[tool] Separate module for report hash generation

.travis.yml

@@ -146,6 +146,7 @@ script:
       make package &&
       make -C tools/plist_to_html test &&
       make -C tools/report-converter package test &&
+      make -C tools/codechecker_report_hash test &&
       if [[ "$CC_MODULE" = "analyzer" ]]; then
         BUILD_DIR=$TRAVIS_BUILD_DIR/build make -C analyzer \
           test_unit \

Makefile

@@ -63,7 +63,13 @@ package_report_converter: build_report_converter package_dir_structure
 	cd $(CC_BUILD_DIR) && \
 	ln -sf ../lib/python3/codechecker_report_converter/cli.py bin/report-converter
 
-package: package_dir_structure set_git_commit_template package_plist_to_html package_tu_collector package_report_converter
+build_report_hash:
+	$(MAKE) -C $(ROOT)/tools/codechecker_report_hash build
+
+package_report_hash: build_report_hash package_dir_structure
+	cp -r $(CC_TOOLS)/codechecker_report_hash/build/codechecker_report_hash/codechecker_report_hash $(CC_BUILD_LIB_DIR)
+
+package: package_dir_structure set_git_commit_template package_plist_to_html package_tu_collector package_report_converter package_report_hash
 	BUILD_DIR=$(BUILD_DIR) BUILD_LOGGER_64_BIT_ONLY=$(BUILD_LOGGER_64_BIT_ONLY) $(MAKE) -C $(CC_ANALYZER) package_analyzer
 	BUILD_DIR=$(BUILD_DIR) $(MAKE) -C $(CC_WEB) package_web
 
@@ -147,7 +153,7 @@ clean_venv_dev:
 
 clean: clean_package clean_vendor
 
-clean_package: clean_plist_to_html clean_tu_collector clean_report_converter
+clean_package: clean_plist_to_html clean_tu_collector clean_report_converter clean_report_hash
 	rm -rf $(BUILD_DIR)
 	find . -name "*.pyc" -delete
 
@@ -163,6 +169,9 @@ clean_tu_collector:
 clean_report_converter:
 	$(MAKE) -C $(CC_TOOLS)/report-converter clean
 
+clean_report_hash:
+	$(MAKE) -C $(CC_TOOLS)/codechecker_report_hash clean
+
 clean_travis:
 	# Clean CodeChecker config files stored in the users home directory.
 	rm -rf ~/.codechecker*

analyzer/codechecker_analyzer/analyzers/clangsa/result_handler.py

@@ -8,8 +8,8 @@
 """
 
 
-from codechecker_common import report
 from codechecker_common.logger import get_logger
+from codechecker_report_hash.hash import HashType, replace_report_hash
 
 from ..result_handler_base import ResultHandler
 
@@ -27,4 +27,5 @@ def postprocess_result(self):
         context insensitive if it is enabled during analysis.
         """
         if self.report_hash_type in ['context-free', 'context-free-v2']:
-            report.use_context_free_hashes(self.analyzer_result_file)
+            replace_report_hash(self.analyzer_result_file,
+                                HashType.CONTEXT_FREE)

analyzer/codechecker_analyzer/analyzers/clangtidy/output_converter.py

@@ -16,7 +16,7 @@
 import re
 
 from codechecker_common.logger import get_logger
-from codechecker_common.report import generate_report_hash
+from codechecker_report_hash.hash import get_report_hash, HashType
 
 LOG = get_logger('analyzer.tidy')
 
@@ -300,10 +300,9 @@ def _create_diag(message, fmap, files):
         diag['path'].append(PListConverter._create_event_from_note(message,
                                                                    fmap))
 
+        source_file = files[diag['location']['file']]
         diag['issue_hash_content_of_line_in_context'] \
-            = generate_report_hash(diag['path'],
-                                   files[diag['location']['file']],
-                                   message.checker)
+            = get_report_hash(diag, source_file, HashType.PATH_SENSITIVE)
 
         return diag
 

analyzer/codechecker_analyzer/analyzers/clangtidy/result_handler.py

@@ -8,8 +8,8 @@
 """
 
