From 3d8ae967ff3a5725832f2630afb314ff981c62aa Mon Sep 17 00:00:00 2001
From: ryanjameskennedy <ryanjameskennedy@icloud.com>
Date: Wed, 16 Oct 2024 11:46:44 +0200
Subject: [PATCH 1/8] Update to v0.11.0

---
 CHANGELOG.md       | 8 ++++++++
 prp/__version__.py | 2 +-
 2 files changed, 9 insertions(+), 1 deletion(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index b398c8a..094873a 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,5 +1,13 @@
 ## [Unreleased]
 
+### Added
+
+### Fixed
+
+### Changed
+
+## [0.11.0]
+
 ### Added
 
  - Added emmtyper and parser
diff --git a/prp/__version__.py b/prp/__version__.py
index 5351733..721fd46 100644
--- a/prp/__version__.py
+++ b/prp/__version__.py
@@ -1,2 +1,2 @@
 """PRP version"""
-VERSION = "0.10.1"
+VERSION = "0.11.0"

From 161dd2d8d44bf5473301cf1a454ab159749c7fd7 Mon Sep 17 00:00:00 2001
From: ryanjameskennedy <ryanjameskennedy@icloud.com>
Date: Mon, 11 Nov 2024 16:40:54 +0100
Subject: [PATCH 2/8] Handle alt types for emmtyper

---
 prp/models/typing.py            | 6 +++---
 prp/parse/phenotype/emmtyper.py | 7 ++++---
 2 files changed, 7 insertions(+), 6 deletions(-)

diff --git a/prp/models/typing.py b/prp/models/typing.py
index 4d04e00..4307ee8 100644
--- a/prp/models/typing.py
+++ b/prp/models/typing.py
@@ -103,9 +103,9 @@ class TypingResultEmm(RWModel):
     """Container for emmtype gene information"""
 
     cluster_count: int
-    emmtype: str
-    emm_like_alleles: list[str]
-    emm_cluster: str
+    emmtype: str | None = None
+    emm_like_alleles: list[str] | None = None
+    emm_cluster: str | None = None
 
 
 class EmmTypingMethodIndex(RWModel):
diff --git a/prp/parse/phenotype/emmtyper.py b/prp/parse/phenotype/emmtyper.py
index 0e4049a..667948b 100644
--- a/prp/parse/phenotype/emmtyper.py
+++ b/prp/parse/phenotype/emmtyper.py
@@ -2,6 +2,7 @@
 
 import logging
 import pandas as pd
+import numpy as np
 
 from typing import Any
 
@@ -16,6 +17,7 @@ def parse_emmtyper_pred(path: str) -> EmmTypingMethodIndex:
     pred_result = []
     df = pd.read_csv(path, sep='\t', header=None)
     df.columns = ["sample_name", "cluster_count", "emmtype", "emm_like_alleles", "emm_cluster"]
+    df.replace(["-", ""], None, inplace=True)
     df_loa = df.to_dict(orient="records")
     for emmtype_array in df_loa:
         emmtype_results = _parse_emmtyper_results(emmtype_array)
@@ -31,10 +33,9 @@ def parse_emmtyper_pred(path: str) -> EmmTypingMethodIndex:
 
 def _parse_emmtyper_results(info: dict[str, Any]) -> TypingResultEmm:
     """Parse emm gene prediction results."""
-    emm_like_alleles = info["emm_like_alleles"].split(";")
+    emm_like_alleles = info["emm_like_alleles"].split(";") if not pd.isna(info["emm_like_alleles"]) else None
     return TypingResultEmm(
-        # info
-        cluster_count=info["cluster_count"],
+        cluster_count=int(info["cluster_count"]),
         emmtype=info["emmtype"],
         emm_like_alleles=emm_like_alleles,
         emm_cluster=info["emm_cluster"],

From 69326bc2de76d8a7fd2b821632e9b7e7cf904800 Mon Sep 17 00:00:00 2001
From: ryanjameskennedy <ryanjameskennedy@icloud.com>
Date: Mon, 11 Nov 2024 16:45:13 +0100
Subject: [PATCH 3/8] Handle emmtyper alt types

