static void requireStartEndLines(char *fileName, char *startLine, char *endLine)
/* Make sure first real line in file is startLine, and last is endLine. Tolerate
* empty lines and white space. */
{
char *reportFileName = fileNameOnly(fileName);
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *line;
/* Get first real line and make sure it is not empty and matches start line. */
if (!lineFileNextReal(lf, &line))
errAbort("%s is empty", reportFileName);
line = trimSpaces(line);
if (!sameString(line, startLine))
errAbort("%s doesn't start with %s as expected", reportFileName, startLine);
boolean lastSame = FALSE;
for (;;)
{
if (!lineFileNextReal(lf, &line))
break;
line = trimSpaces(line);
lastSame = sameString(line, endLine);
}
if (!lastSame)
errAbort("%s doesn't end with %s as expected", reportFileName, endLine);
lineFileClose(&lf);
}
struct wigSection *wigSectionRead(struct lineFile *lf)
/* Parse out next section of wig. */
{
static double *vals = NULL;
static int valAlloc = 0;
/* Get "fixedStep" line and parse it. */
char *line;
if (!lineFileNextReal(lf, &line))
return NULL;
char *pattern = "fixedStep ";
int patSize = 10;
if (!startsWith(pattern, line))
errAbort("Expecting fixedStep line %d of %s", lf->lineIx, lf->fileName);
line += patSize;
struct hash *varHash = hashVarLine(line, lf->lineIx);
int step = sqlUnsigned(requiredVal(lf, varHash, "step"));
int start = sqlUnsigned(requiredVal(lf, varHash, "start"));
char *chrom = cloneString(requiredVal(lf, varHash, "chrom"));
hashFree(&varHash);
/* Parse out numbers until next fixedStep. */
int valCount = 0;
int i;
for (;;)
{
if (!lineFileNextReal(lf, &line))
break;
if (startsWith(pattern, line))
{
lineFileReuse(lf);
break;
}
for (i=0; i<step; ++i)
{
if (valCount >= valAlloc)
{
int newAlloc = valAlloc + 1024;
ExpandArray(vals, valAlloc, newAlloc);
valAlloc = newAlloc;
}
vals[valCount] = lineFileNeedDouble(lf, &line, 0);
++valCount;
}
}
/* Create wigSection. */
struct wigSection *section;
AllocVar(section);
section->chrom = chrom;
section->chromStart = start;
section->chromEnd = start + valCount;
section->vals = CloneArray(vals, valCount);
return section;
}
void makeConditions(char *input, char *output)
/* Parse input in form:
* transcriptionFactor list, of, conditions
* into
* transcriptionFactor<tab>list
* transcriptionFactor<tab>of
* transcriptionFactor<tab>conditions
*/
{
struct lineFile *lf = lineFileOpen(input, TRUE);
FILE *f = mustOpen(output, "w");
char *line;
while (lineFileNextReal(lf, &line))
{
char *tf, *cond;
tf = nextWord(&line);
while ((cond = nextWord(&line)) != NULL)
{
stripChar(cond, ',');
fprintf(f, "%s\t%s\n", tf, cond);
}
}
carefulClose(&f);
lineFileClose(&lf);
}
void wordStoreLoadMonomerOrder(struct wordStore *store, char *readsFile, char *fileName)
/* Read in a file with one line for each monomer type, containing a word for each
* monomer variant. Requires all variants already be in store. The readsFile is passed
* just for nicer error reporting. */
{
/* Stuff for processing file a line at a time. */
