Merge branch 'main' into NEXUS-703/remove-nestio-loop

Author: pawel-kmiecik

.speakeasy/gen.lock

@@ -1,12 +1,12 @@
 lockVersion: 2.0.0
 id: 8b5fa338-9106-4734-abf0-e30d67044a90
 management:
-  docChecksum: 83f88cfa4bd77e1d3065b11a97119361
+  docChecksum: 8befe9665bd102e8cc2241c80080859a
   docVersion: 1.0.78
-  speakeasyVersion: 1.490.0
-  generationVersion: 2.512.0
-  releaseVersion: 0.30.6
-  configChecksum: c946e26d3a3b2fe2756ff169416b448c
+  speakeasyVersion: 1.509.0
+  generationVersion: 2.539.0
+  releaseVersion: 0.31.0
+  configChecksum: faf5191e78f1597b78cd7600af4f1d96
   repoURL: https://github.com/Unstructured-IO/unstructured-python-client.git
   repoSubDirectory: .
   installationURL: https://github.com/Unstructured-IO/unstructured-python-client.git
@@ -16,22 +16,23 @@ features:
     acceptHeaders: 3.0.0
     additionalDependencies: 1.0.0
     constsAndDefaults: 1.0.5
-    core: 5.11.0
+    core: 5.12.2
     defaultEnabledRetries: 0.2.0
     enumUnions: 0.1.0
     envVarSecurityUsage: 0.3.2
     examples: 3.0.1
     globalSecurity: 3.0.3
     globalSecurityCallbacks: 1.0.0
     globalSecurityFlattening: 1.0.0
+    globalServerURLs: 3.1.0
     methodServerURLs: 3.1.1
     multipartFileContentType: 1.0.0
     nameOverrides: 3.0.1
-    nullables: 1.0.0
+    nullables: 1.0.1
     openEnums: 1.0.0
     responseFormat: 1.0.1
     retries: 3.0.2
-    sdkHooks: 1.0.0
+    sdkHooks: 1.0.1
     serverIDs: 3.0.0
     unions: 3.0.4
     uploadStreams: 1.0.0
@@ -124,6 +125,7 @@ generatedFiles:
   - docs/models/shared/googledrivesourceconnectorconfig.md
   - docs/models/shared/googledrivesourceconnectorconfiginput.md
   - docs/models/shared/jobinformation.md
+  - docs/models/shared/jobstatus.md
   - docs/models/shared/kafkaclouddestinationconnectorconfig.md
   - docs/models/shared/kafkaclouddestinationconnectorconfiginput.md
   - docs/models/shared/kafkacloudsourceconnectorconfig.md
@@ -269,6 +271,7 @@ generatedFiles:
   - src/unstructured_client/models/shared/googledrivesourceconnectorconfig.py
   - src/unstructured_client/models/shared/googledrivesourceconnectorconfiginput.py
   - src/unstructured_client/models/shared/jobinformation.py
+  - src/unstructured_client/models/shared/jobstatus.py
   - src/unstructured_client/models/shared/kafkaclouddestinationconnectorconfig.py
   - src/unstructured_client/models/shared/kafkaclouddestinationconnectorconfiginput.py
   - src/unstructured_client/models/shared/kafkacloudsourceconnectorconfig.py
@@ -412,16 +415,16 @@ examples:
         header: {}
       responses:
         "200":
-          application/json: {"created_at": "2025-01-14T00:09:32.358Z", "id": "d18f330b-506f-455c-9cb8-0cee2386ac29", "runtime": "<value>", "status": "<value>", "workflow_id": "136fd799-4c45-4417-a632-f460d8ebb2a2", "workflow_name": "<value>"}
+          application/json: {"created_at": "2025-01-14T00:09:32.358Z", "id": "d18f330b-506f-455c-9cb8-0cee2386ac29", "runtime": "<value>", "status": "IN_PROGRESS", "workflow_id": "136fd799-4c45-4417-a632-f460d8ebb2a2", "workflow_name": "<value>"}
         "422":
-          application/json: {"detail": [{"loc": ["<value>", 701982], "msg": "<value>", "type": "<value>"}, {"loc": ["<value>"], "msg": "<value>", "type": "<value>"}]}
+          application/json: {"detail": "<value>"}
   list_jobs:
     speakeasy-default-list-jobs:
       responses:
         "200":
-          application/json: [{"created_at": "2023-09-01T17:59:19.211Z", "id": "7098ff76-419f-4a5a-a313-d159507b629d", "runtime": "<value>", "status": "<value>", "workflow_id": "75cf761e-8c10-48e4-b3d7-6ae90ad4fd3e", "workflow_name": "<value>"}]
+          application/json: [{"created_at": "2023-09-01T17:59:19.211Z", "id": "7098ff76-419f-4a5a-a313-d159507b629d", "runtime": "<value>", "status": "SCHEDULED", "workflow_id": "75cf761e-8c10-48e4-b3d7-6ae90ad4fd3e", "workflow_name": "<value>"}]
         "422":
-          application/json: {"detail": []}
+          application/json: {"detail": [{"loc": [], "msg": "<value>", "type": "<value>"}]}
   create_source:
     speakeasy-default-create-source:
       requestBody:

