relax pydantic job object (#784)

* parse: loosen PageItem.rows type hint

* bump version to 0.6.41

.github/workflows/publish_release.yml

@@ -41,7 +41,7 @@ jobs:
 
       - name: Wait for PyPI to update
         run: |
-          sleep 60
+          sleep 120
 
       - name: Update llama-parse lock file
         run: |

examples/extract/data/insider_transactions/.gitignore

@@ -0,0 +1 @@
+sec_form_4_dump.json

llama_cloud_services/extract/extract.py

@@ -8,6 +8,12 @@ import secrets
 import warnings
 import httpx
 from pydantic import BaseModel
+from tenacity import (
+    retry_if_exception,
+    stop_after_attempt,
+    wait_exponential_jitter,
+    AsyncRetrying,
+)
 from llama_cloud import (
     ExtractAgent as CloudExtractAgent,
     ExtractConfig,
@@ -17,12 +23,12 @@ from llama_cloud import (
     File,
     ExtractMode,
     StatusEnum,
-    Project,
     ExtractTarget,
     LlamaExtractSettings,
     PaginatedExtractRunsResponse,
 )
 from llama_cloud.client import AsyncLlamaCloud
+from llama_cloud.core.api_error import ApiError
 from llama_cloud_services.extract.utils import (
     JSONObjectType,
     augment_async_errors,
@@ -45,6 +51,17 @@ DEFAULT_EXTRACT_CONFIG = ExtractConfig(
 )
 
 
+def _is_retryable_error(exception: BaseException) -> bool:
+    """Check if an exception is retryable."""
+    if isinstance(exception, ApiError):
+        return exception.status_code in (502, 503, 504, 425, 408)
+    elif isinstance(
+        exception, (httpx.HTTPStatusError, httpx.RequestError, httpx.TimeoutException)
+    ):
+        return True
+    return False
+
+
 class SourceText:
     def __init__(
         self,
@@ -120,9 +137,6 @@ def run_in_thread(
 
 
 def _extraction_config_warning(config: ExtractConfig) -> None:
-    if config.extraction_mode == ExtractMode.ACCURATE:
-        warnings.warn("ACCURATE extraction mode is deprecated. Using BALANCED instead.")
-        config.extraction_mode = ExtractMode.BALANCED
     if config.use_reasoning:
         warnings.warn(
             "`use_reasoning` is an experimental feature. Results will be available in "
@@ -232,9 +246,8 @@ class ExtractionAgent:
             ValueError: If filename is not provided for bytes input or for file-like objects
                        without a name attribute.
         """
+        file_contents: Optional[Union[BufferedIOBase, BytesIO]] = None
         try:
-            file_contents: Union[BufferedIOBase, BytesIO]
-
             if file_input.text_content is not None:
                 # Handle direct text content
                 file_contents = BytesIO(file_input.text_content.encode("utf-8"))
@@ -261,7 +274,7 @@ class ExtractionAgent:
                 project_id=self._project_id, upload_file=file_contents
             )
         finally:
-            if isinstance(file_contents, BufferedReader):
+            if file_contents is not None and isinstance(file_contents, BufferedReader):
                 file_contents.close()
 
     async def _upload_file(self, file_input: FileInput) -> File:
@@ -289,35 +302,60 @@ class ExtractionAgent:
 
         return await self.upload_file(source_text)
 
+    async def _get_job_with_retry(self, job_id: str) -> ExtractJob:
+        """Get job with retry logic for transient errors."""
+        async for attempt in AsyncRetrying(
+            retry=retry_if_exception(_is_retryable_error),
+            stop=stop_after_attempt(5),
+            wait=wait_exponential_jitter(initial=1, max=60, jitter=5),
+            reraise=True,
+        ):
+            with attempt:
+                return await self._client.llama_extract.get_job(job_id=job_id)
+
+    async def _get_run_with_retry(self, job_id: str) -> ExtractRun:
+        """Get extraction run with retry logic for transient errors."""
+        async for attempt in AsyncRetrying(
+            retry=retry_if_exception(_is_retryable_error),
+            stop=stop_after_attempt(3),
+            wait=wait_exponential_jitter(initial=1, max=20, jitter=3),
+            reraise=True,
+        ):
+            with attempt:
+                return await self._client.llama_extract.get_run_by_job_id(job_id=job_id)
+
     async def _wait_for_job_result(self, job_id: str) -> Optional[ExtractRun]:
         """Wait for and return the results of an extraction job."""
         start = time.perf_counter()
         tries = 0
+
         while True:
             await asyncio.sleep(self.check_interval)
             tries += 1
-            job = await self._client.llama_extract.get_job(
-                job_id=job_id,
-            )
 
-            if job.status == StatusEnum.SUCCESS:
-                return await self._client.llama_extract.get_run_by_job_id(
-                    job_id=job_id,
-                )
-            elif job.status == StatusEnum.PENDING:
-                end = time.perf_counter()
-                if end - start > self.max_timeout:
-                    raise Exception(f"Timeout while extracting the file: {job_id}")
-                if self._verbose and tries % 10 == 0:
-                    print(".", end="", flush=True)
-                continue
-            else:
-                warnings.warn(
-                    f"Failure in job: {job_id}, status: {job.status}, error: {job.error}"
-                )
-                return await self._client.llama_extract.get_run_by_job_id(
-                    job_id=job_id,
-                )
+            try:
+                job = await self._get_job_with_retry(job_id)
+
+                if job.status == StatusEnum.SUCCESS:
+                    return await self._get_run_with_retry(job_id)
+                elif job.status == StatusEnum.PENDING:
+                    end = time.perf_counter()
+                    if end - start > self.max_timeout:
+                        raise Exception(f"Timeout while extracting the file: {job_id}")
+                    if self._verbose and tries % 10 == 0:
+                        print(".", end="", flush=True)
+                    continue
+                else:
+                    warnings.warn(
+                        f"Failure in job: {job_id}, status: {job.status}, error: {job.error}"
+                    )
+                    return await self._get_run_with_retry(job_id)
+
+            except Exception as e:
+                # If we get a non-retryable error or all retries are exhausted, re-raise
+                if self._verbose:
+                    print(f"\nError in job polling for {job_id}: {e}")
+                raise e
 
     def save(self) -> None:
         """Persist the extraction agent's schema and config to the database.
@@ -633,21 +671,8 @@ class LlamaExtract(BaseComponent):
         self._thread_pool = ThreadPoolExecutor(
             max_workers=min(10, (os.cpu_count() or 1) + 4)
         )
-        # Fetch default project id if not provided
         if not project_id:
             project_id = os.getenv("LLAMA_CLOUD_PROJECT_ID", None)
-            if not project_id:
-                print("No project_id provided, fetching default project.")
-                projects: List[Project] = self._run_in_thread(
-                    self._async_client.projects.list_projects()
-                )
-                default_project = [p for p in projects if p.is_default]
-                if not default_project:
-                    raise ValueError(
-                        "No default project found. Please provide a project_id."
-                    )
-                project_id = default_project[0].id
-
         self._project_id = project_id
         self._organization_id = organization_id
 
@@ -711,6 +736,8 @@ 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 get_agent(

llama_cloud_services/parse/base.py

@@ -2,6 +2,7 @@ import asyncio
 import mimetypes
 import os
 import time
+import warnings
 from contextlib import asynccontextmanager
 from copy import deepcopy
 from enum import Enum
@@ -13,21 +14,29 @@ from urllib.parse import urlparse
 import httpx
 from fsspec import AbstractFileSystem
 from llama_index.core.async_utils import asyncio_run, run_jobs
-from llama_index.core.bridge.pydantic import Field, PrivateAttr, field_validator
+from llama_index.core.bridge.pydantic import (
+    Field,
+    PrivateAttr,
+    field_validator,
+    model_validator,
+)
 from llama_index.core.constants import DEFAULT_BASE_URL
 from llama_index.core.readers.base import BasePydanticReader
 from llama_index.core.readers.file.base import get_default_fs
 from llama_index.core.schema import Document
 
+from llama_cloud_services.utils import check_extra_params
 from llama_cloud_services.parse.types import JobResult
 from llama_cloud_services.parse.utils import (
     SUPPORTED_FILE_TYPES,
     ResultType,
     ParsingMode,
     FailedPageMode,
+    expand_target_pages,
     nest_asyncio_err,
     nest_asyncio_msg,
     make_api_request,
+    partition_pages,
 )
 
 # can put in a path to the file or the file bytes itself
@@ -57,6 +66,36 @@ def build_url(
     return base_url
 
 
+class JobFailedException(Exception):
+    """Parse job failed exception."""
+
+    def __init__(
+        self,
+        job_id: str,
+        status: str,
+        error_code: Optional[str] = None,
+        error_message: Optional[str] = None,
+    ):
+        exception_str = (
+            f"Job ID: {job_id} failed with status: {status}, "
+            f'Error code: {error_code or "No error code found"}, '
+            f'Error message: {error_message or "No error message found"}'
+        )
+        super().__init__(exception_str)
+        self.job_id = job_id
+        self.status = status
+        self.error_code = error_code
+        self.error_message = error_message
+
+    @classmethod
+    def from_result(cls, result_json: Dict[str, Any]) -> "JobFailedException":
+        job_id = result_json["id"]
+        status = result_json["status"]
+        error_code = result_json.get("error_code")
+        error_message = result_json.get("error_message")
+        return cls(job_id, status, error_code=error_code, error_message=error_message)
+
+
 class BackoffPattern(str, Enum):
     """Backoff pattern for polling."""
 
