static void processBlock(struct blastBlock *bb, unsigned flags,
FILE* pslFh, FILE* scoreFh)
/* process one gapped alignment block */
{
struct blastGappedAli* ba = bb->gappedAli;
struct blastQuery *bq = ba->query;
struct psl *psl = pslBuildFromHsp(firstWordInLine(bq->query), bq->queryBaseCount, bb->qStart, bb->qEnd, ((bb->qStrand > 0) ? '+' : '-'), bb->qSym,
firstWordInLine(ba->targetName), ba->targetSize, bb->tStart, bb->tEnd, ((bb->tStrand > 0) ? '+' : '-'), bb->tSym,
flags);
if (psl->blockCount > 0 && (bb->eVal <= eVal || eVal == -1))
outputPsl(bb, flags, psl, pslFh, scoreFh);
pslFree(&psl);
}
struct hash *hashFaNames(char *fileName)
/* Return a hash full of the names (but not sequence) of all
* records in fa file. */
{
struct hash *hash = hashNew(0);
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *line;
while (lineFileNext(lf, &line, NULL))
{
line = skipLeadingSpaces(line);
if (line[0] == '>')
{
line += 1;
char *name = firstWordInLine(line);
if (name == NULL || name[0] == 0)
errAbort("Empty name in fasta file line %d of %s", lf->lineIx, lf->fileName);
if (hashLookup(hash, name))
errAbort("Duplicate name %s in fasta file line %d of %s",
name, lf->lineIx, lf->fileName);
hashAdd(hash, name, NULL);
}
}
lineFileClose(&lf);
return hash;
}
void oneGenieFile(char *fileName, struct hash *uniq, FILE *f)
/* Process one genie peptide prediction file into known and alt tab files. */
{
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *line;
int lineSize;
boolean firstTime = TRUE;
char *trans;
boolean skip = FALSE;
/* Do cursory sanity check. */
if (!lineFileNext(lf, &line, &lineSize))
errAbort("%s is empty", fileName);
if (line[0] != '>')
errAbort("%s is badly formatted, doesn't begin with '>'", fileName);
lineFileReuse(lf);
while (lineFileNext(lf, &line, &lineSize))
{
if (line[0] == '>')
{
/* End last line. */
if (firstTime)
firstTime = FALSE;
else
fputc('\n', f);
trans = firstWordInLine(line+1);
if (abbr != NULL && startsWith(abbr, trans))
trans += strlen(abbr);
if (hashLookupUpperCase(uniq, trans) != NULL)
{
warn("Duplicate (case insensitive) '%s' line %d of %s. Ignoring all but first.", trans, lf->lineIx, lf->fileName);
skip = TRUE;
}
else
{
char *upperCase;
upperCase = cloneString(trans);
touppers(upperCase);
hashAdd(uniq, upperCase, NULL);
freeMem(upperCase);
fprintf(f, "%s\t", trans);
skip = FALSE;
}
}
else if (!skip)
{
mustWrite(f, line, lineSize-1);
}
}
fputc('\n', f);
lineFileClose(&lf);
}
void genericOne(char *fileName, struct hash *uniq, FILE *f)
/* Process one ensemble peptide prediction file into tab delimited
* output f, using uniq hash to make sure no dupes. */
{
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *line;
int lineSize;
boolean firstTime = TRUE;
char *trans, transBuf[128];
/* Do cursory sanity check. */
if (!lineFileNext(lf, &line, &lineSize))
errAbort("%s is empty", fileName);
if (line[0] != '>')
errAbort("%s is badly formatted, doesn't begin with '>'", fileName);
lineFileReuse(lf);
while (lineFileNext(lf, &line, &lineSize))
{
if (line[0] == '>')
{
char *upperCase;
/* End last line. */
if (firstTime)
firstTime = FALSE;
else
fputc('\n', f);
trans = firstWordInLine(line+1);
if (abbr != NULL && startsWith(abbr, trans))
trans += strlen(abbr);
if (suffix != NULL)
{
