moculus
parent
90aab85f25
commit
30dcf827aa
|
|
@ -485,14 +485,8 @@ class Application:
|
|||
elif token.x >= x + slot.width:
|
||||
break
|
||||
|
||||
c = w.mode.colors.get(token.name, w.mode.default_color)
|
||||
# c = w.mode.default_color
|
||||
# name_parts = token.name.split('.')
|
||||
# for i in range(0, len(name_parts)):
|
||||
# name = '.'.join(name_parts[i:])
|
||||
# if name in w.mode.colors:
|
||||
# c = w.mode.colors[name]
|
||||
# break
|
||||
name = token.fqname()
|
||||
c = w.mode.colors.get(name, w.mode.default_color)
|
||||
|
||||
if DARK_BACKGROUND:
|
||||
c |= curses.A_BOLD
|
||||
|
|
|
|||
|
|
@ -20,7 +20,6 @@ class Highlighter:
|
|||
def __init__(self, lexer):
|
||||
self.lexer = lexer
|
||||
self.tokens = []
|
||||
self.line_contexts = {}
|
||||
|
||||
def dump(self, fmt='(%3s, %2s) | %s'):
|
||||
print fmt % ('y', 'x', 'string')
|
||||
|
|
@ -50,20 +49,36 @@ class Highlighter:
|
|||
sys.stdout.write(token.string)
|
||||
sys.stdout.write('\n')
|
||||
|
||||
def delete_token(self, y, i):
|
||||
assert y < len(self.tokens), "%d < %d" % (y, len(self.tokens))
|
||||
assert i < len(self.tokens[y]), "%d < %d" % (i, len(self.tokens[i]))
|
||||
deleted = []
|
||||
deleted.append(self.tokens[y].pop(i))
|
||||
while y < len(self.tokens):
|
||||
while i < len(self.tokens[y]):
|
||||
while deleted and self.tokens[y][i].parent is not deleted[-1]:
|
||||
del deleted[-1]
|
||||
if not deleted:
|
||||
return
|
||||
elif self.tokens[y][i].parent is deleted[-1]:
|
||||
deleted.append(self.tokens[y].pop(i))
|
||||
else:
|
||||
raise Exception, "huh?? %r %r" % (self.tokens[y][i].parent,
|
||||
deleted)
|
||||
i = 0
|
||||
y += 1
|
||||
def highlight(self, lines):
|
||||
self.tokens = [[] for l in lines]
|
||||
self.lexer.lex(lines, y=0, x=0)
|
||||
for token in self.lexer:
|
||||
self.tokens[token.y].append(token)
|
||||
self.line_contexts = dict(self.lexer.line_contexts)
|
||||
|
||||
# relexing
|
||||
# ======================
|
||||
def relex(self, lines, y1, x1, y2, x2):
|
||||
# start the relexing process
|
||||
#self.lexer.lex(lines, y1, 0)
|
||||
rulecontexts = self.line_contexts[y1]
|
||||
self.lexer.resume(lines, y1, 0, rulecontexts)
|
||||
token = self.tokens[y1][0]
|
||||
self.lexer.resume(lines, y1, 0, token)
|
||||
|
||||
# these keep track of the current y coordinate, the current token index
|
||||
# on line[y], and the current "new token", respectively.
|
||||
|
|
|
|||
458
lex2.py
458
lex2.py
|
|
@ -3,56 +3,66 @@ import re
|
|||
valid_name_re = re.compile('^[a-zA-Z_][a-zA-Z0-9_]*$')
|
||||
reserved_names = ['start', 'middle', 'end', 'null']
|
||||
|
||||
class RuleContext:
|
||||
# to be clear:
|
||||
# x, y: where the rule processing began
|
||||
# rule: the rule which began
|
||||
# flag: a signal to be used to resume the rule correctly
|
||||
# context: the previous rule namespace(s)
|
||||
# matchd: the dictionary returned by the rule's matching
|
||||
def __init__(self, y, x, rule, flag, context, matchd):
|
||||
self.y = y
|
||||
self.x = x
|
||||
self.rule = rule
|
||||
self.flag = flag
|
||||
self.context = context
|
||||
self.matchd = matchd
|
||||
|
||||
class Token(object):
|
||||
def __init__(self, name, rule, y, x, s, **vargs):
|
||||
def __init__(self, name, rule=None, y=0, x=0, s="", parent=None, matchd={}):
|
||||
self.name = name
|
||||
self.rule = rule
|
||||
self.y = y
|
||||
self.x = x
|
||||
self.string = s
|
||||
self.vargs = vargs
|
||||
self.parent = parent
|
||||
self.matchd = matchd
|
||||
def parents(self):
|
||||
if self.parent is not None:
|
||||
parents = self.parent.parents()
|
||||
parents.append(self.parent)
|
||||
return parents
|
||||