@@ -234,6 +273,10 @@ class LlamaParse(BasePydanticReader):
         default=False,
         description="Whether to guess the sheet names of the xlsx file.",
     )
+    high_res_ocr: Optional[bool] = Field(
+        default=False,
+        description="If set to true, the parser will use high resolution OCR to extract text from images. This will increase the accuracy of the parsing job, but reduce the speed.",
+    )
     html_make_all_elements_visible: Optional[bool] = Field(
         default=False,
         description="If set to true, when parsing HTML the parser will consider all elements display not element as display block.",
@@ -297,6 +340,10 @@ class LlamaParse(BasePydanticReader):
         default=False,
         description="If set to true, the parser will output tables as HTML in the markdown.",
     )
+    outlined_table_extraction: Optional[bool] = Field(
+        default=False,
+        description="If set to true, the parser will use a dedicated approach to extract tables with outlined cells. This is useful for documents with spreadsheet-like tables where cells are outlined with borders. This could lead to false positives, so use with caution.",
+    )
     page_error_tolerance: Optional[float] = Field(
         default=None,
         description="The error tolerance for the number of pages with error in a doc (percentage express as 0-1). If we fail to parse a greater percentage of pages than the tolerance value we fail the job.",
@@ -410,10 +457,42 @@ class LlamaParse(BasePydanticReader):
         default=None,
         description="The model name for the vendor multimodal API.",
     )
+    model: Optional[str] = Field(
+        default=None,
+        description="The document model name to be used with `parse_with_agent`.",
+    )
     webhook_url: Optional[str] = Field(
         default=None,
         description="A URL that needs to be called at the end of the parsing job.",
     )
+    partition_pages: Optional[int] = Field(
+        default=None,
+        description="If set, documents will automatically be partitioned into segments containing the specified number of pages at most. Parsing will be split into separate jobs for each partition segment. Can be used in combination with targetPages and maxPages.",
+    )
+    hide_headers: Optional[bool] = Field(
+        default=False,
+        description="Whether to hide page header in output markdown.",
+    )
+    hide_footers: Optional[bool] = Field(
+        default=False,
+        description="Whether to hide page footers in output markdown.",
+    )
+    page_header_suffix: Optional[str] = Field(
+        default=None,
+        description="A suffix to add to the page header in the output markdown.",
+    )
+    page_header_prefix: Optional[str] = Field(
+        default=None,
+        description="A prefix to add to the page header in the output markdown.",
+    )
+    page_footer_suffix: Optional[str] = Field(
+        default=None,
+        description="A suffix to add to the page footer in the output markdown.",
+    )
+    page_footer_prefix: Optional[str] = Field(
+        default=None,
+        description="A prefix to add to the page footer in the output markdown.",
+    )
 
     # Deprecated
     bounding_box: Optional[str] = Field(
@@ -453,6 +532,21 @@ class LlamaParse(BasePydanticReader):
         description="Whether to use the vendor multimodal API.",
     )
 
+    @model_validator(mode="before")
+    @classmethod
+    def warn_extra_params(cls, data: Dict[str, Any]) -> Dict[str, Any]:
+        extra_params, suggestions = check_extra_params(cls, data)
+        if extra_params:
+            suggestions = [f"\n - {suggestion}" for suggestion in suggestions]
+            suggestions_str = "".join(suggestions)
+            warnings.warn(
+                "The following parameters are unused: "
+                + ", ".join(extra_params)
+                + f".\n{suggestions_str}",
+            )
+
+        return data
+
     @field_validator("api_key", mode="before", check_fields=True)
     @classmethod
     def validate_api_key(cls, v: str) -> str:
@@ -540,6 +634,7 @@ class LlamaParse(BasePydanticReader):
         file_input: FileInput,
         extra_info: Optional[dict] = None,
         fs: Optional[AbstractFileSystem] = None,
+        partition_target_pages: Optional[str] = None,
     ) -> str:
         files = None
         file_handle = None
@@ -690,6 +785,9 @@ class LlamaParse(BasePydanticReader):
         if self.html_make_all_elements_visible:
             data["html_make_all_elements_visible"] = self.html_make_all_elements_visible
 
+        if self.high_res_ocr:
+            data["high_res_ocr"] = self.high_res_ocr
+
         if self.html_remove_fixed_elements:
             data["html_remove_fixed_elements"] = self.html_remove_fixed_elements
 
@@ -752,12 +850,33 @@ class LlamaParse(BasePydanticReader):
         if self.output_tables_as_HTML:
             data["output_tables_as_HTML"] = self.output_tables_as_HTML
 
+        if self.outlined_table_extraction:
+            data["outlined_table_extraction"] = self.outlined_table_extraction
+
         if self.page_error_tolerance is not None:
             data["page_error_tolerance"] = self.page_error_tolerance
 
         if self.page_prefix is not None:
             data["page_prefix"] = self.page_prefix
 
+        if self.hide_headers:
+            data["hide_headers"] = self.hide_headers
+
+        if self.hide_footers:
+            data["hide_footers"] = self.hide_footers
+
+        if self.page_header_suffix is not None:
+            data["page_header_suffix"] = self.page_header_suffix
+
+        if self.page_header_prefix is not None:
+            data["page_header_prefix"] = self.page_header_prefix
+
+        if self.page_footer_suffix is not None:
+            data["page_footer_suffix"] = self.page_footer_suffix
+
+        if self.page_footer_prefix is not None:
+            data["page_footer_prefix"] = self.page_footer_prefix
+
         # only send page separator to server if it is not None
         # as if a null, "" string is sent the server will then ignore the page separator instead of using the default
         if self.page_separator is not None:
@@ -834,7 +953,9 @@ class LlamaParse(BasePydanticReader):
         if self.take_screenshot:
             data["take_screenshot"] = self.take_screenshot
 
-        if self.target_pages is not None:
+        if partition_target_pages is not None:
+            data["target_pages"] = partition_target_pages
+        elif 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
@@ -847,6 +968,9 @@ class LlamaParse(BasePydanticReader):
         if self.vendor_multimodal_model_name is not None:
             data["vendor_multimodal_model_name"] = self.vendor_multimodal_model_name
 
+        if self.model is not None:
+            data["model"] = self.model
+
         if self.webhook_url is not None:
             data["webhook_url"] = self.webhook_url
 
@@ -928,15 +1052,7 @@ class LlamaParse(BasePydanticReader):
                         print(".", end="", flush=True)
                     current_interval = self._calculate_backoff(current_interval)
                 else:
-                    error_code = result_json.get("error_code", "No error code found")
-                    error_message = result_json.get(
-                        "error_message", "No error message found"
-                    )
-                    exception_str = (
-                        f"Job ID: {job_id} failed with status: {status}, "
-                        f"Error code: {error_code}, Error message: {error_message}"
-                    )
-                    raise Exception(exception_str)
+                    raise JobFailedException.from_result(result_json)
             except (
                 httpx.ConnectError,
                 httpx.ReadError,
@@ -945,6 +1061,7 @@ class LlamaParse(BasePydanticReader):
                 httpx.ReadTimeout,
                 httpx.WriteTimeout,
                 httpx.HTTPStatusError,
+                httpx.RemoteProtocolError,
             ) as err:
                 error_count += 1
                 end = time.time()
@@ -965,9 +1082,39 @@ class LlamaParse(BasePydanticReader):
         extra_info: Optional[dict] = None,
         fs: Optional[AbstractFileSystem] = None,
         result_type: Optional[str] = None,
+        num_workers: Optional[int] = None,
+    ) -> List[Tuple[str, Dict[str, Any]]]:
+        if self.partition_pages is None:
+            job_results = [
+                await self._parse_one_unpartitioned(
+                    file_path,
+                    extra_info=extra_info,
+                    fs=fs,
+                    result_type=result_type,
+                )
+            ]
+        else:
+            job_results = await self._parse_one_partitioned(
+                file_path,
+                extra_info,
+                fs=fs,
+                result_type=result_type,
+                num_workers=num_workers,
+            )
+        return job_results
+
+    async def _parse_one_unpartitioned(
+        self,
+        file_path: FileInput,
+        extra_info: Optional[dict] = None,
+        fs: Optional[AbstractFileSystem] = None,
+        result_type: Optional[str] = None,
+        **create_kwargs: Any,
     ) -> Tuple[str, Dict[str, Any]]:
         """Create one parse job and wait for the result."""
-        job_id = await self._create_job(file_path, extra_info=extra_info, fs=fs)
+        job_id = await self._create_job(
+            file_path, extra_info=extra_info, fs=fs, **create_kwargs
+        )
         if self.verbose:
             print("Started parsing the file under job_id %s" % job_id)
         result = await self._get_job_result(
@@ -975,21 +1122,105 @@ class LlamaParse(BasePydanticReader):
         )
         return job_id, result
 
