f72d8b57a77755e64ef594ae2d4395a12c19712d hiram Tue Mar 19 21:46:37 2019 -0700 now showing a sample getData URL for supported tracks refs #18869 diff --git src/hg/hubApi/hubApi.c src/hg/hubApi/hubApi.c index f9a2b0c..a58944f 100644 --- src/hg/hubApi/hubApi.c +++ src/hg/hubApi/hubApi.c @@ -20,33 +20,87 @@ static struct cart *cart; /* CGI and other variables */ static struct hash *oldVars = NULL; static struct hash *trackCounter = NULL; static long totalTracks = 0; static boolean measureTiming = FALSE; /* set by CGI parameters */ static boolean allTrackSettings = FALSE; /* checkbox setting */ static char **shortLabels = NULL; /* public hub short labels in array */ // struct hubPublic *publicHubList = NULL; static int publicHubCount = 0; static char *defaultHub = "Plants"; static char *defaultDb = "ce11"; static long enteredMainTime = 0; /* will become = clock1000() on entry */ /* to allow calculation of when to bail out, taking too long */ static long timeOutSeconds = 100; static boolean timedOut = FALSE; +static char *urlPrefix = ""; /* initalized to support self references */ +static struct slName *supportedTypes = NULL; + /* will be initialized to a known supported set */ + +static void initSupportedTypes() +/* initalize the list of supported track types, + */ +{ +struct slName *el = newSlName("bed"); +slAddHead(&supportedTypes, el); +el = newSlName("wig"); +slAddHead(&supportedTypes, el); +el = newSlName("broadPeak"); +slAddHead(&supportedTypes, el); +el = newSlName("narrowPeak"); +slAddHead(&supportedTypes, el); +el = newSlName("bigBed"); +slAddHead(&supportedTypes, el); +el = newSlName("bigWig"); +slAddHead(&supportedTypes, el); +el = newSlName("bigNarrowPeak"); +slAddHead(&supportedTypes, el); +el = newSlName("bigGenePred"); +slAddHead(&supportedTypes, el); +el = newSlName("bigPsl"); +slAddHead(&supportedTypes, el); +el = newSlName("bigBarChart"); +slAddHead(&supportedTypes, el); +el = newSlName("bigInteract"); +slAddHead(&supportedTypes, el); +el = newSlName("bigMaf"); +slAddHead(&supportedTypes, el); +el = newSlName("bigChain"); +slAddHead(&supportedTypes, el); +slNameSort(&supportedTypes); +} + +static boolean isSupportedType(char *type) +/* is given type in the supportedTypes list ? */ +{ +boolean ret = FALSE; +if (startsWith("wigMaf", type)) + return ret; +struct slName *el; +for (el = supportedTypes; el; el = el->next) + { + if (startsWith(el->name, type)) + { + ret = TRUE; + break; + } + } +return ret; +} static int publicHubCmpCase(const void *va, const void *vb) -/* Compare two slNames, ignore case. */ +/* Compare two shortLabels, ignore case. */ { const struct hubPublic *a = *((struct hubPublic **)va); const struct hubPublic *b = *((struct hubPublic **)vb); return strcasecmp(a->shortLabel, b->shortLabel); } static void publicHubSortCase(struct hubPublic **pList) /* Sort slName list, ignore case. */ { slSort(pList, publicHubCmpCase); } static struct hubPublic *hubPublicLoad(char **row) /* Load a hubPublic from row fetched with select * from hubPublic * from database. Dispose of this with hubPublicFree(). */ @@ -152,46 +206,72 @@ hashIncInt(countTracks, "superTrack child"); hashIncInt(countTracks, stripType); hashIncInt(countTracks, "track count"); } else if (isEmpty(tdb->type)) hashIncInt(countTracks, "no type specified"); else { hashIncInt(countTracks, stripType); hashIncInt(countTracks, "track count"); } freeMem(stripType); // showCounts(countTracks); } -static void showSubTracks(struct trackDb *tdb, struct hash *countTracks) +static void sampleUrl(char *db, struct trackDb *tdb, char *chrom, long long chromSize) +/* print