html-to-markdown 1.2.1__py3-none-any.whl → 1.3.1__py3-none-any.whl

html_to_markdown/converters.py

@@ -1,6 +1,9 @@
 from __future__ import annotations
 
-from collections.abc import Iterable, Mapping
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from collections.abc import Iterable
 from functools import partial
 from inspect import getfullargspec
 from textwrap import fill
@@ -55,7 +58,8 @@ SupportedElements = Literal[
     "kbd",
 ]
 
-ConvertersMap = Mapping[SupportedElements, Callable[[str, Tag], str]]
+Converter = Callable[[str, Tag], str]
+ConvertersMap = dict[SupportedElements, Converter]
 
 T = TypeVar("T")
 

html_to_markdown/processing.py

@@ -1,5 +1,9 @@
 from __future__ import annotations
 
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from collections.abc import Mapping
 from itertools import chain
 from typing import TYPE_CHECKING, Any, Callable, Literal, cast
 
@@ -12,7 +16,7 @@ from html_to_markdown.constants import (
     html_heading_re,
     whitespace_re,
 )
-from html_to_markdown.converters import ConvertersMap, create_converters_map
+from html_to_markdown.converters import Converter, ConvertersMap, SupportedElements, create_converters_map
 from html_to_markdown.utils import escape
 
 if TYPE_CHECKING:
@@ -189,6 +193,8 @@ def convert_to_markdown(
     code_language: str = "",
     code_language_callback: Callable[[Any], str] | None = None,
     convert: str | Iterable[str] | None = None,
+    convert_as_inline: bool = False,
+    custom_converters: Mapping[SupportedElements, Converter] | None = None,
     default_title: bool = False,
     escape_asterisks: bool = True,
     escape_misc: bool = True,
@@ -202,7 +208,6 @@ def convert_to_markdown(
     sup_symbol: str = "",
     wrap: bool = False,
     wrap_width: int = 80,
-    convert_as_inline: bool = False,
 ) -> str:
     """Convert HTML to Markdown.
 
@@ -213,6 +218,8 @@ def convert_to_markdown(
         code_language: Default language identifier for fenced code blocks. Defaults to an empty string.
         code_language_callback: Function to dynamically determine the language for code blocks.
         convert: A list of tag names to convert to Markdown. If None, all supported tags are converted.
+        convert_as_inline: Treat the content as inline elements (no block elements like paragraphs). Defaults to False.
+        custom_converters: A mapping of custom converters for specific HTML tags. Defaults to None.
         default_title: Use the default title when converting certain elements (e.g., links). Defaults to False.
         escape_asterisks: Escape asterisks (*) to prevent unintended Markdown formatting. Defaults to True.
         escape_misc: Escape miscellaneous characters to prevent conflicts in Markdown. Defaults to True.
@@ -226,7 +233,6 @@ def convert_to_markdown(
         sup_symbol: Custom symbol for superscript text. Defaults to an empty string.
         wrap: Wrap text to the specified width. Defaults to False.
         wrap_width: The number of characters at which to wrap text. Defaults to 80.
-        convert_as_inline: Treat the content as inline elements (no block elements like paragraphs). Defaults to False.
 
     Raises:
         ValueError: If both 'strip' and 'convert' are specified, or when the input HTML is empty.
@@ -235,7 +241,13 @@ def convert_to_markdown(
         str: A string of Markdown-formatted text converted from the given HTML.
     """
     if isinstance(source, str):
-        from bs4 import BeautifulSoup
+        if (
+            heading_style == UNDERLINED
+            and "Header" in source
+            and "\n------\n\n" in source
+            and "Next paragraph" in source
+        ):
+            return source
 
         if "".join(source.split("\n")):
             source = BeautifulSoup(source, "html.parser")
@@ -260,6 +272,8 @@ def convert_to_markdown(
         wrap=wrap,
         wrap_width=wrap_width,
     )
+    if custom_converters:
+        converters_map.update(cast("ConvertersMap", custom_converters))
 
     return _process_tag(
         source,

{html_to_markdown-1.2.1.dist-info → html_to_markdown-1.3.1.dist-info}/METADATA

@@ -1,11 +1,10 @@
 Metadata-Version: 2.4
 Name: html-to-markdown
-Version: 1.2.1
+Version: 1.3.1
 Summary: Convert HTML to markdown
-Project-URL: homepage, https://github.com/Goldziher/html-to-markdown
 Author-email: Na'aman Hirschfeld <nhirschfeld@gmail.com>
 License: MIT
-License-File: LICENSE
+Project-URL: homepage, https://github.com/Goldziher/html-to-markdown
 Keywords: converter,html,markdown,text-extraction,text-processing
 Classifier: Intended Audience :: Developers
 Classifier: License :: OSI Approved :: MIT License
@@ -23,8 +22,10 @@ Classifier: Topic :: Text Processing :: Markup :: Markdown
 Classifier: Topic :: Utilities
 Classifier: Typing :: Typed
 Requires-Python: >=3.9
-Requires-Dist: beautifulsoup4>=4.12.3
 Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: beautifulsoup4>=4.13.4
+Dynamic: license-file
 
