void webPushErrHandlers(void)
/* Push warn and abort handler for errAbort(). */
{
if (webInTextMode)
pushWarnHandler(textVaWarn);
else
pushWarnHandler(webVaWarn);
pushAbortHandler(softAbort);
hDumpStackPushAbortHandler();
}
boolean errCatchPushHandlers(struct errCatch *errCatch)
/* Push error handlers. Not usually called directly. */
{
pushAbortHandler(errCatchAbortHandler);
pushWarnHandler(errCatchWarnHandler);
struct errCatch **pErrCatchStack = getStack();
slAddHead(pErrCatchStack, errCatch);
return TRUE;
}
void _pf_punt_init()
/* Initialize punt/catch system. Mostly just redirects the
* errAbort handler. */
{
punter.message = dyStringNew(0);
punter.source = dyStringNew(0);
pushWarnHandler(puntWarnHandler);
pushAbortHandler(puntAbortHandler);
}
static void cartJsonPushErrHandlers()
/* Push warn and abort handlers for errAbort. */
{
if (dyWarn == NULL)
dyWarn = dyStringNew(0);
else
dyStringClear(dyWarn);
pushWarnHandler(cartJsonVaWarn);
pushAbortHandler(cartJsonAbort);
}
int main(int argc, char *argv[])
/* Process command line. */
{
if (!cgiIsOnWeb())
usage();
cgiSpoof(&argc, argv);
pushWarnHandler(badRequestWarn);
edwScriptSubmitStatus();
return 0;
}
int test(int argc, char *argv[])
{
int start, stop;
if (argc != 9 && argc != 10)
usageErr();
inName = argv[0];
cdnaName = argv[1];
chromDir = argv[2];
goodLogName = argv[3];
badLogName = argv[4];
unusualName = argv[5];
errName = argv[6];
start = atoi(argv[7]);
stop = atoi(argv[8]);
if (start >= stop)
usageErr();
if (argc == 10)
c2gName = argv[9];
inFile = mustOpen(inName, "r");
goodLogFile = mustOpen(goodLogName, "w");
badLogFile = mustOpen(badLogName, "w");
unusualFile = mustOpen(unusualName, "w");
errFile = mustOpen(errName, "w");
pushWarnHandler(reportWarning);
dnaUtilOpen();
printf("Loading chromosomes\n");
loadGenome(chromDir, &chroms, &chromNames, &chromCount);
startRedoHash();
printf("Analysing %s\n", inName);
if (weAreWeb())
htmlHorizontalLine();
analyse(start, stop);
//endRedoHash();
freeGenome(&chroms, &chromNames, chromCount);
popWarnHandler();
fclose(inFile);
fclose(goodLogFile);
fclose(badLogFile);
fclose(unusualFile);
fclose(errFile);
return 0;
}
struct pipeline *textOutInit(char *fileName, char *compressType)
/* Set up stdout to be HTTP text, file (if fileName is specified), or
* compressed file (if both fileName and compressType are specified --
* see textOut.h for supported compression types).
* Return NULL if no compression, otherwise a pipeline handle on which
* textOutClose should be called when we're done writing stdout. */
{
struct pipeline *compressPipeline = NULL;
trimSpaces(fileName);
if (isEmpty(fileName))
{
printf("Content-Type: text/plain\n\n");
}
else if (isEmpty(compressType) || sameWord(compressType, textOutCompressNone))
{
printf("Content-Type: application/octet-stream\n");
printf("Content-Disposition: attachment; filename=%s\n\n", fileName);
}
else
{
char *suffix = getCompressSuffix(compressType);
printf("Content-Type: application/x-%s\n", compressType);
if (endsWith(fileName, suffix))
printf("Content-Disposition: attachment; filename=%s\n\n", fileName);
else
printf("Content-Disposition: attachment; filename=%s%s\n\n",
fileName, suffix);
/* Send the Content header uncompressed! */
fflush(stdout);
/* Make sure no environment variables interfere with compressor. */
cleanEnvVars(compressType);
/* Redirect stdout to compressor pipeline object. */
compressPipeline = pipelineOpen1(getCompressor(compressType),
pipelineWrite, NULL, NULL);
if (-1 == dup2(pipelineFd(compressPipeline), STDOUT_FILENO))
errnoAbort("dup2(pipelineFd %d, stdout %d) failed in textOpen()",
pipelineFd(compressPipeline), STDOUT_FILENO);
}
pushWarnHandler(textOutWarnHandler);
pushAbortHandler(textOutAbortHandler);
return(compressPipeline);
}
void doMiddle()
/* doMiddle - put up middle part of web page, not including http and html headers/footers */
{
pushWarnHandler(localWarn);
printf("<FORM ACTION=\"../cgi-bin/edwWebSubmit\" METHOD=GET>\n");
struct sqlConnection *conn = edwConnectReadWrite(edwDatabase);
