From 266848d3385efc9eb62a26759cf559de94d722df Mon Sep 17 00:00:00 2001
From: Villon CHEN <villon.chen@oppida.fr>
Date: Mon, 17 Nov 2025 12:09:45 +0100
Subject: [PATCH 1/4] chore(release): bump project version

This commit is a follow-up to PR #14.

Refs #14.
---
 pyproject.toml | 2 +-
 uv.lock        | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/pyproject.toml b/pyproject.toml
index 168d496..51097ec 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,6 +1,6 @@
 [project]
 name = "CodeSecTools"
-version = "0.12.2"
+version = "0.12.3"
 description = "A framework for code security that provides abstractions for static analysis tools and datasets to support their integration, testing, and evaluation."
 readme = "README.md"
 license = "AGPL-3.0-only"
diff --git a/uv.lock b/uv.lock
index 4eff334..efa70d9 100644
--- a/uv.lock
+++ b/uv.lock
@@ -221,7 +221,7 @@ wheels = [
 
 [[package]]
 name = "codesectools"
-version = "0.12.2"
+version = "0.12.3"
 source = { editable = "." }
 dependencies = [
     { name = "gitpython" },

From f6b4a4478c00ba3004757dd39fbff1a4c774d5d9 Mon Sep 17 00:00:00 2001
From: Villon CHEN <villon.chen@oppida.fr>
Date: Mon, 17 Nov 2025 11:53:15 +0100
Subject: [PATCH 2/4] fix(BenchmarkJava): check if file exist to avoid error

---
 .../datasets/BenchmarkJava/dataset.py         | 21 ++++++++++---------
 1 file changed, 11 insertions(+), 10 deletions(-)

diff --git a/codesectools/datasets/BenchmarkJava/dataset.py b/codesectools/datasets/BenchmarkJava/dataset.py
index ec25095..1dd192a 100644
--- a/codesectools/datasets/BenchmarkJava/dataset.py
+++ b/codesectools/datasets/BenchmarkJava/dataset.py
@@ -148,16 +148,17 @@ def load_dataset(self) -> list[TestCode]:
         for row in reader:
             filename = f"{row[0]}.java"
             filepath = testcode_dir / filename
-            content = filepath.read_text()
-            cwes = [CWEs.from_id(int(row[3]))]
-            has_vuln = True if row[2] == "true" else False
-            files.append(
-                TestCode(
-                    filepath.relative_to(self.directory),
-                    content,
-                    cwes,
-                    has_vuln,
+            if filepath.is_file():
+                content = filepath.read_text()
+                cwes = [CWEs.from_id(int(row[3]))]
+                has_vuln = True if row[2] == "true" else False
+                files.append(
+                    TestCode(
+                        filepath.relative_to(self.directory),
+                        content,
+                        cwes,
+                        has_vuln,
+                    )
                 )
-            )
 
         return files

From f43a316400e4f7a94d467fd6fbcae655b2d612fc Mon Sep 17 00:00:00 2001
From: Villon CHEN <villon.chen@oppida.fr>
Date: Mon, 17 Nov 2025 12:07:02 +0100
Subject: [PATCH 3/4] feat(datasets): load dataset only when necessary to
 reduce blocking time

---
 codesectools/datasets/core/dataset.py | 11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)

diff --git a/codesectools/datasets/core/dataset.py b/codesectools/datasets/core/dataset.py
index 97ed03e..e582023 100644
--- a/codesectools/datasets/core/dataset.py
+++ b/codesectools/datasets/core/dataset.py
@@ -57,12 +57,17 @@ def __init__(self, lang: str | None = None) -> None:
         """
         self.directory = USER_CACHE_DIR / self.name
         self.lang = lang
+        self._files = []
         if self.lang:
             self.full_name = f"{self.name}_{self.lang}"
             assert self.full_name in self.list_dataset_full_names()
-            self.files: list[File] = self.load_dataset()
-        else:
-            self.files = []
+
+    @property
+    def files(self) -> list:
+        """Get the list of dataset files, loading them if necessary."""
+        if self.lang:
+            self._files = self.load_dataset()
+        return self._files
 
     @classmethod
     def is_cached(cls) -> bool:

From 29689652d44b672e5825925f68149fd1a3c6f6e3 Mon Sep 17 00:00:00 2001
From: Villon CHEN <villon.chen@oppida.fr>
Date: Mon, 17 Nov 2025 12:12:03 +0100
Subject: [PATCH 4/4] chore(release): bump project version

---
 pyproject.toml | 2 +-
 uv.lock        | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/pyproject.toml b/pyproject.toml
index 51097ec..a997f14 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,6 +1,6 @@
 [project]
 name = "CodeSecTools"
-version = "0.12.3"
+version = "0.12.4"
 description = "A framework for code security that provides abstractions for static analysis tools and datasets to support their integration, testing, and evaluation."
 readme = "README.md"
 license = "AGPL-3.0-only"
diff --git a/uv.lock b/uv.lock
index efa70d9..de3b98f 100644
--- a/uv.lock
+++ b/uv.lock
@@ -221,7 +221,7 @@ wheels = [
 
 [[package]]
 name = "codesectools"
-version = "0.12.3"
+version = "0.12.4"
 source = { editable = "." }
 dependencies = [
     { name = "gitpython" },