static struct chromInfo *loadChromInfo(struct sqlConnection *conn)
{
struct chromInfo *ret = NULL;
char **row;
char query[256];
int chromCount = 0;
cInfoHash = newHash(0);
if (workChr)
sqlSafef(query, ArraySize(query), "SELECT * FROM chromInfo WHERE "
"chrom='%s' ORDER BY chrom DESC", workChr);
else
sqlSafef(query, ArraySize(query),
"SELECT * FROM chromInfo ORDER BY chrom DESC");
struct sqlResult *sr = sqlGetResult(conn, query);
while ((row = sqlNextRow(sr)) != NULL)
{
struct chromInfo *el;
struct chromInfo *ci;
AllocVar(ci);
el = chromInfoLoad(row);
ci->chrom = cloneString(el->chrom);
ci->size = el->size;
slAddHead(&ret, ci);
hashAddInt(cInfoHash, el->chrom, el->size);
++chromCount;
}
sqlFreeResult(&sr);
verbose(2,"#\tchrom count: %d\n", chromCount);
return (ret);
}
struct raToStructReader *raToStructReaderNew(char *name, int fieldCount, char **fields,
int requiredFieldCount, char **requiredFields)
/* Create a helper object for parsing an ra file into a C structure. This structure will
* contain */
{
struct raToStructReader *reader;
AllocVar(reader);
reader->name = cloneString(name);
reader->fieldCount = fieldCount;
reader->fields = fields;
reader->requiredFieldCount = requiredFieldCount;
reader->requiredFields = requiredFields;
struct hash *fieldIds = reader->fieldIds = hashNew(4);
int i;
for (i=0; i<fieldCount; ++i)
hashAddInt(fieldIds, fields[i], i);
if (requiredFieldCount > 0)
{
AllocArray(reader->requiredFieldIds, requiredFieldCount);
for (i=0; i<requiredFieldCount; ++i)
{
char *required = requiredFields[i];
struct hashEl *hel = hashLookup(fieldIds, required);
if (hel == NULL)
errAbort("Required field %s is not in field list", required);
reader->requiredFieldIds[i] = ptToInt(hel->val);
}
}
AllocArray(reader->fieldsObserved, fieldCount);
return reader;
}
void gensatFixFull(char *captionFile)
/* Fix missing captions. */
{
struct lineFile *lf = lineFileOpen(captionFile, TRUE);
char *row[2];
struct dyString *sql = dyStringNew(0);
struct sqlConnection *conn = sqlConnect(database);
struct hash *capHash = newHash(16);
while (lineFileRowTab(lf, row))
{
int captionId;
char *submitId = row[0];
char *caption = row[1];
captionId = hashIntValDefault(capHash, caption, 0);
if (captionId == 0)
{
dyStringClear(sql);
dyStringAppend(sql, "insert into caption values(default, \"");
dyStringAppend(sql, caption);
dyStringAppend(sql, "\")");
sqlUpdate(conn, sql->string);
verbose(1, "%s\n", sql->string);
captionId = sqlLastAutoId(conn);
hashAddInt(capHash, caption, captionId);
}
dyStringClear(sql);
dyStringPrintf(sql, "update imageFile set caption=%d ", captionId);
dyStringPrintf(sql, "where submissionSet=%d ", gensatId);
dyStringPrintf(sql, "and submitId = \"%s\"", submitId);
sqlUpdate(conn, sql->string);
verbose(1, "%s\n", sql->string);
}
dyStringFree(&sql);
}
void chainIndex(char *inChain, char *outIndex)
/* chainIndex - Create simple two column file index for chain. */
{
struct lineFile *lf = lineFileOpen(inChain, TRUE);
FILE *f = mustOpen(outIndex, "w");
struct chain *chain, *lastChain = NULL;
long pos = 0;
struct hash *uniqHash = hashNew(16);
while ((chain = chainRead(lf)) != NULL)
{
if (lastChain == NULL || !sameString(chain->tName, lastChain->tName))
{
if (hashLookup(uniqHash, chain->tName))
{
errAbort("%s is not sorted, %s repeated with intervening %s",
