Daniel Veillard | 3371ff8 | 2002-10-01 13:37:48 +0000 | [diff] [blame] | 1 | #!/usr/bin/python -u |
| 2 | # |
| 3 | # imports the API description and fills up a database with |
| 4 | # name relevance to modules, functions or web pages |
| 5 | # |
Daniel Veillard | 2c77cd7 | 2002-10-01 13:54:14 +0000 | [diff] [blame^] | 6 | # Operation needed: |
| 7 | # ================= |
| 8 | # |
| 9 | # install mysqld, the python wrappers for mysql and libxml2, start mysqld |
| 10 | # Change the root passwd of mysql: |
| 11 | # mysqladmin -u root password new_password |
| 12 | # Create the new database xmlsoft |
| 13 | # mysqladmin -p create xmlsoft |
| 14 | # Create a database user 'veillard' and give him passord access |
| 15 | # change veillard and abcde with the right user name and passwd |
| 16 | # mysql -p |
| 17 | # password: |
| 18 | # mysql> GRANT ALL PRIVILEGES ON xmlsoft TO veillard@localhost |
| 19 | # IDENTIFIED BY 'abcde' WITH GRANT OPTION; |
| 20 | # |
| 21 | # As the user check the access: |
| 22 | # mysql -p xmlsoft |
| 23 | # Enter password: |
| 24 | # Welcome to the MySQL monitor.... |
| 25 | # mysql> use xmlsoft |
| 26 | # Database changed |
| 27 | # mysql> quit |
| 28 | # Bye |
| 29 | # |
| 30 | # Then run the script in the doc subdir, it will create the symbols and |
| 31 | # word tables and populate them with informations extracted from |
| 32 | # the libxml2-api.xml API description, and make then accessible read-only |
| 33 | # by nobody@loaclhost the user expected to be Apache's one |
| 34 | # |
| 35 | # On the Apache configuration, make sure you have php support enabled |
| 36 | # |
| 37 | |
Daniel Veillard | 3371ff8 | 2002-10-01 13:37:48 +0000 | [diff] [blame] | 38 | import MySQLdb |
| 39 | import libxml2 |
| 40 | import sys |
| 41 | import string |
| 42 | import os |
| 43 | |
| 44 | # |
| 45 | # The dictionnary of tables required and the SQL command needed |
| 46 | # to create them |
| 47 | # |
| 48 | TABLES={ |
| 49 | "symbols" : """CREATE TABLE symbols ( |
| 50 | name varchar(255) NOT NULL, |
| 51 | module varchar(255) NOT NULL, |
| 52 | type varchar(25) NOT NULL, |
| 53 | descr varchar(255), |
| 54 | UNIQUE KEY name (name), |
| 55 | KEY module (module))""", |
| 56 | "words" : """CREATE TABLE words ( |
| 57 | name varchar(50) NOT NULL, |
| 58 | symbol varchar(255) NOT NULL, |
| 59 | relevance int, |
| 60 | KEY name (name), |
| 61 | KEY symbol (symbol), |
| 62 | UNIQUE KEY ID (name, symbol))""", |
| 63 | } |
| 64 | |
| 65 | # |
| 66 | # The XML API description file to parse |
| 67 | # |
| 68 | API="libxml2-api.xml" |
| 69 | DB=None |
| 70 | |
| 71 | ######################################################################### |
| 72 | # # |
| 73 | # MySQL database interfaces # |
| 74 | # # |
| 75 | ######################################################################### |
| 76 | def createTable(db, name): |
| 77 | global TABLES |
| 78 | |
| 79 | if db == None: |
| 80 | return -1 |
| 81 | if name == None: |
| 82 | return -1 |
| 83 | c = db.cursor() |
| 84 | |
| 85 | ret = c.execute("DROP TABLE IF EXISTS %s" % (name)) |
