Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1 | #!/usr/bin/python -u |
| 2 | # |
| 3 | # This is the API builder, it parses the C sources and build the |
| 4 | # API formal description in XML. |
| 5 | # |
| 6 | # See Copyright for the status of this software. |
| 7 | # |
| 8 | # daniel@veillard.com |
| 9 | # |
Daniel Veillard | 540a31a | 2003-01-21 11:21:07 +0000 | [diff] [blame] | 10 | import os, sys |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 11 | import string |
| 12 | import glob |
| 13 | |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 14 | debug=0 |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 15 | #debugsym='ignorableWhitespaceSAXFunc' |
| 16 | debugsym=None |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 17 | |
Daniel Veillard | de0a0a5 | 2003-04-24 17:12:57 +0000 | [diff] [blame] | 18 | # |
| 19 | # C parser analysis code |
| 20 | # |
| 21 | ignored_files = { |
| 22 | "trio": "too many non standard macros", |
| 23 | "trio.c": "too many non standard macros", |
| 24 | "trionan.c": "too many non standard macros", |
| 25 | "triostr.c": "too many non standard macros", |
| 26 | "acconfig.h": "generated portability layer", |
| 27 | "config.h": "generated portability layer", |
| 28 | "libxml.h": "internal only", |
Daniel Veillard | 92fc02c | 2003-04-24 23:12:35 +0000 | [diff] [blame] | 29 | "testOOM.c": "out of memory tester", |
| 30 | "testOOMlib.h": "out of memory tester", |
| 31 | "testOOMlib.c": "out of memory tester", |
Daniel Veillard | 189f46b | 2004-01-25 21:03:04 +0000 | [diff] [blame] | 32 | "rngparser.c": "not yet integrated", |
| 33 | "rngparser.h": "not yet integrated", |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 34 | "elfgcchack.h": "not a normal header", |
Daniel Veillard | 9517501 | 2005-07-03 16:09:51 +0000 | [diff] [blame] | 35 | "testHTML.c": "test tool", |
| 36 | "testReader.c": "test tool", |
| 37 | "testSchemas.c": "test tool", |
| 38 | "testXPath.c": "test tool", |
| 39 | "testAutomata.c": "test tool", |
| 40 | "testModule.c": "test tool", |
| 41 | "testRegexp.c": "test tool", |
| 42 | "testThreads.c": "test tool", |
| 43 | "testC14N.c": "test tool", |
| 44 | "testRelax.c": "test tool", |
| 45 | "testThreadsWin32.c": "test tool", |
| 46 | "testSAX.c": "test tool", |
| 47 | "testURI.c": "test tool", |
Daniel Veillard | e43cc57 | 2004-11-03 11:50:29 +0000 | [diff] [blame] | 48 | "testapi.c": "generated regression tests", |
Daniel Veillard | 9517501 | 2005-07-03 16:09:51 +0000 | [diff] [blame] | 49 | "runtest.c": "regression tests program", |
| 50 | "runsuite.c": "regression tests program", |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 51 | "tst.c": "not part of the library", |
Daniel Veillard | 6ad5c4a | 2006-10-11 16:43:06 +0000 | [diff] [blame] | 52 | "test.c": "not part of the library", |
Daniel Veillard | f6b71bd | 2005-01-04 17:50:14 +0000 | [diff] [blame] | 53 | "testdso.c": "test for dynamid shared libraries", |
Daniel Veillard | de0a0a5 | 2003-04-24 17:12:57 +0000 | [diff] [blame] | 54 | } |
| 55 | |
| 56 | ignored_words = { |
| 57 | "WINAPI": (0, "Windows keyword"), |
| 58 | "LIBXML_DLL_IMPORT": (0, "Special macro to flag external keywords"), |
William M. Brack | cdfa286 | 2003-08-29 06:03:38 +0000 | [diff] [blame] | 59 | "XMLPUBVAR": (0, "Special macro for extern vars for win32"), |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 60 | "XSLTPUBVAR": (0, "Special macro for extern vars for win32"), |
| 61 | "EXSLTPUBVAR": (0, "Special macro for extern vars for win32"), |
William M. Brack | cdfa286 | 2003-08-29 06:03:38 +0000 | [diff] [blame] | 62 | "XMLPUBFUN": (0, "Special macro for extern funcs for win32"), |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 63 | "XSLTPUBFUN": (0, "Special macro for extern funcs for win32"), |
| 64 | "EXSLTPUBFUN": (0, "Special macro for extern funcs for win32"), |
William M. Brack | cdfa286 | 2003-08-29 06:03:38 +0000 | [diff] [blame] | 65 | "XMLCALL": (0, "Special macro for win32 calls"), |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 66 | "XSLTCALL": (0, "Special macro for win32 calls"), |
Daniel Veillard | 5eee767 | 2005-08-22 21:22:27 +0000 | [diff] [blame] | 67 | "XMLCDECL": (0, "Special macro for win32 calls"), |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 68 | "EXSLTCALL": (0, "Special macro for win32 calls"), |
Daniel Veillard | de0a0a5 | 2003-04-24 17:12:57 +0000 | [diff] [blame] | 69 | "__declspec": (3, "Windows keyword"), |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 70 | "__stdcall": (0, "Windows keyword"), |
Daniel Veillard | de0a0a5 | 2003-04-24 17:12:57 +0000 | [diff] [blame] | 71 | "ATTRIBUTE_UNUSED": (0, "macro keyword"), |
| 72 | "LIBEXSLT_PUBLIC": (0, "macro keyword"), |
| 73 | "X_IN_Y": (5, "macro function builder"), |
| 74 | } |
| 75 | |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 76 | def escape(raw): |
| 77 | raw = string.replace(raw, '&', '&') |
| 78 | raw = string.replace(raw, '<', '<') |
| 79 | raw = string.replace(raw, '>', '>') |
| 80 | raw = string.replace(raw, "'", ''') |
| 81 | raw = string.replace(raw, '"', '"') |
| 82 | return raw |
| 83 | |
Daniel Veillard | 2925c0a | 2003-11-17 13:58:17 +0000 | [diff] [blame] | 84 | def uniq(items): |
| 85 | d = {} |
| 86 | for item in items: |
| 87 | d[item]=1 |
| 88 | return d.keys() |
| 89 | |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 90 | class identifier: |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 91 | def __init__(self, name, header=None, module=None, type=None, lineno = 0, |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 92 | info=None, extra=None, conditionals = None): |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 93 | self.name = name |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 94 | self.header = header |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 95 | self.module = module |
| 96 | self.type = type |
| 97 | self.info = info |
| 98 | self.extra = extra |
| 99 | self.lineno = lineno |
| 100 | self.static = 0 |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 101 | if conditionals == None or len(conditionals) == 0: |
| 102 | self.conditionals = None |
| 103 | else: |
| 104 | self.conditionals = conditionals[:] |
| 105 | if self.name == debugsym: |
| 106 | print "=> define %s : %s" % (debugsym, (module, type, info, |
| 107 | extra, conditionals)) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 108 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 109 | def __repr__(self): |
| 110 | r = "%s %s:" % (self.type, self.name) |
| 111 | if self.static: |
| 112 | r = r + " static" |
| 113 | if self.module != None: |
| 114 | r = r + " from %s" % (self.module) |
| 115 | if self.info != None: |
| 116 | r = r + " " + `self.info` |
| 117 | if self.extra != None: |
| 118 | r = r + " " + `self.extra` |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 119 | if self.conditionals != None: |
| 120 | r = r + " " + `self.conditionals` |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 121 | return r |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 122 | |
| 123 | |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 124 | def set_header(self, header): |
| 125 | self.header = header |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 126 | def set_module(self, module): |
| 127 | self.module = module |
| 128 | def set_type(self, type): |
| 129 | self.type = type |
| 130 | def set_info(self, info): |
| 131 | self.info = info |
| 132 | def set_extra(self, extra): |
| 133 | self.extra = extra |
| 134 | def set_lineno(self, lineno): |
| 135 | self.lineno = lineno |
| 136 | def set_static(self, static): |
| 137 | self.static = static |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 138 | def set_conditionals(self, conditionals): |
| 139 | if conditionals == None or len(conditionals) == 0: |
| 140 | self.conditionals = None |
| 141 | else: |
| 142 | self.conditionals = conditionals[:] |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 143 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 144 | def get_name(self): |
| 145 | return self.name |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 146 | def get_header(self): |
| 147 | return self.module |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 148 | def get_module(self): |
| 149 | return self.module |
| 150 | def get_type(self): |
| 151 | return self.type |
| 152 | def get_info(self): |
| 153 | return self.info |
| 154 | def get_lineno(self): |
| 155 | return self.lineno |
| 156 | def get_extra(self): |
| 157 | return self.extra |
| 158 | def get_static(self): |
| 159 | return self.static |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 160 | def get_conditionals(self): |
| 161 | return self.conditionals |
Daniel Veillard | d8cf906 | 2003-11-11 21:12:36 +0000 | [diff] [blame] | 162 | |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 163 | def update(self, header, module, type = None, info = None, extra=None, |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 164 | conditionals=None): |
| 165 | if self.name == debugsym: |
| 166 | print "=> update %s : %s" % (debugsym, (module, type, info, |
| 167 | extra, conditionals)) |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 168 | if header != None and self.header == None: |
| 169 | self.set_header(module) |
| 170 | if module != None and (self.module == None or self.header == self.module): |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 171 | self.set_module(module) |
| 172 | if type != None and self.type == None: |
| 173 | self.set_type(type) |
| 174 | if info != None: |
| 175 | self.set_info(info) |
| 176 | if extra != None: |
| 177 | self.set_extra(extra) |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 178 | if conditionals != None: |
| 179 | self.set_conditionals(conditionals) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 180 | |
| 181 | class index: |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 182 | def __init__(self, name = "noname"): |
| 183 | self.name = name |
| 184 | self.identifiers = {} |
| 185 | self.functions = {} |
| 186 | self.variables = {} |
| 187 | self.includes = {} |
| 188 | self.structs = {} |
| 189 | self.enums = {} |
| 190 | self.typedefs = {} |
| 191 | self.macros = {} |
| 192 | self.references = {} |
| 193 | self.info = {} |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 194 | |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 195 | def add_ref(self, name, header, module, static, type, lineno, info=None, extra=None, conditionals = None): |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 196 | if name[0:2] == '__': |
| 197 | return None |
| 198 | d = None |
| 199 | try: |
| 200 | d = self.identifiers[name] |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 201 | d.update(header, module, type, lineno, info, extra, conditionals) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 202 | except: |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 203 | d = identifier(name, header, module, type, lineno, info, extra, conditionals) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 204 | self.identifiers[name] = d |
Daniel Veillard | d8cf906 | 2003-11-11 21:12:36 +0000 | [diff] [blame] | 205 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 206 | if d != None and static == 1: |
| 207 | d.set_static(1) |
Daniel Veillard | d8cf906 | 2003-11-11 21:12:36 +0000 | [diff] [blame] | 208 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 209 | if d != None and name != None and type != None: |
| 210 | self.references[name] = d |
Daniel Veillard | d8cf906 | 2003-11-11 21:12:36 +0000 | [diff] [blame] | 211 | |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 212 | if name == debugsym: |
| 213 | print "New ref: %s" % (d) |
| 214 | |
| 215 | return d |
| 216 | |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 217 | def add(self, name, header, module, static, type, lineno, info=None, extra=None, conditionals = None): |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 218 | if name[0:2] == '__': |
| 219 | return None |
| 220 | d = None |
| 221 | try: |
| 222 | d = self.identifiers[name] |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 223 | d.update(header, module, type, lineno, info, extra, conditionals) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 224 | except: |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 225 | d = identifier(name, header, module, type, lineno, info, extra, conditionals) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 226 | self.identifiers[name] = d |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 227 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 228 | if d != None and static == 1: |