inChain, chain->tName, lastChain->tName);
}
hashAddInt(uniqHash, chain->tName, pos);
fprintf(f, "%lx\t%s\n", pos, chain->tName);
}
chainFree(&lastChain);
lastChain = chain;
pos = lineFileTell(lf);
}
}
static void idAdd(struct hash *hash, void *obj)
/* Add to object hash */
{
char buf[17];
safef(buf, sizeof(buf), "%p", obj);
hashAddInt(hash, buf, hash->elCount+1);
}
static struct hash *createIntHash(SEXP v) {
struct hash *hash = hashNew(0);
SEXP names = getAttrib(v, R_NamesSymbol);
for (int i = 0; i < length(v); i++)
hashAddInt(hash, (char *)CHAR(STRING_ELT(names, i)), INTEGER(v)[i]);
return hash;
}
static struct hash *buildSymHash(char **values, boolean isEnum)
/* build a hash of values for either enum or set symbolic column */
{
struct hash *valHash = hashNew(0);
unsigned setVal = 1; /* not used for enum */
int iVal;
for (iVal = 0; values[iVal] != NULL; iVal++)
{
if (isEnum)
hashAddInt(valHash, values[iVal], iVal);
else
{
hashAddInt(valHash, values[iVal], setVal);
setVal = setVal << 1;
}
}
return valHash;
}
struct hash *initCutterCountHash(struct cutter *cutters)
/* Return one of those hashes (keyed on enz name) of ints all set to zero. */
{
struct hash *countHash = newHash(12);
struct cutter *cutter;
for (cutter = cutters; cutter != NULL; cutter = cutter->next)
hashAddInt(countHash, cutter->name, 0);
return countHash;
}
struct hash *ccdsStatusValLoad(struct sqlConnection *conn)
/* load values from the imported ccdsStatusVals table. Table hashes
* status name to uid. Names are loaded both as-is and lower-case */
{
struct hash *statusVals = hashNew(0);
struct sqlResult *sr = sqlGetResult(conn,
"NOSQLINJ SELECT ccds_status_val_uid, ccds_status FROM CcdsStatusVals");
char **row;
while ((row = sqlNextRow(sr)) != NULL)
{
int uid = sqlSigned(row[0]);
char *stat = row[1];
hashAddInt(statusVals, stat, uid);
tolowers(stat);
hashAddInt(statusVals, stat, uid);
}
sqlFreeResult(&sr);
return statusVals;
}
struct hash *bbiChromSizesFromFile(char *fileName)
/* Read two column file into hash keyed by chrom. */
{
struct hash *hash = hashNew(0);
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *row[2];
while (lineFileRow(lf, row))
hashAddInt(hash, row[0], sqlUnsigned(row[1]));
return hash;
}
static struct hash* bbiChromSizes(struct bbiFile* bbi)
/* return the hash of chrom sizes from the bigBed/bigWig */
{
struct bbiChromInfo* cList = bbiChromList(bbi);
struct bbiChromInfo* c;
struct hash* cHash = newHash(10);
for (c = cList; c != NULL; c = c->next)
hashAddInt(cHash, c->name, (int)c->size);
bbiChromInfoFreeList(&cList);
return cHash;
}
static struct hash *loadSizes(char *szFile)
/* load sizes into a hash */
{
struct hash *sizes = hashNew(0);
struct lineFile *lf = lineFileOpen(szFile, TRUE);
char *row[2];
while (lineFileNextRowTab(lf, row, 2))
hashAddInt(sizes, row[0], sqlSigned(row[1]));
return sizes;
}
struct hash *readCsizeHash(char *filename)
/* read in a chrom sizes file */
{
struct lineFile *lf = lineFileOpen(filename, TRUE);
struct hash *cHash = hashNew(10);
char *words[2];
while (lineFileRowTab(lf, words))
hashAddInt(cHash, words[0], sqlSigned(words[1]));
lineFileClose(&lf);
return cHash;
}
static struct hash *loadChromSizes(char *chromSizesFile)
/* read the chromosome sizes file. */
{
struct lineFile *lf = lineFileOpen(chromSizesFile, TRUE);
