xref: /aosp_15_r20/prebuilts/build-tools/common/py3-stdlib/_markupbase.py (revision cda5da8d549138a6648c5ee6d7a49cf8f4a657be) 
1*cda5da8dSAndroid Build Coastguard Worker"""Shared support for scanning document type declarations in HTML and XHTML.
2*cda5da8dSAndroid Build Coastguard Worker
3*cda5da8dSAndroid Build Coastguard WorkerThis module is used as a foundation for the html.parser module.  It has no
4*cda5da8dSAndroid Build Coastguard Workerdocumented public API and should not be used directly.
5*cda5da8dSAndroid Build Coastguard Worker
6*cda5da8dSAndroid Build Coastguard Worker"""
7*cda5da8dSAndroid Build Coastguard Worker
8*cda5da8dSAndroid Build Coastguard Workerimport re
9*cda5da8dSAndroid Build Coastguard Worker
10*cda5da8dSAndroid Build Coastguard Worker_declname_match = re.compile(r'[a-zA-Z][-_.a-zA-Z0-9]*\s*').match
11*cda5da8dSAndroid Build Coastguard Worker_declstringlit_match = re.compile(r'(\'[^\']*\'|"[^"]*")\s*').match
12*cda5da8dSAndroid Build Coastguard Worker_commentclose = re.compile(r'--\s*>')
13*cda5da8dSAndroid Build Coastguard Worker_markedsectionclose = re.compile(r']\s*]\s*>')
14*cda5da8dSAndroid Build Coastguard Worker
15*cda5da8dSAndroid Build Coastguard Worker# An analysis of the MS-Word extensions is available at
16*cda5da8dSAndroid Build Coastguard Worker# http://www.planetpublish.com/xmlarena/xap/Thursday/WordtoXML.pdf
17*cda5da8dSAndroid Build Coastguard Worker
18*cda5da8dSAndroid Build Coastguard Worker_msmarkedsectionclose = re.compile(r']\s*>')
19*cda5da8dSAndroid Build Coastguard Worker
20*cda5da8dSAndroid Build Coastguard Workerdel re
21*cda5da8dSAndroid Build Coastguard Worker
22*cda5da8dSAndroid Build Coastguard Worker
23*cda5da8dSAndroid Build Coastguard Workerclass ParserBase:
24*cda5da8dSAndroid Build Coastguard Worker    """Parser base class which provides some common support methods used
25*cda5da8dSAndroid Build Coastguard Worker    by the SGML/HTML and XHTML parsers."""
26*cda5da8dSAndroid Build Coastguard Worker
27*cda5da8dSAndroid Build Coastguard Worker    def __init__(self):
28*cda5da8dSAndroid Build Coastguard Worker        if self.__class__ is ParserBase:
29*cda5da8dSAndroid Build Coastguard Worker            raise RuntimeError(
30*cda5da8dSAndroid Build Coastguard Worker                "_markupbase.ParserBase must be subclassed")
31*cda5da8dSAndroid Build Coastguard Worker
32*cda5da8dSAndroid Build Coastguard Worker    def reset(self):
33*cda5da8dSAndroid Build Coastguard Worker        self.lineno = 1
34*cda5da8dSAndroid Build Coastguard Worker        self.offset = 0
35*cda5da8dSAndroid Build Coastguard Worker
36*cda5da8dSAndroid Build Coastguard Worker    def getpos(self):
37*cda5da8dSAndroid Build Coastguard Worker        """Return current line number and offset."""
38*cda5da8dSAndroid Build Coastguard Worker        return self.lineno, self.offset
39*cda5da8dSAndroid Build Coastguard Worker
40*cda5da8dSAndroid Build Coastguard Worker    # Internal -- update line number and offset.  This should be
41*cda5da8dSAndroid Build Coastguard Worker    # called for each piece of data exactly once, in order -- in other
42*cda5da8dSAndroid Build Coastguard Worker    # words the concatenation of all the input strings to this
43*cda5da8dSAndroid Build Coastguard Worker    # function should be exactly the entire input.