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *line, *word;
/* Set up variables we'll put results in in store. */
store->typeHash = hashNew(0);
store->typeList = NULL;
while (lineFileNextReal(lf, &line))
{
struct wordType *type;
AllocVar(type);
slAddHead(&store->typeList, type);
while ((word = nextWord(&line)) != NULL)
{
struct wordInfo *info = hashFindVal(store->infoHash, word);
if (info == NULL)
errAbort("%s is in %s but not %s", word, lf->fileName, readsFile);
struct wordInfoRef *ref;
AllocVar(ref);
ref->val = info;
slAddHead(&type->list, ref);
hashAddUnique(store->typeHash, word, type);
}
}
slReverse(&store->typeList);
lineFileClose(&lf);
verbose(2, "Added %d types containing %d words from %s\n",
slCount(store->typeList), store->typeHash->elCount, fileName);
}
long long currentVmPeak()
/* return value of peak Vm memory usage (if /proc/ business exists) */
{
long long vmPeak = 0;
pid_t pid = getpid();
char temp[256];
safef(temp, sizeof(temp), "/proc/%d/status", (int) pid);
struct lineFile *lf = lineFileMayOpen(temp, TRUE);
if (lf)
{
char *line;
while (lineFileNextReal(lf, &line))
{ // typical line: 'VmPeak: 62646196 kB'
// seems to always be kB
if (stringIn("VmPeak", line))
{
char *words[3];
chopByWhite(line, words, 3);
vmPeak = sqlLongLong(words[1]); // assume always 2nd word
break;
}
}
lineFileClose(&lf);
}
return vmPeak;
}
boolean mgcStatusTblCopyRow(struct lineFile *inLf, FILE *outFh)
/* read a copy one row of a status table tab file without
* fully parsing. Expand if optional fields are missing */
{
char *line;
int numCols, i;
char *row[MGCSTATUS_NUM_COLS];
if (!lineFileNextReal(inLf, &line))
return FALSE;
numCols = chopTabs(line, row);
numCols = min(numCols, MGCSTATUS_NUM_COLS);
lineFileExpectAtLeast(inLf, MGCSTATUS_MIN_NUM_COLS, numCols);
for (i = 0; i < numCols; i++)
{
if (i > 0)
fputc('\t', outFh);
fputs(row[i], outFh);
}
/* pad */
for (; i < MGCSTATUS_NUM_COLS; i++)
fputc('\t', outFh);
fputc('\n', outFh);
return TRUE;
}
void setupTable(struct sqlConnection *conn)
/* Set up the autoTest table for testing. Call dropTable() later
to remove table. */
{
struct lineFile *lf = lineFileOpen("output/newTest.sql", TRUE);
char *update = NULL;
char *line = NULL;
struct dyString *ds = newDyString(256);
if(sqlTableExists(conn, testTableName))
errAbort("dbLinkTest.c::setupTable() - Table %s.%s already exists. Can't create another.",
sqlGetDatabase(conn), testTableName);
while(lineFileNextReal(lf, &line))
{
char *tmp = strstr(line, "not null");
if(tmp != NULL)
{
*tmp = ',';
tmp++;
*tmp = '\0';
}
subChar(line, '\t', ' ');
dyStringPrintf(ds, "%s", line);
}
update = replaceChars(ds->string, "PRIMARY KEY", "UNIQUE");
sqlUpdate(conn, update);
freez(&update);
dyStringFree(&ds);
lineFileClose(&lf);
}
static void readTaggedNumLine(struct lineFile *lf, char *tag,
int count, int *intOut, double *floatOut)
/* Read in a line that starts with tag and then has count numbers.
* Complain and die if tag is unexpected or other problem occurs.