safef(transBuf, sizeof(transBuf), "%s%s", trans, suffix);
trans = transBuf;
}
if (hashLookupUpperCase(uniq, trans) != NULL)
errAbort("Duplicate (case insensitive) '%s' line %d of %s", trans, lf->lineIx, lf->fileName);
upperCase = cloneString(trans);
touppers(upperCase);
hashAdd(uniq, upperCase, NULL);
freeMem(upperCase);
fprintf(f, "%s\t", trans);
}
else
{
mustWrite(f, line, lineSize-1);
}
}
fputc('\n', f);
lineFileClose(&lf);
}
static Fastq* fastq_processNextSequence (int freeMemory, int truncateName)
{
char *line;
static Fastq* currFQ = NULL;
int count;
Seq* currSeq = NULL;
if (ls_isEof (lsFastq)) {
if (freeMemory) {
fastq_freeFastq (currFQ);
}
return NULL;
}
count = 0;
while ( (line=ls_nextLine (lsFastq)) && (count<4) ) {
if (line[0] == '\0') {
continue;
}
if (line[0] == '@') {
if (freeMemory) {
fastq_freeFastq (currFQ);
}
count++;
AllocVar (currFQ);
AllocVar (currFQ->seq);
currSeq = currFQ->seq;
currSeq->name = hlr_strdup (line + 1);
if (truncateName) {
currSeq->name = firstWordInLine (skipLeadingSpaces (currSeq->name));
}
line = ls_nextLine (lsFastq); // reading sequence
currSeq->sequence = hlr_strdup ( line );
currSeq->size = strlen (currSeq->sequence);
count++;
line = ls_nextLine (lsFastq); // reading quality ID
if( line[0] != '+' )
die("Expected quality ID: '+' or '+%s'", currSeq->name );
count++;
line = ls_nextLine (lsFastq); // reading quality
currFQ->quality = hlr_strdup( line );
count++;
}
}
ls_back (lsFastq,1);
return currFQ;
}
boolean faMixedSpeedReadNext(struct lineFile *lf, DNA **retDna, int *retSize, char **retName)
/* Read in DNA or Peptide FA record in mixed case. Allow any upper or lower case
* letter, or the dash character in. */
{
char c;
int bufIx = 0;
static char name[512];
int lineSize, i;
char *line;
dnaUtilOpen();
/* Read first line, make sure it starts with '>', and read first word
* as name of sequence. */
name[0] = 0;
if (!lineFileNext(lf, &line, &lineSize))
{
*retDna = NULL;
*retSize = 0;
return FALSE;
}
if (line[0] == '>')
{
line = firstWordInLine(skipLeadingSpaces(line+1));
if (line == NULL)
errAbort("Expecting sequence name after '>' line %d of %s", lf->lineIx, lf->fileName);
strncpy(name, line, sizeof(name));
name[sizeof(name)-1] = '\0'; /* Just to make sure name is NULL terminated. */
}
else
{
errAbort("Expecting '>' line %d of %s", lf->lineIx, lf->fileName);
}
/* Read until next '>' */
for (;;)
{
if (!lineFileNext(lf, &line, &lineSize))
break;
if (line[0] == '>')
{
lineFileReuse(lf);
break;
}
if (bufIx + lineSize >= faFastBufSize)
expandFaFastBuf(bufIx, lineSize);
for (i=0; i<lineSize; ++i)
{
c = line[i];
if (isalpha(c) || c == '-')
faFastBuf[bufIx++] = c;
}
}
if (bufIx >= faFastBufSize)
expandFaFastBuf(bufIx, 0);
faFastBuf[bufIx] = 0;
*retDna = faFastBuf;
*retSize = bufIx;
*retName = name;
if (bufIx == 0)
{
warn("Invalid fasta format: sequence size == 0 for element %s",name);
}
return TRUE;
}
struct trackHubSettingSpec *trackHubSettingsForVersion(char *specHost, char *version)
/* Return list of settings with support level. Version can be version string or spec url */
{
struct htmlPage *page = NULL;
if (version == NULL)
{
version = trackHubVersionDefault(specHost, &page);
if (version == NULL)
errAbort("Can't get default spec from host %s", specHost);
}
/* Retrieve specs from file url.