| 229 | d.set_static(1) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 230 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 231 | if d != None and name != None and type != None: |
| 232 | if type == "function": |
| 233 | self.functions[name] = d |
| 234 | elif type == "functype": |
| 235 | self.functions[name] = d |
| 236 | elif type == "variable": |
| 237 | self.variables[name] = d |
| 238 | elif type == "include": |
| 239 | self.includes[name] = d |
| 240 | elif type == "struct": |
| 241 | self.structs[name] = d |
| 242 | elif type == "enum": |
| 243 | self.enums[name] = d |
| 244 | elif type == "typedef": |
| 245 | self.typedefs[name] = d |
| 246 | elif type == "macro": |
| 247 | self.macros[name] = d |
| 248 | else: |
| 249 | print "Unable to register type ", type |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 250 | |
| 251 | if name == debugsym: |
| 252 | print "New symbol: %s" % (d) |
| 253 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 254 | return d |
| 255 | |
| 256 | def merge(self, idx): |
| 257 | for id in idx.functions.keys(): |
| 258 | # |
| 259 | # macro might be used to override functions or variables |
| 260 | # definitions |
| 261 | # |
| 262 | if self.macros.has_key(id): |
| 263 | del self.macros[id] |
| 264 | if self.functions.has_key(id): |
| 265 | print "function %s from %s redeclared in %s" % ( |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 266 | id, self.functions[id].header, idx.functions[id].header) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 267 | else: |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 268 | self.functions[id] = idx.functions[id] |
| 269 | self.identifiers[id] = idx.functions[id] |
| 270 | for id in idx.variables.keys(): |
| 271 | # |
| 272 | # macro might be used to override functions or variables |
| 273 | # definitions |
| 274 | # |
| 275 | if self.macros.has_key(id): |
| 276 | del self.macros[id] |
| 277 | if self.variables.has_key(id): |
| 278 | print "variable %s from %s redeclared in %s" % ( |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 279 | id, self.variables[id].header, idx.variables[id].header) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 280 | else: |
| 281 | self.variables[id] = idx.variables[id] |
| 282 | self.identifiers[id] = idx.variables[id] |
| 283 | for id in idx.structs.keys(): |
| 284 | if self.structs.has_key(id): |
| 285 | print "struct %s from %s redeclared in %s" % ( |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 286 | id, self.structs[id].header, idx.structs[id].header) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 287 | else: |
| 288 | self.structs[id] = idx.structs[id] |
| 289 | self.identifiers[id] = idx.structs[id] |
| 290 | for id in idx.typedefs.keys(): |
| 291 | if self.typedefs.has_key(id): |
| 292 | print "typedef %s from %s redeclared in %s" % ( |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 293 | id, self.typedefs[id].header, idx.typedefs[id].header) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 294 | else: |
| 295 | self.typedefs[id] = idx.typedefs[id] |
| 296 | self.identifiers[id] = idx.typedefs[id] |
| 297 | for id in idx.macros.keys(): |
| 298 | # |
| 299 | # macro might be used to override functions or variables |
| 300 | # definitions |
| 301 | # |
| 302 | if self.variables.has_key(id): |
| 303 | continue |
| 304 | if self.functions.has_key(id): |
| 305 | continue |
| 306 | if self.enums.has_key(id): |
| 307 | continue |
| 308 | if self.macros.has_key(id): |
| 309 | print "macro %s from %s redeclared in %s" % ( |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 310 | id, self.macros[id].header, idx.macros[id].header) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 311 | else: |
| 312 | self.macros[id] = idx.macros[id] |
| 313 | self.identifiers[id] = idx.macros[id] |
| 314 | for id in idx.enums.keys(): |
| 315 | if self.enums.has_key(id): |
| 316 | print "enum %s from %s redeclared in %s" % ( |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 317 | id, self.enums[id].header, idx.enums[id].header) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 318 | else: |
| 319 | self.enums[id] = idx.enums[id] |
| 320 | self.identifiers[id] = idx.enums[id] |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 321 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 322 | def merge_public(self, idx): |
| 323 | for id in idx.functions.keys(): |
| 324 | if self.functions.has_key(id): |
William M. Brack | 21e4ef2 | 2005-01-02 09:53:13 +0000 | [diff] [blame] | 325 | # check that function condition agrees with header |
| 326 | if idx.functions[id].conditionals != \ |
| 327 | self.functions[id].conditionals: |
| 328 | print "Header condition differs from Function for %s:" \ |
| 329 | % id |
| 330 | print " H: %s" % self.functions[id].conditionals |
| 331 | print " C: %s" % idx.functions[id].conditionals |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 332 | up = idx.functions[id] |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 333 | self.functions[id].update(None, up.module, up.type, up.info, up.extra) |
Daniel Veillard | c1eed32 | 2002-12-12 11:01:32 +0000 | [diff] [blame] | 334 | # else: |
| 335 | # print "Function %s from %s is not declared in headers" % ( |
William M. Brack | 21e4ef2 | 2005-01-02 09:53:13 +0000 | [diff] [blame] | 336 | # id, idx.functions[id].module) |
Daniel Veillard | c1eed32 | 2002-12-12 11:01:32 +0000 | [diff] [blame] | 337 | # TODO: do the same for variables. |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 338 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 339 | def analyze_dict(self, type, dict): |
| 340 | count = 0 |
| 341 | public = 0 |
| 342 | for name in dict.keys(): |
| 343 | id = dict[name] |
| 344 | count = count + 1 |
| 345 | if id.static == 0: |
| 346 | public = public + 1 |
| 347 | if count != public: |
| 348 | print " %d %s , %d public" % (count, type, public) |
| 349 | elif count != 0: |
| 350 | print " %d public %s" % (count, type) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 351 | |
| 352 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 353 | def analyze(self): |
| 354 | self.analyze_dict("functions", self.functions) |
| 355 | self.analyze_dict("variables", self.variables) |
| 356 | self.analyze_dict("structs", self.structs) |
| 357 | self.analyze_dict("typedefs", self.typedefs) |
| 358 | self.analyze_dict("macros", self.macros) |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 359 | |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 360 | class CLexer: |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 361 | """A lexer for the C language, tokenize the input by reading and |
| 362 | analyzing it line by line""" |
| 363 | def __init__(self, input): |
| 364 | self.input = input |
| 365 | self.tokens = [] |
| 366 | self.line = "" |
| 367 | self.lineno = 0 |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 368 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 369 | def getline(self): |
| 370 | line = '' |
| 371 | while line == '': |
| 372 | line = self.input.readline() |
| 373 | if not line: |
| 374 | return None |
| 375 | self.lineno = self.lineno + 1 |
| 376 | line = string.lstrip(line) |
| 377 | line = string.rstrip(line) |
| 378 | if line == '': |
| 379 | continue |
| 380 | while line[-1] == '\\': |
| 381 | line = line[:-1] |
| 382 | n = self.input.readline() |
| 383 | self.lineno = self.lineno + 1 |
| 384 | n = string.lstrip(n) |
| 385 | n = string.rstrip(n) |
| 386 | if not n: |
| 387 | break |
| 388 | else: |
| 389 | line = line + n |
| 390 | return line |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 391 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 392 | def getlineno(self): |
| 393 | return self.lineno |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 394 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 395 | def push(self, token): |
| 396 | self.tokens.insert(0, token); |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 397 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 398 | def debug(self): |
| 399 | print "Last token: ", self.last |
| 400 | print "Token queue: ", self.tokens |
| 401 | print "Line %d end: " % (self.lineno), self.line |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 402 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 403 | def token(self): |
| 404 | while self.tokens == []: |
| 405 | if self.line == "": |
| 406 | line = self.getline() |
| 407 | else: |
| 408 | line = self.line |
| 409 | self.line = "" |
| 410 | if line == None: |
| 411 | return None |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 412 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 413 | if line[0] == '#': |
| 414 | self.tokens = map((lambda x: ('preproc', x)), |
| 415 | string.split(line)) |
| 416 | break; |
| 417 | l = len(line) |
| 418 | if line[0] == '"' or line[0] == "'": |
| 419 | end = line[0] |
| 420 | line = line[1:] |
| 421 | found = 0 |
| 422 | tok = "" |
| 423 | while found == 0: |
| 424 | i = 0 |
| 425 | l = len(line) |
| 426 | while i < l: |
| 427 | if line[i] == end: |
| 428 | self.line = line[i+1:] |
| 429 | line = line[:i] |
| 430 | l = i |
| 431 | found = 1 |
| 432 | break |
| 433 | if line[i] == '\\': |
| 434 | i = i + 1 |
| 435 | i = i + 1 |
| 436 | tok = tok + line |
| 437 | if found == 0: |
| 438 | line = self.getline() |
| 439 | if line == None: |
| 440 | return None |
| 441 | self.last = ('string', tok) |
| 442 | return self.last |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 443 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 444 | if l >= 2 and line[0] == '/' and line[1] == '*': |
| 445 | line = line[2:] |
| 446 | found = 0 |
| 447 | tok = "" |
| 448 | while found == 0: |
| 449 | i = 0 |
| 450 | l = len(line) |
| 451 | while i < l: |
| 452 | if line[i] == '*' and i+1 < l and line[i+1] == '/': |
| 453 | self.line = line[i+2:] |
| 454 | line = line[:i-1] |
| 455 | l = i |
| 456 | found = 1 |
| 457 | break |
| 458 | i = i + 1 |
| 459 | if tok != "": |
| 460 | tok = tok + "\n" |
| 461 | tok = tok + line |
| 462 | if found == 0: |
| 463 | line = self.getline() |
| 464 | if line == None: |
| 465 | return None |
| 466 | self.last = ('comment', tok) |
| 467 | return self.last |
| 468 | if l >= 2 and line[0] == '/' and line[1] == '/': |
| 469 | line = line[2:] |
| 470 | self.last = ('comment', line) |
| 471 | return self.last |
| 472 | i = 0 |
| 473 | while i < l: |
| 474 | if line[i] == '/' and i+1 < l and line[i+1] == '/': |
| 475 | self.line = line[i:] |
| 476 | line = line[:i] |
| 477 | break |
| 478 | if line[i] == '/' and i+1 < l and line[i+1] == '*': |
| 479 | self.line = line[i:] |
| 480 | line = line[:i] |
| 481 | break |
| 482 | if line[i] == '"' or line[i] == "'": |
| 483 | self.line = line[i:] |
| 484 | line = line[:i] |
| 485 | break |
| 486 | i = i + 1 |
| 487 | l = len(line) |
| 488 | i = 0 |
| 489 | while i < l: |
| 490 | if line[i] == ' ' or line[i] == '\t': |
| 491 | i = i + 1 |
| 492 | continue |
| 493 | o = ord(line[i]) |
| 494 | if (o >= 97 and o <= 122) or (o >= 65 and o <= 90) or \ |
| 495 | (o >= 48 and o <= 57): |
| 496 | s = i |
| 497 | while i < l: |
| 498 | o = ord(line[i]) |
| 499 | if (o >= 97 and o <= 122) or (o >= 65 and o <= 90) or \ |
| 500 | (o >= 48 and o <= 57) or string.find( |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 501 | " \t(){}:;,+-*/%&!|[]=><", line[i]) == -1: |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 502 | i = i + 1 |
| 503 | else: |
| 504 | break |
| 505 | self.tokens.append(('name', line[s:i])) |
| 506 | continue |
| 507 | if string.find("(){}:;,[]", line[i]) != -1: |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 508 | # if line[i] == '(' or line[i] == ')' or line[i] == '{' or \ |
| 509 | # line[i] == '}' or line[i] == ':' or line[i] == ';' or \ |
| 510 | # line[i] == ',' or line[i] == '[' or line[i] == ']': |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 511 | self.tokens.append(('sep', line[i])) |
| 512 | i = i + 1 |
| 513 | continue |
| 514 | if string.find("+-*><=/%&!|.", line[i]) != -1: |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 515 | # if line[i] == '+' or line[i] == '-' or line[i] == '*' or \ |
