gentest.py

libxml,在UNIX/LINUX下非常重要的一个库,为XML相关应用提供方便.目前上载的是最新版本,若要取得最新版本,请参考里面的readme.

#!/usr/bin/python -u
#
# generate a tester program for the API
#
import sys
import os
import string
try:
    import libxml2
except:
    print "libxml2 python bindings not available, skipping testapi.c generation"
    sys.exit(0)

if len(sys.argv) > 1:
    srcPref = sys.argv[1] + '/'
else:
    srcPref = ''

#
# Modules we want to skip in API test
#
skipped_modules = [ "SAX", "xlink", "threads", "globals",
  "xmlmemory", "xmlversion", "xmlexports",
  #deprecated
  "DOCBparser",
]

#
# defines for each module
#
modules_defines = {
    "HTMLparser": "LIBXML_HTML_ENABLED",
    "catalog": "LIBXML_CATALOG_ENABLED",
    "xmlreader": "LIBXML_READER_ENABLED",
    "relaxng": "LIBXML_SCHEMAS_ENABLED",
    "schemasInternals": "LIBXML_SCHEMAS_ENABLED",
    "xmlschemas": "LIBXML_SCHEMAS_ENABLED",
    "xmlschemastypes": "LIBXML_SCHEMAS_ENABLED",
    "xpath": "LIBXML_XPATH_ENABLED",
    "xpathInternals": "LIBXML_XPATH_ENABLED",
    "xinclude": "LIBXML_XINCLUDE_ENABLED",
    "xpointer": "LIBXML_XPTR_ENABLED",
    "xmlregexp" : "LIBXML_REGEXP_ENABLED",
    "xmlautomata" : "LIBXML_AUTOMATA_ENABLED",
    "xmlsave" : "LIBXML_OUTPUT_ENABLED",
    "DOCBparser" : "LIBXML_DOCB_ENABLED",
    "xmlmodule" : "LIBXML_MODULES_ENABLED",
    "pattern" : "LIBXML_PATTERN_ENABLED",
}

#
# defines for specific functions
#
function_defines = {
    "htmlDefaultSAXHandlerInit": "LIBXML_HTML_ENABLED",
    "xmlSAX2EndElement" : "LIBXML_SAX1_ENABLED",
    "xmlSAX2StartElement" : "LIBXML_SAX1_ENABLED",
    "xmlSAXDefaultVersion" : "LIBXML_SAX1_ENABLED",
    "UTF8Toisolat1" : "LIBXML_OUTPUT_ENABLED",
    "xmlCleanupPredefinedEntities": "LIBXML_LEGACY_ENABLED",
    "xmlInitializePredefinedEntities": "LIBXML_LEGACY_ENABLED",
    "xmlSetFeature": "LIBXML_LEGACY_ENABLED",
    "xmlGetFeature": "LIBXML_LEGACY_ENABLED",
    "xmlGetFeaturesList": "LIBXML_LEGACY_ENABLED",
    "xmlIOParseDTD": "LIBXML_VALID_ENABLED",
    "xmlParseDTD": "LIBXML_VALID_ENABLED",
    "xmlParseDoc": "LIBXML_SAX1_ENABLED",
    "xmlParseMemory": "LIBXML_SAX1_ENABLED",
    "xmlRecoverDoc": "LIBXML_SAX1_ENABLED",
    "xmlParseFile": "LIBXML_SAX1_ENABLED",
    "xmlRecoverFile": "LIBXML_SAX1_ENABLED",
    "xmlRecoverMemory": "LIBXML_SAX1_ENABLED",
    "xmlSAXParseFileWithData": "LIBXML_SAX1_ENABLED",
    "xmlSAXParseMemory": "LIBXML_SAX1_ENABLED",
    "xmlSAXUserParseMemory": "LIBXML_SAX1_ENABLED",
    "xmlSAXParseDoc": "LIBXML_SAX1_ENABLED",
    "xmlSAXParseDTD": "LIBXML_SAX1_ENABLED",
    "xmlSAXUserParseFile": "LIBXML_SAX1_ENABLED",
    "xmlParseEntity": "LIBXML_SAX1_ENABLED",
    "xmlParseExternalEntity": "LIBXML_SAX1_ENABLED",
    "xmlSAXParseMemoryWithData": "LIBXML_SAX1_ENABLED",
    "xmlParseBalancedChunkMemory": "LIBXML_SAX1_ENABLED",
    "xmlParseBalancedChunkMemoryRecover": "LIBXML_SAX1_ENABLED",
    "xmlSetupParserForBuffer": "LIBXML_SAX1_ENABLED",
    "xmlStopParser": "LIBXML_PUSH_ENABLED",
    "xmlAttrSerializeTxtContent": "LIBXML_OUTPUT_ENABLED",
    "xmlSAXParseFile": "LIBXML_SAX1_ENABLED",
    "xmlSAXParseEntity": "LIBXML_SAX1_ENABLED",
    "xmlNewTextChild": "LIBXML_TREE_ENABLED",
    "xmlNewDocRawNode": "LIBXML_TREE_ENABLED",
    "xmlNewProp": "LIBXML_TREE_ENABLED",
    "xmlReconciliateNs": "LIBXML_TREE_ENABLED",
    "xmlValidateNCName": "LIBXML_TREE_ENABLED",
    "xmlValidateNMToken": "LIBXML_TREE_ENABLED",
    "xmlValidateName": "LIBXML_TREE_ENABLED",
    "xmlNewChild": "LIBXML_TREE_ENABLED",
    "xmlValidateQName": "LIBXML_TREE_ENABLED",
    "xmlSprintfElementContent": "LIBXML_OUTPUT_ENABLED",
    "xmlValidGetPotentialChildren" : "LIBXML_VALID_ENABLED",
    "xmlValidGetValidElements" : "LIBXML_VALID_ENABLED",
    "docbDefaultSAXHandlerInit" : "LIBXML_DOCB_ENABLED",
    "xmlTextReaderPreservePattern" : "LIBXML_PATTERN_ENABLED",
}