* Settings are the first text word within any <code> tag having class="level-" attribute.
* The level represents the level of support for the setting (e.g. base, full, deprecated)
* The support level ('level-*') is the class value of the * <code> tag.
* E.g. <code class="level-full">boxedConfig on</code> produces:
* setting=boxedConfig, class=full */
if (page == NULL)
page = trackHubVersionSpecMustGet(specHost, version);
if (page == NULL)
errAbort("Can't get settings spec for version %s from host %s", version, specHost);
verbose(5, "Found %d tags\n", slCount(page->tags));
struct trackHubSettingSpec *spec, *savedSpec;
struct hash *specHash = hashNew(0);
struct htmlTag *tag;
struct htmlAttribute *attr;
char buf[256];
for (tag = page->tags; tag != NULL; tag = tag->next)
{
if (differentWord(tag->name, "code"))
continue;
attr = tag->attributes;
if (attr == NULL || differentString(attr->name, "class") || !startsWith("level-", attr->val))
continue;
AllocVar(spec);
int len = min(tag->next->start - tag->end, sizeof buf - 1);
memcpy(buf, tag->end, len);
buf[len] = 0;
verbose(6, "Found spec: %s\n", buf);
spec->name = cloneString(firstWordInLine(buf));
if (spec->name == NULL || strlen(spec->name) == 0)
{
warn("ERROR: format problem with trackDbHub.html -- contact UCSC.");
continue;
}
spec->level = cloneString(chopPrefixAt(attr->val, '-'));
verbose(6, "spec: name=%s, level=%s\n", spec->name, spec->level);
savedSpec = (struct trackHubSettingSpec *)hashFindVal(specHash, spec->name);
if (savedSpec != NULL)
verbose(6, "found spec %s level %s in hash\n", savedSpec->name, savedSpec->level);
if (savedSpec == NULL)
{
hashAdd(specHash, spec->name, spec);
verbose(6, "added spec %s at level %s\n", spec->name, spec->level);
}
else if (trackHubSettingLevelCmp(spec, savedSpec) > 0)
{
hashReplace(specHash, spec->name, spec);
verbose(6, "replaced spec %s at level %s, was %s\n",
spec->name, spec->level, savedSpec->level);
}
}
struct hashEl *el, *list = hashElListHash(specHash);
int settingsCt = slCount(list);
verbose(5, "Found %d settings's\n", slCount(list));
if (settingsCt == 0)
errAbort("Can't find hub setting info for version %s (host %s)."
" Use -version to indicate a different version number or url.", version, specHost);
slSort(&list, hashElCmp);
struct trackHubSettingSpec *specs = NULL;
int baseCt = 0;
int requiredCt = 0;
int deprecatedCt = 0;
for (el = list; el != NULL; el = el->next)
{
if (sameString(((struct trackHubSettingSpec *)el->val)->level, "base"))
baseCt++;
else if (sameString(((struct trackHubSettingSpec *)el->val)->level, "required"))
requiredCt++;
else if (sameString(((struct trackHubSettingSpec *)el->val)->level, "deprecated"))
deprecatedCt++;
slAddHead(&specs, el->val);
}
slReverse(&specs);
verbose(3,
"Found %d supported settings for this version (%d required, %d base, %d deprecated)\n",
slCount(specs), requiredCt, baseCt, deprecatedCt);
return specs;
}
void ctgFaToFa(char *ctgFa, char *ctgCoords, char *ntDir)
/* ctgFaToFa - Convert from one big file with all NT contigs to one contig per file.. */
{
struct lineFile *lf;
char fileName[512], *line;
char *ntName, *hsName;
char *parts[6];
int lineSize, partCount;
struct hash *uniqHash = newHash(0);
FILE *f = NULL;
int dotMod = 0;
struct hash *ntHash = newHash(0);