44*cda5da8dSAndroid Build Coastguard Worker    def updatepos(self, i, j):
45*cda5da8dSAndroid Build Coastguard Worker        if i >= j:
46*cda5da8dSAndroid Build Coastguard Worker            return j
47*cda5da8dSAndroid Build Coastguard Worker        rawdata = self.rawdata
48*cda5da8dSAndroid Build Coastguard Worker        nlines = rawdata.count("\n", i, j)
49*cda5da8dSAndroid Build Coastguard Worker        if nlines:
50*cda5da8dSAndroid Build Coastguard Worker            self.lineno = self.lineno + nlines
51*cda5da8dSAndroid Build Coastguard Worker            pos = rawdata.rindex("\n", i, j) # Should not fail
52*cda5da8dSAndroid Build Coastguard Worker            self.offset = j-(pos+1)
53*cda5da8dSAndroid Build Coastguard Worker        else:
54*cda5da8dSAndroid Build Coastguard Worker            self.offset = self.offset + j-i
55*cda5da8dSAndroid Build Coastguard Worker        return j
56*cda5da8dSAndroid Build Coastguard Worker
57*cda5da8dSAndroid Build Coastguard Worker    _decl_otherchars = ''
58*cda5da8dSAndroid Build Coastguard Worker
59*cda5da8dSAndroid Build Coastguard Worker    # Internal -- parse declaration (for use by subclasses).
60*cda5da8dSAndroid Build Coastguard Worker    def parse_declaration(self, i):
61*cda5da8dSAndroid Build Coastguard Worker        # This is some sort of declaration; in "HTML as
62*cda5da8dSAndroid Build Coastguard Worker        # deployed," this should only be the document type
63*cda5da8dSAndroid Build Coastguard Worker        # declaration ("<!DOCTYPE html...>").
64*cda5da8dSAndroid Build Coastguard Worker        # ISO 8879:1986, however, has more complex
65*cda5da8dSAndroid Build Coastguard Worker        # declaration syntax for elements in <!...>, including:
66*cda5da8dSAndroid Build Coastguard Worker        # --comment--
67*cda5da8dSAndroid Build Coastguard Worker        # [marked section]
68*cda5da8dSAndroid Build Coastguard Worker        # name in the following list: ENTITY, DOCTYPE, ELEMENT,
69*cda5da8dSAndroid Build Coastguard Worker        # ATTLIST, NOTATION, SHORTREF, USEMAP,
70*cda5da8dSAndroid Build Coastguard Worker        # LINKTYPE, LINK, IDLINK, USELINK, SYSTEM
71*cda5da8dSAndroid Build Coastguard Worker        rawdata = self.rawdata
72*cda5da8dSAndroid Build Coastguard Worker        j = i + 2
73*cda5da8dSAndroid Build Coastguard Worker        assert rawdata[i:j] == "<!", "unexpected call to parse_declaration"
74*cda5da8dSAndroid Build Coastguard Worker        if rawdata[j:j+1] == ">":
75*cda5da8dSAndroid Build Coastguard Worker            # the empty comment <!>
76*cda5da8dSAndroid Build Coastguard Worker            return j + 1
77*cda5da8dSAndroid Build Coastguard Worker        if rawdata[j:j+1] in ("-", ""):
78*cda5da8dSAndroid Build Coastguard Worker            # Start of comment followed by buffer boundary,
79*cda5da8dSAndroid Build Coastguard Worker            # or just a buffer boundary.
80*cda5da8dSAndroid Build Coastguard Worker            return -1
81*cda5da8dSAndroid Build Coastguard Worker        # A simple, practical version could look like: ((name|stringlit) S*) + '>'
82*cda5da8dSAndroid Build Coastguard Worker        n = len(rawdata)
83*cda5da8dSAndroid Build Coastguard Worker        if rawdata[j:j+2] == '--': #comment
84*cda5da8dSAndroid Build Coastguard Worker            # Locate --.*-- as the body of the comment
85*cda5da8dSAndroid Build Coastguard Worker            return self.parse_comment(i)
86*cda5da8dSAndroid Build Coastguard Worker        elif rawdata[j] == '[': #marked section
87*cda5da8dSAndroid Build Coastguard Worker            # Locate [statusWord [...arbitrary SGML...]] as the body of the marked section
88*cda5da8dSAndroid Build Coastguard Worker            # Where statusWord is one of TEMP, CDATA, IGNORE, INCLUDE, RCDATA
89*cda5da8dSAndroid Build Coastguard Worker            # Note that this is extended by Microsoft Office "Save as Web" function
90*cda5da8dSAndroid Build Coastguard Worker            # to include [if...] and [endif].