else:
|
||||
return []
|
||||
def domain(self):
|
||||
names = []
|
||||
if self.parent is not None:
|
||||
names.extend(self.parent.domain())
|
||||
names.append(self.rule.name)
|
||||
return names
|
||||
def fqlist(self):
|
||||
names = []
|
||||
if self.parent is not None:
|
||||
names.extend(self.parent.domain())
|
||||
names.append(self.name)
|
||||
return names
|
||||
def fqname(self):
|
||||
if self.name == 'start':
|
||||
names = self.domain()
|
||||
names.append(self.name)
|
||||
else:
|
||||
names = self.fqlist()
|
||||
return '.'.join(names)
|
||||
def copy(self):
|
||||
return Token(self.name, None, self.y, self.x, self.string, **self.vargs)
|
||||
return Token(self.name, self.rule, self.y, self.x, self.string,
|
||||
self.parent, self.matchd)
|
||||
def add_to_string(self, s):
|
||||
self.string += s
|
||||
def end_x(self):
|
||||
return self.x + len(self.string)
|
||||
def __eq__(self, other):
|
||||
return (self.y == other.y and
|
||||
self.x == other.x and
|
||||
self.string == other.string and
|
||||
self.name == other.name and
|
||||
self.vargs == other.vargs)
|
||||
return (self.y == other.y and self.x == other.x
|
||||
and self.name == other.name and self.parent is other.parent and
|
||||
self.string == other.string)
|
||||
def __repr__(self):
|
||||
if len(self.string) < 10:
|
||||
s = self.string
|
||||
else:
|
||||
s = self.string[:10] + '...'
|
||||
return "<Token(%r, %r, %d, %d, %r)>" % (self.name, self.rule, self.y, self.x, s)
|
||||
def render(self):
|
||||
return (self,)
|
||||
fields = (self.fqname(), self.rule, self.y, self.x, s)
|
||||
return "<Token(%r, %r, %d, %d, %r)>" % fields
|
||||
|
||||
class Rule:
|
||||
name = 'abstract'
|
||||
def match(self, lexer, context=[], d={}):
|
||||
def match(self, lexer, parent):
|
||||
raise Exception, "%s rule cannot match!" % self.name
|
||||
def make_token(self, lexer, s, name, **vargs):
|
||||
return Token(name, self, lexer.y, lexer.x, s, **vargs)
|
||||
def make_token(self, lexer, s, name, parent=None, matchd={}):
|
||||
return Token(name, self, lexer.y, lexer.x, s, parent, matchd)
|
||||
|
||||
class ConstantRule(Rule):
|
||||
def __init__(self, name, constant):
|
||||
|
|
@ -60,11 +70,12 @@ class ConstantRule(Rule):
|
|||
assert name not in reserved_names, "reserved rule name: %r" % name
|
||||
self.name = name
|
||||
self.constant = constant
|
||||
def match(self, lexer, context=[], d={}):
|
||||
self.lenth = len(self.constant)
|
||||
def match(self, lexer, parent):
|
||||
if lexer.lines[lexer.y][lexer.x:].startswith(self.constant):
|
||||
name = '.'.join(context + [self.name])
|
||||
lexer.add_token(self.make_token(lexer, self.constant, name, grammar=lexer.grammar))
|
||||
lexer.x += len(self.constant)
|
||||
token = self.make_token(lexer, self.constant, self.name, parent)
|
||||
lexer.add_token(token)
|
||||
lexer.x += self.length
|
||||
return True
|
||||
else:
|
||||
return False
|
||||
|
|
@ -76,17 +87,20 @@ class PatternRule(Rule):
|
|||
self.name = name
|
||||
self.pattern = pattern
|
||||
self.re = re.compile(pattern)
|
||||
def match(self, lexer, context=[], d={}):
|
||||
def _match(self, lexer, parent, m):
|
||||
s = m.group(0)
|
||||
token = self.make_token(lexer, s, self.name, parent)
|
||||
lexer.add_token(token)
|
||||
lexer.x += len(s)
|
||||
def match(self, lexer, parent):
|
||||
m = self.re.match(lexer.lines[lexer.y], lexer.x)
|
||||
if m:
|
||||
name = '.'.join(context + [self.name])
|
||||
lexer.add_token(self.make_token(lexer, m.group(0), name, grammar=lexer.grammar))
|
||||
lexer.x += len(m.group(0))
|
||||
self._match(lexer, parent, m)
|
||||
return True
|
||||
else:
|
||||
return False
|
||||
