void doUploadIdentifiers(struct sqlConnection *conn)
/* Respond to upload identifiers button. */
{
struct hTableInfo *hti = maybeGetHti(database, curTable, conn);
char *idField = getIdField(database, curTrack, curTable, hti);
htmlOpen("Upload Identifiers for %s", curTableLabel());
if (idField == NULL)
errAbort("Sorry, I can't tell which field of table %s to treat as the "
"identifier field.", curTable);
hPrintf("<FORM ACTION=\"%s\" METHOD=POST ENCTYPE=\"multipart/form-data\">\n",
getScriptName());
cartSaveSession(cart);
hPrintf("Please enter the name of a file from your computer that contains a ");
hPrintf("space, tab, or ");
hPrintf("line separated list of the items you want to include.\n");
explainIdentifiers(database, conn, idField);
hPrintf("<BR>\n");
hPrintf("<INPUT TYPE=FILE NAME=\"%s\"> ", hgtaPastedIdentifiers);
hPrintf("<BR>\n");
cgiMakeButton(hgtaDoPastedIdentifiers, "submit");
hPrintf(" ");
cgiMakeButton(hgtaDoMainPage, "cancel");
hPrintf("</FORM>");
cgiDown(0.9);
htmlClose();
}
struct bed *dbGetFilteredBedsOnRegions(struct sqlConnection *conn,
char *db, char *dbVarName, char *table, char *tableVarName,
struct region *regionList, struct lm *lm,
int *retFieldCount)
/* Get list of beds from database, in all regions, that pass filtering. */
{
/* A joining query may be required if the filter incorporates linked tables. */
struct hTableInfo *hti = getHti(db, table, conn);
struct slName *fieldList = getBedFieldSlNameList(hti, db, table);
struct joinerDtf *dtfList = NULL;
struct joinerDtf *filterTables = NULL;
boolean doJoin = joinRequired(db, table,
fieldList, &dtfList, &filterTables);
struct region *region;
struct bed *bedList = NULL;
char *idField = getIdField(db, curTrack, table, hti);
struct hash *idHash = identifierHash(db, table);
if (! doJoin)
{
for (region = regionList; region != NULL; region = region->next)
{
char *identifierFilter = identifierWhereClause(idField, idHash);
char *filter = filterClause(dbVarName, tableVarName, region->chrom, identifierFilter);
struct bed *bedListRegion = getRegionAsMergedBed(dbVarName, tableVarName,
region, filter, idHash, lm, retFieldCount);
struct bed *bed, *nextBed;
for (bed = bedListRegion; bed != NULL; bed = nextBed)
{
nextBed = bed->next;
slAddHead(&bedList, bed);
}
freez(&filter);
}
slReverse(&bedList);
}
else
{
struct joiner *joiner = allJoiner;
struct joinedTables *joined = joinedTablesCreate(joiner,
db, table, dtfList, filterTables, 1000000, regionList);
int bedFieldCount = hTableInfoBedFieldCount(hti);
if (retFieldCount != NULL)
*retFieldCount = bedFieldCount;
bedList = joinedTablesToBed(joined, hti, bedFieldCount, lm);
joinedTablesFree(&joined);
}
joinerDtfFreeList(&dtfList);
joinerDtfFreeList(&filterTables);
hashFree(&idHash);
return bedList;
}
struct joinedTables *tjLoadFirst(struct region *regionList,
struct tableJoiner *tj, int totalFieldCount,
int totalKeyCount, int maxRowCount)
/* Load up first table in series of joined tables. This allocates
* field and key arrays big enough for all. */
{
struct joinedTables *joined = joinedTablesNew(totalFieldCount,
totalKeyCount, maxRowCount);
struct hash *idHash = NULL;
struct hTableInfo *hti = getHtiOnDb(tj->database, tj->table);
char *idField = getIdField(tj->database, curTrack, tj->table, hti);
if (idField != NULL)
idHash = identifierHash(tj->database, tj->table);
tjLoadSome(regionList, joined, 0, 0,
idField, idHash, NULL, NULL, tj,
isPositional(tj->database, tj->table), TRUE);
hashFree(&idHash);
return joined;
}
void doPasteIdentifiers(struct sqlConnection *conn)
/* Respond to paste identifiers button. */
{
struct sqlConnection *alternateConn = conn;
char *actualDb = database;
if (sameWord(curTable, WIKI_TRACK_TABLE))
{
alternateConn = wikiConnect();
actualDb = wikiDbName();
}
char *oldPasted = forCurTable() ?
