remove --no-pan and report better the errors

pangenome · Nov 1, 2023 · 24b423c · 24b423c
1 parent fa7a6ca
commit 24b423c
Show file tree

Hide file tree

Showing 2 changed files with 37 additions and 39 deletions.
diff --git a/partition-before-pggb b/partition-before-pggb
@@ -67,19 +67,17 @@ vcf_spec=false
 multiqc=false
 
 # default values
-OUTPUT_DIR=$(pwd)
 THREADS=$(getconf _NPROCESSORS_ONLN 2>/dev/null || getconf NPROCESSORS_ONLN 2>/dev/null || echo 1)
 
 # general parameters
-output_dir=$OUTPUT_DIR
+output_dir=false
 input_temp_dir=false
 input_paf=false
 resume=false
 threads=$THREADS
 poa_threads=0
 keep_intermediate_files=false
 compress=false
-respect_pansn=true
 show_version=false
 show_help=false
 
@@ -118,7 +116,7 @@ fi
 
 # read the options
 cmd=$0" "$@
-TEMP=`getopt -o i:o:D:a:p:c:s:l:K:F:k:x:f:B:Xn:j:P:O:Me:t:T:vhASY:G:Q:d:I:R:NbrmZzV: --long input-fasta:,output-dir:,temp-dir:,input-paf:,map-pct-id:,n-mappings:,segment-length:,block-length-min:,mash-kmer:,mash-kmer-thres:,min-match-length:,sparse-map:,sparse-factor:,transclose-batch:,skip-normalization,n-haplotypes:,path-jump-max:,subpath-min:,edge-jump-max:,threads:,poa-threads:,skip-viz,do-layout,help,no-merge-segments,stats,exclude-delim:,poa-length-target:,poa-params:,poa-padding:,run-abpoa,global-poa,write-maf,consensus-spec:,consensus-prefix:,pad-max-depth:,block-id-min:,block-ratio-min:,no-splits,resume,keep-temp-files,multiqc,compress,vcf-spec:,no-pansn,version -n 'pggb' -- "$@"`
+TEMP=`getopt -o i:o:D:a:p:c:s:l:K:F:k:x:f:B:Xn:j:P:O:Me:t:T:vhASY:G:Q:d:I:R:NbrmZzV: --long input-fasta:,output-dir:,temp-dir:,input-paf:,map-pct-id:,n-mappings:,segment-length:,block-length-min:,mash-kmer:,mash-kmer-thres:,min-match-length:,sparse-map:,sparse-factor:,transclose-batch:,skip-normalization,n-haplotypes:,path-jump-max:,subpath-min:,edge-jump-max:,threads:,poa-threads:,skip-viz,do-layout,help,no-merge-segments,stats,exclude-delim:,poa-length-target:,poa-params:,poa-padding:,run-abpoa,global-poa,write-maf,consensus-spec:,consensus-prefix:,pad-max-depth:,block-id-min:,block-ratio-min:,no-splits,resume,keep-temp-files,multiqc,compress,vcf-spec:,version -n 'pggb' -- "$@"`
 eval set -- "$TEMP"
 
