apibuild.py

xml开源解析代码.版本为libxml2-2.6.29,可支持GB3212.网络消息发送XML时很有用.

			    self.line = line[i+1:]
			    line = line[:i]
			    l = i
			    found = 1
			    break
			if line[i] == '\\':
			    i = i + 1
			i = i + 1
		    tok = tok + line
		    if found == 0:
		        line = self.getline()
			if line == None:
			    return None
		self.last = ('string', tok)
		return self.last

	    if l >= 2 and line[0] == '/' and line[1] == '*':
	        line = line[2:]
		found = 0
		tok = ""
		while found == 0:
		    i = 0
		    l = len(line)
		    while i < l:
			if line[i] == '*' and i+1 < l and line[i+1] == '/':
			    self.line = line[i+2:]
			    line = line[:i-1]
			    l = i
			    found = 1
			    break
			i = i + 1
	            if tok != "":
		        tok = tok + "\n"
		    tok = tok + line
		    if found == 0:
		        line = self.getline()
			if line == None:
			    return None
		self.last = ('comment', tok)
		return self.last
	    if l >= 2 and line[0] == '/' and line[1] == '/':
	        line = line[2:]
		self.last = ('comment', line)
		return self.last
	    i = 0
	    while i < l:
	        if line[i] == '/' and i+1 < l and line[i+1] == '/':
		    self.line = line[i:]
		    line = line[:i]
		    break
	        if line[i] == '/' and i+1 < l and line[i+1] == '*':
		    self.line = line[i:]
		    line = line[:i]
		    break
		if line[i] == '"' or line[i] == "'":
		    self.line = line[i:]
		    line = line[:i]
		    break
		i = i + 1
	    l = len(line)
	    i = 0
	    while i < l:
	        if line[i] == ' ' or line[i] == '\t':
		    i = i + 1
		    continue
		o = ord(line[i])
		if (o >= 97 and o <= 122) or (o >= 65 and o <= 90) or \
		   (o >= 48 and o <= 57):
		    s = i
		    while i < l:
			o = ord(line[i])
			if (o >= 97 and o <= 122) or (o >= 65 and o <= 90) or \
			   (o >= 48 and o <= 57) or string.find(
			       " \t(){}:;,+-*/%&!|[]=><", line[i]) == -1:
			    i = i + 1
			else:
			    break
		    self.tokens.append(('name', line[s:i]))
		    continue
		if string.find("(){}:;,[]", line[i]) != -1:
#                 if line[i] == '(' or line[i] == ')' or line[i] == '{' or \
#		    line[i] == '}' or line[i] == ':' or line[i] == ';' or \
#		    line[i] == ',' or line[i] == '[' or line[i] == ']':
		    self.tokens.append(('sep', line[i]))
		    i = i + 1
		    continue
		if string.find("+-*><=/%&!|.", line[i]) != -1:
#                 if line[i] == '+' or line[i] == '-' or line[i] == '*' or \
#		    line[i] == '>' or line[i] == '<' or line[i] == '=' or \
#		    line[i] == '/' or line[i] == '%' or line[i] == '&' or \
#		    line[i] == '!' or line[i] == '|' or line[i] == '.':
		    if line[i] == '.' and  i + 2 < l and \
		       line[i+1] == '.' and line[i+2] == '.':
			self.tokens.append(('name', '...'))
			i = i + 3
			continue

		    j = i + 1
		    if j < l and (
		       string.find("+-*><=/%&!|", line[j]) != -1):
#		        line[j] == '+' or line[j] == '-' or line[j] == '*' or \
#			line[j] == '>' or line[j] == '<' or line[j] == '=' or \
#			line[j] == '/' or line[j] == '%' or line[j] == '&' or \
#			line[j] == '!' or line[j] == '|'):
			self.tokens.append(('op', line[i:j+1]))
			i = j + 1
		    else:
			self.tokens.append(('op', line[i]))
			i = i + 1
		    continue
		s = i
		while i < l:
		    o = ord(line[i])
		    if (o >= 97 and o <= 122) or (o >= 65 and o <= 90) or \
		       (o >= 48 and o <= 57) or (
		        string.find(" \t(){}:;,+-*/%&!|[]=><", line[i]) == -1):
#		         line[i] != ' ' and line[i] != '\t' and
#			 line[i] != '(' and line[i] != ')' and
#			 line[i] != '{'  and line[i] != '}' and
#			 line[i] != ':' and line[i] != ';' and
#			 line[i] != ',' and line[i] != '+' and
#			 line[i] != '-' and line[i] != '*' and
#			 line[i] != '/' and line[i] != '%' and
#			 line[i] != '&' and line[i] != '!' and
#			 line[i] != '|' and line[i] != '[' and
#			 line[i] != ']' and line[i] != '=' and
#			 line[i] != '*' and line[i] != '>' and
#			 line[i] != '<'):
			i = i + 1
		    else:
		        break
		self.tokens.append(('name', line[s:i]))
             