.speakeasy/workflow.lock

@@ -1,29 +1,29 @@
-speakeasyVersion: 1.505.0
+speakeasyVersion: 1.509.0
 sources:
     my-source:
         sourceNamespace: my-source
-        sourceRevisionDigest: sha256:8cba0cf748e247efccefbb5d1839dbeb6e2e3210bb2c90f7b75b9034c22ae24a
-        sourceBlobDigest: sha256:69fef8ed681f1d68013e71f67927821d64fc1045d3956ace3775fe62ed76b8bb
+        sourceRevisionDigest: sha256:b62d80c1b902f285b4eaeaf67a4b1d2edb21053eded12c247daecbfeea1d8a30
+        sourceBlobDigest: sha256:8d5c6a31539e2e1047e529fb6561fd4172fa80506241bc25dcf51d7c76f8b6e6
         tags:
             - latest
-            - speakeasy-sdk-regen-1740605410
-            - 1.0.68
+            - speakeasy-sdk-regen-1741046993
+            - 1.0.78
 targets:
     unstructured-python:
         source: my-source
         sourceNamespace: my-source
-        sourceRevisionDigest: sha256:8cba0cf748e247efccefbb5d1839dbeb6e2e3210bb2c90f7b75b9034c22ae24a
-        sourceBlobDigest: sha256:69fef8ed681f1d68013e71f67927821d64fc1045d3956ace3775fe62ed76b8bb
+        sourceRevisionDigest: sha256:b62d80c1b902f285b4eaeaf67a4b1d2edb21053eded12c247daecbfeea1d8a30
+        sourceBlobDigest: sha256:8d5c6a31539e2e1047e529fb6561fd4172fa80506241bc25dcf51d7c76f8b6e6
         codeSamplesNamespace: my-source-code-samples
-        codeSamplesRevisionDigest: sha256:d7a20fe1c8d687f377099cdb412c6d0dd733dd982ec4662052df0894da211da2
+        codeSamplesRevisionDigest: sha256:77255b7d9241b11e147927beb6e60a27654bd6089e121356a3d5ca06e5f0f2f1
 workflow:
     workflowVersion: 1.0.0
     speakeasyVersion: latest
     sources:
         my-source:
             inputs:
                 - location: https://platform.unstructuredapp.io/openapi.json
-                - location: https://api.unstructured.io/general/openapi.json
+                - location: https://api.unstructuredapp.io/general/openapi.json
             overlays:
                 - location: ./overlay_client.yaml
             registry:

README.md

@@ -113,6 +113,7 @@ from unstructured_client import UnstructuredClient
 from unstructured_client.models import shared
 from unstructured_client.utils import BackoffStrategy, RetryConfig
 
+
 with UnstructuredClient() as uc_client:
 
     res = uc_client.destinations.create_destination(request={
@@ -141,6 +142,7 @@ from unstructured_client import UnstructuredClient
 from unstructured_client.models import shared
 from unstructured_client.utils import BackoffStrategy, RetryConfig
 
+
 with UnstructuredClient(
     retry_config=RetryConfig("backoff", BackoffStrategy(1, 50, 1.1, 100), False),
 ) as uc_client:
@@ -193,6 +195,7 @@ When custom error responses are specified for an operation, the SDK may also rai
 from unstructured_client import UnstructuredClient
 from unstructured_client.models import errors, shared
 
