merge: Sync main into development - includes Task 1.1 and 1.2 fixes

src/skill_seekers/cli/github_scraper.py

@@ -33,7 +33,7 @@ except ImportError:
 
 # Import code analyzer for deep code analysis
 try:
-    from code_analyzer import CodeAnalyzer
+    from .code_analyzer import CodeAnalyzer
     CODE_ANALYZER_AVAILABLE = True
 except ImportError:
     CODE_ANALYZER_AVAILABLE = False

src/skill_seekers/cli/pdf_scraper.py

@@ -19,7 +19,7 @@ import argparse
 from pathlib import Path
 
 # Import the PDF extractor
-from pdf_extractor_poc import PDFExtractor
+from .pdf_extractor_poc import PDFExtractor
 
 
 class PDFToSkillConverter:

src/skill_seekers/cli/test_unified_simple.py

@@ -14,7 +14,7 @@ from pathlib import Path
 # Add CLI to path
 sys.path.insert(0, str(Path(__file__).parent))
 
-from config_validator import validate_config
+from .config_validator import validate_config
 
 def test_validate_existing_unified_configs():
     """Test that all existing unified configs are valid"""

tests/test_github_scraper.py

@@ -24,9 +24,6 @@ from pathlib import Path
 from unittest.mock import Mock, patch, MagicMock
 from datetime import datetime
 
-# Add parent directory to path for imports
-sys.path.insert(0, str(Path(__file__).parent.parent / "cli"))
-
 try:
     from github import Github, GithubException
     PYGITHUB_AVAILABLE = True
@@ -40,7 +37,7 @@ class TestGitHubScraperInitialization(unittest.TestCase):
     def setUp(self):
         if not PYGITHUB_AVAILABLE:
             self.skipTest("PyGithub not installed")
-        from github_scraper import GitHubScraper
+        from skill_seekers.cli.github_scraper import GitHubScraper
         self.GitHubScraper = GitHubScraper
 
         # Create temporary directory for test output
@@ -74,7 +71,7 @@ class TestGitHubScraperInitialization(unittest.TestCase):
             'github_token': 'test_token_123'
         }
 
-        with patch('github_scraper.Github') as mock_github:
+        with patch('skill_seekers.cli.github_scraper.Github') as mock_github:
             scraper = self.GitHubScraper(config)
             mock_github.assert_called_once_with('test_token_123')
 
@@ -87,7 +84,7 @@ class TestGitHubScraperInitialization(unittest.TestCase):
         }
 
         with patch.dict(os.environ, {'GITHUB_TOKEN': 'env_token_456'}):
-            with patch('github_scraper.Github') as mock_github:
+            with patch('skill_seekers.cli.github_scraper.Github') as mock_github:
                 scraper = self.GitHubScraper(config)
                 mock_github.assert_called_once_with('env_token_456')
 
@@ -99,7 +96,7 @@ class TestGitHubScraperInitialization(unittest.TestCase):
             'github_token': None
         }
 
-        with patch('github_scraper.Github') as mock_github:
+        with patch('skill_seekers.cli.github_scraper.Github') as mock_github:
             with patch.dict(os.environ, {}, clear=True):
                 scraper = self.GitHubScraper(config)
                 # Should create unauthenticated client
@@ -125,7 +122,7 @@ class TestREADMEExtraction(unittest.TestCase):
     def setUp(self):
         if not PYGITHUB_AVAILABLE:
             self.skipTest("PyGithub not installed")
-        from github_scraper import GitHubScraper
+        from skill_seekers.cli.github_scraper import GitHubScraper
         self.GitHubScraper = GitHubScraper
 
     def test_extract_readme_success(self):
@@ -139,7 +136,7 @@ class TestREADMEExtraction(unittest.TestCase):
         mock_content = Mock()
         mock_content.decoded_content = b'# React\n\nA JavaScript library'
 
-        with patch('github_scraper.Github'):
+        with patch('skill_seekers.cli.github_scraper.Github'):
             scraper = self.GitHubScraper(config)
             scraper.repo = Mock()
             scraper.repo.get_contents.return_value = mock_content
@@ -157,7 +154,7 @@ class TestREADMEExtraction(unittest.TestCase):
             'github_token': None
         }
 
