void tagStormQueryMain(char *query)
/* tagStormQuery - Find stanzas in tag storm based on SQL-like query.. */
{
/* Get parsed out query */
struct lineFile *lf = lineFileOnString("query", TRUE, cloneString(query));
struct rqlStatement *rql = rqlStatementParse(lf);
int stormCount = slCount(rql->tableList);
if (stormCount != 1)
errAbort("Can only handle one tag storm file in query, got %d", stormCount);
char *tagsFileName = rql->tableList->name;
/* Read in tags */
struct tagStorm *tags = tagStormFromFile(tagsFileName);
/* Expand any field names with wildcards. */
struct slName *allFieldList = tagStormFieldList(tags);
rql->fieldList = wildExpandList(allFieldList, rql->fieldList, TRUE);
/* Traverse tree applying query */
struct lm *lm = lmInit(0);
doSelect = sameWord(rql->command, "select");
traverse(tags, tags->forest, rql, lm);
tagStormFree(&tags);
if (sameWord(rql->command, "count"))
printf("%d\n", matchCount);
}
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));
}
char *quotedPrintableDecode(char *input)
/* Use Quoted-Printable standard to decode a string. Return decoded
* string which will be freeMem'd. */
{
size_t inplen = strlen(input);
char *result = (char *)needMem(inplen+1);
size_t j=0;
char *line = NULL;
int size = 0;
int i = 0;
boolean newLine = FALSE;
struct lineFile *lf = lineFileOnString("", TRUE, cloneString(input));
while (lineFileNext(lf, &line, &size))
{
newLine = quotedPCollapse(line);
size = strlen(line);
for (i = 0; i < size; )
result[j++] = line[i++];
if (newLine)
result[j++] = '\n';
}
lineFileClose(&lf); /* frees cloned string */
result[j] = 0; /* terminate text string */
return result;
}
struct gapCalc *gapCalcFromString(char *s)
/* Return gapCalc from description string. */
{
struct lineFile *lf = lineFileOnString("string", TRUE, cloneString(s));
struct gapCalc *gapCalc = gapCalcRead(lf);
lineFileClose(&lf);
return gapCalc;
}
struct rqlStatement *rqlStatementParseString(char *string)
/* Return a parsed-out RQL statement based on string */
{
struct lineFile *lf = lineFileOnString("query", TRUE, cloneString(string));
struct rqlStatement *rql = rqlStatementParse(lf);
lineFileClose(&lf);
return rql;
}
struct asObject *asParseText(char *text)
/* Parse autoSql from text (as opposed to file). */
{
char *dupe = cloneString(text);
struct lineFile *lf = lineFileOnString("text", TRUE, dupe);
struct asObject *objList = asParseLineFile(lf);
freez(&dupe);
return objList;
}
void trySubmitUpload(struct sqlConnection *conn, char *rawText)
/* Called when they've submitted from uploads page */
{
struct lineFile *lf = lineFileOnString("uploaded data", TRUE, rawText);
struct customPp *cpp = customPpNew(lf);
struct hash *settings = hashNew(8);
addIfNonempty(settings, hggMinVal, "minVal");
addIfNonempty(settings, hggMaxVal, "maxVal");
addIfNonempty(settings, hggMaxGapToFill, "maxGapToFill");
addIfNonempty(settings, hggLabelVals, "linesAt");
struct customTrack *trackList = chromGraphParser(database, cpp,
cartUsualString(cart, hggFormatType, formatNames[0]),
cartUsualString(cart, hggMarkerType, cgfMarkerGenomic),
cartUsualString(cart, hggColumnLabels, colLabelNames[0]),
nullIfAllSpace(cartUsualString(cart, hggDataSetName, NULL)),
nullIfAllSpace(cartUsualString(cart, hggDataSetDescription, NULL)),
settings, TRUE);
updateCustomTracks(trackList);
}
static char *limitText(char *text)
/* read text string and limit to 1000 actual data lines */
{
struct dyString *limitedText = dyStringNew(0);
/* yes, opening with FALSE so as not to destroy the original string */
struct lineFile *lf = lineFileOnString("limitText", FALSE, text);
char *lineStart = NULL;
int lineLength = 0;
int legitimateLineCount = 0;
while (legitimateLineCount < 1000 && lineFileNext(lf, &lineStart, &lineLength))
{
char *s, c;
s = skipLeadingSpaces(lineStart);
c = s[0];
if (c != 0 && c != '#')
++legitimateLineCount;
dyStringAppendN(limitedText, lineStart, lineLength);
}
if ((legitimateLineCount == 1000) && lineFileNext(lf, &lineStart, &lineLength))
warn("WARNING: defined regions limit of 1000 definitions reached at line %d<BR>\n",
lf->lineIx-1);
lineFileClose(&lf);
return (dyStringCannibalize(&limitedText));
}
static struct bed4 *parseRegionInput(char *db, char *inputString, int maxRegions, int maxErrs,
struct dyString *dyWarn)
/* scan the user region definition, turn into a bed list */
{
int regionCount = 0;
int errCount = 0;
struct bed4 *bedList = NULL;
struct lineFile *lf = lineFileOnString("userData", TRUE, inputString);
char *line = NULL;
while (lineFileNextReal(lf, &line))
{
char *chromName = NULL;
int chromStart = 0;
int chromEnd = 0;
char *regionName = NULL;
// Chop a copy of line so we can display line if there's an error.