struct hash *sizes = newHash(12);
char *words[2];
while (lineFileRow(lf, words))
hashAddInt(sizes, words[0], sqlSigned(words[1]));
lineFileClose(&lf);
return sizes;
}
struct hash *loadIntHash(char *fileName)
/* Load up file with lines of name<space>size into hash. */
{
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *row[2];
struct hash *hash = hashNew(0);
while (lineFileRow(lf, row))
hashAddInt(hash, row[0], lineFileNeedNum(lf, row, 1));
return hash;
}
struct hash *hashNameIntFile(char *fileName)
/* Given a two column file (name, integer value) return a
* hash keyed by name with integer values */
{
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *row[2];
struct hash *hash = hashNew(16);
while (lineFileRow(lf, row))
hashAddInt(hash, row[0], lineFileNeedNum(lf, row, 1));
lineFileClose(&lf);
return hash;
}
void annoFormatTabSetColumnVis(struct annoFormatter *vSelf, char *sourceName, char *colName,
boolean enabled)
/* Explicitly include or exclude column in output. sourceName must be the same
* as the corresponding annoStreamer source's name. */
{
struct annoFormatTab *self = (struct annoFormatTab *)vSelf;
if (! self->columnVis)
self->columnVis = hashNew(0);
char fullName[PATH_LEN];
makeFullColumnName(fullName, sizeof(fullName), sourceName, colName);
hashAddInt(self->columnVis, fullName, enabled);
}
struct hash *getTotTagsHash(char *libsFile)
/* Read in the library file and hash up the total tags. */
{
struct hash *totTagsHash = newHash(9);
struct cgapSageLib *libs = cgapSageLibLoadAllByTab(libsFile);
struct cgapSageLib *lib;
for (lib = libs; lib != NULL; lib = lib->next)
{
char buf[16];
safef(buf, sizeof(buf), "%d", lib->libId);
hashAddInt(totTagsHash, buf, (int)lib->totalTags);
}
return totTagsHash;
}
void hashIncInt(struct hash *hash, char *name)
/* Increment integer value in hash */
{
struct hashEl *hel = hashLookup(hash, name);
if (hel == NULL)
{
hashAddInt(hash, name, 1);
}
else
{
hel->val = ((char *)hel->val)+1;
/* The much simpler ++hel->val works for gnu C, but really adding one to a void pointer
* I think is not well defined. */
}
}
struct hash *getOrderHash(char *file)
/* Parse species order file into a hash. */
{
struct lineFile *lf = lineFileOpen(file, TRUE);
char *row[1];
int count = 0;
struct hash *hash = newHash(0);
while (lineFileRow(lf, row))
{
hashAddInt(hash, row[0], count);
count++;
}
lineFileClose(&lf);
return hash;
}
void initKillList()
/* Load up a hash of the accessions to avoid. */
{
struct lineFile *lf = NULL;
char *killFile = optionVal("killList", NULL);
char *words[1];
assert(killFile);
killHash = newHash(10);
lf = lineFileOpen(killFile, TRUE);
while(lineFileNextRow(lf, words, ArraySize(words)))
{
hashAddInt(killHash, words[0], 1);
}
lineFileClose(&lf);
}
void fillInExpHash(char *expFileName, struct hash **expHash, struct slName **expNames, int *expCount)
/** Read all of the names from the expFileName and store them in a hash. */
{
struct lineFile *lf = lineFileOpen(expFileName, TRUE);
char *line = NULL;
int lineSize = 0;
struct slName *name = NULL;
*expCount = 0;
*expHash = newHash(5);
while(lineFileNextReal(lf, &line))
{
hashAddInt(*expHash, line, (*expCount)++);
name = newSlName(line);
slAddHead(expNames, name);
}
slReverse(expNames);
}
struct hash *qSizeHash(char *chainfile)