#
# Some functions really need to be skipped for the tests.
#
skipped_functions = [
# block on I/O
"xmlFdRead", "xmlReadFd", "xmlCtxtReadFd",
"htmlFdRead", "htmlReadFd", "htmlCtxtReadFd",
"xmlReaderNewFd", "xmlReaderForFd",
"xmlIORead", "xmlReadIO", "xmlCtxtReadIO",
"htmlIORead", "htmlReadIO", "htmlCtxtReadIO",
"xmlReaderNewIO", "xmlBufferDump", "xmlNanoFTPConnect",
"xmlNanoFTPConnectTo", "xmlNanoHTTPMethod", "xmlNanoHTTPMethodRedir",
# Complex I/O APIs
"xmlCreateIOParserCtxt", "xmlParserInputBufferCreateIO",
"xmlRegisterInputCallbacks", "xmlReaderForIO",
"xmlOutputBufferCreateIO", "xmlRegisterOutputCallbacks",
"xmlSaveToIO", "xmlIOHTTPOpenW",
# library state cleanup, generate false leak informations and other
# troubles, heavillyb tested otherwise.
"xmlCleanupParser", "xmlRelaxNGCleanupTypes", "xmlSetListDoc",
"xmlSetTreeDoc", "xmlUnlinkNode",
# hard to avoid leaks in the tests
"xmlStrcat", "xmlStrncat", "xmlCatalogAddLocal", "xmlNewTextWriterDoc",
"xmlXPathNewValueTree", "xmlXPathWrapString",
# unimplemented
"xmlTextReaderReadInnerXml", "xmlTextReaderReadOuterXml",
"xmlTextReaderReadString",
# destructor
"xmlListDelete", "xmlOutputBufferClose", "xmlNanoFTPClose", "xmlNanoHTTPClose",
# deprecated
"xmlCatalogGetPublic", "xmlCatalogGetSystem", "xmlEncodeEntities",
"xmlNewGlobalNs", "xmlHandleEntity", "xmlNamespaceParseNCName",
"xmlNamespaceParseNSDef", "xmlNamespaceParseQName",
"xmlParseNamespace", "xmlParseQuotedString", "xmlParserHandleReference",
"xmlScanName",
"xmlDecodeEntities", 
# allocators
"xmlMemFree",
# verbosity
"xmlCatalogSetDebug", "xmlShellPrintXPathError", "xmlShellPrintNode",
# Internal functions, no user space should really call them
"xmlParseAttribute", "xmlParseAttributeListDecl", "xmlParseName",
"xmlParseNmtoken", "xmlParseEntityValue", "xmlParseAttValue",
"xmlParseSystemLiteral", "xmlParsePubidLiteral", "xmlParseCharData",
"xmlParseExternalID", "xmlParseComment", "xmlParsePITarget", "xmlParsePI",
"xmlParseNotationDecl", "xmlParseEntityDecl", "xmlParseDefaultDecl",
"xmlParseNotationType", "xmlParseEnumerationType", "xmlParseEnumeratedType",
"xmlParseAttributeType", "xmlParseAttributeListDecl",
"xmlParseElementMixedContentDecl", "xmlParseElementChildrenContentDecl",
"xmlParseElementContentDecl", "xmlParseElementDecl", "xmlParseMarkupDecl",
"xmlParseCharRef", "xmlParseEntityRef", "xmlParseReference",
"xmlParsePEReference", "xmlParseDocTypeDecl", "xmlParseAttribute",
"xmlParseStartTag", "xmlParseEndTag", "xmlParseCDSect", "xmlParseContent",
"xmlParseElement", "xmlParseVersionNum", "xmlParseVersionInfo",
"xmlParseEncName", "xmlParseEncodingDecl", "xmlParseSDDecl",
"xmlParseXMLDecl", "xmlParseTextDecl", "xmlParseMisc",
"xmlParseExternalSubset", "xmlParserHandlePEReference",
"xmlSkipBlankChars",
]