-        with patch('github_scraper.Github'):
+        with patch('skill_seekers.cli.github_scraper.Github'):
             scraper = self.GitHubScraper(config)
             scraper.repo = Mock()
 
@@ -184,7 +181,7 @@ class TestREADMEExtraction(unittest.TestCase):
             'github_token': None
         }
 
-        with patch('github_scraper.Github'):
+        with patch('skill_seekers.cli.github_scraper.Github'):
             scraper = self.GitHubScraper(config)
             scraper.repo = Mock()
             scraper.repo.get_contents.side_effect = GithubException(404, 'Not found')
@@ -201,7 +198,7 @@ class TestLanguageDetection(unittest.TestCase):
     def setUp(self):
         if not PYGITHUB_AVAILABLE:
             self.skipTest("PyGithub not installed")
-        from github_scraper import GitHubScraper
+        from skill_seekers.cli.github_scraper import GitHubScraper
         self.GitHubScraper = GitHubScraper
 
     def test_extract_languages_success(self):
@@ -212,7 +209,7 @@ class TestLanguageDetection(unittest.TestCase):
             'github_token': None
         }
 
-        with patch('github_scraper.Github'):
+        with patch('skill_seekers.cli.github_scraper.Github'):
             scraper = self.GitHubScraper(config)
             scraper.repo = Mock()
             scraper.repo.get_languages.return_value = {
@@ -243,7 +240,7 @@ class TestLanguageDetection(unittest.TestCase):
             'github_token': None
         }
 
-        with patch('github_scraper.Github'):
+        with patch('skill_seekers.cli.github_scraper.Github'):
             scraper = self.GitHubScraper(config)
             scraper.repo = Mock()
             scraper.repo.get_languages.return_value = {}
@@ -260,7 +257,7 @@ class TestIssuesExtraction(unittest.TestCase):
     def setUp(self):
         if not PYGITHUB_AVAILABLE:
             self.skipTest("PyGithub not installed")
-        from github_scraper import GitHubScraper
+        from skill_seekers.cli.github_scraper import GitHubScraper
         self.GitHubScraper = GitHubScraper
 
     def test_extract_issues_success(self):
@@ -310,7 +307,7 @@ class TestIssuesExtraction(unittest.TestCase):
         mock_issue2.body = 'Feature description'
         mock_issue2.pull_request = None
 
-        with patch('github_scraper.Github'):
+        with patch('skill_seekers.cli.github_scraper.Github'):
             scraper = self.GitHubScraper(config)
             scraper.repo = Mock()
             scraper.repo.get_issues.return_value = [mock_issue1, mock_issue2]
@@ -361,7 +358,7 @@ class TestIssuesExtraction(unittest.TestCase):
         mock_pr.title = 'Pull request'
         mock_pr.pull_request = Mock()  # Has pull_request attribute
 
-        with patch('github_scraper.Github'):
+        with patch('skill_seekers.cli.github_scraper.Github'):
             scraper = self.GitHubScraper(config)
             scraper.repo = Mock()
             scraper.repo.get_issues.return_value = [mock_issue, mock_pr]
@@ -399,7 +396,7 @@ class TestIssuesExtraction(unittest.TestCase):
             mock_issue.pull_request = None
             mock_issues.append(mock_issue)
 
-        with patch('github_scraper.Github'):
+        with patch('skill_seekers.cli.github_scraper.Github'):
             scraper = self.GitHubScraper(config)
             scraper.repo = Mock()
             scraper.repo.get_issues.return_value = mock_issues
@@ -417,7 +414,7 @@ class TestChangelogExtraction(unittest.TestCase):
     def setUp(self):
         if not PYGITHUB_AVAILABLE:
             self.skipTest("PyGithub not installed")
-        from github_scraper import GitHubScraper
+        from skill_seekers.cli.github_scraper import GitHubScraper
         self.GitHubScraper = GitHubScraper
 
     def test_extract_changelog_success(self):
@@ -431,7 +428,7 @@ class TestChangelogExtraction(unittest.TestCase):
         mock_content = Mock()
         mock_content.decoded_content = b'# Changelog\n\n## v1.0.0\n- Initial release'
 