userEmail = edwGetEmailAndVerify();
if (userEmail == NULL)
logIn();
else if (cgiVarExists(stopButtonName))
stopUpload(conn);
else if (cgiVarExists("submitUrl"))
submitUrl(conn);
else if (cgiVarExists("monitor"))
monitorSubmission(conn);
else
getUrl(conn);
printf("</FORM>");
}
int main(int argc, char *argv[])
{
char *sangerName, *jimName, *updateName, *errName;
struct g2cFile *sangerGenes, *jimGenes;
if (argc != 5)
{
errAbort("c2gcheck - compares two gene-to-cdna files, notes differences\n"
"and writes out a third merged file.\n"
"Usage:\n"
" c2gcheck Sanger Jim Update errs\n");
}
memPool = lmInit(1<<16);
pushWarnHandler(reportWarning);
sangerName = argv[1];
jimName = argv[2];
updateName = argv[3];
errName = argv[4];
errFile = mustOpen(errName, "w");
sangerGenes = loadG2cFile(sangerName);
jimGenes = loadG2cFile(jimName);
checkOneFile(sangerGenes, sangerName);
checkOneFile(jimGenes, jimName);
checkTwoFiles(sangerGenes, jimGenes, "Jim unique");
checkTwoFiles(jimGenes, sangerGenes, "Sanger unique");
update(sangerGenes, jimGenes);
saveG2cFile(sangerGenes, updateName);
lmCleanup(&memPool);
return 0;
}
char *checkParams(char *database, char *prefix, char *type)
/* If we don't have valid CGI parameters, quit with a Bad Request HTTP response. */
{
pushWarnHandler(htmlVaBadRequestAbort);
pushAbortHandler(htmlVaBadRequestAbort);
if(prefix == NULL || database == NULL)
errAbort("%s", "Missing prefix and/or db CGI parameter");
if (! hDbIsActive(database))
errAbort("'%s' is not a valid, active database", htmlEncode(database));
if (isNotEmpty(type) && differentString(type, ALT_OR_PATCH))
errAbort("'%s' is not a valid type", type);
char *table = NULL;
if (! sameOk(type, ALT_OR_PATCH))
{
struct sqlConnection *conn = hAllocConn(database);
table = connGeneSuggestTable(conn);
hFreeConn(&conn);
if(table == NULL)
errAbort("gene autosuggest is not supported for db '%s'", database);
}
popWarnHandler();
popAbortHandler();
return table;
}
void fqToQa(char *inFile, char *outDir, char *outTabName)
/* fqToQa - convert from fq format with one big file to
* format with one file per clone. */
{
struct lineFile *in;
FILE *out = NULL, *tab;
int lineSize;
char *line;
char ucscName[128];
char path[512];
static char lastPath[512];
int outFileCount = 0;
struct hash *uniqClone = newHash(16);
struct hash *uniqFrag = newHash(19);
boolean ignore = FALSE;
makeDir(outDir);
errLog = mustOpen("fqToQa.err", "w");
pushWarnHandler(warnHandler);
tab = mustOpen(outTabName, "w");
printf("Converting %s", inFile);
fflush(stdout);
in = lineFileOpen(inFile, TRUE);
while (lineFileNext(in, &line, &lineSize))
{
if (line[0] == '>')
{
ignore = FALSE;
gsToUcsc(line+1, ucscName);
faRecNameToQaFileName(outDir, ucscName, path);
if (hashLookup(uniqFrag, ucscName))
{
ignore = TRUE;
warn("Duplicate %s in %s, ignoring all but first",
ucscName, inFile);
}
else
{
hashAdd(uniqFrag, ucscName, NULL);
}
if (!sameString(path, lastPath))
{
strcpy(lastPath, path);
carefulClose(&out);
if (hashLookup(uniqClone, path))
{
warn("Duplicate %s in %s ignoring all but first",
ucscName, inFile);
}
else
{
hashAdd(uniqClone, path, NULL);
out = mustOpen(path, "w");
++outFileCount;
if ((outFileCount&7) == 0)
{
putc('.', stdout);
fflush(stdout);
}
}
}
if (out != NULL && !ignore)
{
fprintf(out, ">%s\n", ucscName);
fprintf(tab, "%s\t%s\n", ucscName, line+1);
}
}
else
{
if (out != NULL && !ignore)
{
fputs(line, out);
fputc('\n', out);
}
}
}
carefulClose(&out);
fclose(tab);
lineFileClose(&in);
printf("Made %d .qa files in %s\n", outFileCount, outDir);
}
static void vcfGenotypesDetails(struct vcfRecord *rec, struct trackDb *tdb, char **displayAls)
/* Print summary of allele and genotype frequency, plus collapsible section
* with table of genotype details. */
{
struct vcfFile *vcff = rec->file;
if (vcff->genotypeCount == 0)
return;
// Wrapper table for collapsible section:
puts("<TABLE>");
pushWarnHandler(ignoreEm);
vcfParseGenotypes(rec);
popWarnHandler();
// Tally genotypes and alleles for summary:
int refs = 0, alts = 0, unks = 0;