+    async def _parse_one_partitioned(
+        self,
+        file_path: FileInput,
+        extra_info: Optional[dict] = None,
+        fs: Optional[AbstractFileSystem] = None,
+        result_type: Optional[str] = None,
+        num_workers: Optional[int] = None,
+    ) -> List[Tuple[str, Dict[str, Any]]]:
+        """Partition a file and run separate parse jobs per partition segment."""
+        assert self.partition_pages is not None
+
+        num_workers = num_workers or self.num_workers
+        if num_workers < 1:
+            raise ValueError("Invalid number of workers")
+        if self.target_pages is not None:
+            jobs = [
+                self._parse_one_unpartitioned(
+                    file_path,
+                    extra_info=extra_info,
+                    fs=fs,
+                    result_type=result_type,
+                    partition_target_pages=target_pages,
+                )
+                for target_pages in partition_pages(
+                    expand_target_pages(self.target_pages),
+                    self.partition_pages,
+                    max_pages=self.max_pages,
+                )
+            ]
+            return await run_jobs(
+                jobs,
+                workers=num_workers,
+                desc="Getting job results",
+                show_progress=self.show_progress,
+            )
+
+        total = 0
+        results: List[Tuple[str, Dict[str, Any]]] = []
+        while self.max_pages is None or total < self.max_pages:
+            if (
+                self.max_pages is not None
+                and total + self.partition_pages >= self.max_pages
+            ):
+                size = self.max_pages - total
+            else:
+                size = self.partition_pages
+            if not size:
+                break
+            try:
+                # Fetch JSON result type first to get accurate pagination data
+                # and then fetch the user's desired result type if needed
+                job_id, json_result = await self._parse_one_unpartitioned(
+                    file_path,
+                    extra_info=extra_info,
+                    fs=fs,
+                    result_type=ResultType.JSON.value,
+                    partition_target_pages=f"{total}-{total + size - 1}",
+                )
+                result_type = result_type or self.result_type.value
+                if result_type == ResultType.JSON.value:
+                    job_result = json_result
+                else:
+                    job_result = await self._get_job_result(
+                        job_id, result_type, verbose=self.verbose
+                    )
+            except JobFailedException as e:
+                if results and e.error_code == "NO_DATA_FOUND_IN_FILE":
+                    # Expected when we try to read past the end of the file
+                    return results
+                raise
+            results.append((job_id, job_result))
+            if len(json_result["pages"]) < size:
+                break
+            total += size
+        return results
+
     async def _aload_data(
         self,
         file_path: FileInput,
         extra_info: Optional[dict] = None,
         fs: Optional[AbstractFileSystem] = None,
         verbose: bool = False,
+        num_workers: Optional[int] = None,
     ) -> List[Document]:
         """Load data from the input path."""
         try:
-            _job_id, result = await self._parse_one(
-                file_path, extra_info=extra_info, fs=fs
-            )
+            results = [
+                job_result
+                for _, job_result in await self._parse_one(
+                    file_path, extra_info, fs=fs, num_workers=num_workers
+                )
+            ]
+            # Flatten the resulting doc if it was partitioned
+            separator = self.page_separator or _DEFAULT_SEPARATOR
             docs = [
                 Document(
-                    text=result[self.result_type.value],
+                    text=separator.join(
+                        result[self.result_type.value] for result in results
+                    ),
                     metadata=extra_info or {},
                 )
             ]
@@ -1012,7 +1243,11 @@ class LlamaParse(BasePydanticReader):
         extra_info: Optional[dict] = None,
         fs: Optional[AbstractFileSystem] = None,
     ) -> List[Document]:
-        """Load data from the input path."""
+        """Load data from the input path.
+
+        File(s) which were partitioned before parsing will be loaded as a single
+        re-assembled Document.
+        """
         if isinstance(file_path, (str, PurePosixPath, Path, bytes, BufferedIOBase)):
             return await self._aload_data(
                 file_path, extra_info=extra_info, fs=fs, verbose=self.verbose
@@ -1024,6 +1259,7 @@ class LlamaParse(BasePydanticReader):
                     extra_info=extra_info,
                     fs=fs,
                     verbose=self.verbose and not self.show_progress,
+                    num_workers=1,
                 )
                 for f in file_path
             ]
@@ -1062,6 +1298,34 @@ class LlamaParse(BasePydanticReader):
             else:
                 raise e
 
+    async def _aparse_one(
+        self,
+        file_path: FileInput,
+        file_name: str,
+        extra_info: Optional[dict] = None,
+        fs: Optional[AbstractFileSystem] = None,
+        num_workers: Optional[int] = None,
+    ) -> List[JobResult]:
+        job_results = await self._parse_one(
+            file_path,
+            extra_info,
+            fs=fs,
+            result_type=ResultType.JSON.value,
+            num_workers=num_workers,
+        )
+        return [
+            JobResult(
+                job_id=job_id,
+                file_name=file_name,
+                job_result=job_result,
+                api_key=self.api_key,
+                base_url=self.base_url,
+                client=self.aclient,
+                page_separator=self.page_separator or _DEFAULT_SEPARATOR,
+            )
+            for job_id, job_result in job_results
+        ]
+
     async def aparse(
         self,
         file_path: Union[List[FileInput], FileInput],
@@ -1080,7 +1344,7 @@ class LlamaParse(BasePydanticReader):
             fs: Optional filesystem to use for reading files.
 
         Returns:
-            JobResult object or list of JobResult objects if multiple files were provided
+            JobResult object or list of JobResult objects if either multiple files were provided or file(s) were partitioned before parsing.
         """
 
         if isinstance(file_path, (str, PurePosixPath, Path, bytes, BufferedIOBase)):
@@ -1092,22 +1356,10 @@ class LlamaParse(BasePydanticReader):
                 file_name = extra_info["file_name"]
             else:
                 file_name = str(file_path)
-
-            job_id, job_result = await self._parse_one(
-                file_path,
-                extra_info=extra_info,
-                fs=fs,
-                result_type=ResultType.JSON.value,
-            )
-            return JobResult(
-                job_id=job_id,
-                file_name=file_name,
-                job_result=job_result,
-                api_key=self.api_key,
-                base_url=self.base_url,
-                client=self.aclient,
-                page_separator=self.page_separator or _DEFAULT_SEPARATOR,
+            result = await self._aparse_one(
+                file_path, file_name, extra_info=extra_info, fs=fs
             )
+            return result[0] if len(result) == 1 else result
 
         elif isinstance(file_path, list):
             file_names = []
@@ -1121,35 +1373,25 @@ class LlamaParse(BasePydanticReader):
                 else:
                     file_names.append(str(f))
 
+            job_results = []
             try:
-                job_results = await run_jobs(
+                for result in await run_jobs(
                     [
-                        self._parse_one(
+                        self._aparse_one(
                             f,
+                            file_names[i],
                             extra_info=extra_info,
                             fs=fs,
-                            result_type=ResultType.JSON.value,
+                            num_workers=1,
                         )
-                        for f in file_path
+                        for i, f in enumerate(file_path)
                     ],
                     workers=self.num_workers,
                     desc="Getting job results",
                     show_progress=self.show_progress,
-                )
-
-                # Create JobResults just using the job_ids and job_results
-                return [
-                    JobResult(
-                        job_id=job_id,
-                        file_name=file_names[i],
-                        job_result=job_result,
-                        api_key=self.api_key,
-                        base_url=self.base_url,
-                        client=self.aclient,
-                        page_separator=self.page_separator or _DEFAULT_SEPARATOR,
-                    )
-                    for i, (job_id, job_result) in enumerate(job_results)
-                ]
+                ):
+                    job_results.extend(result)
+                return job_results
 
             except RuntimeError as e:
                 if nest_asyncio_err in str(e):
@@ -1190,21 +1432,27 @@ class LlamaParse(BasePydanticReader):
                 raise e
 
     async def _aget_json(
-        self, file_path: FileInput, extra_info: Optional[dict] = None
+        self,
+        file_path: FileInput,
+        extra_info: Optional[dict] = None,
+        num_workers: Optional[int] = None,
     ) -> List[dict]:
         """Load data from the input path."""
         try:
-            job_id, result = await self._parse_one(
+            job_results = await self._parse_one(
                 file_path,
                 extra_info=extra_info,
                 result_type=ResultType.JSON.value,
+                num_workers=num_workers,
             )
-            result["job_id"] = job_id
 