| 516 | # line[i] == '>' or line[i] == '<' or line[i] == '=' or \ |
| 517 | # line[i] == '/' or line[i] == '%' or line[i] == '&' or \ |
| 518 | # line[i] == '!' or line[i] == '|' or line[i] == '.': |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 519 | if line[i] == '.' and i + 2 < l and \ |
| 520 | line[i+1] == '.' and line[i+2] == '.': |
| 521 | self.tokens.append(('name', '...')) |
| 522 | i = i + 3 |
| 523 | continue |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 524 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 525 | j = i + 1 |
| 526 | if j < l and ( |
| 527 | string.find("+-*><=/%&!|", line[j]) != -1): |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 528 | # line[j] == '+' or line[j] == '-' or line[j] == '*' or \ |
| 529 | # line[j] == '>' or line[j] == '<' or line[j] == '=' or \ |
| 530 | # line[j] == '/' or line[j] == '%' or line[j] == '&' or \ |
| 531 | # line[j] == '!' or line[j] == '|'): |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 532 | self.tokens.append(('op', line[i:j+1])) |
| 533 | i = j + 1 |
| 534 | else: |
| 535 | self.tokens.append(('op', line[i])) |
| 536 | i = i + 1 |
| 537 | continue |
| 538 | s = i |
| 539 | while i < l: |
| 540 | o = ord(line[i]) |
| 541 | if (o >= 97 and o <= 122) or (o >= 65 and o <= 90) or \ |
| 542 | (o >= 48 and o <= 57) or ( |
| 543 | string.find(" \t(){}:;,+-*/%&!|[]=><", line[i]) == -1): |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 544 | # line[i] != ' ' and line[i] != '\t' and |
| 545 | # line[i] != '(' and line[i] != ')' and |
| 546 | # line[i] != '{' and line[i] != '}' and |
| 547 | # line[i] != ':' and line[i] != ';' and |
| 548 | # line[i] != ',' and line[i] != '+' and |
| 549 | # line[i] != '-' and line[i] != '*' and |
| 550 | # line[i] != '/' and line[i] != '%' and |
| 551 | # line[i] != '&' and line[i] != '!' and |
| 552 | # line[i] != '|' and line[i] != '[' and |
| 553 | # line[i] != ']' and line[i] != '=' and |
| 554 | # line[i] != '*' and line[i] != '>' and |
| 555 | # line[i] != '<'): |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 556 | i = i + 1 |
| 557 | else: |
| 558 | break |
| 559 | self.tokens.append(('name', line[s:i])) |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 560 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 561 | tok = self.tokens[0] |
| 562 | self.tokens = self.tokens[1:] |
| 563 | self.last = tok |
| 564 | return tok |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 565 | |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 566 | class CParser: |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 567 | """The C module parser""" |
| 568 | def __init__(self, filename, idx = None): |
| 569 | self.filename = filename |
| 570 | if len(filename) > 2 and filename[-2:] == '.h': |
| 571 | self.is_header = 1 |
| 572 | else: |
| 573 | self.is_header = 0 |
| 574 | self.input = open(filename) |
| 575 | self.lexer = CLexer(self.input) |
| 576 | if idx == None: |
| 577 | self.index = index() |
| 578 | else: |
| 579 | self.index = idx |
| 580 | self.top_comment = "" |
| 581 | self.last_comment = "" |
| 582 | self.comment = None |
| 583 | self.collect_ref = 0 |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 584 | self.no_error = 0 |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 585 | self.conditionals = [] |
| 586 | self.defines = [] |
Daniel Veillard | d8cf906 | 2003-11-11 21:12:36 +0000 | [diff] [blame] | 587 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 588 | def collect_references(self): |
| 589 | self.collect_ref = 1 |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 590 | |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 591 | def stop_error(self): |
| 592 | self.no_error = 1 |
| 593 | |
| 594 | def start_error(self): |
| 595 | self.no_error = 0 |
| 596 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 597 | def lineno(self): |
| 598 | return self.lexer.getlineno() |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 599 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 600 | def index_add(self, name, module, static, type, info=None, extra = None): |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 601 | if self.is_header == 1: |
| 602 | self.index.add(name, module, module, static, type, self.lineno(), |
| 603 | info, extra, self.conditionals) |
| 604 | else: |
| 605 | self.index.add(name, None, module, static, type, self.lineno(), |
| 606 | info, extra, self.conditionals) |
Daniel Veillard | d8cf906 | 2003-11-11 21:12:36 +0000 | [diff] [blame] | 607 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 608 | def index_add_ref(self, name, module, static, type, info=None, |
| 609 | extra = None): |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 610 | if self.is_header == 1: |
| 611 | self.index.add_ref(name, module, module, static, type, |
| 612 | self.lineno(), info, extra, self.conditionals) |
| 613 | else: |
| 614 | self.index.add_ref(name, None, module, static, type, self.lineno(), |
| 615 | info, extra, self.conditionals) |
Daniel Veillard | d8cf906 | 2003-11-11 21:12:36 +0000 | [diff] [blame] | 616 | |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 617 | def warning(self, msg): |
| 618 | if self.no_error: |
| 619 | return |
| 620 | print msg |
| 621 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 622 | def error(self, msg, token=-1): |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 623 | if self.no_error: |
| 624 | return |
| 625 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 626 | print "Parse Error: " + msg |
| 627 | if token != -1: |
| 628 | print "Got token ", token |
| 629 | self.lexer.debug() |
| 630 | sys.exit(1) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 631 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 632 | def debug(self, msg, token=-1): |
| 633 | print "Debug: " + msg |
| 634 | if token != -1: |
| 635 | print "Got token ", token |
| 636 | self.lexer.debug() |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 637 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 638 | def parseTopComment(self, comment): |
| 639 | res = {} |
| 640 | lines = string.split(comment, "\n") |
| 641 | item = None |
| 642 | for line in lines: |
| 643 | while line != "" and (line[0] == ' ' or line[0] == '\t'): |
| 644 | line = line[1:] |
| 645 | while line != "" and line[0] == '*': |
| 646 | line = line[1:] |
| 647 | while line != "" and (line[0] == ' ' or line[0] == '\t'): |
| 648 | line = line[1:] |
| 649 | try: |
| 650 | (it, line) = string.split(line, ":", 1) |
| 651 | item = it |
| 652 | while line != "" and (line[0] == ' ' or line[0] == '\t'): |
| 653 | line = line[1:] |
| 654 | if res.has_key(item): |
| 655 | res[item] = res[item] + " " + line |
| 656 | else: |
| 657 | res[item] = line |
| 658 | except: |
| 659 | if item != None: |
| 660 | if res.has_key(item): |
| 661 | res[item] = res[item] + " " + line |
| 662 | else: |
| 663 | res[item] = line |
| 664 | self.index.info = res |
| 665 | |
| 666 | def parseComment(self, token): |
| 667 | if self.top_comment == "": |
| 668 | self.top_comment = token[1] |
| 669 | if self.comment == None or token[1][0] == '*': |
| 670 | self.comment = token[1]; |
| 671 | else: |
| 672 | self.comment = self.comment + token[1] |
| 673 | token = self.lexer.token() |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 674 | |
| 675 | if string.find(self.comment, "DOC_DISABLE") != -1: |
| 676 | self.stop_error() |
| 677 | |
| 678 | if string.find(self.comment, "DOC_ENABLE") != -1: |
| 679 | self.start_error() |
| 680 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 681 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 682 | |
Daniel Veillard | 99b7850 | 2005-09-11 22:04:08 +0000 | [diff] [blame] | 683 | # |
| 684 | # Parse a comment block associate to a typedef |
| 685 | # |
| 686 | def parseTypeComment(self, name, quiet = 0): |
| 687 | if name[0:2] == '__': |
| 688 | quiet = 1 |
| 689 | |
| 690 | args = [] |
| 691 | desc = "" |
| 692 | |
| 693 | if self.comment == None: |
| 694 | if not quiet: |
| 695 | self.warning("Missing comment for type %s" % (name)) |
| 696 | return((args, desc)) |
| 697 | if self.comment[0] != '*': |
| 698 | if not quiet: |
| 699 | self.warning("Missing * in type comment for %s" % (name)) |
| 700 | return((args, desc)) |
| 701 | lines = string.split(self.comment, '\n') |
| 702 | if lines[0] == '*': |
| 703 | del lines[0] |
| 704 | if lines[0] != "* %s:" % (name): |
| 705 | if not quiet: |
| 706 | self.warning("Misformatted type comment for %s" % (name)) |
| 707 | self.warning(" Expecting '* %s:' got '%s'" % (name, lines[0])) |
| 708 | return((args, desc)) |
| 709 | del lines[0] |
| 710 | while len(lines) > 0 and lines[0] == '*': |
| 711 | del lines[0] |
| 712 | desc = "" |
| 713 | while len(lines) > 0: |
| 714 | l = lines[0] |
| 715 | while len(l) > 0 and l[0] == '*': |
| 716 | l = l[1:] |
| 717 | l = string.strip(l) |
| 718 | desc = desc + " " + l |
| 719 | del lines[0] |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 720 | |
Daniel Veillard | 99b7850 | 2005-09-11 22:04:08 +0000 | [diff] [blame] | 721 | desc = string.strip(desc) |
| 722 | |
| 723 | if quiet == 0: |
| 724 | if desc == "": |
| 725 | self.warning("Type comment for %s lack description of the macro" % (name)) |
| 726 | |
| 727 | return(desc) |
| 728 | # |
| 729 | # Parse a comment block associate to a macro |
| 730 | # |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 731 | def parseMacroComment(self, name, quiet = 0): |
| 732 | if name[0:2] == '__': |
| 733 | quiet = 1 |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 734 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 735 | args = [] |
| 736 | desc = "" |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 737 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 738 | if self.comment == None: |
| 739 | if not quiet: |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 740 | self.warning("Missing comment for macro %s" % (name)) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 741 | return((args, desc)) |
| 742 | if self.comment[0] != '*': |
| 743 | if not quiet: |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 744 | self.warning("Missing * in macro comment for %s" % (name)) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 745 | return((args, desc)) |
| 746 | lines = string.split(self.comment, '\n') |
| 747 | if lines[0] == '*': |
| 748 | del lines[0] |
| 749 | if lines[0] != "* %s:" % (name): |
| 750 | if not quiet: |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 751 | self.warning("Misformatted macro comment for %s" % (name)) |
| 752 | self.warning(" Expecting '* %s:' got '%s'" % (name, lines[0])) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 753 | return((args, desc)) |
| 754 | del lines[0] |
| 755 | while lines[0] == '*': |
| 756 | del lines[0] |
| 757 | while len(lines) > 0 and lines[0][0:3] == '* @': |
| 758 | l = lines[0][3:] |
| 759 | try: |
| 760 | (arg, desc) = string.split(l, ':', 1) |
| 761 | desc=string.strip(desc) |
| 762 | arg=string.strip(arg) |
| 763 | except: |
| 764 | if not quiet: |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 765 | self.warning("Misformatted macro comment for %s" % (name)) |
| 766 | self.warning(" problem with '%s'" % (lines[0])) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 767 | del lines[0] |
| 768 | continue |
| 769 | del lines[0] |
| 770 | l = string.strip(lines[0]) |
| 771 | while len(l) > 2 and l[0:3] != '* @': |
| 772 | while l[0] == '*': |
| 773 | l = l[1:] |
| 774 | desc = desc + ' ' + string.strip(l) |
| 775 | del lines[0] |
| 776 | if len(lines) == 0: |
| 777 | break |
| 778 | l = lines[0] |
| 779 | args.append((arg, desc)) |
| 780 | while len(lines) > 0 and lines[0] == '*': |
| 781 | del lines[0] |
| 782 | desc = "" |
| 783 | while len(lines) > 0: |
| 784 | l = lines[0] |
| 785 | while len(l) > 0 and l[0] == '*': |
| 786 | l = l[1:] |
| 787 | l = string.strip(l) |
| 788 | desc = desc + " " + l |
| 789 | del lines[0] |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 790 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 791 | desc = string.strip(desc) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 792 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 793 | if quiet == 0: |