| 86 | if ret == 1: |
| 87 | print "Removed table %s" % (name) |
| 88 | print "Creating table %s" % (name) |
| 89 | try: |
| 90 | ret = c.execute(TABLES[name]) |
| 91 | except: |
| 92 | print "Failed to create table %s" % (name) |
| 93 | return -1 |
| 94 | return ret |
| 95 | |
| 96 | def checkTables(db): |
| 97 | global TABLES |
| 98 | |
| 99 | if db == None: |
| 100 | return -1 |
| 101 | c = db.cursor() |
| 102 | nbtables = c.execute("show tables") |
| 103 | print "Found %d tables" % (nbtables) |
| 104 | tables = {} |
| 105 | i = 0 |
| 106 | while i < nbtables: |
| 107 | l = c.fetchone() |
| 108 | name = l[0] |
| 109 | tables[name] = {} |
| 110 | i = i + 1 |
| 111 | |
| 112 | for table in TABLES.keys(): |
| 113 | if not tables.has_key(table): |
| 114 | print "table %s missing" % (table) |
| 115 | createTable(db, table) |
| 116 | print "checkTables finished" |
| 117 | |
| 118 | # make sure apache can access the tables read-only |
| 119 | try: |
| 120 | ret = c.execute("GRANT SELECT ON xmlsoft.* TO nobody@localhost") |
| 121 | except: |
| 122 | pass |
| 123 | return 0 |
| 124 | |
| 125 | def openMySQL(db="xmlsoft", passwd=None): |
| 126 | global DB |
| 127 | |
| 128 | if passwd == None: |
| 129 | passwd = "ducon" |
| 130 | DB = MySQLdb.connect(passwd=passwd, db=db) |
| 131 | if DB == None: |
| 132 | return -1 |
| 133 | ret = checkTables(DB) |
| 134 | return ret |
| 135 | |
| 136 | def updateWord(name, symbol, relevance): |
| 137 | global DB |
| 138 | |
| 139 | if DB == None: |
| 140 | openMySQL() |
| 141 | if DB == None: |
| 142 | return -1 |
| 143 | if name == None: |
| 144 | return -1 |
| 145 | if symbol == None: |
| 146 | return -1 |
| 147 | |
| 148 | c = DB.cursor() |
| 149 | try: |
| 150 | ret = c.execute( |
| 151 | """INSERT INTO words (name, symbol, relevance) VALUES ('%s','%s', %d)""" % |
| 152 | (name, symbol, relevance)) |
| 153 | except: |
| 154 | try: |
| 155 | ret = c.execute( |
| 156 | """UPDATE words SET relevance = %d where name = '%s' and symbol = '%s'""" % |
| 157 | (relevance, name, symbol)) |
| 158 | except: |
| 159 | print "Update word (%s, %s, %s) failed command" % (name, symbol, relevance) |
| 160 | print "UPDATE words SET relevance = %d where name = '%s' and symbol = '%s'" % (relevance, name, symbol) |
| 161 | print sys.exc_type, sys.exc_value |
| 162 | return -1 |
| 163 | |
| 164 | return ret |
| 165 | |
| 166 | def updateSymbol(name, module, type, desc): |
| 167 | global DB |
| 168 | |
| 169 | updateWord(name, name, 50) |
| 170 | if DB == None: |
| 171 | openMySQL() |
| 172 | if DB == None: |
| 173 | return -1 |
| 174 | if name == None: |
| 175 | return -1 |
| 176 | if module == None: |
| 177 | return -1 |
| 178 | if type == None: |
| 179 | return -1 |
| 180 | |
| 181 | try: |
| 182 | desc = string.replace(desc, "'", " ") |
| 183 | l = string.split(desc, ".") |
| 184 | desc = l[0] |
| 185 | desc = desc[0:99] |
| 186 | except: |
| 187 | desc = "" |
| 188 | |
| 189 | c = DB.cursor() |
| 190 | try: |
| 191 | ret = c.execute( |
| 192 | """INSERT INTO symbols (name, module, type, descr) VALUES ('%s','%s', '%s', '%s')""" % |
