adding clang-tidy as a linter tool

ceph-pr-clang-tidy/build/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2018 PSPDFKit
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

ceph-pr-clang-tidy/build/build

@@ -0,0 +1,35 @@
+#!/bin/bash
+set -ex
+
+docs_pr_only
+if [ "$DOCS_ONLY" = true ]; then
+    echo "Only the doc/ dir changed.  No need to run make check or API tests."
+    mkdir -p $WORKSPACE/build/out
+    echo "File created to avoid Jenkins' Artifact Archiving plugin from hanging" > $WORKSPACE/build/out/mgr.foo.log
+    exit 0
+fi
+
+sudo apt-get install -y clang-tidy
+
+sudo apt-get install -y parallel
+
+build_debs ${VENV} ${vers} ${debian_version}
+
+output_file="clang-tidy-result"
+file_list="files_to_check.txt"
+
+# Store the list of files from both rgw and osd directories
+find "$WORKSPACE/src/rgw" \( -name '*.cpp' -o -name '*.hpp' -o -name '*.cc' \) > "$file_list"
+find "$WORKSPACE/src/osd" \( -name '*.cpp' -o -name '*.hpp' -o -name '*.cc' \) >> "$file_list"
+
+# Run clang-tidy and save output
+{
+  echo "Files being checked:"
+  cat "$file_list"
+  echo
+  parallel -m clang-tidy -checks="-*,bugprone-use-after-move" -p "$WORKSPACE/build" {} < "$file_list"
+} | tee "$WORKSPACE/build/$output_file"
+
+sudo chmod +x clang-tidy-to-junit.py
+
+clang-tidy-to-junit.py $WORKSPACE/src < "$WORKSPACE/build/$output_file" > "$WORKSPACE/report.xml"

