock_msg, None, mock_flow)
mock_output_flow.send.assert_called_once()

View file

@ -254,8 +254,7 @@ class TestUniversalProcessor(IsolatedAsyncioTestCase):
"triples": mock_triples_flow,
}.get(name))
# Mock save_child_document and magic
processor.librarian.save_child_document = AsyncMock(return_value="mock-id")
mock_flow.librarian.save_child_document = AsyncMock(return_value="mock-id")
with patch('trustgraph.decoding.universal.processor.magic') as mock_magic:
mock_magic.from_buffer.return_value = "text/markdown"
@ -310,7 +309,7 @@ class TestUniversalProcessor(IsolatedAsyncioTestCase):
"triples": mock_triples_flow,
}.get(name))
processor.librarian.save_child_document = AsyncMock(return_value="mock-id")
mock_flow.librarian.save_child_document = AsyncMock(return_value="mock-id")
with patch('trustgraph.decoding.universal.processor.magic') as mock_magic:
mock_magic.from_buffer.return_value = "application/pdf"
@ -361,7 +360,7 @@ class TestUniversalProcessor(IsolatedAsyncioTestCase):
"triples": mock_triples_flow,
}.get(name))
processor.librarian.save_child_document = AsyncMock(return_value="mock-id")
mock_flow.librarian.save_child_document = AsyncMock(return_value="mock-id")
with patch('trustgraph.decoding.universal.processor.magic') as mock_magic:
mock_magic.from_buffer.return_value = "application/pdf"
@ -374,7 +373,7 @@ class TestUniversalProcessor(IsolatedAsyncioTestCase):
assert mock_triples_flow.send.call_count == 2
# save_child_document called twice (page + image)
assert processor.librarian.save_child_document.call_count == 2
assert mock_flow.librarian.save_child_document.call_count == 2
@patch('trustgraph.base.flow_processor.FlowProcessor.add_args')
def test_add_args(self, mock_parent_add_args):