| 193 | (name, module, type, desc)) |
| 194 | except: |
| 195 | try: |
| 196 | ret = c.execute( |
| 197 | """UPDATE symbols SET module='%s', type='%s', descr='%s' where name='%s'""" % |
| 198 | (module, type, desc, name)) |
| 199 | except: |
| 200 | print "Update symbol (%s, %s, %s) failed command" % (name, module, type) |
| 201 | print """UPDATE symbols SET module='%s', type='%s', descr='%s' where name='%s'""" % (module, type, desc, name) |
| 202 | print sys.exc_type, sys.exc_value |
| 203 | return -1 |
| 204 | |
| 205 | return ret |
| 206 | |
| 207 | def addFunction(name, module, desc = ""): |
| 208 | return updateSymbol(name, module, 'function', desc) |
| 209 | |
| 210 | def addMacro(name, module, desc = ""): |
| 211 | return updateSymbol(name, module, 'macro', desc) |
| 212 | |
| 213 | def addEnum(name, module, desc = ""): |
| 214 | return updateSymbol(name, module, 'enum', desc) |
| 215 | |
| 216 | def addStruct(name, module, desc = ""): |
| 217 | return updateSymbol(name, module, 'struct', desc) |
| 218 | |
| 219 | def addConst(name, module, desc = ""): |
| 220 | return updateSymbol(name, module, 'const', desc) |
| 221 | |
| 222 | def addType(name, module, desc = ""): |
| 223 | return updateSymbol(name, module, 'type', desc) |
| 224 | |
| 225 | def addFunctype(name, module, desc = ""): |
| 226 | return updateSymbol(name, module, 'functype', desc) |
| 227 | |
| 228 | ######################################################################### |
| 229 | # # |
| 230 | # Word dictionnary and analysis routines # |
| 231 | # # |
| 232 | ######################################################################### |
| 233 | |
| 234 | wordsDict = {} |
| 235 | |
| 236 | def splitIdentifier(str): |
| 237 | ret = [] |
| 238 | while str != "": |
| 239 | cur = string.lower(str[0]) |
| 240 | str = str[1:] |
| 241 | if ((cur < 'a') or (cur > 'z')): |
| 242 | continue |
| 243 | while (str != "") and (str[0] >= 'A') and (str[0] <= 'Z'): |
| 244 | cur = cur + string.lower(str[0]) |
| 245 | str = str[1:] |
| 246 | while (str != "") and (str[0] >= 'a') and (str[0] <= 'z'): |
| 247 | cur = cur + str[0] |
| 248 | str = str[1:] |
| 249 | while (str != "") and (str[0] >= '0') and (str[0] <= '9'): |
| 250 | str = str[1:] |
| 251 | ret.append(cur) |
| 252 | return ret |
| 253 | |
| 254 | def addWord(word, module, symbol, relevance): |
| 255 | global wordsDict |
| 256 | |
| 257 | if word == None or len(word) < 3: |
| 258 | return -1 |
| 259 | if module == None or symbol == None: |
| 260 | return -1 |
| 261 | if wordsDict.has_key(word): |
| 262 | d = wordsDict[word] |
| 263 | if d == None: |
| 264 | return 0 |
| 265 | if len(d) > 500: |
| 266 | wordsDict[word] = None |
| 267 | return 0 |
| 268 | try: |
| 269 | relevance = relevance + d[(module, symbol)] |
| 270 | except: |
| 271 | pass |
| 272 | else: |
| 273 | wordsDict[word] = {} |
| 274 | wordsDict[word][(module, symbol)] = relevance |
| 275 | return relevance |
| 276 | |
| 277 | def addString(str, module, symbol, relevance): |
| 278 | if str == None or len(str) < 3: |
