Update to tidy-html5 v5.7.45

Author: Kijewski

.github/workflows/build_wheels.yml

@@ -0,0 +1,64 @@
+name: Build wheels
+
+on:
+  workflow_dispatch:
+    inputs:
+      os:
+        description: OS
+        required: true
+        default: ubuntu-20.04
+      python:
+        description: Python
+        required: true
+        default: 3.8
+      manylinux:
+        description: Manylinux
+        required: true
+        default: manylinux2014
+
+jobs:
+  build_wheels:
+    name: Build compat:${{ github.event.inputs.manylinux }}, python:${{ github.event.inputs.python }}, os:${{ github.event.inputs.os }}
+    runs-on: ${{ github.event.inputs.os }}
+
+    steps:
+      - uses: actions/checkout@v2
+        with:
+          submodules: true
+
+      - name: Cache pip
+        uses: actions/cache@v2
+        with:
+          key: cache--${{ github.event.inputs.os }}--${{ github.event.inputs.python }}--${{ hashFiles('./requirements.txt') }}
+          path: ~/.cache/pip
+
+      - uses: actions/setup-python@v2
+        name: Install Python
+
+      - name: Setup python
+        uses: actions/setup-python@v2
+        with:
+          python-version: ${{ github.event.inputs.python }}
+
+      - name: Update pip
+        run: python -m pip install -U pip wheel setuptools
+
+      - name: Install requirements
+        run: python -m pip install -Ur requirements.txt
+
+      - name: Convert Cython code
+        run: make _pytidyhtml5.cpp
+
+      - name: Build wheels
+        run: python -m cibuildwheel --output-dir wheelhouse
+        env:
+          CIBW_BEFORE_ALL: make tidy-html5/build/cmake/libtidys.a
+          CIBW_SKIP: "cp27-* pp27-* cp35-* pp35-*"  # skip Python 2.7 & 3.5 wheels
+          CIBW_MANYLINUX_X86_64_IMAGE: ${{ github.event.inputs.manylinux }}
+          CIBW_MANYLINUX_I686_IMAGE: ${{ github.event.inputs.manylinux }}
+
+      - name: Store artifacts
+        uses: actions/upload-artifact@v2
+        with:
+          name: Wheelhouse
+          path: ./wheelhouse/*.whl

.gitmodules

@@ -1,4 +1,3 @@
 [submodule "tidy-html5"]
 	path = tidy-html5
-	url = https://github.com/htacg/tidy-html5.git
-	branch = 3a30f6a4300417674026f6dddea5973debc6b808
+	url = git://github.com/htacg/tidy-html5.git

Makefile

@@ -4,10 +4,10 @@ all: sdist bdist_wheel docs
 
 NAME := pytidyhtml5
 
-.PHONY: all sdist bdist_wheel clean docs
+.PHONY: all sdist bdist_wheel clean docs prepare
 
 FILES := Makefile MANIFEST.in _${NAME}.pyx README.rst setup.py \
-         lib/native.hpp lib/VERSION
+         lib/native.hpp lib/VERSION tidy-html5/build/cmake/libtidys.a
 
 
 TIDY_CFLAGS := -O2 -fomit-frame-pointer -flto
@@ -45,13 +45,15 @@ tidy-html5/build/cmake/libtidys.a: | tidy-html5/.git
 
 	$(RANLIB) $@
 
-lib/_import_tidy_enum.pyx: tidy-html5/build/cmake/libtidys.a
+lib/_import_tidy_enum.pyx:
 	./generate_imports.py
 
-_${NAME}.cpp: _${NAME}.pyx $(wildcard lib/*.pyx) lib/_import_tidy_enum.pyx
-	rm -f -- dist/*.so _${NAME}.cpp
-	rm -f ./_${NAME}.cpp
-	cythonize -f $<
+_${NAME}.cpp: _${NAME}.pyx $(wildcard lib/*.pyx) | lib/_import_tidy_enum.pyx
+	-rm -- dist/*.so
+	-rm ./_${NAME}.cpp
+	python -m Cython.Build.Cythonize -f $<
+
+prepare: _${NAME}.cpp ${FILES}
 
 sdist: _${NAME}.cpp ${FILES}
 	rm -f -- dist/${NAME}-*.tar.gz
@@ -68,6 +70,7 @@ docs: bdist_wheel $(wildcard docs/* docs/*/*)
 	python -m sphinx -M html docs/ dist/
 
 clean:
+	-rm -- _${NAME}.cpp
 	-rm -- ./lib/_import_tidy_enum.pyx
 	-rm -- ./lib/_tidy_enum.pyx
 	-rm -r -- ./build/

