Moved completers to new dir; improved completers and added word completion

This commit is contained in:
2026-02-16 01:39:23 -06:00
parent 5273c58ed6
commit b922415f98
35 changed files with 669 additions and 44 deletions

View File

@@ -0,0 +1,3 @@
from .parser import load, loads
from .writer import dump, dumps
from .speg import ParseError

View File

@@ -0,0 +1,295 @@
from .speg import peg
import re, sys
if sys.version_info[0] == 2:
_chr = unichr
else:
_chr = chr
def load(fin):
return loads(fin.read())
def loads(s):
if isinstance(s, bytes):
s = s.decode('utf-8')
if s.startswith(u'\ufeff'):
s = s[1:]
return peg(s.replace('\r\n', '\n'), _p_root)
def _p_ws(p):
p('[ \t]*')
def _p_nl(p):
p(r'([ \t]*(?:#[^\n]*)?\r?\n)+')
def _p_ews(p):
with p:
p(_p_nl)
p(_p_ws)
def _p_id(p):
return p(r'[$a-zA-Z_][$0-9a-zA-Z_]*')
_escape_table = {
'r': '\r',
'n': '\n',
't': '\t',
'f': '\f',
'b': '\b',
}
def _p_unescape(p):
esc = p('\\\\(?:u[0-9a-fA-F]{4}|[^\n])')
if esc[1] == 'u':
return _chr(int(esc[2:], 16))
return _escape_table.get(esc[1:], esc[1:])
_re_indent = re.compile(r'[ \t]*')
def _p_block_str(p, c):
p(r'{c}{c}{c}'.format(c=c))
lines = [['']]
with p:
while True:
s = p(r'(?:{c}(?!{c}{c})|[^{c}\\])*'.format(c=c))
l = s.split('\n')
lines[-1].append(l[0])
lines.extend([x] for x in l[1:])
if p(r'(?:\\\n[ \t]*)*'):
continue
p.commit()
lines[-1].append(p(_p_unescape))
p(r'{c}{c}{c}'.format(c=c))
lines = [''.join(l) for l in lines]
strip_ws = len(lines) > 1
if strip_ws and all(c in ' \t' for c in lines[-1]):
lines.pop()
indent = None
for line in lines[1:]:
if not line:
continue
if indent is None:
indent = _re_indent.match(line).group(0)
continue
for i, (c1, c2) in enumerate(zip(indent, line)):
if c1 != c2:
indent = indent[:i]
break
ind_len = len(indent or '')
if strip_ws and all(c in ' \t' for c in lines[0]):
lines = [line[ind_len:] for line in lines[1:]]
else:
lines[1:] = [line[ind_len:] for line in lines[1:]]
return '\n'.join(lines)
_re_mstr_nl = re.compile(r'(?:[ \t]*\n)+[ \t]*')
_re_mstr_trailing_nl = re.compile(_re_mstr_nl.pattern + r'\Z')
def _p_multiline_str(p, c):
p('{c}(?!{c}{c})(?:[ \t]*\n[ \t]*)?'.format(c=c))
string_parts = []
with p:
while True:
string_parts.append(p(r'[^{c}\\]*'.format(c=c)))
if p(r'(?:\\\n[ \t]*)*'):
string_parts.append('')
continue
p.commit()
string_parts.append(p(_p_unescape))
p(c)
string_parts[-1] = _re_mstr_trailing_nl.sub('', string_parts[-1])
string_parts[::2] = [_re_mstr_nl.sub(' ', part) for part in string_parts[::2]]
