struct mapPos *readInfoFile(char *mapName)
/* Read maps from file. */
{
struct lineFile *lf = lineFileOpen(mapName, TRUE);
int lineSize, wordCount;
char *line, *words[16];
struct mapPos *list = NULL, *el;
lineFileNeedNext(lf, &line, &lineSize);
while (lineFileNext(lf, &line, &lineSize))
{
if (line[0] == '#')
continue;
wordCount = chopLine(line, words);
lineFileExpectWords(lf, 3, wordCount);
AllocVar(el);
el->cloneName = cloneString(words[0]);
el->pos = atoi(words[1]);
el->phase = atoi(words[2]);
slAddHead(&list, el);
}
lineFileClose(&lf);
slReverse(&list);
return list;
}
void makeMotifs(char *inFile, struct hash *tfHash, char *outFile)
/* Parse input motifs and save them to outFile in dnaMotif format. */
{
struct lineFile *lf = lineFileOpen(inFile, TRUE);
FILE *f = mustOpen(outFile, "w");
struct hashEl *hel;
for (;;)
{
char *line;
char *words[256], *word;
int wordCount;
struct dnaMotif *motif;
if (!lineFileSkipTo(lf, "Probability matrix for"))
break;
lineFileNeedNext(lf, &line, NULL);
wordCount = chopLine(line, words);
if (wordCount >= ArraySize(words))
errAbort("Line %d of %s is too long\n", lf->lineIx, lf->fileName);
if (!sameString(words[0], "#"))
badFormat(lf);
AllocVar(motif);
motif->columnCount = wordCount-1;
readBaseProbs(lf, words, "#A", &motif->aProb, motif->columnCount);
readBaseProbs(lf, words, "#C", &motif->cProb, motif->columnCount);
readBaseProbs(lf, words, "#T", &motif->tProb, motif->columnCount);
readBaseProbs(lf, words, "#G", &motif->gProb, motif->columnCount);
if (!lineFileSkipTo(lf, "Source:"))
lineFileUnexpectedEnd(lf);
lineFileReuse(lf);
lineFileNeedNext(lf, &line, NULL);
word = nextWord(&line);
word = nextWord(&line);
if (word == NULL)
errAbort("Short Source: line %d of %s", lf->lineIx, lf->fileName);
motif->name = cloneString(word);
hel = hashLookup(tfHash, motif->name);
if (hel == NULL)
errAbort("%s in %s but not GFFs", motif->name, lf->fileName);
hel->val = motif;
dnaMotifTabOut(motif, f);
}
carefulClose(&f);
lineFileClose(&lf);
}
void checkInputOpenFiles(struct inInfo *array, int count)
/* Make sure all of the input is there and of right format before going forward. Since
* this is going to take a while we want to fail fast. */
{
int i;
for (i=0; i<count; ++i)
{
struct inInfo *in = &array[i];
switch (in->type)
{
case itBigWig:
{
/* Just open and close, it will abort if any problem. */
in->bbi = bigWigFileOpen(in->fileName);
break;
}
case itPromoterBed:
case itUnstrandedBed:
case itBlockedBed:
{
struct lineFile *lf = in->lf = lineFileOpen(in->fileName, TRUE);
char *line;
lineFileNeedNext(lf, &line, NULL);
char *dupe = cloneString(line);
char *row[256];
int wordCount = chopLine(dupe, row);
struct bed *bed = NULL;
switch (in->type)
{
case itPromoterBed:
lineFileExpectAtLeast(lf, 6, wordCount);
bed = bedLoadN(row, 6);
char strand = bed->strand[0];
if (strand != '+' && strand != '-')
errAbort("%s must be stranded, got %s in that field", lf->fileName, row[6]);
break;
case itUnstrandedBed:
lineFileExpectAtLeast(lf, 4, wordCount);
bed = bedLoadN(row, 4);
break;
case itBlockedBed:
lineFileExpectAtLeast(lf, 4, wordCount);
bed = bedLoadN(row, 12);
break;
default:
internalErr();
break;
}
bedFree(&bed);
freez(&dupe);
lineFileReuse(lf);
break;
}
default:
internalErr();
break;
}
}
}
int countWcDiff(char *fileName)
/** Count how many lines counted are reported by 'wc'. */
{
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *line;
int lineSize;
int numLineDiff=0;
char *row[3];
lineFileNeedNext(lf, &line, &lineSize);
line = trimSpaces(line);
chopString(line, " ", row, 3);
numLineDiff=atoi(row[0]);
lineFileClose(&lf);
return numLineDiff;
}
void rmskOut2OpenVerify(char *fileName, struct lineFile **retFile, boolean *retEmpty)
/* Open repeat masker .out file and verify that it is good.
