Add new parameter, fix parse_mode (#660)

* update with new parameters

* lint

.github/workflows/publish_release.yml

@@ -14,7 +14,7 @@ env:
 jobs:
   build-n-publish:
     name: Build and publish to PyPI
-    if: github.repository == 'run-llama/llama_parse'
+    if: github.repository == 'run-llama/llama_cloud_services'
     runs-on: ubuntu-latest
 
     steps:

README.md

@@ -36,7 +36,7 @@ See the quickstart guides for each service for more information:
 
 - [LlamaParse](./parse.md)
 - [LlamaReport (beta/invite-only)](./report.md)
-- [LlamaExtract (coming soon!)]()
+- [LlamaExtract (beta/invite-only)](./extract.md)
 
 ## Documentation
 

llama_cloud_services/extract/extract.py

@@ -9,16 +9,20 @@ import httpx
 from pydantic import BaseModel
 from llama_cloud import (
     ExtractAgent as CloudExtractAgent,
+    ExtractAgentCreate,
     ExtractConfig,
     ExtractJob,
     ExtractJobCreate,
     ExtractRun,
+    ExtractSchemaValidateRequest,
+    ExtractAgentUpdate,
     File,
     ExtractMode,
     StatusEnum,
     Project,
     ExtractTarget,
     LlamaExtractSettings,
+    PaginatedExtractRunsResponse,
 )
 from llama_cloud.client import AsyncLlamaCloud
 from llama_cloud_services.extract.utils import JSONObjectType, augment_async_errors
@@ -53,6 +57,8 @@ class ExtractionAgent:
         num_workers: int = 4,
         show_progress: bool = True,
         verbose: bool = False,
+        verify: Optional[bool] = True,
+        httpx_timeout: Optional[float] = 60,
     ):
         self._client = client
         self._agent = agent
@@ -62,6 +68,8 @@ class ExtractionAgent:
         self.max_timeout = max_timeout
         self.num_workers = num_workers
         self.show_progress = show_progress
+        self.verify = verify
+        self.httpx_timeout = httpx_timeout
         self._verbose = verbose
         self._data_schema: Union[JSONObjectType, None] = None
         self._config: Union[ExtractConfig, None] = None
@@ -74,14 +82,20 @@ class ExtractionAgent:
 
         def run_coro() -> T:
             async def wrapped_coro() -> T:
+                # Get the original client to preserve its configuration
+                original_client = self._client._client_wrapper.httpx_client
+
+                # Create a new client with the same configuration as the original
                 async with httpx.AsyncClient(
-                    timeout=self._client._client_wrapper.httpx_client.timeout,
+                    verify=self.verify,
+                    timeout=self.httpx_timeout,
                 ) as client:
-                    original_client = self._client._client_wrapper.httpx_client
+                    # Temporarily replace the client
                     self._client._client_wrapper.httpx_client = client
                     try:
                         return await coro
                     finally:
+                        # Restore the original client
                         self._client._client_wrapper.httpx_client = original_client
 
             return asyncio.run(wrapped_coro())
@@ -114,7 +128,7 @@ class ExtractionAgent:
             )
         validated_schema = self._run_in_thread(
             self._client.llama_extract.validate_extraction_schema(
-                data_schema=processed_schema
+                request=ExtractSchemaValidateRequest(data_schema=processed_schema)
             )
         )
         self._data_schema = validated_schema.data_schema
@@ -187,8 +201,10 @@ class ExtractionAgent:
         self._agent = self._run_in_thread(
             self._client.llama_extract.update_extraction_agent(
                 extraction_agent_id=self.id,
-                data_schema=self.data_schema,
-                config=self.config,
+                request=ExtractAgentUpdate(
+                    data_schema=self.data_schema,
+                    config=self.config,
+                ),
             )
         )
 
@@ -375,15 +391,29 @@ class ExtractionAgent:
             )
         )
 
-    def list_extraction_runs(self) -> List[ExtractRun]:
+    def delete_extraction_run(self, run_id: str) -> None:
+        """Delete an extraction run by ID.
+
+        Args:
+            run_id (str): The ID of the extraction run to delete
+        """
+        self._run_in_thread(
+            self._client.llama_extract.delete_extraction_run(run_id=run_id)
+        )
+
+    def list_extraction_runs(
+        self, page: int = 0, limit: int = 100
+    ) -> PaginatedExtractRunsResponse:
         """List extraction runs for the extraction agent.
 