91*cda5da8dSAndroid Build Coastguard Worker            return self.parse_marked_section(i)
92*cda5da8dSAndroid Build Coastguard Worker        else: #all other declaration elements
93*cda5da8dSAndroid Build Coastguard Worker            decltype, j = self._scan_name(j, i)
94*cda5da8dSAndroid Build Coastguard Worker        if j < 0:
95*cda5da8dSAndroid Build Coastguard Worker            return j
96*cda5da8dSAndroid Build Coastguard Worker        if decltype == "doctype":
97*cda5da8dSAndroid Build Coastguard Worker            self._decl_otherchars = ''
98*cda5da8dSAndroid Build Coastguard Worker        while j < n:
99*cda5da8dSAndroid Build Coastguard Worker            c = rawdata[j]
100*cda5da8dSAndroid Build Coastguard Worker            if c == ">":
101*cda5da8dSAndroid Build Coastguard Worker                # end of declaration syntax
102*cda5da8dSAndroid Build Coastguard Worker                data = rawdata[i+2:j]
103*cda5da8dSAndroid Build Coastguard Worker                if decltype == "doctype":
104*cda5da8dSAndroid Build Coastguard Worker                    self.handle_decl(data)
105*cda5da8dSAndroid Build Coastguard Worker                else:
106*cda5da8dSAndroid Build Coastguard Worker                    # According to the HTML5 specs sections "8.2.4.44 Bogus
107*cda5da8dSAndroid Build Coastguard Worker                    # comment state" and "8.2.4.45 Markup declaration open
108*cda5da8dSAndroid Build Coastguard Worker                    # state", a comment token should be emitted.
109*cda5da8dSAndroid Build Coastguard Worker                    # Calling unknown_decl provides more flexibility though.
110*cda5da8dSAndroid Build Coastguard Worker                    self.unknown_decl(data)
111*cda5da8dSAndroid Build Coastguard Worker                return j + 1
112*cda5da8dSAndroid Build Coastguard Worker            if c in "\"'":
113*cda5da8dSAndroid Build Coastguard Worker                m = _declstringlit_match(rawdata, j)
114*cda5da8dSAndroid Build Coastguard Worker                if not m:
115*cda5da8dSAndroid Build Coastguard Worker                    return -1 # incomplete
116*cda5da8dSAndroid Build Coastguard Worker                j = m.end()
117*cda5da8dSAndroid Build Coastguard Worker            elif c in "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ":
118*cda5da8dSAndroid Build Coastguard Worker                name, j = self._scan_name(j, i)
119*cda5da8dSAndroid Build Coastguard Worker            elif c in self._decl_otherchars:
120*cda5da8dSAndroid Build Coastguard Worker                j = j + 1
121*cda5da8dSAndroid Build Coastguard Worker            elif c == "[":
122*cda5da8dSAndroid Build Coastguard Worker                # this could be handled in a separate doctype parser
123*cda5da8dSAndroid Build Coastguard Worker                if decltype == "doctype":
124*cda5da8dSAndroid Build Coastguard Worker                    j = self._parse_doctype_subset(j + 1, i)
125*cda5da8dSAndroid Build Coastguard Worker                elif decltype in {"attlist", "linktype", "link", "element"}:
126*cda5da8dSAndroid Build Coastguard Worker                    # must tolerate []'d groups in a content model in an element declaration
127*cda5da8dSAndroid Build Coastguard Worker                    # also in data attribute specifications of attlist declaration
128*cda5da8dSAndroid Build Coastguard Worker                    # also link type declaration subsets in linktype declarations
129*cda5da8dSAndroid Build Coastguard Worker                    # also link attribute specification lists in link declarations
130*cda5da8dSAndroid Build Coastguard Worker                    raise AssertionError("unsupported '[' char in %s declaration" % decltype)
131*cda5da8dSAndroid Build Coastguard Worker                else:
132*cda5da8dSAndroid Build Coastguard Worker                    raise AssertionError("unexpected '[' char in declaration")