int refRefs = 0, refAlts = 0, altAlts = 0, gtUnk = 0, gtOther = 0, phasedGts = 0;
int i;
for (i = 0; i < vcff->genotypeCount; i++)
{
struct vcfGenotype *gt = &(rec->genotypes[i]);
if (gt->isPhased)
phasedGts++;
if (gt->hapIxA == 0)
refs++;
else if (gt->hapIxA > 0)
alts++;
else
unks++;
if (!gt->isHaploid)
{
if (gt->hapIxB == 0)
refs++;
else if (gt->hapIxB > 0)
alts++;
else
unks++;
if (gt->hapIxA == 0 && gt->hapIxB == 0)
refRefs++;
else if (gt->hapIxA == 1 && gt->hapIxB == 1)
altAlts++;
else if ((gt->hapIxA == 1 && gt->hapIxB == 0) ||
(gt->hapIxA == 0 && gt->hapIxB == 1))
refAlts++;
else if (gt->hapIxA < 0 || gt->hapIxB < 0)
gtUnk++;
else
gtOther++;
}
}
printf("<B>Genotype count:</B> %d", vcff->genotypeCount);
if (differentString(seqName, "chrY"))
printf(" (%d phased)", phasedGts);
else
printf(" (haploid)");
puts("<BR>");
int totalAlleles = refs + alts + unks;
double refAf = (double)refs/totalAlleles;
double altAf = (double)alts/totalAlleles;
printf("<B>Alleles:</B> %s: %d (%.3f%%); %s: %d (%.3f%%)",
displayAls[0], refs, 100*refAf, displayAls[1], alts, 100*altAf);
if (unks > 0)
printf("; unknown: %d (%.3f%%)", unks, 100 * (double)unks/totalAlleles);
puts("<BR>");
// Should be a better way to detect haploid chromosomes than comparison with "chrY":
if (vcff->genotypeCount > 1 && differentString(seqName, "chrY"))
{
printf("<B>Genotypes:</B> %s/%s: %d (%.3f%%); %s/%s: %d (%.3f%%); %s/%s: %d (%.3f%%)",
displayAls[0], displayAls[0], refRefs, 100*(double)refRefs/vcff->genotypeCount,
displayAls[0], displayAls[1], refAlts, 100*(double)refAlts/vcff->genotypeCount,
displayAls[1], displayAls[1], altAlts, 100*(double)altAlts/vcff->genotypeCount);
if (gtUnk > 0)
printf("; unknown: %d (%.3f%%)", gtUnk, 100*(double)gtUnk/vcff->genotypeCount);
if (gtOther > 0)
printf("; other: %d (%.3f%%)", gtOther, 100*(double)gtOther/vcff->genotypeCount);
printf("<BR>\n");
if (rec->alleleCount == 2)
{
boolean showHW = cartOrTdbBoolean(cart, tdb, VCF_SHOW_HW_VAR, FALSE);
if (showHW)
printf("<B><A HREF=\"http://en.wikipedia.org/wiki/Hardy%%E2%%80%%93Weinberg_principle\" "
"TARGET=_BLANK>Hardy-Weinberg equilibrium</A>:</B> "
"P(%s/%s) = %.3f%%; P(%s/%s) = %.3f%%; P(%s/%s) = %.3f%%<BR>",
displayAls[0], displayAls[0], 100*refAf*refAf,
displayAls[0], displayAls[1], 100*2*refAf*altAf,
displayAls[1], displayAls[1], 100*altAf*altAf);
}
}
puts("<BR>");
vcfGenotypeTable(rec, tdb->track, displayAls);
puts("</TABLE>");
}
int main(int argc, char *argv[])
{
long enteredMainTime = clock1000();
struct dyString *output = newDyString(10000);
setUdcCacheDir();
cgiSpoof(&argc, argv);
pushWarnHandler(htmlVaBadRequestAbort);
pushAbortHandler(htmlVaBadRequestAbort);
char *database = cgiString("db");
char *cmd = cgiString("cmd");
char *jsonp = cgiOptionalString("jsonp");
if (!hDbExists(database))
errAbort("Invalid database '%s'", database);
if (!strcmp(cmd, "defaultPos"))
{
dyStringPrintf(output, "{\"pos\": \"%s\"}", hDefaultPos(database));
}
else if (!strcmp(cmd, "metaDb"))
{
// Return list of values for given metaDb var
// e.g. http://genome.ucsc.edu/hgApi?db=hg18&cmd=metaDb&var=cell
struct sqlConnection *conn = hAllocConn(database);
boolean metaDbExists = sqlTableExists(conn, "metaDb");
if (metaDbExists)
{
char *var = cgiOptionalString("var");
if (!var)
errAbort("Missing var parameter");
boolean fileSearch = (cgiOptionalInt("fileSearch",0) == 1);
struct slPair *pairs = mdbValLabelSearch(conn, var, MDB_VAL_STD_TRUNCATION, FALSE,
!fileSearch, fileSearch);
struct slPair *pair;
dyStringPrintf(output, "[\n");
for (pair = pairs; pair != NULL; pair = pair->next)
{
if (pair != pairs)
dyStringPrintf(output, ",\n");
dyStringPrintf(output, "['%s','%s']", javaScriptLiteralEncode(mdbPairLabel(pair)),
javaScriptLiteralEncode(mdbPairVal(pair)));
}
dyStringPrintf(output, "\n]\n");
}
else
errAbort("Assembly does not support metaDb");
}
// TODO: move to lib since hgTracks and hgApi share
#define METADATA_VALUE_PREFIX "hgt_mdbVal"
else if (startsWith(METADATA_VALUE_PREFIX, cmd))
{
// Returns metaDb value control: drop down or free text, with or without help link.