-            if not isinstance(file_path, (bytes, BufferedIOBase)):
-                result["file_path"] = str(file_path)
-
-            return [result]
+            results = []
+            for job_id, job_result in job_results:
+                job_result["job_id"] = job_id
+                if not isinstance(file_path, (bytes, BufferedIOBase)):
+                    job_result["file_path"] = str(file_path)
+                results.append(job_result)
+            return results
         except Exception as e:
             file_repr = file_path if isinstance(file_path, str) else "<bytes/buffer>"
             print(f"Error while parsing the file '{file_repr}':", e)
@@ -1219,7 +1467,7 @@ class LlamaParse(BasePydanticReader):
         extra_info: Optional[dict] = None,
     ) -> List[dict]:
         """Load data from the input path."""
-        if isinstance(file_path, (str, Path)):
+        if isinstance(file_path, (str, PurePosixPath, Path, bytes, BufferedIOBase)):
             return await self._aget_json(file_path, extra_info=extra_info)
         elif isinstance(file_path, list):
             jobs = [self._aget_json(f, extra_info=extra_info) for f in file_path]
@@ -1240,7 +1488,7 @@ class LlamaParse(BasePydanticReader):
                     raise e
         else:
             raise ValueError(
-                "The input file_path must be a string or a list of strings."
+                "The input file_path must be a string, Path, bytes, BufferedIOBase, or a list of these types."
             )
 
     def get_json_result(

llama_cloud_services/parse/types.py

@@ -14,11 +14,14 @@ PAGE_REGEX = r"page[-_](\d+)\.jpg$"
 class JobMetadata(BaseModel):
     """Metadata about the job."""
 
-    job_pages: int = Field(description="The number of pages in the job.")
-    job_auto_mode_triggered_pages: int = Field(
-        description="The number of pages that triggered auto mode (thus increasing the cost)."
+    job_pages: int = Field(default=0, description="The number of pages in the job.")
+    job_auto_mode_triggered_pages: Optional[int] = Field(
+        default=None,
+        description="The number of pages that triggered auto mode (thus increasing the cost).",
+    )
+    job_is_cache_hit: bool = Field(
+        default=False, description="Whether the job was a cache hit."
     )
-    job_is_cache_hit: bool = Field(description="Whether the job was a cache hit.")
 
 
 class BBox(BaseModel):
@@ -43,7 +46,7 @@ class PageItem(BaseModel):
     md: Optional[str] = Field(
         default=None, description="The markdown-formatted content of the item."
     )
-    rows: Optional[List[List[str]]] = Field(
+    rows: Optional[List[List[Any]]] = Field(
         default=None, description="The rows of the item."
     )
     bBox: Optional[BBox] = Field(
@@ -124,32 +127,43 @@ class Page(BaseModel):
     items: List[PageItem] = Field(
         default_factory=list, description="The items in the page."
     )
-    status: str = Field(description="The status of the page.")
+    status: Optional[str] = Field(default=None, description="The status of the page.")
     links: List[SerializeAsAny[Any]] = Field(
         default_factory=list, description="The links in the page."
     )
     width: Optional[float] = Field(default=None, description="The width of the page.")
     height: Optional[float] = Field(default=None, description="The height of the page.")
-    triggeredAutoMode: bool = Field(
-        description="Whether the page triggered auto mode (thus increasing the cost)."
+    triggeredAutoMode: Optional[bool] = Field(
+        default=False,
+        description="Whether the page triggered auto mode (thus increasing the cost).",
+    )
+    parsingMode: str = Field(
+        default="", description="The parsing mode used for the page."
     )
-    parsingMode: str = Field(description="The parsing mode used for the page.")
     structuredData: Optional[Dict[str, Any]] = Field(
-        description="The structured data of the page."
+        default=None, description="The structured data of the page."
     )
     noStructuredContent: bool = Field(
-        description="Whether the page has no structured data."
+        default=True, description="Whether the page has no structured data."
+    )
+    noTextContent: bool = Field(
+        default=False, description="Whether the page has no text content."
     )
-    noTextContent: bool = Field(description="Whether the page has no text content.")
 
 
 class JobResult(BaseModel):
     """The raw JSON result from the LlamaParse API."""
 
-    pages: List[Page] = Field(description="The pages of the document.")
-    job_metadata: JobMetadata = Field(description="The metadata of the job.")
-    file_name: str = Field(description="The path to the file that was parsed.")
-    job_id: str = Field(description="The ID of the job.")
+    pages: List[Page] = Field(
+        default_factory=list, description="The pages of the document."
+    )
+    job_metadata: JobMetadata = Field(
+        default_factory=JobMetadata, description="The metadata of the job."
+    )
+    file_name: str = Field(
+        default="", description="The path to the file that was parsed."
+    )
+    job_id: str = Field(default="", description="The ID of the job.")
     is_done: bool = Field(default=False, description="Whether the job is done.")
     error: Optional[str] = Field(
         default=None, description="The error message if the job failed."

llama_cloud_services/parse/utils.py

@@ -1,4 +1,5 @@
 import httpx
+import itertools
 import logging
 from enum import Enum
 from tenacity import (
@@ -8,7 +9,7 @@ from tenacity import (
     retry_if_exception,
     before_sleep_log,
 )
-from typing import Any
+from typing import Any, Iterable, Iterator, Optional
 
 logger = logging.getLogger(__name__)
 
@@ -297,3 +298,56 @@ async def make_api_request(
         return response
 
     return await _make_request(url, **httpx_kwargs)
+
+
+def expand_target_pages(target_pages: str) -> Iterator[int]:
+    """Yield all values in target_pages."""
+    for target in target_pages.strip().split(","):
+        if "-" in target:
+            try:
+                start, end = map(int, target.strip().split("-"))
+                if start > end:
+                    raise ValueError
+                yield from range(start, end + 1)
+            except ValueError as e:
+                raise ValueError(f"Invalid page range: {target}") from e
+        else:
+            try:
+                yield int(target)
+            except ValueError as e:
+                raise ValueError(f"Invalid page number: {target}") from e
+
+
+def partition_pages(
+    pages: Iterable[int], size: int, max_pages: Optional[int] = None
+) -> Iterator[str]:
+    """Yield partitioned target_pages segments."""
+    if size < 1:
+        raise ValueError(f"Invalid partition segment size: {size}")
+    if max_pages is not None and max_pages < 1:
+        raise ValueError("Max pages must be > 0")
+    it = iter(pages)
+    total = 0
+    while max_pages is None or total < max_pages:
+        segment = tuple(itertools.islice(it, size))
+        if segment:
+            targets = []
+            for _k, g in itertools.groupby(enumerate(segment), lambda x: x[0] - x[1]):
+                group = [item[1] for item in g]
+                if len(group) > 1:
+                    start, end = group[0], group[-1]
+                    group_size = end - start + 1
+                    if max_pages is not None and total + group_size > max_pages:
+                        end -= total + group_size - max_pages
+                        group_size = end - start + 1
+                    if group_size > 1:
+                        targets.append(f"{start}-{end}")
+                    else:
+                        targets.append(str(start))
+                    total += group_size
+                else:
+                    targets.append(str(group[0]))
+                    total += 1
+            yield ",".join(targets)
+        else:
+            return

llama_cloud_services/utils.py

@@ -0,0 +1,29 @@
+import difflib
+from pydantic import BaseModel
+from typing import Any, Dict, List, Tuple, Type
+
+
+def check_extra_params(
+    model_cls: Type[BaseModel], data: Dict[str, Any]
+) -> Tuple[List[str], List[str]]:
+    # check if one of the parameters is unused, and warn the user
+    model_attributes = set(model_cls.model_fields.keys())
+    extra_params = [param for param in data.keys() if param not in model_attributes]
+
+    suggestions: List[str] = []
+    if extra_params:
+        # for each unused parameter, check if it is similar to a valid parameter and suggest a typo correction, else suggest to check the documentation / update the package
+        for param in extra_params:
+            similar_params = difflib.get_close_matches(
+                param, model_attributes, n=1, cutoff=0.8
+            )
+            if similar_params:
+                suggestions.append(
+                    f"'{param}' is not a valid parameter. Did you mean '{similar_params[0]}' instead of '{param}'?"
+                )
+            else:
+                suggestions.append(
+                    f"'{param}' is not a valid parameter. Please check the documentation or update the package."
+                )
+
+    return extra_params, suggestions

llama_parse/poetry.lock

@@ -1,4 +1,4 @@
-# This file is automatically @generated by Poetry 2.1.2 and should not be changed by hand.
+# This file is automatically @generated by Poetry 2.1.1 and should not be changed by hand.
 