| 279 | return -1 |
| 280 | ret = 0 |
| 281 | str = string.replace(str, ".", " ") |
| 282 | str = string.replace(str, ",", " ") |
| 283 | str = string.replace(str, "'", " ") |
| 284 | str = string.replace(str, '"', " ") |
| 285 | str = string.replace(str, ";", " ") |
| 286 | str = string.replace(str, "-", " ") |
| 287 | l = string.split(str) |
| 288 | for word in l: |
| 289 | if len(word) > 2: |
| 290 | ret = ret + addWord(word, module, symbol, 5) |
| 291 | |
| 292 | return ret |
| 293 | |
| 294 | |
| 295 | ######################################################################### |
| 296 | # # |
| 297 | # XML API description analysis # |
| 298 | # # |
| 299 | ######################################################################### |
| 300 | |
| 301 | def loadAPI(filename): |
| 302 | doc = libxml2.parseFile(filename) |
| 303 | print "loaded %s" % (filename) |
| 304 | return doc |
| 305 | |
| 306 | def foundExport(file, symbol): |
| 307 | if file == None: |
| 308 | return 0 |
| 309 | if symbol == None: |
| 310 | return 0 |
| 311 | addFunction(symbol, file) |
| 312 | l = splitIdentifier(symbol) |
| 313 | for word in l: |
| 314 | addWord(word, file, symbol, 10) |
| 315 | return 1 |
| 316 | |
| 317 | def analyzeAPIFile(top): |
| 318 | count = 0 |
| 319 | name = top.prop("name") |
| 320 | cur = top.children |
| 321 | while cur != None: |
| 322 | if cur.type == 'text': |
| 323 | cur = cur.next |
| 324 | continue |
| 325 | if cur.name == "exports": |
| 326 | count = count + foundExport(name, cur.prop("symbol")) |
| 327 | else: |
| 328 | print "unexpected element %s in API doc <file name='%s'>" % (name) |
| 329 | cur = cur.next |
| 330 | return count |
| 331 | |
| 332 | def analyzeAPIFiles(top): |
| 333 | count = 0 |
| 334 | cur = top.children |
| 335 | |
| 336 | while cur != None: |
| 337 | if cur.type == 'text': |
| 338 | cur = cur.next |
| 339 | continue |
| 340 | if cur.name == "file": |
| 341 | count = count + analyzeAPIFile(cur) |
| 342 | else: |
| 343 | print "unexpected element %s in API doc <files>" % (cur.name) |
| 344 | cur = cur.next |
| 345 | return count |
| 346 | |
| 347 | def analyzeAPIEnum(top): |
| 348 | file = top.prop("file") |
| 349 | if file == None: |
| 350 | return 0 |
| 351 | symbol = top.prop("name") |
| 352 | if symbol == None: |
| 353 | return 0 |
| 354 | |
| 355 | addEnum(symbol, file) |
| 356 | l = splitIdentifier(symbol) |
| 357 | for word in l: |
| 358 | addWord(word, file, symbol, 10) |
| 359 | |
| 360 | return 1 |
| 361 | |
| 362 | def analyzeAPIConst(top): |
| 363 | file = top.prop("file") |
| 364 | if file == None: |
| 365 | return 0 |
| 366 | symbol = top.prop("name") |
| 367 | if symbol == None: |
| 368 | return 0 |
| 369 | |
| 370 | addConst(symbol, file) |
| 371 | l = splitIdentifier(symbol) |
| 372 | for word in l: |
| 373 | addWord(word, file, symbol, 10) |
| 374 | |
| 375 | return 1 |
| 376 | |
| 377 | def analyzeAPIType(top): |
| 378 | file = top.prop("file") |
| 379 | if file == None: |
| 380 | return 0 |
| 381 | symbol = top.prop("name") |
| 382 | if symbol == None: |