// e.g. http://genome.ucsc.edu/hgApi?db=hg18&cmd=hgt_mdbVal3&var=cell
// TODO: Move guts to lib, so that hgTracks::searchTracks.c and hgApi.c can share
struct sqlConnection *conn = hAllocConn(database);
boolean metaDbExists = sqlTableExists(conn, "metaDb");
if (metaDbExists)
{
char *var = cgiOptionalString("var");
if (!var)
errAbort("Missing var parameter");
int ix = atoi(cmd+strlen(METADATA_VALUE_PREFIX)); // 1 based index
if (ix == 0) //
errAbort("Unsupported 'cmd' parameter");
enum cvSearchable searchBy = cvSearchMethod(var);
char name[128];
safef(name,sizeof name,"%s%i",METADATA_VALUE_PREFIX,ix);
if (searchBy == cvSearchBySingleSelect || searchBy == cvSearchByMultiSelect)
{
boolean fileSearch = (cgiOptionalInt("fileSearch",0) == 1);
struct slPair *pairs = mdbValLabelSearch(conn, var, MDB_VAL_STD_TRUNCATION, FALSE,
!fileSearch, fileSearch);
if (slCount(pairs) > 0)
{
char *dropDownHtml =
cgiMakeSelectDropList((searchBy == cvSearchByMultiSelect),
name, pairs, NULL, ANYLABEL, "mdbVal",
"style='min-width: 200px; font-size: .9em;' "
"onchange='findTracksMdbValChanged(this);'");
if (dropDownHtml)
{
dyStringAppend(output,dropDownHtml);
freeMem(dropDownHtml);
}
slPairFreeList(&pairs);
}
}
else if (searchBy == cvSearchByFreeText)
{
dyStringPrintf(output,"<input type='text' name='%s' value='' class='mdbVal freeText' "
"onchange='findTracksMdbValChanged(this);' style='max-width:310px; "
"width:310px; font-size:.9em;'>", name);
}
else if (searchBy == cvSearchByWildList)
{
dyStringPrintf(output,"<input type='text' name='%s' value='' class='mdbVal wildList' "
"title='enter comma separated list of values' "
"onchange='findTracksMdbValChanged(this);' style='max-width:310px; "
"width:310px; font-size:.9em;'>", name);
}
else if (searchBy == cvSearchByDateRange || searchBy == cvSearchByIntegerRange)
{
// TO BE IMPLEMENTED
}
else
errAbort("Metadata variable not searchable");
dyStringPrintf(output,"<span id='helpLink%i'> </span>",ix);
}
else
errAbort("Assembly does not support metaDb");
}
else if (!strcmp(cmd, "tableMetadata"))
{ // returns an html table with metadata for a given track
char *trackName = cgiOptionalString("track");
boolean showLonglabel = (NULL != cgiOptionalString("showLonglabel"));
boolean showShortLabel = (NULL != cgiOptionalString("showShortLabel"));
if (trackName != NULL)
{
// hTrackDbForTrackAndAncestors avoids overhead of getting whole track list!
struct trackDb *tdb = hTrackDbForTrackAndAncestors(database, trackName);
if (tdb != NULL)
{
char * html = metadataAsHtmlTable(database,tdb,showLonglabel,showShortLabel);
if (html)
{
dyStringAppend(output,html);
freeMem(html);
}
else
dyStringPrintf(output,"No metadata found for track %s.",trackName);
}
else
dyStringPrintf(output,"Track %s not found",trackName);
}
else
dyStringAppend(output,"No track variable found");
}
else if (sameString(cmd, "codonToPos") || sameString(cmd, "exonToPos"))
{
char query[256];
struct sqlResult *sr;
char **row;
struct genePred *gp;
char *name = cgiString("name");
char *table = cgiString("table");
int num = cgiInt("num");
struct sqlConnection *conn = hAllocConn(database);
sqlSafef(query, sizeof(query), "select name, chrom, strand, txStart, txEnd, cdsStart, cdsEnd, exonCount, exonStarts, exonEnds from %s where name = '%s'", table, name);
sr = sqlGetResult(conn, query);
if ((row = sqlNextRow(sr)) != NULL)
{
gp = genePredLoad(row);
boolean found;
int start, end;
if (sameString(cmd, "codonToPos"))
found = codonToPos(gp, num, &start, &end);
else
found = exonToPos(gp, num, &start, &end);
if (found)
dyStringPrintf(output, "{\"pos\": \"%s:%d-%d\"}", gp->chrom, start + 1, end);
else
dyStringPrintf(output, "{\"error\": \"%d is an invalid %s for this gene\"}", num, sameString(cmd, "codonToPos") ? "codon" : "exon");
}
else
dyStringPrintf(output, "{\"error\": \"Couldn't find item: %s\"}", name);
sqlFreeResult(&sr);
hFreeConn(&conn);
}
else
{
warn("unknown cmd: %s",cmd);
errAbort("Unsupported 'cmd' parameter");
}
apiOut(dyStringContents(output), jsonp);
cgiExitTime("hgApi", enteredMainTime);
return 0;
}
int main(int argc, char *argv[])
{
char *genoListName;
char *cdnaListName;
char *oocFileName;
char *pairFileName;
struct patSpace *patSpace;
long startTime, endTime;
char **genoList;
int genoListSize;
char *genoListBuf;
char **cdnaList;
int cdnaListSize;
char *cdnaListBuf;
char *genoName;
int i;
int estIx = 0;
struct dnaSeq **seqListList = NULL, *seq;
static char hitFileName[512], mergerFileName[512], okFileName[512];