cartUsualString(cart, hgtaPastedIdentifiers, "") : "";
struct hTableInfo *hti = maybeGetHti(actualDb, curTable, conn);
char *idField = getIdField(actualDb, curTrack, curTable, hti);
htmlOpen("Paste In Identifiers for %s", curTableLabel());
if (idField == NULL)
errAbort("Sorry, I can't tell which field of table %s to treat as the "
"identifier field.", curTable);
hPrintf("<FORM ACTION=\"%s\" METHOD=%s>\n", getScriptName(),
cartUsualString(cart, "formMethod", "POST"));
cartSaveSession(cart);
hPrintf("Please paste in the identifiers you want to include.\n");
if (sqlDatabaseExists("hgTemp"))
explainIdentifiers(actualDb, alternateConn, idField);
else
warn("No hgTemp database found for temporary tables.<br>Please src/product/README.mysql.setup for more information.");
hPrintf("<BR>\n");
cgiMakeTextArea(hgtaPastedIdentifiers, oldPasted, 10, 70);
hPrintf("<BR>\n");
cgiMakeButton(hgtaDoPastedIdentifiers, "submit");
hPrintf(" ");
cgiMakeButton(hgtaDoClearPasteIdentifierText, "clear");
hPrintf(" ");
cgiMakeButton(hgtaDoMainPage, "cancel");
hPrintf("</FORM>");
cgiDown(0.9);
htmlClose();
if (sameWord(curTable, WIKI_TRACK_TABLE))
wikiDisconnect(&alternateConn);
}
void RelationalCollection::setFilterFromRelationFilters() {
// FIXME escape the id field for sql ?
if (m_relation_filters.empty()) {
if (displayMode() == CollectionDisplay::Filtered)
setFilter("false");
else
setFilter("");
} else {
QString sql_filter_string = "";
QStringList sql_filters;
foreach(RelationalObjectRef filter, m_relation_filters) {
sql_filters << QString("%1 == %2").arg(getIdField(filter), QVariant(filter.id()).toString());
}
foreach(QSqlField filter, m_filters) {
sql_filters << QString("%1 == %2").arg(filter.name(), filter.value().toString());
}
QString getIdField(RelationalObject *object) {
return getIdField(object->fieldName());
}
void bamTabOut(char *db, char *table, struct sqlConnection *conn, char *fields, FILE *f)
/* Print out selected fields from BAM. If fields is NULL, then print out all fields. */
{
struct hTableInfo *hti = NULL;
hti = getHti(db, table, conn);
struct hash *idHash = NULL;
char *idField = getIdField(db, curTrack, table, hti);
int idFieldNum = 0;
/* if we know what field to use for the identifiers, get the hash of names */
if (idField != NULL)
idHash = identifierHash(db, table);
if (f == NULL)
f = stdout;
/* Convert comma separated list of fields to array. */
int fieldCount = chopByChar(fields, ',', NULL, 0);
char **fieldArray;
AllocArray(fieldArray, fieldCount);
chopByChar(fields, ',', fieldArray, fieldCount);
/* Get list of all fields in big bed and turn it into a hash of column indexes keyed by
* column name. */
struct hash *fieldHash = hashNew(0);
struct slName *bb, *bbList = bamGetFields();
int i;
for (bb = bbList, i=0; bb != NULL; bb = bb->next, ++i)
{
/* if we know the field for identifiers, save it away */
if ((idField != NULL) && sameString(idField, bb->name))
idFieldNum = i;
hashAddInt(fieldHash, bb->name, i);
}
/* Create an array of column indexes corresponding to the selected field list. */
int *columnArray;
AllocArray(columnArray, fieldCount);
for (i=0; i<fieldCount; ++i)
{
columnArray[i] = hashIntVal(fieldHash, fieldArray[i]);
}
/* Output row of labels */
fprintf(f, "#%s", fieldArray[0]);
for (i=1; i<fieldCount; ++i)
fprintf(f, "\t%s", fieldArray[i]);
fprintf(f, "\n");
struct asObject *as = bamAsObj();