struct hash *hsHash = newHash(0);
struct ntContig *nt;
char *words[8];
printf("Loading %s\n", ctgCoords);
lf = lineFileOpen(ctgCoords, TRUE);
while (lineFileRow(lf, words))
{
ntName = words[0];
if ((nt = hashFindVal(ntHash, ntName)) != NULL)
++nt->cloneCount;
else
{
AllocVar(nt);
hashAddSaveName(ntHash, ntName, nt, &nt->name);
hashAddSaveName(hsHash, words[1], nt, &nt->hsName);
nt->cloneCount = 1;
}
}
lineFileClose(&lf);
lf = lineFileOpen(ctgFa, FALSE);
makeDir(ntDir);
while (lineFileNext(lf, &line, &lineSize))
{
if ((++dotMod&0x1ffff) == 0)
{
printf(".");
fflush(stdout);
}
if (line[0] == '>')
{
carefulClose(&f);
line[lineSize-1] = 0;
partCount = chopByChar(line, '|',parts,ArraySize(parts));
if (partCount < 3)
{
uglyf("partCount = %d\n", partCount);
errAbort("Expecting | separated header line %d of %s", lf->lineIx, lf->fileName);
}
ntName = parts[1];
nt = hashFindVal(ntHash, ntName);
hsName = parts[2];
if (nt == NULL)
{
hsName = firstWordInLine(ntName);
nt = hashMustFindVal(hsHash, hsName);
ntName = nt->name;
}
if (nt->cloneCount > 1)
{
if (!startsWith("Hs", hsName))
errAbort("Expecting %s to start with 'Hs' line %d of %s",
hsName, lf->lineIx, lf->fileName);
if (hashLookup(uniqHash, ntName))
ntName = nextFakeNtName(hsName, ntName);
hashAddUnique(uniqHash, ntName, NULL);
if (!startsWith("NT_", ntName))
errAbort("Expecting NT_ name line %d of %s", lf->lineIx, lf->fileName);
sprintf(fileName, "%s/%s.fa", ntDir, ntName);
f = mustOpen(fileName, "w");
fprintf(f, ">%s.1_1\n", ntName);
}
}
else
{
if (f != NULL)
mustWrite(f, line, lineSize);
}
}
printf("\n");
carefulClose(&f);
lineFileClose(&lf);
}
boolean faSpeedReadNextKeepCase(struct lineFile *lf,
DNA **retDna, int *retSize, char **retName)
/* Read in next FA entry as fast as we can. Faster than that old,
* pokey faFastReadNext. Return FALSE at EOF.
* The returned DNA and name will be overwritten by the next call
* to this function. */
{
int c;
int bufIx = 0;
static char name[256];
int lineSize, i;
char *line;
dnaUtilOpen();
/* Read first line, make sure it starts wiht '>', and read first word
* as name of sequence. */
name[0] = 0;
if (!lineFileNext(lf, &line, &lineSize))
{
*retDna = NULL;
*retSize = 0;
return FALSE;
}
if (line[0] == '>')
{
line = firstWordInLine(skipLeadingSpaces(line+1));
if (line == NULL)
errAbort("Expecting sequence name after '>' line %d of %s", lf->lineIx, lf->fileName);
strncpy(name, line, sizeof(name));
}
else
{
errAbort("Expecting '>' line %d of %s", lf->lineIx, lf->fileName);
}
/* Read until next '>' */
for (;;)
{
if (!lineFileNext(lf, &line, &lineSize))
break;
if (line[0] == '>')
{
lineFileReuse(lf);
break;
}
if (bufIx + lineSize >= faFastBufSize)
expandFaFastBuf(bufIx);
for (i=0; i<lineSize; ++i)
{
c = line[i];
if (isalpha(c))
{
faFastBuf[bufIx++] = c;
}
}
}
if (bufIx >= faFastBufSize)
expandFaFastBuf(bufIx);
faFastBuf[bufIx] = 0;
*retDna = faFastBuf;
*retSize = bufIx;
*retName = name;
return TRUE;
}
static struct jsonWrite *rTdbToJw(struct trackDb *tdb, struct hash *fieldHash,
struct hash *excludeTypesHash, int depth, int maxDepth)
/* Recursively build and return a new jsonWrite object with JSON for tdb and its children,
* or NULL if tdb or all children have been filtered out by excludeTypesHash.