* Put output as integers and/or floating point into intOut and
* floatOut. */
{
char *line;
int i = 0;
char *word;
if (!lineFileNextReal(lf, &line))
lineFileUnexpectedEnd(lf);
word = nextWord(&line);
if (!sameWord(tag, word))
errAbort("Expecting %s got %s line %d of %s",
tag, word, lf->lineIx, lf->fileName);
for (i = 0; i < count; ++i)
{
word = nextWord(&line);
if (word == NULL)
errAbort("Not enough numbers line %d of %s", lf->lineIx, lf->fileName);
if (!isdigit(word[0]))
errAbort("Expecting number got %s line %d of %s",
word, lf->lineIx, lf->fileName);
if (intOut)
intOut[i] = atoi(word);
if (floatOut)
floatOut[i] = atof(word);
}
word = nextWord(&line);
if (word != NULL)
errAbort("Too many numbers line %d of %s", lf->lineIx, lf->fileName);
}
int findBedSize(char *fileName, struct lineFile **retLf)
/* Read first line of file and figure out how many words in it. */
/* Input file could be stdin, in which case we really don't want to open,
* read, and close it here. So if retLf is non-NULL, return the open
* linefile (having told it to reuse the line we just read). */
{
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *words[64], *line;
int wordCount;
if (!lineFileNextReal(lf, &line))
if (ignoreEmpty)
return(0);
line = cloneString(line);
if (strictTab)
wordCount = chopTabs(line, words);
else
wordCount = chopLine(line, words);
if (wordCount == 0)
errAbort("%s appears to be empty", fileName);
if (retLf != NULL)
{
lineFileReuse(lf);
*retLf = lf;
}
else
lineFileClose(&lf);
freeMem(line);
return wordCount;
}
struct cnFill *cnFillRead(struct chainNet *net, struct lineFile *lf)
/* Recursively read in list and children from file. */
{
char *line;
int d, depth = 0;
struct cnFill *fillList = NULL;
struct cnFill *fill = NULL;
for (;;)
{
if (!lineFileNextReal(lf, &line))
break;
d = countLeadingChars(line, ' ');
if (fill == NULL)
depth = d;
if (d < depth)
{
lineFileReuse(lf);
break;
}
if (d > depth)
{
lineFileReuse(lf);
fill->children = cnFillRead(net, lf);
}
else
{
fill = cnFillFromLine(net->nameHash, lf, line);
slAddHead(&fillList, fill);
}
}
slReverse(&fillList);
return fillList;
}
struct hash *readDescriptionFile(char *fileName)
/* Return two column file keyed by first column with
* values in second column. Strip any tabs from values. */
{
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *line;
struct hash *hash = newHash(16);
while (lineFileNextReal(lf, &line))
{
char *key = nextWord(&line);
char *val = skipLeadingSpaces(line);
if (val != NULL && val[0] != 0)
{
char *desc = hashFindVal(hash, key);
subChar(val, '\t', ' ');
stripChar(val, '\r');
/* if gene exists in hash */
if (desc != NULL)
{
struct dyString *dy = dyStringNew(1024);
dyStringPrintf(dy, "%s ", desc);
dyStringAppend(dy, val);
val = cloneString(dy->string);
dyStringFree(&dy);
}
/* add to gene and description to hash */
hashAdd(hash, key, cloneString(val));
}
}
lineFileClose(&lf);
return hash;
}
void fillInBioHash(char *fileName, struct hash *bioHash)
/* Fill in the bioHash with key/value pairs from file. */
{
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *line = NULL;
int regSize = 0;
while(lineFileNextReal(lf, &line))
{
char *key = NULL;
char *val = NULL;
char *mark = NULL;
mark = strchr(line, '=');
if(mark == NULL) // Error: not in boulder IO format.
errAbort("pickCassettePcrPrimers::fillInBioHash() - ",
"Couldn't find '=' in line %s. File %s doesn't appear to be in boulderIO format.",
line, fileName);
if(mark == line) // First character is '=' means end of record.