 [[package]]
 name = "aiohappyeyeballs"
@@ -257,7 +257,7 @@ version = "2025.1.31"
 description = "Python package for providing Mozilla's CA Bundle."
 optional = false
 python-versions = ">=3.6"
-groups = ["main"]
+groups = ["main", "dev"]
 files = [
     {file = "certifi-2025.1.31-py3-none-any.whl", hash = "sha256:ca78db4565a652026a4db2bcdf68f2fb589ea80d0be70e03929ed730746b84fe"},
     {file = "certifi-2025.1.31.tar.gz", hash = "sha256:3d5da6925056f6f18f119200434a4780a94263f10d1c21d032a6f6b2baa20651"},
@@ -350,7 +350,7 @@ version = "3.4.1"
 description = "The Real First Universal Charset Detector. Open, modern and actively maintained alternative to Chardet."
 optional = false
 python-versions = ">=3.7"
-groups = ["main"]
+groups = ["main", "dev"]
 files = [
     {file = "charset_normalizer-3.4.1-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:91b36a978b5ae0ee86c394f5a54d6ef44db1de0815eb43de826d41d21e4af3de"},
     {file = "charset_normalizer-3.4.1-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:7461baadb4dc00fd9e0acbe254e3d7d2112e7f92ced2adc96e54ef6501c5f176"},
@@ -593,7 +593,7 @@ description = "Like `typing._eval_type`, but lets older Python versions use newe
 optional = false
 python-versions = ">=3.8"
 groups = ["main"]
-markers = "python_version == \"3.9\""
+markers = "python_version < \"3.10\""
 files = [
     {file = "eval_type_backport-0.2.2-py3-none-any.whl", hash = "sha256:cb6ad7c393517f476f96d456d0412ea80f0a8cf96f6892834cd9340149111b0a"},
     {file = "eval_type_backport-0.2.2.tar.gz", hash = "sha256:f0576b4cf01ebb5bd358d02314d31846af5e07678387486e2c798af0e7d849c1"},
@@ -892,31 +892,31 @@ colorama = ">=0.4"
 
 [[package]]
 name = "h11"
-version = "0.14.0"
+version = "0.16.0"
 description = "A pure-Python, bring-your-own-I/O implementation of HTTP/1.1"
 optional = false
-python-versions = ">=3.7"
-groups = ["main"]
+python-versions = ">=3.8"
+groups = ["main", "dev"]
 files = [
-    {file = "h11-0.14.0-py3-none-any.whl", hash = "sha256:e3fe4ac4b851c468cc8363d500db52c2ead036020723024a109d37346efaa761"},
-    {file = "h11-0.14.0.tar.gz", hash = "sha256:8f19fbbe99e72420ff35c00b27a34cb9937e902a8b810e2c88300c6f0a3b699d"},
+    {file = "h11-0.16.0-py3-none-any.whl", hash = "sha256:63cf8bbe7522de3bf65932fda1d9c2772064ffb3dae62d55932da54b31cb6c86"},
+    {file = "h11-0.16.0.tar.gz", hash = "sha256:4e35b956cf45792e4caa5885e69fba00bdbc6ffafbfa020300e549b208ee5ff1"},
 ]
 
 [[package]]
 name = "httpcore"
-version = "1.0.8"
+version = "1.0.9"
 description = "A minimal low-level HTTP client."
 optional = false
 python-versions = ">=3.8"
 groups = ["main"]
 files = [
-    {file = "httpcore-1.0.8-py3-none-any.whl", hash = "sha256:5254cf149bcb5f75e9d1b2b9f729ea4a4b883d1ad7379fc632b727cec23674be"},
-    {file = "httpcore-1.0.8.tar.gz", hash = "sha256:86e94505ed24ea06514883fd44d2bc02d90e77e7979c8eb71b90f41d364a1bad"},
+    {file = "httpcore-1.0.9-py3-none-any.whl", hash = "sha256:2d400746a40668fc9dec9810239072b40b4484b640a8c38fd654a024c7a1bf55"},
+    {file = "httpcore-1.0.9.tar.gz", hash = "sha256:6e34463af53fd2ab5d807f399a9b45ea31c3dfa2276f15a2c3f00afff6e176e8"},
 ]
 
 [package.dependencies]
 certifi = "*"
-h11 = ">=0.13,<0.15"
+h11 = ">=0.16"
 
 [package.extras]
 asyncio = ["anyio (>=4.0,<5.0)"]
@@ -955,7 +955,7 @@ version = "3.10"
 description = "Internationalized Domain Names in Applications (IDNA)"
 optional = false
 python-versions = ">=3.6"
-groups = ["main"]
+groups = ["main", "dev"]
 files = [
     {file = "idna-3.10-py3-none-any.whl", hash = "sha256:946d195a0d259cbba61165e88e65941f16e9b36ea6ddb97f00452bae8b1287d3"},
     {file = "idna-3.10.tar.gz", hash = "sha256:12f65c9b470abda6dc35cf8e63cc574b1c52b11df2c86030af0ac09b01b13ea9"},
@@ -971,7 +971,7 @@ description = "Read metadata from Python packages"
 optional = false
 python-versions = ">=3.9"
 groups = ["dev"]
-markers = "python_version == \"3.9\""
+markers = "python_version < \"3.10\""
 files = [
     {file = "importlib_metadata-8.6.1-py3-none-any.whl", hash = "sha256:02a89390c1e15fdfdc0d7c6b25cb3e62650d0494005c97d6f148bf5b9787525e"},
     {file = "importlib_metadata-8.6.1.tar.gz", hash = "sha256:310b41d755445d74569f993ccfc22838295d9fe005425094fad953d7f15c8580"},
@@ -1170,14 +1170,14 @@ test = ["ipykernel", "pre-commit", "pytest (<8)", "pytest-cov", "pytest-timeout"
 
 [[package]]
 name = "llama-cloud"
-version = "0.1.19"
+version = "0.1.23"
 description = ""
 optional = false
 python-versions = "<4,>=3.8"
 groups = ["main"]
 files = [
-    {file = "llama_cloud-0.1.19-py3-none-any.whl", hash = "sha256:d2d551baa4b63f7717f8e04cbb81b0f817e5450a66870c5487dd371f81dab8ec"},
-    {file = "llama_cloud-0.1.19.tar.gz", hash = "sha256:b0a5424ae0099ca27df2a2d7e5aec99066de9ca860ab65987c9f931f1ea7abff"},
+    {file = "llama_cloud-0.1.23-py3-none-any.whl", hash = "sha256:ce95b0705d85c99b3b27b0af0d16a17d9a81b14c96bf13c1063a1bd13d8d0446"},
+    {file = "llama_cloud-0.1.23.tar.gz", hash = "sha256:3d84a24a860f046d39a106c06742ec0ea39a574ac42bbf91706fe025f44e233e"},
 ]
 
 [package.dependencies]
@@ -1187,23 +1187,23 @@ pydantic = ">=1.10"
 
 [[package]]
 name = "llama-cloud-services"
-version = "0.6.18"
+version = "0.6.30"
 description = "Tailored SDK clients for LlamaCloud services."
 optional = false
 python-versions = "<4.0,>=3.9"
 groups = ["main"]
 files = [
-    {file = "llama_cloud_services-0.6.18-py3-none-any.whl", hash = "sha256:ba24ef36f4f78619722e5fe2cac8739175ba63f67cb20f0432be16801c78162c"},
-    {file = "llama_cloud_services-0.6.18.tar.gz", hash = "sha256:070a0e1397e2dd6da59e5f6a0437e3fd27e1c156170eea66001c4bcbec2f8783"},
+    {file = "llama_cloud_services-0.6.30-py3-none-any.whl", hash = "sha256:4d5817a9841fc3ba3409865c52d082090f4ef827931f0e5e4a89f5818c0d4e36"},
+    {file = "llama_cloud_services-0.6.30.tar.gz", hash = "sha256:2cb5004d13127aac52888ae9b3d70f899d598633520b2a2542bb62682d08d776"},
 ]
 
 [package.dependencies]
 click = ">=8.1.7,<9.0.0"
 eval-type-backport = {version = ">=0.2.0,<0.3.0", markers = "python_version < \"3.10\""}
-llama-cloud = "0.1.19"
-llama-index-core = ">=0.11.0"
+llama-cloud = "0.1.23"
+llama-index-core = ">=0.12.0"
 platformdirs = ">=4.3.7,<5.0.0"
-pydantic = "!=2.10"
+pydantic = ">=2.8,<2.10 || >2.10"
 python-dotenv = ">=1.0.1,<2.0.0"
 