#
# These functions have side effects on the global state
# and hence generate errors on memory allocation tests
#
skipped_memcheck = [ "xmlLoadCatalog", "xmlAddEncodingAlias",
   "xmlSchemaInitTypes", "xmlNanoFTPProxy", "xmlNanoFTPScanProxy",
   "xmlNanoHTTPScanProxy", "xmlResetLastError", "xmlCatalogConvert",
   "xmlCatalogRemove", "xmlLoadCatalogs", "xmlCleanupCharEncodingHandlers",
   "xmlInitCharEncodingHandlers", "xmlCatalogCleanup",
   "xmlSchemaGetBuiltInType",
   "htmlParseFile", "htmlCtxtReadFile" # loads the catalogs
]

#
# Extra code needed for some test cases
#
extra_pre_call = {
   "xmlSAXUserParseFile": """
#ifdef LIBXML_SAX1_ENABLED
        if (sax == (xmlSAXHandlerPtr)&xmlDefaultSAXHandler) user_data = NULL;
#endif
""",
   "xmlSAXUserParseMemory": """
#ifdef LIBXML_SAX1_ENABLED
        if (sax == (xmlSAXHandlerPtr)&xmlDefaultSAXHandler) user_data = NULL;
#endif
""",
   "xmlParseBalancedChunkMemory": """
#ifdef LIBXML_SAX1_ENABLED
        if (sax == (xmlSAXHandlerPtr)&xmlDefaultSAXHandler) user_data = NULL;
#endif
""",
   "xmlParseBalancedChunkMemoryRecover": """
#ifdef LIBXML_SAX1_ENABLED
        if (sax == (xmlSAXHandlerPtr)&xmlDefaultSAXHandler) user_data = NULL;
#endif
""",
   "xmlParserInputBufferCreateFd":
       "if (fd >= 0) fd = -1;",
}
extra_post_call = {
   "xmlAddChild": 
       "if (ret_val == NULL) { xmlFreeNode(cur) ; cur = NULL ; }",
   "xmlAddChildList": 
       "if (ret_val == NULL) { xmlFreeNodeList(cur) ; cur = NULL ; }",
   "xmlAddSibling":
       "if (ret_val == NULL) { xmlFreeNode(elem) ; elem = NULL ; }",
   "xmlAddNextSibling":
       "if (ret_val == NULL) { xmlFreeNode(elem) ; elem = NULL ; }",
   "xmlAddPrevSibling": 
       "if (ret_val == NULL) { xmlFreeNode(elem) ; elem = NULL ; }",
   "xmlDocSetRootElement": 
       "if (doc == NULL) { xmlFreeNode(root) ; root = NULL ; }",
   "xmlReplaceNode": 
       """if (cur != NULL) {
              xmlUnlinkNode(cur);
              xmlFreeNode(cur) ; cur = NULL ; }
          if (old != NULL) {
              xmlUnlinkNode(old);
              xmlFreeNode(old) ; old = NULL ; }
	  ret_val = NULL;""",
   "xmlTextMerge": 
       """if ((first != NULL) && (first->type != XML_TEXT_NODE)) {
              xmlUnlinkNode(second);
              xmlFreeNode(second) ; second = NULL ; }""",
   "xmlBuildQName": 
       """if ((ret_val != NULL) && (ret_val != ncname) &&
              (ret_val != prefix) && (ret_val != memory))
              xmlFree(ret_val);
	  ret_val = NULL;""",
   "xmlNewDocElementContent":
       """xmlFreeDocElementContent(doc, ret_val); ret_val = NULL;""",
   "xmlDictReference": "xmlDictFree(dict);",
   # Functions which deallocates one of their parameters
   "xmlXPathConvertBoolean": """val = NULL;""",
   "xmlXPathConvertNumber": """val = NULL;""",
   "xmlXPathConvertString": """val = NULL;""",
   "xmlSaveFileTo": """buf = NULL;""",
   "xmlSaveFormatFileTo": """buf = NULL;""",
   "xmlIOParseDTD": "input = NULL;",
   "xmlRemoveProp": "cur = NULL;",
   "xmlNewNs": "if ((node == NULL) && (ret_val != NULL)) xmlFreeNs(ret_val);",
   "xmlCopyNamespace": "if (ret_val != NULL) xmlFreeNs(ret_val);",
   "xmlCopyNamespaceList": "if (ret_val != NULL) xmlFreeNsList(ret_val);",
   "xmlNewTextWriter": "if (ret_val != NULL) out = NULL;",
   "xmlNewTextWriterPushParser": "if (ctxt != NULL) {xmlFreeDoc(ctxt->myDoc); ctxt->myDoc = NULL;} if (ret_val != NULL) ctxt = NULL;",
   "xmlNewIOInputStream": "if (ret_val != NULL) input = NULL;",
   "htmlParseChunk": "if (ctxt != NULL) {xmlFreeDoc(ctxt->myDoc); ctxt->myDoc = NULL;}",
   "htmlParseDocument": "if (ctxt != NULL) {xmlFreeDoc(ctxt->myDoc); ctxt->myDoc = NULL;}",
   "xmlParseDocument": "if (ctxt != NULL) {xmlFreeDoc(ctxt->myDoc); ctxt->myDoc = NULL;}",
   "xmlParseChunk": "if (ctxt != NULL) {xmlFreeDoc(ctxt->myDoc); ctxt->myDoc = NULL;}",
   "xmlParseExtParsedEnt": "if (ctxt != NULL) {xmlFreeDoc(ctxt->myDoc); ctxt->myDoc = NULL;}",
}

