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/near/knownToHprd/knownToHprd.c src/hg/near/knownToHprd/knownToHprd.c
index 5d2c87a..7740edb 100644
--- src/hg/near/knownToHprd/knownToHprd.c
+++ src/hg/near/knownToHprd/knownToHprd.c
@@ -1,138 +1,140 @@
 /* knownToHprd - Create knownToHprd table by reading HPRD FLAT_FILES and using kgXref.
  *   This utilility should make automating incremental kg build easier. 
  *   It is assumed that the HPRD flat files have been decompressed in
  *   a directory like /cluster/data/hg18/p2p/hprd/FLAT_FILES/ 
  *   The data has this structure:
 [hgwdev:FLAT_FILES> head HPRD_ID_MAPPINGS.txt
 00001   ALDH1A1 NM_000689.3     NP_000680.2     216     100640  P00352  Aldehyde dehydrogenase 1
  */
 
 /* 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 "linefile.h"
 #include "hash.h"
 #include "dystring.h"
 #include "options.h"
 #include "jksql.h"
 #include "hdb.h"
 
 void usage()
 /* Explain usage and exit. */
 {
 errAbort(
   "knownToHprd - Create knownToHprd table using HPRD flat file and kgXref\n"
   "usage:\n"
   "   knownToHprd database hprdFile\n"
   "options:\n"
   "   -keepTab - temp tab file not removed during cleanup.\n"
   );
 }
 
 char *outTable = "knownToHprd";
 
 static struct optionSpec options[] = {
    {"keepTab", OPTION_BOOLEAN},
    {NULL, 0},
 };
 
 void createTable(struct sqlConnection *conn, char *tableName)
 /* Create our name/value table, dropping if it already exists. */
 {
 struct dyString *dy = dyStringNew(512);
 sqlDyStringPrintf(dy,
 "CREATE TABLE  %s (\n"
 "    name varchar(255) not null,\n"
 "    value varchar(255) not null,\n"
 "#Indices\n"
 "    PRIMARY KEY(name),\n"
 "    INDEX(value)\n"
 ")\n",  tableName);
 sqlRemakeTable(conn, tableName, dy->string);
 dyStringFree(&dy);
 }
 
 
 void knownToHprd(char *database, char *hprdFile)
 /* knownToHprd - Create knownToHprd table by using hprd flat file and kgXref. */
 {
 char *tempDir = ".";
 FILE *f = hgCreateTabFile(tempDir, outTable);
 struct lineFile *lf = lineFileOpen(hprdFile, TRUE);
 char *lrow[8];
 
 struct sqlConnection *hConn = sqlConnect(database);
 struct sqlResult *sr;
 char **row;
 struct hash *mRnaHash = newHash(18);
 struct hash *spIdHash = newHash(18);
 struct hash *kgIdHash = newHash(18);
 int missing=0;  // how many recs had no match via kgXref
 
-sr = sqlGetResult(hConn, NOSQLINJ "select kgId, mRNA, spId from kgXref");
+char query[1024];
+sqlSafef(query, sizeof query, "select kgId, mRNA, spId from kgXref");
+sr = sqlGetResult(hConn, query);
 while ((row = sqlNextRow(sr)) != NULL)
     {
     char *kgId = row[0];
     char *mRNA = row[1];
     char *spId = row[2];
     hashAdd(mRnaHash, mRNA, cloneString(kgId));
     if (spId && spId[0])
 	hashAdd(spIdHash, spId, cloneString(kgId));
     }
 sqlFreeResult(&sr);
 
 while (lineFileRowTab(lf, lrow))
     {
     char *hprdId = lrow[0];
     char *cDna = lrow[2];
     char *uniProt = lrow[6];
     struct hashEl *kgId = NULL;
     chopSuffixAt(cDna,'.');
 
     boolean found = FALSE;
     for (kgId = hashLookup(mRnaHash, cDna); kgId; kgId = hashLookupNext(kgId))
 	{
 	found = TRUE;
 	if (!hashLookup(kgIdHash, kgId->val))  // each kgId only once 
 	    {
 	    hashAdd(kgIdHash, kgId->val, NULL);
 	    fprintf(f,"%s\t%s\n",(char *)kgId->val,hprdId);
 	    }
 	}
 
     for (kgId = hashLookup(spIdHash, uniProt); kgId; kgId = hashLookupNext(kgId))
 	{
 	found = TRUE;
 	if (!hashLookup(kgIdHash, kgId->val))  // each kgId only once 
 	    {
 	    hashAdd(kgIdHash, kgId->val, NULL);
 	    fprintf(f,"%s\t%s\n",(char *)kgId->val,hprdId);
 	    }
 	}
 
     if (!found)
 	{
 	++missing;
 	verbose(2, "missing %s %s %s\n", hprdId, cDna, uniProt);
 	}
     }
 lineFileClose(&lf);
 carefulClose(&f);
 
 verbose(2, "missing=%d\n",missing);
 
 createTable(hConn, outTable);
 hgLoadTabFile(hConn, tempDir, outTable, &f);
 if (!optionExists("keepTab"))
     hgRemoveTabFile(tempDir, outTable);
 }
 
 int main(int argc, char *argv[])
 /* Process command line. */
 {
 optionInit(&argc, argv, options);
 if (argc != 3)
     usage();
 knownToHprd(argv[1],argv[2]);
 return 0;
 }