struct asFilter *filter = NULL;
if (anyFilter())
{
filter = asFilterFromCart(cart, db, table, as);
if (filter)
{
fprintf(f, "# Filtering on %d columns\n", slCount(filter->columnList));
}
}
/* Loop through outputting each region */
struct region *region, *regionList = getRegions();
int maxOut = bigFileMaxOutput();
for (region = regionList; region != NULL && (maxOut > 0); region = region->next)
{
struct lm *lm = lmInit(0);
char *fileName = bamFileName(table, conn, region->chrom);
struct samAlignment *sam, *samList = bamFetchSamAlignment(fileName, region->chrom,
region->start, region->end, lm);
char *row[SAMALIGNMENT_NUM_COLS];
char numBuf[BAM_NUM_BUF_SIZE];
for (sam = samList; sam != NULL && (maxOut > 0); sam = sam->next)
{
samAlignmentToRow(sam, numBuf, row);
if (asFilterOnRow(filter, row))
{
/* if we're looking for identifiers, check if this matches */
if ((idHash != NULL)&&(hashLookup(idHash, row[idFieldNum]) == NULL))
continue;
int i;
fprintf(f, "%s", row[columnArray[0]]);
for (i=1; i<fieldCount; ++i)
fprintf(f, "\t%s", row[columnArray[i]]);
fprintf(f, "\n");
maxOut --;
}
}
freeMem(fileName);
lmCleanup(&lm);
}
if (maxOut == 0)
warn("Reached output limit of %d data values, please make region smaller,\n\tor set a higher output line limit with the filter settings.", bigFileMaxOutput());
/* Clean up and exit. */
hashFree(&fieldHash);
freeMem(fieldArray);
freeMem(columnArray);
}
void showMainControlTable(struct sqlConnection *conn)
/* Put up table with main controls for main page. */
{
struct grp *selGroup;
boolean isWig = FALSE, isPositional = FALSE, isMaf = FALSE, isBedGr = FALSE,
isChromGraphCt = FALSE, isPal = FALSE, isArray = FALSE, isBam = FALSE, isVcf = FALSE, isHalSnake = FALSE, isLongTabix = FALSE;
boolean gotClade = hGotClade();
struct hTableInfo *hti = NULL;
hPrintf("<TABLE BORDER=0>\n");
/* Print clade, genome and assembly line. */
{
if (gotClade)
{
hPrintf("<TR><TD><B>clade:</B>\n");
printCladeListHtml(hGenome(database), onChangeClade());
nbSpaces(3);
hPrintf("<B>genome:</B>\n");
printGenomeListForCladeHtml(database, onChangeOrg());
}
else
{
hPrintf("<TR><TD><B>genome:</B>\n");
printGenomeListHtml(database, onChangeOrg());
}
nbSpaces(3);
hPrintf("<B>assembly:</B>\n");
printAssemblyListHtml(database, onChangeDb());
hPrintf("</TD></TR>\n");
}
/* Print group and track line. */
{
hPrintf("<TR><TD>");
selGroup = showGroupField(hgtaGroup, onChangeGroupOrTrack(), conn, hAllowAllTables());
nbSpaces(3);
curTrack = showTrackField(selGroup, hgtaTrack, onChangeGroupOrTrack(), FALSE);
nbSpaces(3);
boolean hasCustomTracks = FALSE;
struct trackDb *t;
for (t = fullTrackList; t != NULL; t = t->next)
{
if (isCustomTrack(t->table))
{
hasCustomTracks = TRUE;
break;
}
}
hOnClickButton("document.customTrackForm.submit();return false;",
hasCustomTracks ? CT_MANAGE_BUTTON_LABEL : CT_ADD_BUTTON_LABEL);
hPrintf(" ");
if (hubConnectTableExists())
hOnClickButton("document.trackHubForm.submit();return false;", "track hubs");
hPrintf("</TD></TR>\n");
}
/* Print table line. */
{
hPrintf("<TR><TD>");
curTable = showTableField(curTrack, hgtaTable, TRUE);
if (isHubTrack(curTable) || (strchr(curTable, '.') == NULL)) /* In same database */
{
hti = getHti(database, curTable, conn);
isPositional = htiIsPositional(hti);
}
isLongTabix = isLongTabixTable( curTable);
isBam = isBamTable( curTable);