modules = []

def is_skipped_module(name):
    for mod in skipped_modules:
        if mod == name:
	    return 1
    return 0

def is_skipped_function(name):
    for fun in skipped_functions:
        if fun == name:
	    return 1
    # Do not test destructors
    if string.find(name, 'Free') != -1:
        return 1
    return 0

def is_skipped_memcheck(name):
    for fun in skipped_memcheck:
        if fun == name:
	    return 1
    return 0

missing_types = {}
def add_missing_type(name, func):
    try:
        list = missing_types[name]
	list.append(func)
    except:
        missing_types[name] = [func]

generated_param_types = []
def add_generated_param_type(name):
    generated_param_types.append(name)

generated_return_types = []
def add_generated_return_type(name):
    generated_return_types.append(name)

missing_functions = {}
missing_functions_nr = 0
def add_missing_functions(name, module):
    global missing_functions_nr

    missing_functions_nr = missing_functions_nr + 1
    try:
        list = missing_functions[module]
	list.append(name)
    except:
        missing_functions[module] = [name]

#
# Provide the type generators and destructors for the parameters
#

def type_convert(str, name, info, module, function, pos):
#    res = string.replace(str, "    ", " ")
#    res = string.replace(str, "   ", " ")
#    res = string.replace(str, "  ", " ")
    res = string.replace(str, " *", "_ptr")
#    res = string.replace(str, "*", "_ptr")
    res = string.replace(res, " ", "_")
    if res == 'const_char_ptr':
        if string.find(name, "file") != -1 or \
           string.find(name, "uri") != -1 or \
           string.find(name, "URI") != -1 or \
           string.find(info, "filename") != -1 or \
           string.find(info, "URI") != -1 or \
           string.find(info, "URL") != -1:
	    if string.find(function, "Save") != -1 or \
	       string.find(function, "Create") != -1 or \
	       string.find(function, "Write") != -1 or \
	       string.find(function, "Fetch") != -1:
	        return('fileoutput')
	    return('filepath')
    if res == 'void_ptr':
        if module == 'nanoftp' and name == 'ctx':
	    return('xmlNanoFTPCtxtPtr')
        if function == 'xmlNanoFTPNewCtxt' or \
	   function == 'xmlNanoFTPConnectTo' or \
	   function == 'xmlNanoFTPOpen':
	    return('xmlNanoFTPCtxtPtr')
        if module == 'nanohttp' and name == 'ctx':
	    return('xmlNanoHTTPCtxtPtr')
	if function == 'xmlNanoHTTPMethod' or \
	   function == 'xmlNanoHTTPMethodRedir' or \
	   function == 'xmlNanoHTTPOpen' or \
	   function == 'xmlNanoHTTPOpenRedir':
	    return('xmlNanoHTTPCtxtPtr');
        if function == 'xmlIOHTTPOpen':
	    return('xmlNanoHTTPCtxtPtr')
	if string.find(name, "data") != -1:
	    return('userdata')
	if string.find(name, "user") != -1:
	    return('userdata')
    if res == 'xmlDoc_ptr':
        res = 'xmlDocPtr'
    if res == 'xmlNode_ptr':
        res = 'xmlNodePtr'
    if res == 'xmlDict_ptr':
        res = 'xmlDictPtr'
    if res == 'xmlNodePtr' and pos != 0:
        if (function == 'xmlAddChild' and pos == 2) or \
	   (function == 'xmlAddChildList' and pos == 2) or \
