html-to-markdown 2.14.0__cp310-abi3-macosx_11_0_arm64.whl → 2.14.5__cp310-abi3-macosx_11_0_arm64.whl

html_to_markdown/__init__.py

@@ -15,16 +15,16 @@ V1 API (backward compatibility):
     markdown = convert_to_markdown(html, heading_style="atx")
 """
 
-import contextlib
-
 from html_to_markdown.api import (
     InlineImage,
     InlineImageConfig,
     InlineImageWarning,
+    MetadataConfig,
     OptionsHandle,
     convert,
     convert_with_handle,
     convert_with_inline_images,
+    convert_with_metadata,
     create_options_handle,
 )
 from html_to_markdown.exceptions import (
@@ -37,10 +37,6 @@ from html_to_markdown.exceptions import (
 from html_to_markdown.options import ConversionOptions, PreprocessingOptions
 from html_to_markdown.v1_compat import convert_to_markdown, markdownify
 
-# Optional metadata support
-with contextlib.suppress(ImportError):
-    from html_to_markdown.api import MetadataConfig, convert_with_metadata
-
 __all__ = [
     "ConflictingOptionsError",
     "ConversionOptions",
@@ -63,4 +59,4 @@ __all__ = [
     "markdownify",
 ]
 
-__version__ = "2.14.0"
+__version__ = "2.14.5"

html_to_markdown/_html_to_markdown.pyi

@@ -141,13 +141,13 @@ class DocumentMetadata(TypedDict):
     canonical_url: str | None
     base_href: str | None
     language: str | None
-    text_direction: str | None  # "ltr" | "rtl" | "auto" | None
+    text_direction: str | None
     open_graph: dict[str, str]
     twitter_card: dict[str, str]
     meta_tags: dict[str, str]
 
 class HeaderMetadata(TypedDict):
-    level: int  # 1-6
+    level: int
     text: str
     id: str | None
     depth: int
@@ -157,7 +157,7 @@ class LinkMetadata(TypedDict):
     href: str
     text: str
     title: str | None
-    link_type: str  # "anchor" | "internal" | "external" | "email" | "phone" | "other"
+    link_type: str
     rel: list[str]
     attributes: dict[str, str]
 
@@ -165,12 +165,12 @@ class ImageMetadata(TypedDict):
     src: str
     alt: str | None
     title: str | None
-    dimensions: tuple[int, int] | None  # (width, height)
-    image_type: str  # "data_uri" | "inline_svg" | "external" | "relative"
+    dimensions: tuple[int, int] | None
+    image_type: str
     attributes: dict[str, str]
 
 class StructuredData(TypedDict):
-    data_type: str  # "json_ld" | "microdata" | "rdfa"
+    data_type: str
     raw_json: str
     schema_type: str | None
 

html_to_markdown/api.py

@@ -8,20 +8,16 @@ import html_to_markdown._html_to_markdown as _rust
 from html_to_markdown._html_to_markdown import (
     ConversionOptionsHandle as OptionsHandle,
 )
-from html_to_markdown._html_to_markdown import InlineImageConfig
+from html_to_markdown._html_to_markdown import (
+    InlineImageConfig,
+    MetadataConfig,
+)
 from html_to_markdown.options import ConversionOptions, PreprocessingOptions
 
-_HAS_METADATA = False
-try:
-    from html_to_markdown._html_to_markdown import ExtendedMetadata, MetadataConfig
-
-    _HAS_METADATA = True
-except ImportError:
-    MetadataConfig = None  # type: ignore[misc,assignment]
-    if TYPE_CHECKING:
-        from html_to_markdown._html_to_markdown import ExtendedMetadata  # pragma: no cover
-    else:
-        ExtendedMetadata = dict[str, object]  # type: ignore[assignment]
+if TYPE_CHECKING:
+    from html_to_markdown._html_to_markdown import ExtendedMetadata  # pragma: no cover
+else:
+    ExtendedMetadata = dict[str, object]  # type: ignore[assignment]
 
 
 class InlineImage(TypedDict):
@@ -146,40 +142,43 @@ def convert_with_handle(html: str, handle: OptionsHandle) -> str:
     return _rust.convert_with_options_handle(html, handle)
 
