From 211ebf0d04f8851cd8c89f261d9c330ae74c5347 Mon Sep 17 00:00:00 2001
From: Shadi Zaheri <szaheri@broadinstitute.org>
Date: Wed, 24 Jul 2024 08:57:50 -0400
Subject: [PATCH 1/3] Add BarcodeExtractor workflow to extract the barcodes
 from each inut_bam

---
 .../TechAgnostic/Utility/BarcodeExtractor.wdl | 74 +++++++++++++++++++
 1 file changed, 74 insertions(+)
 create mode 100644 wdl/pipelines/TechAgnostic/Utility/BarcodeExtractor.wdl

diff --git a/wdl/pipelines/TechAgnostic/Utility/BarcodeExtractor.wdl b/wdl/pipelines/TechAgnostic/Utility/BarcodeExtractor.wdl
new file mode 100644
index 000000000..7043beb69
--- /dev/null
+++ b/wdl/pipelines/TechAgnostic/Utility/BarcodeExtractor.wdl
@@ -0,0 +1,74 @@
+version 1.0
+import "../../../structs/Structs.wdl"
+
+
+workflow BarcodeExtractor {
+    meta {
+        description: "Extract barcode from BAM file and create movie_name.barcode for each sample."
+    }
+    input {
+        File input_file
+        String id
+        String movie_name
+    }
+
+    call ExtractBarcode {
+        input: input_bam = input_file
+    }
+
+    output {
+        String barcode = ExtractBarcode.barcode
+        String movie_name_barcode = movie_name + "." + ExtractBarcode.barcode
+    }
+}
+
+task ExtractBarcode {
+    input {
+        File input_bam
+        RuntimeAttr? runtime_attr_override
+    }
+
+    command <<<
+    python3 <<CODE
+import re
+import sys
+
+def extract_barcode(input_bam):
+    match = re.search(r'\\.hifi_reads\\.(bc\\d+)\\.bam', input_bam)
+    return match.group(1) if match else None
+
+input_bam = "${input_bam}"
+barcode = extract_barcode(input_bam)
+if barcode:
+    print(barcode)
+else:
+    print("No barcode found", file=sys.stderr)
+CODE
+    >>>
+
+    output {
+        String barcode = read_string(stdout())
+    }
+
+    #########################
+    RuntimeAttr default_attr = object {
+        cpu_cores:          2,
+        mem_gb:             8,
+        disk_gb:            50,
+        boot_disk_gb:       10,
+        preemptible_tries:  3,
+        max_retries:        0,
+        docker:             "us.gcr.io/broad-dsp-lrma/lr-compile-sv-stats:0.0.0"
+    }
+    RuntimeAttr runtime_attr = select_first([runtime_attr_override, default_attr])
+
+    runtime {
+        cpu:                    select_first([runtime_attr.cpu_cores, default_attr.cpu_cores])
+        memory:                 select_first([runtime_attr.mem_gb, default_attr.mem_gb]) + " GiB"
+        disks:                  "local-disk " + select_first([runtime_attr.disk_gb, default_attr.disk_gb]) + " HDD"
+        bootDiskSizeGb:         select_first([runtime_attr.boot_disk_gb, default_attr.boot_disk_gb])
+        preemptible:            select_first([runtime_attr.preemptible_tries, default_attr.preemptible_tries])
+        maxRetries:             select_first([runtime_attr.max_retries, default_attr.max_retries])
+        docker:                 select_first([runtime_attr.docker, default_attr.docker])
+    }
+}

From b974cb21e9900f7c7af4e791486ea4ec495b0421 Mon Sep 17 00:00:00 2001
From: Shadi Zaheri <szaheri@broadinstitute.org>
Date: Wed, 24 Jul 2024 10:32:16 -0400
Subject: [PATCH 2/3] Modifying the extract funtion

---
 .../TechAgnostic/Utility/BarcodeExtractor.wdl | 23 ++++++++++---------
 1 file changed, 12 insertions(+), 11 deletions(-)

diff --git a/wdl/pipelines/TechAgnostic/Utility/BarcodeExtractor.wdl b/wdl/pipelines/TechAgnostic/Utility/BarcodeExtractor.wdl
index 7043beb69..bf89b1b30 100644
--- a/wdl/pipelines/TechAgnostic/Utility/BarcodeExtractor.wdl
+++ b/wdl/pipelines/TechAgnostic/Utility/BarcodeExtractor.wdl
@@ -1,6 +1,6 @@
 version 1.0
-import "../../../structs/Structs.wdl"
 
