fix segfault on empty SAM tag [CW-4299]

CHANGELOG.md

@@ -4,6 +4,10 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [v0.18.4]
+### Fixed
+- Segfault on SAM-style tags without values in the FASTQ header.
+
 ## [v0.18.3]
 ### Fixed
 - Bug causing segfault on unlikely RG SAM tags in FASTQ header comments.

Makefile

@@ -214,7 +214,6 @@ regression_test_parse_rg_fastq: fastcat
 	done;
 	rm -r test/test-tmp
 
-
 .PHONY: regression_test_parse_rg_bam
 regression_test_parse_rg_bam: bamstats
 	if [ -d test/test-tmp ]; then rm -r test/test-tmp; fi
@@ -233,6 +232,10 @@ regression_test_parse_rd_fastq: fastcat
 	if [ -d test/test-tmp ]; then rm -r test/test-tmp; fi
 	mkdir test/test-tmp && \
 	cd test/test-tmp && \
-	../../fastcat ../parse_rd/RD-first-tag-and-no-RG.fastq -i rd > /dev/null && \
-	grep -q dummy_run_id rd
+	for i in ../parse_rd/*.fastq; do \
+		echo $$i; \
+		../../fastcat $$i --histograms hist -i rd > /dev/null; \
+		diff rd $$i.runids; \
+		rm -rf hist rg; \
+	done;
 	rm -r test/test-tmp

src/fastqcomments.c

@@ -87,6 +87,9 @@ read_meta parse_read_meta(kstring_t comment) {
             key = strtok_r(pch, ":", &p2);
             keytype = strtok_r(NULL, ":", &p2);
             value = strtok_r(NULL, "", &p2);
+            // we allow empty tags (e.g. 'RG:Z:'); in this case, `keytype` will be
+            // non-null, but `value` will be null; we set it to ""
+            if (keytype != NULL && value == NULL) value = "";
         }
         else {
             // split words on `=`
@@ -96,7 +99,7 @@ read_meta parse_read_meta(kstring_t comment) {
         }
 
         // if there was no delimiter in the word, value will be NULL --> add word to `rest`
-        if (value == NULL && !sam_tags) {
+        if (value == NULL) {
             ksprintf_with_opt_delim(meta->rest, " ", "%s", key);
         } else {
             if (!strcmp(key, "runid") || !strcmp(key, "RD")) {

src/version.h

@@ -1,2 +1,2 @@
 
-const char *argp_program_version = "0.18.3";
+const char *argp_program_version = "0.18.4";

test/parse_rd/RD-first-tag-and-no-RG-CW-4285.fastq.runids

@@ -0,0 +1,2 @@
+filename	run_id	count
+../parse_rd/RD-first-tag-and-no-RG-CW-4285.fastq	dummy_run_id	1

test/parse_rd/empty-RD-CW-4299.fastq

@@ -0,0 +1,4 @@
+@9ae4818c-61e9-4011-bf40-ca5721922da4	RD:Z:
+AAAAAAAAAAAAAAAAAA
++
+AAAAAAAAAAAAAAAAAA

test/parse_rd/empty-RD-CW-4299.fastq.runids

@@ -0,0 +1,2 @@
+filename	run_id	count
+../parse_rd/empty-RD-CW-4299.fastq		1