-        with patch('github_scraper.Github'):
+        with patch('skill_seekers.cli.github_scraper.Github'):
             scraper = self.GitHubScraper(config)
             scraper.repo = Mock()
             scraper.repo.get_contents.return_value = mock_content
@@ -449,7 +446,7 @@ class TestChangelogExtraction(unittest.TestCase):
             'github_token': None
         }
 
-        with patch('github_scraper.Github'):
+        with patch('skill_seekers.cli.github_scraper.Github'):
             scraper = self.GitHubScraper(config)
             scraper.repo = Mock()
 
@@ -479,7 +476,7 @@ class TestChangelogExtraction(unittest.TestCase):
             'github_token': None
         }
 
-        with patch('github_scraper.Github'):
+        with patch('skill_seekers.cli.github_scraper.Github'):
             scraper = self.GitHubScraper(config)
             scraper.repo = Mock()
             scraper.repo.get_contents.side_effect = GithubException(404, 'Not found')
@@ -496,7 +493,7 @@ class TestReleasesExtraction(unittest.TestCase):
     def setUp(self):
         if not PYGITHUB_AVAILABLE:
             self.skipTest("PyGithub not installed")
-        from github_scraper import GitHubScraper
+        from skill_seekers.cli.github_scraper import GitHubScraper
         self.GitHubScraper = GitHubScraper
 
     def test_extract_releases_success(self):
@@ -532,7 +529,7 @@ class TestReleasesExtraction(unittest.TestCase):
         mock_release2.tarball_url = 'https://github.com/facebook/react/archive/v18.0.0-rc.0.tar.gz'
         mock_release2.zipball_url = 'https://github.com/facebook/react/archive/v18.0.0-rc.0.zip'
 
-        with patch('github_scraper.Github'):
+        with patch('skill_seekers.cli.github_scraper.Github'):
             scraper = self.GitHubScraper(config)
             scraper.repo = Mock()
             scraper.repo.get_releases.return_value = [mock_release1, mock_release2]
@@ -562,7 +559,7 @@ class TestReleasesExtraction(unittest.TestCase):
             'github_token': None
         }
 
-        with patch('github_scraper.Github'):
+        with patch('skill_seekers.cli.github_scraper.Github'):
             scraper = self.GitHubScraper(config)
             scraper.repo = Mock()
             scraper.repo.get_releases.return_value = []
@@ -579,7 +576,7 @@ class TestGitHubToSkillConverter(unittest.TestCase):
     def setUp(self):
         if not PYGITHUB_AVAILABLE:
             self.skipTest("PyGithub not installed")
-        from github_scraper import GitHubToSkillConverter
+        from skill_seekers.cli.github_scraper import GitHubToSkillConverter
         self.GitHubToSkillConverter = GitHubToSkillConverter
 
         # Create temporary directory for test output
@@ -646,7 +643,7 @@ class TestGitHubToSkillConverter(unittest.TestCase):
         }
 
         # Override data file path
-        with patch('github_scraper.GitHubToSkillConverter.__init__') as mock_init:
+        with patch('skill_seekers.cli.github_scraper.GitHubToSkillConverter.__init__') as mock_init:
             mock_init.return_value = None
             converter = self.GitHubToSkillConverter(config)
             converter.data_file = str(self.data_file)
@@ -669,7 +666,7 @@ class TestGitHubToSkillConverter(unittest.TestCase):
         }
 
         # Patch the paths to use our temp directory
-        with patch('github_scraper.GitHubToSkillConverter._load_data') as mock_load:
+        with patch('skill_seekers.cli.github_scraper.GitHubToSkillConverter._load_data') as mock_load:
             mock_load.return_value = self.mock_data
             converter = self.GitHubToSkillConverter(config)
             converter.skill_dir = str(self.output_dir / 'test_skill')
@@ -689,7 +686,7 @@ class TestErrorHandling(unittest.TestCase):
     def setUp(self):
         if not PYGITHUB_AVAILABLE:
             self.skipTest("PyGithub not installed")
-        from github_scraper import GitHubScraper
+        from skill_seekers.cli.github_scraper import GitHubScraper
         self.GitHubScraper = GitHubScraper
 
     def test_invalid_repo_name(self):
@@ -700,7 +697,7 @@ class TestErrorHandling(unittest.TestCase):
             'github_token': None
         }
 