+
 with UnstructuredClient() as uc_client:
     res = None
     try:
@@ -325,6 +328,7 @@ Generally, the SDK will work well with most IDEs out of the box. However, when u
 from unstructured_client import UnstructuredClient
 from unstructured_client.models import shared
 
+
 with UnstructuredClient() as uc_client:
 
     res = uc_client.destinations.create_destination(request={
@@ -355,6 +359,7 @@ from unstructured_client import UnstructuredClient
 from unstructured_client.models import shared
 
 async def main():
+
     async with UnstructuredClient() as uc_client:
 
         res = await uc_client.destinations.create_destination_async(request={
@@ -451,6 +456,7 @@ Certain SDK methods accept file objects as part of a request body or multi-part
 from unstructured_client import UnstructuredClient
 from unstructured_client.models import shared
 
+
 with UnstructuredClient() as uc_client:
 
     res = uc_client.general.partition(request={
@@ -486,12 +492,14 @@ The `UnstructuredClient` class implements the context manager protocol and regis
 ```python
 from unstructured_client import UnstructuredClient
 def main():
+
     with UnstructuredClient() as uc_client:
         # Rest of application here...
 
 
 # Or when using async:
 async def amain():
+
     async with UnstructuredClient() as uc_client:
         # Rest of application here...
 ```

RELEASES.md

@@ -835,3 +835,14 @@ Based on:
 - [python v0.30.6] .
 ### Releases
 - [PyPI v0.30.6] https://pypi.org/project/unstructured-client/0.30.6 - .
+
+
+## 2025-03-04 00:09:38
+### Changes
+Based on:
+- OpenAPI Doc  
+- Speakeasy CLI 1.509.0 (2.539.0) https://github.com/speakeasy-api/speakeasy
+### Generated
+- [python v0.31.0] .
+### Releases
+- [PyPI v0.31.0] https://pypi.org/project/unstructured-client/0.31.0 - .

USAGE.md

@@ -4,6 +4,7 @@
 from unstructured_client import UnstructuredClient
 from unstructured_client.models import shared
 
+
 with UnstructuredClient() as uc_client:
 
     res = uc_client.destinations.create_destination(request={
@@ -34,6 +35,7 @@ from unstructured_client import UnstructuredClient
 from unstructured_client.models import shared
 
 async def main():
+
     async with UnstructuredClient() as uc_client:
 
         res = await uc_client.destinations.create_destination_async(request={

_sample_docs/embedded-images-tables.jpg

@@ -13,9 +13,6 @@
 from unstructured_client.utils.retries import BackoffStrategy, RetryConfig
 
 
-FREEMIUM_URL = "https://api.unstructured.io"
-
-
 @pytest.fixture(scope="module")
 def client() -> UnstructuredClient:
     _client = UnstructuredClient(api_key_auth=os.getenv("UNSTRUCTURED_API_KEY"))
@@ -47,7 +44,6 @@ def test_partition_strategies(split_pdf, strategy, client, doc_path):
     )
 
     response = client.general.partition(
-        server_url=FREEMIUM_URL,
         request=req
     )
     assert response.status_code == 200
@@ -109,7 +105,6 @@ def test_partition_handling_server_error(error, split_pdf, monkeypatch, doc_path
 
     with pytest.raises(sdk_raises):
         response = client.general.partition(
-            server_url=FREEMIUM_URL,
             request=req
         )
 
@@ -220,7 +215,6 @@ async def call_api():
         )
 
         resp = client.general.partition(
-            server_url=FREEMIUM_URL,
             request=req
         )
 