         Returns:
-            List[ExtractRun]: List of extraction runs
+            PaginatedExtractRunsResponse: Paginated list of extraction runs
         """
         return self._run_in_thread(
             self._client.llama_extract.list_extract_runs(
                 extraction_agent_id=self.id,
+                skip=page * limit,
+                limit=limit,
             )
         )
 
@@ -416,6 +446,12 @@ class LlamaExtract(BaseComponent):
     verbose: bool = Field(
         default=False, description="Show verbose output when extracting files."
     )
+    verify: Optional[bool] = Field(
+        default=True, description="Simple SSL verification option."
+    )
+    httpx_timeout: Optional[float] = Field(
+        default=60, description="Timeout for the httpx client."
+    )
     _async_client: AsyncLlamaCloud = PrivateAttr()
     _thread_pool: ThreadPoolExecutor = PrivateAttr()
     _project_id: Optional[str] = PrivateAttr()
@@ -431,6 +467,8 @@ class LlamaExtract(BaseComponent):
         show_progress: bool = True,
         project_id: Optional[str] = None,
         organization_id: Optional[str] = None,
+        verify: Optional[bool] = True,
+        httpx_timeout: Optional[float] = 60,
         verbose: bool = False,
     ):
         if not api_key:
@@ -448,11 +486,18 @@ class LlamaExtract(BaseComponent):
             max_timeout=max_timeout,
             num_workers=num_workers,
             show_progress=show_progress,
+            verify=verify,
+            httpx_timeout=httpx_timeout,
             verbose=verbose,
         )
+        self._httpx_client = httpx.AsyncClient(verify=verify, timeout=httpx_timeout)
+        self.verify = verify
+        self.httpx_timeout = httpx_timeout
 
         self._async_client = AsyncLlamaCloud(
-            token=self.api_key, base_url=self.base_url, timeout=None
+            token=self.api_key,
+            base_url=self.base_url,
+            httpx_client=self._httpx_client,
         )
         self._thread_pool = ThreadPoolExecutor(
             max_workers=min(10, (os.cpu_count() or 1) + 4)
@@ -481,17 +526,22 @@ class LlamaExtract(BaseComponent):
         def run_coro() -> T:
             # Create a new client for this thread
             async def wrapped_coro() -> T:
+                assert (
+                    self._httpx_client is not None
+                ), "httpx_client should be initialized"
+                # Create a new client with the same configuration as the original
                 async with httpx.AsyncClient(
-                    timeout=self._async_client._client_wrapper.httpx_client.timeout,
+                    verify=self.verify,
+                    timeout=self.httpx_timeout,
                 ) as client:
-                    # Replace the client in the coro's context
-                    original_client = self._async_client._client_wrapper.httpx_client
+                    # Temporarily replace the client
                     self._async_client._client_wrapper.httpx_client = client
                     try:
                         return await coro
                     finally:
+                        # Restore the original client
                         self._async_client._client_wrapper.httpx_client = (
-                            original_client
+                            self._httpx_client
                         )
 
             return asyncio.run(wrapped_coro())
@@ -526,11 +576,13 @@ class LlamaExtract(BaseComponent):
 
         agent = self._run_in_thread(
             self._async_client.llama_extract.create_extraction_agent(
-                name=name,
-                data_schema=data_schema,
-                config=config or DEFAULT_EXTRACT_CONFIG,
                 project_id=self._project_id,
                 organization_id=self._organization_id,
+                request=ExtractAgentCreate(
+                    name=name,
+                    data_schema=data_schema,
+                    config=config or DEFAULT_EXTRACT_CONFIG,
+                ),
             )
         )
 
@@ -592,6 +644,8 @@ class LlamaExtract(BaseComponent):
             num_workers=self.num_workers,
             show_progress=self.show_progress,
             verbose=self.verbose,
+            verify=self.verify,
+            httpx_timeout=self.httpx_timeout,
         )
 
     def list_agents(self) -> List[ExtractionAgent]:

llama_cloud_services/parse/base.py

@@ -111,6 +111,10 @@ class LlamaParse(BasePydanticReader):
     )
 