* Set retEmpty if it has header characteristic of an empty file. */
{
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *line;
int lineSize;
lineFileNeedNext(lf, &line, &lineSize);
if (startsWith("There were no", line))
*retEmpty = TRUE;
line = skipLeadingSpaces(line);
if (! ( startsWith("SW", line) || startsWith("bit", line) ) )
errAbort("%s doesn't seem to be a RepeatMasker .out file", fileName);
lineFileSkip(lf, 2);
*retEmpty = FALSE;
*retFile = lf;
}
void readBaseProbs(struct lineFile *lf, char **words,
char *firstWord, float **pArray, int colCount)
/* Allocate and read base probabilities. */
{
char *line;
int wordCount;
float *array;
int i;
lineFileNeedNext(lf, &line, NULL);
wordCount = chopByWhite(line, words, colCount+1);
lineFileExpectWords(lf, colCount+1, wordCount);
if (!sameString(words[0], firstWord))
errAbort("Expecting %s, got %s line %d of %s", firstWord, words[0],
lf->lineIx, lf->fileName);
AllocArray(array, colCount);
for (i=0; i<colCount; ++i)
array[i] = atof(words[i+1]);
*pArray = array;
}
struct segFile *segMayOpen(char *fileName)
/* Open up a segment file for reading. Read header and verify. Prepare
* for subsequent calls to segNext(). Return NULL if file does not exist. */
{
struct segFile *sf;
struct lineFile *lf;
char *line, *name, *val, *word;
char *sig = "##seg";
/* Open fileName. */
if ((lf = lineFileMayOpen(fileName, TRUE)) == NULL)
return NULL;
AllocVar(sf);
sf->lf = lf;
/* Check for a valid signature. */
lineFileNeedNext(lf, &line, NULL);
if (!startsWith(sig, line))
errAbort("%s does not start with %s", fileName, sig);
line += strlen(sig);
/* parse name=val. */
while ((word = nextWord(&line)) != NULL)
{
name = word;
val = strchr(word, '=');
if (val == NULL)
errAbort("Missing = after %s line 1 of %s", name, fileName);
*val++ = 0;
if (sameString(name, "version"))
sf->version = atoi(val);
}
if (sf->version == 0)
errAbort("No version line 1 of %s", fileName);
return sf;
}
struct mafFile *mafMayOpen(char *fileName)
/* Open up a maf file and verify header. */
{
struct mafFile *mf;
struct lineFile *lf;
char *line, *word;
char *sig = "##maf";
if ((lf = lineFileMayOpen(fileName, TRUE)) == NULL)
return NULL;
AllocVar(mf);
mf->lf = lf;
lineFileNeedNext(lf, &line, NULL);
if (!startsWith(sig, line))
{
errAbort("%s does not start with %s", fileName, sig);
}
line += strlen(sig);
while ((word = nextWord(&line)) != NULL)
{
/* Parse name=val. */
char *name = word;
char *val = strchr(word, '=');
if (val == NULL)
errAbort("Missing = after %s line 1 of %s\n", name, fileName);
*val++ = 0;
if (sameString(name, "version"))
mf->version = atoi(val);
else if (sameString(name, "scoring"))
mf->scoring = cloneString(val);
}
if (mf->version == 0)
errAbort("No version line 1 of %s\n", fileName);
return mf;
}
void hgLoadRnaFold(char *database, char *table, char *foldDir)
/* hgLoadRnaFold - Load a directory full of RNA fold files into database. */
{
char path[PATH_LEN];
struct slName *dirList, *dirEl;
struct lineFile *lf;
char *line, *word, *s, c;
FILE *f = hgCreateTabFile(tabDir, table);
int count = 0;
dirList = listDir(foldDir, "*");
for (dirEl = dirList; dirEl != NULL; dirEl = dirEl->next)
{
char *name = dirEl->name;
if (sameString(name, "CVS"))
continue;
safef(path, sizeof(path), "%s/%s", foldDir, name);
lf = lineFileOpen(path, TRUE);
if (!lineFileNext(lf, &line, NULL))
{
if (warnEmpty)
{
warn("%s is empty, skipping\n", name);
lineFileClose(&lf);
continue;
}
else
errAbort("%s is empty\n", name);
}
if (!isupper(line[0]))
notFold(path, 1);
fprintf(f, "%s\t", name); /* Save name */
fprintf(f, "%s\t", line); /* Save sequence */
lineFileNeedNext(lf, &line, NULL);
c = line[0];
if (c != '.' && c != '(')
notFold(path, 2);
word = nextWord(&line);
fprintf(f, "%s\t", word); /* Save nested parenthesis */
/* Parse out (energy) term at end of line. */
s = strchr(line, '(');
if (s == NULL)
notFold(path, 3);
word = skipLeadingSpaces(s+1);
if (word == NULL || (!word[0] == '-' && !isdigit(word[0])))
notFold(path, 4);
if ((s = strchr(word, ')')) == NULL)
notFold(path, 5);
*s = 0;
fprintf(f, "%s\n", word);
lineFileClose(&lf);
++count;
}
printf("Parsed %d files\n", count);
if (doLoad)
{
struct sqlConnection *conn = sqlConnect(database);
rnaFoldCreateTable(conn, table);
hgLoadTabFile(conn, tabDir, table, &f);
hgRemoveTabFile(tabDir, table);
sqlDisconnect(&conn);
}
}