|
||||
class ContextPatternRule(Rule):
|
||||
class ContextPatternRule(PatternRule):
|
||||
def __init__(self, name, pattern, fallback):
|
||||
assert valid_name_re.match(name), 'invalid name %r' % name
|
||||
assert name not in reserved_names, "reserved rule name: %r" % name
|
||||
|
|
@ -94,16 +108,14 @@ class ContextPatternRule(Rule):
|
|||
self.pattern = pattern
|
||||
self.fallback = fallback
|
||||
self.fallback_re = re.compile(fallback)
|
||||
def match(self, lexer, context=[], d={}):
|
||||
def match(self, lexer, parent):
|
||||
try:
|
||||
r = re.compile(self.pattern % d)
|
||||
r = re.compile(self.pattern % parent.matchd)
|
||||
except KeyError:
|
||||
r = self.fallback_re
|
||||
m = r.match(lexer.lines[lexer.y], lexer.x)
|
||||
if m:
|
||||
name = '.'.join(context + [self.name])
|
||||
lexer.add_token(self.make_token(lexer, m.group(0), name, grammar=lexer.grammar))
|
||||
lexer.x += len(m.group(0))
|
||||
self._match(lexer, parent, m)
|
||||
return True
|
||||
else:
|
||||
return False
|
||||
|
|
@ -117,42 +129,50 @@ class RegionRule(Rule):
|
|||
self.grammar = grammar
|
||||
self.end = end
|
||||
self.start_re = re.compile(start)
|
||||
def _add_from_regex(self, context, name, lexer, m, grammar):
|
||||
t_name = '.'.join(context + [self.name, name])
|
||||
t = self.make_token(lexer, m.group(0), t_name, grammar=grammar)
|
||||
lexer.add_token(t)
|
||||
lexer.x += len(m.group(0))
|
||||
|
||||
def resume(self, lexer, context, flag, d, rulecontexts):
|
||||
assert rulecontexts, "can't resume without rulecontexts!"
|
||||
self._match(lexer, context, d, None, rulecontexts)
|
||||
def resume(self, lexer, toresume):
|
||||
assert toresume, "can't resume without tokens to resume!"
|
||||
self._match(lexer, None, None, toresume)
|
||||
return True
|
||||
|
||||
def match(self, lexer, context=[], d={}):
|
||||
# see if we can match our start token
|
||||
def match(self, lexer, parent):
|
||||
m = self.start_re.match(lexer.lines[lexer.y], lexer.x)
|
||||
if m:
|
||||
# region was match, so let's do this
|
||||
return self._match(lexer, context, m.groupdict(), m, [])
|
||||
self._match(lexer, parent, m, [])
|
||||
return True
|
||||
else:
|
||||
# region was not matched; we never started. so return false
|
||||
return False
|
||||
|
||||
def _match(self, lexer, context, d, m, rulecontext=[]):
|
||||
# if we have been given rulecontext, then we are going to "resume" a
|
||||
# parse that can already be assumed to have started
|
||||
reenter = len(rulecontext) > 0
|
||||
assert m or reenter
|
||||
def _add_from_regex(self, name, lexer, parent, m, matchd={}):
|
||||
s = m.group(0)
|
||||
token = self.make_token(lexer, s, name, parent, matchd)
|
||||
lexer.add_token(token)
|
||||
lexer.x += len(s)
|
||||
return token
|
||||
|
||||
# first let's do some bookkeeping
|
||||
lexer.context.append(RuleContext(lexer.y, lexer.x, self, 'start',
|
||||
list(context), dict(d)))
|
||||
if m is not None:
|
||||
self._add_from_regex(context, 'start', lexer, m, lexer.grammar)
|
||||
null_t_name = '.'.join(context + [self.name, 'null'])
|
||||
def _match(self, lexer, parent, m, toresume=[]):
|
||||
# we either need a match object, or a token to resume
|
||||
assert m or len(toresume) > 0
|
||||
|
||||
if m:
|
||||
# if we had a match, then it becomes the parent, and we save its
|
||||
# subgroup dict
|
||||
d = m.groupdict()
|
||||
parent = self._add_from_regex('start', lexer, parent, m, d)
|
||||
else:
|
||||
# otherwise, we should be resuming the start token, so let's pull
|
||||
# the relevant info out of the token
|
||||
parent = toresume[0]
|
||||
d = parent.matchd
|
||||
assert parent.name == 'start'