---
 CHANGELOG.md | 1 +
 1 file changed, 1 insertion(+)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 094873a..dea76e4 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -3,6 +3,7 @@
 ### Added
 
 ### Fixed
+ - Handle alt types for emmtyper
 
 ### Changed
 

From 1a2f147ef43cbc6a1d1a5fe3e63c616ef785bfc2 Mon Sep 17 00:00:00 2001
From: ryanjameskennedy <ryanjameskennedy@icloud.com>
Date: Mon, 11 Nov 2024 16:47:15 +0100
Subject: [PATCH 4/8] Rm numpy import

---
 prp/parse/phenotype/emmtyper.py | 1 -
 1 file changed, 1 deletion(-)

diff --git a/prp/parse/phenotype/emmtyper.py b/prp/parse/phenotype/emmtyper.py
index 667948b..65eefa3 100644
--- a/prp/parse/phenotype/emmtyper.py
+++ b/prp/parse/phenotype/emmtyper.py
@@ -2,7 +2,6 @@
 
 import logging
 import pandas as pd
-import numpy as np
 
 from typing import Any
 

From 253ad8dd1b797db0653ed41fcdf62e9fe30cae62 Mon Sep 17 00:00:00 2001
From: ryanjameskennedy <ryanjameskennedy@icloud.com>
Date: Mon, 11 Nov 2024 16:59:47 +0100
Subject: [PATCH 5/8] Update to v0.11.1

---
 CHANGELOG.md       | 9 +++++++++
 prp/__version__.py | 2 +-
 2 files changed, 10 insertions(+), 1 deletion(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index dea76e4..2c35279 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -3,6 +3,15 @@
 ### Added
 
 ### Fixed
+
+### Changed
+
+## [0.11.1]
+
+### Added
+
+### Fixed
+
  - Handle alt types for emmtyper
 
 ### Changed
diff --git a/prp/__version__.py b/prp/__version__.py
index 721fd46..697e539 100644
--- a/prp/__version__.py
+++ b/prp/__version__.py
@@ -1,2 +1,2 @@
 """PRP version"""
-VERSION = "0.11.0"
+VERSION = "0.11.1"

From b9a646262ec376c70d90f1e9609eaf0c907e9ab3 Mon Sep 17 00:00:00 2001
From: Markus Johansson <markus.h.johansson@skane.se>
Date: Tue, 19 Nov 2024 14:56:34 +0100
Subject: [PATCH 6/8] Added mypy as dev dependency

---
 CHANGELOG.md   | 2 ++
 pyproject.toml | 1 +
 2 files changed, 3 insertions(+)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 2c35279..8407c5d 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -6,6 +6,8 @@
 
 ### Changed
 
+- Added mypy as test dependency
+
 ## [0.11.1]
 
 ### Added
diff --git a/pyproject.toml b/pyproject.toml
index 61aff58..4dbbef4 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -46,6 +46,7 @@ dev = [
 ]
 test = [
     "pytest-cov ~=4.1.0",
+    "mypy == 1.13.0"
 ]
 
 [build-system]

From 301a5f065c2912c00a973a96aa07298b45ba5564 Mon Sep 17 00:00:00 2001
From: Markus Johansson <markus.h.johansson@skane.se>
Date: Tue, 19 Nov 2024 15:47:19 +0100
Subject: [PATCH 7/8] Fixed crash caused by mlst results with no calls

---
 prp/parse/typing.py             |  5 ++++-
 tests/fixtures/__init__.py      |  5 +++++
 tests/fixtures/mlst.nocall.json |  9 +++++++++
 tests/parse/test_typing.py      | 28 ++++++++++++++++++++++++++--
 4 files changed, 44 insertions(+), 3 deletions(-)
 create mode 100644 tests/fixtures/mlst.nocall.json

