Merge branch 'lxml-5.2'

.github/workflows/ci.yml

@@ -129,6 +129,10 @@ jobs:
         exclude:
           - os: ubuntu-latest
             python-version: "3.6"
+          - os: macos-latest
+            python-version: "3.6"
+          - os: macos-latest
+            python-version: "3.7"
 
           # Windows sub-jobs
           # ==============
@@ -213,42 +217,3 @@ jobs:
           name: pycoverage_html
           path: coverage*
           if-no-files-found: ignore
-
-      - name: Upload Wheel
-        uses: actions/upload-artifact@65462800fd760344b1a7b4382951275a0abb4808 # v4.3.3
-        if: matrix.env.STATIC_DEPS && !matrix.extra_hash
-        with:
-          name: wheels-${{ runner.os }}-${{ matrix.python-version }}
-          path: dist/*.whl
-          if-no-files-found: ignore
-
-  collect-wheels:
-    needs: [ci]
-    runs-on: ubuntu-latest
-    steps:
-      - name: Collect wheels
-        uses: actions/download-artifact@65a9edc5881444af0b9093a5e628f2fe47ea3b2e # v4.1.7
-        with:
-          path: ~/downloads
-          merge-multiple: true
-
-      - name: List downloaded artifacts
-        run: ls -la ~/downloads
-
-      - name: Upload Linux wheels
-        uses: actions/upload-artifact@65462800fd760344b1a7b4382951275a0abb4808 # v4.3.3
-        with:
-          name: wheels-linux
-          path: ~/downloads/*linux*.whl
-
-      - name: Upload macOS wheels
-        uses: actions/upload-artifact@65462800fd760344b1a7b4382951275a0abb4808 # v4.3.3
-        with:
-          name: wheels-macosx
-          path: ~/downloads/*macosx*.whl
-
-      - name: Upload Windows wheels
-        uses: actions/upload-artifact@65462800fd760344b1a7b4382951275a0abb4808 # v4.3.3
-        with:
-          name: wheels-windows
-          path: ~/downloads/*-win*.whl

CHANGES.txt

@@ -25,6 +25,11 @@ Bugs fixed
 * GH#417: The ``test_feed_parser`` test could fail if ``lxml_html_clean`` was not installed.
   It is now skipped in that case.
 
+* LP#2059910: The minimum CPU architecture for the Linux x86 binary wheels was set back to
+  "core2", without SSE 4.2.
+
+* If libxml2 uses iconv, the compile time version is available as `etree.ICONV_COMPILED_VERSION`.
+
 
 5.2.1 (2024-04-02)
 ==================

Makefile

@@ -65,7 +65,7 @@ wheel_%: dist/lxml-$(LXMLVERSION).tar.gz
 		-e AR=gcc-ar \
 		-e NM=gcc-nm \
 		-e RANLIB=gcc-ranlib \
-		-e CFLAGS="$(MANYLINUX_CFLAGS) $(if $(patsubst %aarch64,,$@),-march=core2 -msse4.1 -msse4.2,-march=armv8-a -mtune=cortex-a72)" \
+		-e CFLAGS="$(MANYLINUX_CFLAGS) $(if $(patsubst %aarch64,,$@),-march=core2,-march=armv8-a -mtune=cortex-a72)" \
 		-e LDFLAGS="$(MANYLINUX_LDFLAGS)" \
 		-e STATIC_DEPS="${STATIC_DEPS}" \
 		-e LIBXML2_VERSION="$(MANYLINUX_LIBXML2_VERSION)" \

pyproject.toml

@@ -8,7 +8,7 @@ skip = [
     "pp*-manylinux_i686",
     "*-musllinux_i686",
     # Py3.8 wheel for macos is not universal: https://bugs.launchpad.net/lxml/+bug/2055404
-    "cp38_macosx_universal2",
+    "cp38-macosx_universal2",
     # Reduce job load and HTTP hit rate on library servers.
     "cp36-manylinux_aarch64",
     "cp37-manylinux_aarch64",
@@ -51,7 +51,7 @@ environment.CFLAGS = "-O3 -g1 -pipe -fPIC -flto -march=core2 -mtune=generic"
 [[tool.cibuildwheel.overrides]]
 select = "*linux_x86_64"
 inherit.environment = "append"
-environment.CFLAGS = "-O3 -g1 -pipe -fPIC -flto -march=core2 -msse4.1 -msse4.2 -mtune=generic"
+environment.CFLAGS = "-O3 -g1 -pipe -fPIC -flto -march=core2 -mtune=generic"
 
 [[tool.cibuildwheel.overrides]]
 select = "*aarch64"

src/lxml/etree.pyx

@@ -232,7 +232,7 @@ cdef class C14NError(LxmlError):
     """
 
 # version information
-cdef __unpackDottedVersion(version):
+cdef tuple __unpackDottedVersion(version):
     version_list = []
     l = (version.decode("ascii").replace('-', '.').split('.') + [0]*4)[:4]
     for item in l:
@@ -255,11 +255,11 @@ cdef __unpackDottedVersion(version):
         version_list.append(item)
     return tuple(version_list)
 
