44ccfacbe3a3d4b300f80d48651c77837a4b571e
galt
  Tue Apr 26 11:12:02 2022 -0700
SQL INJECTION Prevention Version 2 - this improves our methods by making subclauses of SQL that get passed around be both easy and correct to use. The way that was achieved was by getting rid of the obscure and not well used functions sqlSafefFrag and sqlDyStringPrintfFrag and replacing them with the plain versions of those functions, since these are not needed anymore. The new version checks for NOSQLINJ in unquoted %-s which is used to include SQL clauses, and will give an error the NOSQLINJ clause is not present, and this will automatically require the correct behavior by developers. sqlDyStringPrint is a very useful function, however because it was not enforced, users could use various other dyString functions and they operated without any awareness or checking for SQL correct use. Now those dyString functions are prohibited and it will produce an error if you try to use a dyString function on a SQL string, which is simply detected by the presence of the NOSQLINJ prefix.

diff --git src/hg/protein/kgGetCds/kgGetCds.c src/hg/protein/kgGetCds/kgGetCds.c
index e8dde73..dcebc83 100644
--- src/hg/protein/kgGetCds/kgGetCds.c
+++ src/hg/protein/kgGetCds/kgGetCds.c
@@ -1,191 +1,191 @@
 /* kgGetCds - create a gene candidate table with CDS info */
 
 /* Copyright (C) 2013 The Regents of the University of California 
  * See kent/LICENSE or http://genome.ucsc.edu/license/ for licensing information. */
 
 #include "common.h"
 #include "genePred.h"
 #include "genePredReader.h"
 #include "jksql.h"
 #include "bits.h"
 #include "hdb.h"
 
 
 char cdsBloc[2000][30];
 void usage(char *msg)
 /* Explain usage and exit. */
 {
 errAbort(
     "kgGetCds - create a gene candidate table with CDS info\n"
     "\nUsage:\n"
     "   kgGetCds tempKgDb geneTable outfile\n"
     "\n"
     "   tempKgDb: 	temp KG build DB\n"
     "   geneTable:	KG gene candidate table\n"
     "   outfile:	output file\n"
     "\n"
     "Example:\n"
     "  kgGetCds kgHg17FTemp kgCandidateX jY.tmp\n"
     );
 }
 
 void processAlign(char *kgTempDb, char *spDb, char *alignID, int cdsCnt, FILE *outf)
 {
 struct sqlConnection *conn2, *conn3, *conn4;
 char query2[256], query3[256];
 struct sqlResult *sr2, *sr3;
 char **row2, **row3;
 char *score;
 char *chrom;
 char *protAcc;
 char *mrnaID;
 char *ranking;
 int  protDbId;
 char condStr[255];
 int  i;
 char *chp;
 char *isCurated;
 
 conn2= hAllocConn(kgTempDb);
 conn3= hAllocConn(kgTempDb);
 conn4= hAllocConn(kgTempDb);
 
 sqlSafef(query2, sizeof(query2), "select * from %s.kgCandidate where alignID='%s'", kgTempDb, alignID);
 sr2 = sqlMustGetResult(conn2, query2);
 row2 = sqlNextRow(sr2);
 while (row2 != NULL)
     {
     mrnaID = row2[0];
     chrom = row2[1];
     ranking = row2[11];
     
     /* check if it is a composite mrnaID */
     /* if yes, select from entries with both protein and mrna specified */
     if (alignID[0] == 'U') 
     	{
 	chp = strstr(row2[0], "_");
 	*chp = '\0';
 	protAcc = row2[0];
 	chp ++;
 	mrnaID = chp;
     	sqlSafef(query3, sizeof(query3), 
     	      "select protAcc, score from %s.protMrnaScore where mrnaAcc='%s' and protAcc='%s'",
 	      kgTempDb, mrnaID, protAcc);
 	}
     else
     	{
     	sqlSafef(query3, sizeof(query3), 
     	      "select protAcc, score from %s.protMrnaScore where mrnaAcc='%s' order by score desc",
 	      kgTempDb, mrnaID);
 	}
 	
     sr3  = sqlMustGetResult(conn3, query3);
     row3 = sqlNextRow(sr3);
 	      
     while(row3 != NULL)
         {
 	protAcc = row3[0];
 	score   = row3[1];
 
 	chp = strstr(protAcc, "-");
 	if (chp == NULL)
 	    {
-            sqlSafefFrag(condStr, sizeof(condStr), "acc='%s'", protAcc);
+            sqlSafef(condStr, sizeof(condStr), "acc='%s'", protAcc);
 	    isCurated = sqlGetField(spDb, "info", "isCurated", condStr);
 	    if (sameWord(isCurated, "1"))
 	    	{
 		protDbId = 1;
 		}
 	    else
 	    	{
 		protDbId = 2;
 		}
 	    }
    	else
 	    {
 	    protDbId = 4;
 	    }
 	    
 	fprintf(outf, "%s:", chrom);
 	for (i=0; i<cdsCnt; i++) fprintf(outf, "%s", cdsBloc[i]);
 	fprintf(outf, "\t%s\t%d\t%8s\t%s\t%s\t%s\n", 
 		ranking, protDbId, score, mrnaID, protAcc, alignID);
 
 	/* for composite type, process just one record */ 
         if (alignID[0] == 'U') break; 
 	row3 = sqlNextRow(sr3);
 	}
     sqlFreeResult(&sr3);
     row2 = sqlNextRow(sr2);
     }
 sqlFreeResult(&sr2);
 hFreeConn(&conn2);
 hFreeConn(&conn3);
 hFreeConn(&conn4);
 }
 
 void kgGetCds(char *db, char *spDb, char *geneTable, FILE *outf)
 /* get CDS info */
 {
 struct sqlConnection *conn = NULL;
 struct genePred *gp;
 int cdsCnt;
 struct genePredReader *gpr;
 int iExon, exonStart, exonEnd;
 
 if (db != NULL)
     conn = sqlConnect(db);
 
 gpr = genePredReaderQuery(conn, geneTable, NULL);
 while ((gp = genePredReaderNext(gpr)) != NULL)
     {
     cdsCnt = 0;
     for (iExon = 0; iExon < gp->exonCount; iExon++)
     	{
     	if (genePredCdsExon(gp, iExon, &exonStart, &exonEnd))
     	    {
 	    sprintf(cdsBloc[cdsCnt], "%d-%d;", exonStart, exonEnd);
 	    
 	    cdsCnt++;
 	    }
     	}
     if (cdsCnt > 0) 
     	{
 	processAlign(db, spDb, gp->name, cdsCnt, outf);
 	}
     else
     	{
 	fprintf(stderr, "%s does not have cds.\n", gp->name);
 	}
     }
 sqlDisconnect(&conn);
 }
 
 int main(int argc, char *argv[])
 /* Process command line. */
 {
 char *kgTempDb;
 char *kgCandidateTable;
 char *outFileName;
 FILE *outf;
 char *spDbDate, spDb[100];
 
 if (argc != 5)
     {
     usage("wrong number of args");
     }
 else
     {
     kgTempDb         = argv[1];
     spDbDate   	     = argv[2];
     kgCandidateTable = argv[3];
     outFileName	     = argv[4];
 
     sprintf(spDb, "sp%s", spDbDate);
     outf = mustOpen(outFileName, "w");
 
     kgGetCds(kgTempDb, spDb, kgCandidateTable, outf);
     fclose(outf);
     }
 return 0;
 }