isVcf = isVcfTable(curTable, NULL);
isWig = isWiggle(database, curTable);
if (isBigWigTable(curTable))
{
isPositional = TRUE;
isWig = TRUE;
}
isHalSnake = isHalTable( curTable);
isMaf = isMafTable(database, curTrack, curTable);
isBedGr = isBedGraph(curTable);
isArray = isMicroarray(curTrack, curTable);
struct trackDb *tdb = findTdbForTable(database, curTrack, curTable, ctLookupName);
isPal = isPalCompatible(conn, tdb, curTable);
nbSpaces(1);
if (isCustomTrack(curTable))
{
isChromGraphCt = isChromGraph(tdb);
}
cgiMakeButton(hgtaDoSchema, "describe table schema");
hPrintf("</TD></TR>\n");
}
if (curTrack == NULL)
{
struct trackDb *tdb = hTrackDbForTrack(database, curTable);
struct trackDb *cTdb = hCompositeTrackDbForSubtrack(database, tdb);
if (cTdb)
curTrack = cTdb;
else
curTrack = tdb;
isMaf = isMafTable(database, curTrack, curTable);
}
/* Region line */
{
char *regionType = cartUsualString(cart, hgtaRegionType, hgtaRegionTypeGenome);
char *range = cartUsualString(cart, hgtaRange, "");
if (isPositional)
{
boolean doEncode = FALSE;
if (!trackHubDatabase(database))
doEncode = sqlTableExists(conn, "encodeRegions");
hPrintf("<TR><TD><B>region:</B>\n");
/* If regionType not allowed force it to "genome". */
if ((sameString(regionType, hgtaRegionTypeUserRegions) &&
userRegionsFileName() == NULL) ||
(sameString(regionType, hgtaRegionTypeEncode) && !doEncode))
regionType = hgtaRegionTypeGenome;
// Is "genome" is not allowed because of tdb 'tableBrowser noGenome'?
boolean disableGenome = ((curTrack && cartTrackDbIsNoGenome(database, curTrack->table)) ||
(curTable && cartTrackDbIsNoGenome(database, curTable)));
// If "genome" is selected but not allowed, force it to "range":
if (sameString(regionType, hgtaRegionTypeGenome) && disableGenome)
regionType = hgtaRegionTypeRange;
jsTrackingVar("regionType", regionType);
if (disableGenome)
{
makeRegionButtonExtraHtml(hgtaRegionTypeGenome, regionType, "DISABLED");
hPrintf(" <span"NO_GENOME_CLASS">genome (unavailable for selected track)</span>"
" ");
}
else
{
makeRegionButton(hgtaRegionTypeGenome, regionType);
hPrintf(" genome ");
}
if (doEncode)
{
makeRegionButton(hgtaRegionTypeEncode, regionType);
hPrintf(" ENCODE Pilot regions ");
}
makeRegionButton(hgtaRegionTypeRange, regionType);
hPrintf(" position ");
hPrintf("<INPUT TYPE=TEXT NAME=\"%s\" SIZE=26 VALUE=\"%s\" onFocus=\"%s\">\n",
hgtaRange, range, jsRadioUpdate(hgtaRegionType, "regionType", "range"));
cgiMakeButton(hgtaDoLookupPosition, "lookup");
hPrintf(" ");
if (userRegionsFileName() != NULL)
{
makeRegionButton(hgtaRegionTypeUserRegions, regionType);
hPrintf(" defined regions ");
cgiMakeButton(hgtaDoSetUserRegions, "change");
hPrintf(" ");
cgiMakeButton(hgtaDoClearUserRegions, "clear");
}
else
cgiMakeButton(hgtaDoSetUserRegions, "define regions");
hPrintf("</TD></TR>\n");
}
else
{
/* Need to put at least stubs of cgi variables in for JavaScript to work. */
jsTrackingVar("regionType", regionType);
cgiMakeHiddenVar(hgtaRange, range);
cgiMakeHiddenVar(hgtaRegionType, regionType);
}
/* Select identifiers line (if applicable). */
if (!isWig && getIdField(database, curTrack, curTable, hti) != NULL)
{
hPrintf("<TR><TD><B>identifiers (names/accessions):</B>\n");
cgiMakeButton(hgtaDoPasteIdentifiers, "paste list");
hPrintf(" ");
cgiMakeButton(hgtaDoUploadIdentifiers, "upload list");
if (identifierFileName() != NULL)
{