char *outRoot;
struct hash *pairHash;
if (dumpMe)
{
bigHtmlFile = mustOpen("C:\\inetpub\\wwwroot\\test\\patAli.html", "w");
littleHtmlFile = mustOpen("C:\\inetpub\\wwwroot\\test\\patSpace.html", "w");
htmStart(bigHtmlFile, "PatSpace Alignments");
htmStart(littleHtmlFile, "PatSpace Index");
}
if ((hostName = getenv("HOST")) == NULL)
hostName = "";
if (argc != 6)
usage();
pushWarnHandler(patSpaceWarnHandler);
startTime = clock1000();
dnaUtilOpen();
makePolys();
genoListName = argv[1];
cdnaListName = argv[2];
oocFileName = argv[3];
pairFileName = argv[4];
outRoot = argv[5];
sprintf(hitFileName, "%s.hit", outRoot);
sprintf(mergerFileName, "%s.glu", outRoot);
sprintf(okFileName, "%s.ok", outRoot);
readAllWords(genoListName, &genoList, &genoListSize, &genoListBuf);
readAllWords(cdnaListName, &cdnaList, &cdnaListSize, &cdnaListBuf);
pairHash = makePairHash(pairFileName);
hitOut = mustOpen(hitFileName, "w");
mergerOut = mustOpen(mergerFileName, "w");
dumpOut = mustOpen("dump.out", "w");
seqListList = needMem(genoListSize*sizeof(seqListList[0]) );
fprintf(hitOut, "Pattern space 0.2 cDNA matcher\n");
fprintf(hitOut, "cDNA files: ", cdnaListSize);
for (i=0; i<cdnaListSize; ++i)
fprintf(hitOut, " %s", cdnaList[i]);
fprintf(hitOut, "\n");
fprintf(hitOut, "%d genomic files\n", genoListSize);
for (i=0; i<genoListSize; ++i)
{
genoName = genoList[i];
if (!startsWith("//", genoName) )
{
seqListList[i] = seq = faReadAllDna(genoName);
fprintf(hitOut, "%d els in %s ", slCount(seq), genoList[i]);
for (; seq != NULL; seq = seq->next)
fprintf(hitOut, "%d ", seq->size);
fprintf(hitOut, "\n");
}
}
patSpace = makePatSpace(seqListList, genoListSize, oocFileName);
for (i=0; i<cdnaListSize; ++i)
{
FILE *f;
char *estFileName;
DNA *dna;
char *estName;
int size;
int c;
int maxSizeForFuzzyFind = 20000;
int dotCount = 0;
estFileName = cdnaList[i];
if (startsWith("//", estFileName) )
continue;
f = mustOpen(estFileName, "rb");
while ((c = fgetc(f)) != EOF)
if (c == '>')
break;
printf("%s", cdnaList[i]);
fflush(stdout);
while (fastFaReadNext(f, &dna, &size, &estName))
{
aliSeqName = estName;
if (size < maxSizeForFuzzyFind) /* Some day need to fix this somehow... */
{
struct hashEl *hel;
struct cdnaAliList *calList = NULL;
hel = hashLookup(pairHash, estName);
if (hel != NULL) /* Do pair processing. */
{
struct estPair *ep;
struct seq *thisSeq, *otherSeq;
ep = hel->val;
if (hel->name == ep->name3)
{
thisSeq = &ep->seq3;
otherSeq = &ep->seq5;
}
else
{
thisSeq = &ep->seq5;
otherSeq = &ep->seq3;
}
if (otherSeq->dna == NULL) /* First in pair - need to save sequence. */
{
thisSeq->size = size;
thisSeq->dna = needMem(size);
memcpy(thisSeq->dna, dna, size);
}
else /* Second in pair - do gluing and free partner. */
{
char mergedName[64];
thisSeq->dna = dna;
thisSeq->size = size;
sprintf(mergedName, "%s_AND_%s", ep->name5, ep->name3);
patSpaceFindOne(patSpace, ep->seq5.dna, ep->seq5.size,
'+', '5', ep->name5, &calList);
reverseComplement(ep->seq5.dna, ep->seq5.size);
patSpaceFindOne(patSpace, ep->seq5.dna, ep->seq5.size,
'-', '5', ep->name5, &calList);
patSpaceFindOne(patSpace, ep->seq3.dna, ep->seq3.size,
'+', '3', ep->name3, &calList);
reverseComplement(ep->seq3.dna, ep->seq3.size);
patSpaceFindOne(patSpace, ep->seq3.dna, ep->seq3.size,
'-', '3', ep->name3, &calList);
slReverse(&calList);
writeMergers(calList, mergedName, genoList);
freez(&otherSeq->dna);
thisSeq->dna = NULL;
thisSeq->size =otherSeq->size = 0;
}
}
else
{
patSpaceFindOne(patSpace, dna, size, '+', '5', estName, &calList);
reverseComplement(dna, size);
patSpaceFindOne(patSpace, dna, size, '-', '5', estName, &calList);
slReverse(&calList);
writeMergers(calList, estName, genoList);
}
++estIx;
if ((estIx & 0xfff) == 0)
{
printf(".");
++dotCount;
fflush(stdout);
}
}
}
printf("\n");
}
aliSeqName = "";
printf("ffSubmitted %3d ffAccepted %3d ffOkScore %3d ffSolidMatch %2d\n",
ffSubmitted, ffAccepted, ffOkScore, ffSolidMatch);
endTime = clock1000();
printf("Total time is %4.2f\n", 0.001*(endTime-startTime));
/* Write out file who's presense say's we succeeded */
{
FILE *f = mustOpen(okFileName, "w");
fputs("ok", f);
fclose(f);
}
if (dumpMe)
{
htmEnd(bigHtmlFile);
htmEnd(littleHtmlFile);
}
return 0;
}
struct pipeline *textOutInit(char *fileName, char *compressType, int *saveStdout)
/* Set up stdout to be HTTP text, file (if fileName is specified), or
* compressed file (if both fileName and compressType are specified --
* see textOut.h for supported compression types).
