"""Provide advanced parsing abilities for ParenMatch and other extensions.
HyperParser uses PyParser. PyParser mostly gives information on the
proper indentation of code. HyperParser gives additional information on
from idlelib import PyParse
def __init__(self, editwin, index):
"To initialize, analyze the surroundings of the given index."
self.text = text = editwin.text
parser = PyParse.Parser(editwin.indentwidth, editwin.tabwidth)
lno = index2line(text.index(index))
if not editwin.context_use_ps1:
for context in editwin.num_context_lines:
startat = max(lno - context, 1)
startatindex = repr(startat) + ".0"
stopatindex = "%d.end" % lno
# We add the newline because PyParse requires a newline
# at end. We add a space so that index won't be at end
# of line, so that its status will be the same as the
# char before it, if should.
parser.set_str(text.get(startatindex, stopatindex)+' \n')
bod = parser.find_good_parse_start(
editwin._build_char_in_string_func(startatindex))
if bod is not None or startat == 1:
r = text.tag_prevrange("console", index)
stopatindex = "%d.end" % lno
# We add the newline because PyParse requires it. We add a
# space so that index won't be at end of line, so that its
# status will be the same as the char before it, if should.
parser.set_str(text.get(startatindex, stopatindex)+' \n')
# We want what the parser has, minus the last newline and space.
self.rawtext = parser.str[:-2]
# Parser.str apparently preserves the statement we are in, so
# that stopatindex can be used to synchronize the string with
self.stopatindex = stopatindex
self.bracketing = parser.get_last_stmt_bracketing()
# find which pairs of bracketing are openers. These always
# correspond to a character of rawtext.
self.isopener = [i>0 and self.bracketing[i][1] >
for i in range(len(self.bracketing))]
def set_index(self, index):
"""Set the index to which the functions relate.
The index must be in the same statement.
indexinrawtext = (len(self.rawtext) -
len(self.text.get(index, self.stopatindex)))
raise ValueError("Index %s precedes the analyzed statement"
self.indexinrawtext = indexinrawtext
# find the rightmost bracket to which index belongs
while (self.indexbracket < len(self.bracketing)-1 and
self.bracketing[self.indexbracket+1][0] < self.indexinrawtext):
if (self.indexbracket < len(self.bracketing)-1 and
self.bracketing[self.indexbracket+1][0] == self.indexinrawtext and
not self.isopener[self.indexbracket+1]):
"""Is the index given to the HyperParser in a string?"""
# The bracket to which we belong should be an opener.
# If it's an opener, it has to have a character.
return (self.isopener[self.indexbracket] and
self.rawtext[self.bracketing[self.indexbracket][0]]
"""Is the index given to the HyperParser in normal code?"""
return (not self.isopener[self.indexbracket] or
self.rawtext[self.bracketing[self.indexbracket][0]]
def get_surrounding_brackets(self, openers='([{', mustclose=False):
"""Return bracket indexes or None.
If the index given to the HyperParser is surrounded by a
bracket defined in openers (or at least has one before it),
return the indices of the opening bracket and the closing
bracket (or the end of line, whichever comes first).
If it is not surrounded by brackets, or the end of line comes
before the closing bracket and mustclose is True, returns None.
bracketinglevel = self.bracketing[self.indexbracket][1]
before = self.indexbracket
while (not self.isopener[before] or
self.rawtext[self.bracketing[before][0]] not in openers or
self.bracketing[before][1] > bracketinglevel):
bracketinglevel = min(bracketinglevel, self.bracketing[before][1])
after = self.indexbracket + 1
while (after < len(self.bracketing) and
self.bracketing[after][1] >= bracketinglevel):
beforeindex = self.text.index("%s-%dc" %
(self.stopatindex, len(self.rawtext)-self.bracketing[before][0]))
if (after >= len(self.bracketing) or
self.bracketing[after][0] > len(self.rawtext)):
afterindex = self.stopatindex
# We are after a real char, so it is a ')' and we give the
afterindex = self.text.index(
"%s-%dc" % (self.stopatindex,
len(self.rawtext)-(self.bracketing[after][0]-1)))
return beforeindex, afterindex
# Ascii chars that may be in a white space
_whitespace_chars = " \t\n\\"
# Ascii chars that may be in an identifier
_id_chars = string.ascii_letters + string.digits + "_"
# Ascii chars that may be the first char of an identifier
_id_first_chars = string.ascii_letters + "_"
# Given a string and pos, return the number of chars in the
# identifier which ends at pos, or 0 if there is no such one. Saved
# words are not identifiers.
def _eat_identifier(self, str, limit, pos):
while i > limit and str[i-1] in self._id_chars:
if (i < pos and (str[i] not in self._id_first_chars or
keyword.iskeyword(str[i:pos]))):
def get_expression(self):
"""Return a string with the Python expression which ends at the
given index, which is empty if there is no real one.
if not self.is_in_code():
raise ValueError("get_expression should only be called "
"if index is inside a code.")
bracketing = self.bracketing
brck_index = self.indexbracket
brck_limit = bracketing[brck_index][0]
pos = self.indexinrawtext
last_identifier_pos = pos
# Eat whitespaces, comments, and if postdot_phase is False - a dot
if pos>brck_limit and rawtext[pos-1] in self._whitespace_chars:
elif (not postdot_phase and
pos > brck_limit and rawtext[pos-1] == '.'):
# The next line will fail if we are *inside* a comment,
elif (pos == brck_limit and brck_index > 0 and
rawtext[bracketing[brck_index-1][0]] == '#'):
brck_limit = bracketing[brck_index][0]
pos = bracketing[brck_index+1][0]
# If we didn't eat anything, quit.
# We didn't find a dot, so the expression end at the
ret = self._eat_identifier(rawtext, brck_limit, pos)
# There is an identifier to eat
last_identifier_pos = pos
# Now, to continue the search, we must find a dot.
# (the loop continues now)
# We are at a bracketing limit. If it is a closing
# bracket, eat the bracket, otherwise, stop the search.
level = bracketing[brck_index][1]
while brck_index > 0 and bracketing[brck_index-1][1] > level:
if bracketing[brck_index][0] == brck_limit:
# We were not at the end of a closing bracket
pos = bracketing[brck_index][0]
brck_limit = bracketing[brck_index][0]
last_identifier_pos = pos
# [] and () may be used after an identifier, so we
# continue. postdot_phase is True, so we don't allow a dot.
# We can't continue after other types of brackets
if rawtext[pos] in "'\"":
while pos > 0 and rawtext[pos - 1] in "rRbBuU":
last_identifier_pos = pos
# We've found an operator or something.
return rawtext[last_identifier_pos:self.indexinrawtext]
if __name__ == '__main__':
unittest.main('idlelib.idle_test.test_hyperparser', verbosity=2)
It is recommended that you Edit text format, this type of Fix handles quite a lot in one request