 [[package]]
@@ -2469,19 +2469,19 @@ files = [
 
 [[package]]
 name = "requests"
-version = "2.32.3"
+version = "2.32.4"
 description = "Python HTTP for Humans."
 optional = false
 python-versions = ">=3.8"
-groups = ["main"]
+groups = ["main", "dev"]
 files = [
-    {file = "requests-2.32.3-py3-none-any.whl", hash = "sha256:70761cfe03c773ceb22aa2f671b4757976145175cdfca038c02654d061d6dcc6"},
-    {file = "requests-2.32.3.tar.gz", hash = "sha256:55365417734eb18255590a9ff9eb97e9e1da868d4ccd6402399eaf68af20a760"},
+    {file = "requests-2.32.4-py3-none-any.whl", hash = "sha256:27babd3cda2a6d50b30443204ee89830707d396671944c998b5975b031ac2b2c"},
+    {file = "requests-2.32.4.tar.gz", hash = "sha256:27d0316682c8a29834d3264820024b62a36942083d52caf2f14c0591336d3422"},
 ]
 
 [package.dependencies]
 certifi = ">=2017.4.17"
-charset-normalizer = ">=2,<4"
+charset_normalizer = ">=2,<4"
 idna = ">=2.5,<4"
 urllib3 = ">=1.21.1,<3"
 
@@ -2738,23 +2738,24 @@ files = [
 
 [[package]]
 name = "tornado"
-version = "6.4.2"
+version = "6.5.1"
 description = "Tornado is a Python web framework and asynchronous networking library, originally developed at FriendFeed."
 optional = false
-python-versions = ">=3.8"
+python-versions = ">=3.9"
 groups = ["dev"]
 files = [
-    {file = "tornado-6.4.2-cp38-abi3-macosx_10_9_universal2.whl", hash = "sha256:e828cce1123e9e44ae2a50a9de3055497ab1d0aeb440c5ac23064d9e44880da1"},
-    {file = "tornado-6.4.2-cp38-abi3-macosx_10_9_x86_64.whl", hash = "sha256:072ce12ada169c5b00b7d92a99ba089447ccc993ea2143c9ede887e0937aa803"},
-    {file = "tornado-6.4.2-cp38-abi3-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:1a017d239bd1bb0919f72af256a970624241f070496635784d9bf0db640d3fec"},
-    {file = "tornado-6.4.2-cp38-abi3-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:c36e62ce8f63409301537222faffcef7dfc5284f27eec227389f2ad11b09d946"},
-    {file = "tornado-6.4.2-cp38-abi3-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:bca9eb02196e789c9cb5c3c7c0f04fb447dc2adffd95265b2c7223a8a615ccbf"},
-    {file = "tornado-6.4.2-cp38-abi3-musllinux_1_2_aarch64.whl", hash = "sha256:304463bd0772442ff4d0f5149c6f1c2135a1fae045adf070821c6cdc76980634"},
-    {file = "tornado-6.4.2-cp38-abi3-musllinux_1_2_i686.whl", hash = "sha256:c82c46813ba483a385ab2a99caeaedf92585a1f90defb5693351fa7e4ea0bf73"},
-    {file = "tornado-6.4.2-cp38-abi3-musllinux_1_2_x86_64.whl", hash = "sha256:932d195ca9015956fa502c6b56af9eb06106140d844a335590c1ec7f5277d10c"},
-    {file = "tornado-6.4.2-cp38-abi3-win32.whl", hash = "sha256:2876cef82e6c5978fde1e0d5b1f919d756968d5b4282418f3146b79b58556482"},
-    {file = "tornado-6.4.2-cp38-abi3-win_amd64.whl", hash = "sha256:908b71bf3ff37d81073356a5fadcc660eb10c1476ee6e2725588626ce7e5ca38"},
-    {file = "tornado-6.4.2.tar.gz", hash = "sha256:92bad5b4746e9879fd7bf1eb21dce4e3fc5128d71601f80005afa39237ad620b"},
+    {file = "tornado-6.5.1-cp39-abi3-macosx_10_9_universal2.whl", hash = "sha256:d50065ba7fd11d3bd41bcad0825227cc9a95154bad83239357094c36708001f7"},
+    {file = "tornado-6.5.1-cp39-abi3-macosx_10_9_x86_64.whl", hash = "sha256:9e9ca370f717997cb85606d074b0e5b247282cf5e2e1611568b8821afe0342d6"},
+    {file = "tornado-6.5.1-cp39-abi3-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:b77e9dfa7ed69754a54c89d82ef746398be82f749df69c4d3abe75c4d1ff4888"},
+    {file = "tornado-6.5.1-cp39-abi3-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:253b76040ee3bab8bcf7ba9feb136436a3787208717a1fb9f2c16b744fba7331"},
+    {file = "tornado-6.5.1-cp39-abi3-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:308473f4cc5a76227157cdf904de33ac268af770b2c5f05ca6c1161d82fdd95e"},
+    {file = "tornado-6.5.1-cp39-abi3-musllinux_1_2_aarch64.whl", hash = "sha256:caec6314ce8a81cf69bd89909f4b633b9f523834dc1a352021775d45e51d9401"},
+    {file = "tornado-6.5.1-cp39-abi3-musllinux_1_2_i686.whl", hash = "sha256:13ce6e3396c24e2808774741331638ee6c2f50b114b97a55c5b442df65fd9692"},
+    {file = "tornado-6.5.1-cp39-abi3-musllinux_1_2_x86_64.whl", hash = "sha256:5cae6145f4cdf5ab24744526cc0f55a17d76f02c98f4cff9daa08ae9a217448a"},
+    {file = "tornado-6.5.1-cp39-abi3-win32.whl", hash = "sha256:e0a36e1bc684dca10b1aa75a31df8bdfed656831489bc1e6a6ebed05dc1ec365"},
+    {file = "tornado-6.5.1-cp39-abi3-win_amd64.whl", hash = "sha256:908e7d64567cecd4c2b458075589a775063453aeb1d2a1853eedb806922f568b"},
+    {file = "tornado-6.5.1-cp39-abi3-win_arm64.whl", hash = "sha256:02420a0eb7bf617257b9935e2b754d1b63897525d8a289c9d65690d580b4dcf7"},
+    {file = "tornado-6.5.1.tar.gz", hash = "sha256:84ceece391e8eb9b2b95578db65e920d2a61070260594819589609ba9bc6308c"},
 ]
 
 [[package]]
@@ -2806,7 +2807,7 @@ files = [
     {file = "typing_extensions-4.13.2-py3-none-any.whl", hash = "sha256:a439e7c04b49fec3e5d3e2beaa21755cadbbdc391694e28ccdd36ca4a1408f8c"},
     {file = "typing_extensions-4.13.2.tar.gz", hash = "sha256:e6c81219bd689f51865d9e372991c540bda33a0379d5573cddb9a3a23f7caaef"},
 ]
-markers = {dev = "python_version == \"3.9\""}
+markers = {dev = "python_version < \"3.10\""}
 
 [[package]]
 name = "typing-inspect"
@@ -2845,7 +2846,7 @@ version = "2.4.0"
 description = "HTTP library with thread-safe connection pooling, file post, and more."
 optional = false
 python-versions = ">=3.9"
-groups = ["main"]
+groups = ["main", "dev"]
 files = [
     {file = "urllib3-2.4.0-py3-none-any.whl", hash = "sha256:4e16665048960a0900c702d4a66415956a584919c03361cac9f1df5c5dd7e813"},
     {file = "urllib3-2.4.0.tar.gz", hash = "sha256:414bc6535b787febd7567804cc015fee39daab8ad86268f1310a9250697de466"},
@@ -3067,7 +3068,7 @@ description = "Backport of pathlib-compatible object wrapper for zip files"
 optional = false
 python-versions = ">=3.9"
 groups = ["dev"]
-markers = "python_version == \"3.9\""
+markers = "python_version < \"3.10\""
 files = [
     {file = "zipp-3.21.0-py3-none-any.whl", hash = "sha256:ac1bbe05fd2991f160ebce24ffbac5f6d11d83dc90891255885223d42b3cd931"},
     {file = "zipp-3.21.0.tar.gz", hash = "sha256:2c9958f6430a2040341a52eb608ed6dd93ef4392e02ffe219417c1b28b5dd1f4"},
@@ -3084,4 +3085,4 @@ type = ["pytest-mypy"]
 [metadata]
 lock-version = "2.1"
 python-versions = ">=3.9,<4.0"
-content-hash = "5630c89272551940c2c689fc2c78f5ced7b6dd904ae239ac8987e2e5fb2c0186"
+content-hash = "1be767f62c8a8e1c9acb76347055a837ea2c44130004703361d0303b049bede7"

llama_parse/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "llama-parse"
-version = "0.6.22"
+version = "0.6.43"
 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 = ">=0.6.22"
+llama-cloud-services = ">=0.6.43"
 
 [tool.poetry.group.dev.dependencies]
 pytest = "^8.0.0"

poetry.lock

@@ -1,4 +1,4 @@
-# This file is automatically @generated by Poetry 2.1.2 and should not be changed by hand.
+# This file is automatically @generated by Poetry 2.1.3 and should not be changed by hand.
 