/* read the chain file and figure out what the chromosome sizes are on the query end */
{
struct lineFile *lf = lineFileOpen(chainfile, TRUE);
struct chain *ch;
struct hash *csizes = hashNew(10);
while ((ch = chainRead(lf)) != NULL)
{
char *chrom = ch->qName;
int size = ch->qSize;
if (!hashLookup(csizes, chrom))
hashAddInt(csizes, chrom, size);
chainFree(&ch);
}
lineFileClose(&lf);
return csizes;
}
static void buildChromSizes(char *db)
/* build table of chromosome sizes and list of chromosomes */
{
struct sqlConnection *conn = hAllocConn(db);
struct sqlResult *sr;
char **row;
gChromSizes = hashNew(8);
sr = sqlGetResult(conn, NOSQLINJ "SELECT chrom,size FROM chromInfo");
while ((row = sqlNextRow(sr)) != NULL)
{
unsigned sz = gbParseUnsigned(NULL, row[1]);
hashAddInt(gChromSizes, row[0], sz);
slSafeAddHead(&gChroms, newSlName(row[0]));
}
sqlFreeResult(&sr);
hFreeConn(&conn);
}
static char *findUniqueName(struct hash *dupeHash, char *root)
/* If root name is already in hash, return root_1, root_2
* or something like that. */
{
struct hashEl *hel;
if ((hel = hashLookup(dupeHash, root)) == NULL)
{
hashAddInt(dupeHash, root, 1);
return root;
}
else
{
static char buf[256];
int val = ptToInt(hel->val) + 1;
hel->val = intToPt(val);
safef(buf, sizeof(buf), "%s_%d", root, val);
return buf;
}
}
void consolidateTheCounts(char *inputFile, char *outputFile)
/* Read the cat'ed file in, and either make a new hash item for each enzyme */
/* encountered on each line, or add to an existing one. Then output the hash. */
{
struct lineFile *lf = lineFileOpen(inputFile, TRUE);
struct hash *countHash = newHash(12);
char *words[2];
while (lineFileRow(lf, words))
{
char *name = words[0];
int count = lineFileNeedFullNum(lf, words, 1);
struct hashEl *el = hashLookup(countHash, name);
if (!el)
hashAddInt(countHash, name, count);
else
el->val = intToPt(ptToInt(el->val) + count);
}
writeHashToFile(countHash, outputFile);
freeHash(&countHash);
}
struct hash *makeExpsTable(char *database, char *expTable, char *expFile,
int *expCount)
/* Open experiment file and use it to create experiment table.
Use optional fields if present, otherwise defaults.
Return a hash of expId's, keyed by name */
{
struct lineFile *lf = lineFileOpen(expFile, TRUE);
FILE *f = hgCreateTabFile(tabDir, expTable);
int expId = 0;
char *words[6];
int wordCt;
struct hash *expHash = newHash(0);
while ((wordCt = lineFileChopNext(lf, words, ArraySize(words))))
{
char *name = words[0];
hashAddInt(expHash, name, expId);
fprintf(f, "%d\t%s\t", expId++, name);
fprintf(f, "%s\t", wordCt > 1 ? words[1] : name);
fprintf(f, "%s\t", wordCt > 2 ? words[2] : expUrl);
fprintf(f, "%s\t", wordCt > 3 ? words[3] : expRef);
fprintf(f, "%s\t", wordCt > 4 ? words[4] : expCredit);
fprintf(f, "0\n"); /* extras */
}
if (expId <= 0)
errAbort("No experiments in %s", lf->fileName);
verbose(2, "%d experiments\n", expId);
if (doLoad)
{
struct sqlConnection *conn = sqlConnect(database);
expRecordCreateTable(conn, expTable);
hgLoadTabFile(conn, tabDir, expTable, &f);
sqlDisconnect(&conn);
}
lineFileClose(&lf);
if (expCount)
*expCount = expId;
return expHash;
}
void mafSplitPos(char *database, char *size, char *outFile)
/* Pick best positions for split close to size.
* Use middle of a gap as preferred site.
