def FormatCode(unformatted_source,
filename='<unknown>',
style_config=None,
lines=None,
print_diff=False,
verify=False):
"""Format a string of Python code.
This provides an alternative entry point to YAPF.
Arguments:
unformatted_source: (unicode) The code to format.
filename: (unicode) The name of the file being reformatted.
remaining arguments: see comment at the top of this module.
Returns:
Tuple of (reformatted_source, changed). reformatted_source conforms to the
desired formatting style. changed is True if the source changed.
"""
_CheckPythonVersion()
style.SetGlobalStyle(style.CreateStyleFromConfig(style_config))
if not unformatted_source.endswith('\n'):
unformatted_source += '\n'
try:
tree = pytree_utils.ParseCodeToTree(unformatted_source)
except parse.ParseError as e:
raise parse.ParseError(filename + ': ' + e.message)
# Run passes on the tree, modifying it in place.
comment_splicer.SpliceComments(tree)
continuation_splicer.SpliceContinuations(tree)
subtype_assigner.AssignSubtypes(tree)
split_penalty.ComputeSplitPenalties(tree)
blank_line_calculator.CalculateBlankLines(tree)
uwlines = pytree_unwrapper.UnwrapPyTree(tree)
_RecalTokenColumn(uwlines, unformatted_source)
for uwl in uwlines:
uwl.CalculateFormattingInformation()
lines = _LineRangesToSet(lines)
_MarkLinesToFormat(uwlines, lines)
reformatted_source = reformatter.Reformat(_SplitSemicolons(uwlines),
verify, lines)
if unformatted_source == reformatted_source:
return '' if print_diff else reformatted_source, False
code_diff = _GetUnifiedDiff(unformatted_source,
reformatted_source,
filename=filename)
if print_diff:
return code_diff, code_diff.strip() != '' # pylint: disable=g-explicit-bool-comparison
return reformatted_source, True
def parse_tokens(self, tokens, debug=False):
"""Parse a series of tokens and return the syntax tree."""
# XXX Move the prefix computation into a wrapper around tokenize.
p = parse.Parser(self.grammar, self.convert)
p.setup()
lineno = 1
column = 0
type = value = start = end = line_text = None
prefix = ""
for quintuple in tokens:
type, value, start, end, line_text = quintuple
if start != (lineno, column):
assert (lineno, column) <= start, ((lineno, column), start)
s_lineno, s_column = start
if lineno < s_lineno:
prefix += "\n" * (s_lineno - lineno)
lineno = s_lineno
column = 0
if column < s_column:
prefix += line_text[column:s_column]
column = s_column
if type in (tokenize.COMMENT, tokenize.NL):
prefix += value
lineno, column = end
if value.endswith("\n"):
lineno += 1
column = 0
continue
if type == token.OP:
type = grammar.opmap[value]
if debug:
self.logger.debug("%s %r (prefix=%r)",
token.tok_name[type], value, prefix)
if p.addtoken(type, value, (prefix, start)):
if debug:
self.logger.debug("Stop.")
break
prefix = ""
lineno, column = end
if value.endswith("\n"):
lineno += 1
column = 0
else:
# We never broke out -- EOF is too soon (how can this happen???)
raise parse.ParseError("incomplete input",
type, value, (prefix, start))
return p.rootnode