     # Parsing specific configurations (Alphabetical order)
+    adaptive_long_table: Optional[bool] = Field(
+        default=False,
+        description="If set to true, LlamaParse will try to detect long table and adapt the output.",
+    )
     annotate_links: Optional[bool] = Field(
         default=False,
         description="Annotate links found in the document to extract their URL.",
@@ -163,14 +167,7 @@ class LlamaParse(BasePydanticReader):
         default=None,
         description="The top margin of the bounding box to use to extract text from documents expressed as a float between 0 and 1 representing the percentage of the page height.",
     )
-    complemental_formatting_instruction: Optional[str] = Field(
-        default=None,
-        description="The complemental formatting instruction for the parser. Tell llamaParse how some thing should to be formatted, while retaining the markdown output.",
-    )
-    content_guideline_instruction: Optional[str] = Field(
-        default=None,
-        description="The content guideline for the parser. Tell LlamaParse how the content should be changed / transformed.",
-    )
+
     continuous_mode: Optional[bool] = Field(
         default=False,
         description="Parse documents continuously, leading to better results on documents where tables span across two pages.",
@@ -203,10 +200,7 @@ class LlamaParse(BasePydanticReader):
         default=False,
         description="Note: Non compatible with gpt-4o. If set to true, the parser will use a faster mode to extract text from documents. This mode will skip OCR of images, and table/heading reconstruction.",
     )
-    formatting_instruction: Optional[str] = Field(
-        default=None,
-        description="The Formatting instruction for the parser. Override default llamaParse behavior. In most case you want to use complemental_formatting_instruction instead.",
-    )
+
     guess_xlsx_sheet_names: Optional[bool] = Field(
         default=False,
         description="Whether to guess the sheet names of the xlsx file.",
@@ -282,10 +276,18 @@ class LlamaParse(BasePydanticReader):
         default=None,
         description="A templated suffix to add to the beginning of each page. If it contain `{page_number}`, it will be replaced by the page number.",
     )
+    parse_mode: Optional[str] = Field(
+        default=None,
+        description="The parsing mode to use, see ParsingMode enum for possible values ",
+    )
     premium_mode: Optional[bool] = Field(
         default=False,
         description="Use our best parser mode if set to True.",
     )
+    preserve_layout_alignment_across_pages: Optional[bool] = Field(
+        default=False,
+        description="Preserve grid alignment across page in text mode.",
+    )
     skip_diagonal_text: Optional[bool] = Field(
         default=False,
         description="If set to true, the parser will ignore diagonal text (when the text rotation in degrees modulo 90 is not 0).",
@@ -327,6 +329,14 @@ class LlamaParse(BasePydanticReader):
         default=None,
         description="The named JSON Schema to use to structure the output of the parsing job. For convenience / testing, LlamaParse provides a few named JSON Schema that can be used directly. Use 'imFeelingLucky' to let llamaParse dream the schema.",
     )
+    system_prompt: Optional[str] = Field(
+        default=None,
+        description="The system prompt. Replace llamaParse default system prompt, may impact accuracy",
+    )
+    system_prompt_append: Optional[str] = Field(
+        default=None,
+        description="String to append to default system prompt.",
+    )
     take_screenshot: Optional[bool] = Field(
         default=False,
         description="Whether to take screenshot of each page of the document.",
@@ -335,9 +345,9 @@ class LlamaParse(BasePydanticReader):
         default=None,
         description="The target pages to extract text from documents. Describe as a comma separated list of page numbers. The first page of the document is page 0",
     )
-    use_vendor_multimodal_model: Optional[bool] = Field(
-        default=False,
-        description="Whether to use the vendor multimodal API.",
+    user_prompt: Optional[str] = Field(
+        default=None,
+        description="The user prompt. Replace llamaParse default user prompt",
     )
     vendor_multimodal_api_key: Optional[str] = Field(
         default=None,
@@ -357,6 +367,18 @@ class LlamaParse(BasePydanticReader):
         default=None,
         description="The bounding box to use to extract text from documents describe as a string containing the bounding box margins",
     )
+    complemental_formatting_instruction: Optional[str] = Field(
+        default=None,
+        description="The complemental formatting instruction for the parser. Tell llamaParse how some thing should to be formatted, while retaining the markdown output.",
+    )
+    content_guideline_instruction: Optional[str] = Field(
+        default=None,
+        description="The content guideline for the parser. Tell LlamaParse how the content should be changed / transformed.",
+    )
+    formatting_instruction: Optional[str] = Field(
+        default=None,
+        description="The Formatting instruction for the parser. Override default llamaParse behavior. In most case you want to use complemental_formatting_instruction instead.",
+    )
     gpt4o_mode: Optional[bool] = Field(
         default=False,
         description="Whether to use gpt-4o extract text from documents.",
@@ -373,6 +395,11 @@ class LlamaParse(BasePydanticReader):
         default="", description="The parsing instruction for the parser."
     )
 
