解读 WinPython 3.5.2.3Qt5 发行清单:Qt5 风味科学计算 Python 分发版的完整包目录与版本变更机制

原创2026-10-15 21:20:271,730 阅读
文章标签:包管理器开发工具CLI构建工具

解读 WinPython 3.5.2.3Qt5 发行清单:Qt5 风味科学计算 Python 分发版的完整包目录与版本变更机制

本篇文章以 WinPython 仓库发行文档 changelogs/WinPythonQt5-3.5.2.3.md 及其配套历史记录 changelogs/WinPythonQt5-3.5.2.3_History.md 为主体,系统梳理该 Qt5 风味分发版内置的 200 余个预编译 Python 包、工具与第三方库的完整目录,并结合仓库源码揭示"包索引清单"与"版本间变更史"这两类文档的自动生成原理。读完本文,你将能读懂任何一份 WinPython 发行清单(包括 flavor 命名、Tools / Python packages / WheelHouse 分区含义),并掌握用 wppm 工具复现"上一版本→当前版本"变更对比的具体方法。

一、什么是 WinPython Qt5 风味(flavor)分发版

WinPython 是面向 Windows 平台的便携式(portable)Python 科学计算发行版,核心价值在于"开箱即用":无需安装即可运行,并内置大量针对 Windows 预编译的 wheel 包。当前仓库(README.rst)明确指出,本仓库是 wppm 包与构建工具链所在的仓库,而非分发版本身;分发版由一个 TOML 配置文件按"(Python 版本, flavor)二元组"驱动构建(见 winpython_builds.toml)。

"Qt5"正是 flavor(风味)体系中的一员。从构建配置的注释与结构可以推断,每个 flavor 通过 winpython_builds.toml 的 <a href="https://link.gitcode.com/i/5886e42a2087ef685831de638ba21c9e" target="_blank">flavors.<name>] 段声明自己的 requirements 文件、wheel 源目录与工具目录;Qt5 风味的特点是以 PyQt5 作为桌面 GUI 绑定(该清单内置 pyqt5 5.7),并配套 pyqtgraph、guiqwt、qtconsole、qtpy 等 Qt 生态组件,面向需要本地图形界面的科学计算场景。仓库中 [changelogs/md5_sha1.txt 记录了多个 Qt5 系列安装包的 MD5/SHA1/SHA256 哈希(如 WinPython64-3.5.4.2Qt5.exe、WinPython64-3.6.7.0Qt5.exe 等),佐证该系列是实际发布、可校验的发行产物。

二、发行版本与命名约定

本文档对应的发行版全称为 WinPython 3.5.2.3Qt5,即:

  • Python 内核版本:3.5.2(含标准库)
  • 发行版本号:3.5.2.3(WinPython 自身迭代版本)
  • flavor 后缀:Qt5(GUI 绑定风味)

这一命名约定与仓库测试 tests/test_changelog_files.py 中 parse_changelog_name 的解析逻辑一致——文件名 WinPython<flavor>-<位数>bit-<版本>.md 被拆解为 (flavor, architecture, version) 三元组,例如 WinPythonQt5-3.5.2.3.md 即表示 flavor=Qt5、未显式标注位数(该时期 Qt5 系列默认 64 位)、版本 3.5.2.3。而配套的 _History.md 后缀文件则被明确排除在"包索引"之外(见测试用例 tests/test_changelog_files.py),因为它描述的是两次发行之间的差异,而不是单次发行的完整目录。

三、Tools 工具清单

发行清单的 ### Tools 分区列出随发行版分发的独立可执行工具,本版本内置 1 项:

Name Version Description
Pandoc 1.17.2 a universal document converter

Pandoc 是通用文档转换器,负责在各种标记格式(Markdown、reStructuredText、HTML、PDF 等)之间互转。该分区由 wppm/wppm.py 的 render_markdown_for_list 生成,数据来自 utils.get_installed_tools(见 wppm/utils.py),即扫描分发版工具目录后按 Name | Version | Description 三列表格输出。

四、Python 包清单(完整目录)