* Return NULL if no compression, otherwise a pipeline handle on which
* textOutClose should be called when we're done writing stdout. */
{
struct pipeline *compressPipeline = NULL;
// if path contains a slash, we are outputting to a local file
boolean outToFile = (strchr(fileName, '/') != NULL);
if (outToFile)
{
FILE *f;
f = fopen(fileName, "w");
/* We want to capture stdout output to a file */
fflush(stdout);
int tempOut = dup(STDOUT_FILENO);
if (saveStdout)
*saveStdout = tempOut;
dup2(fileno(f),STDOUT_FILENO); /* closes STDOUT before setting it again */
fclose(f);
}
trimSpaces(fileName);
if (isEmpty(fileName))
{
printf("Content-Type: text/plain\n\n");
}
else if (isEmpty(compressType) || sameWord(compressType, textOutCompressNone))
{
if (!outToFile)
{
printf("Content-Type: application/octet-stream\n");
printf("Content-Disposition: attachment; filename=%s\n\n", fileName);
}
}
else
{
if (!outToFile)
{
char *suffix = getCompressSuffix(compressType);
printf("Content-Type: application/x-%s\n", compressType);
if (endsWith(fileName, suffix))
printf("Content-Disposition: attachment; filename=%s\n\n", fileName);
else
printf("Content-Disposition: attachment; filename=%s%s\n\n",
fileName, suffix);
/* Send the Content header uncompressed! */
fflush(stdout);
}
/* Make sure no environment variables interfere with compressor. */
cleanEnvVars(compressType);
/* Redirect stdout to compressor pipeline object. */
compressPipeline = pipelineOpen1(getCompressor(compressType),
pipelineWrite, NULL, NULL);
if (-1 == dup2(pipelineFd(compressPipeline), STDOUT_FILENO))
errnoAbort("dup2(pipelineFd %d, stdout %d) failed in textOpen()",
pipelineFd(compressPipeline), STDOUT_FILENO);
}
pushWarnHandler(textOutWarnHandler);
pushAbortHandler(textOutAbortHandler);
return(compressPipeline);
}
void doMiddle()
/* Write what goes between BODY and /BODY */
{
pushWarnHandler(localWarn);
if (!cgiServerHttpsIsOn())
usage();
struct sqlConnection *conn = edwConnectReadWrite();
printf("<FORM ACTION=\"edwWebRegisterScript\" METHOD=POST>\n");
printf("<B>Register Script with ENCODE Data Warehouse</B><BR>\n");
#ifdef SOON
uglyf("HTTP_AUTHENTICATION: '%s'<BR>\n", getenv("HTTP_AUTHENTICATION"));
uglyf("HTTP_AUTHORIZATION: '%s'<BR>\n", getenv("HTTP_AUTHORIZATION"));
dumpEnv(mainEnv);
#endif
if (userEmail == NULL)
{
printf("Please sign in:");
printf("<INPUT TYPE=BUTTON NAME=\"signIn\" VALUE=\"sign in\" id=\"signin\">");
}
else if (cgiVarExists("description"))
{
struct edwUser *user = edwUserFromEmail(conn, userEmail);
if (user == NULL)
edwWarnUnregisteredUser(userEmail);
else
{
char password[HEXED_32_SIZE];
edwRandomHexed32(password);
char babyName[HEXED_32_SIZE];
edwRandomBabble(babyName, sizeof(babyName));
edwRegisterScript(conn, user, babyName, password, cgiString("description"));
printf("Script now registered.<BR>\n");
printf("The script user name is %s.<BR>\n", babyName);
printf("The script password is %s.<BR>\n", password);
printf("Please save the script user name and password somewhere. ");
puts("Please pass these two and the URL");
puts(" of your validated manifest file (validated.txt) to our server to submit data.");
puts("Construct a URL of the form:<BR>");
printf("<PRE>https://encodedcc.sdsc.edu/cgi-bin/edwScriptSubmit"
"?user=%s&password=%s&url=%s\n</PRE>",
babyName, password,
cgiEncode("http://your.host.edu/your_dir/validated.txt"));
puts("That is pass the CGI encoded variables user, password, and url to the ");
puts("web services CGI at");
puts("https://encodedcc.sdsc.edu/cgi-bin/edwScriptSubmit. ");
puts("You can use the http://encodedcc.sdsc.edu/cgi-bin/edwWebBrowse site to ");
puts("monitor your submission interactively. Please contact your wrangler if you ");
puts("have any questions.<BR>");
cgiMakeButton("submit", "Register another script");
}
printf(" ");
edwPrintLogOutButton();
}
else
{
struct edwUser *user = edwUserFromEmail(conn, userEmail);
edwPrintLogOutButton();
if (user == NULL)
edwWarnUnregisteredUser(userEmail);
else
{
printf("%s is authorized to register a new script<BR>\n", userEmail);
printf("<BR>Script description:\n");
cgiMakeTextVar("description", NULL, 80);
cgiMakeSubmitButton();
}
}
printf("</FORM>\n");
}
static void vcfHapClusterDraw(struct track *tg, int seqStart, int seqEnd,
struct hvGfx *hvg, int xOff, int yOff, int width,
MgFont *font, Color color, enum trackVisibility vis)
/* Split samples' chromosomes (haplotypes), cluster them by center-weighted
* alpha similarity, and draw in the order determined by clustering. */
{
struct vcfFile *vcff = tg->extraUiData;
if (vcff->records == NULL)
return;
purple = hvGfxFindColorIx(hvg, 0x99, 0x00, 0xcc);
undefYellow = hvGfxFindRgb(hvg, &undefinedYellowColor);
enum hapColorMode colorMode = getColorMode(tg->tdb);
pushWarnHandler(ignoreEm);
struct vcfRecord *rec;
for (rec = vcff->records; rec != NULL; rec = rec->next)
vcfParseGenotypes(rec);
popWarnHandler();
unsigned short gtHapCount = 0;
int nRecords = slCount(vcff->records);
int centerIx = getCenterVariantIx(tg, seqStart, seqEnd, vcff->records);
// Limit the number of variants that we compare, to keep from timing out:
// (really what we should limit is the number of distinct haplo's passed to hacTree!)
