def p_subshell(p):
'''subshell : LEFT_PAREN compound_list RIGHT_PAREN'''
lparen = ast.node(kind='reservedword', word=p[1], pos=p.lexspan(1))
rparen = ast.node(kind='reservedword', word=p[3], pos=p.lexspan(3))
parts = [lparen, p[2], rparen]
p[0] = ast.node(kind='compound', list=parts, redirects=[],
pos=_partsspan(parts))
def p_group_command(p):
'''group_command : LEFT_CURLY compound_list RIGHT_CURLY'''
lcurly = ast.node(kind='reservedword', word=p[1], pos=p.lexspan(1))
rcurly = ast.node(kind='reservedword', word=p[3], pos=p.lexspan(3))
parts = [lcurly, p[2], rcurly]
p[0] = ast.node(kind='compound', list=parts, redirects=[],
pos=_partsspan(parts))
def p_shell_command(p):
'''shell_command : for_command
| case_command
| WHILE compound_list DO compound_list DONE
| UNTIL compound_list DO compound_list DONE
| select_command
| if_command
| subshell
| group_command
| arith_command
| cond_command
| arith_for_command'''
if len(p) == 2:
p[0] = p[1]
else:
# while or until
assert p[2].kind == 'list'
parts = _makeparts(p)
kind = parts[0].word
assert kind in ('while', 'until')
p[0] = ast.node(kind='compound',
redirects=[],
list=[ast.node(kind=kind, parts=parts, pos=_partsspan(parts))],
pos=_partsspan(parts))
assert p[0].kind == 'compound'
def p_shell_command(p):
'''shell_command : for_command
| case_command
| WHILE compound_list DO compound_list DONE
| UNTIL compound_list DO compound_list DONE
| select_command
| if_command
| subshell
| group_command
| arith_command
| cond_command
| arith_for_command'''
if len(p) == 2:
p[0] = p[1]
else:
# while or until
assert p[2].kind == 'list'
parts = _makeparts(p)
kind = parts[0].word
assert kind in ('while', 'until')
p[0] = ast.node(kind='compound',
redirects=[],
list=[ast.node(kind=kind, parts=parts, pos=_partsspan(parts))],
pos=_partsspan(parts))
assert p[0].kind == 'compound'
def p_subshell(p):
'''subshell : LEFT_PAREN compound_list RIGHT_PAREN'''
lparen = ast.node(kind='reservedword', word=p[1], pos=p.lexspan(1))
rparen = ast.node(kind='reservedword', word=p[3], pos=p.lexspan(3))
parts = [lparen, p[2], rparen]
p[0] = ast.node(kind='compound', list=parts, redirects=[],
pos=_partsspan(parts))
def p_group_command(p):
'''group_command : LEFT_CURLY compound_list RIGHT_CURLY'''
lcurly = ast.node(kind='reservedword', word=p[1], pos=p.lexspan(1))
rcurly = ast.node(kind='reservedword', word=p[3], pos=p.lexspan(3))
parts = [lcurly, p[2], rcurly]
p[0] = ast.node(kind='compound', list=parts, redirects=[],
pos=_partsspan(parts))
def _expandword(parser, tokenword):
if parser._expansionlimit == -1:
# we enter this branch in the following conditions:
# - currently parsing a substitution as a result of an expansion
# - the previous expansion had limit == 0
#
# this means that this node is a descendant of a substitution in an
# unexpanded word and will be filtered in the limit == 0 condition below
#
# (the reason we even expand when limit == 0 is to get quote removal)
node = ast.node(kind='word', word=tokenword,
pos=(tokenword.lexpos, tokenword.endlexpos), parts=[])
return node
else:
quoted = bool(tokenword.flags & flags.word.QUOTED)
doublequoted = quoted and tokenword.value[0] == '"'
# TODO set qheredocument
parts, expandedword = subst._expandwordinternal(parser,
tokenword, 0,
doublequoted, 0, 0)
# limit reached, don't include substitutions (still expanded to get
# quote removal though)
if parser._expansionlimit == 0:
parts = [node for node in parts if 'substitution' not in node.kind]
node = ast.node(kind='word', word=expandedword,
pos=(tokenword.lexpos, tokenword.endlexpos), parts=parts)
return node
def _expandword(parser, tokenword):
if parser._expansionlimit == -1:
# we enter this branch in the following conditions:
# - currently parsing a substitution as a result of an expansion
# - the previous expansion had limit == 0
#
# this means that this node is a descendant of a substitution in an
# unexpanded word and will be filtered in the limit == 0 condition below
#
# (the reason we even expand when limit == 0 is to get quote removal)
node = ast.node(kind='word', word=tokenword,
pos=(tokenword.lexpos, tokenword.endlexpos), parts=[])
return node
else:
quoted = bool(tokenword.flags & flags.word.QUOTED)
doublequoted = quoted and tokenword.value[0] == '"'
# TODO set qheredocument
parts, expandedword = subst._expandwordinternal(parser,
tokenword, 0,
doublequoted, 0, 0)
# limit reached, don't include substitutions (still expanded to get
# quote removal though)
if parser._expansionlimit == 0:
parts = [node for node in parts if 'substitution' not in node.kind]
node = ast.node(kind='word', word=expandedword,
pos=(tokenword.lexpos, tokenword.endlexpos), parts=parts)
return node
def p_list0(p):
'''list0 : list1 NEWLINE newline_list
| list1 AMPERSAND newline_list
| list1 SEMICOLON newline_list'''
parts = p[1]
if len(parts) > 1 or p.slice[2].ttype != tokenizer.tokentype.NEWLINE:
parts.append(ast.node(kind='operator', op=p[2], pos=p.lexspan(2)))
p[0] = ast.node(kind='list', parts=parts, pos=_partsspan(parts))
else:
p[0] = parts[0]
def p_list0(p):
'''list0 : list1 NEWLINE newline_list
| list1 AMPERSAND newline_list
| list1 SEMICOLON newline_list'''
parts = p[1]
if len(parts) > 1 or p.slice[2].ttype != tokenizer.tokentype.NEWLINE:
parts.append(ast.node(kind='operator', op=p[2], pos=p.lexspan(2)))
p[0] = ast.node(kind='list', parts=parts, pos=_partsspan(parts))
else:
p[0] = parts[0]
def p_if_command(p):
'''if_command : IF compound_list THEN compound_list FI
| IF compound_list THEN compound_list ELSE compound_list FI
| IF compound_list THEN compound_list elif_clause FI'''
# we currently don't distinguish the various lists that make up the
# command, because it's not needed later on. if there will be a need
# we can always add different nodes for elif/else.
parts = _makeparts(p)
p[0] = ast.node(kind='compound',
redirects=[],
list=[ast.node(kind='if', parts=parts, pos=_partsspan(parts))],
pos=_partsspan(parts))
def p_if_command(p):
'''if_command : IF compound_list THEN compound_list FI
| IF compound_list THEN compound_list ELSE compound_list FI
| IF compound_list THEN compound_list elif_clause FI'''
# we currently don't distinguish the various lists that make up the
# command, because it's not needed later on. if there will be a need
# we can always add different nodes for elif/else.
parts = _makeparts(p)
p[0] = ast.node(kind='compound',
redirects=[],
list=[ast.node(kind='if', parts=parts, pos=_partsspan(parts))],
pos=_partsspan(parts))
def p_redirection(p):
'''redirection : GREATER WORD
| LESS WORD
| NUMBER GREATER WORD
| NUMBER LESS WORD
| REDIR_WORD GREATER WORD
| REDIR_WORD LESS WORD
| GREATER_GREATER WORD
| NUMBER GREATER_GREATER WORD
| REDIR_WORD GREATER_GREATER WORD
| GREATER_BAR WORD
| NUMBER GREATER_BAR WORD
| REDIR_WORD GREATER_BAR WORD
| LESS_GREATER WORD
| NUMBER LESS_GREATER WORD
| REDIR_WORD LESS_GREATER WORD
| LESS_LESS_LESS WORD
| NUMBER LESS_LESS_LESS WORD
| REDIR_WORD LESS_LESS_LESS WORD
| LESS_AND NUMBER
| NUMBER LESS_AND NUMBER
| REDIR_WORD LESS_AND NUMBER
| GREATER_AND NUMBER
| NUMBER GREATER_AND NUMBER
| REDIR_WORD GREATER_AND NUMBER
| LESS_AND WORD
| NUMBER LESS_AND WORD
| REDIR_WORD LESS_AND WORD
| GREATER_AND WORD
| NUMBER GREATER_AND WORD
| REDIR_WORD GREATER_AND WORD
| GREATER_AND DASH
| NUMBER GREATER_AND DASH
| REDIR_WORD GREATER_AND DASH
| LESS_AND DASH
| NUMBER LESS_AND DASH
| REDIR_WORD LESS_AND DASH
| AND_GREATER WORD
| AND_GREATER_GREATER WORD'''
parserobj = p.context
if len(p) == 3:
output = p[2]
if p.slice[2].ttype == tokenizer.tokentype.WORD:
output = _expandword(parserobj, p.slice[2])
p[0] = ast.node(kind='redirect', input=None, type=p[1], heredoc=None,