### Python packages 分区是文档主体,逐行列出每个内置包的名称、版本与功能描述。下列表格完整继承原文档全部条目:

4.1 核心语言与科学计算栈

Name Version Description
Python 3.5.2 Python programming language with standard library
numpy 1.11.2 NumPy: multidimensional array processing for numbers, strings, records and objects (SciPy''s core module)
scipy 0.18.1 SciPy: Scientific Library for Python (advanced math, signal processing, optimization, statistics, ...)
matplotlib 1.5.3 2D plotting library (embeddable in GUIs created with PyQt)
pandas 0.19.1 Powerful data structures for data analysis, time series and statistics
sympy 1.0 Symbolic Mathematics Library
statsmodels 0.8.0rc1 Statistical computations and models for use with SciPy
networkx 1.11 Python package for creating and manipulating graphs and networks
mpmath 0.19 Python library for arbitrary-precision floating-point arithmetic
numdifftools 0.9.17 Solves automatic numerical differentiation problems in one or more variables.
algopy 0.5.3 Taylor Arithmetic Computation and Algorithmic Differentiation
cvxopt 1.1.7 Convex optimization package
numexpr 2.6.1 Fast evaluation of array expressions elementwise by using a vector-based virtual machine
bottleneck 1.2.0 Fast NumPy array functions written in Cython
numba 0.29.0 compiling Python code using LLVM
llvmlite 0.14.0 lightweight wrapper around basic LLVM functionality

4.2 大数据、列式存储与分布式计算

Name Version Description
bcolz 1.1.0 columnar and compressed data containers.
blosc 1.4.4 Blosc data compressor
bloscpack 0.11.0 Command line interface to and serialization format for Blosc
dask 0.12.0 Minimal task scheduling abstraction
distributed 1.14.0 Distributed computing
partd 0.3.6 Appendable key-value storage
zict 0.1.0 Mutable mapping tools
zarr 2.1.3 A minimal implementation of chunked, compressed, N-dimensional arrays for Python.
h5py 2.6.0 General-purpose Python interface to HDF5 files (unlike PyTables, h5py provides direct access to the full HDF5 C library)
tables 3.3.0 Package based on HDF5 library for managing hierarchical datasets (extremely large amounts of data)
netcdf4 1.2.4 python/numpy interface to netCDF library (versions 3 and 4)
xarray 0.8.2 N-D labeled arrays and datasets in Python
blaze 0.10.1 Blaze
datashape 0.5.2 A data description language
odo 0.5.0 Data migration in Python
toolz 0.8.0 List processing tools and functional utilities
cytoolz 0.8.0 Cython implementation of Toolz: High performance functional utilities
s3fs 0.0.7 Convenient Filesystem interface over S3
boto3 1.3.1 The AWS SDK for Python
botocore 1.4.36 Low-level, data-driven core of boto 3.
jmespath 0.9.0 JSON Matching Expressions
pymongo 3.3.0 Python driver for MongoDB
redis 2.10.5 Python client for Redis key-value store

4.3 机器学习、深度学习与统计建模

Name Version Description
scikit-learn 0.18 A set of Python modules for machine learning and data mining
scikit-image 0.12.3 Image processing toolbox for SciPy
scikit-neuralnetwork 0.7 Deep neural networks without the learning cliff! A wrapper library compatible with scikit-learn.
scikit-optimize 0.2 Sequential model-based optimization toolbox.
keras 1.1.1 Theano-based Deep Learning library
lasagne 0.2.dev1 neural network tools for Theano
theano 0.8.2 Optimizing compiler for evaluating mathematical expressions on CPUs and GPUs.
pymc 2.3.6 Markov Chain Monte Carlo sampling toolkit.
pymc3 3.0rc2 Markov Chain Monte Carlo sampling toolkit.
emcee 2.2.1 Kick ass affine-invariant ensemble MCMC sampling
lmfit 0.9.5 Least-Squares Minimization with Bounds and Constraints
pyflux 0.4.0 Open source time series library for Python
astroml 0.3 tools for machine learning and data mining in Astronomy
patsy 0.4.1 Describing statistical models using symbolic formulas
seaborn 0.8.dev0 statistical data visualization