 [[package]]
 name = "aiohappyeyeballs"
@@ -1925,14 +1925,14 @@ rapidfuzz = ">=3.9.0,<4.0.0"
 
 [[package]]
 name = "llama-cloud"
-version = "0.1.19"
+version = "0.1.32"
 description = ""
 optional = false
 python-versions = "<4,>=3.8"
 groups = ["main"]
 files = [
-    {file = "llama_cloud-0.1.19-py3-none-any.whl", hash = "sha256:d2d551baa4b63f7717f8e04cbb81b0f817e5450a66870c5487dd371f81dab8ec"},
-    {file = "llama_cloud-0.1.19.tar.gz", hash = "sha256:b0a5424ae0099ca27df2a2d7e5aec99066de9ca860ab65987c9f931f1ea7abff"},
+    {file = "llama_cloud-0.1.32-py3-none-any.whl", hash = "sha256:c42b2d5fb24acc8595bcc3626fb84c872909a16ab6d6879a1cb1101b21c238bd"},
+    {file = "llama_cloud-0.1.32.tar.gz", hash = "sha256:cea98241127311ea91f191c3c006aa6558f01d16f9539ed93b24d716b888f10e"},
 ]
 
 [package.dependencies]
@@ -4623,4 +4623,4 @@ type = ["pytest-mypy"]
 [metadata]
 lock-version = "2.1"
 python-versions = ">=3.9,<4.0"
-content-hash = "02cf83e2b3d328f05e17cfe994d324d30273458fded20bac745928cbce6dac32"
+content-hash = "112c1ccc4187a295dc07bebb14753f27f2e153dfdab895f763297f5391d3fb10"

pyproject.toml

@@ -8,7 +8,7 @@ python_version = "3.10"
 
 [tool.poetry]
 name = "llama-cloud-services"
-version = "0.6.22"
+version = "0.6.43"
 description = "Tailored SDK clients for LlamaCloud services."
 authors = ["Logan Markewich <logan@runllama.ai>"]
 license = "MIT"
@@ -18,12 +18,13 @@ packages = [{include = "llama_cloud_services"}]
 [tool.poetry.dependencies]
 python = ">=3.9,<4.0"
 llama-index-core = ">=0.12.0"
-llama-cloud = "==0.1.19"
+llama-cloud = "==0.1.32"
 pydantic = ">=2.8,!=2.10"
 click = "^8.1.7"
 python-dotenv = "^1.0.1"
 eval-type-backport = {python = "<3.10", version = "^0.2.0"}
 platformdirs = "^4.3.7"
+tenacity = ">=8.5.0, <10.0"
 
 [tool.poetry.group.dev.dependencies]
 pytest = "^8.0.0"

tests/extract/conftest.py

@@ -0,0 +1,41 @@
+import os
+from typing import List
+from llama_cloud_services.extract import LlamaExtract
+
+# Global storage for agents to cleanup
+_TEST_AGENTS_TO_CLEANUP: List[str] = []
+
+
+def pytest_sessionfinish(session, exitstatus):
+    """Hook that runs after all tests complete - cleanup agents here"""
+    print(
+        f"pytest_sessionfinish hook called! Agents to cleanup: {_TEST_AGENTS_TO_CLEANUP}"
+    )
+
+    if _TEST_AGENTS_TO_CLEANUP:
+        print("Creating cleanup client...")
+        # Create a fresh client just for cleanup
+        cleanup_client = LlamaExtract(
+            api_key=os.getenv("LLAMA_CLOUD_API_KEY"),
+            base_url=os.getenv("LLAMA_CLOUD_BASE_URL"),
+            project_id=os.getenv("LLAMA_CLOUD_PROJECT_ID"),
+            verbose=True,
+        )
+
+        for agent_id in _TEST_AGENTS_TO_CLEANUP:
+            try:
+                print(f"Deleting agent {agent_id}...")
+                cleanup_client.delete_agent(agent_id)
+                print(f"Cleaned up agent {agent_id}")
+            except Exception as e:
+                print(f"Warning: Failed to delete agent {agent_id}: {e}")
+
+        _TEST_AGENTS_TO_CLEANUP.clear()
+        print("Agent cleanup completed")
+    else:
+        print("No agents to cleanup")
+
+
+def register_agent_for_cleanup(agent_id: str):
+    """Register an agent ID for cleanup at the end of the test session"""
+    _TEST_AGENTS_TO_CLEANUP.append(agent_id)

tests/extract/test_extract_api.py

@@ -5,6 +5,7 @@ from pydantic import BaseModel
 
 from llama_cloud_services.extract import LlamaExtract, ExtractionAgent, SourceText
 from tests.extract.util import load_test_dotenv
+from .conftest import register_agent_for_cleanup
 
 load_test_dotenv()
 
@@ -27,7 +28,7 @@ class TestSchema(BaseModel):
 
 # Test data paths
 TEST_DIR = Path(__file__).parent / "data"
-TEST_PDF = TEST_DIR / "slide" / "saas_slide.pdf"
+TEST_PDF = TEST_DIR / "api_test" / "noisebridge_receipt.pdf"
 
 
 @pytest.fixture
@@ -58,7 +59,7 @@ def test_schema_dict():
 
 @pytest.fixture
 def test_agent(llama_extract, test_agent_name, test_schema_dict, request):
-    """Creates a test agent and cleans it up after the test"""
+    """Creates a test agent and collects it for cleanup at the end of all tests"""
     test_id = request.node.nodeid
     test_hash = hex(hash(test_id))[-8:]
     base_name = test_agent_name
@@ -86,13 +87,11 @@ def test_agent(llama_extract, test_agent_name, test_schema_dict, request):
         print(f"Warning: Failed to cleanup existing agent: {e}")
 
     agent = llama_extract.create_agent(name=name, data_schema=schema)
-    yield agent
 
-    # Cleanup after test
-    try:
-        llama_extract.delete_agent(agent.id)
-    except Exception as e:
-        print(f"Warning: Failed to delete agent {agent.id}: {e}")
+    # Add agent to cleanup list via conftest helper
+    register_agent_for_cleanup(agent.id)
+
+    yield agent
 
 
 class TestLlamaExtract:

tests/parse/test_llama_parse.py

@@ -1,6 +1,7 @@
 import os
 import pytest
 import shutil
+from typing import Optional, cast
 from fsspec.implementations.local import LocalFileSystem
 from httpx import AsyncClient
 
@@ -20,11 +21,15 @@ def test_simple_page_text() -> None:
     assert len(result[0].text) > 0
 
 
-@pytest.fixture
-def markdown_parser() -> LlamaParse:
+@pytest.fixture(params=[None, 2])
+def markdown_parser(request: pytest.FixtureRequest) -> LlamaParse:
     if os.environ.get("LLAMA_CLOUD_API_KEY", "") == "":
         pytest.skip("LLAMA_CLOUD_API_KEY not set")
-    return LlamaParse(result_type="markdown", ignore_errors=False)
+    return LlamaParse(
+        result_type="markdown",
+        ignore_errors=False,
+        partition_pages=cast(Optional[int], request.param),
+    )
 
 
 def test_simple_page_markdown(markdown_parser: LlamaParse) -> None:
@@ -35,8 +40,6 @@ def test_simple_page_markdown(markdown_parser: LlamaParse) -> None:
 
 
 def test_simple_page_markdown_bytes(markdown_parser: LlamaParse) -> None:
-    markdown_parser = LlamaParse(result_type="markdown", ignore_errors=False)
-
     filepath = "tests/test_files/attention_is_all_you_need.pdf"
     with open(filepath, "rb") as f:
         file_bytes = f.read()
@@ -51,8 +54,6 @@ def test_simple_page_markdown_bytes(markdown_parser: LlamaParse) -> None:
 
 
 def test_simple_page_markdown_buffer(markdown_parser: LlamaParse) -> None:
-    markdown_parser = LlamaParse(result_type="markdown", ignore_errors=False)
-
     filepath = "tests/test_files/attention_is_all_you_need.pdf"
     with open(filepath, "rb") as f:
         # client must provide extra_info with file_name
@@ -161,9 +162,12 @@ async def test_mixing_input_types() -> None:
     os.environ.get("LLAMA_CLOUD_API_KEY", "") == "",
     reason="LLAMA_CLOUD_API_KEY not set",
 )
+@pytest.mark.parametrize("partition_pages", [None, 2])
 @pytest.mark.asyncio
-async def test_download_images() -> None:
-    parser = LlamaParse(result_type="markdown", take_screenshot=True)
+async def test_download_images(partition_pages: Optional[int]) -> None:
+    parser = LlamaParse(
+        result_type="markdown", take_screenshot=True, partition_pages=partition_pages
+    )
     filepath = "tests/test_files/attention_is_all_you_need.pdf"
     json_result = await parser.aget_json([filepath])
 