|
||||
null_t = None
|
||||
|
||||
# if we have an end token, then build it here. notice that it can
|
||||
# this determines whether we are still reentering. if len(toresume) == 1
|
||||
# then it means that we have been reentering but will not continue, so
|
||||
# reenter will be false.
|
||||
reenter = len(toresume) > 1
|
||||
|
||||
# if we have an end regex, then build it here. notice that it can
|
||||
# reference named groups from the start token. if we have no end,
|
||||
# well, then, we're never getting out of here alive!
|
||||
if self.end:
|
||||
|
|
@ -167,21 +187,19 @@ class RegionRule(Rule):
|
|||
# if this line is empty, then we skip it, but here we insert
|
||||
# an empty null token just so we have something
|
||||
if not reenter and len(lexer.lines[lexer.y]) == 0:
|
||||
null_t = Token(null_t_name, None, lexer.y, lexer.x, '')
|
||||
null_t = Token('null', None, lexer.y, lexer.x, '', parent)
|
||||
lexer.add_token(null_t)
|
||||
null_t = None
|
||||
|
||||
# ok, as long as we haven't found the end token, and have more
|
||||
# data on the current line to read, we will process tokens
|
||||
while not done and lexer.y == old_y and lexer.x < len(lexer.lines[lexer.y]):
|
||||
while (not done and lexer.y == old_y and
|
||||
lexer.x < len(lexer.lines[lexer.y])):
|
||||
# if we are reentering mid-parse, then that takes precedence
|
||||
if reenter:
|
||||
reenter = False
|
||||
rule2 = rulecontext[0].rule
|
||||
context2 = rulecontext[0].context
|
||||
d2 = rulecontext[0].matchd
|
||||
assert rule2.resume(lexer, context2, d2, rulecontext[1:]), \
|
||||
"%r %r %r %r" % (lexer, context2, d2, rulecontext[1:])
|
||||
rule2 = toresume[1].rule
|
||||
rule2.resume(lexer, toresume[1:])
|
||||
found = True
|
||||
null_t = None
|
||||
break
|
||||
|
|
@ -191,7 +209,7 @@ class RegionRule(Rule):
|
|||
if self.end:
|
||||
m = end_re.match(lexer.lines[lexer.y], lexer.x)
|
||||
if m:
|
||||
self._add_from_regex(context, 'end', lexer, m, None)
|
||||
self._add_from_regex('end', lexer, parent, m, {})
|
||||
done = True
|
||||
break
|
||||
|
||||
|
|
@ -199,7 +217,7 @@ class RegionRule(Rule):
|
|||
# find a token, note that we found one and exit the loop
|
||||
found = False
|
||||
for rule in self.grammar.rules:
|
||||
if rule.match(lexer, context + [self.name], d):
|
||||
if rule.match(lexer, parent):
|
||||
found = True
|
||||
null_t = None
|
||||
break
|
||||
|
|
@ -209,7 +227,7 @@ class RegionRule(Rule):
|
|||
# create if it isn't set).
|
||||
if not found:
|
||||
if null_t is None:
|
||||
null_t = Token(null_t_name, None, lexer.y, lexer.x, '')
|
||||
null_t = Token('null', None, lexer.y, lexer.x, '', parent)
|
||||
lexer.add_token(null_t)
|
||||
null_t.add_to_string(lexer.lines[lexer.y][lexer.x])
|
||||
lexer.x += 1
|
||||
|
|
@ -223,12 +241,9 @@ class RegionRule(Rule):
|
|||
# then that means we're finished with the line and should move
|
||||
# on to the next one here
|
||||
if not done and old_y == lexer.y:
|
||||
lexer.save_context()
|
||||
lexer.y += 1
|
||||
lexer.x = 0
|
||||
|
||||
# alright, we're finally done procesing the region, so return true
|
||||
lexer.context.pop(-1)
|
||||
return True
|
||||
|
||||
class DualRegionRule(Rule):
|
||||
|
|
@ -242,52 +257,49 @@ class DualRegionRule(Rule):
|
|||
self.grammar2 = grammar2
|
||||
self.end = end
|
||||
self.start_re = re.compile(start)
|
||||
def _add_from_regex(self, context, name, lexer, m, grammar=None):
|
||||
t_name = '.'.join(context + [self.name, name])
|
||||
t = self.make_token(lexer, m.group(0), t_name, grammar=grammar)
|
||||
lexer.add_token(t)
|
||||
lexer.x += len(m.group(0))
|
||||
|
||||
def resume(self, lexer, context, flag, d, rulecontexts):
|
||||
if flag == 'start':
|
||||
d2 = self._match_first(lexer, context, d, None, rulecontexts)
|
||||
d3 = dict(d.items() + d2.items())
|
||||
self._match_second(lexer, context, d3, None, rulecontexts)
|
||||
return True
|
||||
elif flag == 'middle':
|
||||
self._match_second(lexer, context, flag, d, None, rulecontexts)
|
||||
return True
|
||||
def _add_from_regex(self, name, lexer, parent, m, matchd={}):
|
||||
s = m.group(0)
|
||||
token = self.make_token(lexer, s, name, parent, matchd)
|
