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/snp/snpLoad/snpSplitSimple.c src/hg/snp/snpLoad/snpSplitSimple.c
index 466f5cd..4f00011 100644
--- src/hg/snp/snpLoad/snpSplitSimple.c
+++ src/hg/snp/snpLoad/snpSplitSimple.c
@@ -1,182 +1,182 @@
 /* snpSplitSimple - second step in dbSNP processing.
  * Split the ContigLocFilter table by distinct chrom from ContigInfo. 
  * ContigLocFilter contains chrom (added by snpContigLocFilter, the first step) */
 
 /* 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 "dystring.h"
 #include "hash.h"
 #include "hdb.h"
 
 
 static char *snpDb = NULL;
 static char *contigGroup = NULL;
 static struct hash *chromHash = NULL;
 
 void usage()
 /* Explain usage and exit. */
 {
 errAbort(
     "snpSplitSimple - split the ContigLocFilter table by chrom\n"
     "usage:\n"
     "    snpSplitSimple snpDb contigGroup\n");
 }
 
 
 struct hash *loadChroms(char *contigGroup)
 /* hash all chromNames that match contigGroup and open file for each */
 {
 struct hash *ret;
 char query[512];
 struct sqlConnection *conn = hAllocConn();
 struct sqlResult *sr;
 char **row;
 char *chromName;
 FILE *f;
 char fileName[64];
 
 ret = newHash(0);
 verbose(1, "getting chroms...\n");
 sqlSafef(query, sizeof(query), "select distinct(contig_chr) from ContigInfo where group_term = '%s' and contig_end != 0", contigGroup);
 sr = sqlGetResult(conn, query);
 while ((row = sqlNextRow(sr)) != NULL)
     {
     chromName = cloneString(row[0]);
     safef(fileName, ArraySize(fileName), "chr%s_snpTmp.tab", chromName);
     f = mustOpen(fileName, "w");
     hashAdd(ret, chromName, f);
     }
 sqlFreeResult(&sr);
 hFreeConn(&conn);
 return ret;
 }
 
 void writeSplitTables()
 /* sequentially read ContigLocFilter, writing to appropriate chrom file */
 /* we are storing chromName in the table just as a sanity check */
 /* It is dropped in the next step of the pipeline */
 {
 char query[512];
 struct sqlConnection *conn = hAllocConn();
 struct sqlResult *sr;
 char **row;
 struct hashEl *hel;
 char *chromName;
 
 verbose(1, "reading ContigLocFilter...\n");
 
 sqlSafef(query, sizeof(query), 
     "select snp_id, ctg_id, chromName, loc_type, phys_pos_from, phys_pos, orientation, allele from ContigLocFilter");
 
 sr = sqlGetResult(conn, query);
 while ((row = sqlNextRow(sr)) != NULL)
     {
     hel = hashLookup(chromHash,row[2]);
     if (hel == NULL)
         {
 	verbose(1, "%s not found\n", row[2]);
 	continue;
 	}
     fprintf(hel->val, "%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\n", row[0], row[1], row[2], row[3], row[4], row[5], row[6], row[7]);
     }
 sqlFreeResult(&sr);
 sqlDisconnect(&conn);
 }
 
 
 void createTable(char *chromName)
 /* create a chrN_snpTmp table */
 {
 struct sqlConnection *conn = hAllocConn();
 char tableName[64];
 char *createString =
 "CREATE TABLE %s (\n"
 "    snp_id int(11) not null,\n"
 "    ctg_id int(11) not null,\n"
 "    chromName char(32) not null,\n"
 "    loc_type tinyint(4) not null,\n"
 "    phys_pos_from int(11) not null,\n"
 "    phys_pos varchar(32),\n"
 "    orientation tinyint(4) not null,\n"
 "    allele blob\n"
 ");\n";
 
-struct dyString *dy = newDyString(1024);
+struct dyString *dy = dyStringNew(1024);
 
 safef(tableName, ArraySize(tableName), "chr%s_snpTmp", chromName);
 sqlDyStringPrintf(dy, createString, tableName);
 sqlRemakeTable(conn, tableName, dy->string);
 dyStringFree(&dy);
 hFreeConn(&conn);
 }
 
 
 void loadDatabase(char *chromName)
 /* load one table into database */
 {
 FILE *f;
 struct sqlConnection *conn = hAllocConn();
 char tableName[64], fileName[64];
 
 safef(tableName, ArraySize(tableName), "chr%s_snpTmp", chromName);
 safef(fileName, ArraySize(fileName), "chr%s_snpTmp.tab", chromName);
 
 f = mustOpen(fileName, "r");
 hgLoadTabFile(conn, ".", tableName, &f);
 
 hFreeConn(&conn);
 }
 
 
 
 int main(int argc, char *argv[])
 /* read ContigLocFilter, writing to individual chrom tables */
 {
 struct hashCookie cookie;
 struct hashEl *hel;
 char *chromName;
 
 if (argc != 3)
     usage();
 
 snpDb = argv[1];
 contigGroup = argv[2];
 hSetDb(snpDb);
 
 /* check for needed tables */
 if(!hTableExistsDb(snpDb, "ContigLocFilter"))
     errAbort("no ContigLocFilter table in %s\n", snpDb);
 if(!hTableExistsDb(snpDb, "ContigInfo"))
     errAbort("no ContigInfo table in %s\n", snpDb);
 
 chromHash = loadChroms(contigGroup);
 if (chromHash == NULL) 
     {
     verbose(1, "couldn't get chrom info\n");
     return 1;
     }
 
 writeSplitTables();
 
 verbose(1, "closing files...\n");
 cookie = hashFirst(chromHash);
 while (hel = hashNext(&cookie))
     fclose(hel->val);
 
 verbose(1, "creating tables...\n");
 cookie = hashFirst(chromHash);
 while ((chromName = hashNextName(&cookie)) != NULL)
     createTable(chromName);
 
 verbose(1, "loading database...\n");
 cookie = hashFirst(chromHash);
 while ((chromName = hashNextName(&cookie)) != NULL)
     {
     verbose(1, "chrom = %s\n", chromName);
     loadDatabase(chromName);
     }
 
 return 0;
 }