133*cda5da8dSAndroid Build Coastguard Worker            else:
134*cda5da8dSAndroid Build Coastguard Worker                raise AssertionError("unexpected %r char in declaration" % rawdata[j])
135*cda5da8dSAndroid Build Coastguard Worker            if j < 0:
136*cda5da8dSAndroid Build Coastguard Worker                return j
137*cda5da8dSAndroid Build Coastguard Worker        return -1 # incomplete
138*cda5da8dSAndroid Build Coastguard Worker
139*cda5da8dSAndroid Build Coastguard Worker    # Internal -- parse a marked section
140*cda5da8dSAndroid Build Coastguard Worker    # Override this to handle MS-word extension syntax <![if word]>content<![endif]>
141*cda5da8dSAndroid Build Coastguard Worker    def parse_marked_section(self, i, report=1):
142*cda5da8dSAndroid Build Coastguard Worker        rawdata= self.rawdata
143*cda5da8dSAndroid Build Coastguard Worker        assert rawdata[i:i+3] == '<![', "unexpected call to parse_marked_section()"
144*cda5da8dSAndroid Build Coastguard Worker        sectName, j = self._scan_name( i+3, i )
145*cda5da8dSAndroid Build Coastguard Worker        if j < 0:
146*cda5da8dSAndroid Build Coastguard Worker            return j
147*cda5da8dSAndroid Build Coastguard Worker        if sectName in {"temp", "cdata", "ignore", "include", "rcdata"}:
148*cda5da8dSAndroid Build Coastguard Worker            # look for standard ]]> ending
149*cda5da8dSAndroid Build Coastguard Worker            match= _markedsectionclose.search(rawdata, i+3)
150*cda5da8dSAndroid Build Coastguard Worker        elif sectName in {"if", "else", "endif"}:
151*cda5da8dSAndroid Build Coastguard Worker            # look for MS Office ]> ending
152*cda5da8dSAndroid Build Coastguard Worker            match= _msmarkedsectionclose.search(rawdata, i+3)
153*cda5da8dSAndroid Build Coastguard Worker        else:
154*cda5da8dSAndroid Build Coastguard Worker            raise AssertionError(
155*cda5da8dSAndroid Build Coastguard Worker                'unknown status keyword %r in marked section' % rawdata[i+3:j]
156*cda5da8dSAndroid Build Coastguard Worker            )
157*cda5da8dSAndroid Build Coastguard Worker        if not match:
158*cda5da8dSAndroid Build Coastguard Worker            return -1
159*cda5da8dSAndroid Build Coastguard Worker        if report:
160*cda5da8dSAndroid Build Coastguard Worker            j = match.start(0)
161*cda5da8dSAndroid Build Coastguard Worker            self.unknown_decl(rawdata[i+3: j])
162*cda5da8dSAndroid Build Coastguard Worker        return match.end(0)
163*cda5da8dSAndroid Build Coastguard Worker
164*cda5da8dSAndroid Build Coastguard Worker    # Internal -- parse comment, return length or -1 if not terminated
165*cda5da8dSAndroid Build Coastguard Worker    def parse_comment(self, i, report=1):
166*cda5da8dSAndroid Build Coastguard Worker        rawdata = self.rawdata
167*cda5da8dSAndroid Build Coastguard Worker        if rawdata[i:i+4] != '<!--':
168*cda5da8dSAndroid Build Coastguard Worker            raise AssertionError('unexpected call to parse_comment()')
169*cda5da8dSAndroid Build Coastguard Worker        match = _commentclose.search(rawdata, i+4)
170*cda5da8dSAndroid Build Coastguard Worker        if not match:
171*cda5da8dSAndroid Build Coastguard Worker            return -1
172*cda5da8dSAndroid Build Coastguard Worker        if report:
173*cda5da8dSAndroid Build Coastguard Worker            j = match.start(0)
174*cda5da8dSAndroid Build Coastguard Worker            self.handle_comment(rawdata[i+4: j])
175*cda5da8dSAndroid Build Coastguard Worker        return match.end(0)
176*cda5da8dSAndroid Build Coastguard Worker
177*cda5da8dSAndroid Build Coastguard Worker    # Internal -- scan past the internal subset in a <!DOCTYPE declaration,
178*cda5da8dSAndroid Build Coastguard Worker    # returning the index just past any whitespace following the trailing ']'.