@@ -175,3 +179,17 @@ async def test_download_images() -> None:
 
     await parser.aget_images(json_result, download_path)
     assert len(os.listdir(download_path)) == len(json_result[0]["pages"][0]["images"])
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize("split_by_page,expected", [(True, 4), (False, 1)])
+async def test_multiple_page_markdown(
+    markdown_parser: LlamaParse,
+    split_by_page: bool,
+    expected: int,
+) -> None:
+    markdown_parser.split_by_page = split_by_page
+    filepath = "tests/test_files/TOS.pdf"
+    result = await markdown_parser.aload_data(filepath)
+    assert len(result) == expected
+    assert all(len(doc.text) > 0 for doc in result)

tests/parse/test_llama_parse_result.py

@@ -1,6 +1,7 @@
 import tempfile
 import os
 import pytest
+from typing import Optional
 from llama_cloud_services import LlamaParse
 from llama_cloud_services.parse.types import JobResult
 
@@ -15,16 +16,23 @@ def chart_file_path() -> str:
     return "tests/test_files/attention_is_all_you_need_chart.pdf"
 
 
+@pytest.fixture
+def multiple_page_path() -> str:
+    return "tests/test_files/TOS.pdf"
+
+
 @pytest.mark.asyncio
 @pytest.mark.skipif(
     os.environ.get("LLAMA_CLOUD_API_KEY", "") == "",
     reason="LLAMA_CLOUD_API_KEY not set",
 )
-async def test_basic_parse_result(file_path: str):
+@pytest.mark.parametrize("partition_pages", [None, 2])
+async def test_basic_parse_result(file_path: str, partition_pages: Optional[int]):
     parser = LlamaParse(
         take_screenshot=True,
         auto_mode=True,
         fast_mode=False,
+        partition_pages=partition_pages,
     )
     result = await parser.aparse(file_path)
 
@@ -96,10 +104,12 @@ async def test_link_parse_result(file_path: str):
     os.environ.get("LLAMA_CLOUD_API_KEY", "") == "",
     reason="LLAMA_CLOUD_API_KEY not set",
 )
+@pytest.mark.skip(reason="TODO: Needs to be fixed in prod. Raising 500 error.")
 async def test_parse_structured_output(file_path: str):
     parser = LlamaParse(
         structured_output=True,
         structured_output_json_schema_name="imFeelingLucky",
+        invalidate_cache=True,
     )
     result = await parser.aparse(file_path)
     assert isinstance(result, JobResult)
@@ -142,8 +152,11 @@ async def test_parse_layout(file_path: str):
     os.environ.get("LLAMA_CLOUD_API_KEY", "") == "",
     reason="LLAMA_CLOUD_API_KEY not set",
 )
-def test_parse_multiple_files(file_path: str, chart_file_path: str):
-    parser = LlamaParse()
+@pytest.mark.parametrize("partition_pages", [None, 2])
+def test_parse_multiple_files(
+    file_path: str, chart_file_path: str, partition_pages: Optional[int]
+):
+    parser = LlamaParse(partition_pages=partition_pages)
     result = parser.parse([file_path, chart_file_path])
 
     assert isinstance(result, list)
@@ -152,3 +165,40 @@ def test_parse_multiple_files(file_path: str, chart_file_path: str):
     assert isinstance(result[1], JobResult)
     assert result[0].file_name == file_path
     assert result[1].file_name == chart_file_path
+
+
+@pytest.mark.asyncio
+@pytest.mark.skipif(
+    os.environ.get("LLAMA_CLOUD_API_KEY", "") == "",
+    reason="LLAMA_CLOUD_API_KEY not set",
+)
+@pytest.mark.parametrize("partition_pages", [None, 2])
+async def test_multiple_page_parse_result(
+    multiple_page_path: str, partition_pages: Optional[int]
+):
+    parser = LlamaParse(
+        take_screenshot=True,
+        auto_mode=True,
+        fast_mode=False,
+        partition_pages=partition_pages,
+    )
+    results = await parser.aparse(multiple_page_path)
+    if partition_pages is None:
+        assert isinstance(results, JobResult)
+        results = [results]
+    else:
+        assert isinstance(results, list)
+
+    for result in results:
+        assert isinstance(result, JobResult)
+        assert result.job_id is not None
+        assert result.file_name == multiple_page_path
+        assert len(result.pages) > 0
+
+        assert result.pages[0].text is not None
+        assert len(result.pages[0].text) > 0
+
+        assert result.pages[0].md is not None
+        assert len(result.pages[0].md) > 0
+
+        assert result.pages[0].md != result.pages[0].text

tests/parse/test_utils.py

@@ -0,0 +1,30 @@
+import pytest
+
+
+from llama_cloud_services.parse.utils import expand_target_pages, partition_pages
+
+
+def test_expand_target_pages() -> None:
+    with pytest.raises(ValueError):
+        list(expand_target_pages("x"))
+    with pytest.raises(ValueError):
+        list(expand_target_pages("1-2-3"))
+    with pytest.raises(ValueError):
+        list(expand_target_pages("2-1"))
+    result = list(expand_target_pages("0,2-3,5,8-10"))
+    assert result == [0, 2, 3, 5, 8, 9, 10]
+
+
+def test_partion_pages() -> None:
+    pages = [0, 2, 3, 5, 8, 9, 10]
+    with pytest.raises(ValueError):
+        list(partition_pages(pages, 0))
+    result = list(partition_pages(pages, 3))
+    assert result == ["0,2-3", "5,8-9", "10"]
+
+    with pytest.raises(ValueError):
+        list(partition_pages(pages, 3, 0))
+    result = list(partition_pages(pages, 3, max_pages=5))
+    assert result == ["0,2-3", "5,8"]
+    result = list(partition_pages(pages, 3, max_pages=10))
+    assert result == ["0,2-3", "5,8-9", "10"]

tests/test_utils.py

@@ -0,0 +1,66 @@
+from pydantic import BaseModel
+from llama_cloud_services.utils import check_extra_params
+
+
+class MyModel(BaseModel):
+    name: str
+    age: int
+    email: str
+    is_active: bool
+
+
+def test_check_extra_params_no_extra():
+    """Test when all parameters are valid - should return empty lists."""
+    data = {"name": "John", "age": 25, "email": "john@example.com", "is_active": True}
+
+    extra_params, suggestions = check_extra_params(MyModel, data)
+
+    assert extra_params == []
+    assert suggestions == []
+
+
+def test_check_extra_params_with_typos():
+    """Test when there are extra parameters that are close to valid ones (typos)."""
+    data = {
+        "name": "John",
+        "age": 25,
+        "emial": "john@example.com",  # typo: emial instead of email
+        "is_activ": True,  # typo: is_activ instead of is_active
+        "address": "123 Main St",  # completely different parameter
+    }
+
+    extra_params, suggestions = check_extra_params(MyModel, data)
+
+    assert len(extra_params) == 3
+    assert "emial" in extra_params
+    assert "is_activ" in extra_params
+    assert "address" in extra_params
+
+    # Check that typo suggestions are provided
+    assert len(suggestions) == 3
+    assert "Did you mean 'email' instead of 'emial'?" in suggestions[0]
+    assert "Did you mean 'is_active' instead of 'is_activ'?" in suggestions[1]
+    assert "check the documentation or update the package" in suggestions[2]
+
+
+def test_check_extra_params_completely_invalid():
+    """Test when there are extra parameters with no close matches."""
+    data = {
+        "name": "John",
+        "xyz": "invalid",
+        "random_field": 123,
+        "completely_different": True,
+    }
+
+    extra_params, suggestions = check_extra_params(MyModel, data)
+
+    assert len(extra_params) == 3
+    assert "xyz" in extra_params
+    assert "random_field" in extra_params
+    assert "completely_different" in extra_params
+
+    # All suggestions should be generic (no close matches)
+    assert len(suggestions) == 3
+    for suggestion in suggestions:
+        assert "check the documentation or update the package" in suggestion
+        assert "Did you mean" not in suggestion