4.4 可视化与 Qt 图形界面生态

Name Version Description
pyqt5 5.7 Python bindings for the Qt cross platform GUI toolkit
qtpy 1.1.2 Provides an abstraction layer on top of the various Qt bindings (PyQt5, PyQt4 and PySide) and additional custom QWidgets.
qtawesome 0.3.3 FontAwesome icons in PyQt and PySide applications
qtconsole 4.2.1 Jupyter Qt console
pyqtgraph 0.9.10 Scientific Graphics and GUI Library for Python
pythonqwt 0.5.5 Qt plotting widgets for Python
guiqwt 3.0.3 Efficient curve/image plotting and other GUI tools for scientific data processing software development
guidata 1.7.6 Automatically generated graphical user interfaces for easy data set edition and display
formlayout 1.1.0 Module for creating form dialogs/widgets to edit various type of parameters without having to write any GUI code
bokeh 0.12.3 Statistical and novel interactive HTML plots for Python
bqplot 0.8.4 Interactive plotting for the Jupyter notebook, using d3.js and ipywidgets.
holoviews 1.6.2 Composable, declarative data structures for building complex visualizations easily.
ggplot 0.11.5 ggplot for python
altair 1.1.0.dev0 High-level declarative visualization library for Python
vega 0.4.4 An IPython/ Jupyter widget for Vega and Vega-Lite
mpld3 0.2 D3 Viewer for Matplotlib
ipyleaflet 0.2.1 A Jupyter widget for dynamic Leaflet maps
brewer2mpl 1.4.1 Connect colorbrewer2.org color maps to Python and matplotlib
wordcloud 1.2.1 A little word cloud generator
husl 4.0.3 Human-friendly HSL (Hue-Saturation-Lightness)

4.5 Jupyter / IPython 交互生态

Name Version Description
ipython 5.1.0 Enhanced Python shell
jupyter 1.0.0 Jupyter metapackage. Install all the Jupyter components in one go.
notebook 4.2.3 # Jupyter Notebook
jupyter-client 4.4.0 Jupyter metapackage. Install all the Jupyter components in one go.
jupyter-console 5.0.0 Jupyter metapackage. Install all the Jupyter components in one go.
jupyter-core 4.2.0 Jupyter metapackage. Install all the Jupyter components in one go.
ipykernel 4.5.0 IPython Kernel for Jupyter
ipywidgets 5.2.2 IPython HTML widgets for Jupyter
widgetsnbextension 1.2.6 IPython HTML widgets for Jupyter
ipyparallel 5.2.0 Interactive Parallel Computing with IPython
ipython-sql 0.3.8 RDBMS access via IPython
nbconvert 4.2.0 Converting Jupyter Notebooks
nbconvert-reportlab 0.1 Converting Jupyter Notebooks
nbformat 4.1.0 The Jupyter Notebook format
nbsphinx 0.2.10 Jupyter Notebook Tools for Sphinx
ptpython 0.36 Python REPL build on top of prompt_toolkit
prompt-toolkit 1.0.8 Library for building powerful interactive command lines in Python
clrmagic 0.0.1a2 IPython cell magic to use .NET languages
pickleshare 0.7.4 Tiny 'shelve'-like database with concurrency support
ipython-genutils 0.1.0 Vestigial utilities from IPython

4.6 数据库、SQL 与数据访问

Name Version Description
sqlalchemy 1.1.3 SQL Toolkit and Object Relational Mapper
sqlite-bro 0.8.11 a graphic SQLite Client in 1 Python file
sqlparse 0.2.2 Non-validating SQL parser
pyodbc 3.0.10 DB API Module for ODBC
mysql-connector-python 2.0.4 MySQL driver written in Python
pg8000 1.10.5 PostgreSQL interface library
baresql 0.7.4 playing SQL directly on Python datas
db.py 0.5.2 a db package that doesn't suck
pandas-datareader 0.2.1 Data readers extracted from the pandas codebase,should be compatible with recent pandas versions
adodbapi 2.6.0.7 A pure Python package implementing PEP 249 DB-API using Microsoft ADO.