output=output, pos=(p.lexpos(1), p.endlexpos(2)))
else:
output = p[3]
if p.slice[3].ttype == tokenizer.tokentype.WORD:
output = _expandword(parserobj, p.slice[3])
p[0] = ast.node(kind='redirect', input=p[1], type=p[2], heredoc=None,
output=output, pos=(p.lexpos(1), p.endlexpos(3)))
def p_redirection(p):
'''redirection : GREATER WORD
| LESS WORD
| NUMBER GREATER WORD
| NUMBER LESS WORD
| REDIR_WORD GREATER WORD
| REDIR_WORD LESS WORD
| GREATER_GREATER WORD
| NUMBER GREATER_GREATER WORD
| REDIR_WORD GREATER_GREATER WORD
| GREATER_BAR WORD
| NUMBER GREATER_BAR WORD
| REDIR_WORD GREATER_BAR WORD
| LESS_GREATER WORD
| NUMBER LESS_GREATER WORD
| REDIR_WORD LESS_GREATER WORD
| LESS_LESS_LESS WORD
| NUMBER LESS_LESS_LESS WORD
| REDIR_WORD LESS_LESS_LESS WORD
| LESS_AND NUMBER
| NUMBER LESS_AND NUMBER
| REDIR_WORD LESS_AND NUMBER
| GREATER_AND NUMBER
| NUMBER GREATER_AND NUMBER
| REDIR_WORD GREATER_AND NUMBER
| LESS_AND WORD
| NUMBER LESS_AND WORD
| REDIR_WORD LESS_AND WORD
| GREATER_AND WORD
| NUMBER GREATER_AND WORD
| REDIR_WORD GREATER_AND WORD
| GREATER_AND DASH
| NUMBER GREATER_AND DASH
| REDIR_WORD GREATER_AND DASH
| LESS_AND DASH
| NUMBER LESS_AND DASH
| REDIR_WORD LESS_AND DASH
| AND_GREATER WORD
| AND_GREATER_GREATER WORD'''
parserobj = p.context
if len(p) == 3:
output = p[2]
if p.slice[2].ttype == tokenizer.tokentype.WORD:
output = _expandword(parserobj, p.slice[2])
p[0] = ast.node(kind='redirect', input=None, type=p[1], heredoc=None,
output=output, pos=(p.lexpos(1), p.endlexpos(2)))
else:
output = p[3]
if p.slice[3].ttype == tokenizer.tokentype.WORD:
output = _expandword(parserobj, p.slice[3])
p[0] = ast.node(kind='redirect', input=p[1], type=p[2], heredoc=None,
output=output, pos=(p.lexpos(1), p.endlexpos(3)))
def p_case_command(p):
'''case_command : CASE WORD newline_list IN newline_list ESAC
| CASE WORD newline_list IN case_clause_sequence newline_list ESAC
| CASE WORD newline_list IN case_clause ESAC'''
parts = _makeparts(p)
p[0] = ast.node(kind='compound',
redirects=[],
list=[
ast.node(kind='case',
parts=parts,
pos=_partsspan(parts),
lineno=parts[-1].lineno)
],
pos=_partsspan(parts),
lineno=parts[-1].lineno)
def compoundnode(s, *parts, **kwargs):
redirects = kwargs.pop('redirects', [])
assert not kwargs
return ast.node(kind='compound',
s=s,
list=list(parts),
redirects=redirects)
def redirectnode(s, input, type, output, heredoc=None):
return ast.node(kind='redirect',
input=input,
type=type,
output=output,
heredoc=heredoc,
s=s)
def listnode(s, *parts):
for i in range(len(parts)):
if i % 2 == 0:
assert parts[i].kind in ('command', 'pipeline', 'compound'), parts[i].kind
else:
assert parts[i].kind == 'operator', parts[i].kind
return ast.node(kind='list', parts=list(parts), s=s)
def _extractcommandsubst(parserobj, string, sindex, sxcommand=False):
if string[sindex] == '(':
raise NotImplementedError('arithmetic expansion')
#return _extractdelimitedstring(parserobj, string, sindex, '$(', '(', '(', sxcommand=True)
else:
node, si = _parsedolparen(parserobj, string, sindex)
si += 1
return ast.node(kind='commandsubstitution', command=node, pos=(sindex-2, si), lineno=node.lineno), si
def _extractcommandsubst(parserobj, string, sindex, sxcommand=False):
if string[sindex] == '(':
raise NotImplementedError('arithmetic expansion')
#return _extractdelimitedstring(parserobj, string, sindex, '$(', '(', '(', sxcommand=True)
else:
node, si = _parsedolparen(parserobj, string, sindex)
si += 1
return ast.node(kind='commandsubstitution', command=node, pos=(sindex-2, si)), si
def p_list_terminator(p):
'''list_terminator : NEWLINE
| SEMICOLON
| EOF'''
if p[1] == ';':
p[0] = ast.node(kind='operator',
op=';',
pos=p.lexspan(1),
lineno=p.lineno(1))
def _make_pattern(part_, action_):
if action_.kind == 'list':
action_ = action_.parts