hPrintf(" ");
cgiMakeButton(hgtaDoClearIdentifiers, "clear list");
}
hPrintf("</TD></TR>\n");
}
}
/* microarray options */
/* button for option page here (median/log-ratio, etc) */
/* Filter line. */
{
hPrintf("<TR><TD><B>filter:</B>\n");
if (anyFilter())
{
cgiMakeButton(hgtaDoFilterPage, "edit");
hPrintf(" ");
cgiMakeButton(hgtaDoClearFilter, "clear");
if (isWig || isBedGr)
wigShowFilter(conn);
}
else
{
cgiMakeButton(hgtaDoFilterPage, "create");
}
hPrintf("</TD></TR>\n");
}
/* Composite track subtrack merge line. */
boolean canSubtrackMerge = (curTrack && tdbIsComposite(curTrack) && !isBam && !isVcf && !isLongTabix);
if (canSubtrackMerge)
{
hPrintf("<TR><TD><B>subtrack merge:</B>\n");
if (anySubtrackMerge(database, curTable))
{
cgiMakeButton(hgtaDoSubtrackMergePage, "edit");
hPrintf(" ");
cgiMakeButton(hgtaDoClearSubtrackMerge, "clear");
}
else
{
cgiMakeButton(hgtaDoSubtrackMergePage, "create");
}
hPrintf("</TD></TR>\n");
}
/* Intersection line. */
if (isPositional)
{
if (anyIntersection())
{
hPrintf("<TR><TD><B>intersection with %s:</B>\n",
cartString(cart, hgtaIntersectTable));
cgiMakeButton(hgtaDoIntersectPage, "edit");
hPrintf(" ");
cgiMakeButton(hgtaDoClearIntersect, "clear");
hPrintf("</TD></TR>\n");
}
else if (canIntersect(database, curTable))
{
hPrintf("<TR><TD><B>intersection:</B>\n");
cgiMakeButton(hgtaDoIntersectPage, "create");
hPrintf("</TD></TR>\n");
}
}
/* Correlation line. */
struct trackDb *tdb = findTdbForTable(database, curTrack, curTable, ctLookupName);
if (correlateTrackTableOK(tdb, curTable))
{
char *table2 = cartUsualString(cart, hgtaCorrelateTable, "none");
hPrintf("<TR><TD><B>correlation:</B>\n");
if (differentWord(table2, "none") && strlen(table2) && ! isNoGenomeDisabled(database, table2))
{
struct grp *groupList = fullGroupList;
struct grp *selGroup = findSelectedGroup(groupList, hgtaCorrelateGroup);
struct trackDb *tdb2 = findSelectedTrack(fullTrackList, selGroup,hgtaCorrelateTrack);
if (tdbIsComposite(tdb2))
{
struct slRef *tdbRefList = trackDbListGetRefsToDescendantLeaves(tdb2->subtracks);
struct slRef *tdbRef;
for (tdbRef = tdbRefList; tdbRef != NULL; tdbRef = tdbRef->next)
{
struct trackDb *subTdb = tdbRef->val;
if (sameString(table2, subTdb->table))
{
tdb2 = subTdb;
break;
}
}
slFreeList(&tdbRefList);
}
cgiMakeButton(hgtaDoCorrelatePage, "calculate");
cgiMakeButton(hgtaDoClearCorrelate, "clear");
if (tdb2 && tdb2->shortLabel)
hPrintf(" (with: %s)", tdb2->shortLabel);
#ifdef NOT_YET
/* debugging dbg vvvvv */
if (curTrack && curTrack->type) /* dbg */
{
hPrintf("<BR> (debug: '%s', '%s(%s)')",
curTrack->type, tdb2->type, table2);
}
/* debugging debug ^^^^^ */
#endif
}
else
cgiMakeButton(hgtaDoCorrelatePage, "create");
hPrintf("</TD></TR>\n");
}
/* Print output type line. */
showOutputTypeRow(isWig, isBedGr, isPositional, isMaf, isChromGraphCt, isPal, isArray, isHalSnake);
/* Print output destination line. */
{
char *compressType =
cartUsualString(cart, hgtaCompressType, textOutCompressNone);
char *fileName = cartUsualString(cart, hgtaOutFileName, "");
hPrintf("<TR><TD>\n");
hPrintf("<B>output file:</B> ");
cgiMakeTextVar(hgtaOutFileName, fileName, 29);
hPrintf(" (leave blank to keep output in browser)</TD></TR>\n");
hPrintf("<TR><TD>\n");
hPrintf("<B>file type returned: </B>");
cgiMakeRadioButton(hgtaCompressType, textOutCompressNone,
sameWord(textOutCompressNone, compressType));