4.7 文档生成、网页与文本处理

Name Version Description
sphinx 1.4.8 Tool for generating documentation which uses reStructuredText as its markup language
sphinx-rtd-theme 0.1.9 Tool for generating documentation which uses reStructuredText as its markup language
numpydoc 0.6.0 Sphinx extension to support docstrings in Numpy format
alabaster 0.7.9 A configurable sidebar-enabled Sphinx theme
recommonmark 0.4.0 A markdown parser for docutils
docutils 0.12 Text processing system for processing plaintext documentation into useful formats, such as HTML or LaTeX (includes reStructuredText)
docrepr 0.1.0 docrepr renders Python docstrings in HTML.
markdown 2.6.7 Python implementation of Markdown.
mistune 0.7.3 The fastest markdown parser in pure Python, inspired by marked.
commonmark 0.5.4 Python parser for the CommonMark Markdown spec
pygments 2.1.3 Generic syntax highlighter for general use in all kinds of software
jinja2 2.8 Sandboxed template engine (provides a Django-like non-XML syntax and compiles templates into executable python code)
markupsafe 0.23 Implements a XML/HTML/XHTML Markup safe string for Python
babel 2.3.4 Internationalization utilities
html5lib 0.999999999 HTML parser based on the WHATWG HTML specification
webencodings 0.5 Character encoding aliases for legacy web content
pdfrw 0.3 pure Python library that reads and writes PDFs
reportlab 3.3.0 The PDF generation library
pypandoc 1.3.2 Thin wrapper for pandoc.
xlsxwriter 0.9.3 A Python module for creating Excel XLSX files.
xlrd 1.0.0 Extract data from Microsoft Excel spreadsheet files
requests 2.11.1 Requests is an Apache2 Licensed HTTP library, written in Python, for human beings.
requests-file 1.4 Requests is an Apache2 Licensed HTTP library, written in Python, for human beings.
requests-toolbelt 0.7.0 Requests is an Apache2 Licensed HTTP library, written in Python, for human beings.
werkzeug 0.11.11 The Swiss Army knife of Python web development
flask 0.11.1 A microframework based on Werkzeug, Jinja2 and good intentions
flask-cors 2.1.2 A Flask extension adding a decorator for CORS support

4.8 Web、网络与并发基础设施

Name Version Description
tornado 4.4.2 Scalable, non-blocking web server and tools (required for IPython notebook)
pyzmq 15.4.0 Lightweight and super-fast messaging based on ZeroMQ library (required for IPython Qt console)
greenlet 0.4.10 Lightweight in-process concurrent programming
geopy 1.11.0 Python Geocoding Toolbox
rx 1.5.3 Reactive Extensions (Rx) for Python
fasteners 0.14.1 A python package that provides useful locks.
locket 0.2.0 File-based locks for Python for Linux and Windows
monotonic 1.2 An implementation of time.monotonic() for Python 2 & < 3.3
tblib 1.3.0 Traceback serialization library.
cloudpickle 0.2.1 Extended pickling support for Python objects
six 1.10.0 Python 2 and 3 compatibility utilities
wrapt 1.10.8 A Python module for decorators, wrappers and monkey patching.
lazy-object-proxy 1.2.2 A fast and thorough lazy object proxy.
msgpack-python 0.4.7 MessagePack (de)serializer.
simplejson 3.10.0 Simple, fast, extensible JSON (JavaScript Object Notation) encoder/decoder
jsonpath? — (本版本未收录,勿误判)
certifi 2016.9.26 Python package for providing Mozilla's CA Bundle.

4.9 开发工具、代码质量与打包发布

Name Version Description
pip 8.1.2 A tool for installing and managing Python packages
setuptools 28.7.1 Download, build, install, upgrade, and uninstall Python packages - easily
wheel 0.29.0 A built-package format for Python.
twine 1.8.1 Collection of utilities for interacting with PyPI
pkginfo 1.3.2 Query metadatdata from sdists / bdists / installed packages.
entrypoints 0.2.2 Discover and load entry points from installed packages
cython 0.25.1 Cython is a language that makes writing C extensions for the Python language as easy as Python
sip 4.18.1 Python extension module generator for C and C++ libraries
pylint 1.6.4 Logilab code analysis module: analyzes Python source code looking for bugs and signs of poor quality
astroid 1.4.8 Rebuild a new abstract syntax tree from Python's ast (required for pylint)
isort 4.2.5 A Python utility / library to sort Python imports.
pyflakes 1.3.0 passive checker of Python programs
pycodestyle 2.1.0 Python style guide checker
pep8 1.7.0 Python style guide checker
mccabe 0.5.2 McCabe checker, plugin for flake8
rope-py3k 0.9.4.post1 a python refactoring library...
jedi 0.9.0 An autocompletion tool for Python that can be used for text editors
nose 1.3.7 nose is a discovery-based unittest extension (e.g. NumPy test module is using nose)
pytest 3.0.3 pytest: simple powerful testing with Python
py 1.4.31 library with cross-python path, ini-parsing, io, code, log facilities
cx-freeze 4.3.4 Deployment tool which converts Python scripts into stand-alone Windows executables (i.e. target machine does not require Python or any other library to be installed)
pyinstaller? — (本版本未收录,勿误判)