||||
lexer.add_token(token)
|
||||
lexer.x += len(s)
|
||||
return token
|
||||
def resume(self, lexer, toresume):
|
||||
assert toresume
|
||||
token = toresume[0]
|
||||
d = token.matchd
|
||||
if token.name == 'start':
|
||||
stoken = toresume[0]
|
||||
mtoken = self._match_first(lexer, stoken, None, toresume)
|
||||
self._match_second(lexer, mtoken, [])
|
||||
elif token.name == 'middle':
|
||||
d3 = token.matchd
|
||||
self._match_second(lexer, token.parent, d3, toresume)
|
||||
else:
|
||||
raise Exception, "invalid flag %r" % flag
|
||||
|
||||
def match(self, lexer, context=[], d={}):
|
||||
return True
|
||||
def match(self, lexer, parent):
|
||||
# see if we can match our start token
|
||||
m = self.start_re.match(lexer.lines[lexer.y], lexer.x)
|
||||
if m:
|
||||
# region was match, so let's do this
|
||||
d1 = m.groupdict()
|
||||
d2 = self._match_first(lexer, context, d1, m, [])
|
||||
d2 = self._match_first(lexer, parent, m, [])
|
||||
d3 = dict(d1.items() + d2.items())
|
||||
self._match_second(lexer, context, d3, None, [])
|
||||
self._match_second(lexer, parent, d3, None, [])
|
||||
return True
|
||||
else:
|
||||
# region was not matched; we never started. so return false
|
||||
return False
|
||||
|
||||
def _match_first(self, lexer, context, d1, m1, rulecontext=[]):
|
||||
def _match_first(self, lexer, context, d1, m1, toresume=[]):
|
||||
# if we have been given rulecontext, then we are going to "resume" a
|
||||
# parse that can already be assumed to have started
|
||||
reenter = len(rulecontext) > 0
|
||||
reenter = len(toresume) > 1
|
||||
assert m1 or reenter
|
||||
|
||||
# first let's do some bookkeeping
|
||||
lexer.context.append(RuleContext(lexer.y, lexer.x, self, 'start',
|
||||
list(context), dict(d1)))
|
||||
|
||||
# ok, so create our start token, and get ready to start reading data
|
||||
if m1 is not None:
|
||||
self._add_from_regex(context, 'start', lexer, m1, lexer.grammar)
|
||||
null_t_name = '.'.join(context + [self.name, 'null'])
|
||||
self._add_from_regex('start', lexer, parent, m1, m1.groupdict())
|
||||
null_t = None
|
||||
|
||||
middle_re = re.compile(self.middle % d1)
|
||||
|
|
@ -302,7 +314,7 @@ class DualRegionRule(Rule):
|
|||
# if this line is empty, then we will skip it, but here weinsert
|
||||
# an empty null token just so we have something
|
||||
if len(lexer.lines[lexer.y]) == 0:
|
||||
null_t = Token(null_t_name, None, lexer.y, lexer.x, '')
|
||||
null_t = Token('null', None, lexer.y, lexer.x, '')
|
||||
lexer.add_token(null_t)
|
||||
null_t = None
|
||||
|
||||
|
|
@ -311,22 +323,22 @@ class DualRegionRule(Rule):
|
|||
while not done and lexer.y == old_y and lexer.x < len(lexer.lines[lexer.y]):
|
||||
# if we are reentering mid-parse, then that takes precedence
|
||||
if reenter:
|
||||
reenter = False
|
||||
xrule = rulecontext[0].rule
|
||||
xcontext = rulecontext[0].context
|
||||
xd = rulecontext[0].matchd
|
||||
assert rule2.resume(lexer, xcontext, xd, rulecontext[1:]), \
|
||||
"%r %r %r %r" % (lexer, xcontext, xd, rulecontext[1:])
|
||||
found = True
|
||||
null_t = None
|
||||
break
|
||||
raise Exception, "aw damn"
|
||||
#reenter = False
|
||||
#xrule = rulecontext[0].rule
|
||||
#xd = rulecontext[0].matchd
|
||||
#assert rule2.resume(lexer, xcontext, xd, rulecontext[1:]), \
|
||||
# "%r %r %r %r" % (lexer, xcontext, xd, rulecontext[1:])
|
||||
#found = True
|
||||
#null_t = None
|
||||
#break
|
||||
|
||||
# see if we have found the middle token. if so, we can then
|
||||
# proceed to "stage 2"
|
||||
m2 = middle_re.match(lexer.lines[lexer.y], lexer.x)
|
||||
if m2:
|
||||
d2 = m2.groupdict()
|
||||
self._add_from_regex(context, 'middle', lexer, m2, None)
|
||||
self._add_from_regex('middle', lexer, parent, m2, {})
|
||||
done = True
|
||||
break
|
||||
|
||||
|
|
@ -334,7 +346,7 @@ class DualRegionRule(Rule):
|
|||
# find a token, note that we found one and exit the loop
|
||||
found = False
|
||||
for rule in self.grammar1.rules:
|
||||
if rule.match(lexer, context + [self.name], d1):
|
||||
if rule.match(lexer, parent):
|
||||
found = True
|
||||
null_t = None
|
||||
break
|
||||
|
|
@ -344,7 +356,7 @@ class DualRegionRule(Rule):