@@ -233,3 +227,124 @@ async def call_api():
     uvloop.install()
     elements = asyncio.run(call_api())
     assert len(elements) > 0
+
+
+@pytest.mark.parametrize("split_pdf", [True, False])
+@pytest.mark.parametrize("vlm_model", ["gpt-4o"])
+@pytest.mark.parametrize("vlm_model_provider", ["openai"])
+@pytest.mark.parametrize(
+    "filename", 
+    [
+        "layout-parser-paper-fast.pdf", 
+        "fake-power-point.ppt",
+        "embedded-images-tables.jpg",
+    ]
+)
+def test_partition_strategy_vlm_openai(split_pdf, vlm_model, vlm_model_provider, client, doc_path, filename):
+    with open(doc_path / filename, "rb") as f:
+        files = shared.Files(
+            content=f.read(),
+            file_name=filename,
+        )
+
+    req = operations.PartitionRequest(
+        partition_parameters=shared.PartitionParameters(
+            files=files,
+            strategy="vlm",
+            vlm_model=vlm_model,
+            vlm_model_provider=vlm_model_provider,
+            languages=["eng"],
+            split_pdf_page=split_pdf,
+        )
+    )
+
+    response = client.general.partition(
+        request=req
+    )
+    assert response.status_code == 200
+    assert len(response.elements) > 0
+    assert response.elements[0]["metadata"]["partitioner_type"] == "vlm_partition"
+
+
+@pytest.mark.parametrize("split_pdf", [True, False])
+@pytest.mark.parametrize("vlm_model", 
+    [
+        "us.amazon.nova-pro-v1:0", 
+        "us.amazon.nova-lite-v1:0", 
+        "us.anthropic.claude-3-5-sonnet-20241022-v2:0",
+        "us.anthropic.claude-3-opus-20240229-v1:0",
+        "us.anthropic.claude-3-haiku-20240307-v1:0",
+        "us.anthropic.claude-3-sonnet-20240229-v1:0",
+        "us.meta.llama3-2-90b-instruct-v1:0",
+        "us.meta.llama3-2-11b-instruct-v1:0",
+    ]
+)
+@pytest.mark.parametrize("vlm_model_provider", ["bedrock"])
+@pytest.mark.parametrize(
+    "filename", 
+    [
+        "layout-parser-paper-fast.pdf", 
+        "fake-power-point.ppt",
+        "embedded-images-tables.jpg",
+    ]
+)
+def test_partition_strategy_vlm_bedrock(split_pdf, vlm_model, vlm_model_provider, client, doc_path, filename):
+    with open(doc_path / filename, "rb") as f:
+        files = shared.Files(
+            content=f.read(),
+            file_name=filename,
+        )
+
+    req = operations.PartitionRequest(
+        partition_parameters=shared.PartitionParameters(
+            files=files,
+            strategy="vlm",
+            vlm_model=vlm_model,
+            vlm_model_provider=vlm_model_provider,
+            languages=["eng"],
+            split_pdf_page=split_pdf,
+        )
+    )
+
+    response = client.general.partition(
+        request=req
+    )
+    assert response.status_code == 200
+    assert len(response.elements) > 0
+    assert response.elements[0]["metadata"]["partitioner_type"] == "vlm_partition"
+
+@pytest.mark.parametrize("split_pdf", [True, False])
+@pytest.mark.parametrize("vlm_model", ["claude-3-5-sonnet-20241022",])
+@pytest.mark.parametrize("vlm_model_provider", ["anthropic"])
+@pytest.mark.parametrize(
+    "filename", 
+    [
+        "layout-parser-paper-fast.pdf", 
+        "fake-power-point.ppt",
+        "embedded-images-tables.jpg",
+    ]
+)
+def test_partition_strategy_vlm_anthropic(split_pdf, vlm_model, vlm_model_provider, client, doc_path, filename):
+    with open(doc_path / filename, "rb") as f:
+        files = shared.Files(
+            content=f.read(),
+            file_name=filename,
+        )
+
+    req = operations.PartitionRequest(
+        partition_parameters=shared.PartitionParameters(
+            files=files,
+            strategy="vlm",
+            vlm_model=vlm_model,
+            vlm_model_provider=vlm_model_provider,
+            languages=["eng"],
+            split_pdf_page=split_pdf,
+        )
+    )
+
+    response = client.general.partition(
+        request=req
+    )
+    assert response.status_code == 200
+    assert len(response.elements) > 0
+    assert response.elements[0]["metadata"]["partitioner_type"] == "vlm_partition"

_sample_docs/fake-power-point.ppt

@@ -211,12 +211,6 @@ def test_unit_clean_server_url_fixes_malformed_localhost_url(server_url: str):
     assert client.general.sdk_configuration.server_url == "http://localhost:8000"
 
 
-
-def test_unit_clean_server_url_returns_None_given_no_server_url():
-    client = UnstructuredClient(api_key_auth=FAKE_KEY)
-    assert client.general.sdk_configuration.server_url == None
-
-
 @pytest.mark.parametrize(
     "server_url",
     [