return ''.join(string_parts)
def _p_string(p):
with p:
return p(_p_block_str, '"')
with p:
return p(_p_block_str, "'")
with p:
return p(_p_multiline_str, '"')
return p(_p_multiline_str, "'")
def _p_array_value(p):
with p:
p(_p_nl)
return p(_p_object)
with p:
p(_p_ws)
return p(_p_line_object)
p(_p_ews)
return p(_p_simple_value)
def _p_key(p):
with p:
return p(_p_id)
return p(_p_string)
def _p_flow_kv(p):
k = p(_p_key)
p(_p_ews)
p(':')
with p:
p(_p_nl)
return k, p(_p_object)
with p:
p(_p_ws)
return k, p(_p_line_object)
p(_p_ews)
return k, p(_p_simple_value)
def _p_flow_obj_sep(p):
with p:
p(_p_ews)
p(',')
p(_p_ews)
return
p(_p_nl)
p(_p_ws)
def _p_simple_value(p):
with p:
p('null')
return None
with p:
p('false')
return False
with p:
p('true')
return True
with p:
return int(p('0b[01]+')[2:], 2)
with p:
return int(p('0o[0-7]+')[2:], 8)
with p:
return int(p('0x[0-9a-fA-F]+')[2:], 16)
with p:
return float(p(r'-?(?:[1-9][0-9]*|0)?\.[0-9]+(?:[Ee][\+-]?[0-9]+)?|(?:[1-9][0-9]*|0)(?:\.[0-9]+)?[Ee][\+-]?[0-9]+'))
with p:
return int(p('-?[1-9][0-9]*|0'), 10)
with p:
return p(_p_string)
with p:
p(r'\[')
r = []
with p:
p.set('I', '')
r.append(p(_p_array_value))
with p:
while True:
with p:
p(_p_ews)
p(',')
rr = p(_p_array_value)
if not p:
p(_p_nl)
with p:
rr = p(_p_object)
if not p:
p(_p_ews)
rr = p(_p_simple_value)
r.append(rr)
p.commit()
with p:
p(_p_ews)
p(',')
p(_p_ews)
p(r'\]')
return r
p(r'\{')
r = {}
p(_p_ews)
with p:
p.set('I', '')
k, v = p(_p_flow_kv)
r[k] = v
with p:
while True:
p(_p_flow_obj_sep)
k, v = p(_p_flow_kv)
r[k] = v
p.commit()
p(_p_ews)
with p:
p(',')
p(_p_ews)
p(r'\}')
return r
def _p_line_kv(p):
k = p(_p_key)
p(_p_ws)
p(':')
p(_p_ws)
with p:
p(_p_nl)
p(p.get('I'))
return k, p(_p_indented_object)
with p:
return k, p(_p_line_object)
with p:
return k, p(_p_simple_value)
p(_p_nl)
p(p.get('I'))
p('[ \t]')
p(_p_ws)
return k, p(_p_simple_value)
def _p_line_object(p):
k, v = p(_p_line_kv)
r = { k: v }
with p:
while True:
p(_p_ws)
p(',')
p(_p_ws)
k, v = p(_p_line_kv)
r[k] = v # uniqueness
p.commit()
return r
def _p_object(p):
p.set('I', p.get('I') + p('[ \t]*'))
r = p(_p_line_object)
with p:
while True:
p(_p_ws)
with p:
p(',')
p(_p_nl)
p(p.get('I'))
rr = p(_p_line_object)
r.update(rr) # unqueness
p.commit()
return r
def _p_indented_object(p):
p.set('I', p.get('I') + p('[ \t]'))
return p(_p_object)
def _p_root(p):
with p:
p(_p_nl)
with p:
p.set('I', '')
r = p(_p_object)
p(_p_ws)
with p:
p(',')
if not p:
p(_p_ws)
r = p(_p_simple_value)
p(_p_ews)
p(p.eof)
return r