+    use_vendor_multimodal_model: Optional[bool] = Field(
+        default=False,
+        description="Whether to use the vendor multimodal API.",
+    )
+
     @field_validator("api_key", mode="before", check_fields=True)
     @classmethod
     def validate_api_key(cls, v: str) -> str:
@@ -501,6 +528,9 @@ class LlamaParse(BasePydanticReader):
 
         data["from_python_package"] = True
 
+        if self.adaptive_long_table:
+            data["adaptive_long_table"] = self.adaptive_long_table
+
         if self.annotate_links:
             data["annotate_links"] = self.annotate_links
 
@@ -552,11 +582,17 @@ class LlamaParse(BasePydanticReader):
             data["bbox_top"] = self.bbox_top
 
         if self.complemental_formatting_instruction:
+            print(
+                "WARNING: complemental_formatting_instruction is deprecated and may be remove in a future release. Use system_prompt, system_prompt_append or user_prompt instead."
+            )
             data[
                 "complemental_formatting_instruction"
             ] = self.complemental_formatting_instruction
 
         if self.content_guideline_instruction:
+            print(
+                "WARNING: content_guideline_instruction is deprecated and may be remove in a future release. Use system_prompt, system_prompt_append or user_prompt instead."
+            )
             data["content_guideline_instruction"] = self.content_guideline_instruction
 
         if self.continuous_mode:
@@ -584,6 +620,9 @@ class LlamaParse(BasePydanticReader):
             data["fast_mode"] = self.fast_mode
 
         if self.formatting_instruction:
+            print(
+                "WARNING: formatting_instruction is deprecated and may be remove in a future release. Use system_prompt, system_prompt_append or user_prompt instead."
+            )
             data["formatting_instruction"] = self.formatting_instruction
 
         if self.guess_xlsx_sheet_names:
@@ -623,6 +662,9 @@ class LlamaParse(BasePydanticReader):
             data["invalidate_cache"] = self.invalidate_cache
 
         if self.is_formatting_instruction:
+            print(
+                "WARNING: formatting_instruction is deprecated and may be remove in a future release. Use system_prompt, system_prompt_append or user_prompt instead."
+            )
             data["is_formatting_instruction"] = self.is_formatting_instruction
 
         if self.job_timeout_extra_time_per_page_in_seconds is not None:
@@ -664,13 +706,21 @@ class LlamaParse(BasePydanticReader):
 
         if self.parsing_instruction:
             print(
-                "WARNING: parsing_instruction is deprecated. Use complemental_formatting_instruction or content_guideline_instruction instead."
+                "WARNING: parsing_instruction is deprecated. Use system_prompt, system_prompt_append or user_prompt instead."
             )
             data["parsing_instruction"] = self.parsing_instruction
 
+        if self.parse_mode:
+            data["parse_mode"] = self.parse_mode
+
         if self.premium_mode:
             data["premium_mode"] = self.premium_mode
 
+        if self.preserve_layout_alignment_across_pages:
+            data[
+                "preserve_layout_alignment_across_pages"
+            ] = self.preserve_layout_alignment_across_pages
+
         if self.skip_diagonal_text:
             data["skip_diagonal_text"] = self.skip_diagonal_text
 
