static void test_stSet_insert(CuTest* testCase) {
/*
* Tests inserting already present keys.
*/
testSetup();
CuAssertTrue(testCase, stSet_search(set0, one) == one);
stSet_insert(set0, one);
CuAssertTrue(testCase, stSet_search(set0, one) == one);
stSet_insert(set0, three);
CuAssertTrue(testCase, stSet_search(set0, three) == three);
stIntTuple *seven = stIntTuple_construct2(7, 7);
CuAssertTrue(testCase, stSet_search(set0, seven) == NULL);
stSet_insert(set0, seven);
CuAssertTrue(testCase, stSet_search(set0, seven) == seven);
stIntTuple_destruct(seven);
testTeardown();
}
static void test_stSet_remove(CuTest* testCase) {
testSetup();
CuAssertTrue(testCase, stSet_remove(set0, one) == one);
CuAssertTrue(testCase, stSet_search(set0, one) == NULL);
CuAssertTrue(testCase, stSet_remove(set1, one) == one);
CuAssertTrue(testCase, stSet_search(set1, one) == NULL);
stSet_insert(set1, one);
CuAssertTrue(testCase, stSet_search(set1, one) == one);
testTeardown();
}
stSet *stSet_getUnion(stSet *set1, stSet *set2) {
stSet_verifySetsHaveSameFunctions(set1, set2);
stSet *set3 = stSet_construct3(stSet_getHashFunction(set1),
stSet_getEqualityFunction(set1),
NULL);
// Add everything
stSetIterator *sit= stSet_getIterator(set1);
void *o;
while ((o = stSet_getNext(sit)) != NULL) {
stSet_insert(set3, o);
}
stSet_destructIterator(sit);
sit = stSet_getIterator(set2);
while ((o = stSet_getNext(sit)) != NULL) {
stSet_insert(set3, o);
}
stSet_destructIterator(sit);
return set3;
}
static void test_stSet_getUnion(CuTest* testCase) {
testSetup();
// Check union of empty sets is empty
stSet *set2 = stSet_construct();
stSet *set3 = stSet_construct();
stSet *set4 = stSet_getUnion(set2, set3);
CuAssertTrue(testCase, stSet_size(set4) == 0);
stSet_destruct(set2);
stSet_destruct(set3);
stSet_destruct(set4);
// Check union of non empty set and empty set is non-empty
set2 = stSet_construct();
set3 = stSet_getUnion(set0, set2);
CuAssertTrue(testCase, stSet_size(set3) == 6);
stSet_destruct(set2);
stSet_destruct(set3);
// Check union of two non-empty overlapping sets is correct
set2 = stSet_construct();
set3 = stSet_construct();
stIntTuple **uniqs = (stIntTuple **) st_malloc(sizeof(*uniqs) * 4);
uniqs[0] = stIntTuple_construct2(9, 0);
uniqs[1] = stIntTuple_construct2(9, 1);
uniqs[2] = stIntTuple_construct2(9, 2);
uniqs[3] = stIntTuple_construct2(9, 3);
stIntTuple **common = (stIntTuple **) st_malloc(sizeof(*uniqs) * 5);
common[0] = stIntTuple_construct2(5, 0);
common[1] = stIntTuple_construct2(5, 1);
common[2] = stIntTuple_construct2(5, 2);
common[3] = stIntTuple_construct2(5, 3);
common[4] = stIntTuple_construct2(5, 4);
for (int i = 0; i < 5; ++i) {
stSet_insert(set2, common[i]);
stSet_insert(set3, common[i]);
}
stSet_insert(set2, uniqs[0]);
stSet_insert(set2, uniqs[1]);
stSet_insert(set3, uniqs[2]);
stSet_insert(set3, uniqs[3]);
set4 = stSet_getUnion(set2, set3);
CuAssertTrue(testCase, stSet_size(set4) == 9);
for (int i = 0; i < 4; ++i) {
CuAssertTrue(testCase, stSet_search(set4, uniqs[i]) != NULL);
}
for (int i = 0; i < 5; ++i) {
CuAssertTrue(testCase, stSet_search(set4, common[i]) != NULL);
}
stSet_destruct(set2);
stSet_destruct(set3);
stSet_destruct(set4);
// Check we get an exception with sets with different functions.