hPrintf(" plain text  ");
cgiMakeRadioButton(hgtaCompressType, textOutCompressGzip,
sameWord(textOutCompressGzip, compressType));
hPrintf(" gzip compressed");
hPrintf("</TD></TR>\n");
}
hPrintf("</TABLE>\n");
/* Submit buttons. */
{
hPrintf("<BR>\n");
if (isWig || isBam || isVcf || isLongTabix)
{
char *name;
extern char *maxOutMenu[];
char *maxOutput = maxOutMenu[0];
if (isCustomTrack(curTable))
name=filterFieldVarName("ct", curTable, "_", filterMaxOutputVar);
else
name=filterFieldVarName(database,curTable, "_",filterMaxOutputVar);
maxOutput = cartUsualString(cart, name, maxOutMenu[0]);
if (isWig)
hPrintf(
"<I>Note: to return more than %s lines, change the filter setting"
" (above). The entire data set may be available for download as"
" a very large file that contains the original data values (not"
" compressed into the wiggle format) -- see the Downloads page."
"</I><BR>", maxOutput);
else if (isBam || isVcf || isLongTabix)
hPrintf(
"<I>Note: to return more than %s lines, change the filter setting"
" (above). Please consider downloading the entire data from our Download pages."
"</I><BR>", maxOutput);
}
else if (anySubtrackMerge(database, curTable) || anyIntersection())
{
hPrintf("<I>Note: The all fields and selected fields output formats "
"are not available when a%s has been specified.</I><BR>",
canSubtrackMerge ? " subtrack merge or intersection" : "n intersection");
}
cgiMakeButton(hgtaDoTopSubmit, "get output");
hPrintf(" ");
if (isPositional || isWig)
{
cgiMakeButton(hgtaDoSummaryStats, "summary/statistics");
hPrintf(" ");
}
#ifdef SOMETIMES
hPrintf(" ");
cgiMakeButton(hgtaDoTest, "test");
#endif /* SOMETIMES */
}
hPrintf("<P>"
"To reset <B>all</B> user cart settings (including custom tracks), \n"
"<A HREF=\"/cgi-bin/cartReset?destination=%s\">click here</A>.\n",
getScriptName());
}
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();
}
void vcfTabOut(char *db, char *table, struct sqlConnection *conn, char *fields, FILE *f,
boolean isTabix)
/* Print out selected fields from VCF. If fields is NULL, then print out all fields. */
{
struct hTableInfo *hti = NULL;
hti = getHti(db, table, conn);
struct hash *idHash = NULL;
char *idField = getIdField(db, curTrack, table, hti);
int idFieldNum = 0;
/* if we know what field to use for the identifiers, get the hash of names */
if (idField != NULL)
idHash = identifierHash(db, table);
if (f == NULL)
f = stdout;
/* Convert comma separated list of fields to array. */
int fieldCount = chopByChar(fields, ',', NULL, 0);
char **fieldArray;
AllocArray(fieldArray, fieldCount);
chopByChar(fields, ',', fieldArray, fieldCount);
/* Get list of all fields in big bed and turn it into a hash of column indexes keyed by
* column name. */
struct hash *fieldHash = hashNew(0);
struct slName *bb, *bbList = vcfGetFields();
int i;
for (bb = bbList, i=0; bb != NULL; bb = bb->next, ++i)
{
/* if we know the field for identifiers, save it away */
if ((idField != NULL) && sameString(idField, bb->name))
idFieldNum = i;
hashAddInt(fieldHash, bb->name, i);
}
/* Create an array of column indexes corresponding to the selected field list. */
int *columnArray;
AllocArray(columnArray, fieldCount);
for (i=0; i<fieldCount; ++i)
{
columnArray[i] = hashIntVal(fieldHash, fieldArray[i]);
}
// If we are outputting a subset of fields, invalidate the VCF header.