README.rst

@@ -1,2 +1,4 @@
 PyTidyHTML5
 ===========
+
+HTML clean-up and repair: Statically linked Python interface for [tidy-html5](https://github.com/htacg/tidy-html5).

lib/VERSION

@@ -1 +1 @@
-"2020.7.17.post1"
+"2021.1.11"

lib/_import_tidy_enum.pyx

@@ -117,6 +117,8 @@ cdef extern from 'tidyenum.h' nogil:
     const TidyAttrId TidyAttr_CODEBASE
     const TidyAttrId TidyAttr_CODETYPE
     const TidyAttrId TidyAttr_COLOR
+    const TidyAttrId TidyAttr_COLORINTERPOLATION
+    const TidyAttrId TidyAttr_COLORRENDERING
     const TidyAttrId TidyAttr_COLS
     const TidyAttrId TidyAttr_COLSPAN
     const TidyAttrId TidyAttr_COMPACT
@@ -149,6 +151,9 @@ cdef extern from 'tidyenum.h' nogil:
     const TidyAttrId TidyAttr_ENCTYPE
     const TidyAttrId TidyAttr_EVENT
     const TidyAttrId TidyAttr_FACE
+    const TidyAttrId TidyAttr_FILL
+    const TidyAttrId TidyAttr_FILLOPACITY
+    const TidyAttrId TidyAttr_FILLRULE
     const TidyAttrId TidyAttr_FOR
     const TidyAttrId TidyAttr_FORM
     const TidyAttrId TidyAttr_FORMACTION
@@ -190,6 +195,7 @@ cdef extern from 'tidyenum.h' nogil:
     const TidyAttrId TidyAttr_LEFTMARGIN
     const TidyAttrId TidyAttr_LINK
     const TidyAttrId TidyAttr_LIST
+    const TidyAttrId TidyAttr_LOADING
     const TidyAttrId TidyAttr_LONGDESC
     const TidyAttrId TidyAttr_LOOP
     const TidyAttrId TidyAttr_LOW
@@ -213,6 +219,7 @@ cdef extern from 'tidyenum.h' nogil:
     const TidyAttrId TidyAttr_NOVALIDATE
     const TidyAttrId TidyAttr_NOWRAP
     const TidyAttrId TidyAttr_OBJECT
+    const TidyAttrId TidyAttr_OPACITY
     const TidyAttrId TidyAttr_OPEN
     const TidyAttrId TidyAttr_OPTIMUM
     const TidyAttrId TidyAttr_OnABORT
@@ -331,6 +338,7 @@ cdef extern from 'tidyenum.h' nogil:
     const TidyAttrId TidyAttr_SHOWGRIDY
     const TidyAttrId TidyAttr_SIZE
     const TidyAttrId TidyAttr_SIZES
+    const TidyAttrId TidyAttr_SLOT
     const TidyAttrId TidyAttr_SPAN
     const TidyAttrId TidyAttr_SPELLCHECK
     const TidyAttrId TidyAttr_SRC
@@ -340,6 +348,14 @@ cdef extern from 'tidyenum.h' nogil:
     const TidyAttrId TidyAttr_STANDBY
     const TidyAttrId TidyAttr_START
     const TidyAttrId TidyAttr_STEP
+    const TidyAttrId TidyAttr_STROKE
+    const TidyAttrId TidyAttr_STROKEDASHARRAY
+    const TidyAttrId TidyAttr_STROKEDASHOFFSET
+    const TidyAttrId TidyAttr_STROKELINECAP
+    const TidyAttrId TidyAttr_STROKELINEJOIN
+    const TidyAttrId TidyAttr_STROKEMITERLIMIT
+    const TidyAttrId TidyAttr_STROKEOPACITY
+    const TidyAttrId TidyAttr_STROKEWIDTH
     const TidyAttrId TidyAttr_STYLE
     const TidyAttrId TidyAttr_SUMMARY
     const TidyAttrId TidyAttr_TABINDEX
@@ -682,6 +698,7 @@ cdef extern from 'tidyenum.h' nogil:
     const TidyTagId TidyTag_SELECT
     const TidyTagId TidyTag_SERVER
     const TidyTagId TidyTag_SERVLET
+    const TidyTagId TidyTag_SLOT
     const TidyTagId TidyTag_SMALL
     const TidyTagId TidyTag_SOURCE
     const TidyTagId TidyTag_SPACER

