From 4072e1380b1ac45760cb087c939c75d6c53a4cec Mon Sep 17 00:00:00 2001
From: xvazquezc <xvazquezc@users.noreply.github.com>
Date: Tue, 25 Aug 2026 10:40:57 +1000
Subject: [PATCH] fix: cds IDs to be unique

---
 gfftk/gff.py           |  4 +++-
 tests/unit/test_gff.py | 40 ++++++++++++++++++++++++++++++++++++++--
 2 files changed, 41 insertions(+), 3 deletions(-)

diff --git a/gfftk/gff.py b/gfftk/gff.py
index 301d2a3..d58449c 100755
--- a/gfftk/gff.py
+++ b/gfftk/gff.py
@@ -3414,8 +3414,9 @@ def _sortDict(d):
                 # GFF3 phase is 1 less than flat file
                 current_phase = v["codon_start"][i] - 1
                 for y in range(0, num_cds):
+                    cds_num = y + 1
                     gffout.write(
-                        "{:}\t{:}\tCDS\t{:}\t{:}\t.\t{:}\t{:}\tID={:}.cds;Parent={:};\n".format(
+                        "{:}\t{:}\tCDS\t{:}\t{:}\t.\t{:}\t{:}\tID={:}.cds{:};Parent={:};\n".format(
                             v["contig"],
                             new_source,
                             sortedCDS[y][0],
@@ -3423,6 +3424,7 @@ def _sortDict(d):
                             v["strand"],
                             current_phase,
                             v["ids"][i],
+                            cds_num,
                             v["ids"][i],
                         )
                     )
diff --git a/tests/unit/test_gff.py b/tests/unit/test_gff.py
index ef5a0b9..42d7b27 100644
--- a/tests/unit/test_gff.py
+++ b/tests/unit/test_gff.py
@@ -7,6 +7,7 @@
 
 from gfftk.gff import (
     _detect_format,
+    dict2gff3,
     gff2dict,
     is_combined_gff_fasta,
     simplifyGO,
@@ -19,6 +20,42 @@
 class TestGFFParsing:
     """Tests for GFF parsing functions."""
 
+    def test_dict2gff3_assigns_unique_cds_ids(self):
+        """Each CDS segment emitted for a transcript must receive its own ID."""
+        with tempfile.NamedTemporaryFile(mode="w", delete=False) as gff_temp:
+            gff_temp.write("##gff-version 3\n")
+            gff_temp.write("contig1\ttest\tgene\t1\t12\t.\t+\t.\tID=gene1\n")
+            gff_temp.write(
+                "contig1\ttest\tmRNA\t1\t12\t.\t+\t.\tID=mRNA1;Parent=gene1\n"
+            )
+            gff_temp.write("contig1\ttest\texon\t1\t3\t.\t+\t.\tID=exon1;Parent=mRNA1\n")
+            gff_temp.write("contig1\ttest\texon\t7\t12\t.\t+\t.\tID=exon2;Parent=mRNA1\n")
+            gff_temp.write("contig1\ttest\tCDS\t1\t3\t.\t+\t0\tID=cds1;Parent=mRNA1\n")
+            gff_temp.write("contig1\ttest\tCDS\t7\t12\t.\t+\t0\tID=cds2;Parent=mRNA1\n")
+            gff_name = gff_temp.name
+
+        with tempfile.NamedTemporaryFile(mode="w", delete=False) as fasta_temp:
+            fasta_temp.write(">contig1\nATGAAATTTTAA\n")
+            fasta_name = fasta_temp.name
+
+        with tempfile.NamedTemporaryFile(mode="w", delete=False) as output_temp:
+            output_name = output_temp.name
+
+        try:
+            annotations = gff2dict(gff_name, fasta_name, debug=False)
+            dict2gff3(annotations, output=output_name)
+            with open(output_name) as gff_output:
+                cds_ids = [
+                    line.split("ID=", 1)[1].split(";", 1)[0]
+                    for line in gff_output
+                    if "\tCDS\t" in line
+                ]
+            assert cds_ids == ["mRNA1.cds1", "mRNA1.cds2"]
+        finally:
+            os.unlink(gff_name)
+            os.unlink(fasta_name)
+            os.unlink(output_name)
+
     def test_detect_format(self):
         """Test the _detect_format function."""
         # Create a GFF3-like file for testing
@@ -402,6 +439,7 @@ def test_longest_orf_gapmm2_alignment(self):
             if os.path.exists(gff_file):
                 os.remove(gff_file)
 
+
     def test_longest_orf_exact_boundary(self):
         """Test _longest_orf correctly processes an ORF that exactly ends on an exon boundary."""
         # Create a fasta file
@@ -452,5 +490,3 @@ def test_longest_orf_exact_boundary(self):
                 os.remove(fasta_file)
             if os.path.exists(gff_file):
                 os.remove(gff_file)
-
-