179*cda5da8dSAndroid Build Coastguard Worker    def _parse_doctype_subset(self, i, declstartpos):
180*cda5da8dSAndroid Build Coastguard Worker        rawdata = self.rawdata
181*cda5da8dSAndroid Build Coastguard Worker        n = len(rawdata)
182*cda5da8dSAndroid Build Coastguard Worker        j = i
183*cda5da8dSAndroid Build Coastguard Worker        while j < n:
184*cda5da8dSAndroid Build Coastguard Worker            c = rawdata[j]
185*cda5da8dSAndroid Build Coastguard Worker            if c == "<":
186*cda5da8dSAndroid Build Coastguard Worker                s = rawdata[j:j+2]
187*cda5da8dSAndroid Build Coastguard Worker                if s == "<":
188*cda5da8dSAndroid Build Coastguard Worker                    # end of buffer; incomplete
189*cda5da8dSAndroid Build Coastguard Worker                    return -1
190*cda5da8dSAndroid Build Coastguard Worker                if s != "<!":
191*cda5da8dSAndroid Build Coastguard Worker                    self.updatepos(declstartpos, j + 1)
192*cda5da8dSAndroid Build Coastguard Worker                    raise AssertionError(
193*cda5da8dSAndroid Build Coastguard Worker                        "unexpected char in internal subset (in %r)" % s
194*cda5da8dSAndroid Build Coastguard Worker                    )
195*cda5da8dSAndroid Build Coastguard Worker                if (j + 2) == n:
196*cda5da8dSAndroid Build Coastguard Worker                    # end of buffer; incomplete
197*cda5da8dSAndroid Build Coastguard Worker                    return -1
198*cda5da8dSAndroid Build Coastguard Worker                if (j + 4) > n:
199*cda5da8dSAndroid Build Coastguard Worker                    # end of buffer; incomplete
200*cda5da8dSAndroid Build Coastguard Worker                    return -1
201*cda5da8dSAndroid Build Coastguard Worker                if rawdata[j:j+4] == "<!--":
202*cda5da8dSAndroid Build Coastguard Worker                    j = self.parse_comment(j, report=0)
203*cda5da8dSAndroid Build Coastguard Worker                    if j < 0:
204*cda5da8dSAndroid Build Coastguard Worker                        return j
205*cda5da8dSAndroid Build Coastguard Worker                    continue
206*cda5da8dSAndroid Build Coastguard Worker                name, j = self._scan_name(j + 2, declstartpos)
207*cda5da8dSAndroid Build Coastguard Worker                if j == -1:
208*cda5da8dSAndroid Build Coastguard Worker                    return -1
209*cda5da8dSAndroid Build Coastguard Worker                if name not in {"attlist", "element", "entity", "notation"}:
210*cda5da8dSAndroid Build Coastguard Worker                    self.updatepos(declstartpos, j + 2)
211*cda5da8dSAndroid Build Coastguard Worker                    raise AssertionError(
212*cda5da8dSAndroid Build Coastguard Worker                        "unknown declaration %r in internal subset" % name
213*cda5da8dSAndroid Build Coastguard Worker                    )
214*cda5da8dSAndroid Build Coastguard Worker                # handle the individual names
215*cda5da8dSAndroid Build Coastguard Worker                meth = getattr(self, "_parse_doctype_" + name)
216*cda5da8dSAndroid Build Coastguard Worker                j = meth(j, declstartpos)
217*cda5da8dSAndroid Build Coastguard Worker                if j < 0:
218*cda5da8dSAndroid Build Coastguard Worker                    return j
219*cda5da8dSAndroid Build Coastguard Worker            elif c == "%":
220*cda5da8dSAndroid Build Coastguard Worker                # parameter entity reference
221*cda5da8dSAndroid Build Coastguard Worker                if (j + 1) == n:
222*cda5da8dSAndroid Build Coastguard Worker                    # end of buffer; incomplete
223*cda5da8dSAndroid Build Coastguard Worker                    return -1
224*cda5da8dSAndroid Build Coastguard Worker                s, j = self._scan_name(j + 1, declstartpos)
225*cda5da8dSAndroid Build Coastguard Worker                if j < 0:
226*cda5da8dSAndroid Build Coastguard Worker                    return j
227*cda5da8dSAndroid Build Coastguard Worker                if rawdata[j] == ";":
228*cda5da8dSAndroid Build Coastguard Worker                    j = j + 1
229*cda5da8dSAndroid Build Coastguard Worker            elif c == "]":
230*cda5da8dSAndroid Build Coastguard Worker                j = j + 1