stTry {
stSet_getUnion(set0, set1);
} stCatch(except) {
CuAssertTrue(testCase, stExcept_getId(except) == SET_EXCEPTION_ID);
} stTryEnd
testTeardown();
}
stSet *stSet_getDifference(stSet *set1, stSet *set2) {
stSet_verifySetsHaveSameFunctions(set1, set2);
stSet *set3 = stSet_construct3(stSet_getHashFunction(set1),
stSet_getEqualityFunction(set1),
NULL);
// Add those from set1 only if they are not in set2
stSetIterator *sit= stSet_getIterator(set1);
void *o;
while ((o = stSet_getNext(sit)) != NULL) {
if (stSet_search(set2, o) == NULL) {
stSet_insert(set3, o);
}
}
stSet_destructIterator(sit);
return set3;
}
static void test_stSet_removeAndFreeKey(CuTest* testCase) {
stSet *set2 = stSet_construct2(free);
stList *keys = stList_construct();
int64_t keyNumber = 1000;
for (int64_t i = 0; i < keyNumber; i++) {
int64_t *key = st_malloc(sizeof(*key));
stList_append(keys, key);
stSet_insert(set2, key);
}
for (int64_t i = 0; i < keyNumber; i++) {
int64_t *key = stList_get(keys, i);
CuAssertPtrEquals(testCase, key, stSet_removeAndFreeKey(set2, key));
}
CuAssertIntEquals(testCase, 0, stSet_size(set2));
stSet_destruct(set2);
stList_destruct(keys);
}
stTree *stTree_getMRCA(stTree *node1, stTree *node2) {
// Find all of node 1's parents (inclusive of node 1)
stSet *parents = stSet_construct();
stTree *curNode = node1;
do {
stSet_insert(parents, curNode);
} while ((curNode = stTree_getParent(curNode)) != NULL);
// Find the first parent of node 2 that is a parent of node 1
stTree *ret = NULL;
curNode = node2;
do {
if (stSet_search(parents, curNode) != NULL) {
ret = curNode;
break;
}
} while ((curNode = stTree_getParent(curNode)) != NULL);
stSet_destruct(parents);
return ret;
}
static void test_stSet_testIterator(CuTest *testCase) {
testSetup();
stSetIterator *iterator = stSet_getIterator(set0);
stSetIterator *iteratorCopy = stSet_copyIterator(iterator);
int64_t i = 0;
stSet *seen = stSet_construct();
for (i = 0; i < 6; i++) {
void *o = stSet_getNext(iterator);
CuAssertTrue(testCase, o != NULL);
CuAssertTrue(testCase, stSet_search(set0, o) != NULL);
CuAssertTrue(testCase, stSet_search(seen, o) == NULL);
CuAssertTrue(testCase, stSet_getNext(iteratorCopy) == o);
stSet_insert(seen, o);
}
CuAssertTrue(testCase, stSet_getNext(iterator) == NULL);
CuAssertTrue(testCase, stSet_getNext(iterator) == NULL);
CuAssertTrue(testCase, stSet_getNext(iteratorCopy) == NULL);
stSet_destruct(seen);
stSet_destructIterator(iterator);
stSet_destructIterator(iteratorCopy);
testTeardown();
}
static void testSetup() {
// compare by value of memory address
set0 = stSet_construct();
// compare by value of ints.
set1 = stSet_construct3((uint64_t(*)(const void *)) stIntTuple_hashKey,
(int(*)(const void *, const void *)) stIntTuple_equalsFn,
(void(*)(void *)) stIntTuple_destruct);
one = stIntTuple_construct1( 0);
two = stIntTuple_construct1( 1);
three = stIntTuple_construct1( 2);
four = stIntTuple_construct1( 3);
five = stIntTuple_construct1( 4);
six = stIntTuple_construct1( 5);
stSet_insert(set0, one);
stSet_insert(set0, two);
stSet_insert(set0, three);
stSet_insert(set0, four);
stSet_insert(set0, five);
stSet_insert(set0, six);
stSet_insert(set1, one);
stSet_insert(set1, two);
stSet_insert(set1, three);
stSet_insert(set1, four);
stSet_insert(set1, five);
stSet_insert(set1, six);
}
// Compute the connected components, if they haven't been computed
// already since the last modification.
static void computeConnectedComponents(stNaiveConnectivity *connectivity) {
if (connectivity->connectedComponentCache != NULL) {
// Already computed the connected components.
return;
}
stHashIterator *nodeIt = stHash_getIterator(connectivity->nodesToAdjList);
void *node;
stNaiveConnectedComponent *componentsHead = NULL;
while ((node = stHash_getNext(nodeIt)) != NULL) {
stSet *myNodeSet = stSet_construct();
stSet_insert(myNodeSet, node);
struct adjacency *adjList = stHash_search(connectivity->nodesToAdjList, node);
if (adjList != NULL) {
while (adjList != NULL) {
stSet_insert(myNodeSet, adjList->toNode);
adjList = adjList->next;
}
}
// Now go through the existing connected components and see if
// this overlaps any of them. If it's not a full overlap, then
// this set becomes the union, and we continue looking for
// additional overlaps, then this becomes a new connected
// component. If we find that this is a subset of an existing
// component, we can quit early, since we can't possibly add
// to it or any others.
