Merge pull request #3402 from csordasmarton/diagnostic_message_hash

[analyzer] Create new diagnostic message hash

analyzer/codechecker_analyzer/analyzers/clangsa/result_handler.py

@@ -29,7 +29,12 @@ def postprocess_result(self):
         Override the context sensitive issue hash in the plist files to
         context insensitive if it is enabled during analysis.
         """
-        if self.report_hash_type in ['context-free', 'context-free-v2'] and \
-                os.path.exists(self.analyzer_result_file):
-            replace_report_hash(self.analyzer_result_file,
-                                HashType.CONTEXT_FREE)
+        if os.path.exists(self.analyzer_result_file):
+            if self.report_hash_type in ['context-free', 'context-free-v2']:
+                replace_report_hash(
+                    self.analyzer_result_file,
+                    HashType.CONTEXT_FREE)
+            elif self.report_hash_type == 'diagnostic-message':
+                replace_report_hash(
+                    self.analyzer_result_file,
+                    HashType.DIAGNOSTIC_MESSAGE)

analyzer/codechecker_analyzer/analyzers/clangtidy/result_handler.py

@@ -57,3 +57,5 @@ def postprocess_result(self):
         # context free hash for 'context-free' choice.
         if self.report_hash_type == 'context-free-v2':
             replace_report_hash(output_file, HashType.CONTEXT_FREE)
+        elif self.report_hash_type == 'diagnostic-message':
+            replace_report_hash(output_file, HashType.DIAGNOSTIC_MESSAGE)

analyzer/codechecker_analyzer/cmd/analyze.py

@@ -75,6 +75,13 @@
     from the source content are removed.
   * 'column numbers' from the main diag sections location.
 
+- diagnostic-message:
+  * Same as 'context-free-v2' (file name, checker message etc.)
+  * 'bug step messages' from all events.
+
+  Be careful with this hash because it can change easily for example on
+  variable / function renames.
+
 OUR RECOMMENDATION: we recommend you to use 'context-free-v2' hash because the
 hash will not be changed so easily for example on code indentation or when a
 checker is renamed.
@@ -289,7 +296,10 @@ def add_arguments_to_parser(parser):
                         dest="report_hash",
                         default=argparse.SUPPRESS,
                         required=False,
-                        choices=['context-free', 'context-free-v2'],
+                        choices=[
+                            'context-free',
+                            'context-free-v2',
+                            'diagnostic-message'],
                         help="R|Specify the hash calculation method for "
                              "reports. By default the calculation method for "
                              "Clang Static Analyzer is context sensitive and "
@@ -300,6 +310,9 @@ def add_arguments_to_parser(parser):
                              "(kept for backward compatibility).\n"
                              "- context-free-v2: context free hash is used "
                              "for ClangSA and Clang Tidy.\n"
+                             "- diagnostic-message: context free hash with "
+                             "bug step messages is used for ClangSA and "
+                             "Clang Tidy.\n"
                              "See the 'issue hashes' section of the help "
                              "message of this command below for more "
                              "information.\n"

analyzer/codechecker_analyzer/cmd/check.py

@@ -233,7 +233,10 @@ def add_arguments_to_parser(parser):
                                dest="report_hash",
                                default=argparse.SUPPRESS,
                                required=False,
-                               choices=['context-free', 'context-free-v2'],
+                               choices=[
+                                   'context-free',
+                                   'context-free-v2',
+                                   'diagnostic-message'],
                                help="R|Specify the hash calculation method "
                                     "for reports. By default the calculation "
                                     "method for Clang Static Analyzer is "
@@ -246,6 +249,9 @@ def add_arguments_to_parser(parser):
                                     "compatibility).\n"
                                     "- context-free-v2: context free hash is "
                                     "used for ClangSA and Clang Tidy.\n"
+                                    "- diagnostic-message: context free hash "
+                                    "with bug step messages is used for "
+                                    "ClangSA and Clang Tidy.\n"
                                     "See the 'issue hashes' section of the "
                                     "help message of this command below for "
                                     "more information.\n"

analyzer/tests/functional/analyze_and_parse/test_files/diagnostic_message_hash_clang_tidy.output