| 383 | return 0 |
| 384 | |
| 385 | addType(symbol, file) |
| 386 | l = splitIdentifier(symbol) |
| 387 | for word in l: |
| 388 | addWord(word, file, symbol, 10) |
| 389 | return 1 |
| 390 | |
| 391 | def analyzeAPIFunctype(top): |
| 392 | file = top.prop("file") |
| 393 | if file == None: |
| 394 | return 0 |
| 395 | symbol = top.prop("name") |
| 396 | if symbol == None: |
| 397 | return 0 |
| 398 | |
| 399 | addFunctype(symbol, file) |
| 400 | l = splitIdentifier(symbol) |
| 401 | for word in l: |
| 402 | addWord(word, file, symbol, 10) |
| 403 | return 1 |
| 404 | |
| 405 | def analyzeAPIStruct(top): |
| 406 | file = top.prop("file") |
| 407 | if file == None: |
| 408 | return 0 |
| 409 | symbol = top.prop("name") |
| 410 | if symbol == None: |
| 411 | return 0 |
| 412 | |
| 413 | addStruct(symbol, file) |
| 414 | l = splitIdentifier(symbol) |
| 415 | for word in l: |
| 416 | addWord(word, file, symbol, 10) |
| 417 | |
| 418 | info = top.prop("info") |
| 419 | if info != None: |
| 420 | l = string.split(info) |
| 421 | for word in l: |
| 422 | if len(word) > 2: |
| 423 | addWord(word, file, symbol, 5) |
| 424 | return 1 |
| 425 | |
| 426 | def analyzeAPIMacro(top): |
| 427 | file = top.prop("file") |
| 428 | if file == None: |
| 429 | return 0 |
| 430 | symbol = top.prop("name") |
| 431 | if symbol == None: |
| 432 | return 0 |
| 433 | |
| 434 | info = None |
| 435 | cur = top.children |
| 436 | while cur != None: |
| 437 | if cur.type == 'text': |
| 438 | cur = cur.next |
| 439 | continue |
| 440 | if cur.name == "info": |
| 441 | info = cur.content |
| 442 | break |
| 443 | cur = cur.next |
| 444 | |
| 445 | l = splitIdentifier(symbol) |
| 446 | for word in l: |
| 447 | addWord(word, file, symbol, 10) |
| 448 | |
| 449 | if info == None: |
| 450 | addMacro(symbol, file) |
| 451 | print "Macro %s description has no <info>" % (symbol) |
| 452 | return 0 |
| 453 | |
| 454 | addMacro(symbol, file, info) |
| 455 | l = string.split(info) |
| 456 | for word in l: |
| 457 | if len(word) > 2: |
| 458 | addWord(word, file, symbol, 5) |
| 459 | return 1 |
| 460 | |
| 461 | def analyzeAPIFunction(top): |
| 462 | file = top.prop("file") |
| 463 | if file == None: |
| 464 | return 0 |
| 465 | symbol = top.prop("name") |
| 466 | if symbol == None: |
| 467 | return 0 |
| 468 | |
| 469 | info = None |
| 470 | cur = top.children |
| 471 | while cur != None: |
| 472 | if cur.type == 'text': |
| 473 | cur = cur.next |
| 474 | continue |
| 475 | if cur.name == "info": |
| 476 | info = cur.content |
| 477 | elif cur.name == "return": |
| 478 | rinfo = cur.prop("info") |
| 479 | if rinfo != None: |
| 480 | addString(rinfo, file, symbol, 7) |
| 481 | elif cur.name == "arg": |
| 482 | ainfo = cur.prop("info") |
| 483 | if rinfo != None: |
| 484 | addString(ainfo, file, symbol, 5) |
| 485 | name = cur.prop("name") |
| 486 | if name != None: |
| 487 | addWord(name, file, symbol, 7) |
| 488 | cur = cur.next |
| 489 | if info == None: |
| 490 | print "Function %s description has no <info>" % (symbol) |
| 491 | addFunction(symbol, file, "") |