* If excludeTypesHash is non-NULL, omit any tracks/views/subtracks with type in excludeTypesHash.
* If fieldHash is non-NULL, include only the field names indexed in fieldHash. */
{
if (maxDepth >= 0 && depth > maxDepth)
return NULL;
boolean doSubtracks = (tdb->subtracks && fieldOk("subtracks", fieldHash));
// If excludeTypesHash is given and tdb is a leaf track/subtrack, look up the first word
// of tdb->type in excludeTypesHash; if found, return NULL.
if (excludeTypesHash && !doSubtracks)
{
char typeCopy[PATH_LEN];
safecpy(typeCopy, sizeof(typeCopy), tdb->type);
if (hashLookup(excludeTypesHash, firstWordInLine(typeCopy)))
return NULL;
}
boolean gotSomething = !doSubtracks;
struct jsonWrite *jwNew = jsonWriteNew();
jsonWriteObjectStart(jwNew, NULL);
writeTdbSimple(jwNew, tdb, fieldHash);
if (tdb->parent && fieldOk("parent", fieldHash))
{
// We can't link to an object in JSON and better not recurse here or else infinite loop.
if (tdbIsSuperTrackChild(tdb))
{
// Supertracks have been omitted from fullTrackList, so add the supertrack object's
// non-parent/child info here.
jsonWriteObjectStart(jwNew, "parent");
writeTdbSimple(jwNew, tdb->parent, fieldHash);
jsonWriteObjectEnd(jwNew);
}
else
// Just the name so we don't have infinite loops.
jsonWriteString(jwNew, "parent", tdb->parent->track);
}
if (doSubtracks)
{
jsonWriteListStart(jwNew, "subtracks");
slSort(&tdb->subtracks, trackDbViewCmp);
struct trackDb *subTdb;
for (subTdb = tdb->subtracks; subTdb != NULL; subTdb = subTdb->next)
{
struct jsonWrite *jwSub = rTdbToJw(subTdb, fieldHash, excludeTypesHash, depth+1, maxDepth);
if (jwSub)
{
gotSomething = TRUE;
jsonWriteAppend(jwNew, NULL, jwSub);
jsonWriteFree(&jwSub);
}
}
jsonWriteListEnd(jwNew);
}
jsonWriteObjectEnd(jwNew);
if (! gotSomething)
// All children were excluded; clean up and null out jwNew.
jsonWriteFree(&jwNew);
return jwNew;
}
void splitByBase(char *inName, int splitCount, char *outRoot, off_t estSize)
/* Split into a file base by base. */
{
struct lineFile *lf = lineFileOpen(inName, TRUE);
int lineSize;
char *line;
char c;
char dir[PATH_LEN], seqName[128], outFile[128], outPathName[PATH_LEN];
int digits = digitsBaseTen(splitCount);
boolean warnedMultipleRecords = FALSE;
int fileCount = 0;
off_t nextEnd = 0;
off_t curPos = 0;
FILE *f = NULL;
int linePos = 0;
int outLineSize = 50;
if (!lineFileNext(lf, &line, &lineSize))
errAbort("%s is empty", inName);
if (line[0] == '>')
{
line = firstWordInLine(line+1);
if (line == NULL)
errAbort("Empty initial '>' line in %s", inName);
strncpy(seqName, line, sizeof(seqName));
}
else
{
splitPath(inName, dir, seqName, NULL);
lineFileReuse(lf);
}
splitPath(outRoot, NULL, outFile, NULL);
while (lineFileNext(lf, &line, &lineSize))
{
if (line[0] == '>')
{
if (!warnedMultipleRecords)
{
warnedMultipleRecords = TRUE;
warn("More than one record in FA file line %d of %s",
