Merge pull request #70 from ICGC-TCGA-PanCancer/dev

Looks good, happy for release to go ahead

Changes

@@ -1,3 +1,7 @@
+1.5.4
+  bam_stats C 
+  * Reference file parameter is now optional to replicate bam_stats.pl functionality. 
+  * Warnings in help, and when a cram file is given as reference from header may not be found, and bam stats will fail. 
 1.5.3
   bam_stats C - changed array for khash in insert size calculations in order to make code more robust.
   Header RG line reading now reads anything not a tab or newline as it should when determining what the values of tags are.

MYMETA.json

@@ -4,7 +4,7 @@
       "unknown"
    ],
    "dynamic_config" : 0,
-   "generated_by" : "ExtUtils::MakeMaker version 6.68, CPAN::Meta::Converter version 2.131560",
+   "generated_by" : "ExtUtils::MakeMaker version 6.68, CPAN::Meta::Converter version 2.143240",
    "license" : [
       "gpl_2"
    ],
@@ -55,5 +55,5 @@
       }
    },
    "release_status" : "stable",
-   "version" : "v1.5.3"
+   "version" : "v1.5.4"
 }

MYMETA.yml

@@ -3,38 +3,38 @@ abstract: unknown
 author:
   - unknown
 build_requires:
-  ExtUtils::MakeMaker: 0
+  ExtUtils::MakeMaker: '0'
 configure_requires:
-  ExtUtils::MakeMaker: 0
+  ExtUtils::MakeMaker: '0'
 dynamic_config: 0
-generated_by: 'ExtUtils::MakeMaker version 6.68, CPAN::Meta::Converter version 2.131560'
+generated_by: 'ExtUtils::MakeMaker version 6.68, CPAN::Meta::Converter version 2.143240'
 license: gpl
 meta-spec:
   url: http://module-build.sourceforge.net/META-spec-v1.4.html
-  version: 1.4
+  version: '1.4'
 name: PCAP
 no_index:
   directory:
     - t
     - inc
 requires:
-  Bio::DB::Sam: 1.39
-  Bio::Root::Version: 1.006923
-  Capture::Tiny: 0.24
-  Const::Fast: 0.014
-  Data::UUID: 1.219
-  Devel::Cover: 1.09
-  File::Which: 0.05
-  GD: 2.52
-  IPC::System::Simple: 1.25
-  List::Util: 1.38
-  Math::Gradient: 0.04
-  Module::Build: 0.42
-  Pod::Coverage: 0.23
-  Proc::ProcessTable: 0.5
-  Sub::Exporter::Progressive: 0.001011
-  Term::UI: 0.42
-  Test::Fatal: 0.013
-  Try::Tiny: 0.19
-  XML::Simple: 2.2
-version: v1.5.3
+  Bio::DB::Sam: '1.39'
+  Bio::Root::Version: '1.006923'
+  Capture::Tiny: '0.24'
+  Const::Fast: '0.014'
+  Data::UUID: '1.219'
+  Devel::Cover: '1.09'
+  File::Which: '0.05'
+  GD: '2.52'
+  IPC::System::Simple: '1.25'
+  List::Util: '1.38'
+  Math::Gradient: '0.04'
+  Module::Build: '0.42'
+  Pod::Coverage: '0.23'
+  Proc::ProcessTable: '0.5'
+  Sub::Exporter::Progressive: '0.001011'
+  Term::UI: '0.42'
+  Test::Fatal: '0.013'
+  Try::Tiny: '0.19'
+  XML::Simple: '2.2'
+version: v1.5.4

c/Makefile

@@ -1,4 +1,4 @@
-VERSION=1.5.3
+VERSION=1.5.4
 
 #Compiler
 CC = gcc -O3 -DVERSION='"$(VERSION)"' -g

c/bam_stats.c

@@ -56,11 +56,12 @@ void print_version (int exit_code){
 
 void print_usage (int exit_code){
 
-	printf ("Usage: bam_stats [-i file] [-o file] [-p plots] [-r reference.fa.fai] [-h] [-v]\n\n");
+	printf ("Usage: bam_stats -i file -o file [-p plots] [-r reference.fa.fai] [-h] [-v]\n\n");
   printf ("-i --input     File path to read in.\n");
-  printf ("-r --ref-file  File path to reference index (.fai) file.\n");
   printf ("-o --output    File path to output.\n\n");
 	printf ("Optional:\n");
+	printf ("-r --ref-file  File path to reference index (.fai) file.\n");
+	printf ("               NB. If cram format is supplied via -b and the reference listed in the cram header can't be found bam_stats may fail to work correctly.\n");
 	printf ("-p --plots     Folder to contain quality score plots.\n\n");
 	printf ("Other:\n");
 	printf ("-h --help      Display this usage information.\n");
@@ -69,6 +70,9 @@ void print_usage (int exit_code){
 }
 
 void options(int argc, char *argv[]){
+
+  ref_file = NULL;
+
 	const struct option long_opts[] =
 	  {
              	{"version", no_argument, 0, 'v'},
@@ -128,9 +132,11 @@ void options(int argc, char *argv[]){
    	printf("Input file (-i) %s does not exist.\n",input_file);
    	print_usage(1);
    }
-   if(check_exist(ref_file) != 1){
-   	printf("Reference fasta index file (-r) %s does not exist.\n",ref_file);
-   	print_usage(1);
+   if(ref_file){
+     if(check_exist(ref_file) != 1){
+      printf("Reference fasta index file (-r) %s does not exist.\n",ref_file);
+      print_usage(1);
+     }
    }
 
    return;
@@ -338,7 +344,12 @@ int main(int argc, char *argv[]){
   //Open bam file as object
   input = hts_open(input_file,"r");
   //Set reference index file
-  hts_set_fai_filename(input, ref_file);
+  if(ref_file){
+    hts_set_fai_filename(input, ref_file);
+  }else{
+    if(input->format.format == cram) log_warn("No reference file provided for a cram input file, if the reference described in the cram header can't be located bam_stats may fail.");
+  }
+
   //Read header from bam file
   head = sam_hdr_read(input);
   rg_info_t **grps = parse_header(head, &grps_size, &grp_stats);

lib/PCAP.pm

@@ -24,7 +24,7 @@ use strict;
 use Const::Fast qw(const);
 use base 'Exporter';
 
-our $VERSION = '1.5.3';
+our $VERSION = '1.5.4';
 our @EXPORT = qw($VERSION);
 
 const my $LICENSE =>
@@ -58,6 +58,7 @@ const my %UPGRADE_PATH => ( '0.1.0'  => 'biobambam,bwa,samtools',
                             '1.5.1'  => '',
                             '1.5.2'  => '',
                             '1.5.3'  => '',
+                            '1.5.4'  => '',
                           );
 
 sub license {

setup.sh

@@ -5,7 +5,7 @@ SOURCE_BWA="https://github.com/lh3/bwa/archive/0.7.10.tar.gz"
 SOURCE_SAMTOOLS="https://github.com/samtools/samtools/archive/0.1.20.tar.gz"
 
 # for bamstats
-SOURCE_HTSLIB="https://github.com/samtools/htslib/archive/1.1.tar.gz"
+SOURCE_HTSLIB="https://github.com/samtools/htslib/archive/1.2.1.tar.gz"
 
 # for bigwig
 SOURCE_JKENT_BIN="https://github.com/ENCODE-DCC/kentUtils/raw/master/bin/linux.x86_64"