prepare-commit-msg

#!/usr/bin/env python 
#
# An example hook script to prepare the commit log message.
# Called by "git commit" with the name of the file that has the
# commit message, followed by the description of the commit
# message's source.  The hook's purpose is to edit the commit
# message file.  If the hook fails with a non-zero status,
# the commit is aborted.
#
# To enable this hook, rename this file to "prepare-commit-msg".

# This hook includes three examples.  The first comments out the
# "Conflicts:" part of a merge commit.
#
# The second includes the output of "git diff --name-status -r"
# into the message, just before the "git status" output.  It is
# commented because it doesn't cope with --amend or with squashed
# commits.
#
# The third example adds a Signed-off-by line to the message, that can
# still be edited.  This is rarely a good idea.

import binascii
import random
import os
import re
import string
import subprocess
import sys
import traceback
from collections import Counter


# default open function(python3)
open = open
# using python2 codecs.open func if python version is 2
if sys.version_info.major == 2:
    import codecs
    open = codecs.open

def main():
    try:
        if not auto_commit_msg_enabled():
            return
        commit_msg_file = sys.argv[1]
        changed_folders, changed_paths = process_metadata(commit_msg_file)
        scope = get_scope(changed_folders, changed_paths)
        type = get_type(changed_folders, changed_paths)
        tags = get_tags(changed_folders, changed_paths)
        jiras = get_jiras()
        change_id = get_change_id()
        write_commit_msg(commit_msg_file, type, scope, tags, jiras)
    except Exception as e:
        print("get exception while prepare commit msg: %s" % e)
        traceback.print_exc()


def auto_commit_msg_enabled():
    if len(sys.argv) >= 3 and sys.argv[2] != "template":
        # NOTE(weiw): it's maybe cherry pick or amend or something
        return False
    bashCommand = "git config --get zstack.autoCommitMsg"
    process = subprocess.Popen(bashCommand.split(), stdout=subprocess.PIPE)
    output, error = process.communicate()
    output = output.decode()
    if error != None:
        print("get error while running git config --get zstack.autoCommitMsg: %s" % error)
        return False
    elif "false" in output.lower():
        return False
    return True

def write_commit_msg(commit_msg_filepath, type, scope, tags, jiras):
    template = '''# Possible types: fix/feature/test/refactor/chore
# Possible tags: APIImpact/GlobalConfigImpact/GlobalPropertyImpact/DBImpact/ZQLImpact
# Possible jira-footers: Resolves/Related
# Please describe the commit as detailed as possible!
# 1. Why is this change necessary?
# 2. How does it address the problem?
# 3. Are there any side effects?\n
'''

    with open(commit_msg_filepath, 'r+', encoding='utf8') as msg:
        in_usable_content = False
        useable_contents = []
        for line in msg.readlines():
            if line.startswith("Resolves/Related: ZSTAC"):
                in_usable_content = True
                continue
            if line.startswith("# ------------------------ >8 ------------------------"):
                in_usable_content = True
                # NOTE(weiw): do not continue, it's usable!
            if in_usable_content is False:
                continue
            useable_contents.append(line)

        real_commit_msg = []
        real_commit_msg.append("<%s>[%s]: <description>\n\n" % (type, scope))
        real_commit_msg.append(template)
        if len(tags) > 0:
            real_commit_msg.append("%s\n" % "\n".join(tags))

        real_commit_msg.append("\n")
        if jiras:
            for jira in jiras:
                real_commit_msg.append("Resolves: %s\n" % jira.upper())
        else:
            real_commit_msg.append("Resolves: ZSTAC-XXXX\n")
        
        real_commit_msg.append("\nChange-Id: %s\n" % get_change_id())

    with open(commit_msg_filepath, 'w', encoding='utf8') as msg:
        real_commit_msg.extend(useable_contents)
        msg.writelines(real_commit_msg)


def process_metadata(commit_msg_filepath):
    # @return: 
    # changed_folders:  list of changed first level folder, 
    #                   if root path of code repo, then return root. 
    #                   eg. ["root", "conf", "network", "root"]
    # changed_paths:    dict of paths of changed files with its changed type, 
    #                   eg. {".gitmessage":"new file", "conf/web.xml":"modified"}
    #                   !! changed type not support yet !!