 
-from codechecker_common import report
 from codechecker_common.logger import get_logger
+from codechecker_report_hash.hash import HashType, replace_report_hash
 
 from ..result_handler_base import ResultHandler
 
@@ -54,4 +54,4 @@ def postprocess_result(self):
         # --report-hash option ('context-free-v2') and we still do not use
         # context free hash for 'context-free' choice.
         if self.report_hash_type == 'context-free-v2':
-            report.use_context_free_hashes(output_file)
+            replace_report_hash(output_file, HashType.CONTEXT_FREE)

analyzer/codechecker_analyzer/cmd/parse.py

@@ -29,8 +29,9 @@
 from codechecker_common.source_code_comment_handler import \
     SourceCodeCommentHandler
 from codechecker_common.output_formatters import twodim_to_str
-from codechecker_common.report import Report, get_report_path_hash
+from codechecker_common.report import Report
 from codechecker_common.source_code_comment_handler import skip_suppress_status
+from codechecker_report_hash.hash import get_report_path_hash
 
 LOG = logger.get_logger('system')
 
@@ -150,7 +151,7 @@ def write(self, file_report_map, output=sys.stdout):
                                     key=lambda r: r.main['location']['line'])
 
             for report in sorted_reports:
-                path_hash = get_report_path_hash(report, report.files)
+                path_hash = get_report_path_hash(report.bug_path, report.files)
                 if path_hash in self._processed_path_hashes:
                     LOG.debug("Not showing report because it is a "
                               "deduplication of an already processed report!")
@@ -608,7 +609,7 @@ def skip_html_report_data_handler(report_hash, source_file, report_line,
         comments.
         """
         report = Report(None, diag['path'], files)
-        path_hash = get_report_path_hash(report, files)
+        path_hash = get_report_path_hash(report.bug_path, files)
         if path_hash in processed_path_hashes:
             LOG.debug("Skip report because it is a deduplication of an "
                       "already processed report!")

analyzer/tests/unit/__init__.py

@@ -15,3 +15,4 @@
 PKG_ROOT = os.path.join(REPO_ROOT, 'build', 'CodeChecker')
 
 sys.path.append(REPO_ROOT)
+sys.path.append(os.path.join(REPO_ROOT, 'tools', 'codechecker_report_hash'))

codechecker_common/plist_parser.py

@@ -37,7 +37,8 @@
 from xml.parsers.expat import ExpatError
 
 from codechecker_common.logger import get_logger
-from codechecker_common.report import Report, generate_report_hash
+from codechecker_common.report import Report
+from codechecker_report_hash.hash import get_report_hash, HashType
 
 LOG = get_logger('report')
 
@@ -116,22 +117,6 @@ def get_checker_name(diagnostic, path=""):
     return checker_name
 
 
-def get_report_hash(diagnostic, source_file):
-    """
-    Check if checker name is available in the report.
-    Checker hash was not available in older clang versions before 3.8.
-    """
-
-    report_hash = diagnostic.get('issue_hash_content_of_line_in_context')
-    if not report_hash:
-        # Generate hash value if it is missing from the report.
-        report_hash \
-            = generate_report_hash(diagnostic['path'],
-                                   source_file,
-                                   get_checker_name(diagnostic))
-    return report_hash
-
-
 def parse_plist_file(path, source_root=None, allow_plist_update=True):
     """
     Parse the reports from a plist file.
@@ -168,9 +153,15 @@ def parse_plist_file(path, source_root=None, allow_plist_update=True):
             if source_root:
                 file_path = os.path.join(source_root, file_path.lstrip('/'))
 
-            report_hash = get_report_hash(diag, file_path)
-            main_section['issue_hash_content_of_line_in_context'] = \
-                report_hash
+            report_hash = diag.get('issue_hash_content_of_line_in_context')
+
+            if not report_hash:
+                # Generate hash value if it is missing from the report.
+                report_hash = get_report_hash(diag, file_path,
+                                              HashType.PATH_SENSITIVE)
+
+                main_section['issue_hash_content_of_line_in_context'] = \
+                    report_hash
 
             if 'issue_hash_content_of_line_in_context' not in diag:
                 # If the report hash was not in the plist, we set it in the