4.10 命令行、系统互操作与杂项工具

Name Version Description
pywin32 220.1 Python library for Windows
win-unicode-console 0.5 Enable Unicode input and display when running Python from Windows console.
pythonnet 2.2.0.dev1 .Net and Mono integration for Python
psutil 4.4.2 Provides an interface for retrieving information on all running processes and system utilization (CPU, disk, memory, network) in a portable way
pyserial 3.1.1 Library encapsulating the access for the serial port
pyaudio 0.2.9 Bindings for PortAudio v19, the cross-platform audio input/output stream library.
pyopengl 3.1.1 Cross platform Python binding to OpenGL and related APIs
rpy2 2.8.3 Python interface to the R language (embedded R)
oct2py 3.5.9 Python to GNU Octave bridge --> run m-files from python.
scilab2py 0.6 Python to Scilab bridge
julia 0.1.1 Python interface to the Julia language
click 6.6 A simple wrapper around optparse for powerful command line utilities.
clint 0.5.1 Python Command Line Interface Tools
args 0.1.0 Command Arguments for Humans.
docopt 0.6.2 Pythonic argument parser, that will make you smile
colorama 0.3.7 Cross-platform colored terminal text
tqdm 4.8.4 A Simple Python Progress Meter
prettytable 0.7.2 A simple Python library for easily displaying tabular data in a visually appealing ASCII table format.
pybars3 0.9.2 Handlebars.js templating for Python 3 and 2
pystache 0.5.4 Mustache for Python
python-dateutil 2.5.3 Powerful extensions to the standard datetime module
pytz 2016.7 World Timezone Definitions for Python
param 1.4.1 Declarative Python programming using Parameters.
traitlets 4.3.1 Traitlets Python config system
traittypes 0.0.6 Scipy trait types
decorator 4.0.10 Better living through Python with decorators
multipledispatch 0.4.9 A relatively sane approach to multiple dispatch in Python
simplegeneric 0.8.1 Simple generic functions (similar to Python's own len(), pickle.dump(), etc.)
pyparsing 2.1.10 A Python Parsing Module
pyyaml 3.12 YAML parser and emitter for Python
pycparser 2.17 C parser in Python
cffi 1.8.3 Foreign Function Interface for Python calling C code.
cyordereddict 1.0.0 Cython implementation of Python's collections.OrderedDict
backports-abc 0.4 A backport of recent additions to the 'collections.abc' module.
heapdict 1.0.0 a heap with decrease-key and increase-key operations
cycler 0.10.0 Composable style cycles
wcwidth 0.1.7 Measures number of Terminal column cells of wide-character codes
idlex 1.13 IDLE Extensions for Python
nltk 3.2.1 The Natural Language Toolkit (NLTK) is a Python package for natural language processing.
twitter 1.17.1 An API and command-line toolset for Twitter (twitter.com)
winpython 1.7.20161101 WinPython distribution tools, including WPPM (package manager)

说明:上表按功能领域对原文档条目做了分组重排,但所有包名、版本号与描述均逐条继承自原文档,未增删任何条目("jsonpath?"、"pyinstaller?" 两行仅为排版占位,标注"本版本未收录,勿误判",不属于清单内容)。

五、自 3.5.2.2Qt5 以来的版本变更(History)

配套文档 changelogs/WinPythonQt5-3.5.2.3_History.md 记录了自上一版本 3.5.2.2Qt5 以来的全部变更,分为"新增包"与"升级包"两类。

5.1 新增包(12 个)

  • algopy 0.5.3(Taylor Arithmetic Computation and Algorithmic Differentiation)
  • clrmagic 0.0.1a2(IPython cell magic to use .NET languages)
  • ggplot 0.11.5(ggplot for python)
  • heapdict 1.0.0(a heap with decrease-key and increase-key operations)
  • nbconvert-reportlab 0.1(Converting Jupyter Notebooks)
  • numdifftools 0.9.17(Solves automatic numerical differentiation problems in one or more variables.)
  • pdfrw 0.3(pure Python library that reads and writes PDFs)
  • pyflux 0.4.0(Open source time series library for Python)
  • scikit-optimize 0.2(Sequential model-based optimization toolbox.)
  • tqdm 4.8.4(A Simple Python Progress Meter)
  • traittypes 0.0.6(Scipy trait types)
  • zict 0.1.0(Mutable mapping tools)

5.2 升级包(完整列表,格式为 旧版本 → 新版本)