| 794 | if desc == "": |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 795 | self.warning("Macro comment for %s lack description of the macro" % (name)) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 796 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 797 | return((args, desc)) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 798 | |
| 799 | # |
| 800 | # Parse a comment block and merge the informations found in the |
| 801 | # parameters descriptions, finally returns a block as complete |
| 802 | # as possible |
| 803 | # |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 804 | def mergeFunctionComment(self, name, description, quiet = 0): |
| 805 | if name == 'main': |
| 806 | quiet = 1 |
| 807 | if name[0:2] == '__': |
| 808 | quiet = 1 |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 809 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 810 | (ret, args) = description |
| 811 | desc = "" |
| 812 | retdesc = "" |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 813 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 814 | if self.comment == None: |
| 815 | if not quiet: |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 816 | self.warning("Missing comment for function %s" % (name)) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 817 | return(((ret[0], retdesc), args, desc)) |
| 818 | if self.comment[0] != '*': |
| 819 | if not quiet: |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 820 | self.warning("Missing * in function comment for %s" % (name)) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 821 | return(((ret[0], retdesc), args, desc)) |
| 822 | lines = string.split(self.comment, '\n') |
| 823 | if lines[0] == '*': |
| 824 | del lines[0] |
| 825 | if lines[0] != "* %s:" % (name): |
| 826 | if not quiet: |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 827 | self.warning("Misformatted function comment for %s" % (name)) |
| 828 | self.warning(" Expecting '* %s:' got '%s'" % (name, lines[0])) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 829 | return(((ret[0], retdesc), args, desc)) |
| 830 | del lines[0] |
| 831 | while lines[0] == '*': |
| 832 | del lines[0] |
| 833 | nbargs = len(args) |
| 834 | while len(lines) > 0 and lines[0][0:3] == '* @': |
| 835 | l = lines[0][3:] |
| 836 | try: |
| 837 | (arg, desc) = string.split(l, ':', 1) |
| 838 | desc=string.strip(desc) |
| 839 | arg=string.strip(arg) |
| 840 | except: |
| 841 | if not quiet: |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 842 | self.warning("Misformatted function comment for %s" % (name)) |
| 843 | self.warning(" problem with '%s'" % (lines[0])) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 844 | del lines[0] |
| 845 | continue |
| 846 | del lines[0] |
| 847 | l = string.strip(lines[0]) |
| 848 | while len(l) > 2 and l[0:3] != '* @': |
| 849 | while l[0] == '*': |
| 850 | l = l[1:] |
| 851 | desc = desc + ' ' + string.strip(l) |
| 852 | del lines[0] |
| 853 | if len(lines) == 0: |
| 854 | break |
| 855 | l = lines[0] |
| 856 | i = 0 |
| 857 | while i < nbargs: |
| 858 | if args[i][1] == arg: |
| 859 | args[i] = (args[i][0], arg, desc) |
| 860 | break; |
| 861 | i = i + 1 |
| 862 | if i >= nbargs: |
| 863 | if not quiet: |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 864 | self.warning("Unable to find arg %s from function comment for %s" % ( |
| 865 | arg, name)) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 866 | while len(lines) > 0 and lines[0] == '*': |
| 867 | del lines[0] |
| 868 | desc = "" |
| 869 | while len(lines) > 0: |
| 870 | l = lines[0] |
| 871 | while len(l) > 0 and l[0] == '*': |
| 872 | l = l[1:] |
| 873 | l = string.strip(l) |
| 874 | if len(l) >= 6 and l[0:6] == "return" or l[0:6] == "Return": |
| 875 | try: |
| 876 | l = string.split(l, ' ', 1)[1] |
| 877 | except: |
| 878 | l = "" |
| 879 | retdesc = string.strip(l) |
| 880 | del lines[0] |
| 881 | while len(lines) > 0: |
| 882 | l = lines[0] |
| 883 | while len(l) > 0 and l[0] == '*': |
| 884 | l = l[1:] |
| 885 | l = string.strip(l) |
| 886 | retdesc = retdesc + " " + l |
| 887 | del lines[0] |
| 888 | else: |
| 889 | desc = desc + " " + l |
| 890 | del lines[0] |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 891 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 892 | retdesc = string.strip(retdesc) |
| 893 | desc = string.strip(desc) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 894 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 895 | if quiet == 0: |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 896 | # |
| 897 | # report missing comments |
| 898 | # |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 899 | i = 0 |
| 900 | while i < nbargs: |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 901 | if args[i][2] == None and args[i][0] != "void" and \ |
| 902 | ((args[i][1] != None) or (args[i][1] == '')): |
William M. Brack | 21e4ef2 | 2005-01-02 09:53:13 +0000 | [diff] [blame] | 903 | self.warning("Function comment for %s lacks description of arg %s" % (name, args[i][1])) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 904 | i = i + 1 |
| 905 | if retdesc == "" and ret[0] != "void": |
William M. Brack | 21e4ef2 | 2005-01-02 09:53:13 +0000 | [diff] [blame] | 906 | self.warning("Function comment for %s lacks description of return value" % (name)) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 907 | if desc == "": |
William M. Brack | 21e4ef2 | 2005-01-02 09:53:13 +0000 | [diff] [blame] | 908 | self.warning("Function comment for %s lacks description of the function" % (name)) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 909 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 910 | return(((ret[0], retdesc), args, desc)) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 911 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 912 | def parsePreproc(self, token): |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 913 | if debug: |
| 914 | print "=> preproc ", token, self.lexer.tokens |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 915 | name = token[1] |
| 916 | if name == "#include": |
| 917 | token = self.lexer.token() |
| 918 | if token == None: |
| 919 | return None |
| 920 | if token[0] == 'preproc': |
| 921 | self.index_add(token[1], self.filename, not self.is_header, |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 922 | "include") |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 923 | return self.lexer.token() |
| 924 | return token |
| 925 | if name == "#define": |
| 926 | token = self.lexer.token() |
| 927 | if token == None: |
| 928 | return None |
| 929 | if token[0] == 'preproc': |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 930 | # TODO macros with arguments |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 931 | name = token[1] |
| 932 | lst = [] |
| 933 | token = self.lexer.token() |
| 934 | while token != None and token[0] == 'preproc' and \ |
| 935 | token[1][0] != '#': |
| 936 | lst.append(token[1]) |
| 937 | token = self.lexer.token() |
| 938 | try: |
| 939 | name = string.split(name, '(') [0] |
| 940 | except: |
| 941 | pass |
| 942 | info = self.parseMacroComment(name, not self.is_header) |
| 943 | self.index_add(name, self.filename, not self.is_header, |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 944 | "macro", info) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 945 | return token |
William M. Brack | 21e4ef2 | 2005-01-02 09:53:13 +0000 | [diff] [blame] | 946 | |
| 947 | # |
| 948 | # Processing of conditionals modified by Bill 1/1/05 |
| 949 | # |
| 950 | # We process conditionals (i.e. tokens from #ifdef, #ifndef, |
| 951 | # #if, #else and #endif) for headers and mainline code, |
| 952 | # store the ones from the header in libxml2-api.xml, and later |
| 953 | # (in the routine merge_public) verify that the two (header and |
| 954 | # mainline code) agree. |
| 955 | # |
| 956 | # There is a small problem with processing the headers. Some of |
| 957 | # the variables are not concerned with enabling / disabling of |
| 958 | # library functions (e.g. '__XML_PARSER_H__'), and we don't want |
| 959 | # them to be included in libxml2-api.xml, or involved in |
| 960 | # the check between the header and the mainline code. To |
| 961 | # accomplish this, we ignore any conditional which doesn't include |
| 962 | # the string 'ENABLED' |
| 963 | # |
| 964 | if name == "#ifdef": |
| 965 | apstr = self.lexer.tokens[0][1] |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 966 | try: |
William M. Brack | 21e4ef2 | 2005-01-02 09:53:13 +0000 | [diff] [blame] | 967 | self.defines.append(apstr) |
| 968 | if string.find(apstr, 'ENABLED') != -1: |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 969 | self.conditionals.append("defined(%s)" % apstr) |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 970 | except: |
| 971 | pass |
William M. Brack | 21e4ef2 | 2005-01-02 09:53:13 +0000 | [diff] [blame] | 972 | elif name == "#ifndef": |
| 973 | apstr = self.lexer.tokens[0][1] |
| 974 | try: |
| 975 | self.defines.append(apstr) |
| 976 | if string.find(apstr, 'ENABLED') != -1: |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 977 | self.conditionals.append("!defined(%s)" % apstr) |
William M. Brack | 21e4ef2 | 2005-01-02 09:53:13 +0000 | [diff] [blame] | 978 | except: |
| 979 | pass |
| 980 | elif name == "#if": |
| 981 | apstr = "" |
| 982 | for tok in self.lexer.tokens: |
| 983 | if apstr != "": |
| 984 | apstr = apstr + " " |
| 985 | apstr = apstr + tok[1] |
| 986 | try: |
| 987 | self.defines.append(apstr) |
| 988 | if string.find(apstr, 'ENABLED') != -1: |
| 989 | self.conditionals.append(apstr) |
| 990 | except: |
| 991 | pass |
| 992 | elif name == "#else": |
| 993 | if self.conditionals != [] and \ |
| 994 | string.find(self.defines[-1], 'ENABLED') != -1: |
| 995 | self.conditionals[-1] = "!(%s)" % self.conditionals[-1] |
| 996 | elif name == "#endif": |
| 997 | if self.conditionals != [] and \ |
| 998 | string.find(self.defines[-1], 'ENABLED') != -1: |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 999 | self.conditionals = self.conditionals[:-1] |
| 1000 | self.defines = self.defines[:-1] |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1001 | token = self.lexer.token() |
| 1002 | while token != None and token[0] == 'preproc' and \ |
| 1003 | token[1][0] != '#': |
| 1004 | token = self.lexer.token() |
| 1005 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1006 | |
| 1007 | # |
| 1008 | # token acquisition on top of the lexer, it handle internally |
| 1009 | # preprocessor and comments since they are logically not part of |
| 1010 | # the program structure. |
| 1011 | # |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1012 | def token(self): |
| 1013 | global ignored_words |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1014 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1015 | token = self.lexer.token() |
| 1016 | while token != None: |
| 1017 | if token[0] == 'comment': |
| 1018 | token = self.parseComment(token) |
| 1019 | continue |
| 1020 | elif token[0] == 'preproc': |
| 1021 | token = self.parsePreproc(token) |
| 1022 | continue |
Daniel Veillard | 99dd763 | 2005-05-06 11:40:56 +0000 | [diff] [blame] | 1023 | elif token[0] == "name" and token[1] == "__const": |
| 1024 | token = ("name", "const") |
| 1025 | return token |
| 1026 | elif token[0] == "name" and token[1] == "__attribute": |
| 1027 | token = self.lexer.token() |
| 1028 | while token != None and token[1] != ";": |
| 1029 | token = self.lexer.token() |
| 1030 | return token |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1031 | elif token[0] == "name" and ignored_words.has_key(token[1]): |
| 1032 | (n, info) = ignored_words[token[1]] |
| 1033 | i = 0 |
| 1034 | while i < n: |
| 1035 | token = self.lexer.token() |
| 1036 | i = i + 1 |
| 1037 | token = self.lexer.token() |
| 1038 | continue |
| 1039 | else: |
| 1040 | if debug: |
| 1041 | print "=> ", token |
| 1042 | return token |
| 1043 | return None |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1044 | |
| 1045 | # |
| 1046 | # Parse a typedef, it records the type and its name. |
| 1047 | # |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1048 | def parseTypedef(self, token): |
| 1049 | if token == None: |
| 1050 | return None |
| 1051 | token = self.parseType(token) |
| 1052 | if token == None: |
| 1053 | self.error("parsing typedef") |
| 1054 | return None |
| 1055 | base_type = self.type |