|
|||
# create if it isn't set).
|
||||
if not found:
|
||||
if null_t is None:
|
||||
null_t = Token(null_t_name, None, lexer.y, lexer.x, '')
|
||||
null_t = Token('null', None, lexer.y, lexer.x, '')
|
||||
lexer.add_token(null_t)
|
||||
null_t.add_to_string(lexer.lines[lexer.y][lexer.x])
|
||||
lexer.x += 1
|
||||
|
|
@ -361,21 +373,17 @@ class DualRegionRule(Rule):
|
|||
lexer.save_context()
|
||||
lexer.y += 1
|
||||
lexer.x = 0
|
||||
lexer.context.pop(-1)
|
||||
return d2
|
||||
|
||||
def _match_second(self, lexer, context, d3, m, rulecontext=[]):
|
||||
def _match_second(self, lexer, context, d3, m, toresume=[]):
|
||||
# if we have been given rulecontext, then we are going to "resume" a
|
||||
# parse that can already be assumed to have started
|
||||
reenter = len(rulecontext) > 0
|
||||
reenter = len(toresume) > 1
|
||||
|
||||
# ok stage 2 is like stage 1, only we are looking for end tokens
|
||||
# instead of middle tokens
|
||||
null_t_name = '.'.join(context + [self.name, 'null'])
|
||||
null_t = None
|
||||
end_re = re.compile(self.end % d3)
|
||||
lexer.context.append(RuleContext(lexer.y, lexer.x, self, 'middle',
|
||||
list(context), dict(d3)))
|
||||
|
||||
# ok, so as long as we aren't done (we haven't found an end token),
|
||||
# keep reading input
|
||||
|
|
@ -385,20 +393,20 @@ class DualRegionRule(Rule):
|
|||
|
||||
# if we are reentering mid-parse, then that takes precedence
|
||||
if reenter:
|
||||
reenter = False
|
||||
xrule = rulecontext[0].rule
|
||||
xcontext = rulecontext[0].context
|
||||
xd = rulecontext[0].matchd
|
||||
assert rule2.resume(lexer, xcontext, xd, rulecontext[1:]), \
|
||||
"%r %r %r %r" % (lexer, xcontext, xd, rulecontext[1:])
|
||||
found = True
|
||||
null_t = None
|
||||
break
|
||||
raise Exception, "aw damn"
|
||||
#reenter = False
|
||||
#xrule = rulecontext[0].rule
|
||||
#xd = rulecontext[0].matchd
|
||||
#assert rule2.resume(lexer, xcontext, xd, rulecontext[1:]), \
|
||||
# "%r %r %r %r" % (lexer, xcontext, xd, rulecontext[1:])
|
||||
#found = True
|
||||
#null_t = None
|
||||
#break
|
||||
|
||||
# if this line is empty, then we will skip it, but here weinsert
|
||||
# an empty null token just so we have something
|
||||
if len(lexer.lines[lexer.y]) == 0:
|
||||
null_t = Token(null_t_name, None, lexer.y, lexer.x, '')
|
||||
null_t = Token('null', None, lexer.y, lexer.x, '')
|
||||
lexer.add_token(null_t)
|
||||
null_t = None
|
||||
|
||||
|
|
@ -409,7 +417,7 @@ class DualRegionRule(Rule):
|
|||
# proceed to "stage 2"
|
||||
m3 = end_re.match(lexer.lines[lexer.y], lexer.x)
|
||||
if m3:
|
||||
self._add_from_regex(context, 'end', lexer, m3, None)
|
||||
self._add_from_regex('end', lexer, parent, m3, {})
|
||||
done = True
|
||||
break
|
||||
|
||||
|
|
@ -417,7 +425,7 @@ class DualRegionRule(Rule):
|
|||
# find a token, note that we found one and exit the loop
|
||||
found = False
|
||||
for rule in self.grammar2.rules:
|
||||
if rule.match(lexer, context + [self.name], d3):
|
||||
if rule.match(lexer, parent):
|
||||
found = True
|
||||
null_t = None
|
||||
break
|
||||
|
|
@ -427,7 +435,7 @@ class DualRegionRule(Rule):