ceph-pr-clang-tidy/build/clang-tidy-to-junit.py

@@ -0,0 +1,117 @@
+#!/usr/bin/env python3
+
+import sys
+import collections
+import re
+import logging
+import itertools
+from xml.sax.saxutils import escape
+
+# Create a `ErrorDescription` tuple with all the information we want to keep.
+ErrorDescription = collections.namedtuple(
+    'ErrorDescription', 'file line column error error_identifier description')
+
+
+class ClangTidyConverter:
+    # All the errors encountered.
+    errors = []
+
+    # Parses the error.
+    # Group 1: file path
+    # Group 2: line
+    # Group 3: column
+    # Group 4: error message
+    # Group 5: error identifier
+    error_regex = re.compile(
+        r"^([\w\/\.\-\ ]+):(\d+):(\d+): (.+) (\[[\w\-,\.]+\])$")
+
+    # This identifies the main error line (it has a [the-warning-type] at the end)
+    # We only create a new error when we encounter one of those.
+    main_error_identifier = re.compile(r'\[[\w\-,\.]+\]$')
+
+    def __init__(self, basename):
+        self.basename = basename
+
+    def print_junit_file(self, output_file):
+        # Write the header.
+        output_file.write("""<?xml version="1.0" encoding="UTF-8" ?>
+<testsuites id="1" name="Clang-Tidy" tests="{error_count}" errors="{error_count}" failures="0" time="0">""".format(error_count=len(self.errors)))
+
+        sorted_errors = sorted(self.errors, key=lambda x: x.file)
+
+        # Iterate through the errors, grouped by file.
+        for file, errorIterator in itertools.groupby(sorted_errors, key=lambda x: x.file):
+            errors = list(errorIterator)
+            error_count = len(errors)
+
+            # Each file gets a test-suite
+            output_file.write("""\n    <testsuite errors="{error_count}" name="{file}" tests="{error_count}" failures="0" time="0">\n"""
+                              .format(error_count=error_count, file=file))
+            for error in errors:
+                # Write each error as a test case.
+                output_file.write("""
+        <testcase id="{id}" name="{id}" time="0">
+            <failure message="{message}">
+{htmldata}
+            </failure>
+        </testcase>""".format(id="[{}/{}] {}".format(error.line, error.column, error.error_identifier),
+                              message=escape(error.error, entities={"\"": "&quot;"}),
+                              htmldata=escape(error.description)))
+            output_file.write("\n    </testsuite>\n")
+        output_file.write("</testsuites>\n")
+
+    def process_error(self, error_array):
+        if len(error_array) == 0:
+            return
+
+        result = self.error_regex.match(error_array[0])
+        if result is None:
+            logging.warning(
+                'Could not match error_array to regex: %s', error_array)
+            return
+
+        # We remove the `basename` from the `file_path` to make prettier filenames in the JUnit file.
+        file_path = result.group(1).replace(self.basename, "")
+        error = ErrorDescription(file_path, int(result.group(2)), int(
+            result.group(3)), result.group(4), result.group(5), "\n".join(error_array[1:]))
+        self.errors.append(error)
+
+    def convert(self, input_file, output_file):
+        # Collect all lines related to one error.
+        current_error = []
+        for line in input_file:
+            # If the line starts with a `/`, it is a line about a file.
+            if line[0] == '/':
+                # Look if it is the start of a error
+                if self.main_error_identifier.search(line, re.M):
+                    # If so, process any `current_error` we might have
+                    self.process_error(current_error)
+                    # Initialize `current_error` with the first line of the error.
+                    current_error = [line]
+                else:
+                    # Otherwise, append the line to the error.
+                    current_error.append(line)
+            elif len(current_error) > 0:
+                # If the line didn't start with a `/` and we have a `current_error`, we simply append
+                # the line as additional information.
+                current_error.append(line)
+            else:
+                pass
+
+        # If we still have any current_error after we read all the lines,
+        # process it.
+        if len(current_error) > 0:
+            self.process_error(current_error)
+
+        # Print the junit file.
+        self.print_junit_file(output_file)
+
+
+if __name__ == "__main__":
+    if len(sys.argv) < 2:
+        logging.error("Usage: %s base-filename-path", sys.argv[0])
+        logging.error(
+            "  base-filename-path: Removed from the filenames to make nicer paths.")
+        sys.exit(1)
+    converter = ClangTidyConverter(sys.argv[1])
+    converter.convert(sys.stdin, sys.stdout)

ceph-pr-clang-tidy/config/definitions/ceph-pr-clang-tidy.yml

@@ -0,0 +1,56 @@
+- job:
+    name: ceph-pr-clang-tidy
+    node: small
+    project-type: freestyle
+    defaults: global
+    display-name: 'ceph: Clang-tidy checks'
+    concurrent: true
+    quiet-period: 5
+    block-downstream: false
+    block-upstream: false
+    retry-count: 3
+    properties:
+      - build-discarder:
+          days-to-keep: 15
+          artifact-days-to-keep: 15
+      - github:
+          url: https://github.com/ceph/ceph/
+
+  scm:
+  - git:
+      url: https://github.com/ceph/ceph.git
+      name: origin
+      branches:
+        - origin/pr/${ghprbPullId}/merge
+      refspec: +refs/pull/${ghprbPullId}/*:refs/remotes/origin/pr/${ghprbPullId}/*
+      skip-tag: true
+      shallow-clone: true
+      honor-refspec: true
+      timeout: 20
+
+  triggers:
+  - github-pull-request:
+      allow-whitelist-orgs-as-admins: true
+      org-list:
+        - ceph
+      trigger-phrase: 'jenkins test clang-tidy'
+      only-trigger-phrase: false
+      github-hooks: true
+      permit-all: true
+      auto-close-on-fail: false
+      status-context: "Clang-tidy lint check"
+      started-status: "checking if bugs exist"
+      success-status: "no bugs found"
+      failure-status: "bugs found"
+
+  builders:
+  - shell:
+      !include-raw:
+        - ../../../scripts/build_utils.sh
+        - ../../build/build
+        - ../../build/clang-tidy-to-junit.py
+
+  publishers:
+  - junit:
+      results: report.xml
+      allow-empty-results: true