| 1056 | type = base_type |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1057 | #self.debug("end typedef type", token) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1058 | while token != None: |
| 1059 | if token[0] == "name": |
| 1060 | name = token[1] |
| 1061 | signature = self.signature |
| 1062 | if signature != None: |
| 1063 | type = string.split(type, '(')[0] |
| 1064 | d = self.mergeFunctionComment(name, |
| 1065 | ((type, None), signature), 1) |
| 1066 | self.index_add(name, self.filename, not self.is_header, |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1067 | "functype", d) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1068 | else: |
| 1069 | if base_type == "struct": |
| 1070 | self.index_add(name, self.filename, not self.is_header, |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1071 | "struct", type) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1072 | base_type = "struct " + name |
| 1073 | else: |
Daniel Veillard | 99b7850 | 2005-09-11 22:04:08 +0000 | [diff] [blame] | 1074 | # TODO report missing or misformatted comments |
| 1075 | info = self.parseTypeComment(name, 1) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1076 | self.index_add(name, self.filename, not self.is_header, |
Daniel Veillard | 99b7850 | 2005-09-11 22:04:08 +0000 | [diff] [blame] | 1077 | "typedef", type, info) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1078 | token = self.token() |
| 1079 | else: |
| 1080 | self.error("parsing typedef: expecting a name") |
| 1081 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1082 | #self.debug("end typedef", token) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1083 | if token != None and token[0] == 'sep' and token[1] == ',': |
| 1084 | type = base_type |
| 1085 | token = self.token() |
| 1086 | while token != None and token[0] == "op": |
| 1087 | type = type + token[1] |
| 1088 | token = self.token() |
| 1089 | elif token != None and token[0] == 'sep' and token[1] == ';': |
| 1090 | break; |
| 1091 | elif token != None and token[0] == 'name': |
| 1092 | type = base_type |
| 1093 | continue; |
| 1094 | else: |
| 1095 | self.error("parsing typedef: expecting ';'", token) |
| 1096 | return token |
| 1097 | token = self.token() |
| 1098 | return token |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 1099 | |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1100 | # |
| 1101 | # Parse a C code block, used for functions it parse till |
| 1102 | # the balancing } included |
| 1103 | # |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1104 | def parseBlock(self, token): |
| 1105 | while token != None: |
| 1106 | if token[0] == "sep" and token[1] == "{": |
| 1107 | token = self.token() |
| 1108 | token = self.parseBlock(token) |
| 1109 | elif token[0] == "sep" and token[1] == "}": |
| 1110 | self.comment = None |
| 1111 | token = self.token() |
| 1112 | return token |
| 1113 | else: |
| 1114 | if self.collect_ref == 1: |
| 1115 | oldtok = token |
| 1116 | token = self.token() |
| 1117 | if oldtok[0] == "name" and oldtok[1][0:3] == "xml": |
| 1118 | if token[0] == "sep" and token[1] == "(": |
| 1119 | self.index_add_ref(oldtok[1], self.filename, |
Daniel Veillard | d8cf906 | 2003-11-11 21:12:36 +0000 | [diff] [blame] | 1120 | 0, "function") |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1121 | token = self.token() |
| 1122 | elif token[0] == "name": |
| 1123 | token = self.token() |
| 1124 | if token[0] == "sep" and (token[1] == ";" or |
| 1125 | token[1] == "," or token[1] == "="): |
| 1126 | self.index_add_ref(oldtok[1], self.filename, |
Daniel Veillard | d8cf906 | 2003-11-11 21:12:36 +0000 | [diff] [blame] | 1127 | 0, "type") |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1128 | elif oldtok[0] == "name" and oldtok[1][0:4] == "XML_": |
| 1129 | self.index_add_ref(oldtok[1], self.filename, |
Daniel Veillard | d8cf906 | 2003-11-11 21:12:36 +0000 | [diff] [blame] | 1130 | 0, "typedef") |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1131 | elif oldtok[0] == "name" and oldtok[1][0:7] == "LIBXML_": |
| 1132 | self.index_add_ref(oldtok[1], self.filename, |
Daniel Veillard | d8cf906 | 2003-11-11 21:12:36 +0000 | [diff] [blame] | 1133 | 0, "typedef") |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 1134 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1135 | else: |
| 1136 | token = self.token() |
| 1137 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1138 | |
| 1139 | # |
| 1140 | # Parse a C struct definition till the balancing } |
| 1141 | # |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1142 | def parseStruct(self, token): |
| 1143 | fields = [] |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1144 | #self.debug("start parseStruct", token) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1145 | while token != None: |
| 1146 | if token[0] == "sep" and token[1] == "{": |
| 1147 | token = self.token() |
| 1148 | token = self.parseTypeBlock(token) |
| 1149 | elif token[0] == "sep" and token[1] == "}": |
| 1150 | self.struct_fields = fields |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1151 | #self.debug("end parseStruct", token) |
| 1152 | #print fields |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1153 | token = self.token() |
| 1154 | return token |
| 1155 | else: |
| 1156 | base_type = self.type |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1157 | #self.debug("before parseType", token) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1158 | token = self.parseType(token) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1159 | #self.debug("after parseType", token) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1160 | if token != None and token[0] == "name": |
| 1161 | fname = token[1] |
| 1162 | token = self.token() |
| 1163 | if token[0] == "sep" and token[1] == ";": |
| 1164 | self.comment = None |
| 1165 | token = self.token() |
| 1166 | fields.append((self.type, fname, self.comment)) |
| 1167 | self.comment = None |
| 1168 | else: |
| 1169 | self.error("parseStruct: expecting ;", token) |
| 1170 | elif token != None and token[0] == "sep" and token[1] == "{": |
| 1171 | token = self.token() |
| 1172 | token = self.parseTypeBlock(token) |
| 1173 | if token != None and token[0] == "name": |
| 1174 | token = self.token() |
| 1175 | if token != None and token[0] == "sep" and token[1] == ";": |
| 1176 | token = self.token() |
| 1177 | else: |
| 1178 | self.error("parseStruct: expecting ;", token) |
| 1179 | else: |
| 1180 | self.error("parseStruct: name", token) |
| 1181 | token = self.token() |
| 1182 | self.type = base_type; |
| 1183 | self.struct_fields = fields |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1184 | #self.debug("end parseStruct", token) |
| 1185 | #print fields |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1186 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1187 | |
| 1188 | # |
| 1189 | # Parse a C enum block, parse till the balancing } |
| 1190 | # |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1191 | def parseEnumBlock(self, token): |
| 1192 | self.enums = [] |
| 1193 | name = None |
| 1194 | self.comment = None |
| 1195 | comment = "" |
| 1196 | value = "0" |
| 1197 | while token != None: |
| 1198 | if token[0] == "sep" and token[1] == "{": |
| 1199 | token = self.token() |
| 1200 | token = self.parseTypeBlock(token) |
| 1201 | elif token[0] == "sep" and token[1] == "}": |
| 1202 | if name != None: |
| 1203 | if self.comment != None: |
| 1204 | comment = self.comment |
| 1205 | self.comment = None |
| 1206 | self.enums.append((name, value, comment)) |
| 1207 | token = self.token() |
| 1208 | return token |
| 1209 | elif token[0] == "name": |
| 1210 | if name != None: |
| 1211 | if self.comment != None: |
| 1212 | comment = string.strip(self.comment) |
| 1213 | self.comment = None |
| 1214 | self.enums.append((name, value, comment)) |
| 1215 | name = token[1] |
| 1216 | comment = "" |
| 1217 | token = self.token() |
| 1218 | if token[0] == "op" and token[1][0] == "=": |
| 1219 | value = "" |
| 1220 | if len(token[1]) > 1: |
| 1221 | value = token[1][1:] |
| 1222 | token = self.token() |
| 1223 | while token[0] != "sep" or (token[1] != ',' and |
| 1224 | token[1] != '}'): |
| 1225 | value = value + token[1] |
| 1226 | token = self.token() |
| 1227 | else: |
| 1228 | try: |
| 1229 | value = "%d" % (int(value) + 1) |
| 1230 | except: |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 1231 | self.warning("Failed to compute value of enum %s" % (name)) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1232 | value="" |
| 1233 | if token[0] == "sep" and token[1] == ",": |
| 1234 | token = self.token() |
| 1235 | else: |
| 1236 | token = self.token() |
| 1237 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1238 | |
| 1239 | # |
| 1240 | # Parse a C definition block, used for structs it parse till |
| 1241 | # the balancing } |
| 1242 | # |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1243 | def parseTypeBlock(self, token): |
| 1244 | while token != None: |
| 1245 | if token[0] == "sep" and token[1] == "{": |
| 1246 | token = self.token() |
| 1247 | token = self.parseTypeBlock(token) |
| 1248 | elif token[0] == "sep" and token[1] == "}": |
| 1249 | token = self.token() |
| 1250 | return token |
| 1251 | else: |
| 1252 | token = self.token() |
| 1253 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1254 | |
| 1255 | # |
| 1256 | # Parse a type: the fact that the type name can either occur after |
| 1257 | # the definition or within the definition makes it a little harder |
| 1258 | # if inside, the name token is pushed back before returning |
| 1259 | # |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1260 | def parseType(self, token): |
| 1261 | self.type = "" |
| 1262 | self.struct_fields = [] |
| 1263 | self.signature = None |
| 1264 | if token == None: |
| 1265 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1266 | |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 1267 | while token[0] == "name" and ( |
William M. Brack | fbb619f | 2005-06-06 13:49:18 +0000 | [diff] [blame] | 1268 | token[1] == "const" or \ |
| 1269 | token[1] == "unsigned" or \ |
| 1270 | token[1] == "signed"): |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1271 | if self.type == "": |
| 1272 | self.type = token[1] |
| 1273 | else: |
| 1274 | self.type = self.type + " " + token[1] |
| 1275 | token = self.token() |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1276 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1277 | if token[0] == "name" and (token[1] == "long" or token[1] == "short"): |
| 1278 | if self.type == "": |
| 1279 | self.type = token[1] |
| 1280 | else: |
| 1281 | self.type = self.type + " " + token[1] |
| 1282 | if token[0] == "name" and token[1] == "int": |
| 1283 | if self.type == "": |
| 1284 | self.type = tmp[1] |
| 1285 | else: |
| 1286 | self.type = self.type + " " + tmp[1] |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 1287 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1288 | elif token[0] == "name" and token[1] == "struct": |
| 1289 | if self.type == "": |
| 1290 | self.type = token[1] |
| 1291 | else: |
| 1292 | self.type = self.type + " " + token[1] |
| 1293 | token = self.token() |
| 1294 | nametok = None |
| 1295 | if token[0] == "name": |
| 1296 | nametok = token |
| 1297 | token = self.token() |
| 1298 | if token != None and token[0] == "sep" and token[1] == "{": |
| 1299 | token = self.token() |
| 1300 | token = self.parseStruct(token) |
| 1301 | elif token != None and token[0] == "op" and token[1] == "*": |
| 1302 | self.type = self.type + " " + nametok[1] + " *" |
| 1303 | token = self.token() |
| 1304 | while token != None and token[0] == "op" and token[1] == "*": |
| 1305 | self.type = self.type + " *" |
| 1306 | token = self.token() |
| 1307 | if token[0] == "name": |
| 1308 | nametok = token |
| 1309 | token = self.token() |
| 1310 | else: |
| 1311 | self.error("struct : expecting name", token) |
| 1312 | return token |
| 1313 | elif token != None and token[0] == "name" and nametok != None: |
| 1314 | self.type = self.type + " " + nametok[1] |
| 1315 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1316 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1317 | if nametok != None: |
| 1318 | self.lexer.push(token) |
| 1319 | token = nametok |
| 1320 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1321 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1322 | elif token[0] == "name" and token[1] == "enum": |