 
-if _HAS_METADATA:
-
-    def convert_with_metadata(
-        html: str,
-        options: ConversionOptions | None = None,
-        preprocessing: PreprocessingOptions | None = None,
-        metadata_config: MetadataConfig | None = None,
-    ) -> tuple[str, ExtendedMetadata]:
-        """Convert HTML and extract comprehensive metadata.
-
-        Args:
-            html: HTML string to convert
-            options: Optional conversion configuration
-            preprocessing: Optional preprocessing configuration
-            metadata_config: Optional metadata extraction configuration
-
-        Returns:
-            Tuple of (markdown, metadata_dict) where metadata_dict contains:
-            - document: Document-level metadata (title, description, lang, etc.)
-            - headers: List of header elements with hierarchy
-            - links: List of extracted hyperlinks with classification
-            - images: List of extracted images with metadata
-            - structured_data: List of JSON-LD, Microdata, or RDFa blocks
-        """
-        if options is None:
-            options = ConversionOptions()
-        if preprocessing is None:
-            preprocessing = PreprocessingOptions()
-        if metadata_config is None:
-            metadata_config = MetadataConfig()
-
-        rust_options = _to_rust_options(options, preprocessing)
-        markdown, metadata = _rust.convert_with_metadata(html, rust_options, metadata_config)
-        return markdown, metadata
+def convert_with_metadata(
+    html: str,
+    options: ConversionOptions | None = None,
+    preprocessing: PreprocessingOptions | None = None,
+    metadata_config: MetadataConfig | None = None,
+) -> tuple[str, ExtendedMetadata]:
+    """Convert HTML and extract comprehensive metadata.
+
+    Args:
+        html: HTML string to convert
+        options: Optional conversion configuration
+        preprocessing: Optional preprocessing configuration
+        metadata_config: Optional metadata extraction configuration
+
+    Returns:
+        Tuple of (markdown, metadata_dict) where metadata_dict contains:
+        - document: Document-level metadata (title, description, lang, etc.)
+        - headers: List of header elements with hierarchy
+        - links: List of extracted hyperlinks with classification
+        - images: List of extracted images with metadata
+        - structured_data: List of JSON-LD, Microdata, or RDFa blocks
+    """
+    if not hasattr(_rust, "convert_with_metadata"):
+        raise ImportError(
+            "convert_with_metadata is missing from the native extension; this indicates a broken/partial installation."
+        )
+
+    if options is None:
+        options = ConversionOptions()
+    if preprocessing is None:
+        preprocessing = PreprocessingOptions()
+    if metadata_config is None:
+        metadata_config = MetadataConfig()
+
+    rust_options = _to_rust_options(options, preprocessing)
+    markdown, metadata = _rust.convert_with_metadata(html, rust_options, metadata_config)
+    return markdown, metadata
 
 
 __all__ = [

html_to_markdown/v1_compat.py

@@ -136,7 +136,6 @@ def convert_to_markdown(
         )
 
     # ~keep: v1 used indented code blocks by default, but switched to backticks when a language was set
-    # This maintains v1 behavior for backward compatibility
     code_block_style = "backticks" if code_language else "indented"
 
     options = ConversionOptions(

{html_to_markdown-2.14.0.dist-info → html_to_markdown-2.14.5.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: html-to-markdown
-Version: 2.14.0
+Version: 2.14.5
 Classifier: Development Status :: 5 - Production/Stable
 Classifier: Environment :: Console
 Classifier: Intended Audience :: Developers
@@ -44,7 +44,7 @@ High-performance HTML to Markdown converter with a clean Python API (powered by
 [![Hex.pm](https://img.shields.io/hexpm/v/html_to_markdown.svg)](https://hex.pm/packages/html_to_markdown)
 [![NuGet](https://img.shields.io/nuget/v/Goldziher.HtmlToMarkdown.svg)](https://www.nuget.org/packages/Goldziher.HtmlToMarkdown/)
 [![Maven Central](https://img.shields.io/maven-central/v/io.github.goldziher/html-to-markdown.svg)](https://central.sonatype.com/artifact/io.github.goldziher/html-to-markdown)
-[![Go Reference](https://pkg.go.dev/badge/github.com/Goldziher/html-to-markdown/packages/go/htmltomarkdown.svg)](https://pkg.go.dev/github.com/Goldziher/html-to-markdown/packages/go/htmltomarkdown)
+[![Go Reference](https://pkg.go.dev/badge/github.com/Goldziher/html-to-markdown/packages/go/v2/htmltomarkdown.svg)](https://pkg.go.dev/github.com/Goldziher/html-to-markdown/packages/go/v2/htmltomarkdown)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://github.com/Goldziher/html-to-markdown/blob/main/LICENSE)
 [![Discord](https://img.shields.io/badge/Discord-Join%20our%20community-7289da)](https://discord.gg/pXxagNK2zN)
 

