From 080b851aa6fe72f05e872ac10222eebb6edf65a6 Mon Sep 17 00:00:00 2001
From: chrispiros <valandil285@gmail.com>
Date: Thu, 16 Oct 2025 15:35:53 +0200
Subject: [PATCH 1/6] Update functions.py

---
 emm/preprocessing/functions.py | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/emm/preprocessing/functions.py b/emm/preprocessing/functions.py
index b88c1c8..b81b8ce 100644
--- a/emm/preprocessing/functions.py
+++ b/emm/preprocessing/functions.py
@@ -23,7 +23,11 @@
 from typing import Any, Callable
 
 import cleanco
-from unidecode import unidecode
+
+try:
+    from unidecode import unidecode
+except ImportError:
+    unidecode = None
 
 from emm.preprocessing.abbreviation_util import abbreviations_to_words, legal_abbreviations_to_words
 
@@ -50,7 +54,7 @@ def map_shorthands(name):
 
     return {
         # Replace accented characters by their normalized representation, e.g. replace 'ä' with 'A\xa4'
-        "strip_accents_unicode": F.run_custom_function(unidecode),
+        "strip_accents_unicode": F.run_custom_function(unidecode if unidecode is not None else (lambda x: x)),
         # Replace all dash and underscore characters with a space characters
         "strip_hyphens": F.regex_replace(r"""[-_]""", " ", simple=True),
         # Replace all punctuation characters (e.g. '.', '-', '_', ''', ';') with spaces

From f44f64b896c5389e0b664d19285a8ffcbbef9d1d Mon Sep 17 00:00:00 2001
From: chrispiros <valandil285@gmail.com>
Date: Thu, 16 Oct 2025 15:36:29 +0200
Subject: [PATCH 2/6] Update pyproject.toml

---
 pyproject.toml | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/pyproject.toml b/pyproject.toml
index 4ee6d89..edcf4da 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -31,8 +31,7 @@ dependencies = [
   "sparse-dot-topn>=1.1.1",
   "joblib",
   "pyarrow>=6.0.1", # seems to work with spark 3.1.2 - 3.3.1
-  "requests",
-  "unidecode"
+  "requests"
 ]
 dynamic = ["version"]
 
@@ -51,7 +50,8 @@ dev = [
   "matplotlib",
   "pygments",
   "pandoc",
-  "pympler"
+  "pympler",
+  "unidecode"
 ]
 test = [
   "pytest",

From bdc1089800f87ee2eb26ab95f07fce761be02ed5 Mon Sep 17 00:00:00 2001
From: chrispiros <valandil285@gmail.com>
Date: Mon, 3 Nov 2025 15:58:14 +0100
Subject: [PATCH 3/6] added preprocessing group in pyproject.toml

---
 pyproject.toml | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/pyproject.toml b/pyproject.toml
index edcf4da..57fb37a 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -50,7 +50,9 @@ dev = [
   "matplotlib",
   "pygments",
   "pandoc",
-  "pympler",
+  "pympler"
+]
+preprocessing = [
   "unidecode"
 ]
 test = [

From f8b4567302d8f77716ef9d9ce8aa7d2ff3161a10 Mon Sep 17 00:00:00 2001
From: chrispiros <valandil285@gmail.com>
Date: Mon, 3 Nov 2025 16:03:22 +0100
Subject: [PATCH 4/6] added warning in functions.py

---
 emm/preprocessing/functions.py | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/emm/preprocessing/functions.py b/emm/preprocessing/functions.py
index b81b8ce..1d2163a 100644
--- a/emm/preprocessing/functions.py
+++ b/emm/preprocessing/functions.py
@@ -23,11 +23,17 @@
 from typing import Any, Callable
 
 import cleanco
+import warnings
 
 try:
     from unidecode import unidecode
 except ImportError:
     unidecode = None
+    warnings.warn(
+        "The 'unidecode' module is not installed. 'strip_accents_unicode' will default to an identity function. "
+        "Install 'unidecode' to enable accent stripping functionality.",
+        ImportWarning
+    )
 
 from emm.preprocessing.abbreviation_util import abbreviations_to_words, legal_abbreviations_to_words
 

From 348ef3a0a3685301d1baad2498ed1de30b3a45c6 Mon Sep 17 00:00:00 2001
From: chrispiros <valandil285@gmail.com>
Date: Mon, 24 Nov 2025 14:46:23 +0100
Subject: [PATCH 5/6] pre-commit applied

---
 emm/preprocessing/functions.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/emm/preprocessing/functions.py b/emm/preprocessing/functions.py
index 1d2163a..ed8cf91 100644
--- a/emm/preprocessing/functions.py
+++ b/emm/preprocessing/functions.py
@@ -19,11 +19,11 @@
 
 from __future__ import annotations
 
+import warnings
 from functools import partial
 from typing import Any, Callable
 
 import cleanco
-import warnings
 
 try:
     from unidecode import unidecode
@@ -32,7 +32,7 @@
     warnings.warn(
         "The 'unidecode' module is not installed. 'strip_accents_unicode' will default to an identity function. "
         "Install 'unidecode' to enable accent stripping functionality.",
-        ImportWarning
+        ImportWarning,
     )
 
 from emm.preprocessing.abbreviation_util import abbreviations_to_words, legal_abbreviations_to_words

From 966a09c54c14449838bde9ff32ab14300d31441c Mon Sep 17 00:00:00 2001
From: chrispiros <valandil285@gmail.com>
Date: Tue, 25 Nov 2025 14:25:44 +0100
Subject: [PATCH 6/6] added 'unidecode' to test deps

---
 pyproject.toml | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/pyproject.toml b/pyproject.toml
index 57fb37a..89a8caa 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -58,7 +58,8 @@ preprocessing = [
 test = [
   "pytest",
   "pytest-ordering",
-  "virtualenv"
+  "virtualenv",
+  "unidecode"
 ]
 test-cov = [
   "coverage",