lf->lineIx, lf->fileName);
continue;
}
}
while ((c = *line++) != 0)
{
if (isdigit(c) || isspace(c))
continue;
if (!isalpha(c))
errAbort("Weird %c (0x%x) line %d of %s", c, c, lf->lineIx, lf->fileName);
if (++curPos >= nextEnd)
{
if (f != NULL)
{
if (linePos != 0)
fputc('\n', f);
fclose(f);
}
mkOutPath(outPathName, outRoot, digits, fileCount);
verbose(2, "writing %s\n", outPathName);
f = mustOpen(outPathName, "w");
fprintf(f, ">%s%0*d\n", outFile, digits, fileCount);
++fileCount;
linePos = 0;
nextEnd = calcNextEnd(fileCount, splitCount, estSize);
}
fputc(c, f);
if (++linePos >= outLineSize)
{
fputc('\n', f);
linePos = 0;
}
}
}
if (f != NULL)
{
if (linePos != 0)
fputc('\n', f);
fclose(f);
}
lineFileClose(&lf);
}
void faNcbiToUcsc(char *inFile, char *out)
/* faNcbiToUcsc - Convert FA file from NCBI to UCSC format.. */
{
struct lineFile *lf = lineFileOpen(inFile, TRUE);
char outName[512];
char *line;
boolean split = cgiBoolean("split");
boolean ntLast = cgiBoolean("ntLast");
boolean encode = cgiBoolean("encode");
struct dnaSeq seq;
FILE *f = NULL;
char *wordBefore = cgiUsualString("wordBefore", "gb");
int wordIx = cgiUsualInt("wordIx", -1);
char *e = NULL;
char *nt = NULL;
ZeroVar(&seq);
if (split)
makeDir(out);
else
f = mustOpen(out, "w");
while (lineFileNext(lf, &line, NULL))
{
if (line[0] == '>')
{
if (ntLast || encode)
{
nt = NULL;
if (ntLast)
{
e = NULL;
nt = stringIn("NT_", line);
if (nt == NULL)
nt = stringIn("NG_", line);
if (nt == NULL)
nt = stringIn("NC_", line);
if (nt == NULL)
errAbort("Expecting NT_ NG_ or NC_in '%s'", line);
e = strchr(nt, '|');
if (e != NULL) *e = 0;
e = strchr(nt, ' ');
if (e != NULL) *e = 0;
}
else
{
nt = stringIn("|EN", line);
if (nt == NULL)
errAbort("Expecting EN in %s", line);
nt++;
nt = firstWordInLine(nt);
}
if (split)
{
sprintf(outName, "%s/%s.fa", out, nt);
carefulClose(&f);
f = mustOpen(outName, "w");
}
fprintf(f, ">%s\n", nt);
}
else
{
char *words[32];
int wordCount, i;
char *accession = NULL;
wordCount = chopString(line+1, "|", words, ArraySize(words));
if (wordIx >= 0)
{
if (wordIx >= wordCount)
errAbort("Sorry only %d words", wordCount);
accession = words[wordIx];
}
else
{
for (i=0; i<wordCount-1; ++i)
{
if (sameString(words[i], wordBefore))
{
accession = words[i+1];
break;
}
}
if (accession == NULL)
errAbort("Couldn't find '%s' line %d of %s",
wordBefore, lf->lineIx, lf->fileName);
}
chopSuffix(accession);
fprintf(f, ">%s\n", accession);
}
}
else
{
fprintf(f, "%s\n", line);
}
}
}
void hgFlyBase(char *database, char *genesFile)
/* hgFlyBase - Parse FlyBase genes.txt file and turn it into a couple of
* tables. */
{
char *tGene = "fbGene";
char *tSynonym = "fbSynonym";
char *tAllele = "fbAllele";
char *tRef = "fbRef";
char *tRole = "fbRole";
char *tPhenotype = "fbPhenotype";
char *tTranscript = "fbTranscript";
char *tGo = "fbGo";
char *tUniProt = "fbUniProt";
