html-table-parse 0.2__py3-none-any.whl → 0.2.2__py3-none-any.whl

html_table_parse/__init__.py

@@ -1,4 +1,4 @@
-__version__ = '0.2'
+__version__ = '0.2.2'
 
 from collections import defaultdict as _defaultdict
 from functools import partial as _partial

html_table_parse-0.2.2.dist-info/METADATA

@@ -0,0 +1,71 @@
+Metadata-Version: 2.3
+Name: html-table-parse
+Version: 0.2.2
+Summary: Parse HTML table as Python list or dict
+Author: 5j9
+Author-email: 5j9 <5j9@users.noreply.github.com>
+License: GPL-3.0
+Requires-Dist: lxml>=6.1.1
+Requires-Python: >=3.10
+Project-URL: Homepage, https://github.com/5j9/html-table-parse
+Description-Content-Type: text/markdown
+
+# HTML Table Parse
+
+A lightweight HTML table parser that converts tables to Python data structures without pandas.
+
+## Installation
+
+```bash
+pip install html-table-parse
+```
+
+## Usage
+
+```python
+from html_table_parse import to_list, to_dict, to_dicts
+
+html = """
+<table>
+    <tr><th>Name</th><th>Age</th><th>City</th></tr>
+    <tr><td>Alice</td><td>30</td><td>NYC</td></tr>
+    <tr><td>Bob</td><td>25</td><td>LA</td></tr>
+</table>
+"""
+
+# List of lists
+to_list(html)
+# [['Name', 'Age', 'City'], ['Alice', '30', 'NYC'], ['Bob', '25', 'LA']]
+
+# Dictionary of columns
+to_dict(html)
+# {'Name': ['Alice', 'Bob'], 'Age': ['30', '25'], 'City': ['NYC', 'LA']}
+
+# List of dictionaries
+to_dicts(html)
+# [{'Name': 'Alice', 'Age': '30', 'City': 'NYC'}, 
+#  {'Name': 'Bob', 'Age': '25', 'City': 'LA'}]
+```
+
+## Features
+
+- No pandas required - lightweight alternative to `pandas.read_html()`
+- Supports `colspan` and `rowspan` attributes
+- Handles duplicate headers (auto-numbered)
+- Multiple output formats: lists, dict of columns, or list of dicts
+- Automatic whitespace normalization
+- Fast parsing with ```lxml```
+
+## API
+
+### `to_list(html: str, index: int = 0) -> list[list]`
+
+Parse table as list of rows.
+
+### `to_dict(html: str, index: int = 0) -> dict[str, list]`
+
+Parse table as dictionary of columns (first row = headers).
+
+### `to_dicts(html: str, index: int = 0) -> list[dict]`
+
+Parse table as list of dictionaries (first row = headers).

html_table_parse-0.2.2.dist-info/RECORD

@@ -0,0 +1,4 @@
+html_table_parse/__init__.py,sha256=ucjFk6-e7MpqRw_MTbaSIBJWkfOsfAJq7mfixAd350E,4944
+html_table_parse-0.2.2.dist-info/WHEEL,sha256=eh7sammvW2TypMMMGKgsM83HyA_3qQ5Lgg3ynoecH3M,79
+html_table_parse-0.2.2.dist-info/METADATA,sha256=zjUa1mJaN7X60BRE9RPB_-MINvItiH5mzxOjPuqdY_Y,1837
+html_table_parse-0.2.2.dist-info/RECORD,,

html_table_parse-0.2.dist-info/METADATA

@@ -1,15 +0,0 @@
-Metadata-Version: 2.3
-Name: html-table-parse
-Version: 0.2
-Summary: Parse HTML table as Python list or dict
-Author: 5j9
-Author-email: 5j9 <5j9@users.noreply.github.com>
-License: GPL-3.0
-Requires-Dist: lxml>=6.1.1
-Requires-Python: >=3.14
-Project-URL: Homepage, https://github.com/5j9/html-table-parse
-Description-Content-Type: text/x-rst
-
-Parse HTML table as Python list or dict
-
-Requires Python 3.14+

html_table_parse-0.2.dist-info/RECORD

@@ -1,4 +0,0 @@
-html_table_parse/__init__.py,sha256=WBw2LjxXiqRISTzXeoYlE0_LfUQPW9In2wQVhwpXDLk,4942
-html_table_parse-0.2.dist-info/WHEEL,sha256=eh7sammvW2TypMMMGKgsM83HyA_3qQ5Lgg3ynoecH3M,79
-html_table_parse-0.2.dist-info/METADATA,sha256=FcUDwK2un5zDd6RgyOUEL9RENEmLxuK04Dl1555E5BQ,400
-html_table_parse-0.2.dist-info/RECORD,,