move data to package, be more explicit about package_data

33 files changed, 6 insertions, 1 deletions

diff --git a/pyproject.toml b/pyproject.toml
index e08ac75f..e83a4a6b 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -116,7 +116,12 @@
     max-complexity = 10
 
   [tool.setuptools.package-data]
-    "*" = ["*.yaml", "*.json"]
+    'sec_certs' = ["rules.yaml"]
+    'sec_certs.config' = ["settings.yaml", "settings-schema.json"]
+    'sec_certs.data' = [
+      "reference_annotations/split/*.json",
+      "reference_annotations/manual_annotations/*.csv",
+    ]
 
 
 [tool.setuptools_scm]
diff --git a/data/cert_ids/duplicate_ids.csv b/src/sec_certs/data/cert_ids/duplicate_ids.csv
index b24bc4f3..b24bc4f3 100644
--- a/data/cert_ids/duplicate_ids.csv
+++ b/src/sec_certs/data/cert_ids/duplicate_ids.csv
diff --git a/data/cert_ids/missing_ids.csv b/src/sec_certs/data/cert_ids/missing_ids.csv
index 163bc6dc..163bc6dc 100644
--- a/data/cert_ids/missing_ids.csv
+++ b/src/sec_certs/data/cert_ids/missing_ids.csv
diff --git a/data/cert_ids/random.csv b/src/sec_certs/data/cert_ids/random.csv
index b1b3885a..b1b3885a 100644
--- a/data/cert_ids/random.csv
+++ b/src/sec_certs/data/cert_ids/random.csv
diff --git a/data/cert_ids/readme.md b/src/sec_certs/data/cert_ids/readme.md
index 75c38a48..75c38a48 100644
--- a/data/cert_ids/readme.md
+++ b/src/sec_certs/data/cert_ids/readme.md
diff --git a/data/cert_ids/truth.csv b/src/sec_certs/data/cert_ids/truth.csv
index 80086fc7..80086fc7 100644
--- a/data/cert_ids/truth.csv
+++ b/src/sec_certs/data/cert_ids/truth.csv
diff --git a/data/cpes/label_studio_interface.txt b/src/sec_certs/data/cpes/label_studio_interface.txt
index 3b2a7e20..3b2a7e20 100644
--- a/data/cpes/label_studio_interface.txt
+++ b/src/sec_certs/data/cpes/label_studio_interface.txt
diff --git a/data/cpes/manual_cpe_labels.json b/src/sec_certs/data/cpes/manual_cpe_labels.json
index bd5e1ba6..bd5e1ba6 100644
--- a/data/cpes/manual_cpe_labels.json
+++ b/src/sec_certs/data/cpes/manual_cpe_labels.json
diff --git a/data/cpes/outdated/cc.json b/src/sec_certs/data/cpes/outdated/cc.json
index 8ad9aaf9..8ad9aaf9 100644
--- a/data/cpes/outdated/cc.json
+++ b/src/sec_certs/data/cpes/outdated/cc.json
diff --git a/data/cpes/outdated/fips.json b/src/sec_certs/data/cpes/outdated/fips.json
index 8683da09..8683da09 100644
--- a/data/cpes/outdated/fips.json
+++ b/src/sec_certs/data/cpes/outdated/fips.json
diff --git a/data/cpes/random.csv b/src/sec_certs/data/cpes/random.csv
index efc62f45..efc62f45 100644
--- a/data/cpes/random.csv
+++ b/src/sec_certs/data/cpes/random.csv
diff --git a/data/cpes/readme.md b/src/sec_certs/data/cpes/readme.md
index b96cd4f3..b96cd4f3 100644
--- a/data/cpes/readme.md
+++ b/src/sec_certs/data/cpes/readme.md
diff --git a/data/cpes/validation_test_split/cc/test.json b/src/sec_certs/data/cpes/validation_test_split/cc/test.json
index bee8d8f5..bee8d8f5 100644
--- a/data/cpes/validation_test_split/cc/test.json
+++ b/src/sec_certs/data/cpes/validation_test_split/cc/test.json
diff --git a/data/cpes/validation_test_split/cc/validation.json b/src/sec_certs/data/cpes/validation_test_split/cc/validation.json
index 0ecc3a0e..0ecc3a0e 100644
--- a/data/cpes/validation_test_split/cc/validation.json
+++ b/src/sec_certs/data/cpes/validation_test_split/cc/validation.json
diff --git a/data/cpes/validation_test_split/fips/test.json b/src/sec_certs/data/cpes/validation_test_split/fips/test.json
index 3768ff72..3768ff72 100644
--- a/data/cpes/validation_test_split/fips/test.json
+++ b/src/sec_certs/data/cpes/validation_test_split/fips/test.json
diff --git a/data/cpes/validation_test_split/fips/validation.json b/src/sec_certs/data/cpes/validation_test_split/fips/validation.json
index c06ae9df..c06ae9df 100644
--- a/data/cpes/validation_test_split/fips/validation.json
+++ b/src/sec_certs/data/cpes/validation_test_split/fips/validation.json
diff --git a/data/information_retrieval/split/metadata.csv b/src/sec_certs/data/information_retrieval/split/metadata.csv
index 2c5898d7..2c5898d7 100644
--- a/data/information_retrieval/split/metadata.csv
+++ b/src/sec_certs/data/information_retrieval/split/metadata.csv
diff --git a/data/information_retrieval/split/sampled_images.json b/src/sec_certs/data/information_retrieval/split/sampled_images.json
