1 # -*- coding: iso-8859-1 -*-
2 """A lexical analyzer class for simple shell-like syntaxes."""
4 # Module and documentation by Eric S. Raymond, 21 Dec 1998
5 # Input stacking and error message cleanup added by ESR, March 2000
6 # push_source() and pop_source() made explicit by ESR, January 2001.
7 # Posix compliance, split(), string arguments, and
8 # iterator interface by Gustavo Niemeyer, April 2003.
12 from collections import deque
15 from cStringIO import StringIO
17 from StringIO import StringIO
19 __all__ = ["shlex", "split"]
22 "A lexical analyzer class for simple shell-like syntaxes."
23 def __init__(self, instream=None, infile=None, posix=False):
24 if isinstance(instream, basestring):
25 instream = StringIO(instream)
26 if instream is not None:
27 self.instream = instream
30 self.instream = sys.stdin
38 self.wordchars = ('abcdfeghijklmnopqrstuvwxyz'
39 'ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789_')
41 self.wordchars += ('ßàáâãäåæçèéêëìíîïðñòóôõöøùúûüýþÿ'
42 'ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝÞ')
43 self.whitespace = ' \t\r\n'
44 self.whitespace_split = False
47 self.escapedquotes = '"'
49 self.pushback = deque()
53 self.filestack = deque()
56 print 'shlex: reading from %s, line %d' \
57 % (self.instream, self.lineno)
59 def push_token(self, tok):
60 "Push a token onto the stack popped by the get_token method"
62 print "shlex: pushing token " + repr(tok)
63 self.pushback.appendleft(tok)
65 def push_source(self, newstream, newfile=None):
66 "Push an input source onto the lexer's input source stack."
67 if isinstance(newstream, basestring):
68 newstream = StringIO(newstream)
69 self.filestack.appendleft((self.infile, self.instream, self.lineno))
71 self.instream = newstream
74 if newfile is not None:
75 print 'shlex: pushing to file %s' % (self.infile,)
77 print 'shlex: pushing to stream %s' % (self.instream,)
80 "Pop the input source stack."
82 (self.infile, self.instream, self.lineno) = self.filestack.popleft()
84 print 'shlex: popping to %s, line %d' \
85 % (self.instream, self.lineno)
89 "Get a token from the input stream (or from stack if it's nonempty)"
91 tok = self.pushback.popleft()
93 print "shlex: popping token " + repr(tok)
95 # No pushback. Get a token.
96 raw = self.read_token()
98 if self.source is not None:
99 while raw == self.source:
100 spec = self.sourcehook(self.read_token())
102 (newfile, newstream) = spec
103 self.push_source(newstream, newfile)
104 raw = self.get_token()
105 # Maybe we got EOF instead?
106 while raw == self.eof:
107 if not self.filestack:
111 raw = self.get_token()
112 # Neither inclusion nor EOF
115 print "shlex: token=" + repr(raw)
117 print "shlex: token=EOF"
120 def read_token(self):
124 nextchar = self.instream.read(1)
126 self.lineno = self.lineno + 1
128 print "shlex: in state", repr(self.state), \
129 "I see character:", repr(nextchar)
130 if self.state is None:
131 self.token = '' # past end of file
133 elif self.state == ' ':
135 self.state = None # end of file
137 elif nextchar in self.whitespace:
139 print "shlex: I see whitespace in whitespace state"
140 if self.token or (self.posix and quoted):
141 break # emit current token
144 elif nextchar in self.commenters:
145 self.instream.readline()
146 self.lineno = self.lineno + 1
147 elif self.posix and nextchar in self.escape:
149 self.state = nextchar
150 elif nextchar in self.wordchars:
151 self.token = nextchar
153 elif nextchar in self.quotes:
155 self.token = nextchar
156 self.state = nextchar
157 elif self.whitespace_split:
158 self.token = nextchar
161 self.token = nextchar
162 if self.token or (self.posix and quoted):
163 break # emit current token
166 elif self.state in self.quotes:
168 if not nextchar: # end of file
170 print "shlex: I see EOF in quotes state"
171 # XXX what error should be raised here?
172 raise ValueError, "No closing quotation"
173 if nextchar == self.state:
175 self.token = self.token + nextchar
180 elif self.posix and nextchar in self.escape and \
181 self.state in self.escapedquotes:
182 escapedstate = self.state
183 self.state = nextchar
185 self.token = self.token + nextchar
186 elif self.state in self.escape:
187 if not nextchar: # end of file
189 print "shlex: I see EOF in escape state"
190 # XXX what error should be raised here?
191 raise ValueError, "No escaped character"
192 # In posix shells, only the quote itself or the escape
193 # character may be escaped within quotes.
194 if escapedstate in self.quotes and \
195 nextchar != self.state and nextchar != escapedstate:
196 self.token = self.token + self.state
197 self.token = self.token + nextchar
198 self.state = escapedstate
199 elif self.state == 'a':
201 self.state = None # end of file
203 elif nextchar in self.whitespace:
205 print "shlex: I see whitespace in word state"
207 if self.token or (self.posix and quoted):
208 break # emit current token
211 elif nextchar in self.commenters:
212 self.instream.readline()
213 self.lineno = self.lineno + 1
216 if self.token or (self.posix and quoted):
217 break # emit current token
220 elif self.posix and nextchar in self.quotes:
221 self.state = nextchar
222 elif self.posix and nextchar in self.escape:
224 self.state = nextchar
225 elif nextchar in self.wordchars or nextchar in self.quotes \
226 or self.whitespace_split:
227 self.token = self.token + nextchar
229 self.pushback.appendleft(nextchar)
231 print "shlex: I see punctuation in word state"
234 break # emit current token
239 if self.posix and not quoted and result == '':
243 print "shlex: raw token=" + repr(result)
245 print "shlex: raw token=EOF"
248 def sourcehook(self, newfile):
249 "Hook called on a filename to be sourced."
250 if newfile[0] == '"':
251 newfile = newfile[1:-1]
252 # This implements cpp-like semantics for relative-path inclusion.
253 if isinstance(self.infile, basestring) and not os.path.isabs(newfile):
254 newfile = os.path.join(os.path.dirname(self.infile), newfile)
255 return (newfile, open(newfile, "r"))
257 def error_leader(self, infile=None, lineno=None):
258 "Emit a C-compiler-like, Emacs-friendly error-message leader."
263 return "\"%s\", line %d: " % (infile, lineno)
269 token = self.get_token()
270 if token == self.eof:
274 def split(s, comments=False):
275 lex = shlex(s, posix=True)
276 lex.whitespace_split = True
281 if __name__ == '__main__':
282 if len(sys.argv) == 1:
286 lexer = shlex(open(file), file)
288 tt = lexer.get_token()
290 print "Token: " + repr(tt)