else:
action_ = [action_]
return ast.node(kind='pattern',
pattern=part_,
actions=action_,
pos=(part_[0].pos[0], action_[-1].pos[1]),
lineno=action_[-1].lineno)
def p_function_def(p):
'''function_def : WORD LEFT_PAREN RIGHT_PAREN newline_list function_body
| FUNCTION WORD LEFT_PAREN RIGHT_PAREN newline_list function_body
| FUNCTION WORD newline_list function_body'''
parts = _makeparts(p)
body = parts[-1]
name = parts[ast.findfirstkind(parts, 'word')]
p[0] = ast.node(kind='function', name=name, body=body, parts=parts,
pos=_partsspan(parts))
def p_pipeline(p):
'''pipeline : pipeline BAR newline_list pipeline
| pipeline BAR_AND newline_list pipeline
| command'''
if len(p) == 2:
p[0] = [p[1]]
else:
p[0] = p[1]
p[0].append(ast.node(kind='pipe', pipe=p[2], pos=p.lexspan(2)))
p[0].extend(p[len(p) - 1])
def p_pipeline(p):
'''pipeline : pipeline BAR newline_list pipeline
| pipeline BAR_AND newline_list pipeline
| command'''
if len(p) == 2:
p[0] = [p[1]]
else:
p[0] = p[1]
p[0].append(ast.node(kind='pipe', pipe=p[2], pos=p.lexspan(2)))
p[0].extend(p[len(p) - 1])
def p_function_def(p):
'''function_def : WORD LEFT_PAREN RIGHT_PAREN newline_list function_body
| FUNCTION WORD LEFT_PAREN RIGHT_PAREN newline_list function_body
| FUNCTION WORD newline_list function_body'''
parts = _makeparts(p)
body = parts[-1]
name = parts[ast.findfirstkind(parts, 'word')]
p[0] = ast.node(kind='function', name=name, body=body, parts=parts,
pos=_partsspan(parts))
def pipelinenode(s, *parts):
oldparts = parts
if parts[0].kind == 'reservedword' and parts[0].word == '!':
parts = parts[1:]
for i in range(len(parts)):
if i % 2 == 0:
assert parts[i].kind in ('command', 'compound'), parts[i].kind
else:
assert parts[i].kind == 'pipe', parts[i].kind
return ast.node(kind='pipeline', s=s, parts=list(oldparts))
def p_compound_list(p):
'''compound_list : list
| newline_list list1'''
if len(p) == 2:
p[0] = p[1]
else:
parts = p[2]
if len(parts) > 1:
p[0] = ast.node(kind='list', parts=parts, pos=_partsspan(parts))
else:
p[0] = parts[0]
def p_elif_clause(p):
'''elif_clause : ELIF compound_list THEN compound_list
| ELIF compound_list THEN compound_list ELSE compound_list
| ELIF compound_list THEN compound_list elif_clause'''
parts = []
for i in range(1, len(p)):
if isinstance(p[i], ast.node):
parts.append(p[i])
else:
parts.append(ast.node(kind='reservedword', word=p[i], pos=p.lexspan(i)))
p[0] = parts
def p_elif_clause(p):
'''elif_clause : ELIF compound_list THEN compound_list
| ELIF compound_list THEN compound_list ELSE compound_list
| ELIF compound_list THEN compound_list elif_clause'''
parts = []
for i in range(1, len(p)):
if isinstance(p[i], ast.node):
parts.append(p[i])
else:
parts.append(ast.node(kind='reservedword', word=p[i], pos=p.lexspan(i)))
p[0] = parts
def p_simple_list(p):
'''simple_list : simple_list1
| simple_list1 AMPERSAND
| simple_list1 SEMICOLON'''
tok = p.lexer
heredoc.gatherheredocuments(tok)
if len(p) == 3 or len(p[1]) > 1:
parts = p[1]
if len(p) == 3:
parts.append(ast.node(kind='operator', op=p[2], pos=p.lexspan(2)))
p[0] = ast.node(kind='list', parts=parts, pos=_partsspan(parts))
else:
assert len(p[1]) == 1
p[0] = p[1][0]
if (len(p) == 2 and p.lexer._parserstate & flags.parser.CMDSUBST
and p.lexer._current_token.nopos() == p.lexer._shell_eof_token):
# accept the input
p.accept()
def p_simple_list(p):
'''simple_list : simple_list1
| simple_list1 AMPERSAND
| simple_list1 SEMICOLON'''
tok = p.lexer
heredoc.gatherheredocuments(tok)
if len(p) == 3 or len(p[1]) > 1:
parts = p[1]
if len(p) == 3:
parts.append(ast.node(kind='operator', op=p[2], pos=p.lexspan(2)))
p[0] = ast.node(kind='list', parts=parts, pos=_partsspan(parts))
else:
assert len(p[1]) == 1
p[0] = p[1][0]
if (len(p) == 2 and p.lexer._parserstate & flags.parser.CMDSUBST and
