Fix a segfault on missing tags

If the filtering expression queried keys from a file, for example

    -i 'INFO/TAG=@strings_expected'

the program would segfault if INFO/TAG was not present in the first
VCF record. This is now fixed by making sure NULL is not passed to
khash_str2int_has_key.

Fixes #2111

Author: pd3

filter.c

@@ -1,6 +1,6 @@
 /*  filter.c -- filter expressions.
 
-    Copyright (C) 2013-2023 Genome Research Ltd.
+    Copyright (C) 2013-2024 Genome Research Ltd.
 
     Author: Petr Danecek <[email protected]>
 
@@ -615,7 +615,7 @@ static void filters_cmp_string_hash(token_t *atok, token_t *btok, token_t *rtok,
         // there is only one string value, e.g. STR[1][email protected]
         if ( btok->idx >= 0 )
         {
-            int ret = khash_str2int_has_key(atok->hash, btok->str_value.s);
+            int ret = btok->str_value.s ? khash_str2int_has_key(atok->hash, btok->str_value.s) : 0;
             if ( rtok->tok_type==TOK_NE ) ret = ret ? 0 : 1;
             rtok->pass_site = ret;
             return;

test/filter.string.1.1.out

@@ -0,0 +1,8 @@
+##fileformat=VCFv4.2
+##FILTER=<ID=PASS,Description="All filters passed">
+##INFO=<ID=DP,Number=1,Type=Integer,Description="Approximate read depth (informative and non-informative); some reads may have been filtered based on mapq etc.">
+##INFO=<ID=TAG,Number=1,Type=String,Description="This is an example of a string in info.">
+##FORMAT=<ID=TAG,Number=1,Type=String,Description="This is an example of a string in format.">
+##contig=<ID=chr1,length=248956422>
+#CHROM	POS	ID	REF	ALT	QUAL	FILTER	INFO	FORMAT	sample
+chr1	11558105	.	G	C	.	PASS	DP=61;TAG=Example	TAG	Example

test/filter.string.1.txt

@@ -0,0 +1,2 @@
+Example
+Something

test/filter.string.1.vcf

@@ -0,0 +1,10 @@
+##fileformat=VCFv4.2
+##FILTER=<ID=PASS,Description="All filters passed">
+##INFO=<ID=DP,Number=1,Type=Integer,Description="Approximate read depth (informative and non-informative); some reads may have been filtered based on mapq etc.">
+##INFO=<ID=TAG,Number=1,Type=String,Description="This is an example of a string in info.">
+##FORMAT=<ID=TAG,Number=1,Type=String,Description="This is an example of a string in format.">
+##contig=<ID=chr1,length=248956422>
+#CHROM	POS	ID	REF	ALT	QUAL	FILTER	INFO	FORMAT	sample
+chr1	11558102	.	G	GT	.	PASS	DP=61	.	.
+chr1	11558105	.	G	C	.	PASS	DP=61;TAG=Example	TAG	Example
+chr1	11558108	.	A	T	.	PASS	DP=61;TAG=It	TAG	It

test/test.pl

@@ -305,6 +305,8 @@
 run_test(\&test_vcf_norm,$opts,in=>'norm.split.merge',out=>'norm.split.merge.4.out',args=>['-m -','-m +any']);
 run_test(\&test_vcf_norm,$opts,in=>'norm.merge.4',out=>'norm.merge.4.1.out',args=>'-m +any');
 run_test(\&test_vcf_norm,$opts,in=>'norm.merge.4',out=>'norm.merge.4.2.out',args=>'-m +both');
+run_test(\&test_vcf_view,$opts,in=>'filter.string.1',out=>'filter.string.1.1.out',args=>q[-i 'INFO/TAG=@{PATH}/filter.string.1.txt']);
+run_test(\&test_vcf_view,$opts,in=>'filter.string.1',out=>'filter.string.1.1.out',args=>q[-i 'FMT/TAG=@{PATH}/filter.string.1.txt']);
 run_test(\&test_vcf_view,$opts,in=>'merge.gvcf.2.a',out=>'merge.gvcf.2.a.1.out',args=>'-HA');
 run_test(\&test_vcf_view,$opts,in=>'merge.gvcf.2.a',out=>'merge.gvcf.2.a.2.out',args=>'-HAA');
 run_test(\&test_vcf_view,$opts,in=>'weird-chr-names',out=>'weird-chr-names.1.out',args=>'',reg=>'-r 1');