           (function == 'xmlAddNextSibling' and pos == 2) or \
           (function == 'xmlAddSibling' and pos == 2) or \
           (function == 'xmlDocSetRootElement' and pos == 2) or \
           (function == 'xmlReplaceNode' and pos == 2) or \
           (function == 'xmlTextMerge') or \
	   (function == 'xmlAddPrevSibling' and pos == 2):
	    return('xmlNodePtr_in');
    if res == 'const xmlBufferPtr':
        res = 'xmlBufferPtr'
    if res == 'xmlChar_ptr' and name == 'name' and \
       string.find(function, "EatName") != -1:
        return('eaten_name')
    if res == 'void_ptr*':
        res = 'void_ptr_ptr'
    if res == 'char_ptr*':
        res = 'char_ptr_ptr'
    if res == 'xmlChar_ptr*':
        res = 'xmlChar_ptr_ptr'
    if res == 'const_xmlChar_ptr*':
        res = 'const_xmlChar_ptr_ptr'
    if res == 'const_char_ptr*':
        res = 'const_char_ptr_ptr'
    if res == 'FILE_ptr' and module == 'debugXML':
        res = 'debug_FILE_ptr';
    if res == 'int' and name == 'options':
        if module == 'parser' or module == 'xmlreader':
	    res = 'parseroptions'

    return res

known_param_types = []

def is_known_param_type(name, rtype):
    global test
    for type in known_param_types:
        if type == name:
	    return 1
    for type in generated_param_types:
        if type == name:
	    return 1

    if name[-3:] == 'Ptr' or name[-4:] == '_ptr':
        if rtype[0:6] == 'const ':
	    crtype = rtype[6:]
	else:
	    crtype = rtype

        define = 0
	if modules_defines.has_key(module):
	    test.write("#ifdef %s\n" % (modules_defines[module]))
	    define = 1
        test.write("""
#define gen_nb_%s 1
static %s gen_%s(int no ATTRIBUTE_UNUSED, int nr ATTRIBUTE_UNUSED) {
    return(NULL);
}
static void des_%s(int no ATTRIBUTE_UNUSED, %s val ATTRIBUTE_UNUSED, int nr ATTRIBUTE_UNUSED) {
}
""" % (name, crtype, name, name, rtype))
        if define == 1:
	    test.write("#endif\n\n")
        add_generated_param_type(name)
        return 1

    return 0

#
# Provide the type destructors for the return values
#

known_return_types = []

def is_known_return_type(name):
    for type in known_return_types:
        if type == name:
	    return 1
    return 0

#
# Copy the beginning of the C test program result
#

try:
    input = open("testapi.c", "r")
except:
    input = open(srcPref + "testapi.c", "r")
test = open('testapi.c.new', 'w')

def compare_and_save():
    global test

    test.close()
    try:
        input = open("testapi.c", "r").read()
    except:
        input = ''
    test = open('testapi.c.new', "r").read()
    if input != test:
        try:
            os.system("rm testapi.c; mv testapi.c.new testapi.c")
        except:
	    os.system("mv testapi.c.new testapi.c")
        print("Updated testapi.c")
    else:
        print("Generated testapi.c is identical")

line = input.readline()
while line != "":
    if line == "/* CUT HERE: everything below that line is generated */\n":
        break;
    if line[0:15] == "#define gen_nb_":
        type = string.split(line[15:])[0]
	known_param_types.append(type)
    if line[0:19] == "static void desret_":
        type = string.split(line[19:], '(')[0]