    changed_folders = []
    changed_paths = {}
    with open(commit_msg_filepath, 'r+', encoding='utf8') as msg:
        content = msg.readlines()
        for no, line in enumerate(content):
            line = line.strip()
            if not line.startswith("diff --git a/"):
                continue

            path = line.split(" ")[2][2:]
            changed_folders.append(get_top_folder_from_path(path))
            changed_paths[path] = 'new file' if content[no+1].startswith('new file') else 'modified'

    if changed_folders != []:
        return changed_folders, changed_paths

    bashCommand = "git diff HEAD"
    process = subprocess.Popen(bashCommand.split(), stdout=subprocess.PIPE)
    output, error = process.communicate()
    content = output.decode()
    if error != None:
        raise Exception(error)

    for no, line in enumerate(content.splitlines()):
        line = line.strip()
        if not line.startswith("diff --git a/"):
            continue

        path = line.split(" ")[2][2:]
        changed_folders.append(get_top_folder_from_path(path))
        changed_paths[path] = 'new file' if content[no+1].startswith('new file') else 'modified'
    return changed_folders, changed_paths

def get_top_folder_from_path(filepath):
    if "/" not in filepath and "\\" not in filepath:
        return "root"
    elif "/" in filepath:
        return filepath.split("/")[0]
    else:
        return filepath.split("\\")[0]

def get_scope(changed_folders, changed_paths):
    folder_counter = Counter(changed_folders)
    most_common = folder_counter.most_common(1)[0][0]
    if most_common not in ["plugin", "plugin-premium", "test", "test-premium", 
                           "sdk", "header", "core", "doc", "conf"]:
        return most_common

    scopes = []
    for p in changed_paths.keys():
        if "/org/zstack/sdk/" in p:
            word_after_sdk = p.split("org/zstack/sdk/")[-1].split("/")[0]
            if "." in word_after_sdk:
                # eg. sdk/src/main/java/org/zstack/sdk/CreateSchedulerJobGroupAction.java
                scopes.append("sdk")
            else:
                # eg. sdk/src/main/java/org/zstack/sdk/iam2/api/AddRolesToIAM2VirtualIDGroupAction.java
                scopes.append(word_after_sdk)
        elif "/org/zstack/header/" in p:
            # eg. header/src/main/java/org/zstack/header/image/APIAddImageMsg.java
            scopes.append(p.split("org/zstack/header/")[-1].split("/")[0])
        elif "/org/zstack/core/" in p:
            word_after_core = p.split("org/zstack/core/")[-1].split("/")[0]
            if "." in word_after_core:
                # eg. core/src/main/java/org/zstack/core/rest/RESTFacadeImpl.java
                scopes.append("core")
            else:
                # eg. core/src/main/java/org/zstack/core/Platform.java
                scopes.append(word_after_core)
        elif "plugin/" in p or "plugin-premium/" in p:
            # eg. plugin/kvm/src/main/java/org/zstack/kvm/KVMConstant.java
            # eg. plugin-premium/ticket/src/main/java/org/zstack/ticket/TicketBase.java
            if "/" in p:
                scopes.append(p.split("/")[1])
            else:
                scopes.append("root")
        elif "test/" in p or "test-premium/" in p or "doc/" in p:
            # eg. test-premium/src/test/groovy/org/zstack/test/integration/guesttools/GuestToolsCase.groovy
            # eg. doc/globalconfig/accessControl/enable.request.source.ip.address.check.md
            scope = p.split("/")[-2]
            if scope in ["test", "test-premium", "resources", "groovy"]:
                # eg. test/pom.xml
                # eg. test/src/test/resources/zstack.properties
                # eg. test/src/test/groovy/Test5.groovy
                scope = "test"
            scopes.append(scope)
        elif "conf/" in p:
            # eg. ./conf/errorCodes/hostAllocator.xml
            # eg. ./conf/errorCodes/hybrid/aliyun.xml
            # eg. ./conf/globalConfig/encrypt.xml
            # eg. ./conf/springConfigXml/hybrid/ecsimage.xml
            found_magic = False
            conf_magics = ["errorCodes/", "globalConfig/", "serviceConfig/", "springConfigXml/", "simulatorSpringConfigXml/"]
            for magic in conf_magics:
                if magic not in p:
                    continue