p.lexer._current_token.nopos() == p.lexer._shell_eof_token):
# accept the input
p.accept()
def p_compound_list(p):
'''compound_list : list
| newline_list list1'''
if len(p) == 2:
p[0] = p[1]
else:
parts = p[2]
if len(parts) > 1:
p[0] = ast.node(kind='list', parts=parts, pos=_partsspan(parts))
else:
p[0] = parts[0]
def p_simple_list1(p):
'''simple_list1 : simple_list1 AND_AND newline_list simple_list1
| simple_list1 OR_OR newline_list simple_list1
| simple_list1 AMPERSAND simple_list1
| simple_list1 SEMICOLON simple_list1
| pipeline_command'''
if len(p) == 2:
p[0] = [p[1]]
else:
p[0] = p[1]
p[0].append(ast.node(kind='operator', op=p[2], pos=p.lexspan(2)))
p[0].extend(p[len(p) - 1])
def p_simple_list1(p):
'''simple_list1 : simple_list1 AND_AND newline_list simple_list1
| simple_list1 OR_OR newline_list simple_list1
| simple_list1 AMPERSAND simple_list1
| simple_list1 SEMICOLON simple_list1
| pipeline_command'''
if len(p) == 2:
p[0] = [p[1]]
else:
p[0] = p[1]
p[0].append(ast.node(kind='operator', op=p[2], pos=p.lexspan(2)))
p[0].extend(p[len(p) - 1])
def p_for_command(p):
'''for_command : FOR WORD newline_list DO compound_list DONE
| FOR WORD newline_list LEFT_CURLY compound_list RIGHT_CURLY
| FOR WORD SEMICOLON newline_list DO compound_list DONE
| FOR WORD SEMICOLON newline_list LEFT_CURLY compound_list RIGHT_CURLY
| FOR WORD newline_list IN word_list list_terminator newline_list DO compound_list DONE
| FOR WORD newline_list IN word_list list_terminator newline_list LEFT_CURLY compound_list RIGHT_CURLY
| FOR WORD newline_list IN list_terminator newline_list DO compound_list DONE
| FOR WORD newline_list IN list_terminator newline_list LEFT_CURLY compound_list RIGHT_CURLY'''
parts = _makeparts(p)
# find the operatornode that we might have there due to
# list_terminator/newline_list and convert it to a reservedword so its
# considered as part of the for loop
for i, part in enumerate(parts):
if part.kind == 'operator' and part.op == ';':
parts[i] = ast.node(kind='reservedword', word=';', pos=part.pos)
break # there could be only one in there...
p[0] = ast.node(kind='compound',
redirects=[],
list=[ast.node(kind='for', parts=parts, pos=_partsspan(parts))],
pos=_partsspan(parts))
def p_pipeline_command(p):
'''pipeline_command : pipeline
| BANG pipeline_command
| timespec pipeline_command
| timespec list_terminator
| BANG list_terminator'''
if len(p) == 2:
if len(p[1]) == 1:
p[0] = p[1][0]
else:
p[0] = ast.node(kind='pipeline', parts=p[1],
pos=(p[1][0].pos[0], p[1][-1].pos[1]))
else:
# XXX timespec
node = ast.node(kind='reservedword', word='!', pos=p.lexspan(1))
if p[2].kind == 'pipeline':
p[0] = p[2]
p[0].parts.insert(0, node)
p[0].pos = (p[0].parts[0].pos[0], p[0].parts[-1].pos[1])
else:
p[0] = ast.node(kind='pipeline', parts=[node, p[2]],
pos=(node.pos[0], p[2].pos[1]))
def p_for_command(p):
'''for_command : FOR WORD newline_list DO compound_list DONE
| FOR WORD newline_list LEFT_CURLY compound_list RIGHT_CURLY
| FOR WORD SEMICOLON newline_list DO compound_list DONE
| FOR WORD SEMICOLON newline_list LEFT_CURLY compound_list RIGHT_CURLY
| FOR WORD newline_list IN word_list list_terminator newline_list DO compound_list DONE
| FOR WORD newline_list IN word_list list_terminator newline_list LEFT_CURLY compound_list RIGHT_CURLY
| FOR WORD newline_list IN list_terminator newline_list DO compound_list DONE
| FOR WORD newline_list IN list_terminator newline_list LEFT_CURLY compound_list RIGHT_CURLY'''
parts = _makeparts(p)
# find the operatornode that we might have there due to
# list_terminator/newline_list and convert it to a reservedword so its
# considered as part of the for loop
for i, part in enumerate(parts):
if part.kind == 'operator' and part.op == ';':
parts[i] = ast.node(kind='reservedword', word=';', pos=part.pos)
break # there could be only one in there...