boolean allFields = (fieldCount == VCFDATALINE_NUM_COLS);
if (!allFields)
fprintf(f, "# Only selected columns are included below; output is not valid VCF.\n");
struct asObject *as = vcfAsObj();
struct asFilter *filter = NULL;
if (anyFilter())
filter = asFilterFromCart(cart, db, table, as);
/* Loop through outputting each region */
struct region *region, *regionList = getRegions();
int maxOut = bigFileMaxOutput();
struct trackDb *tdb = hashFindVal(fullTableToTdbHash, table);
// Include the header, absolutely necessary for VCF parsing.
boolean printedHeader = FALSE;
// Temporary storage for row-ification:
struct dyString *dyAlt = newDyString(1024);
struct dyString *dyFilter = newDyString(1024);
struct dyString *dyInfo = newDyString(1024);
struct dyString *dyGt = newDyString(1024);
struct vcfRecord *rec;
for (region = regionList; region != NULL && (maxOut > 0); region = region->next)
{
char *fileName = vcfFileName(tdb, conn, table, region->chrom);
struct vcfFile *vcff;
if (isTabix)
vcff = vcfTabixFileMayOpen(fileName, region->chrom, region->start, region->end,
100, maxOut);
else
vcff = vcfFileMayOpen(fileName, region->chrom, region->start, region->end,
100, maxOut, TRUE);
if (vcff == NULL)
noWarnAbort();
// If we are outputting all fields, but this VCF has no genotype info, omit the
// genotype columns from output:
if (allFields && vcff->genotypeCount == 0)
fieldCount = VCFDATALINE_NUM_COLS - 2;
if (!printedHeader)
{
fprintf(f, "%s", vcff->headerString);
if (filter)
fprintf(f, "# Filtering on %d columns\n", slCount(filter->columnList));
if (!allFields)
{
fprintf(f, "#%s", fieldArray[0]);
for (i=1; i<fieldCount; ++i)
fprintf(f, "\t%s", fieldArray[i]);
fprintf(f, "\n");
}
printedHeader = TRUE;
}
char *row[VCFDATALINE_NUM_COLS];
char numBuf[VCF_NUM_BUF_SIZE];
for (rec = vcff->records; rec != NULL && (maxOut > 0); rec = rec->next)
{
vcfRecordToRow(rec, region->chrom, numBuf, dyAlt, dyFilter, dyInfo, dyGt, row);
if (asFilterOnRow(filter, row))
{
/* if we're looking for identifiers, check if this matches */
if ((idHash != NULL) && (hashLookup(idHash, row[idFieldNum]) == NULL))
continue;
// All fields output: after asFilter'ing, preserve original VCF chrom
if (allFields && !sameString(rec->chrom, region->chrom))
row[0] = rec->chrom;
int i;
fprintf(f, "%s", row[columnArray[0]]);
for (i=1; i<fieldCount; ++i)
{
fprintf(f, "\t%s", row[columnArray[i]]);
}
fprintf(f, "\n");
maxOut --;
}
}
vcfFileFree(&vcff);
freeMem(fileName);
}
if (maxOut == 0)
warn("Reached output limit of %d data values, please make region smaller,\n\tor set a higher output line limit with the filter settings.", bigFileMaxOutput());
/* Clean up and exit. */
dyStringFree(&dyAlt); dyStringFree(&dyFilter); dyStringFree(&dyInfo); dyStringFree(&dyGt);
hashFree(&fieldHash);
freeMem(fieldArray);
freeMem(columnArray);
}