| 1323 | if self.type == "": |
| 1324 | self.type = token[1] |
| 1325 | else: |
| 1326 | self.type = self.type + " " + token[1] |
| 1327 | self.enums = [] |
| 1328 | token = self.token() |
| 1329 | if token != None and token[0] == "sep" and token[1] == "{": |
| 1330 | token = self.token() |
| 1331 | token = self.parseEnumBlock(token) |
| 1332 | else: |
| 1333 | self.error("parsing enum: expecting '{'", token) |
| 1334 | enum_type = None |
| 1335 | if token != None and token[0] != "name": |
| 1336 | self.lexer.push(token) |
| 1337 | token = ("name", "enum") |
| 1338 | else: |
| 1339 | enum_type = token[1] |
| 1340 | for enum in self.enums: |
| 1341 | self.index_add(enum[0], self.filename, |
| 1342 | not self.is_header, "enum", |
| 1343 | (enum[1], enum[2], enum_type)) |
| 1344 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1345 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1346 | elif token[0] == "name": |
| 1347 | if self.type == "": |
| 1348 | self.type = token[1] |
| 1349 | else: |
| 1350 | self.type = self.type + " " + token[1] |
| 1351 | else: |
| 1352 | self.error("parsing type %s: expecting a name" % (self.type), |
| 1353 | token) |
| 1354 | return token |
| 1355 | token = self.token() |
| 1356 | while token != None and (token[0] == "op" or |
| 1357 | token[0] == "name" and token[1] == "const"): |
| 1358 | self.type = self.type + " " + token[1] |
| 1359 | token = self.token() |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1360 | |
| 1361 | # |
| 1362 | # if there is a parenthesis here, this means a function type |
| 1363 | # |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1364 | if token != None and token[0] == "sep" and token[1] == '(': |
| 1365 | self.type = self.type + token[1] |
| 1366 | token = self.token() |
| 1367 | while token != None and token[0] == "op" and token[1] == '*': |
| 1368 | self.type = self.type + token[1] |
| 1369 | token = self.token() |
| 1370 | if token == None or token[0] != "name" : |
| 1371 | self.error("parsing function type, name expected", token); |
| 1372 | return token |
| 1373 | self.type = self.type + token[1] |
| 1374 | nametok = token |
| 1375 | token = self.token() |
| 1376 | if token != None and token[0] == "sep" and token[1] == ')': |
| 1377 | self.type = self.type + token[1] |
| 1378 | token = self.token() |
| 1379 | if token != None and token[0] == "sep" and token[1] == '(': |
| 1380 | token = self.token() |
| 1381 | type = self.type; |
| 1382 | token = self.parseSignature(token); |
| 1383 | self.type = type; |
| 1384 | else: |
| 1385 | self.error("parsing function type, '(' expected", token); |
| 1386 | return token |
| 1387 | else: |
| 1388 | self.error("parsing function type, ')' expected", token); |
| 1389 | return token |
| 1390 | self.lexer.push(token) |
| 1391 | token = nametok |
| 1392 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1393 | |
| 1394 | # |
| 1395 | # do some lookahead for arrays |
| 1396 | # |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1397 | if token != None and token[0] == "name": |
| 1398 | nametok = token |
| 1399 | token = self.token() |
| 1400 | if token != None and token[0] == "sep" and token[1] == '[': |
| 1401 | self.type = self.type + nametok[1] |
| 1402 | while token != None and token[0] == "sep" and token[1] == '[': |
| 1403 | self.type = self.type + token[1] |
| 1404 | token = self.token() |
| 1405 | while token != None and token[0] != 'sep' and \ |
| 1406 | token[1] != ']' and token[1] != ';': |
| 1407 | self.type = self.type + token[1] |
| 1408 | token = self.token() |
| 1409 | if token != None and token[0] == 'sep' and token[1] == ']': |
| 1410 | self.type = self.type + token[1] |
| 1411 | token = self.token() |
| 1412 | else: |
| 1413 | self.error("parsing array type, ']' expected", token); |
| 1414 | return token |
| 1415 | elif token != None and token[0] == "sep" and token[1] == ':': |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1416 | # remove :12 in case it's a limited int size |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1417 | token = self.token() |
| 1418 | token = self.token() |
| 1419 | self.lexer.push(token) |
| 1420 | token = nametok |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1421 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1422 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1423 | |
| 1424 | # |
| 1425 | # Parse a signature: '(' has been parsed and we scan the type definition |
| 1426 | # up to the ')' included |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1427 | def parseSignature(self, token): |
| 1428 | signature = [] |
| 1429 | if token != None and token[0] == "sep" and token[1] == ')': |
| 1430 | self.signature = [] |
| 1431 | token = self.token() |
| 1432 | return token |
| 1433 | while token != None: |
| 1434 | token = self.parseType(token) |
| 1435 | if token != None and token[0] == "name": |
| 1436 | signature.append((self.type, token[1], None)) |
| 1437 | token = self.token() |
| 1438 | elif token != None and token[0] == "sep" and token[1] == ',': |
| 1439 | token = self.token() |
| 1440 | continue |
| 1441 | elif token != None and token[0] == "sep" and token[1] == ')': |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1442 | # only the type was provided |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1443 | if self.type == "...": |
| 1444 | signature.append((self.type, "...", None)) |
| 1445 | else: |
| 1446 | signature.append((self.type, None, None)) |
| 1447 | if token != None and token[0] == "sep": |
| 1448 | if token[1] == ',': |
| 1449 | token = self.token() |
| 1450 | continue |
| 1451 | elif token[1] == ')': |
| 1452 | token = self.token() |
| 1453 | break |
| 1454 | self.signature = signature |
| 1455 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1456 | |
| 1457 | # |
| 1458 | # Parse a global definition, be it a type, variable or function |
| 1459 | # the extern "C" blocks are a bit nasty and require it to recurse. |
| 1460 | # |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1461 | def parseGlobal(self, token): |
| 1462 | static = 0 |
| 1463 | if token[1] == 'extern': |
| 1464 | token = self.token() |
| 1465 | if token == None: |
| 1466 | return token |
| 1467 | if token[0] == 'string': |
| 1468 | if token[1] == 'C': |
| 1469 | token = self.token() |
| 1470 | if token == None: |
| 1471 | return token |
| 1472 | if token[0] == 'sep' and token[1] == "{": |
| 1473 | token = self.token() |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1474 | # print 'Entering extern "C line ', self.lineno() |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1475 | while token != None and (token[0] != 'sep' or |
| 1476 | token[1] != "}"): |
| 1477 | if token[0] == 'name': |
| 1478 | token = self.parseGlobal(token) |
| 1479 | else: |
| 1480 | self.error( |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1481 | "token %s %s unexpected at the top level" % ( |
| 1482 | token[0], token[1])) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1483 | token = self.parseGlobal(token) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1484 | # print 'Exiting extern "C" line', self.lineno() |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1485 | token = self.token() |
| 1486 | return token |
| 1487 | else: |
| 1488 | return token |
| 1489 | elif token[1] == 'static': |
| 1490 | static = 1 |
| 1491 | token = self.token() |
| 1492 | if token == None or token[0] != 'name': |
| 1493 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1494 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1495 | if token[1] == 'typedef': |
| 1496 | token = self.token() |
| 1497 | return self.parseTypedef(token) |
| 1498 | else: |
| 1499 | token = self.parseType(token) |
| 1500 | type_orig = self.type |
| 1501 | if token == None or token[0] != "name": |
| 1502 | return token |
| 1503 | type = type_orig |
| 1504 | self.name = token[1] |
| 1505 | token = self.token() |
| 1506 | while token != None and (token[0] == "sep" or token[0] == "op"): |
| 1507 | if token[0] == "sep": |
| 1508 | if token[1] == "[": |
| 1509 | type = type + token[1] |
| 1510 | token = self.token() |
| 1511 | while token != None and (token[0] != "sep" or \ |
| 1512 | token[1] != ";"): |
| 1513 | type = type + token[1] |
| 1514 | token = self.token() |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1515 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1516 | if token != None and token[0] == "op" and token[1] == "=": |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1517 | # |
| 1518 | # Skip the initialization of the variable |
| 1519 | # |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1520 | token = self.token() |
| 1521 | if token[0] == 'sep' and token[1] == '{': |
| 1522 | token = self.token() |
| 1523 | token = self.parseBlock(token) |
| 1524 | else: |
| 1525 | self.comment = None |
| 1526 | while token != None and (token[0] != "sep" or \ |
| 1527 | (token[1] != ';' and token[1] != ',')): |
| 1528 | token = self.token() |
| 1529 | self.comment = None |
| 1530 | if token == None or token[0] != "sep" or (token[1] != ';' and |
| 1531 | token[1] != ','): |
| 1532 | self.error("missing ';' or ',' after value") |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1533 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1534 | if token != None and token[0] == "sep": |
| 1535 | if token[1] == ";": |
| 1536 | self.comment = None |
| 1537 | token = self.token() |
| 1538 | if type == "struct": |
| 1539 | self.index_add(self.name, self.filename, |
| 1540 | not self.is_header, "struct", self.struct_fields) |
| 1541 | else: |
| 1542 | self.index_add(self.name, self.filename, |
| 1543 | not self.is_header, "variable", type) |
| 1544 | break |
| 1545 | elif token[1] == "(": |
| 1546 | token = self.token() |
| 1547 | token = self.parseSignature(token) |
| 1548 | if token == None: |
| 1549 | return None |
| 1550 | if token[0] == "sep" and token[1] == ";": |
| 1551 | d = self.mergeFunctionComment(self.name, |
| 1552 | ((type, None), self.signature), 1) |
| 1553 | self.index_add(self.name, self.filename, static, |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1554 | "function", d) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1555 | token = self.token() |
| 1556 | elif token[0] == "sep" and token[1] == "{": |
| 1557 | d = self.mergeFunctionComment(self.name, |
| 1558 | ((type, None), self.signature), static) |
| 1559 | self.index_add(self.name, self.filename, static, |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1560 | "function", d) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1561 | token = self.token() |
| 1562 | token = self.parseBlock(token); |
| 1563 | elif token[1] == ',': |
| 1564 | self.comment = None |
| 1565 | self.index_add(self.name, self.filename, static, |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1566 | "variable", type) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1567 | type = type_orig |
| 1568 | token = self.token() |
| 1569 | while token != None and token[0] == "sep": |
| 1570 | type = type + token[1] |
| 1571 | token = self.token() |
| 1572 | if token != None and token[0] == "name": |
| 1573 | self.name = token[1] |
| 1574 | token = self.token() |
| 1575 | else: |
| 1576 | break |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 1577 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1578 | return token |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1579 | |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1580 | def parse(self): |
Daniel Veillard | 1e90661 | 2003-12-05 14:57:46 +0000 | [diff] [blame] | 1581 | self.warning("Parsing %s" % (self.filename)) |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1582 | token = self.token() |
| 1583 | while token != None: |
| 1584 | if token[0] == 'name': |
| 1585 | token = self.parseGlobal(token) |
| 1586 | else: |
| 1587 | self.error("token %s %s unexpected at the top level" % ( |
| 1588 | token[0], token[1])) |
| 1589 | token = self.parseGlobal(token) |
| 1590 | return |
| 1591 | self.parseTopComment(self.top_comment) |
| 1592 | return self.index |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 1593 | |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1594 | |
| 1595 | class docBuilder: |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1596 | """A documentation builder""" |
| 1597 | def __init__(self, name, directories=['.'], excludes=[]): |