231*cda5da8dSAndroid Build Coastguard Worker                while j < n and rawdata[j].isspace():
232*cda5da8dSAndroid Build Coastguard Worker                    j = j + 1
233*cda5da8dSAndroid Build Coastguard Worker                if j < n:
234*cda5da8dSAndroid Build Coastguard Worker                    if rawdata[j] == ">":
235*cda5da8dSAndroid Build Coastguard Worker                        return j
236*cda5da8dSAndroid Build Coastguard Worker                    self.updatepos(declstartpos, j)
237*cda5da8dSAndroid Build Coastguard Worker                    raise AssertionError("unexpected char after internal subset")
238*cda5da8dSAndroid Build Coastguard Worker                else:
239*cda5da8dSAndroid Build Coastguard Worker                    return -1
240*cda5da8dSAndroid Build Coastguard Worker            elif c.isspace():
241*cda5da8dSAndroid Build Coastguard Worker                j = j + 1
242*cda5da8dSAndroid Build Coastguard Worker            else:
243*cda5da8dSAndroid Build Coastguard Worker                self.updatepos(declstartpos, j)
244*cda5da8dSAndroid Build Coastguard Worker                raise AssertionError("unexpected char %r in internal subset" % c)
245*cda5da8dSAndroid Build Coastguard Worker        # end of buffer reached
246*cda5da8dSAndroid Build Coastguard Worker        return -1
247*cda5da8dSAndroid Build Coastguard Worker
248*cda5da8dSAndroid Build Coastguard Worker    # Internal -- scan past <!ELEMENT declarations
249*cda5da8dSAndroid Build Coastguard Worker    def _parse_doctype_element(self, i, declstartpos):
250*cda5da8dSAndroid Build Coastguard Worker        name, j = self._scan_name(i, declstartpos)
251*cda5da8dSAndroid Build Coastguard Worker        if j == -1:
252*cda5da8dSAndroid Build Coastguard Worker            return -1
253*cda5da8dSAndroid Build Coastguard Worker        # style content model; just skip until '>'
254*cda5da8dSAndroid Build Coastguard Worker        rawdata = self.rawdata
255*cda5da8dSAndroid Build Coastguard Worker        if '>' in rawdata[j:]:
256*cda5da8dSAndroid Build Coastguard Worker            return rawdata.find(">", j) + 1
257*cda5da8dSAndroid Build Coastguard Worker        return -1
258*cda5da8dSAndroid Build Coastguard Worker
259*cda5da8dSAndroid Build Coastguard Worker    # Internal -- scan past <!ATTLIST declarations
260*cda5da8dSAndroid Build Coastguard Worker    def _parse_doctype_attlist(self, i, declstartpos):
261*cda5da8dSAndroid Build Coastguard Worker        rawdata = self.rawdata
262*cda5da8dSAndroid Build Coastguard Worker        name, j = self._scan_name(i, declstartpos)
263*cda5da8dSAndroid Build Coastguard Worker        c = rawdata[j:j+1]
264*cda5da8dSAndroid Build Coastguard Worker        if c == "":
265*cda5da8dSAndroid Build Coastguard Worker            return -1
266*cda5da8dSAndroid Build Coastguard Worker        if c == ">":
267*cda5da8dSAndroid Build Coastguard Worker            return j + 1
268*cda5da8dSAndroid Build Coastguard Worker        while 1:
269*cda5da8dSAndroid Build Coastguard Worker            # scan a series of attribute descriptions; simplified:
270*cda5da8dSAndroid Build Coastguard Worker            #   name type [value] [#constraint]
271*cda5da8dSAndroid Build Coastguard Worker            name, j = self._scan_name(j, declstartpos)
272*cda5da8dSAndroid Build Coastguard Worker            if j < 0:
273*cda5da8dSAndroid Build Coastguard Worker                return j
274*cda5da8dSAndroid Build Coastguard Worker            c = rawdata[j:j+1]
275*cda5da8dSAndroid Build Coastguard Worker            if c == "":
276*cda5da8dSAndroid Build Coastguard Worker                return -1
277*cda5da8dSAndroid Build Coastguard Worker            if c == "(":
278*cda5da8dSAndroid Build Coastguard Worker                # an enumerated type; look for ')'
279*cda5da8dSAndroid Build Coastguard Worker                if ")" in rawdata[j:]:
280*cda5da8dSAndroid Build Coastguard Worker                    j = rawdata.find(")", j) + 1
281*cda5da8dSAndroid Build Coastguard Worker                else:
282*cda5da8dSAndroid Build Coastguard Worker                    return -1
283*cda5da8dSAndroid Build Coastguard Worker                while rawdata[j:j+1].isspace():