@@ -0,0 +1,57 @@
+NORMAL#CodeChecker log --output $LOGFILE$ --build "make context_hash" --quiet
+NORMAL#CodeChecker analyze $LOGFILE$ --output $OUTPUT$ --report-hash=diagnostic-message --analyzers clang-tidy
+NORMAL#CodeChecker parse $OUTPUT$ --print-steps
+CHECK#CodeChecker check --build "make context_hash" --output $OUTPUT$ --quiet --print-steps --report-hash=diagnostic-message --analyzers clang-tidy
+--------------------------------------------------------------------------------
+[] - Starting build ...
+[] - Build finished successfully.
+[] - Starting static analysis ...
+[] - [1/1] clang-tidy analyzed context_hash.cpp successfully.
+[] - ----==== Summary ====----
+[] - Successfully analyzed
+[] -   clang-tidy: 1
+[] - Total analyzed compilation commands: 1
+[] - ----=================----
+[] - Analysis finished.
+[] - To view results in the terminal use the "CodeChecker parse" command.
+[] - To store results use the "CodeChecker store" command.
+[] - See --help and the user guide for further options about parsing and storing the reports.
+[] - ----=================----
+[HIGH] context_hash.cpp:23:23: suspicious usage of 'sizeof(K)'; did you mean 'K'? [bugprone-sizeof-expression]
+  std::memset(buf, 0, sizeof(BUFLEN));  // sizeof(42) ==> sizeof(int)
+                      ^
+  Report hash: 449e90b974072a5fc5f5b97e2bae4a3f
+  Steps:
+    1, context_hash.cpp:23:23: suspicious usage of 'sizeof(K)'; did you mean 'K'?
+
+[HIGH] context_hash.cpp:34:23: suspicious usage of 'sizeof(K)'; did you mean 'K'? [bugprone-sizeof-expression]
+  std::memset(buf, 0, sizeof(BUFLEN));  // sizeof(42) ==> sizeof(int)
+                      ^
+  Report hash: 449e90b974072a5fc5f5b97e2bae4a3f
+  Steps:
+    1, context_hash.cpp:34:23: suspicious usage of 'sizeof(K)'; did you mean 'K'?
+
+[HIGH] context_hash.cpp:37:21: suspicious usage of 'sizeof(K)'; did you mean 'K'? [bugprone-sizeof-expression]
+std::memset(buf, 0, sizeof(BUFLEN));  // sizeof(42) ==> sizeof(int)
+                    ^
+  Report hash: 449e90b974072a5fc5f5b97e2bae4a3f
+  Steps:
+    1, context_hash.cpp:37:21: suspicious usage of 'sizeof(K)'; did you mean 'K'?
+
+Found 3 defect(s) in context_hash.cpp
+
+
+----==== Summary ====----
+-------------------------------
+Filename         | Report count
+-------------------------------
+context_hash.cpp |            3
+-------------------------------
+-----------------------
+Severity | Report count
+-----------------------
+HIGH     |            3
+-----------------------
+----=================----
+Total number of reports: 3
+----=================----

analyzer/tests/functional/analyze_and_parse/test_files/diagnostic_message_hash_clangsa.output

@@ -0,0 +1,57 @@
+NORMAL#CodeChecker log --output $LOGFILE$ --build "make context_hash" --quiet
+NORMAL#CodeChecker analyze $LOGFILE$ --output $OUTPUT$ --report-hash=diagnostic-message --analyzers clangsa
+NORMAL#CodeChecker parse $OUTPUT$ --print-steps
+CHECK#CodeChecker check --build "make context_hash" --output $OUTPUT$ --quiet --print-steps --report-hash=diagnostic-message --analyzers clangsa
+--------------------------------------------------------------------------------
+[] - Starting build ...
+[] - Build finished successfully.
+[] - Starting static analysis ...
+[] - [1/1] clangsa analyzed context_hash.cpp successfully.
+[] - ----==== Summary ====----
+[] - Successfully analyzed
+[] -   clangsa: 1
+[] - Total analyzed compilation commands: 1
+[] - ----=================----
+[] - Analysis finished.
+[] - To view results in the terminal use the "CodeChecker parse" command.
+[] - To store results use the "CodeChecker store" command.
+[] - See --help and the user guide for further options about parsing and storing the reports.
+[] - ----=================----
+[LOW] context_hash.cpp:8:3: Value stored to 'x' is never read [deadcode.DeadStores]
+  x = 1;
+  ^
+  Report hash: d101ee6221bb0ee2294619731f8a37de
+  Steps:
+    1, context_hash.cpp:8:3: Value stored to 'x' is never read
+
+[LOW] context_hash.cpp:13:3: Value stored to 'x' is never read [deadcode.DeadStores]
+  x = 1;
+  ^
+  Report hash: d101ee6221bb0ee2294619731f8a37de
+  Steps:
+    1, context_hash.cpp:13:3: Value stored to 'x' is never read
+
+[LOW] context_hash.cpp:19:3: Value stored to 'z' is never read [deadcode.DeadStores]
+  z = 1;
+  ^
+  Report hash: 9bf6918cf70cc58a37ea556c4afed103
+  Steps:
+    1, context_hash.cpp:19:3: Value stored to 'z' is never read
+
+Found 3 defect(s) in context_hash.cpp
+
+
+----==== Summary ====----
+-------------------------------
+Filename         | Report count
+-------------------------------
+context_hash.cpp |            3
+-------------------------------
+-----------------------
+Severity | Report count
+-----------------------
+LOW      |            3
+-----------------------
+----=================----
+Total number of reports: 3
+----=================----