 # extract options and their arguments into variables.
@@ -162,7 +160,6 @@ while true ; do
         -T|--poa-threads) poa_threads=$2 ; shift 2 ;;
         -A|--keep-temp-files) keep_intermediate_files=true ; shift ;;
         -Z|--compress) compress=true ; shift ;;
-        --no-pansn) respect_pansn=false ; shift ;; 
         --version) show_version=true ; shift ;;
         -h|--help) show_help=true ; shift ;;
         --) shift ; break ;;
@@ -182,7 +179,7 @@ fi
 # Check input parameters
 if [ "$show_help" == true ]; then
     padding=$(printf %${#0}s) # prints as many spaces as the length of $0
-    echo "usage: $0 -i <input-fasta> [options]"
+    echo "usage: $0 -i <input-fasta> -o <output-dir> [options]"
     echo "options:"
     echo "   [wfmash]"
     echo "    -i, --input-fasta FILE      input FASTA/FASTQ file"
@@ -248,24 +245,17 @@ if [ "$show_help" == true ]; then
     echo "    -A, --keep-temp-files       keep intermediate graphs"
     echo "    -Z, --compress              compress alignment (.paf), graph (.gfa, .og), and MSA (.maf) outputs with pigz,"
     echo "                                and variant (.vcf) outputs with bgzip"
-    echo "    --no-pansn                  Pangenome Sequence Naming (PanSN) not required for sequence names"
     echo "    --version                   display the version of pggb"
     echo "    -h, --help                  this text"
     echo
     echo "Use wfmash, seqwish, smoothxg, odgi, gfaffix, and vg to build, project and display a pangenome graph."
     exit
-elif [ "$input_fasta" = "false" ]; then
-    >&2 echo "[pggb] ERROR: mandatory argument: -i/--input-fasta"
+elif [ "$input_fasta" = "false" ] || [ "$output_dir" = "false" ]; then
+    >&2 echo "[pggb] ERROR: mandatory argument: -i/--input-fasta and -o/--output-dir"
     exit
 elif [ ! -f "${input_fasta}.fai" ]; then
     echo "[pggb] ERROR: index for $input_fasta does not exist. Please create it using 'samtools faidx $input_fasta'."
     exit 1
-elif [ "$respect_pansn" = "false" ] && { [ "$n_haps" = "false" ] || [ "$n_haps" -lt 1 ]; }; then
-    >&2 echo "[pggb] ERROR: when no-pansn is set, -n/--n-haplotypes must be greater than or equal to 1"
-    exit
-elif [ "$respect_pansn" = "false" -a "$vcf_spec" != "false" ]; then
-    >&2 echo "[pggb] ERROR: -V/--vcf-spec and --no-pansn are incompatible"
-    exit
 elif [ "$n_mappings" -lt 1 ]; then
     >&2 echo "[pggb] ERROR: -c/--n-mappings must be greater than or equal to 1"
     exit
@@ -280,10 +270,19 @@ while IFS= read -r line; do
   fi
 done < <(cut -f 1 "${input_fasta}.fai")
 if [ "$pansn_not_respected" != "false" ]; then
-  if [ "$respect_pansn" = "false" ]; then
-    >&2 echo "[pggb] warning: there are sequence names (like '$pansn_not_respected') that do not match the Pangenome Sequence Naming (PanSN)."
-  else
-    >&2 echo "[pggb] ERROR: there are sequence names (like '$pansn_not_respected') that do not match the Pangenome Sequence Naming (PanSN). Set --pan-sn to disable this requirement."
+  >&2 echo "[pggb] warning: there are sequence names (like '$pansn_not_respected') that do not match the Pangenome Sequence Naming (PanSN)."
+
+  pansn_error=false
+  if [ "$vcf_spec" != "false" ]; then
+      pansn_error=true
+      >&2 echo "[pggb] ERROR: -V/--vcf-spec cannot be used if the Pangenome Sequence Naming (PanSN) is not respected."
+  fi
+  if [ "$n_haps" = "false" ] || [ "$n_haps" -lt 1 ]; then
+      pansn_error=true
+      >&2 echo "[pggb] ERROR: -n/--n-haplotypes must be greater than or equal to 1 when the Pangenome Sequence Naming (PanSN) is not respected."
+  fi
+
+  if [ $pansn_error = true ]; then
     exit
   fi
 fi

diff --git a/pggb b/pggb
@@ -67,19 +67,17 @@ vcf_spec=false
 multiqc=false
 
 # default values
-OUTPUT_DIR=$(pwd)
 THREADS=$(getconf _NPROCESSORS_ONLN 2>/dev/null || getconf NPROCESSORS_ONLN 2>/dev/null || echo 1)
 
 # general parameters
-output_dir=$OUTPUT_DIR
+output_dir=false
 input_temp_dir=false
 input_paf=false
 resume=false
 threads=$THREADS
 poa_threads=0
 keep_intermediate_files=false
 compress=false
-respect_pansn=true
 show_version=false
 show_help=false
 
@@ -118,7 +116,7 @@ fi
 
 # read the options
 cmd=$0" "$@
-TEMP=`getopt -o i:o:D:a:p:c:s:l:K:F:k:x:f:B:Xn:j:P:O:Me:t:T:vhASY:G:Q:d:I:R:NbrmZzV: --long input-fasta:,output-dir:,temp-dir:,input-paf:,map-pct-id:,n-mappings:,segment-length:,block-length-min:,mash-kmer:,mash-kmer-thres:,min-match-length:,sparse-map:,sparse-factor:,transclose-batch:,skip-normalization,n-haplotypes:,path-jump-max:,subpath-min:,edge-jump-max:,threads:,poa-threads:,skip-viz,do-layout,help,no-merge-segments,stats,exclude-delim:,poa-length-target:,poa-params:,poa-padding:,run-abpoa,global-poa,write-maf,consensus-spec:,consensus-prefix:,pad-max-depth:,block-id-min:,block-ratio-min:,no-splits,resume,keep-temp-files,multiqc,compress,vcf-spec:,no-pansn,version -n 'pggb' -- "$@"`
+TEMP=`getopt -o i:o:D:a:p:c:s:l:K:F:k:x:f:B:Xn:j:P:O:Me:t:T:vhASY:G:Q:d:I:R:NbrmZzV: --long input-fasta:,output-dir:,temp-dir:,input-paf:,map-pct-id:,n-mappings:,segment-length:,block-length-min:,mash-kmer:,mash-kmer-thres:,min-match-length:,sparse-map:,sparse-factor:,transclose-batch:,skip-normalization,n-haplotypes:,path-jump-max:,subpath-min:,edge-jump-max:,threads:,poa-threads:,skip-viz,do-layout,help,no-merge-segments,stats,exclude-delim:,poa-length-target:,poa-params:,poa-padding:,run-abpoa,global-poa,write-maf,consensus-spec:,consensus-prefix:,pad-max-depth:,block-id-min:,block-ratio-min:,no-splits,resume,keep-temp-files,multiqc,compress,vcf-spec:,version -n 'pggb' -- "$@"`
 eval set -- "$TEMP"
 