| 1598 | self.name = name |
| 1599 | self.directories = directories |
| 1600 | self.excludes = excludes + ignored_files.keys() |
| 1601 | self.modules = {} |
| 1602 | self.headers = {} |
| 1603 | self.idx = index() |
| 1604 | self.xref = {} |
| 1605 | self.index = {} |
| 1606 | if name == 'libxml2': |
| 1607 | self.basename = 'libxml' |
| 1608 | else: |
| 1609 | self.basename = name |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1610 | |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1611 | def indexString(self, id, str): |
| 1612 | if str == None: |
| 1613 | return |
| 1614 | str = string.replace(str, "'", ' ') |
| 1615 | str = string.replace(str, '"', ' ') |
| 1616 | str = string.replace(str, "/", ' ') |
| 1617 | str = string.replace(str, '*', ' ') |
| 1618 | str = string.replace(str, "[", ' ') |
| 1619 | str = string.replace(str, "]", ' ') |
| 1620 | str = string.replace(str, "(", ' ') |
| 1621 | str = string.replace(str, ")", ' ') |
| 1622 | str = string.replace(str, "<", ' ') |
| 1623 | str = string.replace(str, '>', ' ') |
| 1624 | str = string.replace(str, "&", ' ') |
| 1625 | str = string.replace(str, '#', ' ') |
| 1626 | str = string.replace(str, ",", ' ') |
| 1627 | str = string.replace(str, '.', ' ') |
| 1628 | str = string.replace(str, ';', ' ') |
| 1629 | tokens = string.split(str) |
| 1630 | for token in tokens: |
| 1631 | try: |
| 1632 | c = token[0] |
| 1633 | if string.find(string.letters, c) < 0: |
| 1634 | pass |
| 1635 | elif len(token) < 3: |
| 1636 | pass |
| 1637 | else: |
| 1638 | lower = string.lower(token) |
| 1639 | # TODO: generalize this a bit |
| 1640 | if lower == 'and' or lower == 'the': |
| 1641 | pass |
| 1642 | elif self.xref.has_key(token): |
| 1643 | self.xref[token].append(id) |
| 1644 | else: |
| 1645 | self.xref[token] = [id] |
| 1646 | except: |
| 1647 | pass |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1648 | |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1649 | def analyze(self): |
| 1650 | print "Project %s : %d headers, %d modules" % (self.name, len(self.headers.keys()), len(self.modules.keys())) |
| 1651 | self.idx.analyze() |
| 1652 | |
| 1653 | def scanHeaders(self): |
| 1654 | for header in self.headers.keys(): |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1655 | parser = CParser(header) |
| 1656 | idx = parser.parse() |
| 1657 | self.headers[header] = idx; |
| 1658 | self.idx.merge(idx) |
| 1659 | |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1660 | def scanModules(self): |
| 1661 | for module in self.modules.keys(): |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1662 | parser = CParser(module) |
| 1663 | idx = parser.parse() |
| 1664 | # idx.analyze() |
| 1665 | self.modules[module] = idx |
| 1666 | self.idx.merge_public(idx) |
| 1667 | |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1668 | def scan(self): |
| 1669 | for directory in self.directories: |
| 1670 | files = glob.glob(directory + "/*.c") |
| 1671 | for file in files: |
| 1672 | skip = 0 |
| 1673 | for excl in self.excludes: |
| 1674 | if string.find(file, excl) != -1: |
| 1675 | skip = 1; |
| 1676 | break |
| 1677 | if skip == 0: |
| 1678 | self.modules[file] = None; |
| 1679 | files = glob.glob(directory + "/*.h") |
| 1680 | for file in files: |
| 1681 | skip = 0 |
| 1682 | for excl in self.excludes: |
| 1683 | if string.find(file, excl) != -1: |
| 1684 | skip = 1; |
| 1685 | break |
| 1686 | if skip == 0: |
| 1687 | self.headers[file] = None; |
| 1688 | self.scanHeaders() |
| 1689 | self.scanModules() |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 1690 | |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1691 | def modulename_file(self, file): |
| 1692 | module = os.path.basename(file) |
| 1693 | if module[-2:] == '.h': |
| 1694 | module = module[:-2] |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 1695 | elif module[-2:] == '.c': |
| 1696 | module = module[:-2] |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1697 | return module |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1698 | |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1699 | def serialize_enum(self, output, name): |
| 1700 | id = self.idx.enums[name] |
| 1701 | output.write(" <enum name='%s' file='%s'" % (name, |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 1702 | self.modulename_file(id.header))) |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1703 | if id.info != None: |
| 1704 | info = id.info |
| 1705 | if info[0] != None and info[0] != '': |
| 1706 | try: |
| 1707 | val = eval(info[0]) |
| 1708 | except: |
| 1709 | val = info[0] |
| 1710 | output.write(" value='%s'" % (val)); |
| 1711 | if info[2] != None and info[2] != '': |
| 1712 | output.write(" type='%s'" % info[2]); |
| 1713 | if info[1] != None and info[1] != '': |
| 1714 | output.write(" info='%s'" % escape(info[1])); |
| 1715 | output.write("/>\n") |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1716 | |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1717 | def serialize_macro(self, output, name): |
| 1718 | id = self.idx.macros[name] |
| 1719 | output.write(" <macro name='%s' file='%s'>\n" % (name, |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 1720 | self.modulename_file(id.header))) |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1721 | if id.info != None: |
| 1722 | try: |
| 1723 | (args, desc) = id.info |
| 1724 | if desc != None and desc != "": |
| 1725 | output.write(" <info>%s</info>\n" % (escape(desc))) |
| 1726 | self.indexString(name, desc) |
| 1727 | for arg in args: |
| 1728 | (name, desc) = arg |
| 1729 | if desc != None and desc != "": |
| 1730 | output.write(" <arg name='%s' info='%s'/>\n" % ( |
| 1731 | name, escape(desc))) |
| 1732 | self.indexString(name, desc) |
| 1733 | else: |
| 1734 | output.write(" <arg name='%s'/>\n" % (name)) |
| 1735 | except: |
| 1736 | pass |
| 1737 | output.write(" </macro>\n") |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1738 | |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1739 | def serialize_typedef(self, output, name): |
| 1740 | id = self.idx.typedefs[name] |
| 1741 | if id.info[0:7] == 'struct ': |
| 1742 | output.write(" <struct name='%s' file='%s' type='%s'" % ( |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 1743 | name, self.modulename_file(id.header), id.info)) |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1744 | name = id.info[7:] |
| 1745 | if self.idx.structs.has_key(name) and ( \ |
| 1746 | type(self.idx.structs[name].info) == type(()) or |
Daniel Veillard | c1eed32 | 2002-12-12 11:01:32 +0000 | [diff] [blame] | 1747 | type(self.idx.structs[name].info) == type([])): |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1748 | output.write(">\n"); |
| 1749 | try: |
| 1750 | for field in self.idx.structs[name].info: |
| 1751 | desc = field[2] |
| 1752 | self.indexString(name, desc) |
| 1753 | if desc == None: |
| 1754 | desc = '' |
| 1755 | else: |
| 1756 | desc = escape(desc) |
| 1757 | output.write(" <field name='%s' type='%s' info='%s'/>\n" % (field[1] , field[0], desc)) |
| 1758 | except: |
| 1759 | print "Failed to serialize struct %s" % (name) |
| 1760 | output.write(" </struct>\n") |
| 1761 | else: |
| 1762 | output.write("/>\n"); |
| 1763 | else : |
Daniel Veillard | 99b7850 | 2005-09-11 22:04:08 +0000 | [diff] [blame] | 1764 | output.write(" <typedef name='%s' file='%s' type='%s'" % ( |
| 1765 | name, self.modulename_file(id.header), id.info)) |
| 1766 | try: |
| 1767 | desc = id.extra |
| 1768 | if desc != None and desc != "": |
| 1769 | output.write(">\n <info>%s</info>\n" % (escape(desc))) |
| 1770 | output.write(" </typedef>\n") |
| 1771 | else: |
| 1772 | output.write("/>\n") |
| 1773 | except: |
| 1774 | output.write("/>\n") |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1775 | |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1776 | def serialize_variable(self, output, name): |
| 1777 | id = self.idx.variables[name] |
| 1778 | if id.info != None: |
| 1779 | output.write(" <variable name='%s' file='%s' type='%s'/>\n" % ( |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 1780 | name, self.modulename_file(id.header), id.info)) |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1781 | else: |
| 1782 | output.write(" <variable name='%s' file='%s'/>\n" % ( |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 1783 | name, self.modulename_file(id.header))) |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 1784 | |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1785 | def serialize_function(self, output, name): |
| 1786 | id = self.idx.functions[name] |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 1787 | if name == debugsym: |
| 1788 | print "=>", id |
| 1789 | |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 1790 | output.write(" <%s name='%s' file='%s' module='%s'>\n" % (id.type, |
| 1791 | name, self.modulename_file(id.header), |
| 1792 | self.modulename_file(id.module))) |
William M. Brack | 21e4ef2 | 2005-01-02 09:53:13 +0000 | [diff] [blame] | 1793 | # |
| 1794 | # Processing of conditionals modified by Bill 1/1/05 |
| 1795 | # |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 1796 | if id.conditionals != None: |
William M. Brack | 21e4ef2 | 2005-01-02 09:53:13 +0000 | [diff] [blame] | 1797 | apstr = "" |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 1798 | for cond in id.conditionals: |
William M. Brack | 21e4ef2 | 2005-01-02 09:53:13 +0000 | [diff] [blame] | 1799 | if apstr != "": |
| 1800 | apstr = apstr + " && " |
| 1801 | apstr = apstr + cond |
| 1802 | output.write(" <cond>%s</cond>\n"% (apstr)); |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1803 | try: |
| 1804 | (ret, params, desc) = id.info |
Daniel Veillard | dee2348 | 2008-04-11 12:58:43 +0000 | [diff] [blame] | 1805 | if (desc == None or desc == '') and \ |
| 1806 | name[0:9] != "xmlThrDef" and name != "xmlDllMain": |
| 1807 | print "%s %s from %s has no description" % (id.type, name, |
| 1808 | self.modulename_file(id.module)) |
| 1809 | |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1810 | output.write(" <info>%s</info>\n" % (escape(desc))) |
| 1811 | self.indexString(name, desc) |
| 1812 | if ret[0] != None: |
| 1813 | if ret[0] == "void": |
| 1814 | output.write(" <return type='void'/>\n") |
| 1815 | else: |
| 1816 | output.write(" <return type='%s' info='%s'/>\n" % ( |
| 1817 | ret[0], escape(ret[1]))) |
| 1818 | self.indexString(name, ret[1]) |
| 1819 | for param in params: |
| 1820 | if param[0] == 'void': |
| 1821 | continue |
| 1822 | if param[2] == None: |
| 1823 | output.write(" <arg name='%s' type='%s' info=''/>\n" % (param[1], param[0])) |
| 1824 | else: |
| 1825 | output.write(" <arg name='%s' type='%s' info='%s'/>\n" % (param[1], param[0], escape(param[2]))) |
| 1826 | self.indexString(name, param[2]) |
| 1827 | except: |
| 1828 | print "Failed to save function %s info: " % name, `id.info` |
| 1829 | output.write(" </%s>\n" % (id.type)) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1830 | |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1831 | def serialize_exports(self, output, file): |
| 1832 | module = self.modulename_file(file) |
| 1833 | output.write(" <file name='%s'>\n" % (module)) |
| 1834 | dict = self.headers[file] |
Daniel Veillard | be58697 | 2003-11-18 20:56:51 +0000 | [diff] [blame] | 1835 | if dict.info != None: |
| 1836 | for data in ('Summary', 'Description', 'Author'): |
| 1837 | try: |
| 1838 | output.write(" <%s>%s</%s>\n" % ( |
| 1839 | string.lower(data), |
| 1840 | escape(dict.info[data]), |
| 1841 | string.lower(data))) |
| 1842 | except: |
| 1843 | print "Header %s lacks a %s description" % (module, data) |
| 1844 | if dict.info.has_key('Description'): |
| 1845 | desc = dict.info['Description'] |
| 1846 | if string.find(desc, "DEPRECATED") != -1: |
| 1847 | output.write(" <deprecated/>\n") |
| 1848 | |
Daniel Veillard | 1a79241 | 2003-11-18 23:52:38 +0000 | [diff] [blame] | 1849 | ids = dict.macros.keys() |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1850 | ids.sort() |
| 1851 | for id in uniq(ids): |
Daniel Veillard | 1a79241 | 2003-11-18 23:52:38 +0000 | [diff] [blame] | 1852 | # Macros are sometime used to masquerade other types. |
| 1853 | if dict.functions.has_key(id): |
| 1854 | continue |
| 1855 | if dict.variables.has_key(id): |
| 1856 | continue |
| 1857 | if dict.typedefs.has_key(id): |
| 1858 | continue |
