TEST_F(XmlDocTest, PosdbGetMetaListNewDoc) {
const char *url = "http://www.example.test/index.html";
char contentNew[] = "<html><head><title>my title</title></head><body>new document</body></html>";
XmlDoc xmlDocNew;
initializeDocForPosdb(&xmlDocNew, url, contentNew);
xmlDocNew.getMetaList(false);
auto metaListKeys = parseMetaList(xmlDocNew.m_metaList, xmlDocNew.m_metaListSize);
// make sure positive special key is there (to clear out existing negative special key)
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, POSDB_DELETEDOC_TERMID, false));
EXPECT_FALSE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, POSDB_DELETEDOC_TERMID, true));
// make sure title & body text is indexed
// title
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("title", "my"), false));
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("title", "title"), false));
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("title", "mytitle"), false));
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("my"), false));
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("title"), false));
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("mytitle"), false));
// body
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("new"), false));
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("document"), false));
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("newdocument"), false));
/// @todo ALC add other terms
}
TEST_F(XmlDocTest, PosdbGetMetaListChangedDoc) {
const char *url = "http://www.example.test/index.html";
char contentOld[] = "<html><head><title>my title</title></head><body>old document</body></html>";
char contentNew[] = "<html><head><title>my title</title></head><body>new document</body></html>";
XmlDoc *xmlDocOld = new XmlDoc();
mnew(xmlDocOld, sizeof(*xmlDocOld), "XmlDoc");
initializeDocForPosdb(xmlDocOld, url, contentOld);
XmlDoc xmlDocNew;
initializeDocForPosdb(&xmlDocNew, url, contentNew);
xmlDocNew.m_oldDocValid = true;
xmlDocNew.m_oldDoc = xmlDocOld;
xmlDocNew.getMetaList(false);
auto metaListKeys = parseMetaList(xmlDocNew.m_metaList, xmlDocNew.m_metaListSize);
// make sure no special key is inserted (positive or negative)
EXPECT_FALSE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, POSDB_DELETEDOC_TERMID, false));
EXPECT_FALSE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, POSDB_DELETEDOC_TERMID, true));
// make sure title & body text is indexed (with difference between old & new document deleted)
// title
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("title", "my"), false));
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("title", "title"), false));
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("title", "mytitle"), false));
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("my"), false));
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("title"), false));
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("mytitle"), false));
// body
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("new"), false));
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("document"), false));
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("newdocument"), false));
// deleted terms
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("old"), true));
EXPECT_TRUE(posdbFindRecord(metaListKeys, xmlDocNew.m_docId, hashWord("olddocument"), true));
/// @todo ALC add other terms
}
bool processLoop ( void *state ) {
// cast it
State8 *st = (State8 *)state;
// get the xmldoc
XmlDoc *xd = &st->m_xd;
// error?
if ( g_errno ) return sendErrorReply ( st , g_errno );
// shortcut
SafeBuf *xbuf = &st->m_xbuf;
if ( st->m_u && st->m_u[0] ) {
// . save the ips.txt file if we are the test coll
// . saveTestBuf() is a function in Msge1.cpp
CollectionRec *cr = xd->getCollRec();
if ( xd && cr && cr->m_coll && !strcmp(cr->m_coll,"qatest123"))
// use same dir that XmlDoc::getTestDir() would use
//saveTestBuf ( "test-page-parser" );
saveTestBuf("qa");
// now get the meta list, in the process it will print out a
// bunch of junk into st->m_xbuf
char *metalist = xd->getMetaList ( );
if ( ! metalist ) return sendErrorReply ( st , g_errno );
// return false if it blocked
if ( metalist == (void *)-1 ) return false;
// for debug...
if ( ! xd->m_indexCode ) xd->doConsistencyTest ( false );
// print it out
xd->printDoc( xbuf );
}
// print reason we can't analyze it (or index it)
//if ( st->m_indexCode != 0 ) {
// xbuf->safePrintf ("<br><br><b>indexCode: %s</b>\n<br>",
// mstrerror(st->m_indexCode));
//}
// we are done
g_inPageParser = false;
// print the final tail
//p += g_httpServer.printTail ( p , pend - p );
//log("parser: send sock=%li",st->m_s->m_sd);
// now encapsulate it in html head/tail and send it off
bool status = g_httpServer.sendDynamicPage( st->m_s ,
xbuf->getBufStart(),
xbuf->length() ,
-1, //cachtime
false ,//postreply?
NULL, //ctype
-1 , //httpstatus
NULL,//cookie
"utf-8");
// delete the state now
if ( st->m_freeIt ) {
mdelete ( st , sizeof(State8) , "PageParser" );
delete (st);
}
// return the status
return status;
}