|
|||
# create if it isn't set).
|
||||
if not found:
|
||||
if null_t is None:
|
||||
null_t = Token(null_t_name, None, lexer.y, lexer.x, '')
|
||||
null_t = Token('null', None, lexer.y, lexer.x, '')
|
||||
lexer.add_token(null_t)
|
||||
null_t.add_to_string(lexer.lines[lexer.y][lexer.x])
|
||||
lexer.x += 1
|
||||
|
|
@ -446,150 +454,8 @@ class DualRegionRule(Rule):
|
|||
lexer.x = 0
|
||||
|
||||
# alright, we're finally done processing; return true
|
||||
lexer.context.pop(-1)
|
||||
return True
|
||||
|
||||
# def matchOLD(self, lexer, context=[], d={}):
|
||||
# m1 = self.start_re.match(lexer.lines[lexer.y], lexer.x)
|
||||
# # see if we can match out start token
|
||||
# if m1:
|
||||
# # ok, so create our start token, and get ready to start reading data
|
||||
# self._add_from_regex(context, 'start', lexer, m1, lexer.grammar)
|
||||
# null_t_name = '.'.join(context + [self.name, 'null'])
|
||||
# null_t = None
|
||||
#
|
||||
# d1 = m1.groupdict()
|
||||
# lexer.context.append(RuleContext(lexer.y, lexer.x, self, 'start',
|
||||
# list(context), dict(d1)))
|
||||
# d2 = {}
|
||||
# middle_re = re.compile(self.middle % d1)
|
||||
#
|
||||
# # ok, so as long as we aren't done (we haven't found an end token),
|
||||
# # keep reading input
|
||||
# done = False
|
||||
# while not done and lexer.y < len(lexer.lines):
|
||||
# old_y = lexer.y
|
||||
# # if this line is empty, then we will skip it, but here weinsert
|
||||
# # an empty null token just so we have something
|
||||
# if len(lexer.lines[lexer.y]) == 0:
|
||||
# null_t = Token(null_t_name, None, lexer.y, lexer.x, '')
|
||||
# lexer.add_token(null_t)
|
||||
# null_t = None
|
||||
#
|
||||
# # ok, as long as we haven't found the end token, and have more
|
||||
# # data on the current line to read, we will process tokens
|
||||
# while not done and lexer.y == old_y and lexer.x < len(lexer.lines[lexer.y]):
|
||||
# # see if we have found the middle token. if so, we can then
|
||||
# # proceed to "stage 2"
|
||||
# m2 = middle_re.match(lexer.lines[lexer.y], lexer.x)
|
||||
# if m2:
|
||||
# d2 = m2.groupdict()
|
||||
# self._add_from_regex(context, 'middle', lexer, m2, None)
|
||||
# done = True
|
||||
# break
|
||||
#
|
||||
# # ok, we need to check all our rules now, in order. if we
|
||||
# # find a token, note that we found one and exit the loop
|
||||
# found = False
|
||||
# for rule in self.grammar1.rules:
|
||||
# if rule.match(lexer, context + [self.name], d1):
|
||||
# found = True
|
||||
# null_t = None
|
||||
# break
|
||||
#
|
||||
# # if we never found a token, then we need to add another
|
||||
# # character to the current null token (which we should
|
||||
# # create if it isn't set).
|
||||
# if not found:
|
||||
# if null_t is None:
|
||||
# null_t = Token(null_t_name, None, lexer.y, lexer.x, '')
|
||||
# lexer.add_token(null_t)
|
||||
# null_t.add_to_string(lexer.lines[lexer.y][lexer.x])
|
||||
# lexer.x += 1
|
||||
#
|
||||
# # ok, since we're soon going to be on a different line (or
|
||||
# # already are), we want a new null token. so forget about the
|
||||
# # current one.
|
||||
# null_t = None
|
||||
#
|
||||
# # if we're still on the same line at this point (and not done)
|
||||
# # then that means we're finished with the line and should move
|
||||
# # on to the next one here
|
||||
# if not done and old_y == lexer.y:
|
||||
# lexer.save_context()
|
||||
# lexer.y += 1
|
||||
# lexer.x = 0
|
||||
#
|
||||
# # ok stage 2 is like stage 1, only we are looking for end tokens
|
||||
# # instead of middle tokens
|
||||
# d3 = dict(d1.items() + d2.items())
|
||||
# end_re = re.compile(self.end % d3)
|
||||
# lexer.context.pop(-1)
|
||||
# lexer.context.append(RuleContext(lexer.y, lexer.x, self, 'middle',
|
||||
# list(context), dict(d3)))
|
||||
#
|
||||
# # ok, so as long as we aren't done (we haven't found an end token),
|
||||
# # keep reading input
|
||||
# done = False
|
||||
# while not done and lexer.y < len(lexer.lines):
|
||||
# old_y = lexer.y
|
||||
# # if this line is empty, then we will skip it, but here weinsert
|
||||
# # an empty null token just so we have something
|