out a sample getData URL */ +{ +long long start = chromSize / 4; +long long end = start + 10000; +if (end > chromSize) + end = chromSize; + +if (db) + { + if (tdb->parent) + hPrintf("<li>%s : %s subtrack of parent: %s <a href='%s/getData/track?db=%s&chrom=%s&track=%s&start=%lld&end=%lld' target=_blank>(sample getData)</a></li>\n", tdb->track, tdb->type, tdb->parent->track, urlPrefix, db, chrom, tdb->track, start, end ); + else + hPrintf("<li>%s : %s <a href='%s/getData/track?db=%s&chrom=%s&track=%s&start=%lld&end=%lld' target=_blank>(sample getData)</a></li>\n", tdb->track, tdb->type, urlPrefix, db, chrom, tdb->track, start, end ); + } +} + +static void showSubTracks(char *db, struct trackDb *tdb, struct hash *countTracks, + char *chromName, long long chromSize) /* tdb has subtracks, show only subTracks, no details */ { hPrintf(" <li><ul>\n"); if (tdb->subtracks) { struct trackDb *tdbEl = NULL; for (tdbEl = tdb->subtracks; tdbEl; tdbEl = tdbEl->next) { if (tdbIsCompositeView(tdbEl)) hPrintf("<li>%s : %s : composite view of parent: %s</li>\n", tdbEl->track, tdbEl->type, tdbEl->parent->track); else + { + if (isSupportedType(tdbEl->type)) + { + sampleUrl(db, tdbEl, chromName, chromSize); +// hPrintf("<li>%s : %s : subtrack of parent: %s (with sample)</li>\n", tdbEl->track, tdbEl->type, tdbEl->parent->track, getData); + } + else hPrintf("<li>%s : %s : subtrack of parent: %s</li>\n", tdbEl->track, tdbEl->type, tdbEl->parent->track); + } hashCountTrack(tdbEl, countTracks); if (tdbEl->subtracks) - showSubTracks(tdbEl, countTracks); + showSubTracks(db, tdbEl, countTracks, chromName, chromSize); } } hPrintf(" </ul></li>\n"); } static void trackSettings(struct trackDb *tdb, struct hash *countTracks) /* process the settingsHash for a trackDb, recursive when subtracks */ { hPrintf(" <li><ul>\n"); // if (tdb->children) haven't yet seen a track with children ? // hPrintf(" <li>%s: has children</li>\n", tdb->track); // else // hPrintf(" <li>%s: NO children</li>\n", tdb->track); struct hashEl *hel; struct hashCookie hc = hashFirst(tdb->settingsHash); @@ -302,32 +382,32 @@ retVal = 1; } } errCatchEnd(errCatch); if (errCatch->gotError) { retVal = 1; dyStringPrintf(errors, "%s", errCatch->message->string); } errCatchFree(&errCatch); return retVal; } /* static int bbiBriefMeasure() */ -static void countOneTdb(struct trackDb *tdb, char *bigDataIndex, - struct hash *countTracks) +static void countOneTdb(char *db, struct trackDb *tdb, char *bigDataIndex, + struct hash *countTracks, char *chromName, long long chromSize) { char *bigDataUrl = trackDbSetting(tdb, "bigDataUrl"); // char *compositeTrack = trackDbSetting(tdb, "compositeTrack"); boolean compositeContainer = tdbIsComposite(tdb); boolean compositeView = tdbIsCompositeView(tdb); // char *superTrack = trackDbSetting(tdb, "superTrack"); boolean superChild = tdbIsSuperTrackChild(tdb); boolean depthSearch = cartUsualBoolean(cart, "depthSearch", FALSE); hashCountTrack(tdb, countTracks); if (depthSearch && bigDataUrl) { long chromCount = 0; long itemCount = 0; struct dyString *errors = newDyString(1024); @@ -345,80 +425,89 @@ else hPrintf(" <li>%s : %s : %ld chroms : %ld count</li>\n", tdb->track, tdb->type, chromCount, itemCount); } } else { if (compositeContainer) hPrintf(" <li>%s : %s : composite track container has %d subtracks</li>\n", tdb->track, tdb->type, slCount(tdb->subtracks)); else if (compositeView) hPrintf(" <li>%s : %s : composite view of parent: %s</li>\n", tdb->track, tdb->type, tdb->parent->track); else if (superChild) hPrintf(" <li>%s : %s : superTrack child of parent: %s</li>\n", tdb->track, tdb->type, tdb->parent->track); else if (! depthSearch && bigDataUrl) hPrintf(" <li>%s : %s : %s</li>\n", tdb->track, tdb->type, bigDataUrl); else + { + if (isSupportedType(tdb->type)) + sampleUrl(db, tdb, chromName, chromSize); + else hPrintf(" <li>%s : %s</li>\n", tdb->track, tdb->type); } + } if (allTrackSettings) { hPrintf(" <li><ul>\n"); trackSettings(tdb, countTracks); /* show all settings */ hPrintf(" </ul></li>\n"); } else if (tdb->subtracks) { - showSubTracks(tdb, countTracks); + showSubTracks(db, tdb, countTracks, chromName, chromSize); } return; -} /* static void countOneTdb(struct trackDb *tdb, - * char *bigDataIndex, struct hash *countTracks) +} /* static void countOneTdb(char *db, struct trackDb *tdb, + * char *bigDataIndex, struct hash *countTracks, + * char *chromName, long long chromSize) */ static void hubTrackList(struct trackDb *topTrackDb, struct trackHubGenome *genome) /* process the track list in a hub to show all tracks */ { hPrintf(" <li><ul>\n"); if (topTrackDb) { struct hash *countTracks = hashNew(0); struct trackDb *tdb = NULL; for ( tdb = topTrackDb; tdb; tdb = tdb->next ) { char *bigDataIndex = NULL; char *relIdxUrl = trackDbSetting(topTrackDb, "bigDataIndex"); if (relIdxUrl != NULL) bigDataIndex = trackHubRelativeUrl(genome->trackDbFile, relIdxUrl); - countOneTdb(tdb, bigDataIndex, countTracks); + countOneTdb(NULL, tdb, bigDataIndex, countTracks, NULL, 0); if (timeOutReached()) break; } /* for ( tdb = topTrackDb; tdb; tdb = tdb->next ) */ hPrintf(" <li>%d different track types</li>\n",countTracks->elCount - 1); /* add this single genome count to the overall multi-genome counts */ if (countTracks->elCount) { hPrintf(" <li><ol>\n"); struct hashEl *hel, *helList = hashElListHash(countTracks); slSort(&helList, hashElCmpIntValDesc); for (hel = helList; hel; hel = hel->next) { if (sameOk("track count", hel->name)) continue; int prevCount = ptToInt(hashFindVal(trackCounter, hel->name)); if (differentStringNullOk("track count", hel->name)) totalTracks += ptToInt(hel->val); hashReplace(trackCounter, hel->name, intToPt(prevCount + ptToInt(hel->val))); + if (isSupportedType(hel->name)) + hPrintf(" <li>%d - %s - supported</li>\n", ptToInt(hel->val), hel->name); + else hPrintf(" <li>%d - %s</li>\n", ptToInt(hel->val), hel->name); } hPrintf(" </ol></li>\n"); } } else hPrintf(" <li>no trackTopDb</li>\n"); hPrintf(" </ul><li>\n"); } /* static struct trackDb *hubTrackList() */ static struct trackDb *assemblySettings(struct trackHubGenome *genome) /* display all the assembly 'settingsHash' */ { struct trackDb *tdb = obtainTdb(genome, NULL); @@ -565,85 +654,84 @@ * parse that and return a function pointer and the parsed words * Returns NULL if not recognized */ { char *tmp = cloneString(pathInfo); /* skip the first leading slash to simplify chopByChar parsing */ tmp += 1; int wordCount = chopByChar(tmp, '/', words, MAX_PATH_INFO); if (wordCount < 1) return NULL; struct hashEl *hel = hashLookup(apiFunctionHash, words[0]); return hel; } -#ifdef NOT -static void apiFunctionSwitch(struct hashEl *hel, char *words[MAX_PATH_INFO]) -/* given a pathInfo string: /command/subCommand/etc... - * parse that and decide on which function to acll +static long long largestChrom(char *db, char **nameReturn) +/* return the length and get the chrom name for the largest chrom + * from chromInfo table. For use is sample getData URLs */ { -hPrintDisable(); /* turn off all normal HTML output, doing JSON output */ - -/* the leading slash has been removed from the pathInfo, therefore, the - * chop will have the first word in