feat(parse): support WARC archives in parsing and scraping (#2761)

Stainless-Generated-From: 1093f815e6b080ff11e05431a6bef3055ae7c0fd

Author: stlc-bot

src/context/dev/resources/parse.py

@@ -63,6 +63,8 @@ def handle(
         *,
         client: str | Omit = omit,
         extension: Literal[
+            "warc",
+            "warc.gz",
             "txt",
             "text",
             "md",
@@ -151,14 +153,18 @@ def handle(
         extra_body: Body | None = None,
         timeout: float | httpx.Timeout | None | NotGiven = not_given,
     ) -> ParseHandleResponse:
-        """
-        Convert uploaded file bytes into Markdown and optional HTML.
+        """Convert uploaded file bytes into Markdown and optional HTML.
+
+        WARC and WARC.GZ
+        archives combine HTTP response, resource, and conversion records in archive
+        order. Revisit and segmented records are skipped. Decoded WARC content is
+        limited to 50 MiB.
 
         Args:
           client: Optional client identifier used for usage attribution.
 
           extension: Optional file extension hint, such as pdf, docx, xlsx, pptx, html, json, csv,
-              md, py, rtf, jpg, png, or txt.
+              md, py, rtf, jpg, png, txt, warc, or warc.gz.
 
           include_images: Include image references in Markdown output
 
@@ -244,6 +250,8 @@ async def handle(
         *,
         client: str | Omit = omit,
         extension: Literal[
+            "warc",
+            "warc.gz",
             "txt",
             "text",
             "md",
@@ -332,14 +340,18 @@ async def handle(
         extra_body: Body | None = None,
         timeout: float | httpx.Timeout | None | NotGiven = not_given,
     ) -> ParseHandleResponse:
-        """
-        Convert uploaded file bytes into Markdown and optional HTML.
+        """Convert uploaded file bytes into Markdown and optional HTML.
+
+        WARC and WARC.GZ
+        archives combine HTTP response, resource, and conversion records in archive
+        order. Revisit and segmented records are skipped. Decoded WARC content is
+        limited to 50 MiB.
 
         Args:
           client: Optional client identifier used for usage attribution.
 
           extension: Optional file extension hint, such as pdf, docx, xlsx, pptx, html, json, csv,
-              md, py, rtf, jpg, png, or txt.
+              md, py, rtf, jpg, png, txt, warc, or warc.gz.
 
           include_images: Include image references in Markdown output
 

src/context/dev/types/parse_handle_params.py

@@ -15,6 +15,8 @@ class ParseHandleParams(TypedDict, total=False):
     """Optional client identifier used for usage attribution."""
 
     extension: Literal[
+        "warc",
+        "warc.gz",
         "txt",
         "text",
         "md",
@@ -88,7 +90,7 @@ class ParseHandleParams(TypedDict, total=False):
     ]
     """
     Optional file extension hint, such as pdf, docx, xlsx, pptx, html, json, csv,
-    md, py, rtf, jpg, png, or txt.
+    md, py, rtf, jpg, png, txt, warc, or warc.gz.
     """
 
     include_images: Annotated[bool, PropertyInfo(alias="includeImages")]

src/context/dev/types/parse_handle_response.py

@@ -32,6 +32,7 @@ class ParseHandleResponse(BaseModel):
     """Indicates success"""
 
     type: Literal[
+        "warc",
         "html",
         "xml",
         "json",

tests/api_resources/test_parse.py

@@ -31,7 +31,7 @@ def test_method_handle_with_all_params(self, client: ContextDev) -> None:
         parse = client.parse.handle(
             body=b"Example data",
             client="x",
-            extension="txt",
+            extension="warc",
             include_images=True,
             include_links=True,
             include_numbered_pages=True,
@@ -93,7 +93,7 @@ async def test_method_handle_with_all_params(self, async_client: AsyncContextDev
         parse = await async_client.parse.handle(
             body=b"Example data",
             client="x",
-            extension="txt",
+            extension="warc",
             include_images=True,
             include_links=True,
             include_numbered_pages=True,