stNaiveConnectedComponent *curComponent = componentsHead;
while (curComponent != NULL) {
stNaiveConnectedComponent *next = curComponent->next;
// Find out whether our node set is a subset of this
// connected component, or if it shares any overlap.
bool isSubset = true;
bool overlap = false;
stSetIterator *myNodeIt = stSet_getIterator(myNodeSet);
void *node;
while ((node = stSet_getNext(myNodeIt)) != NULL) {
if (stSet_search(curComponent->nodes, node)) {
overlap = true;
} else {
isSubset = false;
}
}
stSet_destructIterator(myNodeIt);
if (isSubset) {
assert(overlap == true);
// Quit early.
stSet_destruct(myNodeSet);
myNodeSet = NULL;
break;
} else if (overlap) {
stSet *newNodeSet = stSet_getUnion(myNodeSet, curComponent->nodes);
stSet_destruct(myNodeSet);
removeComponent(&componentsHead, curComponent);
myNodeSet = newNodeSet;
}
curComponent = next;
}
if (myNodeSet != NULL) {
// We have a new (or possibly merged) connected component to
// add to the list.
stNaiveConnectedComponent *newComponent = malloc(sizeof(stNaiveConnectedComponent));
newComponent->nodes = myNodeSet;
newComponent->next = componentsHead;
componentsHead = newComponent;
}
}
stHash_destructIterator(nodeIt);
connectivity->connectedComponentCache = componentsHead;
}
int main(int argc, char *argv[]) {
/*
* Open the database.
* Construct a flower.
* Construct an event tree representing the species tree.
* For each sequence contruct two ends each containing an cap.
* Make a file for the sequence.
* Link the two caps.
* Finish!
*/
int64_t key, j;
Group *group;
Flower_EndIterator *endIterator;
End *end;
bool makeEventHeadersAlphaNumeric = 0;
/*
* Arguments/options
*/
char * logLevelString = NULL;
char * speciesTree = NULL;
char * outgroupEvents = NULL;
///////////////////////////////////////////////////////////////////////////
// (0) Parse the inputs handed by genomeCactus.py / setup stuff.
///////////////////////////////////////////////////////////////////////////
while (1) {
static struct option long_options[] = { { "logLevel", required_argument, 0, 'a' }, { "cactusDisk", required_argument, 0, 'b' }, {
"speciesTree", required_argument, 0, 'f' }, { "outgroupEvents", required_argument, 0, 'g' },
{ "help", no_argument, 0, 'h' }, { "makeEventHeadersAlphaNumeric", no_argument, 0, 'i' }, { 0, 0, 0, 0 } };
int option_index = 0;
key = getopt_long(argc, argv, "a:b:f:hg:i", long_options, &option_index);
if (key == -1) {
break;
}
switch (key) {
case 'a':
logLevelString = optarg;
break;
case 'b':
cactusDiskDatabaseString = optarg;
break;
case 'f':
speciesTree = optarg;
break;
case 'g':
outgroupEvents = optarg;
break;
case 'h':
usage();
return 0;
case 'i':
makeEventHeadersAlphaNumeric = 1;
break;
default:
usage();
return 1;
}
}
///////////////////////////////////////////////////////////////////////////
// (0) Check the inputs.