words[0] - */ -#ifdef NOT -char *words[MAX_PATH_INFO];/*expect no more than MAX_PATH_INFO number of words*/ -int wordCount = chopByChar(pathInfo, '/', words, ArraySize(words)); -if (wordCount < 2) - apiErrAbort("unknown endpoint command: '/%s'", pathInfo); - -struct hashEl *hel = hashLookup(apiFunctionHash, words[0]); -if (hel == NULL) - apiErrAbort("no such command: '%s' for endpoint '/%s'", words[0], pathInfo); -#endif -void (*apiFunction)(char **) = hel->val; -// void (*apiFunction)(char **) = hashMustFindVal(apiFunctionHash, words[0]); - -(*apiFunction)(words); - -} /* static void apiFunctionSwitch(char *pathInfo) */ -#endif +char query[1024]; +struct sqlConnection *conn = hAllocConn(db); +// Build a query to select the hubUrl for the given shortLabel +sqlSafef(query, sizeof(query), "select chrom,size from chromInfo order by size desc limit 1"); +struct sqlResult *sr = sqlGetResult(conn, query); +char **row = sqlNextRow(sr); +long long length = 0; +if (row) + { + *nameReturn = cloneString(row[0]); + length = sqlLongLong(row[1]); + } +sqlFreeResult(&sr); +hFreeConn(&conn); +return length; +} static void tracksForUcscDb(char *db) /* scan the specified database for all tracks */ { struct hash *countTracks = hashNew(0); -hPrintf("<p>Tracks in UCSC genome: '%s'<br>\n", db); +char *chromName = NULL; +long long chromSize = largestChrom(db, &chromName); +hPrintf("<p>Tracks in UCSC genome: '%s', longest chrom: %s:%lld<br>\n", db, chromName, chromSize); struct trackDb *tdbList = obtainTdb(NULL, db); struct trackDb *tdb; hPrintf("<ul>\n"); +hPrintf("<li>%s:%lld</li>\n", chromName, chromSize); for (tdb = tdbList; tdb != NULL; tdb = tdb->next ) { - countOneTdb(tdb, NULL, countTracks); + countOneTdb(db, tdb, NULL, countTracks, chromName, chromSize); if (timeOutReached()) break; } int trackCount = ptToInt(hashFindVal(countTracks, "track count")); /* elCount - 1 since the 'track count' element isn't a track */ hPrintf(" <li>%d total tracks counted, %d different track types</li>\n", trackCount, countTracks->elCount - 1); if (countTracks->elCount) { hPrintf(" <ol>\n"); struct hashEl *hel, *helList = hashElListHash(countTracks); slSort(&helList, hashElCmpIntValDesc); for (hel = helList; hel; hel = hel->next) { if (sameOk("track count", hel->name)) continue; + if (isSupportedType(hel->name)) + hPrintf(" <li>%d - %s - supported</li>\n", ptToInt(hel->val), hel->name); + else hPrintf(" <li>%d - %s</li>\n", ptToInt(hel->val), hel->name); } hPrintf(" </ol>\n"); } hPrintf("</ul>\n"); hPrintf("</p>\n"); } // static void tracksForUcscDb(char * db) static void introductionText() /* output explanation text */ { char *scriptUri = getenv("SCRIPT_URI"); hPrintf("<h1>JSON data API interface to U.C. Santa Cruz genome browser data</h1>\n"); hPrintf("<h2>Data access URL: <em>%s</em></h2>\n", scriptUri); @@ -662,44 +750,56 @@ hPrintf("<li>/getData/sequence - return sequence from specified hub or database genome</li>\n"); hPrintf("<li>/getData/track - return data from specified track in hub or database genome</li>\n"); hPrintf("</ol></li>\n"); hPrintf("</ul>\n"); hPrintf("<h3>Parameters to endpoint functions:</h3>\n"); hPrintf("<ul>\n"); hPrintf("<li>hubUrl=<url> - specify track hub or assembly hub URL</li>\n"); hPrintf("<li>genome=<name> - specify genome assemby in track or assembly hub</li>\n"); hPrintf("<li>db=<ucscDb> - specify database (aka genome assembly) in UCSC genome browser</li>\n"); hPrintf("<li>track=<trackName> - specify data track in hub or UCSC database genome assembly</li>\n"); hPrintf("<li>chrom=<chrN> - specify chromosome name for sequence or track