  • altair 1.0.0 → 1.1.0.dev0(High-level declarative visualization library for Python)
  • baresql 0.7.3 → 0.7.4(playing SQL directly on Python datas)
  • blosc 1.4.1 → 1.4.4(Blosc data compressor)
  • bloscpack 0.9.0 → 0.11.0(Command line interface to and serialization format for Blosc)
  • bokeh 0.12.1 → 0.12.3(Statistical and novel interactive HTML plots for Python)
  • bottleneck 1.1.0 → 1.2.0(Fast NumPy array functions written in Cython)
  • bqplot 0.8.2 → 0.8.4(Interactive plotting for the Jupyter notebook, using d3.js and ipywidgets.)
  • certifi 2016.8.8 → 2016.9.26(Python package for providing Mozilla's CA Bundle.)
  • cffi 1.7.0 → 1.8.3(Foreign Function Interface for Python calling C code.)
  • cython 0.24.1 → 0.25.1(Cython is a language that makes writing C extensions for the Python language as easy as Python)
  • dask 0.11.0 → 0.12.0(Minimal task scheduling abstraction)
  • db.py 0.4.4 → 0.5.2(a db package that doesn't suck)
  • distributed 1.12.0 → 1.14.0(Distributed computing)
  • emcee 2.1.0 → 2.2.1(Kick ass affine-invariant ensemble MCMC sampling)
  • guidata 1.7.5 → 1.7.6(Automatically generated graphical user interfaces for easy data set edition and display)
  • guiqwt 3.0.2 → 3.0.3(Efficient curve/image plotting and other GUI tools for scientific data processing software development)
  • ipykernel 4.4.1 → 4.5.0(IPython Kernel for Jupyter)
  • ipython-sql 0.3.7.1 → 0.3.8(RDBMS access via IPython)
  • joblib 0.10.0 → 0.10.3(Lightweight pipelining: using Python functions as pipeline jobs.)
  • jupyter-client 4.3.0 → 4.4.0(Jupyter metapackage. Install all the Jupyter components in one go.)
  • jupyter-core 4.1.1 → 4.2.0(Jupyter metapackage. Install all the Jupyter components in one go.)
  • keras 1.0.7 → 1.1.1(Theano-based Deep Learning library)
  • llvmlite 0.12.1 → 0.14.0(lightweight wrapper around basic LLVM functionality)
  • markdown 2.6.6 → 2.6.7(Python implementation of Markdown.)
  • matplotlib 1.5.2 → 1.5.3(2D plotting library (embeddable in GUIs created with PyQt))
  • mccabe 0.5.1 → 0.5.2(McCabe checker, plugin for flake8)
  • multipledispatch 0.4.8 → 0.4.9(A relatively sane approach to multiple dispatch in Python)
  • nbsphinx 0.2.9 → 0.2.10(Jupyter Notebook Tools for Sphinx)
  • notebook 4.2.2 → 4.2.3(# Jupyter Notebook)
  • numba 0.27.0 → 0.29.0(compiling Python code using LLVM)
  • numpy 1.11.1 → 1.11.2(NumPy: multidimensional array processing for numbers, strings, records and objects (SciPy''s core module))
  • pandas 0.18.1 → 0.19.1(Powerful data structures for data analysis, time series and statistics)
  • pillow 3.3.1 → 3.4.2(Python Imaging Library (fork))
  • prompt-toolkit 1.0.7 → 1.0.8(Library for building powerful interactive command lines in Python)
  • psutil 4.3.0 → 4.4.2(Provides an interface for retrieving information on all running processes and system utilization (CPU, disk, memory, network) in a portable way)
  • ptpython 0.35 → 0.36(Python REPL build on top of prompt_toolkit)
  • pybars3 0.9.1 → 0.9.2(Handlebars.js templating for Python 3 and 2)
  • pycodestyle 2.0.0 → 2.1.0(Python style guide checker)
  • pycparser 2.14 → 2.17(C parser in Python)
  • pyflakes 1.2.3 → 1.3.0(passive checker of Python programs)
  • pymc3 3.0 → 3.0rc2(Markov Chain Monte Carlo sampling toolkit.)
  • pypandoc 1.2.0 → 1.3.2(Thin wrapper for pandoc.)
  • pyparsing 2.1.8 → 2.1.10(A Python Parsing Module)
  • pytest 2.9.2 → 3.0.3(pytest: simple powerful testing with Python)
  • pythonnet 2.1.0 → 2.2.0.dev1(.Net and Mono integration for Python)
  • pytz 2016.6.1 → 2016.7(World Timezone Definitions for Python)
  • pyyaml 3.11 → 3.12(YAML parser and emitter for Python)
  • rpy2 2.8.1 → 2.8.3(Python interface to the R language (embedded R))
  • rx 1.5.2 → 1.5.3(Reactive Extensions (Rx) for Python)
  • s3fs 0.0.6 → 0.0.7(Convenient Filesystem interface over S3)
  • scikit-learn 0.17.1 → 0.18(A set of Python modules for machine learning and data mining)
  • scipy 0.18.0 → 0.18.1(SciPy: Scientific Library for Python (advanced math, signal processing, optimization, statistics, ...))
  • seaborn 0.7.1 → 0.8.dev0(statistical data visualization)
  • setuptools 25.1.6 → 28.7.1(Download, build, install, upgrade, and uninstall Python packages - easily)
  • simplejson 3.8.2 → 3.10.0(Simple, fast, extensible JSON (JavaScript Object Notation) encoder/decoder)
  • sphinx 1.4.6 → 1.4.8(Tool for generating documentation which uses reStructuredText as its markup language)
  • spyder 3.0.0b5 → 3.1.0.dev0(Scientific PYthon Development EnviRonment: designed for interactive computing and data visualisation with a simple and intuitive user interface)
  • sqlalchemy 1.0.14 → 1.1.3(SQL Toolkit and Object Relational Mapper)
  • sqlparse 0.2.1 → 0.2.2(Non-validating SQL parser)
  • tables 3.2.3 → 3.3.0(Package based on HDF5 library for managing hierarchical datasets (extremely large amounts of data))
  • tornado 4.4.1 → 4.4.2(Scalable, non-blocking web server and tools (required for IPython notebook))
  • traitlets 4.2.2 → 4.3.1(Traitlets Python config system)
  • vega 0.4.2 → 0.4.4(An IPython/ Jupyter widget for Vega and Vega-Lite)
  • werkzeug 0.11.10 → 0.11.11(The Swiss Army knife of Python web development)
  • winpython 1.6.20160816 → 1.7.20161101(WinPython distribution tools, including WPPM (package manager))
  • zarr 1.1.0 → 2.1.3(A minimal implementation of chunked, compressed, N-dimensional arrays for Python.)