break;
key = line;
val = mark+1;
*mark = '\0';
hashAddUnique(bioHash, key, cloneString(val));
}
lineFileClose(&lf);
}
static void addExtras(char *extraFile, struct trackHubCheckOptions *checkOptions)
/* Add settings from extra file (e.g. for specific hub display site) */
{
verbose(2, "Accepting extra settings in '%s'\n", extraFile);
checkOptions->extraFile = extraFile;
checkOptions->extra = hashNew(0);
struct lineFile *lf = NULL;
if (startsWith("http", extraFile))
{
struct dyString *ds = netSlurpUrl(extraFile);
char *s = dyStringCannibalize(&ds);
lf = lineFileOnString(extraFile, TRUE, s);
}
else
{
lf = lineFileOpen(extraFile, TRUE);
}
char *line;
while (lineFileNextReal(lf, &line))
{
hashAdd(checkOptions->extra, line, NULL);
}
lineFileClose(&lf);
verbose(3, "Found %d extra settings\n", hashNumEntries(checkOptions->extra));
}
void peakClusterMakerAddFromSource(struct peakClusterMaker *maker, struct peakSource *source)
/* Read through data source and add items to it to rangeTrees in maker */
{
struct hash *chromHash = maker->chromHash;
struct lineFile *lf = lineFileOpen(source->dataSource, TRUE);
struct lm *lm = chromHash->lm; /* Local memory pool - share with hash */
char *row[source->minColCount];
struct peakItem *item;
char *line;
while (lineFileNextReal(lf, &line))
{
char *asciiLine = lmCloneString(lm, line);
int wordCount = chopByWhite(line, row, source->minColCount);
lineFileExpectAtLeast(lf, source->minColCount, wordCount);
char *chrom = row[source->chromColIx];
struct hashEl *hel = hashLookup(chromHash, chrom);
if (hel == NULL)
{
struct rbTree *tree = rangeTreeNewDetailed(lm, maker->stack);
hel = hashAdd(chromHash, chrom, tree);
}
struct rbTree *tree = hel->val;
lmAllocVar(lm, item);
item->chrom = hel->name;
item->chromStart = sqlUnsigned(row[source->startColIx]);
item->chromEnd = sqlUnsigned(row[source->endColIx]);
item->score = sqlDouble(row[source->scoreColIx]) * source->normFactor;
if (item->score > 1000) item->score = 1000;
item->source = source;
item->asciiLine = asciiLine;
rangeTreeAddValList(tree, item->chromStart, item->chromEnd, item);
}
lineFileClose(&lf);
}
struct dyString *readAndReplaceTableName(char *fileName, char *table)
/* Read file into string. While doing so strip any leading comments
* and insist that the first non-comment line contain the words
* "create table" followed by a table name. Replace the table name,
* and copy the rest of the file verbatem. */
{
struct lineFile *lf = lineFileOpen(fileName, TRUE);
struct dyString *dy = dyStringNew(0);
char *line, *word;
if (!lineFileNextReal(lf, &line))
errAbort("No real lines in %s\n", fileName);
word = nextWord(&line);
if (!sameWord(word, "create"))
errAbort("Expecting first word in file to be CREATE. Got %s", word);
word = nextWord(&line);
if (word == NULL || !sameWord(word, "table"))
errAbort("Expecting second word in file to be table. Got %s", emptyForNull(word));
word = nextWord(&line);
if (word == NULL)
errAbort("Expecting table name on same line as CREATE TABLE");
sqlDyStringPrintf(dy, "CREATE TABLE %s ", table);
if (line != NULL)
dyStringAppend(dy, line);
dyStringAppendC(dy, '\n');
while (lineFileNext(lf, &line, NULL))
{
dyStringAppend(dy, line);
dyStringAppendC(dy, '\n');
}
lineFileClose(&lf);
return dy;
}
int fillSizes(char *sizes)
/**
* Fills the hash table of chrom sizes
*
*
*
*/
{
struct lineFile *lf;
lf = lineFileOpen(sizes,TRUE);
char *line;
struct sizes_hash *s;
while(lineFileNextReal(lf,&line)){
char *split[2];
chopByWhite(line,split,2);
s = malloc(sizeof(struct sizes_hash));
strcpy(s->name, split[0]);
s->length = atoi(split[1]);
int i;