-cdef __unpackIntVersion(int c_version):
+cdef tuple __unpackIntVersion(int c_version, int base=100):
     return (
-        ((c_version // (100*100)) % 100),
-        ((c_version // 100)       % 100),
-        (c_version                % 100)
+        ((c_version // (base*base)) % base),
+        ((c_version // base)        % base),
+        (c_version                  % base)
         )
 
 cdef int _LIBXML_VERSION_INT
@@ -276,6 +276,26 @@ LXML_VERSION = __unpackDottedVersion(tree.LXML_VERSION_STRING)
 
 __version__ = tree.LXML_VERSION_STRING.decode("ascii")
 
+cdef extern from *:
+    """
+    #ifdef ZLIB_VERNUM
+      #define __lxml_zlib_version (ZLIB_VERNUM >> 4)
+    #else
+      #define __lxml_zlib_version 0
+    #endif
+    #ifdef _LIBICONV_VERSION
+      #define __lxml_iconv_version (_LIBICONV_VERSION << 8)
+    #else
+      #define __lxml_iconv_version 0
+    #endif
+    """
+    # zlib isn't included automatically by libxml2's headers
+    #long ZLIB_HEX_VERSION "__lxml_zlib_version"
+    long LIBICONV_HEX_VERSION "__lxml_iconv_version"
+
+#ZLIB_COMPILED_VERSION = __unpackIntVersion(ZLIB_HEX_VERSION, base=0x10)
+ICONV_COMPILED_VERSION = __unpackIntVersion(LIBICONV_HEX_VERSION, base=0x100)[:2]
+
 
 # class for temporary storage of Python references,
 # used e.g. for XPath results

src/tests/test_etree.py

@@ -28,18 +28,19 @@
 from .common_imports import canonicalize, _str, _bytes
 from .common_imports import SimpleFSPath
 
-print("""
-TESTED VERSION: %s""" % etree.__version__ + """
-    Python:           %r""" % (sys.version_info,) + """
-    lxml.etree:       %r""" % (etree.LXML_VERSION,) + """
-    libxml used:      %r""" % (etree.LIBXML_VERSION,) + """
-    libxml compiled:  %r""" % (etree.LIBXML_COMPILED_VERSION,) + """
-    libxslt used:     %r""" % (etree.LIBXSLT_VERSION,) + """
-    libxslt compiled: %r""" % (etree.LIBXSLT_COMPILED_VERSION,) + """
-    FS encoding:      %s""" % (sys.getfilesystemencoding(),) + """
-    Default encoding: %s""" % (sys.getdefaultencoding(),) + """
-    Max Unicode:      %s""" % (sys.maxunicode,) + """
-    PyUCS4 encoding:  %s""" % (getattr(etree, '_pyucs4_encoding_name', ''),) + """
+print(f"""
+TESTED VERSION: {etree.__version__}
+    Python:           {tuple(sys.version_info)!r}
+    lxml.etree:       {etree.LXML_VERSION!r}
+    libxml used:      {etree.LIBXML_VERSION!r}
+    libxml compiled:  {etree.LIBXML_COMPILED_VERSION!r}
+    libxslt used:     {etree.LIBXSLT_VERSION!r}
+    libxslt compiled: {etree.LIBXSLT_COMPILED_VERSION!r}
+    iconv compiled:   {etree.ICONV_COMPILED_VERSION!r}
+    FS encoding:      {sys.getfilesystemencoding()}
+    Default encoding: {sys.getdefaultencoding()}
+    Max Unicode:      {sys.maxunicode}
+    PyUCS4 encoding:  {getattr(etree, '_pyucs4_encoding_name', '')}
 """)
 
 

tools/ci-run.sh

@@ -78,7 +78,8 @@ fi
 if [ -z "${PYTHON_VERSION##2*}" ] || [ -z "${PYTHON_VERSION##pypy-2*}" ]; then
   python -m pip install -U beautifulsoup4==4.9.3 cssselect==1.1.0 html5lib==1.1 rnc2rng==2.6.5 ${EXTRA_DEPS} || exit 1
 else
-  python -m pip install -U beautifulsoup4 cssselect html5lib lxml_html_clean rnc2rng ${EXTRA_DEPS} || exit 1
+  python -m pip install -U beautifulsoup4 cssselect html5lib rnc2rng ${EXTRA_DEPS} || exit 1
+  python -m pip install --no-deps lxml_html_clean || exit 1
 fi
 if [[ "$COVERAGE" == "true" ]]; then
   python -m pip install "coverage<5" || exit 1
@@ -93,8 +94,6 @@ GITHUB_API_TOKEN="${SAVED_GITHUB_API_TOKEN}" \
       $(if [[ "$COVERAGE" == "true" ]]; then echo -n " --with-coverage"; fi ) \
       || exit 1
 
-ccache -s || true
-
 # Run tests
 echo "Running the tests ..."
 GITHUB_API_TOKEN="${SAVED_GITHUB_API_TOKEN}" \
@@ -103,12 +102,4 @@ GITHUB_API_TOKEN="${SAVED_GITHUB_API_TOKEN}" \
       PYTHONUNBUFFERED=x \
       make test || exit 1
 
-if [[ "$COVERAGE" != "true" ]]; then
-  echo "Building a clean wheel ..."
-  GITHUB_API_TOKEN="${SAVED_GITHUB_API_TOKEN}" \
-        CFLAGS="$EXTRA_CFLAGS -O3 -g1 -mtune=generic -fPIC -flto" \
-        LDFLAGS="-flto $EXTRA_LDFLAGS" \
-        make clean wheel || exit 1
-fi
-
 ccache -s || true