284*cda5da8dSAndroid Build Coastguard Worker                    j = j + 1
285*cda5da8dSAndroid Build Coastguard Worker                if not rawdata[j:]:
286*cda5da8dSAndroid Build Coastguard Worker                    # end of buffer, incomplete
287*cda5da8dSAndroid Build Coastguard Worker                    return -1
288*cda5da8dSAndroid Build Coastguard Worker            else:
289*cda5da8dSAndroid Build Coastguard Worker                name, j = self._scan_name(j, declstartpos)
290*cda5da8dSAndroid Build Coastguard Worker            c = rawdata[j:j+1]
291*cda5da8dSAndroid Build Coastguard Worker            if not c:
292*cda5da8dSAndroid Build Coastguard Worker                return -1
293*cda5da8dSAndroid Build Coastguard Worker            if c in "'\"":
294*cda5da8dSAndroid Build Coastguard Worker                m = _declstringlit_match(rawdata, j)
295*cda5da8dSAndroid Build Coastguard Worker                if m:
296*cda5da8dSAndroid Build Coastguard Worker                    j = m.end()
297*cda5da8dSAndroid Build Coastguard Worker                else:
298*cda5da8dSAndroid Build Coastguard Worker                    return -1
299*cda5da8dSAndroid Build Coastguard Worker                c = rawdata[j:j+1]
300*cda5da8dSAndroid Build Coastguard Worker                if not c:
301*cda5da8dSAndroid Build Coastguard Worker                    return -1
302*cda5da8dSAndroid Build Coastguard Worker            if c == "#":
303*cda5da8dSAndroid Build Coastguard Worker                if rawdata[j:] == "#":
304*cda5da8dSAndroid Build Coastguard Worker                    # end of buffer
305*cda5da8dSAndroid Build Coastguard Worker                    return -1
306*cda5da8dSAndroid Build Coastguard Worker                name, j = self._scan_name(j + 1, declstartpos)
307*cda5da8dSAndroid Build Coastguard Worker                if j < 0:
308*cda5da8dSAndroid Build Coastguard Worker                    return j
309*cda5da8dSAndroid Build Coastguard Worker                c = rawdata[j:j+1]
310*cda5da8dSAndroid Build Coastguard Worker                if not c:
311*cda5da8dSAndroid Build Coastguard Worker                    return -1
312*cda5da8dSAndroid Build Coastguard Worker            if c == '>':
313*cda5da8dSAndroid Build Coastguard Worker                # all done
314*cda5da8dSAndroid Build Coastguard Worker                return j + 1
315*cda5da8dSAndroid Build Coastguard Worker
316*cda5da8dSAndroid Build Coastguard Worker    # Internal -- scan past <!NOTATION declarations
317*cda5da8dSAndroid Build Coastguard Worker    def _parse_doctype_notation(self, i, declstartpos):
318*cda5da8dSAndroid Build Coastguard Worker        name, j = self._scan_name(i, declstartpos)
319*cda5da8dSAndroid Build Coastguard Worker        if j < 0:
320*cda5da8dSAndroid Build Coastguard Worker            return j
321*cda5da8dSAndroid Build Coastguard Worker        rawdata = self.rawdata
322*cda5da8dSAndroid Build Coastguard Worker        while 1:
323*cda5da8dSAndroid Build Coastguard Worker            c = rawdata[j:j+1]
324*cda5da8dSAndroid Build Coastguard Worker            if not c:
325*cda5da8dSAndroid Build Coastguard Worker                # end of buffer; incomplete
326*cda5da8dSAndroid Build Coastguard Worker                return -1
327*cda5da8dSAndroid Build Coastguard Worker            if c == '>':
328*cda5da8dSAndroid Build Coastguard Worker                return j + 1
329*cda5da8dSAndroid Build Coastguard Worker            if c in "'\"":
330*cda5da8dSAndroid Build Coastguard Worker                m = _declstringlit_match(rawdata, j)
331*cda5da8dSAndroid Build Coastguard Worker                if not m:
332*cda5da8dSAndroid Build Coastguard Worker                    return -1
333*cda5da8dSAndroid Build Coastguard Worker                j = m.end()
334*cda5da8dSAndroid Build Coastguard Worker            else:
335*cda5da8dSAndroid Build Coastguard Worker                name, j = self._scan_name(j, declstartpos)
336*cda5da8dSAndroid Build Coastguard Worker                if j < 0:
337*cda5da8dSAndroid Build Coastguard Worker                    return j
338*cda5da8dSAndroid Build Coastguard Worker
339*cda5da8dSAndroid Build Coastguard Worker    # Internal -- scan past <!ENTITY declarations
