From f6567eae233a98a999423966b60e3983bc068513 Mon Sep 17 00:00:00 2001
From: Christian Diener <ch.diener@gmail.com>
Date: Wed, 4 Sep 2024 20:48:21 +0200
Subject: [PATCH] Support CARVEME and GAPSEQ external IDs (#1404)

* add more external ID patterns

* add RNs

* add tests

* fix isort

* yes isort, I will add some random spaces here, makes perfect sense

* of course black, please join the party

* you drunk black, go home
---
 release-notes/next-release.md            |  3 +++
 src/cobra/medium/annotations.py          |  6 +++++-
 src/cobra/medium/boundary_types.py       |  2 +-
 tests/test_medium/test_boundary_types.py | 15 +++++++++++++++
 4 files changed, 24 insertions(+), 2 deletions(-)

diff --git a/release-notes/next-release.md b/release-notes/next-release.md
index 0cca0da86..c75df3de0 100644
--- a/release-notes/next-release.md
+++ b/release-notes/next-release.md
@@ -9,7 +9,10 @@
 * Fixed a bug with SBML group parsing that affects the Debian package.
 
 ## Other
+
 * Adding a duplicate boundary reaction (with `add_boundary`) no longer errors, but instead just returns the existing reaction.
+* Automatic detection of the external comprtment now recognizes CARVEME and GAPSEQ ids and will no longer
+  show a warning for models generated with those tools.
 
 ## Deprecated features
 
diff --git a/src/cobra/medium/annotations.py b/src/cobra/medium/annotations.py
index bda673e75..461e49089 100644
--- a/src/cobra/medium/annotations.py
+++ b/src/cobra/medium/annotations.py
@@ -55,6 +55,8 @@
         "intracellular",
         "intracellular region",
         "intracellular space",
+        "c0",  # GAPSEQ
+        "C_c",  # CARVEME
     ],
     "er": ["endoplasmic reticulum"],
     "erm": ["endoplasmic reticulum membrane"],
@@ -68,6 +70,8 @@
         "extra-organism",
         "external",
         "external medium",
+        "e0",  # GAPSEQ
+        "C_e",  # CARVEME
     ],
     "f": ["flagellum", "bacterial-type flagellum"],
     "g": ["golgi", "golgi apparatus"],
@@ -78,7 +82,7 @@
     "mm": ["mitochondrial membrane"],
     "m": ["mitochondrion", "mitochondria"],
     "n": ["nucleus"],
-    "p": ["periplasm", "periplasmic space"],
+    "p": ["periplasm", "periplasmic space", "p0", "C_p"],  # GAPSEQ  # CARVEME
     "x": ["peroxisome", "glyoxysome"],
     "u": ["thylakoid"],
     "vm": ["vacuolar membrane"],
diff --git a/src/cobra/medium/boundary_types.py b/src/cobra/medium/boundary_types.py
index 2a0338f9d..971480f57 100644
--- a/src/cobra/medium/boundary_types.py
+++ b/src/cobra/medium/boundary_types.py
@@ -94,7 +94,7 @@ def find_external_compartment(model: "Model") -> str:
             "Consider renaming your compartments using "
             "`Model.compartments` to fix this."
         )
-        return most[0]
+        return most.iloc[0]
 
     # No info in the model, so give up
     raise RuntimeError(
diff --git a/tests/test_medium/test_boundary_types.py b/tests/test_medium/test_boundary_types.py
index 2299d5165..a459ded44 100644
--- a/tests/test_medium/test_boundary_types.py
+++ b/tests/test_medium/test_boundary_types.py
@@ -1,5 +1,7 @@
 """Test functionalities of boundary type detection functions."""
 
+import logging
+
 import pytest
 
 from cobra.core import Metabolite, Model, Reaction
@@ -40,6 +42,19 @@ def test_find_external_compartment_multi(model: Model) -> None:
         find_external_compartment(model)
 
 
+@pytest.mark.parametrize("compartment", ["C_e", "e0"])
+def test_find_external_popular_reconstructions(
+    model: Model, compartment, caplog
+) -> None:
+    """Test some additional id formats."""
+    for ex in model.exchanges:
+        ex.reactants[0].compartment = compartment
+    with caplog.at_level(logging.WARNING):
+        external = find_external_compartment(model)
+    assert external == compartment
+    assert "complete nonsense" not in caplog.text
+
+
 def test_no_names_or_boundary_reactions(empty_model: Model) -> None:
     """Test absence of name or boundary reactions."""
     with pytest.raises(RuntimeError):