                word_after_magic = p.split(magic)[1]
                if "/" in word_after_magic:
                    scopes.append(word_after_magic.split("/")[0])
                else:
                    scopes.append(word_after_magic.split(".")[0])
                found_magic = True
                break
            if not found_magic:
                scopes.append("conf")
        elif "pom.xml" in p:
            # eg. header/pom.xml
            # eg. pom.xml
            scopes.append("dependencies")

    scope_counter = Counter(scopes)
    most_common = scope_counter.most_common(1)[0][0]

    return most_common
        
def get_type(changed_folders, changed_paths):
    if set(changed_folders).issubset(set(["test", "testlib", "test-premium", "testlib-premium"])):
        return "test"
    elif set(changed_folders).issubset(set(["doc"])):
        return "doc"

    new_file = 0
    modified = 0
    doc = 0

    for p, t in changed_paths.items():
        if t == "new file":
            new_file += 1
        else:
            modified += 1
        if "Doc_zh_cn" in p or p.endswith(".md") or p.startswith("doc/"):
            doc += 1

    if doc == len(changed_paths):
        return "doc"
    if new_file >= modified:
        return "feature"
    else:
        return "fix"

def get_tags(changed_folders, changed_paths):
    tags = set()
    for p in changed_paths.keys():
        file_name = os.path.split(p)[1]
        if file_name.endswith("Msg.java") and file_name.startswith("API"):
            tags.add("APIImpact")
        elif file_name.endswith("GlobalConfig.java"):
            tags.add("GlobalConfigImpact")
        elif file_name.endswith("GlobalProperty.java"):
            tags.add("GlobalPropertyImpact")
        elif file_name.endswith(".sql") or file_name.endswith("VO.java") or \
                file_name.endswith("VO_.java") or file_name.endswith("EO.java") or \
                file_name.endswith("AO.java") or file_name.endswith("AO_.java"):
            tags.add("DBImpact")
        elif "zql" in p:
            tags.add("ZQLImpact")
    
    return tags

def get_jiras():
    jiras = []
    jira_patterns = [r"\bZSTAC-\d+\b", r"\bZSTACK-\d+\b", r"\bMINI-\d+\b", 
                     r"\bZOPS-\d+\b", r"\bZHCI-\d+\b", r"\bZSV-\d+\b"]

    bashCommand = "git rev-parse --abbrev-ref HEAD"
    process = subprocess.Popen(bashCommand.split(), stdout=subprocess.PIPE)
    output, error = process.communicate()
    output = output.decode()
    if error != None:
        print("get error while reading current branch: %s" % error)
        return jiras
    for pattern in jira_patterns:
        searchObj = re.search(pattern, output, re.I)
        if searchObj:
            jiras.append(searchObj.group())
    return jiras

def get_change_id():
    letters = string.ascii_lowercase
    data = ''.join(random.choice(letters) for i in range(20)).encode()
    return("I{}".format(binascii.hexlify(data).decode('iso8859-1')))

def get_git_root_path():
    bashCommand = "git rev-parse --show-toplevel"
    process = subprocess.Popen(bashCommand.split(), stdout=subprocess.PIPE)
    output, error = process.communicate()
    output = output.decode()
    if error != None:
        raise Exception(error)
    return output

if __name__ == "__main__":
    main()

#case "$2,$3" in
#  merge,)
#    /usr/bin/perl -i.bak -ne 's/^/# /, s/^# #/#/ if /^Conflicts/ .. /#/; print' "$1" ;;
#
## ,|template,)
##   /usr/bin/perl -i.bak -pe '
##      print "\n" . `git diff --cached --name-status -r`
##      if /^#/ && $first++ == 0' "$1" ;;
#
#  *) ;;
#esac

# SOB=$(git var GIT_AUTHOR_IDENT | sed -n 's/^\(.*>\).*$/Signed-off-by: \1/p')
# grep -qs "^$SOB" "$1" || echo "$SOB" >> "$1"