From b975e0cfd1d0adc5f51e66292d29d4651d3f053f Mon Sep 17 00:00:00 2001
From: Muriithi Frederick Muriuki
Date: Wed, 1 Sep 2021 07:35:40 +0300
Subject: Parse the genotype file's data header

* gn3/db/genotypes.py: parse data header
* tests/unit/db/test_genotypes.py: check that header's parse works correctly.

  Add tests to check that the parser works as expected. Add code to implement
  the parsing and pass the tests.
---
 gn3/db/genotypes.py             | 19 +++++++++++++++++++
 tests/unit/db/test_genotypes.py | 22 +++++++++++++++++++++-
 2 files changed, 40 insertions(+), 1 deletion(-)

diff --git a/gn3/db/genotypes.py b/gn3/db/genotypes.py
index 2be3e1a..be0dfc2 100644
--- a/gn3/db/genotypes.py
+++ b/gn3/db/genotypes.py
@@ -87,3 +87,22 @@ def parse_genotype_labels(lines: list):
     return tuple(
         item for item in (__parse_label(line) for line in lines)
         if item is not None)
+
+def parse_genotype_header(line: str, parlist = tuple()):
+    """
+    Parse the genotype file header line
+
+    DESCRIPTION:
+    Reworks
+    https://github.com/genenetwork/genenetwork1/blob/master/web/webqtl/utility/gen_geno_ob.py#L94-L114
+    """
+    items = [item.strip() for item in line.split("\t")]
+    Mbmap = "Mb" in items
+    prgy = ((parlist + tuple(items[4:])) if Mbmap
+            else (parlist + tuple(items[3:])))
+    return (
+        ("Mbmap", Mbmap),
+        ("cm_column", items.index("cM")),
+        ("mb_column", None if not Mbmap else items.index("Mb")),
+        ("prgy", prgy),
+        ("nprgy", len(prgy)))
diff --git a/tests/unit/db/test_genotypes.py b/tests/unit/db/test_genotypes.py
index 0264764..4fa8a53 100644
--- a/tests/unit/db/test_genotypes.py
+++ b/tests/unit/db/test_genotypes.py
@@ -1,6 +1,6 @@
 """Tests gn3.db.genotypes"""
 from unittest import TestCase
-from gn3.db.genotypes import parse_genotype_labels
+from gn3.db.genotypes import parse_genotype_labels, parse_genotype_header
 
 class TestGenotypes(TestCase):
     """Tests for functions in `gn3.db.genotypes`."""
@@ -15,3 +15,23 @@ class TestGenotypes(TestCase):
         (("group", "test_group"), ("filler", "test_filler"),
          ("type", "test_type"), ("mat", "test_mat"), ("pat", "test_pat"),
          ("het", "test_het"), ("unk", "test_unk")))
+
+    def test_parse_genotype_header(self):
+        for header, expected in [
+                [("Chr\tLocus\tcM\tMb\tBXD1\tBXD2\tBXD5\tBXD6\tBXD8\tBXD9\t"
+                  "BXD11\tBXD12\tBXD13\tBXD14\tBXD15\tBXD16\tBXD18\tBXD19"),
+                 (("Mbmap", True), ("cm_column", 2), ("mb_column", 3),
+                  ("prgy",
+                   ("BXD1", "BXD2", "BXD5", "BXD6", "BXD8", "BXD9", "BXD11",
+                    "BXD12", "BXD13", "BXD14", "BXD15", "BXD16", "BXD18",
+                    "BXD19")),
+                  ("nprgy", 14))],
+                [("Chr\tLocus\tcM\tBXD1\tBXD2\tBXD5\tBXD6\tBXD8\tBXD9\tBXD11"
+                  "\tBXD12\tBXD13\tBXD14\tBXD15\tBXD16\tBXD18"),
+                 (("Mbmap", False), ("cm_column", 2), ("mb_column", None),
+                  ("prgy",
+                   ("BXD1", "BXD2", "BXD5", "BXD6", "BXD8", "BXD9", "BXD11",
+                    "BXD12", "BXD13", "BXD14", "BXD15", "BXD16", "BXD18")),
+                  ("nprgy", 13))]]:
+            with self.subTest(header=header):
+                self.assertEqual(parse_genotype_header(header), expected)
-- 
cgit v1.2.3