Rename corpus_integrity to corpus_data to reflect expanded test scope

Co-authored-by: bact <128572+bact@users.noreply.github.com>

Author: Copilot

.github/workflows/corpus-data.yml

@@ -8,14 +8,14 @@ on:
     paths:
       - ".github/workflows/corpus-data.yml"
       - "pythainlp/corpus/**"
-      - "tests/corpus_integrity/**"
+      - "tests/corpus_data/**"
   pull_request:
     branches:
       - dev
     paths:
       - ".github/workflows/corpus-data.yml"
       - "pythainlp/corpus/**"
-      - "tests/corpus_integrity/**"
+      - "tests/corpus_data/**"
 
 # Avoid duplicate runs for the same source branch and repository
 concurrency:
@@ -50,16 +50,16 @@ jobs:
         env:
           PYTHONIOENCODING: utf-8
         run: |
-          python -m unittest discover -s tests/corpus_integrity -p "test_catalog*.py" -v
+          python -m unittest discover -s tests/corpus_data -p "test_catalog*.py" -v
 
       - name: Test built-in corpus files
         env:
           PYTHONIOENCODING: utf-8
         run: |
-          python -m unittest discover -s tests/corpus_integrity -p "test_builtin_*.py" -v
+          python -m unittest discover -s tests/corpus_data -p "test_builtin_*.py" -v
 
       - name: Test downloadable corpus files
         env:
           PYTHONIOENCODING: utf-8
         run: |
-          python -m unittest discover -s tests/corpus_integrity -p "test_downloadable_*.py" -v
+          python -m unittest discover -s tests/corpus_data -p "test_downloadable_*.py" -v

tests/README.md

@@ -169,20 +169,20 @@ for real-world usage:
 - Multi-engine robustness testing across all core tokenization engines
 - Very long strings that can cause performance issues (issue #893)
 
-## Corpus data tests (corpus_integrity/)
+## Corpus data tests (corpus_data/)
 
 A separate test suite that verifies the integrity, format, parseability, and catalog
 functionality of corpus data in PyThaiNLP. These tests are separate from regular unit tests
 because they test actual file loading and parsing (not mocked), downloadable corpus tests
 require network access, and they verify corpus catalog operations.
 
 For detailed information about corpus data tests, see:
-[tests/corpus_integrity/README.md](corpus_integrity/README.md)
+[tests/corpus_data/README.md](corpus_data/README.md)
 
 The corpus data tests are triggered automatically via GitHub Actions
-when changes are made to `pythainlp/corpus/**` or `tests/corpus_integrity/**`.
+when changes are made to `pythainlp/corpus/**` or `tests/corpus_data/**`.
 
 **Run corpus data tests:**
 ```shell
-python -m unittest tests.corpus_integrity
+python -m unittest tests.corpus_data
 ```

tests/corpus_integrity/README.md tests/corpus_data/README.mdtests/corpus_integrity/README.md renamed to tests/corpus_data/README.md

@@ -49,32 +49,32 @@ size of the downloads.
 Run all corpus data tests:
 
 ```bash
-python -m unittest discover -s tests/corpus_integrity -v
+python -m unittest discover -s tests/corpus_data -v
 ```
 
 Run only catalog tests:
 
 ```bash
-python -m unittest tests.corpus_integrity.test_catalog -v
+python -m unittest tests.corpus_data.test_catalog -v
 ```
 
 Run only built-in corpus tests:
 
 ```bash
-python -m unittest tests.corpus_integrity.test_builtin_corpus -v
+python -m unittest tests.corpus_data.test_builtin_corpus -v
 ```
 
 Run only downloadable corpus tests:
 
 ```bash
-python -m unittest tests.corpus_integrity.test_downloadable_corpus -v
+python -m unittest tests.corpus_data.test_downloadable_corpus -v
 ```
 
 ## CI Integration
 
 The corpus data tests run automatically via GitHub Actions workflow (`.github/workflows/corpus-data.yml`) when:
 - Changes are made to `pythainlp/corpus/**`
-- Changes are made to `tests/corpus_integrity/**`
+- Changes are made to `tests/corpus_data/**`
 - The workflow file itself is modified
 
 ## What is Tested

tests/corpus_data/__init__.py

@@ -0,0 +1,11 @@
+# SPDX-FileCopyrightText: 2016-2026 PyThaiNLP Project
+# SPDX-FileType: SOURCE
+# SPDX-License-Identifier: Apache-2.0
+"""
+Corpus data tests.
+
+These tests verify the integrity, format, parseability, and catalog
+functionality of corpus data in PyThaiNLP. They are separate from
+regular unit tests to avoid slowing down development test cycles
+with large file downloads and network access.
+"""