for(i=0;i<NUMSAMPLES;i++)
s->numHets[i] = 0;
HASH_ADD_STR(chrSizes,name,s);
}
return 0;
}
void makeTableDescriptions(char *database, char *asFile)
/* makeTableDescriptions - Add table descriptions to database.. */
{
struct sqlConnection *conn = sqlConnect(database);
struct lineFile *lf = lineFileOpen(asFile, TRUE);
FILE *f = hgCreateTabFile(".", "tableDescriptions");
/* Open a tab file with name corresponding to tableName in tmpDir. */
char *line;
/* struct asObject *asList = */ asParseFile(asFile); /* Just to check syntax */
if (sqlTableExists(conn, "chromInfo"))
errAbort("%s looks like a genome database, has chromInfo, aborting",
database);
sqlRemakeTable(conn, "tableDescriptions",
"NOSQLINJ CREATE TABLE tableDescriptions (\n"
" tableName varchar(255) not null,\n"
" autoSqlDef longblob not null,\n"
" gbdAnchor varchar(255) not null,\n"
" PRIMARY KEY(tableName(32))\n"
")" );
while (lineFileNextReal(lf, &line))
{
if (startsWith("table", line))
{
struct dyString *as = dyStringNew(0);
char *name = trimSpaces(line + 6); /* Skip over table. */
char *escaped = NULL;
fprintf(f, "%s\t", name);
/* Putting lines into as. */
for (;;)
{
char *s;
dyStringAppend(as, line);
dyStringAppendC(as, '\n');
s = skipLeadingSpaces(line);
if (s[0] == ')')
break;
if (!lineFileNext(lf, &line, NULL))
errAbort("Unexpected end of file, missing closing paren in %s",
lf->fileName);
}
escaped = needMem(2*as->stringSize+1);
fprintf(f, "%s\t", sqlEscapeTabFileString2(escaped, as->string));
fprintf(f, "\n");
freez(&escaped);
dyStringFree(&as);
}
else
errAbort("Expecting table line %d of %s", lf->lineIx, lf->fileName);
}
hgLoadTabFile(conn, ".", "tableDescriptions", &f);
}
void loadAccs(char *accFile, struct hash *accTbl)
/* load accessions from a file */
{
struct lineFile *lf = lineFileOpen(accFile, TRUE);
char *line;
while (lineFileNextReal(lf, &line))
addAcc(accTbl, trimSpaces(line));
lineFileClose(&lf);
}
void edwFixRevoked(char *database, char *inFile)
/* edwFixRevoked - Mark as deprecated files that are revoked in ENCODE2. */
/* inFile is in format:
* metaVariable objStatus revoked [- reason]
* metaObject name */
{
struct sqlConnection *conn = edwConnect();
struct lineFile *lf = lineFileOpen(inFile, TRUE);
char *line;
char *defaultReason = "Revoked in ENCODE2";
char *reason = defaultReason;
while (lineFileNextReal(lf, &line))
{
if (startsWithWord("metaVariable", line))
{
char *pattern = "metaVariable objStatus revoked";
if (startsWithWord(pattern, line))
{
reason = skipLeadingSpaces(line + strlen(pattern));
if (isEmpty(reason))
reason = defaultReason;
else
{
if (reason[0] == '-')
reason = skipLeadingSpaces(reason + 1);
reason = cloneString(reason);
}
}
else
errAbort("??? %s\n", line);
}
else if (startsWithWord("metaObject", line))
{
char *row[3];
int wordCount = chopLine(line, row);
if (wordCount != 2)
errAbort("Strange metaobject line %d of %s\n", lf->lineIx, lf->fileName);
char *prefix = row[1];
if (!startsWith("wgEncode", prefix))
errAbort("Strange object line %d of %s\n", lf->lineIx, lf->fileName);
char query[512];
sqlSafef(query, sizeof(query),
"select * from edwFile where submitFileName like '%s/%%/%s%%'", database, prefix);
struct edwFile *ef, *efList = edwFileLoadByQuery(conn, query);
printf("# %s %s\n", prefix, reason);
for (ef = efList; ef != NULL; ef = ef->next)
{
long long id = ef->id;
printf("update edwFile set deprecated='%s' where id=%lld;\n", reason, id);
}
}
else
errAbort("Unrecognized first word in %s\n", line);
}
}
char *raFoldInOneRetName(struct lineFile *lf, struct hash *hashOfHash)
/* Fold in one record from ra file into hashOfHash.