data</li>\n"); hPrintf("<li>start=<123> - specify start coordinate (0 relative) for data from track or sequence retrieval</li>\n"); hPrintf("<li>end=<456> - specify end coordinate (1 relative) for data from track or sequence retrieval</li>\n"); hPrintf("<li>(see also: <a href='http://genome.ucsc.edu/blog/the-ucsc-genome-browser-coordinate-counting-systems/' target=_blank>UCSC browser coordinate counting systems)</a></li>"); hPrintf("</ul>\n"); +hPrintf("<h3>Supported track types, at this time, for getData functions:</h3>\n"); +hPrintf("<ul>\n"); +static struct slName *el; +for (el = supportedTypes; el; el = el->next) + { + hPrintf("<li>%s</li>\n", el->name); + } +hPrintf("</ul>\n"); } -static void showExamples(char *url, struct trackHubGenome *hubGenome, char *ucscDb) +static void initUrlPrefix() +/* set up urlPrefix for self referenes */ { -char *urlPrefix = ""; char *httpHost = getenv("HTTP_HOST"); -if (! startsWith("hgwdev-api", httpHost)) { +if (! startsWith("hgwdev-api", httpHost)) + { if (startsWith("hgwdev",httpHost) || startsWith("genome-test", httpHost)) { urlPrefix = "../cgi-bin/hubApi"; } } +} +static void showExamples(char *url, struct trackHubGenome *hubGenome, char *ucscDb) +{ hPrintf("<h2>Example URLs to return json data structures:</h2>\n"); hPrintf("<h3>listing functions</h3>\n"); hPrintf("<ol>\n"); hPrintf("<li><a href='%s/list/publicHubs' target=_blank>list public hubs</a> <em>%s/list/publicHubs</em></li>\n", urlPrefix, urlPrefix); hPrintf("<li><a href='%s/list/ucscGenomes' target=_blank>list database genomes</a> <em>%s/list/ucscGenomes</em></li>\n", urlPrefix, urlPrefix); hPrintf("<li><a href='%s/list/hubGenomes?hubUrl=%s' target=_blank>list genomes from specified hub</a> <em>%s/list/hubGenomes?hubUrl=%s</em></li>\n", urlPrefix, url, urlPrefix, url); hPrintf("<li><a href='%s/list/tracks?hubUrl=%s&genome=%s' target=_blank>list tracks from specified hub and genome</a> <em>%s/list/tracks?hubUrl=%s&genome=%s</em></li>\n", urlPrefix, url, hubGenome->name, urlPrefix, url, hubGenome->name); hPrintf("<li><a href='%s/list/tracks?db=%s' target=_blank>list tracks from specified UCSC database</a> <em>%s/list/tracks?db=%s</em></li>\n", urlPrefix, ucscDb, urlPrefix, ucscDb); hPrintf("<li><a href='%s/list/chromosomes?db=%s' target=_blank>list chromosomes from specified UCSC database</a> <em>%s/list/chromosomes?db=%s</em></li>\n", urlPrefix, ucscDb, urlPrefix, ucscDb); hPrintf("<li><a href='%s/list/chromosomes?db=%s&track=gap' target=_blank>list chromosomes from specified track from UCSC databaset</a> <em>%s/list/chromosomes?db=%s&track=gap</em></li>\n", urlPrefix, ucscDb, urlPrefix, ucscDb); hPrintf("</ol>\n"); hPrintf("<h3>getData functions</h3>\n"); hPrintf("<ol>\n"); @@ -736,30 +836,32 @@ } static void doMiddle(struct cart *theCart) /* Set up globals and make web page */ { cart = theCart; measureTiming = hPrintStatus() && isNotEmpty(cartOptionalString(cart, "measureTiming")); measureTiming = TRUE; char *database = NULL; char *genome = NULL; cgiVarSet("ignoreCookie", "1"); getDbAndGenome(cart, &database, &genome, oldVars); initGenbankTableNames(database); +initSupportedTypes(); +initUrlPrefix(); char *docRoot = cfgOptionDefault("browser.documentRoot", DOCUMENT_ROOT); int timeout = cartUsualInt(cart, "udcTimeout", 300); if (udcCacheTimeout() < timeout) udcSetCacheTimeout(timeout); knetUdcInstall(); char *pathInfo = getenv("PATH_INFO"); /* nothing on incoming path, then display the WEB page instead */ if (sameOk("/",pathInfo)) pathInfo = NULL; boolean commandError = FALSE; /*expect no more than MAX_PATH_INFO number of words*/