| 492 | else: |
| 493 | addFunction(symbol, file, info) |
| 494 | addString(info, file, symbol, 5) |
| 495 | |
| 496 | l = splitIdentifier(symbol) |
| 497 | for word in l: |
| 498 | addWord(word, file, symbol, 10) |
| 499 | |
| 500 | return 1 |
| 501 | |
| 502 | def analyzeAPISymbols(top): |
| 503 | count = 0 |
| 504 | cur = top.children |
| 505 | |
| 506 | while cur != None: |
| 507 | if cur.type == 'text': |
| 508 | cur = cur.next |
| 509 | continue |
| 510 | if cur.name == "macro": |
| 511 | count = count + analyzeAPIMacro(cur) |
| 512 | elif cur.name == "function": |
| 513 | count = count + analyzeAPIFunction(cur) |
| 514 | elif cur.name == "const": |
| 515 | count = count + analyzeAPIConst(cur) |
| 516 | elif cur.name == "typedef": |
| 517 | count = count + analyzeAPIType(cur) |
| 518 | elif cur.name == "struct": |
| 519 | count = count + analyzeAPIStruct(cur) |
| 520 | elif cur.name == "enum": |
| 521 | count = count + analyzeAPIEnum(cur) |
| 522 | elif cur.name == "functype": |
| 523 | count = count + analyzeAPIFunctype(cur) |
| 524 | else: |
| 525 | print "unexpected element %s in API doc <files>" % (cur.name) |
| 526 | cur = cur.next |
| 527 | return count |
| 528 | |
| 529 | def analyzeAPI(doc): |
| 530 | count = 0 |
| 531 | if doc == None: |
| 532 | return -1 |
| 533 | root = doc.getRootElement() |
| 534 | if root.name != "api": |
| 535 | print "Unexpected root name" |
| 536 | return -1 |
| 537 | cur = root.children |
| 538 | while cur != None: |
| 539 | if cur.type == 'text': |
| 540 | cur = cur.next |
| 541 | continue |
| 542 | if cur.name == "files": |
| 543 | pass |
| 544 | # count = count + analyzeAPIFiles(cur) |
| 545 | elif cur.name == "symbols": |
| 546 | count = count + analyzeAPISymbols(cur) |
| 547 | else: |
| 548 | print "unexpected element %s in API doc" % (cur.name) |
| 549 | cur = cur.next |
| 550 | return count |
| 551 | |
| 552 | ######################################################################### |
| 553 | # # |
| 554 | # Main code: open the DB, the API XML and analyze it # |
| 555 | # # |
| 556 | ######################################################################### |
| 557 | try: |
| 558 | openMySQL() |
| 559 | except: |
| 560 | print "Failed to open the database" |
| 561 | print sys.exc_type, sys.exc_value |
| 562 | sys.exit(1) |
| 563 | |
| 564 | try: |
| 565 | doc = loadAPI(API) |
| 566 | ret = analyzeAPI(doc) |
| 567 | print "Analyzed %d blocs" % (ret) |
| 568 | doc.freeDoc() |
| 569 | except: |
| 570 | print "Failed to parse and analyze %s" % (API) |
| 571 | print sys.exc_type, sys.exc_value |
| 572 | sys.exit(1) |
| 573 | |
| 574 | print "Indexed %d words" % (len(wordsDict)) |
| 575 | i = 0 |
| 576 | skipped = 0 |
| 577 | for word in wordsDict.keys(): |
| 578 | refs = wordsDict[word] |
| 579 | if refs == None: |
| 580 | skipped = skipped + 1 |
| 581 | continue; |
| 582 | for (module, symbol) in refs.keys(): |
| 583 | updateWord(word, symbol, refs[(module, symbol)]) |
| 584 | i = i + 1 |
| 585 | |
| 586 | print "Found %d associations, skipped %d words" % (i, skipped) |