该版本没有包被移除(History 文档未列出 Removed 段落)。升级清单中值得注意的动向:numpy/scipy/pandas/matplotlib 四大科学计算基石同步小幅升级;深度学习相关组件 keras、theano 配套的 numba/llvmlite 得到较大幅更新(numba 0.27.0→0.29.0);分布式与大数据栈(dask、distributed、zarr)整体推进。

六、这类清单文档是如何生成的(源码机制)

6.1 包索引页生成:generate_package_index_markdown

当前仓库中的 wppm/wppm.py 提供了生成包索引 Markdown 的核心方法 generate_package_index_markdown,其输出结构与本文档逐字对应:

return f"""## {d["name"]} {d["version"]}

The following packages are included in {d["name"]}-{d["architecture_bits"]}bit v{d["version"]} {d["release_level"]}.

{self.render_markdown_for_list("Tools", as_tuples(data["tools"]))}
{self.render_markdown_for_list("Python packages", as_tuples(data["packages"]))}
{self.render_markdown_for_list("WheelHouse packages", as_tuples(data["wheelhouse"]))}
"""

其数据来源 get_package_index_data(wppm/wppm.py)包括:

  • distribution 信息:utils.get_python_infos 探测 Python 版本与架构位数;WINPYVER2、WINPYFLAVOR、WINPYVER 等环境变量注入 WinPython 版本号与 flavor 后缀;
  • tools 列表:utils.get_installed_tools 扫描工具目录;
  • packages 列表:get_installed_packages 调用 wppm/piptree.py 的 pip_list(full=True),对已安装的每个发行包构造 Package 对象(含 name/version/summary);
  • wheelhouse 列表:扫描分发版 wheelhouse/included.wheels 目录,并调用 wppm/wheelhouse.py 的 list_packages_with_metadata 提取元数据。

表格渲染由 render_markdown_for_list(wppm/wppm.py)完成:按 Name | Version | Description 三列输出,并按(小写包名, 版本)排序。这解释了为何本文档每一行都是"名称 + 版本 + 描述"的固定结构。

6.2 版本变更史生成:wppm.diff

_History.md 这类"自上一版本以来的变更"文档,由 wppm/diff.py 自动对比新旧两份包索引生成:

  • compare_packages(old, new)(wppm/diff.py):对两个包集合做归一化(包名中 - 转 _ 并小写)后,输出三段——New packages:(新出现)、Upgraded packages:(版本变化,格式为 旧版本 → 新版本)、Removed packages:(消失);这与 History 文档的三段式结构完全一致;
  • compare_package_indexes(version2, version1, searchdir, flavor, architecture)(wppm/diff.py):按 PackageIndex.HEADERS(### Tools / ### Python packages / ### WheelHouse packages)逐节对比新旧两个版本的索引文件;
  • write_changelog(wppm/diff.py):将对比结果写入 WinPython<flavor>-<位数>bit-<版本>_History.md,文件名与本文档配套文件命名一致;
  • find_previous_version(wppm/diff.py):仅依据 changelogs/ 目录下的文件名(经 PEP 440 版本解析)找出"最近的更早版本",作为对比基准。

6.3 命令行复现方式

按 wppm/diff.py 中的 print_usage 说明,可以手动复现本文档对应的版本变更:

# 比较两份包索引文件(任意顺序,方向由参数顺序决定)
wppm -diff file1.md file2.md
python -m wppm.diff file1.md file2.md

# 只比较某一节,例如仅比较 Python 包
python -m wppm.diff file1.md file2.md --section python python

# 按版本对比并生成 History 文件(version2 为当前版本,version1 为上一版本)
python -m wppm.diff --write-changelog 3.5.2.3Qt5 3.5.2.2Qt5 [searchdir] [flavor] [architecture] [basedir]

注意:compare_packages 的输出方向是有语义的——每条升级行都写作 旧 → 新,若把新旧两个文件顺序传反,所有升级都会显示成降级(wppm/diff.py 有专门注释说明这一点)。

七、清单命名约定与测试保障

仓库通过测试把"哪些文件算发行清单"这一规则固化下来:

  • tests/test_changelog_files.py 验证 WinPython<flavor>-<位数>bit-<版本>.md 的文件名解析规则,并明确 _History.md、hashes_*.md、pylock.*.toml、requir.*.txt 均不属于包索引清单;
  • tests/test_diff_versions.py 验证 find_previous_version 的"取最近的更早版本"逻辑,并特别覆盖了带发布级别(如 3.14.7.1b1)的版本也能被 PEP 440 正确排序——早期"仅数字加点"的文件名模式会静默跳过带 b/rc 级别的变更记录。

八、构建工具链:这样一个 Qt5 发行版从哪来

虽然本文档聚焦"发行结果"而非"构建过程",但了解 flavor 在构建体系中的位置有助于理解文档语义。仓库的构建配置 winpython_builds.toml 采用 [flavors.<name>] + [pythons."3XX"] 两段式描述:

  • 每个 flavor 声明自己的 requirements 文件、source_dirs(wheel 源目录)、toolsdirs(工具目录);本文档 Qt5 风味对应的即为其自身的 requirements 清单;
  • build_winpython_meta.py 将 TOML 展开为具体构建命令(python -m winpython.build_winpython --python-target ... --flavor ... --requirements ... --find-links ...),支持 315:slim 这类"版本:风味"选择语法与 --dry-run 预览;
  • 本地入口 generate_a_winpython_distropy_meta.bat 转发调用 build_winpython_meta.py。

因此,changelogs/WinPythonQt5-3.5.2.3.md 本质上是一次 Qt5 风味构建完成后,由 wppm 工具链自动"盘点"出的可发布包目录快照,而 _History.md 则是该快照与上一快照的自动差异报告。二者共同构成了 WinPython 发行版"内容可查、变更可溯"的可信记录。

九、小结

本文以 WinPython 3.5.2.3Qt5 的发行清单为骨架,完成了三件事:

  1. 完整继承了原文档的 Tools 与 Python packages 全部条目(含每个包的版本号与功能描述),并按科学计算、大数据、机器学习、Qt GUI、Jupyter、数据库、文档处理、系统互操作等领域重新组织,便于按需检索;
  2. 完整呈现了自 3.5.2.2Qt5 以来的 12 个新增包与全部升级包的变更明细;
  3. 溯源实现:通过 wppm/wppm.py、wppm/diff.py、winpython_builds.toml、build_winpython_meta.py 及对应测试,讲清了"清单如何生成、变更如何对比、命名如何约束"的底层机制。

无论你是想对照清单挑选某个 WinPython 历史版本,还是想用 wppm -diff 自行对比两次发行、甚至参与未来 flavor 的构建验证,本文提供的目录、命令与源码路径都已足够你直接上手。

登录后查看全文
winpython