docs/analyzer/report_identification.md

@@ -66,3 +66,13 @@ The hashes are calculated based on the following information:
 - **content of the line** where the bug is if it can be read up. All the
 whitespaces from the source content are removed.
 - **range column numbers** where the bug is.
+
+#### diagnostic-message
+The hashes are calculated based on the following information:
+- Same as `context-free-v2` (**file name**, **checker message** etc.)
+- **bug step messages** (e.g.: *Calling 'foo'*, *Entered call from 'main'*
+etc.).
+
+**Note**: this is an experimental hash and it is not recommended to use it on
+your project because this hash can change very easily for example on variable /
+function renames.

docs/analyzer/user_guide.md

@@ -96,7 +96,7 @@ usage: CodeChecker check [-h] [-o OUTPUT_DIR] [-t {plist}] [-q]
                          [--keep-gcc-include-fixed] [--keep-gcc-intrin]
                          (-b COMMAND | -l LOGFILE) [-j JOBS] [-c]
                          [--compile-uniqueing COMPILE_UNIQUEING]
-                         [--report-hash {context-free,context-free-v2}]
+                         [--report-hash {context-free,context-free-v2,diagnostic-message}]
                          [-i SKIPFILE | --file FILE [FILE ...]]
                          [--analyzers ANALYZER [ANALYZER ...]]
                          [--capture-analysis-output] [--generate-reproducer]
@@ -189,7 +189,7 @@ analyzer arguments:
                         directory. (By default, CodeChecker would keep reports
                         and overwrites only those files that were update by
                         the current build command).
-  --report-hash {context-free,context-free-v2}
+  --report-hash {context-free,context-free-v2,diagnostic-message}
                         Specify the hash calculation method for reports. By
                         default the calculation method for Clang Static
                         Analyzer is context sensitive and for Clang Tidy it is
@@ -200,6 +200,8 @@ analyzer arguments:
                         compatibility).
                         - context-free-v2: context free hash is used for
                         ClangSA and Clang Tidy.
+                        - diagnostic-message: context free hash with bug step
+                        messages is used for ClangSA and Clang Tidy.
                         See the 'issue hashes' section of the help message of
                         this command below for more information.
                         USE WISELY AND AT YOUR OWN RISK!
@@ -506,6 +508,13 @@ generated and not the context free hash (kept for backward compatibility). Use
     from the source content are removed.
   * 'column numbers' from the main diag sections location.
 
+- diagnostic-message:
+  * Same as 'context-free-v2' (file name, checker message etc.)
+  * 'bug step messages' from all events.
+
+  Be careful with this hash because it can change easily for example on
+  variable / function renames.
+
 OUR RECOMMENDATION: we recommend you to use 'context-free-v2' hash because the
 hash will not be changed so easily for example on code indentation or when a
 checker is renamed.
@@ -819,7 +828,7 @@ usage: CodeChecker analyze [-h] [-j JOBS]
                            [--keep-gcc-include-fixed] [--keep-gcc-intrin]
                            [-t {plist}] [-q] [-c]
                            [--compile-uniqueing COMPILE_UNIQUEING]
-                           [--report-hash {context-free,context-free-v2}]
+                           [--report-hash {context-free,context-free-v2,diagnostic-message}]
                            [-n NAME] [--analyzers ANALYZER [ANALYZER ...]]
                            [--add-compiler-defaults]
                            [--capture-analysis-output] [--generate-reproducer]
@@ -909,7 +918,7 @@ optional arguments:
                         python regex. If more than one matches an error is
                         given. The whole compilation action text is searched
                         for match. (default: none)
---report-hash {context-free,context-free-v2}
+  --report-hash {context-free,context-free-v2,diagnostic-message}
                         Specify the hash calculation method for reports. By
                         default the calculation method for Clang Static
                         Analyzer is context sensitive and for Clang Tidy it is
@@ -920,6 +929,8 @@ optional arguments:
                         compatibility).
                         - context-free-v2: context free hash is used for
                         ClangSA and Clang Tidy.
+                        - diagnostic-message: context free hash with bug step
+                        messages is used for ClangSA and Clang Tidy.
                         See the 'issue hashes' section of the help message of
                         this command below for more information.
                         USE WISELY AND AT YOUR OWN RISK!

docs/tools/codechecker_report_hash.md

@@ -3,6 +3,7 @@
 Multiple hash types are available:
 - [`CONTEXT_FREE`](#generate-path-sensitive-report-hash)
 - [`PATH_SENSITIVE`](#generate-context-sensitive-report-hash)
+- [`DIAGNOSTIC_MESSAGE`](#generate-diagnostic-message-hash)
 
 You can use this library to generate report hash for these types by using the
 `get_report_hash` function.
@@ -36,6 +37,18 @@ High level overview of the hash content:
   from the source content are removed.
 * `column numbers` from the main diag sections location.
 