@@ -699,13 +749,17 @@ class LlamaParse(BasePydanticReader):
             data[
                 "structured_output_json_schema_name"
             ] = self.structured_output_json_schema_name
-
+        if self.system_prompt is not None:
+            data["system_prompt"] = self.system_prompt
+        if self.system_prompt_append is not None:
+            data["system_prompt_append"] = self.system_prompt_append
         if self.take_screenshot:
             data["take_screenshot"] = self.take_screenshot
 
         if self.target_pages is not None:
             data["target_pages"] = self.target_pages
-
+        if self.user_prompt is not None:
+            data["user_prompt"] = self.user_prompt
         if self.use_vendor_multimodal_model:
             data["use_vendor_multimodal_model"] = self.use_vendor_multimodal_model
 

llama_cloud_services/parse/utils.py

@@ -14,6 +14,16 @@ class ResultType(str, Enum):
     STRUCTURED = "structured"
 
 
+class ParsingMode(str, Enum):
+    """The parsing mode for the parser."""
+
+    parse_page_without_llm = "parse_page_without_llm"
+    parse_page_with_llm = "parse_page_with_llm"
+    parse_page_with_lvm = "parse_page_with_lvm"
+    parse_page_with_agent = "parse_page_with_agent"
+    parse_document_with_llm = "parse_document_with_llm"
+
+
 class Language(str, Enum):
     BAZA = "abq"
     ADYGHE = "ady"

llama_parse/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "llama-parse"
-version = "0.6.1"
+version = "0.6.8"
 description = "Parse files into RAG-Optimized formats."
 authors = ["Logan Markewich <logan@llamaindex.ai>"]
 license = "MIT"
@@ -13,7 +13,7 @@ packages = [{include = "llama_parse"}]
 
 [tool.poetry.dependencies]
 python = ">=3.9,<4.0"
-llama-cloud-services = ">=0.6.1"
+llama-cloud-services = ">=0.6.7"
 
 [tool.poetry.group.dev.dependencies]
 pytest = "^8.0.0"

pyproject.toml

@@ -8,7 +8,7 @@ python_version = "3.10"
 
 [tool.poetry]
 name = "llama-cloud-services"
-version = "0.6.1"
+version = "0.6.8"
 description = "Tailored SDK clients for LlamaCloud services."
 authors = ["Logan Markewich <logan@runllama.ai>"]
 license = "MIT"
@@ -18,7 +18,7 @@ packages = [{include = "llama_cloud_services"}]
 [tool.poetry.dependencies]
 python = ">=3.9,<4.0"
 llama-index-core = ">=0.11.0"
-llama-cloud = "^0.1.11"
+llama-cloud = "^0.1.16"
 pydantic = "!=2.10"
 click = "^8.1.7"
 python-dotenv = "^1.0.1"

tests/extract/test_extract_api.py

@@ -182,7 +182,14 @@ class TestExtractionAgent:
         assert "new_field" in updated_agent.data_schema["properties"]
 
     def test_list_extraction_runs(self, test_agent: ExtractionAgent):
-        assert len(test_agent.list_extraction_runs()) == 0
+        assert test_agent.list_extraction_runs().total == 0
         test_agent.extract(TEST_PDF)
         runs = test_agent.list_extraction_runs()
-        assert len(runs) > 0
+        assert runs.total > 0
+
+    def test_delete_extraction_run(self, test_agent: ExtractionAgent):
+        assert test_agent.list_extraction_runs().total == 0
+        run = test_agent.extract(TEST_PDF)
+        test_agent.delete_extraction_run(run.id)
+        runs = test_agent.list_extraction_runs()
+        assert runs.total == 0

tests/report/test_llama_report.py

@@ -81,6 +81,11 @@ async def test_create_and_delete_report(
 
 
 @pytest.mark.asyncio
+@pytest.mark.xfail(
+    condition=lambda: os.getenv("CI"),
+    reason="Report plan sometimes times out",
+    raises=TimeoutError,
+)
 async def test_report_plan_workflow(report: ReportClient) -> None:
     """Test the report planning workflow."""
     # Wait for the plan