 # extract options and their arguments into variables.
@@ -162,7 +160,6 @@ while true ; do
         -T|--poa-threads) poa_threads=$2 ; shift 2 ;;
         -A|--keep-temp-files) keep_intermediate_files=true ; shift ;;
         -Z|--compress) compress=true ; shift ;;
-        --no-pansn) respect_pansn=false ; shift ;; 
         --version) show_version=true ; shift ;;
         -h|--help) show_help=true ; shift ;;
         --) shift ; break ;;
@@ -182,7 +179,7 @@ fi
 # Check input parameters
 if [ "$show_help" == true ]; then
     padding=$(printf %${#0}s) # prints as many spaces as the length of $0
-    echo "usage: $0 -i <input-fasta> [options]"
+    echo "usage: $0 -i <input-fasta> -o <output-dir> [options]"
     echo "options:"
     echo "   [wfmash]"
     echo "    -i, --input-fasta FILE      input FASTA/FASTQ file"
@@ -248,24 +245,17 @@ if [ "$show_help" == true ]; then
     echo "    -A, --keep-temp-files       keep intermediate graphs"
     echo "    -Z, --compress              compress alignment (.paf), graph (.gfa, .og), and MSA (.maf) outputs with pigz,"
     echo "                                and variant (.vcf) outputs with bgzip"
-    echo "    --no-pansn                  Pangenome Sequence Naming (PanSN) not required for sequence names"
     echo "    --version                   display the version of pggb"
     echo "    -h, --help                  this text"
     echo
     echo "Use wfmash, seqwish, smoothxg, odgi, gfaffix, and vg to build, project and display a pangenome graph."
     exit
-elif [ "$input_fasta" = "false" ]; then
-    >&2 echo "[pggb] ERROR: mandatory argument: -i/--input-fasta"
+elif [ "$input_fasta" = "false" ] || [ "$output_dir" = "false" ]; then
+    >&2 echo "[pggb] ERROR: mandatory argument: -i/--input-fasta and -o/--output-dir"
     exit
 elif [ ! -f "${input_fasta}.fai" ]; then
     echo "[pggb] ERROR: index for $input_fasta does not exist. Please create it using 'samtools faidx $input_fasta'."
     exit 1
-elif [ "$respect_pansn" = "false" ] && { [ "$n_haps" = "false" ] || [ "$n_haps" -lt 1 ]; }; then
-    >&2 echo "[pggb] ERROR: when no-pansn is set, -n/--n-haplotypes must be greater than or equal to 1"
-    exit
-elif [ "$respect_pansn" = "false" -a "$vcf_spec" != "false" ]; then
-    >&2 echo "[pggb] ERROR: -V/--vcf-spec and --no-pansn are incompatible"
-    exit
 elif [ "$n_mappings" -lt 1 ]; then
     >&2 echo "[pggb] ERROR: -c/--n-mappings must be greater than or equal to 1"
     exit
@@ -280,10 +270,19 @@ while IFS= read -r line; do
   fi
 done < <(cut -f 1 "${input_fasta}.fai")
 if [ "$pansn_not_respected" != "false" ]; then
-  if [ "$respect_pansn" = "false" ]; then
-    >&2 echo "[pggb] warning: there are sequence names (like '$pansn_not_respected') that do not match the Pangenome Sequence Naming (PanSN)."
-  else
-    >&2 echo "[pggb] ERROR: there are sequence names (like '$pansn_not_respected') that do not match the Pangenome Sequence Naming (PanSN). Set --pan-sn to disable this requirement."
+  >&2 echo "[pggb] warning: there are sequence names (like '$pansn_not_respected') that do not match the Pangenome Sequence Naming (PanSN)."
+
+  pansn_error=false
+  if [ "$vcf_spec" != "false" ]; then
+      pansn_error=true
+      >&2 echo "[pggb] ERROR: -V/--vcf-spec cannot be used if the Pangenome Sequence Naming (PanSN) is not respected."
+  fi
+  if [ "$n_haps" = "false" ] || [ "$n_haps" -lt 1 ]; then
+      pansn_error=true
+      >&2 echo "[pggb] ERROR: -n/--n-haplotypes must be greater than or equal to 1 when the Pangenome Sequence Naming (PanSN) is not respected."
+  fi
+
+  if [ $pansn_error = true ]; then
     exit
   fi
 fi
@@ -478,7 +477,7 @@ reporting:
   multiqc:            $multiqc
 EOT
 
-
+#-------------------------------------------------------------------------------
 echo -e "\nRunning pggb\n" >> "$log_file"
 
 if [[ "$input_paf" == false ]]; then