+### Generate diagnostic message hash
+`get_report_hash` function can be used to generate report hash with bug event
+messages if the hash type parameter is `DIAGNOSTIC_MESSAGE`.
+
+High level overview of the hash content:
+* Same as `context-free-v2` (*file name*, *checker message* etc.)
+* `bug step messages` from all events.
+
+**Note**: this is an experimental hash and it is not recommended to use it on
+your project because this hash can change very easily for example on variable /
+function renames.
+
 ### Generate path hash
 `get_report_path_hash` can be used to get path hash for the given bug path
 which can be used to filter deduplications of multiple reports.

tools/codechecker_report_hash/codechecker_report_hash/hash.py

@@ -61,6 +61,7 @@ class HashType(Enum):
     """ Report hash types. """
     CONTEXT_FREE = 1
     PATH_SENSITIVE = 2
+    DIAGNOSTIC_MESSAGE = 3
 
 
 def __get_line(file_path: str, line_no: int, errors: str = 'ignore') -> str:
@@ -121,7 +122,7 @@ def _remove_whitespace(line_content: str, old_col: int) -> Tuple[str, int]:
            old_col - line_strip_len
 
 
-def __get_report_hash_path_sensitive(diag: Diag, file_path: str) -> str:
+def __get_report_hash_path_sensitive(diag: Diag, file_path: str) -> List[str]:
     """ Report hash generation from the given diagnostic.
 
     Hash generation algorithm for older plist versions where no
@@ -237,15 +238,14 @@ def compare_ctrl_sections(
                 col_num = loc['col']
                 hash_content.append(str(col_num))
 
-        return __str_to_hash('|||'.join(hash_content))
-
+        return hash_content
     except Exception as ex:
         LOG.error("Hash generation failed")
         LOG.error(ex)
-        return ''
+        return []
 
 
-def __get_report_hash_context_free(diag: Diag, file_path: str) -> str:
+def __get_report_hash_context_free(diag: Diag, file_path: str) -> List[str]:
     """ Generate report hash without bug path.
 
     !!! NOT Compatible with the old hash generation method
@@ -290,23 +290,51 @@ def __get_report_hash_context_free(diag: Diag, file_path: str) -> str:
                         str(from_col),
                         str(until_col)]
 
-        return __str_to_hash('|||'.join(hash_content))
+        return hash_content
+    except Exception as ex:
+        LOG.error("Hash generation failed")
+        LOG.error(ex)
+        return []
+
+
+def __get_report_hash_diagnostic_message(
+    diag: Diag,
+    file_path: str
+) -> List[str]:
+    """ Generate report hash with bug path messages.
 
+    The hash will contain the same information as the CONTEXT_FREE hash +
+    'bug step messages' from events.
+    """
+    try:
+        hash_content = __get_report_hash_context_free(diag, file_path)
+
+        # Add bug step messages to the hash.
+        for event in [x for x in diag['path'] if x.get('kind') == 'event']:
+            hash_content.append(event['message'])
+
+        return hash_content
     except Exception as ex:
         LOG.error("Hash generation failed")
         LOG.error(ex)
-        return ''
+        return []
 
 
 def get_report_hash(diag: Diag, file_path: str, hash_type: HashType) -> str:
     """ Get report hash for the given diagnostic. """
+    hash_content = None
+
     if hash_type == HashType.CONTEXT_FREE:
-        return __get_report_hash_context_free(diag, file_path)
+        hash_content = __get_report_hash_context_free(diag, file_path)
     elif hash_type == HashType.PATH_SENSITIVE:
-        return __get_report_hash_path_sensitive(diag, file_path)
+        hash_content = __get_report_hash_path_sensitive(diag, file_path)
+    elif hash_type == HashType.DIAGNOSTIC_MESSAGE:
+        hash_content = __get_report_hash_diagnostic_message(diag, file_path)
     else:
         raise Exception("Invalid report hash type: " + str(hash_type))
 
+    return __str_to_hash('|||'.join(hash_content))
+
 
 def get_report_path_hash(report) -> str:
     """ Returns path hash for the given bug path.