#19480: merge with 3.3.

diff --cc Lib/html/parser.py
Simple merge

diff --cc Lib/test/test_htmlparser.py
Simple merge

diff --cc Misc/NEWS
index 2c385391599c6e37e629c0581c72cd9a971b4aaf,e4e2e140b4e9fa8c17796958cc792145b3d82b4e..d76c48464f3931aa36e37f5bc2aa349697dd09d5
--- 1/Misc/NEWS
--- 2/Misc/NEWS
+++ b/Misc/NEWS
@@@ -25,63 -56,75 +25,66 @@@ Core and Builtin
  - Issue #18603: Ensure that PyOS_mystricmp and PyOS_mystrnicmp are in the
    Python executable and not removed by the linker's optimizer.
  
 -- Issue #19279: UTF-7 decoder no more produces illegal strings.
 -
 -- Fix macro expansion of _PyErr_OCCURRED(), and make sure to use it in at
 -  least one place so as to avoid regressions.
 -
 -- Issue #19014: memoryview.cast() is now allowed on zero-length views.
 -
 -- Issue #19098: Prevent overflow in the compiler when the recursion limit is set
 -  absurdly high.
 -
 -- Issue #18942: sys._debugmallocstats() output was damaged on Windows.
 -
 -- Issue #18667: Add missing "HAVE_FCHOWNAT" symbol to posix._have_functions.
 -
 -- Issue #18368: PyOS_StdioReadline() no longer leaks memory when realloc()
 -  fails.
 +- Issue #19306: Add extra hints to the faulthandler module's stack
 +  dumps that these are "upside down".
  
 -- Issue #16741: Fix an error reporting in int().
 +Library
 +-------
  
 -- Issue #17899: Fix rare file descriptor leak in os.listdir().
 +- Issue #19378: Fixed a number of cases in the dis module where the new
 +  "file" parameter was not being honoured correctly
  
 -- Issue #18552: Check return value of PyArena_AddPyObject() in
 -  obj2ast_object().
 +- Issue #19378: Removed the "dis.Bytecode.show_info" method
  
 -- Issue #18560: Fix potential NULL pointer dereference in sum().
 +- Issue #19378: Renamed the "dis.Bytecode.display_code" method to
 +  "dis.Bytecode.dis" and converted it to returning a string rather than
 +  printing output.
  
 -- Issue #15905: Fix theoretical buffer overflow in handling of sys.argv[0],
 -  prefix and exec_prefix if the operation system does not obey MAXPATHLEN.
 +- Issue #19378: the "line_offset" parameter in the new "dis.get_instructions"
 +  API has been renamed to "first_line" (and the default value and usage
 +  changed accordingly). This should reduce confusion with the more common use
 +  of "offset" in the dis docs to refer to bytecode offsets.
  
 -- Issue #18344: Fix potential ref-leaks in _bufferedreader_read_all().
 +- Issue #18678: Corrected spwd struct member names in spwd module:
 +  sp_nam->sp_namp, and sp_pwd->sp_pwdp.  The old names are kept as extra
 +  structseq members, for backward compatibility.
  
 -- Issue #17872: Fix a segfault in marshal.load() when input stream returns
 -  more bytes than requested.
 +- Issue #6157: Fixed tkinter.Text.debug().  tkinter.Text.bbox() now raises
 +  TypeError instead of TclError on wrong number of arguments.  Original patch
 +  by Guilherme Polo.
  
 -- Issue #18426: Fix NULL pointer dereference in C extension import when
 -  PyModule_GetDef() returns an error.
 +- Issue #10197: Rework subprocess.get[status]output to use subprocess
 +  functionality and thus to work on Windows. Patch by Nick Coghlan
  
 -- Issue #18328: Reorder ops in PyThreadState_Delete*() functions. Now the
 -  tstate is first removed from TLS and then deallocated.
 +- Issue #6160: The bbox() method of tkinter.Spinbox now returns a tuple of
 +  integers instead of a string.  Based on patch by Guilherme Polo.
  
 -- Issue #18184: PyUnicode_FromFormat() and PyUnicode_FromFormatV() now raise
 -  OverflowError when an argument of %c format is out of range.
 +- Issue #19403: contextlib.redirect_stdout is now reentrant
  
 -- Issue #18137: Detect integer overflow on precision in float.__format__()
 -  and complex.__format__().
 +- Issue #19286: Directories in ``package_data`` are no longer added to
 +  the filelist, preventing failure outlined in the ticket.
  
 -- Issue #18183: Fix various unicode operations on strings with large unicode
 -  codepoints.
++- Issue #19480: HTMLParser now accepts all valid start-tag names as defined
++  by the HTML5 standard.
+ 
 -- Issue #18180: Fix ref leak in _PyImport_GetDynLoadWindows().
 +- Issue #15114: The html.parser module now raises a DeprecationWarning when the
 +  strict argument of HTMLParser or the HTMLParser.error method are used.
  
 -- Issue #18038: SyntaxError raised during compilation sources with illegal
 -  encoding now always contains an encoding name.
 +- Issue #19410: Undo the special-casing removal of '' for
 +  importlib.machinery.FileFinder.
  
 -- Issue #17644: Fix a crash in str.format when curly braces are used in square
 -  brackets.
 +- Issue #19424: Fix the warnings module to accept filename containing surrogate
 +  characters.
  
 -- Issue #17983: Raise a SyntaxError for a ``global __class__`` statement in a
 -  class body.
 +- Issue #19435: Fix directory traversal attack on CGIHttpRequestHandler.
  
 -- Issue #17927: Frame objects kept arguments alive if they had been copied into
 -  a cell, even if the cell was cleared.
 +- Issue #19227: Remove pthread_atfork() handler. The handler was added to
 +  solve #18747 but has caused issues.
  
 -Library
 --------
 +- Issue #19420: Fix reference leak in module initalization code of
 +  _hashopenssl.c
  
 -- Issue #19395: Raise an exception when attempting to pickle a bz2 or lzma
 -  compressor/decompressor object, rather than creating a pickle that would
 -  cause a segfault when loaded and used.
 +- Issue #19329: Optimized compiling charsets in regular expressions.
  
  - Issue #19227: Try to fix deadlocks caused by re-seeding then OpenSSL
    pseudo-random number generator on fork().