char copy[strlen(line)+1];
safecpy(copy, sizeof(copy), line);
char *words[5];
int wordCount = chopByWhite(copy, words, ArraySize(words));
boolean badFormat = FALSE;
boolean gotError = FALSE;
/* might be something of the form: chrom:start-end optionalRegionName */
if (((1 == wordCount) || (2 == wordCount)) &&
hgParseChromRange(NULL, words[0], &chromName,
&chromStart, &chromEnd))
{
if (2 == wordCount)
regionName = cloneString(words[1]);
}
else if (!((3 == wordCount) || (4 == wordCount)))
{
dyStringPrintf(dyWarn, "line %d: '%s': "
"unrecognized format. Please enter 3- or 4-column BED or "
"a chr:start-end position range optionally followed by a name.\n",
lf->lineIx, line);
badFormat = TRUE;
gotError = TRUE;
}
else
{
chromName = words[0];
// Make sure chromStart and chromEnd are numbers
if (!isNumericString(words[1]))
{
dyStringPrintf(dyWarn, "line %d: '%s': chromStart must be a number but is '%s'\n",
lf->lineIx, line, words[1]);
gotError = TRUE;
}
if (!isNumericString(words[2]))
{
dyStringPrintf(dyWarn, "line %d: '%s': chromEnd must be a number but is '%s'\n",
lf->lineIx, line, words[2]);
gotError = TRUE;
}
if (! gotError)
{
chromStart = atoi(words[1]);
chromEnd = atoi(words[2]);
if (wordCount > 3)
regionName = cloneString(words[3]);
}
}
char *officialChromName = chromName ? hgOfficialChromName(db, chromName) : NULL;
if (! badFormat)
{
if (NULL == officialChromName)
{
dyStringPrintf(dyWarn,
"line %d: '%s': chrom name '%s' not recognized in this assembly\n",
lf->lineIx, line, chromName ? chromName : words[0]);
gotError = TRUE;
}
else if (illegalCoordinate(db, officialChromName, chromStart, chromEnd, line, lf->lineIx,
dyWarn))
{
gotError = TRUE;
}
}
if (gotError)
{
errCount++;
if (errCount > maxErrs && maxErrs > 0)
{
dyStringPrintf(dyWarn, "Exceeded maximum number of errors (%d), quitting\n", maxErrs);
break;
}
else
continue;
}
++regionCount;
if (regionCount > maxRegions && maxRegions > 0)
{
dyStringPrintf(dyWarn,
"line %d: limit of %d region definitions exceeded, skipping the rest\n",
lf->lineIx, maxRegions);
break;
}
struct bed4 *bedEl = bed4New(officialChromName, chromStart, chromEnd, regionName);
slAddHead(&bedList, bedEl);
}
lineFileClose(&lf);
// Keep regions in same order as user entered them:
slReverse(&bedList);
return (bedList);
}
void doPastedIdentifiers(struct sqlConnection *conn)
/* Process submit in paste identifiers page. */
{
char *idText = trimSpaces(cartString(cart, hgtaPastedIdentifiers));
htmlOpen("Table Browser (Input Identifiers)");
if (isNotEmpty(idText))
{
/* Write terms to temp file, checking whether they have matches, and
* save temp file name. */
boolean saveIdText = (strlen(idText) < MAX_IDTEXT);
char *idTextForLf = saveIdText ? cloneString(idText) : idText;
struct lineFile *lf = lineFileOnString("idText", TRUE, idTextForLf);
char *line, *word;
struct tempName tn;
FILE *f;
int totalTerms = 0, foundTerms = 0;
struct slName* missingTerms = NULL;
struct dyString *exampleMissingIds = dyStringNew(256);
char *actualDb = database;
if (sameWord(curTable, WIKI_TRACK_TABLE))
actualDb = wikiDbName();
struct hTableInfo *hti = maybeGetHti(actualDb, curTable, conn);
char *idField = getIdField(actualDb, curTrack, curTable, hti);
if (idField == NULL)
{
warn("Sorry, I can't tell which field of table %s to treat as the "
"identifier field.", curTable);
webNewSection("Table Browser");
cartRemove(cart, hgtaIdentifierDb);
cartRemove(cart, hgtaIdentifierTable);
cartRemove(cart, hgtaIdentifierFile);
mainPageAfterOpen(conn);
htmlClose();