// In the meantime, this should at least be a cart var...
int maxVariantsPerSide = 50;
int startIx = max(0, centerIx - maxVariantsPerSide);
int endIx = min(nRecords, centerIx+1 + maxVariantsPerSide);
struct hacTree *ht = NULL;
unsigned short *gtHapOrder = clusterHaps(vcff, centerIx, startIx, endIx, >HapCount, &ht);
struct vcfRecord *centerRec = NULL;
int ix;
// Unlike drawing order (last drawn is on top), the first mapBox gets priority,
// so map center variant before drawing & mapping other variants!
for (rec = vcff->records, ix=0; rec != NULL; rec = rec->next, ix++)
{
if (ix == centerIx)
{
centerRec = rec;
mapBoxForCenterVariant(rec, hvg, tg, xOff, yOff, width);
break;
}
}
for (rec = vcff->records, ix=0; rec != NULL; rec = rec->next, ix++)
{
boolean isClustered = (ix >= startIx && ix < endIx);
if (ix != centerIx)
drawOneRec(rec, gtHapOrder, gtHapCount, tg, hvg, xOff, yOff, width, isClustered, FALSE,
colorMode);
}
// Draw the center rec on top, outlined with black lines, to make sure it is very visible:
drawOneRec(centerRec, gtHapOrder, gtHapCount, tg, hvg, xOff, yOff, width, TRUE, TRUE,
colorMode);
// Draw as much of the tree as can fit in the left label area:
int extraPixel = (colorMode == altOnlyMode) ? 1 : 0;
int hapHeight = tg->height- CLIP_PAD - 2*extraPixel;
struct yFromNodeHelper yHelper = {0, NULL, NULL};
initYFromNodeHelper(&yHelper, yOff+extraPixel, hapHeight, gtHapCount, gtHapOrder,
vcff->genotypeCount);
struct titleHelper titleHelper = { NULL, 0, 0, 0, 0, NULL, NULL };
initTitleHelper(&titleHelper, tg->track, startIx, centerIx, endIx, nRecords, vcff);
char *treeAngle = cartOrTdbString(cart, tg->tdb, VCF_HAP_TREEANGLE_VAR, VCF_DEFAULT_HAP_TREEANGLE);
boolean drawRectangle = sameString(treeAngle, VCF_HAP_TREEANGLE_RECTANGLE);
drawTreeInLabelArea(ht, hvg, yOff+extraPixel, hapHeight+CLIP_PAD, &yHelper, &titleHelper,
drawRectangle);
}
void tryToDeprecate(struct sqlConnection *conn)
/* CGI variables are set - if possible deprecate, otherwise put up error message. */
{
pushWarnHandler(localWarn);
fileList = cgiString("fileList");
reason = cloneString(trimSpaces(cgiString("reason")));
if (isEmpty(reason))
{
warn("Please enter a reason for deprecation.");
getFileListAndReason(conn);
}
else
{
/* Go through list of accessions and make sure they are all well formed and correspond to files that exist. */
boolean ok = TRUE;
struct slName *accList = slNameListOfUniqueWords(cloneString(fileList), FALSE);
struct slName *acc;
struct slInt *idList = NULL, *idEl;
for (acc = accList; acc != NULL; acc = acc->next)
{
char *licensePlate = acc->name;
if (!startsWith(edwLicensePlatePrefix, licensePlate))
{
ok = FALSE;
warn("%s is not an accession, doesn't start with %s", licensePlate, edwLicensePlatePrefix);
break;
}
char query[256];
sqlSafef(query, sizeof(query), "select fileId from edwValidFile where licensePlate='%s'", licensePlate);
int id = sqlQuickNum(conn, query);
if (id == 0)
{
ok = FALSE;
warn("%s - no such accession. ", licensePlate);
break;
}
/* check to see is it ok tor deprecate this file */
if (!okToDeprecateThisFile(conn, id, userEmail))
{
ok = FALSE;
warn("You can not deprecate %s which was originally uploaded by %s.\n",
licensePlate, edwFindOwnerNameFromFileId(conn, id));
warn("Please click the check box below to override this rule.");
break;
}
idEl = slIntNew(id);
slAddTail(&idList, idEl);
}
if (accList == NULL)
{
warn("Please enter some file accessions");
ok = FALSE;
}
/* If a problem then put up page to try again, otherwise do deprecation. */
if (!ok)
getFileListAndReason(conn);
else
{
deprecateFileList(conn, idList, reason);
printf("Deprecated %d files<BR>\n", slCount(idList));
cgiMakeButton("submit", "Deprecate More Files");
printf(" ");
edwPrintLogOutButton();
}
}
}
void pushWarnAbort()
/* Push handler that will abort on warnings. */
{
pushWarnHandler(warnAbortHandler);
}
void pushSilentWarnHandler()
/* Set warning handler to be quiet. Do a popWarnHandler to restore. */