FILE *fGene = hgCreateTabFile(tabDir, tGene);
FILE *fSynonym = hgCreateTabFile(tabDir, tSynonym);
FILE *fAllele = hgCreateTabFile(tabDir, tAllele);
FILE *fRef = hgCreateTabFile(tabDir, tRef);
FILE *fRole = hgCreateTabFile(tabDir, tRole);
FILE *fPhenotype = hgCreateTabFile(tabDir, tPhenotype);
FILE *fTranscript = NULL;
FILE *fGo = hgCreateTabFile(tabDir, tGo);
FILE *fUniProt = hgCreateTabFile(tabDir, tUniProt);
struct lineFile *lf = lineFileOpen(genesFile, TRUE);
struct hash *refHash = newHash(19);
int nextRefId = 0;
int nextAlleleId = 0;
char *line, sub, type, *rest, *s;
char *geneSym = NULL, *geneName = NULL, *geneId = NULL;
int recordCount = 0;
struct slName *synList = NULL, *syn;
int curAllele = 0, curRef = 0;
struct ref *ref = NULL;
struct sqlConnection *conn;
struct hash *goUniqHash = newHash(18);
/* Make table from flybase genes to BGDP transcripts. */
if (doTranscript)
{
fTranscript = hgCreateTabFile(tabDir, tTranscript);
getAllSplices(database, fTranscript);
}
/* Make dummy reference for flybase itself. */
fprintf(fRef, "0\tFlyBase\n");
/* Loop through parsing and writing tab files. */
while (lineFileNext(lf, &line, NULL))
{
sub = line[0];
if (sub == '#')
{
/* End of record. */
++recordCount;
if (geneId == NULL)
errAbort("Record without *z line ending line %d of %s",
lf->lineIx, lf->fileName);
/* Write out synonyms. */
s = naForNull(geneSym);
geneSym = ungreek(s);
freeMem(s);
s = naForNull(geneName);
geneName = ungreek(s);
if (! sameString(s, "n/a"))
freeMem(s);
if (geneSym != NULL && !sameString(geneSym, "n/a"))
slNameStore(&synList, geneSym);
if (geneName != NULL && !sameString(geneName, "n/a"))
slNameStore(&synList, geneName);
for (syn = synList; syn != NULL; syn = syn->next)
{
s = ungreek(syn->name);
fprintf(fSynonym, "%s\t%s\n", geneId, s);
freeMem(s);
}
/* Write out gene record. */
fprintf(fGene, "%s\t%s\t%s\n", geneId, geneSym, geneName);
/* Clean up. */
freez(&geneSym);
freez(&geneName);
freez(&geneId);
slFreeList(&synList);
ref = NULL;
curRef = curAllele = 0;
continue;
}
else if (sub == 0)
errAbort("blank line %d of %s, not allowed in gene.txt",
lf->lineIx, lf->fileName);
else if (isalnum(sub))
errAbort("line %d of %s begins with %c, not allowed",
lf->lineIx, lf->fileName, sub);
type = line[1];
rest = trimSpaces(line+2);
if (sub == '*' && type == 'a')
geneSym = cloneString(rest);
else if (sub == '*' && type == 'e')
geneName = cloneString(rest);
else if (sub == '*' && type == 'z')
{
geneId = cloneString(rest);
if (!startsWith("FBgn", geneId))
errAbort("Bad FlyBase gene ID %s line %d of %s", geneId,
lf->lineIx, lf->fileName);
}
else if (type == 'i' && (sub == '*' || sub == '$'))
{
if (strlen(rest) > 2) /* Avoid short useless ones. */
slNameStore(&synList, rest);
}
else if (sub == '*' && type == 'A')
{
if (geneId == NULL)
errAbort("Allele before geneId line %d of %s",
lf->lineIx, lf->fileName);
curAllele = ++nextAlleleId;