View File

@@ -0,0 +1 @@
from .peg import peg, ParseError

View File

@@ -0,0 +1,147 @@
import sys, re
class ParseError(Exception):
def __init__(self, msg, text, offset, line, col):
self.msg = msg
self.text = text
self.offset = offset
self.line = line
self.col = col
super(ParseError, self).__init__(msg, offset, line, col)
if sys.version_info[0] == 2:
_basestr = basestring
else:
_basestr = str
def peg(s, r):
p = _Peg(s)
try:
return p(r)
except _UnexpectedError as e:
offset = max(p._errors)
err = p._errors[offset]
raise ParseError(err.msg, s, offset, err.line, err.col)
class _UnexpectedError(RuntimeError):
def __init__(self, state, expr):
self.state = state
self.expr = expr
class _PegState:
def __init__(self, pos, line, col):
self.pos = pos
self.line = line
self.col = col
self.vars = {}
self.commited = False
class _PegError:
def __init__(self, msg, line, col):
self.msg = msg
self.line = line
self.col = col
class _Peg:
def __init__(self, s):
self._s = s
self._states = [_PegState(0, 1, 1)]
self._errors = {}
self._re_cache = {}
def __call__(self, r, *args, **kw):
if isinstance(r, _basestr):
compiled = self._re_cache.get(r)
if not compiled:
compiled = re.compile(r)
self._re_cache[r] = compiled
st = self._states[-1]
m = compiled.match(self._s[st.pos:])
if not m:
self.error(expr=r, err=kw.get('err'))
ms = m.group(0)
st.pos += len(ms)
nl_pos = ms.rfind('\n')
if nl_pos < 0:
st.col += len(ms)
else:
st.col = len(ms) - nl_pos
st.line += ms[:nl_pos].count('\n') + 1
return ms
else:
kw.pop('err', None)
return r(self, *args, **kw)
def __repr__(self):
pos = self._states[-1].pos
vars = {}
for st in self._states:
vars.update(st.vars)
return '_Peg(%r, %r)' % (self._s[:pos] + '*' + self._s[pos:], vars)
@staticmethod
def eof(p):
if p._states[-1].pos != len(p._s):
p.error()
def error(self, err=None, expr=None):
st = self._states[-1]
if err is None:
err = 'expected {!r}, found {!r}'.format(expr, self._s[st.pos:st.pos+4])
self._errors[st.pos] = _PegError(err, st.line, st.col)
raise _UnexpectedError(st, expr)
def get(self, key, default=None):
for state in self._states[::-1]:
if key in state.vars:
return state.vars[key][0]
return default
def set(self, key, value):
self._states[-1].vars[key] = value, False
def set_global(self, key, value):
self._states[-1].vars[key] = value, True
def opt(self, *args, **kw):
with self:
return self(*args, **kw)
def not_(self, s, *args, **kw):
with self:
self(s)
self.error()
def __enter__(self):
self._states[-1].committed = False
self._states.append(_PegState(self._states[-1].pos, self._states[-1].line, self._states[-1].col))
def __exit__(self, type, value, traceback):
if type is None:
self.commit()
self._states.pop()
return type == _UnexpectedError
def commit(self):
cur = self._states[-1]
prev = self._states[-2]
for key in cur.vars:
val, g = cur.vars[key]
if not g:
continue
if key in prev.vars:
prev.vars[key] = val, prev.vars[key][1]
else:
prev.vars[key] = val, True
prev.pos = cur.pos
prev.line = cur.line
prev.col = cur.col
prev.committed = True
def __nonzero__(self):
return self._states[-1].committed
__bool__ = __nonzero__