	tok = self.tokens[0]
	self.tokens = self.tokens[1:]
	self.last = tok
	return tok
      
class CParser:
    """The C module parser"""
    def __init__(self, filename, idx = None):
        self.filename = filename
	if len(filename) > 2 and filename[-2:] == '.h':
	    self.is_header = 1
	else:
	    self.is_header = 0
        self.input = open(filename)
	self.lexer = CLexer(self.input)
	if idx == None:
	    self.index = index()
	else:
	    self.index = idx
	self.top_comment = ""
	self.last_comment = ""
	self.comment = None
	self.collect_ref = 0
	self.no_error = 0
	self.conditionals = []
	self.defines = []

    def collect_references(self):
        self.collect_ref = 1

    def stop_error(self):
        self.no_error = 1

    def start_error(self):
        self.no_error = 0

    def lineno(self):
        return self.lexer.getlineno()

    def index_add(self, name, module, static, type, info=None, extra = None):
	if self.is_header == 1:
	    self.index.add(name, module, module, static, type, self.lineno(),
			   info, extra, self.conditionals)
	else:
	    self.index.add(name, None, module, static, type, self.lineno(),
			   info, extra, self.conditionals)

    def index_add_ref(self, name, module, static, type, info=None,
                      extra = None):
	if self.is_header == 1:
	    self.index.add_ref(name, module, module, static, type,
	                       self.lineno(), info, extra, self.conditionals)
	else:
	    self.index.add_ref(name, None, module, static, type, self.lineno(),
			       info, extra, self.conditionals)

    def warning(self, msg):
        if self.no_error:
	    return
	print msg

    def error(self, msg, token=-1):
        if self.no_error:
	    return

        print "Parse Error: " + msg
	if token != -1:
	    print "Got token ", token
	self.lexer.debug()
	sys.exit(1)

    def debug(self, msg, token=-1):
        print "Debug: " + msg
	if token != -1:
	    print "Got token ", token
	self.lexer.debug()

    def parseTopComment(self, comment):
	res = {}
	lines = string.split(comment, "\n")
	item = None
	for line in lines:
	    while line != "" and (line[0] == ' ' or line[0] == '\t'):
		line = line[1:]
	    while line != "" and line[0] == '*':
		line = line[1:]
	    while line != "" and (line[0] == ' ' or line[0] == '\t'):
		line = line[1:]
	    try:
		(it, line) = string.split(line, ":", 1)
		item = it
		while line != "" and (line[0] == ' ' or line[0] == '\t'):
		    line = line[1:]
		if res.has_key(item):
		    res[item] = res[item] + " " + line
		else:
		    res[item] = line
	    except:
		if item != None:
		    if res.has_key(item):
			res[item] = res[item] + " " + line
		    else:
			res[item] = line
	self.index.info = res

    def parseComment(self, token):
        if self.top_comment == "":
	    self.top_comment = token[1]
	if self.comment == None or token[1][0] == '*':
	    self.comment = token[1];
	else:
	    self.comment = self.comment + token[1]
	token = self.lexer.token()

        if string.find(self.comment, "DOC_DISABLE") != -1:
	    self.stop_error()

        if string.find(self.comment, "DOC_ENABLE") != -1:
	    self.start_error()

	return token

    #
    # Parse a comment block associate to a typedef
    #
    def parseTypeComment(self, name, quiet = 0):
        if name[0:2] == '__':
	    quiet = 1

        args = []
	desc = ""

        if self.comment == None:
	    if not quiet:
		self.warning("Missing comment for type %s" % (name))
	    return((args, desc))
        if self.comment[0] != '*':
	    if not quiet:
		self.warning("Missing * in type comment for %s" % (name))
	    return((args, desc))
	lines = string.split(self.comment, '\n')
	if lines[0] == '*':
	    del lines[0]
	if lines[0] != "* %s:" % (name):
	    if not quiet:
		self.warning("Misformatted type comment for %s" % (name))
		self.warning("  Expecting '* %s:' got '%s'" % (name, lines[0]))
	    return((args, desc))
	del lines[0]
	while len(lines) > 0 and lines[0] == '*':
	    del lines[0]
	desc = ""
	while len(lines) > 0:
	    l = lines[0]
	    while len(l) > 0 and l[0] == '*':
	        l = l[1:]
	    l = string.strip(l)
	    desc = desc + " " + l
	    del lines[0]
		     
