Compile html5lib with Cython

MANIFEST.in

@@ -7,4 +7,5 @@ include .pytest.expect
 include tox.ini
 include pytest.ini
 graft html5lib/tests/testdata
+recursive-include html5lib *.pxd
 recursive-include html5lib/tests *.py

benchmarks/bench_html.py

@@ -4,7 +4,7 @@
 
 import pyperf
 
-sys.path[0:0] = [os.path.join(os.path.dirname(__file__), "..")]
+#sys.path[0:0] = [os.path.join(os.path.dirname(__file__), "..")]
 import html5lib  # noqa: E402
 
 
@@ -49,9 +49,9 @@ def add_cmdline_args(cmd, args):
         source = io.BytesIO(fh.read())
 
     if "parse" in benchmarks:
-        for tb in ("etree", "dom", "lxml"):
+        for tb in ("etree",):
             runner.bench_func("html_parse_%s" % tb, bench_parse, source, tb)
 
-    if "serialize" in benchmarks:
-        for tb in ("etree", "dom", "lxml"):
-            runner.bench_time_func("html_serialize_%s" % tb, bench_serialize, source, tb)
+    # if "serialize" in benchmarks:
+    #     for tb in ("etree",):
+    #         runner.bench_time_func("html_serialize_%s" % tb, bench_serialize, source, tb)

benchmarks/bench_wpt.py

@@ -4,7 +4,7 @@
 
 import pyperf
 
-sys.path[0:0] = [os.path.join(os.path.dirname(__file__), "..")]
+#sys.path[0:0] = [os.path.join(os.path.dirname(__file__), "..")]
 import html5lib  # noqa: E402
 
 

html5lib/_inputstream.pxd

@@ -0,0 +1,109 @@
+# cython: language_level=3
+cimport cython
+from cpython cimport array
+
+ctypedef void (*rCEf)(HTMLUnicodeInputStream, unicode) except *
+
+cdef dict charsUntilCache
+
+cdef class BufferedStream(object):
+    cdef object stream
+    cdef object buffer
+    cdef object position
+    cpdef object tell(self)
+    cpdef object seek(self, object pos)
+    cpdef object read(self, object bytes)
+    cdef object _bufferedBytes(self)
+    cdef object _readStream(self, object bytes)
+    cdef object _readFromBuffer(self, object bytes)
+
+#def HTMLInputStream(source, object **kwargs)
+
+cdef class HTMLUnicodeInputStream(object):
+    cdef rCEf reportCharacterErrors
+    cdef object newLines
+    cdef readonly object charEncoding
+    cdef object dataStream
+    cdef unicode chunk
+    cdef Py_ssize_t chunkSize
+    cdef Py_ssize_t chunkOffset
+    cdef readonly list errors
+
+    # number of (complete) lines in previous chunks
+    cdef Py_ssize_t prevNumLines
+    # number of columns in the last line of the previous chunk
+    cdef Py_ssize_t prevNumCols
+
+    # Deal with CR LF and surrogates split over chunk boundaries
+    cdef unicode _bufferedCharacter
+
+    cdef object reset(self)
+    cdef object openStream(self, object source)
+
+    @cython.locals(nLines=Py_ssize_t, lastLinePos=Py_ssize_t)
+    cdef tuple _position(self, Py_ssize_t offset)
+    cpdef tuple position(self)
+
+    @cython.locals(chunkOffset=Py_ssize_t, char=unicode)
+    cpdef unicode char(self)
+
+    @cython.locals(data=unicode)
+    cdef bint readChunk(self, Py_ssize_t chunkSize=?) except? -1
+
+    @cython.locals(c=ulong)
+    cdef void characterErrorsUCS4(self, unicode data) except *
+    cdef void characterErrorsUCS2(self, unicode data) except *
+
+    cpdef object charsUntil(self, object characters, bint opposite=?)
+    cpdef object unget(self, object char)
+
+cdef class HTMLBinaryInputStream(HTMLUnicodeInputStream):
+    cdef object rawStream
+    cdef readonly object numBytesMeta
+    cdef readonly object numBytesChardet
+    cdef object override_encoding
+    cdef object transport_encoding
+    cdef object same_origin_parent_encoding
+    cdef object likely_encoding
+    cdef object default_encoding
+    cdef object reset(self)
+    cdef object openStream(self, object source)
+    cdef object determineEncoding(self, object chardet=?)
+    cpdef object changeEncoding(self, object newEncoding)
+    @cython.locals(string=bytes)
+    cdef object detectBOM(self)
+    cdef object detectEncodingMeta(self)
+
+# cdef class EncodingBytes(bytes):
+#     cdef object previous(self)
+#     cdef object setPosition(self, object position)
+#     cdef object getPosition(self)
+#     cdef object getCurrentByte(self)
+#     cdef object skip(self, object chars=?)
+#     cdef object skipUntil(self, object chars)
+#     cdef object matchBytes(self, object bytes)
+#     cdef object jumpTo(self, object bytes)
+
+ctypedef bint (*encstate)(EncodingParser) except? -1
+
+cdef class EncodingParser(object):
+    cdef object data
+    cdef object encoding
+
+    @cython.locals(func=encstate, keepParsing=bint)
+    cdef object getEncoding(self)
+    cdef bint handleComment(self) except? -1
+    @cython.locals(hasPragma=bint, name=bytes, value=bytes, tentativeEncoding=bytes)
+    cdef bint handleMeta(self) except? -1
+    cdef bint handlePossibleStartTag(self) except? -1
+    cdef bint handlePossibleEndTag(self) except? -1
+    cdef bint handlePossibleTag(self, bint endTag) except? -1
+    cdef bint handleOther(self) except? -1
+    @cython.locals(c=bytes)
+    cdef tuple getAttribute(self)
+
+cdef class ContentAttrParser(object):
+    cdef object data
+    cpdef object parse(self)  # this needs to be cpdef for tests
+
+cdef object lookupEncoding(object encoding)