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();
}
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();
}
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 assignEventsAndSequences(Event *parentEvent, stTree *tree,
stSet *outgroupNameSet,
char *argv[], int64_t *j) {
Event *myEvent = NULL; // To distinguish from the global "event" variable.
assert(tree != NULL);
totalEventNumber++;
if (stTree_getChildNumber(tree) > 0) {
myEvent = event_construct3(stTree_getLabel(tree),
stTree_getBranchLength(tree), parentEvent,
eventTree);
for (int64_t i = 0; i < stTree_getChildNumber(tree); i++) {
assignEventsAndSequences(myEvent, stTree_getChild(tree, i),
outgroupNameSet, argv, j);
}
}
if (stTree_getChildNumber(tree) == 0 || (stTree_getLabel(tree) != NULL && (stSet_search(outgroupNameSet, (char *)stTree_getLabel(tree)) != NULL))) {
// This event is a leaf and/or an outgroup, so it has
// associated sequence.
assert(stTree_getLabel(tree) != NULL);
assert(stTree_getBranchLength(tree) != INFINITY);
if (stTree_getChildNumber(tree) == 0) {
// Construct the leaf event
myEvent = event_construct3(stTree_getLabel(tree), stTree_getBranchLength(tree), parentEvent, eventTree);
}
char *fileName = argv[*j];
if (!stFile_exists(fileName)) {
st_errAbort("File does not exist: %s\n", fileName);
}
// Set the global "event" variable, which is needed for the
// function provided to fastaReadToFunction.
event = myEvent;
if (stFile_isDir(fileName)) {
st_logInfo("Processing directory: %s\n", fileName);
stList *filesInDir = stFile_getFileNamesInDirectory(fileName);
for (int64_t i = 0; i < stList_length(filesInDir); i++) {
char *absChildFileName = stFile_pathJoin(fileName, stList_get(filesInDir, i));
assert(stFile_exists(absChildFileName));
setCompleteStatus(absChildFileName); //decide if the sequences in the file should be free or attached.
FILE *fileHandle = fopen(absChildFileName, "r");
fastaReadToFunction(fileHandle, processSequence);
fclose(fileHandle);
free(absChildFileName);
}
stList_destruct(filesInDir);
} else {
st_logInfo("Processing file: %s\n", fileName);
setCompleteStatus(fileName); //decide if the sequences in the file should be free or attached.
FILE *fileHandle = fopen(fileName, "r");
fastaReadToFunction(fileHandle, processSequence);
fclose(fileHandle);
}
(*j)++;
}
}
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();
}
stNaiveConnectedComponent *stNaiveConnectivity_getConnectedComponent(stNaiveConnectivity
*connectivity, void *node) {
computeConnectedComponents(connectivity);
stNaiveConnectedComponent *curComponent = connectivity->connectedComponentCache;
while (curComponent != NULL) {
if (stSet_search(curComponent->nodes, node)) {
break;
}
curComponent = curComponent->next;
}
assert(curComponent != NULL);
return curComponent;
}
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;
}
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_search(CuTest* testCase) {
testSetup();
stIntTuple *i = stIntTuple_construct1( 0);
stIntTuple *j = stIntTuple_construct2(10, 0);
stIntTuple *k = stIntTuple_construct1( 5);
//Check search by memory address
CuAssertTrue(testCase, stSet_search(set0, one) == one);
CuAssertTrue(testCase, stSet_search(set0, two) == two);
CuAssertTrue(testCase, stSet_search(set0, three) == three);
CuAssertTrue(testCase, stSet_search(set0, four) == four);
CuAssertTrue(testCase, stSet_search(set0, five) == five);
CuAssertTrue(testCase, stSet_search(set0, six) == six);
//Check not present
CuAssertTrue(testCase, stSet_search(set0, i) == NULL);
CuAssertTrue(testCase, stSet_search(set0, j) == NULL);
CuAssertTrue(testCase, stSet_search(set0, k) == NULL);
//Check search by memory address
CuAssertTrue(testCase, stSet_search(set1, one) == one);
CuAssertTrue(testCase, stSet_search(set1, two) == two);
CuAssertTrue(testCase, stSet_search(set1, three) == three);
CuAssertTrue(testCase, stSet_search(set1, four) == four);
CuAssertTrue(testCase, stSet_search(set1, five) == five);
CuAssertTrue(testCase, stSet_search(set1, six) == six);
//Check not present
CuAssertTrue(testCase, stSet_search(set1, j) == NULL);
//Check is searching by memory
CuAssertTrue(testCase, stSet_search(set1, i) == one);
CuAssertTrue(testCase, stSet_search(set1, k) == six);
stIntTuple_destruct(i);
stIntTuple_destruct(j);
stIntTuple_destruct(k);
testTeardown();
}
// 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;
}
void stSet_insert(stSet *set, void *key) {
if (stSet_search(set, key) != NULL) { // This will ensure we don't end up with duplicate keys..
stSet_remove(set, key);
}
stHash_insert(set->hash, key, key);
}