	desc = string.strip(desc)

	if quiet == 0:
	    if desc == "":
	        self.warning("Type comment for %s lack description of the macro" % (name))

	return(desc)
    #
    # Parse a comment block associate to a macro
    #
    def parseMacroComment(self, name, quiet = 0):
        if name[0:2] == '__':
	    quiet = 1

        args = []
	desc = ""

        if self.comment == None:
	    if not quiet:
		self.warning("Missing comment for macro %s" % (name))
	    return((args, desc))
        if self.comment[0] != '*':
	    if not quiet:
		self.warning("Missing * in macro comment for %s" % (name))
	    return((args, desc))
	lines = string.split(self.comment, '\n')
	if lines[0] == '*':
	    del lines[0]
	if lines[0] != "* %s:" % (name):
	    if not quiet:
		self.warning("Misformatted macro comment for %s" % (name))
		self.warning("  Expecting '* %s:' got '%s'" % (name, lines[0]))
	    return((args, desc))
	del lines[0]
	while lines[0] == '*':
	    del lines[0]
	while len(lines) > 0 and lines[0][0:3] == '* @':
	    l = lines[0][3:]
	    try:
	        (arg, desc) = string.split(l, ':', 1)
		desc=string.strip(desc)
		arg=string.strip(arg)
            except:
		if not quiet:
		    self.warning("Misformatted macro comment for %s" % (name))
		    self.warning("  problem with '%s'" % (lines[0]))
		del lines[0]
		continue
	    del lines[0]
	    l = string.strip(lines[0])
	    while len(l) > 2 and l[0:3] != '* @':
	        while l[0] == '*':
		    l = l[1:]
		desc = desc + ' ' + string.strip(l)
		del lines[0]
		if len(lines) == 0:
		    break
		l = lines[0]
            args.append((arg, desc))
	while len(lines) > 0 and lines[0] == '*':
	    del lines[0]
	desc = ""
	while len(lines) > 0:
	    l = lines[0]
	    while len(l) > 0 and l[0] == '*':
	        l = l[1:]
	    l = string.strip(l)
	    desc = desc + " " + l
	    del lines[0]
		     
	desc = string.strip(desc)

	if quiet == 0:
	    if desc == "":
	        self.warning("Macro comment for %s lack description of the macro" % (name))

	return((args, desc))

     #
     # Parse a comment block and merge the informations found in the
     # parameters descriptions, finally returns a block as complete
     # as possible
     #
    def mergeFunctionComment(self, name, description, quiet = 0):
        if name == 'main':
	    quiet = 1
        if name[0:2] == '__':
	    quiet = 1

	(ret, args) = description
	desc = ""
	retdesc = ""

        if self.comment == None:
	    if not quiet:
		self.warning("Missing comment for function %s" % (name))
	    return(((ret[0], retdesc), args, desc))
        if self.comment[0] != '*':
	    if not quiet:
		self.warning("Missing * in function comment for %s" % (name))
	    return(((ret[0], retdesc), args, desc))
	lines = string.split(self.comment, '\n')
	if lines[0] == '*':
	    del lines[0]
	if lines[0] != "* %s:" % (name):
	    if not quiet:
		self.warning("Misformatted function comment for %s" % (name))
		self.warning("  Expecting '* %s:' got '%s'" % (name, lines[0]))
	    return(((ret[0], retdesc), args, desc))
	del lines[0]
	while lines[0] == '*':
	    del lines[0]
	nbargs = len(args)
	while len(lines) > 0 and lines[0][0:3] == '* @':
	    l = lines[0][3:]
	    try:
	        (arg, desc) = string.split(l, ':', 1)
		desc=string.strip(desc)
		arg=string.strip(arg)
            except:
		if not quiet:
		    self.warning("Misformatted function comment for %s" % (name))
		    self.warning("  problem with '%s'" % (lines[0]))
		del lines[0]
		continue
	    del lines[0]
	    l = string.strip(lines[0])
	    while len(l) > 2 and l[0:3] != '* @':
	        while l[0] == '*':
		    l = l[1:]
		desc = desc + ' ' + string.strip(l)
		del lines[0]
		if len(lines) == 0:
		    break