| 1859 | if dict.structs.has_key(id): |
| 1860 | continue |
| 1861 | if dict.enums.has_key(id): |
| 1862 | continue |
| 1863 | output.write(" <exports symbol='%s' type='macro'/>\n" % (id)) |
| 1864 | ids = dict.enums.keys() |
| 1865 | ids.sort() |
| 1866 | for id in uniq(ids): |
| 1867 | output.write(" <exports symbol='%s' type='enum'/>\n" % (id)) |
| 1868 | ids = dict.typedefs.keys() |
| 1869 | ids.sort() |
| 1870 | for id in uniq(ids): |
| 1871 | output.write(" <exports symbol='%s' type='typedef'/>\n" % (id)) |
| 1872 | ids = dict.structs.keys() |
| 1873 | ids.sort() |
| 1874 | for id in uniq(ids): |
| 1875 | output.write(" <exports symbol='%s' type='struct'/>\n" % (id)) |
| 1876 | ids = dict.variables.keys() |
| 1877 | ids.sort() |
| 1878 | for id in uniq(ids): |
| 1879 | output.write(" <exports symbol='%s' type='variable'/>\n" % (id)) |
| 1880 | ids = dict.functions.keys() |
| 1881 | ids.sort() |
| 1882 | for id in uniq(ids): |
| 1883 | output.write(" <exports symbol='%s' type='function'/>\n" % (id)) |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1884 | output.write(" </file>\n") |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1885 | |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1886 | def serialize_xrefs_files(self, output): |
| 1887 | headers = self.headers.keys() |
| 1888 | headers.sort() |
| 1889 | for file in headers: |
| 1890 | module = self.modulename_file(file) |
| 1891 | output.write(" <file name='%s'>\n" % (module)) |
| 1892 | dict = self.headers[file] |
William M. Brack | a2e844a | 2004-01-06 11:52:13 +0000 | [diff] [blame] | 1893 | ids = uniq(dict.functions.keys() + dict.variables.keys() + \ |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1894 | dict.macros.keys() + dict.typedefs.keys() + \ |
William M. Brack | a2e844a | 2004-01-06 11:52:13 +0000 | [diff] [blame] | 1895 | dict.structs.keys() + dict.enums.keys()) |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1896 | ids.sort() |
William M. Brack | a2e844a | 2004-01-06 11:52:13 +0000 | [diff] [blame] | 1897 | for id in ids: |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1898 | output.write(" <ref name='%s'/>\n" % (id)) |
| 1899 | output.write(" </file>\n") |
| 1900 | pass |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 1901 | |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1902 | def serialize_xrefs_functions(self, output): |
| 1903 | funcs = {} |
| 1904 | for name in self.idx.functions.keys(): |
| 1905 | id = self.idx.functions[name] |
| 1906 | try: |
| 1907 | (ret, params, desc) = id.info |
| 1908 | for param in params: |
| 1909 | if param[0] == 'void': |
| 1910 | continue |
| 1911 | if funcs.has_key(param[0]): |
| 1912 | funcs[param[0]].append(name) |
| 1913 | else: |
| 1914 | funcs[param[0]] = [name] |
| 1915 | except: |
| 1916 | pass |
| 1917 | typ = funcs.keys() |
| 1918 | typ.sort() |
| 1919 | for type in typ: |
| 1920 | if type == '' or type == 'void' or type == "int" or \ |
| 1921 | type == "char *" or type == "const char *" : |
| 1922 | continue |
| 1923 | output.write(" <type name='%s'>\n" % (type)) |
| 1924 | ids = funcs[type] |
| 1925 | ids.sort() |
William M. Brack | cf9eadf | 2003-12-25 13:24:05 +0000 | [diff] [blame] | 1926 | pid = '' # not sure why we have dups, but get rid of them! |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1927 | for id in ids: |
William M. Brack | cf9eadf | 2003-12-25 13:24:05 +0000 | [diff] [blame] | 1928 | if id != pid: |
| 1929 | output.write(" <ref name='%s'/>\n" % (id)) |
| 1930 | pid = id |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1931 | output.write(" </type>\n") |
| 1932 | |
| 1933 | def serialize_xrefs_constructors(self, output): |
| 1934 | funcs = {} |
| 1935 | for name in self.idx.functions.keys(): |
| 1936 | id = self.idx.functions[name] |
| 1937 | try: |
| 1938 | (ret, params, desc) = id.info |
| 1939 | if ret[0] == "void": |
| 1940 | continue |
| 1941 | if funcs.has_key(ret[0]): |
| 1942 | funcs[ret[0]].append(name) |
| 1943 | else: |
| 1944 | funcs[ret[0]] = [name] |
| 1945 | except: |
| 1946 | pass |
| 1947 | typ = funcs.keys() |
| 1948 | typ.sort() |
| 1949 | for type in typ: |
| 1950 | if type == '' or type == 'void' or type == "int" or \ |
| 1951 | type == "char *" or type == "const char *" : |
| 1952 | continue |
| 1953 | output.write(" <type name='%s'>\n" % (type)) |
| 1954 | ids = funcs[type] |
William M. Brack | 09a4d0a | 2004-01-05 14:28:43 +0000 | [diff] [blame] | 1955 | ids.sort() |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1956 | for id in ids: |
| 1957 | output.write(" <ref name='%s'/>\n" % (id)) |
| 1958 | output.write(" </type>\n") |
| 1959 | |
| 1960 | def serialize_xrefs_alpha(self, output): |
| 1961 | letter = None |
| 1962 | ids = self.idx.identifiers.keys() |
| 1963 | ids.sort() |
| 1964 | for id in ids: |
| 1965 | if id[0] != letter: |
| 1966 | if letter != None: |
| 1967 | output.write(" </letter>\n") |
| 1968 | letter = id[0] |
| 1969 | output.write(" <letter name='%s'>\n" % (letter)) |
| 1970 | output.write(" <ref name='%s'/>\n" % (id)) |
| 1971 | if letter != None: |
| 1972 | output.write(" </letter>\n") |
| 1973 | |
| 1974 | def serialize_xrefs_references(self, output): |
| 1975 | typ = self.idx.identifiers.keys() |
| 1976 | typ.sort() |
| 1977 | for id in typ: |
| 1978 | idf = self.idx.identifiers[id] |
Daniel Veillard | 5d4644e | 2005-04-01 13:11:58 +0000 | [diff] [blame] | 1979 | module = idf.header |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 1980 | output.write(" <reference name='%s' href='%s'/>\n" % (id, |
| 1981 | 'html/' + self.basename + '-' + |
| 1982 | self.modulename_file(module) + '.html#' + |
| 1983 | id)) |
| 1984 | |
| 1985 | def serialize_xrefs_index(self, output): |
| 1986 | index = self.xref |
| 1987 | typ = index.keys() |
| 1988 | typ.sort() |
| 1989 | letter = None |
| 1990 | count = 0 |
| 1991 | chunk = 0 |
| 1992 | chunks = [] |
| 1993 | for id in typ: |
| 1994 | if len(index[id]) > 30: |
| 1995 | continue |
| 1996 | if id[0] != letter: |
| 1997 | if letter == None or count > 200: |
| 1998 | if letter != None: |
| 1999 | output.write(" </letter>\n") |
| 2000 | output.write(" </chunk>\n") |
| 2001 | count = 0 |
| 2002 | chunks.append(["chunk%s" % (chunk -1), first_letter, letter]) |
| 2003 | output.write(" <chunk name='chunk%s'>\n" % (chunk)) |
| 2004 | first_letter = id[0] |
| 2005 | chunk = chunk + 1 |
| 2006 | elif letter != None: |
| 2007 | output.write(" </letter>\n") |
| 2008 | letter = id[0] |
| 2009 | output.write(" <letter name='%s'>\n" % (letter)) |
| 2010 | output.write(" <word name='%s'>\n" % (id)) |
| 2011 | tokens = index[id]; |
| 2012 | tokens.sort() |
| 2013 | tok = None |
William M. Brack | a2e844a | 2004-01-06 11:52:13 +0000 | [diff] [blame] | 2014 | for token in tokens: |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 2015 | if tok == token: |
| 2016 | continue |
| 2017 | tok = token |
| 2018 | output.write(" <ref name='%s'/>\n" % (token)) |
| 2019 | count = count + 1 |
| 2020 | output.write(" </word>\n") |
| 2021 | if letter != None: |
| 2022 | output.write(" </letter>\n") |
| 2023 | output.write(" </chunk>\n") |
William M. Brack | 966668a | 2003-12-20 02:10:28 +0000 | [diff] [blame] | 2024 | if count != 0: |
| 2025 | chunks.append(["chunk%s" % (chunk -1), first_letter, letter]) |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 2026 | output.write(" <chunks>\n") |
| 2027 | for ch in chunks: |
| 2028 | output.write(" <chunk name='%s' start='%s' end='%s'/>\n" % ( |
| 2029 | ch[0], ch[1], ch[2])) |
| 2030 | output.write(" </chunks>\n") |
| 2031 | |
| 2032 | def serialize_xrefs(self, output): |
| 2033 | output.write(" <references>\n") |
| 2034 | self.serialize_xrefs_references(output) |
| 2035 | output.write(" </references>\n") |
| 2036 | output.write(" <alpha>\n") |
| 2037 | self.serialize_xrefs_alpha(output) |
| 2038 | output.write(" </alpha>\n") |
| 2039 | output.write(" <constructors>\n") |
| 2040 | self.serialize_xrefs_constructors(output) |
| 2041 | output.write(" </constructors>\n") |
| 2042 | output.write(" <functions>\n") |
| 2043 | self.serialize_xrefs_functions(output) |
| 2044 | output.write(" </functions>\n") |
| 2045 | output.write(" <files>\n") |
| 2046 | self.serialize_xrefs_files(output) |
| 2047 | output.write(" </files>\n") |
| 2048 | output.write(" <index>\n") |
| 2049 | self.serialize_xrefs_index(output) |
| 2050 | output.write(" </index>\n") |
| 2051 | |
| 2052 | def serialize(self): |
| 2053 | filename = "%s-api.xml" % self.name |
| 2054 | print "Saving XML description %s" % (filename) |
| 2055 | output = open(filename, "w") |
| 2056 | output.write('<?xml version="1.0" encoding="ISO-8859-1"?>\n') |
| 2057 | output.write("<api name='%s'>\n" % self.name) |
| 2058 | output.write(" <files>\n") |
| 2059 | headers = self.headers.keys() |
| 2060 | headers.sort() |
| 2061 | for file in headers: |
| 2062 | self.serialize_exports(output, file) |
| 2063 | output.write(" </files>\n") |
| 2064 | output.write(" <symbols>\n") |
| 2065 | macros = self.idx.macros.keys() |
| 2066 | macros.sort() |
| 2067 | for macro in macros: |
| 2068 | self.serialize_macro(output, macro) |
| 2069 | enums = self.idx.enums.keys() |
| 2070 | enums.sort() |
| 2071 | for enum in enums: |
| 2072 | self.serialize_enum(output, enum) |
| 2073 | typedefs = self.idx.typedefs.keys() |
| 2074 | typedefs.sort() |
| 2075 | for typedef in typedefs: |
| 2076 | self.serialize_typedef(output, typedef) |
| 2077 | variables = self.idx.variables.keys() |
| 2078 | variables.sort() |
| 2079 | for variable in variables: |
| 2080 | self.serialize_variable(output, variable) |
| 2081 | functions = self.idx.functions.keys() |
| 2082 | functions.sort() |
| 2083 | for function in functions: |
| 2084 | self.serialize_function(output, function) |
| 2085 | output.write(" </symbols>\n") |
| 2086 | output.write("</api>\n") |
| 2087 | output.close() |
| 2088 | |
| 2089 | filename = "%s-refs.xml" % self.name |
| 2090 | print "Saving XML Cross References %s" % (filename) |
| 2091 | output = open(filename, "w") |
| 2092 | output.write('<?xml version="1.0" encoding="ISO-8859-1"?>\n') |
| 2093 | output.write("<apirefs name='%s'>\n" % self.name) |
| 2094 | self.serialize_xrefs(output) |
| 2095 | output.write("</apirefs>\n") |
| 2096 | output.close() |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 2097 | |
| 2098 | |
| 2099 | def rebuild(): |
| 2100 | builder = None |
Daniel Veillard | e8ba84e | 2003-11-18 13:54:15 +0000 | [diff] [blame] | 2101 | if glob.glob("parser.c") != [] : |
| 2102 | print "Rebuilding API description for libxml2" |
| 2103 | builder = docBuilder("libxml2", [".", "."], |
| 2104 | ["xmlwin32version.h", "tst.c"]) |
| 2105 | elif glob.glob("../parser.c") != [] : |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 2106 | print "Rebuilding API description for libxml2" |
| 2107 | builder = docBuilder("libxml2", ["..", "../include/libxml"], |
Daniel Veillard | d433046 | 2003-04-29 12:40:16 +0000 | [diff] [blame] | 2108 | ["xmlwin32version.h", "tst.c"]) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 2109 | elif glob.glob("../libxslt/transform.c") != [] : |
| 2110 | print "Rebuilding API description for libxslt" |
| 2111 | builder = docBuilder("libxslt", ["../libxslt"], |
Daniel Veillard | 024b570 | 2002-12-12 00:15:55 +0000 | [diff] [blame] | 2112 | ["win32config.h", "libxslt.h", "tst.c"]) |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 2113 | else: |
| 2114 | print "rebuild() failed, unable to guess the module" |
| 2115 | return None |
| 2116 | builder.scan() |
| 2117 | builder.analyze() |
| 2118 | builder.serialize() |
Daniel Veillard | 024b570 | 2002-12-12 00:15:55 +0000 | [diff] [blame] | 2119 | if glob.glob("../libexslt/exslt.c") != [] : |
| 2120 | extra = docBuilder("libexslt", ["../libexslt"], ["libexslt.h"]) |
| 2121 | extra.scan() |
| 2122 | extra.analyze() |
| 2123 | extra.serialize() |
Daniel Veillard | a9b66d0 | 2002-12-11 14:23:49 +0000 | [diff] [blame] | 2124 | return builder |
| 2125 | |
| 2126 | # |
| 2127 | # for debugging the parser |
| 2128 | # |
| 2129 | def parse(filename): |
| 2130 | parser = CParser(filename) |
| 2131 | idx = parser.parse() |
| 2132 | return idx |
| 2133 | |
| 2134 | if __name__ == "__main__": |
Daniel Veillard | a235132 | 2004-06-27 12:08:10 +0000 | [diff] [blame] | 2135 | if len(sys.argv) > 1: |
| 2136 | debug = 1 |
| 2137 | parse(sys.argv[1]) |
| 2138 | else: |
| 2139 | rebuild() |