* This will add ra's and ra fields to whatever already
* exists in the hashOfHash, overriding fields of the
* same name if they exist already. */
{
char *word, *line, *name;
struct hash *ra;
struct hashEl *hel;
/* Get first nonempty non-comment line and make sure
* it contains name. */
if (!lineFileNextReal(lf, &line))
return NULL;
word = nextWord(&line);
if (!sameString(word, "name"))
errAbort("Expecting 'name' line %d of %s, got %s",
lf->lineIx, lf->fileName, word);
name = nextWord(&line);
if (name == NULL)
errAbort("Short name field line %d of %s", lf->lineIx, lf->fileName);
/* Find ra hash associated with name, making up a new
* one if need be. */
if ((ra = hashFindVal(hashOfHash, name)) == NULL)
{
ra = newHash(7);
hashAdd(hashOfHash, name, ra);
hashAdd(ra, "name", lmCloneString(ra->lm, name));
}
/* Fill in fields of ra hash with data up to next
* blank line or end of file. */
for (;;)
{
if (!lineFileNext(lf, &line, NULL))
break;
line = skipLeadingSpaces(line);
if (line[0] == 0)
break;
if (line[0] == '#')
continue;
word = nextWord(&line);
line = skipLeadingSpaces(line);
if (line == NULL)
line = "";
hel = hashLookup(ra, word);
if (hel == NULL)
hel = hashAdd(ra, word, lmCloneString(ra->lm, line));
else
hel->val = lmCloneString(ra->lm, line);
}
return hashFindVal(ra, "name");
}
static boolean rowReaderNextFile(struct rowReader *rr)
/* read the next row into the rowReader from a file. */
{
char *line;
if (!lineFileNextReal(rr->lf, &line))
return FALSE;
rr->numCols = chopByChar(line, '\t', NULL, 0);
if (rr->numCols > rr->colSpace)
growRow(rr, rr->numCols);
chopByChar(line, '\t', rr->row, rr->numCols);
return TRUE;
}
void cnvBedToGenePred(char *bedFile, char *genePredFile)
/* convert bed format files to genePred format */
{
struct lineFile *bedLf = lineFileOpen(bedFile, TRUE);
FILE *gpFh = mustOpen(genePredFile, "w");
char *line;
while (lineFileNextReal(bedLf, &line))
cnvBedRec(line, gpFh);
carefulClose(&gpFh);
lineFileClose(&bedLf);
}
static char *mkChimeraPyScript(char *scriptFile)
/* create chimera python script from commands in scriptFile */
{
struct dyString *buf = dyStringNew(0);
dyStringAppend(buf, "from chimera import runCommand\n");
struct lineFile *lf = lineFileOpen(scriptFile, TRUE);
char *line;
while (lineFileNextReal(lf, &line))
dyStringPrintf(buf, "runCommand(\"%s\")\n", line);
lineFileClose(&lf);
return dyStringCannibalize(&buf);
}
void cnvBedToPsl(char *chromSizesFile, char *bedFile, char *pslFile)
/* convert bed format files to PSL format */
{
struct hash *chromSizes = loadChromSizes(chromSizesFile);
struct lineFile *bedLf = lineFileOpen(bedFile, TRUE);
FILE *pslFh = mustOpen(pslFile, "w");
char *line;
while (lineFileNextReal(bedLf, &line))
cnvBedRec(line, chromSizes, pslFh);
carefulClose(&pslFh);
lineFileClose(&bedLf);
}
struct hash *hashFirstWord(char *fileName)
/* Hash first word in each line. */
{
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *line, *word;
struct hash *hash = newHash(16);
while (lineFileNextReal(lf, &line))
{
word = nextWord(&line);
hashAdd(hash, word, NULL);
}
lineFileClose(&lf);
return hash;
}
struct slName *getWhiteListFromFile()
/* Read in the -justThese file and store the names of the enzymes */
/* in a list of slNames. */
{
struct slName *whiteList = NULL;
struct lineFile *lf = lineFileOpen(justThese, TRUE);
char *line;
while (lineFileNextReal(lf, &line))
{
struct slName *newName = newSlName(line);
slAddHead(&whiteList, newName);
}