diff --git a/prp/parse/typing.py b/prp/parse/typing.py
index 6b4cec8..5512b8a 100644
--- a/prp/parse/typing.py
+++ b/prp/parse/typing.py
@@ -43,6 +43,9 @@ def parse_mlst_results(mlst_fpath: str) -> TypingResultMlst:
     LOG.info("Parsing mlst results")
     with open(mlst_fpath, "r", encoding="utf-8") as jsonfile:
         result = json.load(jsonfile)[0]
+        # get raw allele info
+        alleles = {} if result.get("alleles") is None else result["alleles"]
+        # create typing result object
         result_obj = TypingResultMlst(
             scheme=result["scheme"],
             sequence_type=(
@@ -50,7 +53,7 @@ def parse_mlst_results(mlst_fpath: str) -> TypingResultMlst:
             ),
             alleles={
                 gene: _process_allele_call(allele)
-                for gene, allele in result["alleles"].items()
+                for gene, allele in alleles.items()
             },
         )
     return MethodIndex(
diff --git a/tests/fixtures/__init__.py b/tests/fixtures/__init__.py
index 214fa3e..527fa3c 100644
--- a/tests/fixtures/__init__.py
+++ b/tests/fixtures/__init__.py
@@ -5,3 +5,8 @@
 from .saureus import *
 from .shigella import *
 from .streptococcus import *
+
+@pytest.fixture()
+def mlst_result_path_no_call(data_path):
+    """Get path for mlst file where alleles was not called."""
+    return str(data_path.joinpath("mlst.nocall.json"))
\ No newline at end of file
diff --git a/tests/fixtures/mlst.nocall.json b/tests/fixtures/mlst.nocall.json
new file mode 100644
index 0000000..8aa1b81
--- /dev/null
+++ b/tests/fixtures/mlst.nocall.json
@@ -0,0 +1,9 @@
+[
+   {
+      "sequence_type" : "-",
+      "scheme" : "-",
+      "alleles" : null,
+      "id" : "unique_file_id",
+      "filename" : "assembly_file_name.fasta"
+   }
+]
\ No newline at end of file
diff --git a/tests/parse/test_typing.py b/tests/parse/test_typing.py
index 74000b9..c2015b8 100644
--- a/tests/parse/test_typing.py
+++ b/tests/parse/test_typing.py
@@ -2,7 +2,7 @@
 
 import pytest
 import logging
-from prp.parse.typing import replace_cgmlst_errors
+from prp.parse.typing import replace_cgmlst_errors, parse_mlst_results
 from prp.models.typing import ChewbbacaErrors
 
 # build test cases for handeling chewbacca allele caller errors and annotations
@@ -84,4 +84,28 @@ def test_replace_cgmlst_errors_warnings(caplog):
     # run test that a warning was triggered if input is unknown string
     allele = "A_STRANGE_STRING"
     replace_cgmlst_errors(allele, include_novel_alleles=True, correct_alleles=True)
-    assert allele in caplog.text
\ No newline at end of file
+    assert allele in caplog.text
+
+
+def test_parse_mlst_result(ecoli_mlst_path):
+    """Test parsing of MLST result file."""
+    # FIRST run result parser
+    res_obj = parse_mlst_results(ecoli_mlst_path)
+
+    # THEN verify result type
+    assert res_obj.type == 'mlst'
+    # THEN verify software
+    assert res_obj.software == 'mlst'
+    # THEN verify sequence type and allele assignment
+    assert res_obj.result.sequence_type == 58
+    assert len(res_obj.result.alleles) == 8
+
+
+def test_parse_mlst_result_w_no_call(mlst_result_path_no_call):
+    """Test parsing of MLST results file where the alleles was not called."""
+    # FIRST run result parser
+    res_obj = parse_mlst_results(mlst_result_path_no_call)
+
+    # THEN verify that sequence type is None
+    assert res_obj.result.sequence_type is None
+    assert len(res_obj.result.alleles) == 0
\ No newline at end of file

From dbf5795792dd5744693fc4d3f6ef4bedbf0ce90a Mon Sep 17 00:00:00 2001
From: Markus Johansson <markus.h.johansson@skane.se>
Date: Tue, 19 Nov 2024 15:47:47 +0100
Subject: [PATCH 8/8] Updated CHANGELOG

---
 CHANGELOG.md | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 8407c5d..894d9a1 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -4,6 +4,8 @@
 
 ### Fixed
 
+- Fixed issue where mlst results with no calls crashed PRP.
+
 ### Changed
 
 - Added mypy as test dependency