fprintf(fAllele, "%d\t%s\t%s\n", curAllele, geneId, rest);
if (!sameString(rest, "classical") &&
!sameString(rest, "in vitro") &&
!sameString(rest, "wild-type") )
{
slNameStore(&synList, rest);
}
}
else if (sub == '*' && type == 'm')
{
if (geneId == NULL)
errAbort("*m protein ID before geneId line %d of %s",
lf->lineIx, lf->fileName);
if (startsWith("UniProt", rest))
{
char *ptr = strchr(rest, ':');
if (ptr != NULL)
ptr++;
else
errAbort("Trouble parsing UniProt ID %s like %d of %s",
rest, lf->lineIx, lf->fileName);
fprintf(fUniProt, "%s\t%s\n", geneId, ptr);
}
}
else if (type == 'E')
{
ref = hashFindVal(refHash, rest);
if (ref == NULL)
{
AllocVar(ref);
ref->id = ++nextRefId;
hashAdd(refHash, rest, ref);
subChar(rest, '\t', ' ');
fprintf(fRef, "%d\t%s\n", ref->id, rest);
}
curRef = ref->id;
}
else if ((type == 'k' || type == 'r' || type == 'p') && sub != '@')
{
FILE *f = (type == 'r' ? fRole : fPhenotype);
struct dyString *dy = suckSameLines(lf, line);
subChar(dy->string, '\t', ' ');
if (geneId == NULL)
errAbort("Expecting *z in record before line %d of %s",
lf->lineIx, lf->fileName);
fprintf(f, "%s\t%d\t%d\t%s\n", geneId, curAllele, curRef, dy->string);
dyStringFree(&dy);
}
else if (type == 'd' || type == 'f' || type == 'F')
{
FILE *f = fGo;
char aspect = (type == 'd') ? 'P' : (type == 'f') ? 'C' : 'F';
char *goId = rest;
char *p = strstr(goId, " ; ");
char assoc[128];
if (p == NULL)
continue;
else
goId = firstWordInLine(p + 3);
safef(assoc, sizeof(assoc), "%s.%s", geneId, goId);
if (hashLookup(goUniqHash, assoc) == NULL)
{
hashAddInt(goUniqHash, assoc, 1);
fprintf(f, "%s\t%s\t%c\n", geneId, goId, aspect);
}
}
}
printf("Processed %d records in %d lines\n", recordCount, lf->lineIx);
lineFileClose(&lf);
conn = sqlConnect(database);
remakeTables(conn);
if (doLoad)
{
printf("Loading %s\n", tGene);
hgLoadTabFile(conn, tabDir, tGene, &fGene);
if (doTranscript)
{
printf("Loading %s\n", tTranscript);
hgLoadTabFile(conn, tabDir, tTranscript, &fTranscript);
}
printf("Loading %s\n", tSynonym);
hgLoadTabFile(conn, tabDir, tSynonym, &fSynonym);
printf("Loading %s\n", tAllele);
hgLoadTabFile(conn, tabDir, tAllele, &fAllele);
printf("Loading %s\n", tRef);
hgLoadTabFile(conn, tabDir, tRef, &fRef);
printf("Loading %s\n", tRole);
hgLoadTabFile(conn, tabDir, tRole, &fRole);
printf("Loading %s\n", tPhenotype);
hgLoadTabFile(conn, tabDir, tPhenotype, &fPhenotype);
printf("Loading %s\n", tGo);
hgLoadTabFile(conn, tabDir, tGo, &fGo);
printf("Loading %s\n", tUniProt);
hgLoadTabFile(conn, tabDir, tUniProt, &fUniProt);
hgRemoveTabFile(tabDir, tGene);
if (doTranscript)
hgRemoveTabFile(tabDir, tTranscript);
hgRemoveTabFile(tabDir, tSynonym);
hgRemoveTabFile(tabDir, tAllele);
hgRemoveTabFile(tabDir, tRef);
hgRemoveTabFile(tabDir, tRole);
hgRemoveTabFile(tabDir, tPhenotype);
hgRemoveTabFile(tabDir, tGo);
hgRemoveTabFile(tabDir, tUniProt);
}
}