p[0] = ast.node(kind='compound',
redirects=[],
list=[ast.node(kind='for', parts=parts, pos=_partsspan(parts))],
pos=_partsspan(parts))
def p_pipeline_command(p):
'''pipeline_command : pipeline
| BANG pipeline_command
| timespec pipeline_command
| timespec list_terminator
| BANG list_terminator'''
if len(p) == 2:
if len(p[1]) == 1:
p[0] = p[1][0]
else:
p[0] = ast.node(kind='pipeline', parts=p[1],
pos=(p[1][0].pos[0], p[1][-1].pos[1]))
else:
# XXX timespec
node = ast.node(kind='reservedword', word='!', pos=p.lexspan(1))
if p[2].kind == 'pipeline':
p[0] = p[2]
p[0].parts.insert(0, node)
p[0].pos = (p[0].parts[0].pos[0], p[0].parts[-1].pos[1])
else:
p[0] = ast.node(kind='pipeline', parts=[node, p[2]],
pos=(node.pos[0], p[2].pos[1]))
def _paramexpand(parserobj, string, sindex, lineno=0):
node = None
zindex = sindex + 1
c = string[zindex] if zindex < len(string) else None
if c and c in '0123456789$#?-!*@':
# XXX 7685
node = ast.node(kind='parameter', value=c,
pos=(sindex, zindex+1), lineno=lineno)
elif c == '{':
# XXX 7863
# TODO not start enough, doesn't consider escaping
zindex = string.find('}', zindex + 1)
node = ast.node(kind='parameter', value=string[sindex+2:zindex],
pos=(sindex, zindex+1), lineno=lineno)
# TODO
# return _parameterbraceexpand(string, zindex)
elif c == '(':
return _extractcommandsubst(parserobj, string, zindex + 1)
elif c == '[':
raise NotImplementedError('arithmetic substitution')
#return _extractarithmeticsubst(string, zindex + 1)
else:
tindex = zindex
for zindex in range(tindex, len(string) + 1):
if zindex == len(string):
break
if not string[zindex].isalnum() and not string[zindex] == '_':
break
temp1 = string[sindex:zindex]
if temp1:
return (ast.node(kind='parameter', value=temp1[1:], pos=(sindex, zindex),
lineno=lineno),
zindex)
if zindex < len(string):
zindex += 1
return node, zindex
def p_redirection_heredoc(p):
'''redirection : LESS_LESS WORD
| NUMBER LESS_LESS WORD
| REDIR_WORD LESS_LESS WORD
| LESS_LESS_MINUS WORD
| NUMBER LESS_LESS_MINUS WORD
| REDIR_WORD LESS_LESS_MINUS WORD'''
parserobj = p.context
assert isinstance(parserobj, _parser)
output = ast.node(kind='word', word=p[len(p)-1], parts=[],
pos=p.lexspan(len(p)-1))
if len(p) == 3:
p[0] = ast.node(kind='redirect', input=None, type=p[1], heredoc=None,
output=output, pos=(p.lexpos(1), p.endlexpos(2)))
else:
p[0] = ast.node(kind='redirect', input=p[1], type=p[2], heredoc=None,
output=output, pos=(p.lexpos(1), p.endlexpos(3)))
if p.slice[len(p)-2].ttype == tokenizer.tokentype.LESS_LESS:
parserobj.redirstack.append((p[0], False))
else:
parserobj.redirstack.append((p[0], True))
def p_list1(p):
'''list1 : list1 AND_AND newline_list list1
| list1 OR_OR newline_list list1
| list1 AMPERSAND newline_list list1
| list1 SEMICOLON newline_list list1
| list1 NEWLINE newline_list list1
| pipeline_command'''
if len(p) == 2:
p[0] = [p[1]]
else:
p[0] = p[1]
# XXX newline
p[0].append(ast.node(kind='operator', op=p[2], pos=p.lexspan(2)))
p[0].extend(p[len(p) - 1])
def p_redirection_heredoc(p):
'''redirection : LESS_LESS WORD
| NUMBER LESS_LESS WORD
| REDIR_WORD LESS_LESS WORD
| LESS_LESS_MINUS WORD
| NUMBER LESS_LESS_MINUS WORD
| REDIR_WORD LESS_LESS_MINUS WORD'''
parserobj = p.context
assert isinstance(parserobj, _parser)
output = ast.node(kind='word', word=p[len(p)-1], parts=[],
pos=p.lexspan(len(p)-1))
if len(p) == 3:
p[0] = ast.node(kind='redirect', input=None, type=p[1], heredoc=None,
output=output, pos=(p.lexpos(1), p.endlexpos(2)))
else:
p[0] = ast.node(kind='redirect', input=p[1], type=p[2], heredoc=None,
output=output, pos=(p.lexpos(1), p.endlexpos(3)))
if p.slice[len(p)-2].ttype == tokenizer.tokentype.LESS_LESS:
parserobj.redirstack.append((p[0], False))
else:
parserobj.redirstack.append((p[0], True))
def p_list1(p):
'''list1 : list1 AND_AND newline_list list1
| list1 OR_OR newline_list list1
| list1 AMPERSAND newline_list list1
| list1 SEMICOLON newline_list list1
| list1 NEWLINE newline_list list1
| pipeline_command'''
if len(p) == 2:
p[0] = [p[1]]
else:
p[0] = p[1]
# XXX newline
p[0].append(ast.node(kind='operator', op=p[2], pos=p.lexspan(2)))
p[0].extend(p[len(p) - 1])
def _paramexpand(parserobj, string, sindex):
node = None
zindex = sindex + 1
c = string[zindex] if zindex < len(string) else None
if c and c in '0123456789$#?-!*@':
# XXX 7685
node = ast.node(kind='parameter', value=c,
pos=(sindex, zindex+1))
elif c == '{':
# XXX 7863
# TODO not start enough, doesn't consider escaping
zindex = string.find('}', zindex + 1)
node = ast.node(kind='parameter', value=string[sindex+2:zindex],
pos=(sindex, zindex+1))
# TODO
# return _parameterbraceexpand(string, zindex)
elif c == '(':
return _extractcommandsubst(parserobj, string, zindex + 1)
elif c == '[':
raise NotImplementedError('arithmetic substitution')
#return _extractarithmeticsubst(string, zindex + 1)
else:
tindex = zindex
for zindex in range(tindex, len(string) + 1):
if zindex == len(string):
break
if not string[zindex].isalnum() and not string[zindex] == '_':
break
temp1 = string[sindex:zindex]
if temp1:
return (ast.node(kind='parameter', value=temp1[1:], pos=(sindex, zindex)),
zindex)
if zindex < len(string):
zindex += 1
return node, zindex
def p_command(p):
'''command : simple_command
| shell_command
| shell_command redirection_list
| function_def
| coproc'''
if isinstance(p[1], ast.node):
p[0] = p[1]
if len(p) == 3:
assert p[0].kind == 'compound'
p[0].redirects.extend(p[2])
assert p[0].pos[0] < p[0].redirects[-1].pos[1]
p[0].pos = (p[0].pos[0], p[0].redirects[-1].pos[1])
else:
p[0] = ast.node(kind='command', parts=p[1], pos=_partsspan(p[1]))
def p_command(p):
'''command : simple_command
| shell_command
| shell_command redirection_list
| function_def
| coproc'''
if isinstance(p[1], ast.node):
p[0] = p[1]
if len(p) == 3:
assert p[0].kind == 'compound'
p[0].redirects.extend(p[2])
assert p[0].pos[0] < p[0].redirects[-1].pos[1]
p[0].pos = (p[0].pos[0], p[0].redirects[-1].pos[1])
else:
p[0] = ast.node(kind='command', parts=p[1], pos=_partsspan(p[1]))
def makeheredoc(tokenizer, redirnode, lineno, killleading):
redirword = remove_escape(string_quote_removal(redirnode.output.word))
#redirword = redirnode.output.word
document = []
startpos = tokenizer._shell_input_line_index
#fullline = self.tok.readline(bool(redirword.output.flags & flags.word.QUOTED))
fullline = tokenizer.readline(False)
while fullline:
if killleading:
while fullline[0] == '\t':
fullline = fullline[1:]
if not fullline:
continue
if fullline[:-1] == redirword and fullline[len(redirword)] == '\n':
document.append(fullline[:-1])
# document_done
break
document.append(fullline)
#fullline = self.readline(bool(redirnode.flags & flags.word.QUOTED))
fullline = tokenizer.readline(False)
if not fullline:
raise errors.ParsingError(
"here-document at line %d delimited by end-of-file (wanted %r)" %
(lineno, redirword), tokenizer._shell_input_line,
tokenizer._shell_input_line_index)
document = ''.join(document)
endpos = tokenizer._shell_input_line_index - 1
