From 3a3c9bd6141c64b9b7f4c31a7ab92eae68fd9064 Mon Sep 17 00:00:00 2001
From: lvreynoso <lreynoso@chanzuckerberg.com>
Date: Wed, 8 May 2024 14:22:05 -0700
Subject: [PATCH] Allow multiple arguments to diamond blastx

---
 lib/idseq_utils/idseq_utils/diamond_scatter.py | 10 +++++++++-
 workflows/diamond/Dockerfile                   |  5 ++---
 workflows/diamond/diamond.wdl                  |  4 ++--
 3 files changed, 13 insertions(+), 6 deletions(-)

diff --git a/lib/idseq_utils/idseq_utils/diamond_scatter.py b/lib/idseq_utils/idseq_utils/diamond_scatter.py
index 385301c1d..b3b763f8b 100644
--- a/lib/idseq_utils/idseq_utils/diamond_scatter.py
+++ b/lib/idseq_utils/idseq_utils/diamond_scatter.py
@@ -3,6 +3,7 @@
 """
 
 import os
+import shlex
 import shutil
 import sys
 import errno
@@ -71,8 +72,15 @@ def diamond_blastx(
         database,
         "--out",
         out,
-        f"--{diamond_args}",
     ]
+
+    # backwards compatibility for function calls that expect this function
+    # to automatically append "--" to diamond_args
+    if diamond_args == "long-reads" or diamond_args == "mid-sensitive":
+        diamond_args = "--" + diamond_args
+
+    cmd.extend(shlex.split(diamond_args))
+
     for query in queries:
         cmd += ["--query", query]
     if chunk:
diff --git a/workflows/diamond/Dockerfile b/workflows/diamond/Dockerfile
index 806ef944a..a3d52e129 100644
--- a/workflows/diamond/Dockerfile
+++ b/workflows/diamond/Dockerfile
@@ -19,7 +19,6 @@ WORKDIR /tmp
 RUN git clone https://github.com/chanzuckerberg/czid-workflows
 WORKDIR /tmp/czid-workflows
 RUN pip3 install -r requirements-dev.txt
-RUN git checkout rlim-add-diamond-modification
-RUN cp short-read-mngs/idseq_utils/idseq_utils/diamond_scatter.py /usr/local/bin/
+COPY --from=lib idseq_utils/idseq_utils/diamond_scatter.py /usr/local/bin/
 
-WORKDIR /workdir
\ No newline at end of file
+WORKDIR /workdir
diff --git a/workflows/diamond/diamond.wdl b/workflows/diamond/diamond.wdl
index 35af2d033..7a8f2f814 100644
--- a/workflows/diamond/diamond.wdl
+++ b/workflows/diamond/diamond.wdl
@@ -34,7 +34,7 @@ task RunDiamond {
     }
 
     command <<<
-        python3 /usr/local/bin/diamond_scatter.py blastx-chunk --db ~{db_chunk} --query ~{query_0} ~{if defined(query_1) then '--query ~{query_1}' else ''} --out-dir chunks --diamond-args "~{extra_args}"
+        python3 /usr/local/bin/diamond_scatter.py blastx-chunk --db ~{db_chunk} --query ~{query_0} ~{if defined(query_1) then '--query ~{query_1}' else ''} --out-dir chunks --diamond-args="~{extra_args}"
     >>>
 
     output {
@@ -45,4 +45,4 @@ task RunDiamond {
     runtime {
         docker: docker_image_id
     }
-}
\ No newline at end of file
+}