html_to_markdown-2.14.5.dist-info/RECORD

@@ -0,0 +1,17 @@
+html_to_markdown-2.14.5.data/scripts/html-to-markdown,sha256=pZtJBTQolbGl_VL3mBo8mLIppGODYhfcq6yOZVldWOg,6263872
+html_to_markdown-2.14.5.dist-info/RECORD,,
+html_to_markdown-2.14.5.dist-info/WHEEL,sha256=WvP__evn8XoyZeDO32cKBm5BQTOFbdB1WoQ-d3AzYdw,132
+html_to_markdown-2.14.5.dist-info/METADATA,sha256=_m5URCaMwleUAZpC712kPn5LKdqwwZesMTzlbm0F1MQ,23252
+html_to_markdown-2.14.5.dist-info/licenses/LICENSE,sha256=oQvPC-0UWvfg0WaeUBe11OJMtX60An-TW1ev_oaAA0k,1086
+html_to_markdown/options.py,sha256=vImRfeHAeyAy0Lnt6cTPHGbj7mTdw8AEUgo19u7MAA0,5080
+html_to_markdown/_html_to_markdown.pyi,sha256=IPD6CegtaanBsKTmK30v4nvWZ5HUlCajS6jkiOsoVj8,5875
+html_to_markdown/_html_to_markdown.abi3.so,sha256=hpn9BCm1Z-79e5NsXTxDjhtloCGWp1I_2aEwRy9-KTI,3503168
+html_to_markdown/__init__.py,sha256=7FQJWsnvd_XVwzVEM7rEKI1XKfqIB81oRnV44qUTJeM,1605
+html_to_markdown/api.py,sha256=MsTij04ij6hFhhNxdc5RXf2yobaRmB0BO1P_fjS4VvY,6806
+html_to_markdown/v1_compat.py,sha256=kn5GYvgn3dTW_Zksu9PzWVk-5CYhvXxsqAeyTdDYZSY,8001
+html_to_markdown/cli.py,sha256=Rn-s3FZPea1jgCJtDzH_TFvOEiA_uZFVfgjhr6xyL_g,64
+html_to_markdown/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+html_to_markdown/exceptions.py,sha256=aTASOzbywgfqOYjlw18ZkOWSxKff4EbUbmMua_73TGA,2370
+html_to_markdown/cli_proxy.py,sha256=HPYKH5Mf5OUvkbEQISJvAkxrbjWKxE5GokA44HoQ6z8,3858
+html_to_markdown/__main__.py,sha256=3Ic_EbOt2h6W88q084pkz5IKU6iY5z_woBygH6u9aw0,327
+html_to_markdown/bin/html-to-markdown,sha256=pZtJBTQolbGl_VL3mBo8mLIppGODYhfcq6yOZVldWOg,6263872

html_to_markdown-2.14.0.dist-info/RECORD

@@ -1,17 +0,0 @@
-html_to_markdown-2.14.0.dist-info/RECORD,,
-html_to_markdown-2.14.0.dist-info/WHEEL,sha256=WvP__evn8XoyZeDO32cKBm5BQTOFbdB1WoQ-d3AzYdw,132
-html_to_markdown-2.14.0.dist-info/METADATA,sha256=JG4gq6AeDUutrpmm9yBndbm3rGNfg5TpQH6SpOsuTw4,23246
-html_to_markdown-2.14.0.dist-info/licenses/LICENSE,sha256=oQvPC-0UWvfg0WaeUBe11OJMtX60An-TW1ev_oaAA0k,1086
-html_to_markdown-2.14.0.data/scripts/html-to-markdown,sha256=IuASvb2-OF95pz52ODWCERrDGRIBc4bWjzETXGcP_3w,6263856
-html_to_markdown/options.py,sha256=vImRfeHAeyAy0Lnt6cTPHGbj7mTdw8AEUgo19u7MAA0,5080
-html_to_markdown/_html_to_markdown.pyi,sha256=C4Y80U97JpQhjWZxpDb2Lu2ojbFwAZLtygchVdmhYs0,6093
-html_to_markdown/_html_to_markdown.abi3.so,sha256=jIVAAeoTyJkoHiTwjskEbKnJxvWuDDQ2HMvzs2qZMzs,3486656
-html_to_markdown/__init__.py,sha256=NuQiHxunqCAfbGRYXjgbp31YBXoLlYjXU_Fko4GC5oM,1720
-html_to_markdown/api.py,sha256=zXXoFpdDbMIQXl65NT7BjjYu_1xwEM7VNGNUK2zQNfQ,6934
-html_to_markdown/v1_compat.py,sha256=nZN8hVd3u4vacbfyLsPMIGqSmJENZgx1Ya0SpqVLi-g,8061
-html_to_markdown/cli.py,sha256=Rn-s3FZPea1jgCJtDzH_TFvOEiA_uZFVfgjhr6xyL_g,64
-html_to_markdown/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-html_to_markdown/exceptions.py,sha256=aTASOzbywgfqOYjlw18ZkOWSxKff4EbUbmMua_73TGA,2370
-html_to_markdown/cli_proxy.py,sha256=HPYKH5Mf5OUvkbEQISJvAkxrbjWKxE5GokA44HoQ6z8,3858
-html_to_markdown/__main__.py,sha256=3Ic_EbOt2h6W88q084pkz5IKU6iY5z_woBygH6u9aw0,327
-html_to_markdown/bin/html-to-markdown,sha256=IuASvb2-OF95pz52ODWCERrDGRIBc4bWjzETXGcP_3w,6263856