lineFileClose(&lf);
return whiteList;
}
void lineFileRemoveInitialCustomTrackLines(struct lineFile *lf)
/* remove initial browser and track lines */
{
char *line;
while (lineFileNextReal(lf, &line))
{
if (!(startsWith("browser", line) || startsWith("track", line) ))
{
verbose(2, "found line not browser or track: %s\n", line);
lineFileReuse(lf);
break;
}
verbose(2, "skipping %s\n", line);
}
}
void bedClip(char *inFile, char *chromSizes, char *outFile)
/* bedClip - Remove lines from bed file that refer to off-chromosome places.. */
{
struct hash *chromSizesHash = bbiChromSizesFromFile(chromSizes);
struct lineFile *lf = lineFileOpen(inFile, TRUE);
FILE *f = mustOpen(outFile, "w");
char *line;
while (lineFileNextReal(lf, &line))
{
char *chrom = nextWord(&line);
char *startString = nextWord(&line);
char *endString = nextWord(&line);
if (endString == NULL)
errAbort("Need at least three fields line %d of %s", lf->lineIx, lf->fileName);
if (startString[0] == '-')
{
verbose(2, "Clipping negative line %d of %s\n", lf->lineIx, lf->fileName);
continue; // Clip off negatives
}
if (!isdigit(startString[0]))
errAbort("Expecting number got %s line %d of %s", startString, lf->lineIx, lf->fileName);
if (!isdigit(endString[0]))
errAbort("Expecting number got %s line %d of %s", endString, lf->lineIx, lf->fileName);
int start = sqlUnsigned(startString);
int end = sqlUnsigned(endString);
if (start >= end)
{
verbose(2, "Clipping end <= start line %d of %s\n", lf->lineIx, lf->fileName);
continue;
}
struct hashEl *hel = hashLookup(chromSizesHash, chrom);
if (hel == NULL)
errAbort("Chromosome %s isn't in %s line %d of %s\n", chrom, chromSizes, lf->lineIx, lf->fileName);
int chromSize = ptToInt(hel->val);
if (end > chromSize)
{
verbose(2, "Clipping end > chromSize line %d of %s\n", lf->lineIx, lf->fileName);
continue;
}
fprintf(f, "%s\t%s\t%s", chrom, startString, endString);
line = skipLeadingSpaces(line);
if (line == NULL || line[0] == 0)
fputc('\n', f);
else
fprintf(f, "\t%s\n", line);
}
carefulClose(&f);
}
struct slName* loadAccList(char *fname)
{
struct slName* accList = NULL;
struct lineFile* accFh;
char *line;
accFh = lineFileOpen(fname, TRUE);
while (lineFileNextReal(accFh, &line))
{
line = trimSpaces(line);
slSafeAddHead(&accList, newSlName(line));
}
lineFileClose(&accFh);
return accList;
}
void wigSort(char *input, char *output)
/* wigSort - Sort a wig file.. */
{
struct lineFile *lf = lineFileOpen(input, TRUE);
struct pos *pos, *posList = NULL;
char *line;
while (lineFileNextReal(lf, &line))
{
verbose(2, "processing %s\n", line);
AllocVar(pos);
pos->fileOffset = lineFileTell(lf);
if (posList != NULL)
posList->fileSize = pos->fileOffset - posList->fileOffset;
slAddHead(&posList, pos);
if (stringIn("chrom=", line))
{
parseSteppedSection(lf, line, pos);
}
else
{
/* Check for bed... */
char *words[5];
int wordCount = chopLine(line, words);
if (wordCount != 4)
errAbort("Unrecognized format line %d of %s:\n", lf->lineIx, lf->fileName);
pos->chrom = cloneString(words[0]);
pos->start = lineFileNeedNum(lf, words, 1);
}
}
if (posList != NULL)
{
posList->fileSize = lineFileTell(lf) - posList->fileOffset;
slReverse(&posList);
slSort(&posList, posCmp);
}
lineFileClose(&lf);
FILE *in = mustOpen(input, "r");
FILE *out = mustOpen(output, "w");
for (pos = posList; pos != NULL; pos = pos->next)
{
fseek(in, pos->fileOffset, SEEK_SET);
copyFileBytes(in, out, pos->fileSize);
}
carefulClose(&in);
carefulClose(&out);
}