assert hasattr(redirnode, 'heredoc')
num_of_lines = document.count('\n') + 1
redirnode.heredoc = ast.node(kind='heredoc',
value=document,
pos=(startpos, endpos),
lineno=num_of_lines)
# if the heredoc immediately follows this node, fix its end pos
if redirnode.pos[1] + 1 == startpos:
redirnode.pos = (redirnode.pos[0], endpos)
return document
def _makeparts(p):
parts = []
for i in range(1, len(p)):
if isinstance(p[i], ast.node):
parts.append(p[i])
elif isinstance(p[i], list):
parts.extend(p[i])
elif isinstance(p.slice[i], tokenizer.token):
if p.slice[i].ttype == tokenizer.tokentype.WORD:
parserobj = p.context
parts.append(_expandword(parserobj, p.slice[i]))
else:
parts.append(
ast.node(kind='reservedword', word=p[i], pos=p.lexspan(i)))
else:
pass
return parts
def _makeparts(p):
parts = []
for i in range(1, len(p)):
if isinstance(p[i], ast.node):
parts.append(p[i])
elif isinstance(p[i], list):
parts.extend(p[i])
elif isinstance(p.slice[i], tokenizer.token):
if p.slice[i].ttype == tokenizer.tokentype.WORD:
parserobj = p.context
parts.append(_expandword(parserobj, p.slice[i]))
else:
parts.append(ast.node(kind='reservedword', word=p[i],
pos=p.lexspan(i)))
else:
pass
return parts
def makeheredoc(tokenizer, redirnode, lineno, killleading):
# redirword = string_quote_removal(redirectnode.word)
redirword = redirnode.output.word
document = []
startpos = tokenizer._shell_input_line_index
# fullline = self.tok.readline(bool(redirword.output.flags & flags.word.QUOTED))
fullline = tokenizer.readline(False)
while fullline:
if killleading:
while fullline[0] == "\t":
fullline = fullline[1:]
if not fullline:
continue
if fullline[:-1] == redirword and fullline[len(redirword)] == "\n":
document.append(fullline[:-1])
# document_done
break
document.append(fullline)
# fullline = self.readline(bool(redirnode.flags & flags.word.QUOTED))
fullline = tokenizer.readline(False)
if not fullline:
raise errors.ParsingError(
"here-document at line %d delimited by end-of-file (wanted %r)" % (lineno, redirword),
tokenizer._shell_input_line,
tokenizer._shell_input_line_index,
)
document = "".join(document)
endpos = tokenizer._shell_input_line_index - 1
assert hasattr(redirnode, "heredoc")
redirnode.heredoc = ast.node(kind="heredoc", value=document, pos=(startpos, endpos))
# if the heredoc immediately follows this node, fix its end pos
if redirnode.pos[1] + 1 == startpos:
redirnode.pos = (redirnode.pos[0], endpos)
return document
def p_inputunit(p):
'''inputunit : simple_list simple_list_terminator
| NEWLINE
| error NEWLINE
| EOF'''
# XXX
if p.lexer._parserstate & flags.parser.CMDSUBST:
p.lexer._parserstate.add(flags.parser.EOFTOKEN)
if isinstance(p[1], ast.node):
p[0] = p[1]
# accept right here in case the input contains more lines that are
# not part of the current command
p.accept()
if p.slice[1].ttype == tokenizer.tokentype.NEWLINE:
p[0] = ast.node(kind='newline',
pos=(p.lexpos(1), p.lexpos(1)),
lineno=1)
p.accept()
def commandnode(s, *parts):
return ast.node(kind='command', s=s, parts=list(parts))
def functionnode(s, name, body, *parts):
return ast.node(kind='function', name=name, body=body, parts=list(parts), s=s)
def whilenode(s, *parts):
return ast.node(kind='while', parts=list(parts), s=s)
def fornode(s, *parts):
return ast.node(kind='for', parts=list(parts), s=s)
def ifnode(s, *parts):
return ast.node(kind='if', parts=list(parts), s=s)
def comsubnode(s, command):
return ast.node(kind='commandsubstitution', s=s, command=command)