* If not gaps are in range, use recent repeats (0% diverged) */
{
int splitSize = 0;
int chromSize = 0;
struct hash *chromHash;
struct hashCookie hc;
struct hashEl *hel;
struct sqlConnection *conn = sqlConnect(database);
FILE *f;
db = database;
verbose(1, "Finding split positions for %s at ~%s Mbp intervals\n",
database, size);
splitSize = sqlSigned(size) * 1000000;
if (chrom == NULL)
{
chromHash = hChromSizeHash(database);
}
else
{
chromHash = hashNew(6);
hashAddInt(chromHash, chrom, hChromSize(database, chrom));
}
conn = sqlConnect(database);
f = mustOpen(outFile, "w");
hc = hashFirst(chromHash);
while ((hel = hashNext(&hc)) != NULL)
{
chrom = hel->name;
chromSize = ptToInt(hel->val);
chromSplits(chrom, chromSize, splitSize, conn, f);
}
sqlDisconnect(&conn);
carefulClose(&f);
}
struct hash* searchForKeywords(struct sqlConnection* conn, char *articleTable, char *keywords)
/* return hash with the articleIds that contain a given keyword in the abstract/title/authors */
{
if (isEmpty(keywords))
return NULL;
char query[12000];
sqlSafef(query, sizeof(query), "SELECT articleId FROM %s WHERE "
"MATCH (citation, title, authors, abstract) AGAINST ('%s' IN BOOLEAN MODE)", articleTable, keywords);
//printf("query %s", query);
struct slName *artIds = sqlQuickList(conn, query);
if (artIds==NULL || slCount(artIds)==0)
return NULL;
// convert list to hash
struct hash *hashA = hashNew(0);
struct slName *el;
for (el = artIds; el != NULL; el = el->next)
hashAddInt(hashA, el->name, 1);
freeMem(keywords);
slFreeList(artIds);
return hashA;
}
static struct hash *loadAllGaps(struct sqlConnection *conn, char *db)
/* working on all chroms, fetch all per-chrom gap counts at once
* returns hash by chrom name to gap counts for that chrom
*/
{
struct chromInfo *cInfo;
struct sqlResult *sr;
char **row;
struct hash *ret;
int totalGapSize = 0;
int gapCount = 0;
ret = newHash(0);
/* If not split, read in whole gulp, create per-chrom hash of sizes */
if (hTableExists(db, "gap"))
{
char *prevChrom = NULL;
int totalGapsThisChrom = 0;
sr = sqlGetResult(conn,
NOSQLINJ "select chrom,chromStart,chromEnd from gap order by chrom");
while ((row = sqlNextRow(sr)) != NULL)
{
int gapSize = sqlUnsigned(row[2]) - sqlUnsigned(row[1]);
++gapCount;
if (prevChrom && sameWord(prevChrom,row[0]))
{
totalGapsThisChrom += gapSize;
totalGapSize += gapSize;
}
else
{
if (prevChrom)
{
hashAddInt(ret, prevChrom, totalGapsThisChrom);
freeMem(prevChrom);
prevChrom = cloneString(row[0]);
totalGapsThisChrom = gapSize;
totalGapSize += gapSize;
}
else
{
prevChrom = cloneString(row[0]);
totalGapsThisChrom = gapSize;
totalGapSize += gapSize;
}
}
}
/* and the last one */
if (prevChrom && (totalGapsThisChrom > 0))
{
hashAddInt(ret, prevChrom, totalGapsThisChrom);
freeMem(prevChrom);
}
sqlFreeResult(&sr);
}
else
{
/* for each chrom name, fetch the gap count */
for (cInfo = chromInfoList; cInfo != NULL; cInfo = cInfo->next)
{
int rowOffset;
int totalGapsThisChrom = 0;
sr = hChromQuery(conn, "gap", cInfo->chrom, NULL, &rowOffset);
while ((row = sqlNextRow(sr)) != NULL)
{
int gapSize;
struct agpGap gap;
++gapCount;
agpGapStaticLoad(row+rowOffset, &gap);
gapSize = gap.chromEnd - gap.chromStart;
totalGapsThisChrom += gapSize;
totalGapSize += gapSize;
}
sqlFreeResult(&sr);
hashAddInt(ret, cInfo->chrom, totalGapsThisChrom);
}
}
verbose(2,"#\tloaded %d gaps covering %d bases\n", gapCount, totalGapSize);
return ret;
}