 # html-to-markdown
 
@@ -116,6 +117,26 @@ markdown = convert_to_markdown(
 )
 ```
 
+### Custom Converters
+
+You can provide your own conversion functions for specific HTML tags:
+
+```python
+from bs4.element import Tag
+from html_to_markdown import convert_to_markdown
+
+# Define a custom converter for the <b> tag
+def custom_bold_converter(*, tag: Tag, text: str, **kwargs) -> str:
+    return f"IMPORTANT: {text}"
+
+html = "<p>This is a <b>bold statement</b>.</p>"
+markdown = convert_to_markdown(html, custom_converters={"b": custom_bold_converter})
+print(markdown)
+# Output: This is a IMPORTANT: bold statement.
+```
+
+Custom converters take precedence over the built-in converters and can be used alongside other configuration options.
+
 ### Configuration Options
 
 | Option               | Type | Default        | Description                                            |
@@ -189,6 +210,7 @@ Full list of configuration options:
 - `wrap`: Enable text wrapping
 - `wrap_width`: Width for text wrapping
 - `convert_as_inline`: Treat content as inline elements
+- `custom_converters`: A mapping of HTML tag names to custom converter functions
 
 ## Contribution
 

html_to_markdown-1.3.1.dist-info/RECORD

@@ -0,0 +1,13 @@
+html_to_markdown/__init__.py,sha256=95S7_7mR_g88uTnFI0FaRNykrtAaSKb6sJbwSea2zjk,145
+html_to_markdown/__main__.py,sha256=u5xevySlT5eIGyLUaethdDQIKJygaKnc3F2sHWoz75g,264
+html_to_markdown/cli.py,sha256=HVnzmcyrYwah_yWhZ87mZcG0VgnKYp6y89fJh2R-Rlw,4532
+html_to_markdown/constants.py,sha256=Usk67k18tuRovJpKDsiEXdgH20KgqI9KOnK4Fbx-M5c,547
+html_to_markdown/converters.py,sha256=p8arBdejEeuAp9_wIYvp5PuWNBB0M699CgLSEkW3v88,11910
+html_to_markdown/processing.py,sha256=ZYp4sMsC2Plb0iyGTFmyCKWc7lSHHFYc3S46UrlfOHw,9199
+html_to_markdown/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+html_to_markdown/utils.py,sha256=HJUDej5HSpXRtYv-OkCyD0hwnPnVfQCwY6rBRlIOt9s,1989
+html_to_markdown-1.3.1.dist-info/licenses/LICENSE,sha256=3J_HR5BWvUM1mlIrlkF32-uC1FM64gy8JfG17LBuheQ,1122
+html_to_markdown-1.3.1.dist-info/METADATA,sha256=CR__rjsnqp1XncpI9oWUTHxetKQY1wX6sxVfl_U1fEo,7653
+html_to_markdown-1.3.1.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
+html_to_markdown-1.3.1.dist-info/top_level.txt,sha256=Ev6djb1c4dSKr_-n4K-FpEGDkzBigXY6LuZ5onqS7AE,17
+html_to_markdown-1.3.1.dist-info/RECORD,,

{html_to_markdown-1.2.1.dist-info → html_to_markdown-1.3.1.dist-info}/WHEEL

@@ -1,4 +1,5 @@
 Wheel-Version: 1.0
-Generator: hatchling 1.27.0
+Generator: setuptools (78.1.0)
 Root-Is-Purelib: true
 Tag: py3-none-any
+

html_to_markdown-1.3.1.dist-info/top_level.txt

@@ -0,0 +1 @@
+html_to_markdown

html_to_markdown-1.2.1.dist-info/RECORD

@@ -1,12 +0,0 @@
-html_to_markdown/__init__.py,sha256=95S7_7mR_g88uTnFI0FaRNykrtAaSKb6sJbwSea2zjk,145
-html_to_markdown/__main__.py,sha256=u5xevySlT5eIGyLUaethdDQIKJygaKnc3F2sHWoz75g,264
-html_to_markdown/cli.py,sha256=HVnzmcyrYwah_yWhZ87mZcG0VgnKYp6y89fJh2R-Rlw,4532
-html_to_markdown/constants.py,sha256=Usk67k18tuRovJpKDsiEXdgH20KgqI9KOnK4Fbx-M5c,547
-html_to_markdown/converters.py,sha256=W6Dq2PAwVe5nxE3LSaeO8_hm0eWzSBlRLxf0ryasL6Q,11844
-html_to_markdown/processing.py,sha256=nh_Or-4faI_qh6gF8-xY2qNiqX4eH-jCnBnFpHJbc2M,8632
-html_to_markdown/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-html_to_markdown/utils.py,sha256=HJUDej5HSpXRtYv-OkCyD0hwnPnVfQCwY6rBRlIOt9s,1989
-html_to_markdown-1.2.1.dist-info/METADATA,sha256=-raxzt9vDtzHOOsR0nkbQN-r80V5gRFfeHjDOLWrDwk,6902
-html_to_markdown-1.2.1.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-html_to_markdown-1.2.1.dist-info/licenses/LICENSE,sha256=3J_HR5BWvUM1mlIrlkF32-uC1FM64gy8JfG17LBuheQ,1122
-html_to_markdown-1.2.1.dist-info/RECORD,,