lib/_imports.pyx

@@ -71,6 +71,7 @@ cdef extern from 'native.hpp' namespace 'PyTidyHtml5' nogil:
     PyObject **ref_document[Cls](PyObject *self)
     Out _reinterpret_cast[Out, In](In ptr)
     int call_bool_fn_1(object fn, object arg1)
+    void reset_hash[T](T *obj)
 
     ctypedef boolean AlwaysTrue
 

lib/_output_buffer.pyx

@@ -141,7 +141,7 @@ cdef class StringBuffer(Buffer):
             result = <object> <PyObject*> ObjectInitVar(
                 (<PyVarObject*> self.tidy_buffer.bp), bytes, length,
             )
-            (<PyBytesObject*> result).ob_shash = -1
+            reset_hash(<PyBytesObject*> result)
 
         else:
             s = &(<signed char*> self.tidy_buffer.bp)[start]
@@ -199,7 +199,7 @@ cdef class StringBuffer(Buffer):
             result = ObjectInit(<PyObject*> self.tidy_buffer.bp, unicode)
 
             (<PyASCIIObject*> result).length = length
-            (<PyASCIIObject*> result).hash = -1
+            reset_hash(<PyASCIIObject*> result)
             (<PyASCIIObject*> result).wstr = NULL
             (<PyASCIIObject*> result).state.interned = SSTATE_NOT_INTERNED
             (<PyASCIIObject*> result).state.kind = PyUnicode_1BYTE_KIND