{
pushWarnHandler(silentVaWarn);
}
int main(int argc, char *argv[])
{
char *genoListName;
char *otherListName;
char *oocFileName;
char *typeName;
char *outName;
struct patSpace *patSpace;
long startTime, endTime;
char **genoList;
int genoListSize;
char *genoListBuf;
char **otherList;
int otherListSize;
char *otherListBuf;
char *genoName;
int i;
int blockCount = 0;
struct dnaSeq **seqListList = NULL, *seq = NULL;
char *outRoot;
struct sqlConnection *conn;
enum ffStringency stringency = ffCdna;
int seedSize = 10;
FILE *out;
boolean noHead = FALSE;
struct repeatTracker *rt;
struct hash *repeatHash = newHash(10);
hostName = getenv("HOST");
pushWarnHandler(warnHandler);
startTime = clock1();
cgiSpoof(&argc, argv);
minMatch = cgiOptionalInt("minMatch", minMatch);
maxBad = cgiOptionalInt("maxBad", maxBad);
minBases = cgiOptionalInt("minBases", minBases);
dnaUtilOpen();
#ifdef DEBUG
/* Hard wire command line input so don't have to type it in each
* time run the stupid Gnu debugger. */
genoListName = "pFoo/geno.lst";
otherListName = "pFoo/bacend.lst";
typeName = "genomic";
oocFileName = "/d/biodata/human/10.ooc";
outName = "pFoo/pFoo.psl";
#else
if (argc != 6 && argc != 7)
usage();
genoListName = argv[1];
otherListName = argv[2];
typeName = argv[3];
oocFileName = argv[4];
if (sameWord(oocFileName, "none"))
oocFileName = NULL;
outName = argv[5];
if (argc == 7)
{
if (sameWord("noHead", argv[6]))
noHead = TRUE;
else
usage();
}
#endif
if (sameWord(typeName, "mRNA") || sameWord(typeName, "cDNA"))
{
stringency = ffCdna;
}
else if (sameWord(typeName, "genomic"))
{
stringency = ffTight;
}
else if (sameWord(typeName, "g2g"))
{
stringency = ffTight;
veryTight = TRUE;
seedSize = 11;
}
else if (sameString(typeName, "asm"))
{
stringency = ffTight;
avoidSelfSelf = TRUE;
}
else
{
warn("Unrecognized otherType %s\n", typeName);
usage();
}
readAllWordsOrFa(genoListName, &genoList, &genoListSize, &genoListBuf);
filterMissingFiles(genoList, &genoListSize);
if (genoListSize <= 0)
errAbort("There are no files that exist in %s\n", genoListName);
readAllWordsOrFa(otherListName, &otherList, &otherListSize, &otherListBuf);
if (otherListSize <= 0)
errAbort("There are no files that exist in %s\n", otherListName);
filterMissingFiles(otherList, &otherListSize);
out = mustOpen(outName, "w");
if (!noHead)
pslWriteHead(out);
AllocArray(seqListList, genoListSize);
for (i=0; i<genoListSize; ++i)
{
genoName = genoList[i];
if (!startsWith("#", genoName) )
seqListList[i] = seq = faReadAllDna(genoName);
for (;seq != NULL; seq = seq->next)
{
int size = seq->size;
char *name = seq->name;
struct hashEl *hel;
AllocVar(rt);
AllocArray(rt->repBytes, size);
rt->seq = seq;
if ((hel = hashLookup(repeatHash, name)) != NULL)
errAbort("Duplicate %s in %s\n", name, genoName);
hashAdd(repeatHash, name, rt);
}
storeMasked(repeatHash, genoName);
}
patSpace = makePatSpace(seqListList, genoListSize, seedSize, oocFileName, minMatch, 2000);
endTime = clock1();
printf("Made index in %ld seconds\n", (endTime-startTime));
startTime = endTime;
for (i=0; i<otherListSize; ++i)
{
FILE *f;
char *otherName;
int c;
int dotCount = 0;
struct dnaSeq otherSeq;
ZeroVar(&otherSeq);
otherName = otherList[i];
if (startsWith("#", otherName) )
continue;
f = mustOpen(otherName, "r");
while ((c = fgetc(f)) != EOF)
if (c == '>')
break;
printf("%s\n", otherName);
fflush(stdout);
while (faFastReadNext(f, &otherSeq.dna, &otherSeq.size, &otherSeq.name))
{
aliSeqName = otherSeq.name;
oneStrand(patSpace, repeatHash, &otherSeq, FALSE, stringency, out);
reverseComplement(otherSeq.dna, otherSeq.size);
oneStrand(patSpace, repeatHash, &otherSeq, TRUE, stringency, out);
aliSeqName = NULL;
}
fclose(f);
}
freePatSpace(&patSpace);
endTime = clock1();
printf("Alignment time is %ld sec\n", (endTime-startTime));
startTime = endTime;
fclose(out);
return 0;
}
void pushRHandlers() {
pushAbortHandler(R_abortHandler);
pushWarnHandler(R_warnHandler);
}