-        with patch('github_scraper.Github'):
+        with patch('skill_seekers.cli.github_scraper.Github'):
             scraper = self.GitHubScraper(config)
             scraper.repo = None
             scraper.github.get_repo = Mock(side_effect=GithubException(404, 'Not found'))
@@ -720,7 +717,7 @@ class TestErrorHandling(unittest.TestCase):
             'max_issues': 10
         }
 
-        with patch('github_scraper.Github'):
+        with patch('skill_seekers.cli.github_scraper.Github'):
             scraper = self.GitHubScraper(config)
             scraper.repo = Mock()
             scraper.repo.get_issues.side_effect = GithubException(403, 'Rate limit exceeded')

tests/test_pdf_scraper.py

@@ -19,9 +19,6 @@ import shutil
 from pathlib import Path
 from unittest.mock import Mock, patch, MagicMock
 
-# Add parent directory to path for imports
-sys.path.insert(0, str(Path(__file__).parent.parent / "cli"))
-
 try:
     import fitz  # PyMuPDF
     PYMUPDF_AVAILABLE = True
@@ -35,7 +32,7 @@ class TestPDFToSkillConverter(unittest.TestCase):
     def setUp(self):
         if not PYMUPDF_AVAILABLE:
             self.skipTest("PyMuPDF not installed")
-        from pdf_scraper import PDFToSkillConverter
+        from skill_seekers.cli.pdf_scraper import PDFToSkillConverter
         self.PDFToSkillConverter = PDFToSkillConverter
 
         # Create temporary directory for test output
@@ -88,7 +85,7 @@ class TestCategorization(unittest.TestCase):
     def setUp(self):
         if not PYMUPDF_AVAILABLE:
             self.skipTest("PyMuPDF not installed")
-        from pdf_scraper import PDFToSkillConverter
+        from skill_seekers.cli.pdf_scraper import PDFToSkillConverter
         self.PDFToSkillConverter = PDFToSkillConverter
         self.temp_dir = tempfile.mkdtemp()
 
@@ -196,7 +193,7 @@ class TestSkillBuilding(unittest.TestCase):
     def setUp(self):
         if not PYMUPDF_AVAILABLE:
             self.skipTest("PyMuPDF not installed")
-        from pdf_scraper import PDFToSkillConverter
+        from skill_seekers.cli.pdf_scraper import PDFToSkillConverter
         self.PDFToSkillConverter = PDFToSkillConverter
         self.temp_dir = tempfile.mkdtemp()
 
@@ -308,7 +305,7 @@ class TestCodeBlockHandling(unittest.TestCase):
     def setUp(self):
         if not PYMUPDF_AVAILABLE:
             self.skipTest("PyMuPDF not installed")
-        from pdf_scraper import PDFToSkillConverter
+        from skill_seekers.cli.pdf_scraper import PDFToSkillConverter
         self.PDFToSkillConverter = PDFToSkillConverter
         self.temp_dir = tempfile.mkdtemp()
 
@@ -402,7 +399,7 @@ class TestImageHandling(unittest.TestCase):
     def setUp(self):
         if not PYMUPDF_AVAILABLE:
             self.skipTest("PyMuPDF not installed")
-        from pdf_scraper import PDFToSkillConverter
+        from skill_seekers.cli.pdf_scraper import PDFToSkillConverter
         self.PDFToSkillConverter = PDFToSkillConverter
         self.temp_dir = tempfile.mkdtemp()
 
@@ -501,7 +498,7 @@ class TestErrorHandling(unittest.TestCase):
     def setUp(self):
         if not PYMUPDF_AVAILABLE:
             self.skipTest("PyMuPDF not installed")
-        from pdf_scraper import PDFToSkillConverter
+        from skill_seekers.cli.pdf_scraper import PDFToSkillConverter
         self.PDFToSkillConverter = PDFToSkillConverter
         self.temp_dir = tempfile.mkdtemp()
 
@@ -541,7 +538,7 @@ class TestJSONWorkflow(unittest.TestCase):
     def setUp(self):
         if not PYMUPDF_AVAILABLE:
             self.skipTest("PyMuPDF not installed")
-        from pdf_scraper import PDFToSkillConverter
+        from skill_seekers.cli.pdf_scraper import PDFToSkillConverter
         self.PDFToSkillConverter = PDFToSkillConverter
         self.temp_dir = tempfile.mkdtemp()