lib/_tidy_enum.pyx

@@ -165,6 +165,8 @@ _AttrId = IntEnum('AttrId', {
     'codebase': <unsigned int> TidyAttr_CODEBASE,
     'codetype': <unsigned int> TidyAttr_CODETYPE,
     'color': <unsigned int> TidyAttr_COLOR,
+    'colorinterpolation': <unsigned int> TidyAttr_COLORINTERPOLATION,
+    'colorrendering': <unsigned int> TidyAttr_COLORRENDERING,
     'cols': <unsigned int> TidyAttr_COLS,
     'colspan': <unsigned int> TidyAttr_COLSPAN,
     'compact': <unsigned int> TidyAttr_COMPACT,
@@ -197,6 +199,9 @@ _AttrId = IntEnum('AttrId', {
     'enctype': <unsigned int> TidyAttr_ENCTYPE,
     'event': <unsigned int> TidyAttr_EVENT,
     'face': <unsigned int> TidyAttr_FACE,
+    'fill': <unsigned int> TidyAttr_FILL,
+    'fillopacity': <unsigned int> TidyAttr_FILLOPACITY,
+    'fillrule': <unsigned int> TidyAttr_FILLRULE,
     'for_': <unsigned int> TidyAttr_FOR,
     'form': <unsigned int> TidyAttr_FORM,
     'formaction': <unsigned int> TidyAttr_FORMACTION,
@@ -238,6 +243,7 @@ _AttrId = IntEnum('AttrId', {
     'leftmargin': <unsigned int> TidyAttr_LEFTMARGIN,
     'link': <unsigned int> TidyAttr_LINK,
     'list': <unsigned int> TidyAttr_LIST,
+    'loading': <unsigned int> TidyAttr_LOADING,
     'longdesc': <unsigned int> TidyAttr_LONGDESC,
     'loop': <unsigned int> TidyAttr_LOOP,
     'low': <unsigned int> TidyAttr_LOW,
@@ -261,6 +267,7 @@ _AttrId = IntEnum('AttrId', {
     'novalidate': <unsigned int> TidyAttr_NOVALIDATE,
     'nowrap': <unsigned int> TidyAttr_NOWRAP,
     'object': <unsigned int> TidyAttr_OBJECT,
+    'opacity': <unsigned int> TidyAttr_OPACITY,
     'open': <unsigned int> TidyAttr_OPEN,
     'optimum': <unsigned int> TidyAttr_OPTIMUM,
     'on_abort': <unsigned int> TidyAttr_OnABORT,
@@ -379,6 +386,7 @@ _AttrId = IntEnum('AttrId', {
     'showgridy': <unsigned int> TidyAttr_SHOWGRIDY,
     'size': <unsigned int> TidyAttr_SIZE,
     'sizes': <unsigned int> TidyAttr_SIZES,
+    'slot': <unsigned int> TidyAttr_SLOT,
     'span': <unsigned int> TidyAttr_SPAN,
     'spellcheck': <unsigned int> TidyAttr_SPELLCHECK,
     'src': <unsigned int> TidyAttr_SRC,
@@ -388,6 +396,14 @@ _AttrId = IntEnum('AttrId', {
     'standby': <unsigned int> TidyAttr_STANDBY,
     'start': <unsigned int> TidyAttr_START,
     'step': <unsigned int> TidyAttr_STEP,
+    'stroke': <unsigned int> TidyAttr_STROKE,
+    'strokedasharray': <unsigned int> TidyAttr_STROKEDASHARRAY,
+    'strokedashoffset': <unsigned int> TidyAttr_STROKEDASHOFFSET,
+    'strokelinecap': <unsigned int> TidyAttr_STROKELINECAP,
+    'strokelinejoin': <unsigned int> TidyAttr_STROKELINEJOIN,
+    'strokemiterlimit': <unsigned int> TidyAttr_STROKEMITERLIMIT,
+    'strokeopacity': <unsigned int> TidyAttr_STROKEOPACITY,
+    'strokewidth': <unsigned int> TidyAttr_STROKEWIDTH,
     'style': <unsigned int> TidyAttr_STYLE,
     'summary': <unsigned int> TidyAttr_SUMMARY,
     'tabindex': <unsigned int> TidyAttr_TABINDEX,
@@ -1188,6 +1204,7 @@ _TagId = IntEnum('TagId', {
     'select': <unsigned int> TidyTag_SELECT,
     'server': <unsigned int> TidyTag_SERVER,
     'servlet': <unsigned int> TidyTag_SERVLET,
+    'slot': <unsigned int> TidyTag_SLOT,
     'small': <unsigned int> TidyTag_SMALL,
     'source': <unsigned int> TidyTag_SOURCE,
     'spacer': <unsigned int> TidyTag_SPACER,

lib/native.hpp

@@ -132,5 +132,49 @@ int call_bool_fn_1(PyObject *fn, PyObject *arg1) {
     return truthy;
 }
 
+
+template <typename T>
+struct has_ob_shash {
+    template <typename C> static std::uint8_t test(decltype(&C::ob_shash)) ;
+    template <typename C> static std::uint64_t test(...);
+    enum { value = sizeof(test<T>(0)) == sizeof(std::uint8_t) };
+};
+
+template <typename T>
+struct has_hash {
+    template <typename C> static std::uint8_t test(decltype(&C::hash)) ;
+    template <typename C> static std::uint64_t test(...);
+    enum { value = sizeof(test<T>(0)) == sizeof(std::uint8_t) };
+};
+
+template<class T, bool ob_shash = has_ob_shash<T>::value, bool hash = has_hash<T>::value>
+struct ResetHash_;
+
+template<class T>
+struct ResetHash_ <T, true, false> {
+    static inline void reset(T *obj) {
+        obj->ob_shash = -1;  // CPython: str
+    }
+};
+
+template<class T>
+struct ResetHash_ <T, false, true> {
+    static inline void reset(T *obj) {
+        obj->hash = -1;  // CPython: bytes
+    }
+};
+
+template<class T>
+struct ResetHash_ <T, false, false> {
+    static inline void reset(T *obj) {
+        (void) 0;  // PyPy
+    }
+};
+
+template <class T>
+static inline void reset_hash(T *obj) {
+    ResetHash_<T>::reset(obj);
+}
+
 }
 }