///////////////////////////////////////////////////////////////////////////
//assert(logLevelString == NULL || strcmp(logLevelString, "CRITICAL") == 0 || strcmp(logLevelString, "INFO") == 0 || strcmp(logLevelString, "DEBUG") == 0);
assert(cactusDiskDatabaseString != NULL);
assert(speciesTree != NULL);
//////////////////////////////////////////////
//Set up logging
//////////////////////////////////////////////
st_setLogLevelFromString(logLevelString);
//////////////////////////////////////////////
//Log (some of) the inputs
//////////////////////////////////////////////
st_logInfo("Flower disk name : %s\n", cactusDiskDatabaseString);
for (j = optind; j < argc; j++) {
st_logInfo("Sequence file/directory %s\n", argv[j]);
}
//////////////////////////////////////////////
//Load the database
//////////////////////////////////////////////
stKVDatabaseConf *kvDatabaseConf = kvDatabaseConf = stKVDatabaseConf_constructFromString(cactusDiskDatabaseString);
if (stKVDatabaseConf_getType(kvDatabaseConf) == stKVDatabaseTypeTokyoCabinet || stKVDatabaseConf_getType(kvDatabaseConf)
== stKVDatabaseTypeKyotoTycoon) {
assert(stKVDatabaseConf_getDir(kvDatabaseConf) != NULL);
cactusDisk = cactusDisk_construct2(kvDatabaseConf, "cactusSequences");
} else {
cactusDisk = cactusDisk_construct(kvDatabaseConf, 1);
}
st_logInfo("Set up the flower disk\n");
//////////////////////////////////////////////
//Construct the flower
//////////////////////////////////////////////
if (cactusDisk_getFlower(cactusDisk, 0) != NULL) {
cactusDisk_destruct(cactusDisk);
st_logInfo("The first flower already exists\n");
return 0;
}
flower = flower_construct2(0, cactusDisk);
assert(flower_getName(flower) == 0);
st_logInfo("Constructed the flower\n");
//////////////////////////////////////////////
//Construct the event tree
//////////////////////////////////////////////
st_logInfo("Going to build the event tree with newick string: %s\n", speciesTree);
stTree *tree = stTree_parseNewickString(speciesTree);
st_logInfo("Parsed the tree\n");
if (makeEventHeadersAlphaNumeric) {
makeEventHeadersAlphaNumericFn(tree);
}
stTree_setBranchLength(tree, INT64_MAX);
checkBranchLengthsAreDefined(tree);
eventTree = eventTree_construct2(flower); //creates the event tree and the root even
totalEventNumber = 1;
st_logInfo("Constructed the basic event tree\n");
// Construct a set of outgroup names so that ancestral outgroups
// get recognized.
stSet *outgroupNameSet = stSet_construct3(stHash_stringKey,
stHash_stringEqualKey,
free);
if(outgroupEvents != NULL) {
stList *outgroupNames = stString_split(outgroupEvents);
for(int64_t i = 0; i < stList_length(outgroupNames); i++) {
char *outgroupName = stList_get(outgroupNames, i);
stSet_insert(outgroupNameSet, stString_copy(outgroupName));
}
stList_destruct(outgroupNames);
}
//now traverse the tree
j = optind;
assignEventsAndSequences(eventTree_getRootEvent(eventTree), tree,
outgroupNameSet, argv, &j);
char *eventTreeString = eventTree_makeNewickString(eventTree);
st_logInfo(
"Constructed the initial flower with %" PRIi64 " sequences and %" PRIi64 " events with string: %s\n",
totalSequenceNumber, totalEventNumber, eventTreeString);
assert(event_getSubTreeBranchLength(eventTree_getRootEvent(eventTree)) >= 0.0);
free(eventTreeString);
//assert(0);
//////////////////////////////////////////////
//Label any outgroup events.
//////////////////////////////////////////////
if (outgroupEvents != NULL) {
stList *outgroupEventsList = stString_split(outgroupEvents);
for (int64_t i = 0; i < stList_length(outgroupEventsList); i++) {
char *outgroupEvent = makeEventHeadersAlphaNumeric ? makeAlphaNumeric(stList_get(outgroupEventsList, i)) : stString_copy(stList_get(outgroupEventsList, i));
Event *event = eventTree_getEventByHeader(eventTree, outgroupEvent);
if (event == NULL) {
st_errAbort("Got an outgroup string that does not match an event, outgroup string %s", outgroupEvent);
}
assert(!event_isOutgroup(event));
event_setOutgroupStatus(event, 1);
assert(event_isOutgroup(event));
free(outgroupEvent);
}
stList_destruct(outgroupEventsList);
}
//////////////////////////////////////////////
//Construct the terminal group.
//////////////////////////////////////////////
if (flower_getEndNumber(flower) > 0) {
group = group_construct2(flower);
endIterator = flower_getEndIterator(flower);
while ((end = flower_getNextEnd(endIterator)) != NULL) {
end_setGroup(end, group);
}
flower_destructEndIterator(endIterator);
assert(group_isLeaf(group));
// Create a one link chain if there is only one pair of attached ends..
group_constructChainForLink(group);
assert(!flower_builtBlocks(flower));
} else {
flower_setBuiltBlocks(flower, 1);
}
///////////////////////////////////////////////////////////////////////////
// Write the flower to disk.
///////////////////////////////////////////////////////////////////////////
//flower_check(flower);
cactusDisk_write(cactusDisk);
st_logInfo("Updated the flower on disk\n");
///////////////////////////////////////////////////////////////////////////
// Cleanup.
///////////////////////////////////////////////////////////////////////////
cactusDisk_destruct(cactusDisk);
return 0; //Exit without clean up is quicker, enable cleanup when doing memory leak detection.
stSet_destruct(outgroupNameSet);
stTree_destruct(tree);
stKVDatabaseConf_destruct(kvDatabaseConf);
return 0;
}