||||
# if len(lexer.lines[lexer.y]) == 0:
|
||||
# null_t = Token(null_t_name, None, lexer.y, lexer.x, '')
|
||||
# lexer.add_token(null_t)
|
||||
# null_t = None
|
||||
#
|
||||
# # ok, as long as we haven't found the end token, and have more
|
||||
# # data on the current line to read, we will process tokens
|
||||
# while not done and lexer.y == old_y and lexer.x < len(lexer.lines[lexer.y]):
|
||||
# # see if we have found the middle token. if so, we can then
|
||||
# # proceed to "stage 2"
|
||||
# m3 = end_re.match(lexer.lines[lexer.y], lexer.x)
|
||||
# if m3:
|
||||
# self._add_from_regex(context, 'end', lexer, m3, None)
|
||||
# done = True
|
||||
# break
|
||||
#
|
||||
# # ok, we need to check all our rules now, in order. if we
|
||||
# # find a token, note that we found one and exit the loop
|
||||
# found = False
|
||||
# for rule in self.grammar2.rules:
|
||||
# if rule.match(lexer, context + [self.name], d3):
|
||||
# found = True
|
||||
# null_t = None
|
||||
# break
|
||||
#
|
||||
# # if we never found a token, then we need to add another
|
||||
# # character to the current null token (which we should
|
||||
# # create if it isn't set).
|
||||
# if not found:
|
||||
# if null_t is None:
|
||||
# null_t = Token(null_t_name, None, lexer.y, lexer.x, '')
|
||||
# lexer.add_token(null_t)
|
||||
# null_t.add_to_string(lexer.lines[lexer.y][lexer.x])
|
||||
# lexer.x += 1
|
||||
#
|
||||
# # ok, since we're soon going to be on a different line (or
|
||||
# # already are), we want a new null token. so forget about the
|
||||
# # current one.
|
||||
# null_t = None
|
||||
#
|
||||
# # if we're still on the same line at this point (and not done)
|
||||
# # then that means we're finished with the line and should move
|
||||
# # on to the next one here
|
||||
# if not done and old_y == lexer.y:
|
||||
# lexer.save_context()
|
||||
# lexer.y += 1
|
||||
# lexer.x = 0
|
||||
#
|
||||
# # alright, we're finally done processing; return true
|
||||
# lexer.context.pop(-1)
|
||||
# return True
|
||||
# else:
|
||||
# # dual region was not matched; we never started. so return false
|
||||
# return False
|
||||
|
||||
class Grammar:
|
||||
rules = []
|
||||
def __init__(self):
|
||||
|
|
@ -606,9 +472,6 @@ class Lexer:
|
|||
self.lines = None
|
||||
self.tokens = []
|
||||
|
||||
self.context = []
|
||||
self.line_contexts = {}
|
||||
|
||||
def add_token(self, t):
|
||||
self.tokens.append(t)
|
||||
|
||||
|
|
@ -618,30 +481,24 @@ class Lexer:
|
|||
self.lines = lines
|
||||
self.tokens = []
|
||||
|
||||
self.context = []
|
||||
self.line_contexts = {}
|
||||
|
||||
def resume(self, lines, y=0, x=0, rulecontexts=[]):
|
||||
if len(rulecontexts) == 0:
|
||||
self.lex(lines, y, x)
|
||||
else:
|
||||
def resume(self, lines, y, x, token):
|
||||
self.y = y
|
||||
self.x = x
|
||||
#self.x = 0
|
||||
self.lines = lines
|
||||
self.tokens = []
|
||||
rc = rulecontexts[0]
|
||||
rc.rule.resume(self, rc.context, rc.flag, rc.matchd, rulecontexts[1:])
|
||||
|
||||
if token:
|
||||
toresume = token.parents()
|
||||
if toresume:
|
||||
raise Exception, "aw damn"
|
||||
|
||||
def __iter__(self):
|
||||
if self.lines is None:
|
||||
raise Exception, "no lines to lex"
|
||||
return self
|
||||
|
||||
def save_context(self):
|
||||
self.line_contexts[self.y] = list(self.context)
|
||||
|
||||
def next(self):
|
||||
null_t_name = 'null'
|
||||
null_t = None
|
||||
|
||||
if self.tokens:
|
||||
|
|
@ -652,15 +509,14 @@ class Lexer:
|
|||
while self.x < len(line):
|
||||
curr_t = None
|
||||
for rule in self.grammar.rules:
|
||||
if rule.match(self):
|
||||
if rule.match(self, None):
|
||||
assert self.tokens, "match rendered no tokens?"
|
||||
return self.tokens.pop(0)
|
||||
if null_t is None:
|
||||
null_t = Token(null_t_name, None, self.y, self.x, '')
|
||||
null_t = Token('null', None, self.y, self.x, '')
|
||||
self.add_token(null_t)
|
||||
null_t.add_to_string(line[self.x])
|
||||
self.x += 1
|
||||
self.save_context()
|
||||
null_t = None
|
||||
self.y += 1
|
||||
self.x = 0
|
||||
|
|
|
|||
Loading…
Reference in New Issue