+import "../../../structs/Structs.wdl" as Structs
 
 workflow BarcodeExtractor {
     meta {
@@ -13,7 +13,7 @@ workflow BarcodeExtractor {
     }
 
     call ExtractBarcode {
-        input: input_bam = input_file
+        input: input_bam_path = input_file
     }
 
     output {
@@ -24,21 +24,23 @@ workflow BarcodeExtractor {
 
 task ExtractBarcode {
     input {
-        File input_bam
-        RuntimeAttr? runtime_attr_override
+        String input_bam_path
+        Structs.RuntimeAttr? runtime_attr_override
     }
 
     command <<<
     python3 <<CODE
 import re
-import sys
+import os
 
 def extract_barcode(input_bam):
-    match = re.search(r'\\.hifi_reads\\.(bc\\d+)\\.bam', input_bam)
+    # Extract the filename if the input is a full path
+    filename = os.path.basename(input_bam)
+    match = re.search(r'\\.hifi_reads\\.(bc\\d+)\\.bam', filename)
     return match.group(1) if match else None
 
-input_bam = "${input_bam}"
-barcode = extract_barcode(input_bam)
+input_bam_path = "${input_bam_path}"
+barcode = extract_barcode(input_bam_path)
 if barcode:
     print(barcode)
 else:
@@ -50,8 +52,7 @@ CODE
         String barcode = read_string(stdout())
     }
 
-    #########################
-    RuntimeAttr default_attr = object {
+    Structs.RuntimeAttr default_attr = object {
         cpu_cores:          2,
         mem_gb:             8,
         disk_gb:            50,
@@ -60,7 +61,7 @@ CODE
         max_retries:        0,
         docker:             "us.gcr.io/broad-dsp-lrma/lr-compile-sv-stats:0.0.0"
     }
-    RuntimeAttr runtime_attr = select_first([runtime_attr_override, default_attr])
+    Structs.RuntimeAttr runtime_attr = select_first([runtime_attr_override, default_attr])
 
     runtime {
         cpu:                    select_first([runtime_attr.cpu_cores, default_attr.cpu_cores])

From 526741c5a8076e66e37dbbd46f801b95ad05c3b7 Mon Sep 17 00:00:00 2001
From: Shadi Zaheri <74751641+shadizaheri@users.noreply.github.com>
Date: Wed, 24 Jul 2024 10:51:02 -0400
Subject: [PATCH 3/3] Update BarcodeExtractor.wdl

---
 .../TechAgnostic/Utility/BarcodeExtractor.wdl   | 17 +++++++++++++----
 1 file changed, 13 insertions(+), 4 deletions(-)

diff --git a/wdl/pipelines/TechAgnostic/Utility/BarcodeExtractor.wdl b/wdl/pipelines/TechAgnostic/Utility/BarcodeExtractor.wdl
index bf89b1b30..c25a20b08 100644
--- a/wdl/pipelines/TechAgnostic/Utility/BarcodeExtractor.wdl
+++ b/wdl/pipelines/TechAgnostic/Utility/BarcodeExtractor.wdl
@@ -1,6 +1,14 @@
 version 1.0
 
-import "../../../structs/Structs.wdl" as Structs
+struct RuntimeAttr {
+    Int cpu_cores
+    Int mem_gb
+    Int disk_gb
+    Int boot_disk_gb
+    Int preemptible_tries
+    Int max_retries
+    String docker
+}
 
 workflow BarcodeExtractor {
     meta {
@@ -25,7 +33,7 @@ workflow BarcodeExtractor {
 task ExtractBarcode {
     input {
         String input_bam_path
-        Structs.RuntimeAttr? runtime_attr_override
+        RuntimeAttr? runtime_attr_override
     }
 
     command <<<
@@ -52,7 +60,7 @@ CODE
         String barcode = read_string(stdout())
     }
 
-    Structs.RuntimeAttr default_attr = object {
+    RuntimeAttr default_attr = object {
         cpu_cores:          2,
         mem_gb:             8,
         disk_gb:            50,
@@ -61,7 +69,8 @@ CODE
         max_retries:        0,
         docker:             "us.gcr.io/broad-dsp-lrma/lr-compile-sv-stats:0.0.0"
     }
-    Structs.RuntimeAttr runtime_attr = select_first([runtime_attr_override, default_attr])
+
+    RuntimeAttr runtime_attr = select_first([runtime_attr_override, default_attr])
 
     runtime {
         cpu:                    select_first([runtime_attr.cpu_cores, default_attr.cpu_cores])