Merge branch 'cw-4704' into 'dev'

Support protocol_run_id in RG.ID [CW-4704]

See merge request epi2melabs/fastcat!72

CHANGELOG.md

@@ -4,7 +4,9 @@ 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).
 
-
+## [v0.18.6]
+### Changed
+- Ill advised parsing of RG ID field has been extended to additionally allow for protocol_run_id style (uuid) Run IDs, as well as standalone acquisition_id (sha1) Run IDs
 
 ## [v0.18.5]
 ### Changed

src/common.c

@@ -253,7 +253,9 @@ void destroy_rg_info(readgroup* rg) {
 // <runid>_<basecalling_model>_<barcode_arrangement>
 //
 // where:
-//   - runid is a 40 character string
+//   - runid is either (see CW-4704):
+//        - a 40 character string representing an acquisition_id sha
+//        - a 36 character string representing a protocol_run_id uuid
 //   - basecalling_model is a string maybe containing `_`, and containing one or more `@`
 //   -    mod_caller is optional part of this starting with `_` after the first `@`
 //   - barcode_arrangement is a optional(!) string with an unknown format, but hopefully no `@`
@@ -284,7 +286,9 @@ readgroup* create_rg_info(char* rg) {
         return rg_info;
     }
     delim[0] = '\0';
-    if (strlen(rg_info->runid) != 40) {
+    // ensure runid is long enough to be an acquisition sha or protocol uuid
+    int runid_l = strlen(rg_info->runid);
+    if (runid_l != 36 && runid_l != 40) {
         // free the mutated copy, and reset
         free(rg_info->readgroup);
         rg_info->readgroup = strdup(rg);

src/version.h

@@ -1,2 +1,2 @@
 
-const char *argp_program_version = "0.18.5";
+const char *argp_program_version = "0.18.6";

test/rg_parse.c

@@ -26,25 +26,34 @@ int compare(char* str1, char* str2) {
 
 
 int main() {
-    char *runid = "ef1af1ab8967cb20ca30dbeca93fd66592bf4619";
+    char *runid_acquisition = "ef1af1ab8967cb20ca30dbeca93fd66592bf4619";
+    char *runid_protocol = "c886531d-28f5-41f6-b948-948e8cb78e5e";
     char *basecall_model = "[email protected]";
     char *mod_model_name = "[email protected]_5mCG_5hmCG@v1";
     char *barcode = "barcode01";
     char *suffix = "-1A2B3C4D";
 
     TestCase cases[] = {
-        {runid, basecall_model, mod_model_name, barcode, suffix},
-        {runid, basecall_model, mod_model_name, barcode, NULL},
-        {runid, basecall_model, mod_model_name, NULL, suffix},
-        {runid, basecall_model, mod_model_name, NULL, NULL},
-        {runid, basecall_model, NULL, barcode, suffix},
-        {runid, basecall_model, NULL, barcode, NULL},
-        {runid, basecall_model, NULL, NULL, suffix},
-        {runid, basecall_model, NULL, NULL, NULL},
+        {runid_acquisition, basecall_model, mod_model_name, barcode, suffix},
+        {runid_acquisition, basecall_model, mod_model_name, barcode, NULL},
+        {runid_acquisition, basecall_model, mod_model_name, NULL, suffix},
+        {runid_acquisition, basecall_model, mod_model_name, NULL, NULL},
+        {runid_acquisition, basecall_model, NULL, barcode, suffix},
+        {runid_acquisition, basecall_model, NULL, barcode, NULL},
+        {runid_acquisition, basecall_model, NULL, NULL, suffix},
+        {runid_acquisition, basecall_model, NULL, NULL, NULL},
+        {runid_protocol, basecall_model, mod_model_name, barcode, suffix},
+        {runid_protocol, basecall_model, mod_model_name, barcode, NULL},
+        {runid_protocol, basecall_model, mod_model_name, NULL, suffix},
+        {runid_protocol, basecall_model, mod_model_name, NULL, NULL},
+        {runid_protocol, basecall_model, NULL, barcode, suffix},
+        {runid_protocol, basecall_model, NULL, barcode, NULL},
+        {runid_protocol, basecall_model, NULL, NULL, suffix},
+        {runid_protocol, basecall_model, NULL, NULL, NULL},
     };
 
     int fails = 0;
-    for (int i = 0; i < 8; i++) {
+    for (int i = 0; i < sizeof(cases)/sizeof(TestCase); i++) {
 
         char* read_group = calloc(400, sizeof(char));
         read_group = strcpy(read_group, cases[i].runid);