return;
}
struct slName *allTerms = NULL, *term;
while (lineFileNext(lf, &line, NULL))
{
while ((word = nextWord(&line)) != NULL)
{
term = slNameNew(word);
slAddHead(&allTerms, term);
totalTerms++;
}
}
slReverse(&allTerms);
lineFileClose(&lf);
char *extraWhere = NULL;
int maxIdsInWhere = cartUsualInt(cart, "hgt_maxIdsInWhere", DEFAULT_MAX_IDS_IN_WHERE);
if (totalTerms > 0 && totalTerms <= maxIdsInWhere)
extraWhere = slNameToInExpression(idField, allTerms);
struct lm *lm = lmInit(0);
struct hash *matchHash = getAllPossibleIds(conn, lm, idField, extraWhere);
trashDirFile(&tn, "hgtData", "identifiers", ".key");
f = mustOpen(tn.forCgi, "w");
for (term = allTerms; term != NULL; term = term->next)
{
struct slName *matchList = NULL, *match;
if (matchHash == NULL)
{
matchList = slNameNew(term->name);
}
else
{
/* Support multiple alias->id mappings: */
char upcased[1024];
safecpy(upcased, sizeof(upcased), term->name);
touppers(upcased);
struct hashEl *hel = hashLookup(matchHash, upcased);
if (hel != NULL)
{
matchList = slNameNew((char *)hel->val);
while ((hel = hashLookupNext(hel)) != NULL)
{
match = slNameNew((char *)hel->val);
slAddHead(&matchList, match);
}
}
}
if (matchList != NULL)
{
foundTerms++;
for (match = matchList; match != NULL; match = match->next)
{
mustWrite(f, match->name, strlen(match->name));
mustWrite(f, "\n", 1);
}
}
else
{
slAddHead(&missingTerms, slNameNew(term->name));
}
}
slReverse(&missingTerms);
carefulClose(&f);
cartSetString(cart, hgtaIdentifierDb, database);
cartSetString(cart, hgtaIdentifierTable, curTable);
cartSetString(cart, hgtaIdentifierFile, tn.forCgi);
if (saveIdText)
freez(&idTextForLf);
else
cartRemove(cart, hgtaPastedIdentifiers);
int missingCount = totalTerms - foundTerms;
if (missingCount > 0)
{
char *xrefTable, *aliasField;
getXrefInfo(conn, &xrefTable, NULL, &aliasField);
boolean xrefIsSame = xrefTable && sameString(curTable, xrefTable);
struct tempName tn;
trashDirFile(&tn, "hgt/missingIds", cartSessionId(cart), ".tmp");
FILE *f = mustOpen(tn.forCgi, "w");
int exampleCount = 0;
for (term = missingTerms; term != NULL; term = term->next)
{
if (exampleCount < 10)
{
++exampleCount;
dyStringPrintf(exampleMissingIds, "%s\n", term->name);
}
fprintf(f, "%s\n", term->name);
}
carefulClose(&f);
dyStringPrintf(exampleMissingIds, "\n<a href=%s>Complete list of missing identifiers<a>\n", tn.forHtml);
warn("Note: %d of the %d given identifiers have no match in "
"table %s, field %s%s%s%s%s. "
"Try the \"describe table schema\" button for more "
"information about the table and field.\n"
"%d %smissing identifier(s):\n"
"%s\n",
(totalTerms - foundTerms), totalTerms,
curTable, idField,
(xrefTable ? (xrefIsSame ? "" : " or in alias table ") : ""),
(xrefTable ? (xrefIsSame ? "" : xrefTable) : ""),
(xrefTable ? (xrefIsSame ? " or in field " : ", field ") : ""),
(xrefTable ? aliasField : ""),
exampleCount,
exampleCount < missingCount ? "example " : "",
exampleMissingIds->string
);
webNewSection("Table Browser");
}
lmCleanup(&lm);
hashFree(&matchHash);
}
else
{
cartRemove(cart, hgtaIdentifierFile);
}
mainPageAfterOpen(conn);
htmlClose();
}
static void readPartHeaderMB(struct mimeBuf *b, struct mimePart *p, char *altHeader)
/* Reads the header lines of the mimePart,
saves the header settings in a hash. */
{
struct dyString *fullLine = dyStringNew(0);
char *key=NULL, *val=NULL;
struct lineFile *lf = NULL;
char *line = NULL;
char *lineAhead = NULL;
int size = 0;
p->hdr = newHash(3);
//debug
//fprintf(stderr,"headers dumpMB: ");