index a4c6b1be..a4c6b1be 100644
--- a/data/information_retrieval/split/sampled_images.json
+++ b/src/sec_certs/data/information_retrieval/split/sampled_images.json
diff --git a/data/information_retrieval/split/test.json b/src/sec_certs/data/information_retrieval/split/test.json
index be91674f..be91674f 100644
--- a/data/information_retrieval/split/test.json
+++ b/src/sec_certs/data/information_retrieval/split/test.json
diff --git a/data/information_retrieval/split/train.json b/src/sec_certs/data/information_retrieval/split/train.json
index 16d8e45c..16d8e45c 100644
--- a/data/information_retrieval/split/train.json
+++ b/src/sec_certs/data/information_retrieval/split/train.json
diff --git a/data/information_retrieval/split/valid.json b/src/sec_certs/data/information_retrieval/split/valid.json
index fc91875d..fc91875d 100644
--- a/data/information_retrieval/split/valid.json
+++ b/src/sec_certs/data/information_retrieval/split/valid.json
diff --git a/data/readme.md b/src/sec_certs/data/readme.md
index a4556b70..a4556b70 100644
--- a/data/readme.md
+++ b/src/sec_certs/data/readme.md
diff --git a/data/reference_annotations/manual_annotations/test.csv b/src/sec_certs/data/reference_annotations/manual_annotations/test.csv
index fd43357f..fd43357f 100644
--- a/data/reference_annotations/manual_annotations/test.csv
+++ b/src/sec_certs/data/reference_annotations/manual_annotations/test.csv
diff --git a/data/reference_annotations/manual_annotations/train.csv b/src/sec_certs/data/reference_annotations/manual_annotations/train.csv
index b29c609e..b29c609e 100644
--- a/data/reference_annotations/manual_annotations/train.csv
+++ b/src/sec_certs/data/reference_annotations/manual_annotations/train.csv
diff --git a/data/reference_annotations/manual_annotations/valid.csv b/src/sec_certs/data/reference_annotations/manual_annotations/valid.csv
index 09368347..09368347 100644
--- a/data/reference_annotations/manual_annotations/valid.csv
+++ b/src/sec_certs/data/reference_annotations/manual_annotations/valid.csv
diff --git a/data/reference_annotations/outdated_manually_annotated_references.csv b/src/sec_certs/data/reference_annotations/outdated_manually_annotated_references.csv
index 346ed3af..346ed3af 100644
--- a/data/reference_annotations/outdated_manually_annotated_references.csv
+++ b/src/sec_certs/data/reference_annotations/outdated_manually_annotated_references.csv
diff --git a/data/reference_annotations/readme.md b/src/sec_certs/data/reference_annotations/readme.md
index c17c55dd..c17c55dd 100644
--- a/data/reference_annotations/readme.md
+++ b/src/sec_certs/data/reference_annotations/readme.md
diff --git a/data/reference_annotations/split/test.json b/src/sec_certs/data/reference_annotations/split/test.json
index 41ac06ee..41ac06ee 100644
--- a/data/reference_annotations/split/test.json
+++ b/src/sec_certs/data/reference_annotations/split/test.json
diff --git a/data/reference_annotations/split/train.json b/src/sec_certs/data/reference_annotations/split/train.json
index b2e65750..b2e65750 100644
--- a/data/reference_annotations/split/train.json
+++ b/src/sec_certs/data/reference_annotations/split/train.json
diff --git a/data/reference_annotations/split/valid.json b/src/sec_certs/data/reference_annotations/split/valid.json
index 939924f0..939924f0 100644
--- a/data/reference_annotations/split/valid.json
+++ b/src/sec_certs/data/reference_annotations/split/valid.json
diff --git a/data/sar_correlations/all_certs_sar_cve_corr.csv b/src/sec_certs/data/sar_correlations/all_certs_sar_cve_corr.csv
index 9a459b6b..9a459b6b 100644
--- a/data/sar_correlations/all_certs_sar_cve_corr.csv
+++ b/src/sec_certs/data/sar_correlations/all_certs_sar_cve_corr.csv
diff --git a/data/sar_correlations/readme.md b/src/sec_certs/data/sar_correlations/readme.md
index 6695d4f1..6695d4f1 100644
--- a/data/sar_correlations/readme.md
+++ b/src/sec_certs/data/sar_correlations/readme.md
diff --git a/data/sar_correlations/vuln_rich_certs_sar_cve_corr.csv b/src/sec_certs/data/sar_correlations/vuln_rich_certs_sar_cve_corr.csv
index a5bc6df2..a5bc6df2 100644
--- a/data/sar_correlations/vuln_rich_certs_sar_cve_corr.csv
+++ b/src/sec_certs/data/sar_correlations/vuln_rich_certs_sar_cve_corr.csv