View File

@@ -0,0 +1,191 @@
import re, json, sys
if sys.version_info[0] == 2:
def _is_num(o):
return isinstance(o, int) or isinstance(o, long) or isinstance(o, float)
def _stringify(o):
if isinstance(o, str):
return unicode(o)
if isinstance(o, unicode):
return o
return None
else:
def _is_num(o):
return isinstance(o, int) or isinstance(o, float)
def _stringify(o):
if isinstance(o, bytes):
return o.decode()
if isinstance(o, str):
return o
return None
_id_re = re.compile(r'[$a-zA-Z_][$0-9a-zA-Z_]*\Z')
class CSONEncoder:
def __init__(self, skipkeys=False, ensure_ascii=True, check_circular=True, allow_nan=True, sort_keys=False,
indent=None, default=None):
self._skipkeys = skipkeys
self._ensure_ascii = ensure_ascii
self._allow_nan = allow_nan
self._sort_keys = sort_keys
self._indent = ' ' * (indent or 4)
self._default = default
if check_circular:
self._obj_stack = set()
else:
self._obj_stack = None
def _format_simple_val(self, o):
if o is None:
return 'null'
if isinstance(o, bool):
return 'true' if o else 'false'
if _is_num(o):
return str(o)
s = _stringify(o)
if s is not None:
return self._escape_string(s)
return None
def _escape_string(self, s):
r = json.dumps(s, ensure_ascii=self._ensure_ascii)
return u"'{}'".format(r[1:-1].replace("'", r"\'"))
def _escape_key(self, s):
if s is None or isinstance(s, bool) or _is_num(s):
s = str(s)
s = _stringify(s)
if s is None:
if self._skipkeys:
return None
raise TypeError('keys must be a string')
if not _id_re.match(s):
return self._escape_string(s)
return s
def _push_obj(self, o):
if self._obj_stack is not None:
if id(o) in self._obj_stack:
raise ValueError('Circular reference detected')
self._obj_stack.add(id(o))
def _pop_obj(self, o):
if self._obj_stack is not None:
self._obj_stack.remove(id(o))
def _encode(self, o, obj_val=False, indent='', force_flow=False):
if isinstance(o, list):
if not o:
if obj_val:
yield ' []\n'
else:
yield indent
yield '[]\n'
else:
if obj_val:
yield ' [\n'
else:
yield indent
yield '[\n'
indent = indent + self._indent
self._push_obj(o)
for v in o:
for chunk in self._encode(v, obj_val=False, indent=indent, force_flow=True):
yield chunk
self._pop_obj(o)
yield indent[:-len(self._indent)]
yield ']\n'
elif isinstance(o, dict):
items = [(self._escape_key(k), v) for k, v in o.items()]
if self._skipkeys:
items = [(k, v) for k, v in items if k is not None]
if self._sort_keys:
items.sort()
if force_flow or not items:
if not items:
if obj_val:
yield ' {}\n'
else:
yield indent
yield '{}\n'
else:
if obj_val:
yield ' {\n'
else:
yield indent
yield '{\n'
indent = indent + self._indent
self._push_obj(o)
for k, v in items:
yield indent
yield k
yield ':'
for chunk in self._encode(v, obj_val=True, indent=indent + self._indent, force_flow=False):
yield chunk
self._pop_obj(o)
yield indent[:-len(self._indent)]
yield '}\n'
else:
if obj_val:
yield '\n'
self._push_obj(o)
for k, v in items:
yield indent
yield k
yield ':'
for chunk in self._encode(v, obj_val=True, indent=indent + self._indent, force_flow=False):
yield chunk
self._pop_obj(o)
else:
v = self._format_simple_val(o)
if v is None:
self._push_obj(o)
v = self.default(o)
for chunk in self._encode(v, obj_val=obj_val, indent=indent, force_flow=force_flow):
yield chunk
self._pop_obj(o)
else:
if obj_val:
yield ' '
else:
yield indent
yield v
yield '\n'
def iterencode(self, o):
return self._encode(o)
def encode(self, o):
return ''.join(self.iterencode(o))
def default(self, o):
if self._default is None:
raise TypeError('Cannot serialize an object of type {}'.format(type(o).__name__))
return self._default(o)
def dump(obj, fp, skipkeys=False, ensure_ascii=True, check_circular=True, allow_nan=True, cls=None,
indent=None, default=None, sort_keys=False, **kw):
if indent is None and cls is None:
return json.dump(obj, fp, skipkeys=skipkeys, ensure_ascii=ensure_ascii, check_circular=check_circular,
allow_nan=allow_nan, default=default, sort_keys=sort_keys, separators=(',', ':'))
if cls is None:
cls = CSONEncoder
encoder = cls(skipkeys=skipkeys, ensure_ascii=ensure_ascii, check_circular=check_circular,
allow_nan=allow_nan, sort_keys=sort_keys, indent=indent, default=default, **kw)
for chunk in encoder.iterencode(obj):
fp.write(chunk)
def dumps(obj, skipkeys=False, ensure_ascii=True, check_circular=True, allow_nan=True, cls=None, indent=None,
default=None, sort_keys=False, **kw):
if indent is None and cls is None:
return json.dumps(obj, skipkeys=skipkeys, ensure_ascii=ensure_ascii, check_circular=check_circular,
allow_nan=allow_nan, default=default, sort_keys=sort_keys, separators=(',', ':'))
if cls is None:
cls = CSONEncoder
encoder = cls(skipkeys=skipkeys, ensure_ascii=ensure_ascii, check_circular=check_circular,
allow_nan=allow_nan, sort_keys=sort_keys, indent=indent, default=default, **kw)
return encoder.encode(obj)