//dumpMB(b); //debug
if (altHeader)
{
lf = lineFileOnString("MIME Header", TRUE, altHeader);
}
/* read ahead one line, skipping any leading blanks lines */
do
{
if (altHeader)
lineFileNext(lf, &lineAhead, &size);
else
lineAhead = getLineMB(b);
}
while (sameString(lineAhead,""));
do
{
/* accumulate a full header line - some emailers split into mpl lines */
dyStringClear(fullLine);
do
{
line = lineAhead;
if (altHeader)
lineFileNext(lf, &lineAhead, &size);
else
lineAhead = getLineMB(b);
dyStringAppend(fullLine,line);
if (!altHeader)
freez(&line);
} while (isspace(lineAhead[0]));
line = fullLine->string;
//fprintf(stderr,"found a line! [%s]\n",line); //debug
key = line;
val = strchr(line,':');
if (!val)
errAbort("readPartHeaderMB error - header-line colon not found, line=[%s]",line);
*val = 0;
val++;
key=trimSpaces(key);
// since the hash is case-sensitive, convert to lower case for ease of matching
tolowers(key);
val=trimSpaces(val);
hashAdd(p->hdr,key,cloneString(val));
//debug
//fprintf(stderr,"MIME header: key=[%s], val=[%s]\n",key,val);
//fflush(stderr);
} while (!sameString(lineAhead,""));
if (altHeader)
{
if (nlType == nlt_undet)
nlType = lf->nlType;
lineFileClose(&lf);
}
else
{
freez(&lineAhead);
}
dyStringFree(&fullLine);
}
static struct bed *parseRegionInput(char *inputString)
/* scan the user region definition, turn into a bed list */
{
int itemCount = 0;
struct bed *bedList = NULL;
struct bed *bedEl;
int wordCount;
char *words[5];
struct lineFile *lf;
lf = lineFileOnString("userData", TRUE, inputString);
while (0 != (wordCount = lineFileChopNext(lf, words, ArraySize(words))))
{
char *chromName = NULL;
int chromStart = 0;
int chromEnd = 0;
char *regionName = NULL;
/* might be something of the form: chrom:start-end optionalRegionName */
if (((1 == wordCount) || (2 == wordCount)) &&
hgParseChromRange(NULL, words[0], &chromName,
&chromStart, &chromEnd))
{
if (2 == wordCount)
regionName = cloneString(words[1]);
}
else if (!((3 == wordCount) || (4 == wordCount)))
{
int i;
struct dyString *errMessage = dyStringNew(0);
for (i = 0; i < wordCount; ++i)
dyStringPrintf(errMessage, "%s ", words[i]);
errAbort("line %d: '%s'<BR>\n"
"illegal bed size, expected 3 or 4 fields, found %d\n",
lf->lineIx, dyStringCannibalize(&errMessage), wordCount);
}
else
{
chromName = hgOfficialChromName(database, words[0]);
chromStart = sqlSigned(words[1]);
chromEnd = sqlSigned(words[2]);
if (wordCount > 3)
regionName = cloneString(words[3]);
}
++itemCount;
if (itemCount > 1000)
{
warn("limit 1000 region definitions reached at line %d<BR>\n",
lf->lineIx);
break;
}
AllocVar(bedEl);
bedEl->chrom = chromName;
if (NULL == bedEl->chrom)
errAbort("at line %d, chrom name '%s' %s %s not recognized in this assembly %d",
lf->lineIx, words[0], words[1], words[2], wordCount);
bedEl->chromStart = chromStart;
bedEl->chromEnd = chromEnd;
if (illegalCoordinate(bedEl->chrom, bedEl->chromStart, bedEl->chromEnd))
errAbort("illegal input at line %d: %s %d %d",
lf->lineIx, bedEl->chrom, bedEl->chromStart, bedEl->chromEnd);
if (wordCount > 3)
bedEl->name = regionName;
else
bedEl->name = NULL;
/* if we wanted to give artifical names to each item */
#ifdef NOT
{
char name[128];
safef(name, ArraySize(name), "item_%04d", itemCount);
bedEl->name = cloneString(name);
}
#endif
slAddHead(&bedList, bedEl);
}
lineFileClose(&lf);
// slSort(&bedList, bedCmp); /* this would do chrom,chromStart order */
slReverse(&bedList); /* with no sort, it is in order as user entered */
return (bedList);
}