340*cda5da8dSAndroid Build Coastguard Worker    def _parse_doctype_entity(self, i, declstartpos):
341*cda5da8dSAndroid Build Coastguard Worker        rawdata = self.rawdata
342*cda5da8dSAndroid Build Coastguard Worker        if rawdata[i:i+1] == "%":
343*cda5da8dSAndroid Build Coastguard Worker            j = i + 1
344*cda5da8dSAndroid Build Coastguard Worker            while 1:
345*cda5da8dSAndroid Build Coastguard Worker                c = rawdata[j:j+1]
346*cda5da8dSAndroid Build Coastguard Worker                if not c:
347*cda5da8dSAndroid Build Coastguard Worker                    return -1
348*cda5da8dSAndroid Build Coastguard Worker                if c.isspace():
349*cda5da8dSAndroid Build Coastguard Worker                    j = j + 1
350*cda5da8dSAndroid Build Coastguard Worker                else:
351*cda5da8dSAndroid Build Coastguard Worker                    break
352*cda5da8dSAndroid Build Coastguard Worker        else:
353*cda5da8dSAndroid Build Coastguard Worker            j = i
354*cda5da8dSAndroid Build Coastguard Worker        name, j = self._scan_name(j, declstartpos)
355*cda5da8dSAndroid Build Coastguard Worker        if j < 0:
356*cda5da8dSAndroid Build Coastguard Worker            return j
357*cda5da8dSAndroid Build Coastguard Worker        while 1:
358*cda5da8dSAndroid Build Coastguard Worker            c = self.rawdata[j:j+1]
359*cda5da8dSAndroid Build Coastguard Worker            if not c:
360*cda5da8dSAndroid Build Coastguard Worker                return -1
361*cda5da8dSAndroid Build Coastguard Worker            if c in "'\"":
362*cda5da8dSAndroid Build Coastguard Worker                m = _declstringlit_match(rawdata, j)
363*cda5da8dSAndroid Build Coastguard Worker                if m:
364*cda5da8dSAndroid Build Coastguard Worker                    j = m.end()
365*cda5da8dSAndroid Build Coastguard Worker                else:
366*cda5da8dSAndroid Build Coastguard Worker                    return -1    # incomplete
367*cda5da8dSAndroid Build Coastguard Worker            elif c == ">":
368*cda5da8dSAndroid Build Coastguard Worker                return j + 1
369*cda5da8dSAndroid Build Coastguard Worker            else:
370*cda5da8dSAndroid Build Coastguard Worker                name, j = self._scan_name(j, declstartpos)
371*cda5da8dSAndroid Build Coastguard Worker                if j < 0:
372*cda5da8dSAndroid Build Coastguard Worker                    return j
373*cda5da8dSAndroid Build Coastguard Worker
374*cda5da8dSAndroid Build Coastguard Worker    # Internal -- scan a name token and the new position and the token, or
375*cda5da8dSAndroid Build Coastguard Worker    # return -1 if we've reached the end of the buffer.
376*cda5da8dSAndroid Build Coastguard Worker    def _scan_name(self, i, declstartpos):
377*cda5da8dSAndroid Build Coastguard Worker        rawdata = self.rawdata
378*cda5da8dSAndroid Build Coastguard Worker        n = len(rawdata)
379*cda5da8dSAndroid Build Coastguard Worker        if i == n:
380*cda5da8dSAndroid Build Coastguard Worker            return None, -1
381*cda5da8dSAndroid Build Coastguard Worker        m = _declname_match(rawdata, i)
382*cda5da8dSAndroid Build Coastguard Worker        if m:
383*cda5da8dSAndroid Build Coastguard Worker            s = m.group()
384*cda5da8dSAndroid Build Coastguard Worker            name = s.strip()
385*cda5da8dSAndroid Build Coastguard Worker            if (i + len(s)) == n:
386*cda5da8dSAndroid Build Coastguard Worker                return None, -1  # end of buffer
387*cda5da8dSAndroid Build Coastguard Worker            return name.lower(), m.end()
388*cda5da8dSAndroid Build Coastguard Worker        else:
389*cda5da8dSAndroid Build Coastguard Worker            self.updatepos(declstartpos, i)
390*cda5da8dSAndroid Build Coastguard Worker            raise AssertionError(
391*cda5da8dSAndroid Build Coastguard Worker                "expected name token at %r" % rawdata[declstartpos:declstartpos+20]
392*cda5da8dSAndroid Build Coastguard Worker            )
393*cda5da8dSAndroid Build Coastguard Worker
394*cda5da8dSAndroid Build Coastguard Worker    # To be overridden -- handlers for unknown objects
395*cda5da8dSAndroid Build Coastguard Worker    def unknown_decl(self, data):
396*cda5da8dSAndroid Build Coastguard Worker        pass
397