get_pr_ut.py 8.1 KB
Newer Older
C
chalsliu 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
""" For the PR that only modified the unit test, get cases in pull request. """

import os
import json
18 19 20
import re
import sys
import requests
C
chalsliu 已提交
21 22 23 24 25 26 27 28 29 30 31
from github import Github

PADDLE_ROOT = os.getenv('PADDLE_ROOT', '/paddle/')


class PRChecker(object):
    """ PR Checker. """

    def __init__(self):
        self.github = Github(os.getenv('GITHUB_API_TOKEN'), timeout=60)
        self.repo = self.github.get_repo('PaddlePaddle/Paddle')
32 33 34 35 36 37
        self.py_prog_oneline = re.compile('\d+\|\s*#.*')
        self.py_prog_multiline_a = re.compile('\d+\|\s*""".*?"""', re.DOTALL)
        self.py_prog_multiline_b = re.compile("\d+\|\s*'''.*?'''", re.DOTALL)
        self.cc_prog_online = re.compile('\d+\|\s*//.*')
        self.cc_prog_multiline = re.compile('\d+\|\s*/\*.*?\*/', re.DOTALL)
        self.lineno_prog = re.compile('@@ \-\d+,\d+ \+(\d+),(\d+) @@')
C
chalsliu 已提交
38
        self.pr = None
39
        self.suffix = ''
C
chalsliu 已提交
40
        self.full_case = False
C
chalsliu 已提交
41 42 43 44 45 46 47

    def init(self):
        """ Get pull request. """
        pr_id = os.getenv('GIT_PR_ID')
        if not pr_id:
            print('No PR ID')
            exit(0)
48 49 50
        suffix = os.getenv('PREC_SUFFIX')
        if suffix:
            self.suffix = suffix
C
chalsliu 已提交
51
        self.pr = self.repo.get_pull(int(pr_id))
C
chalsliu 已提交
52 53 54 55 56 57 58 59 60 61 62
        last_commit = None
        ix = 0
        while True:
            commits = self.pr.get_commits().get_page(ix)
            for c in commits:
                last_commit = c.commit
            else:
                break
            ix = ix + 1
        if last_commit.message.find('test=full_case') != -1:
            self.full_case = True
C
chalsliu 已提交
63 64 65 66 67 68 69 70 71 72 73 74 75 76

    def get_pr_files(self):
        """ Get files in pull request. """
        page = 0
        file_list = []
        while True:
            files = self.pr.get_files().get_page(page)
            if not files:
                break
            for f in files:
                file_list.append(PADDLE_ROOT + f.filename)
            page += 1
        return file_list

77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168
    def __get_comment_by_filetype(self, content, filetype):
        result = []
        if filetype == 'py':
            result = self.__get_comment_by_prog(content, self.py_prog_oneline)
            result.extend(
                self.__get_comment_by_prog(content, self.py_prog_multiline_a))
            result.extend(
                self.__get_comment_by_prog(content, self.py_prog_multiline_b))
        if filetype == 'cc':
            result = self.__get_comment_by_prog(content, self.cc_prog_oneline)
            result.extend(
                self.__get_comment_by_prog(content, self.cc_prog_multiline))
        return result

    def __get_comment_by_prog(self, content, prog):
        result = []
        result_list = prog.findall(content)
        if not result_list:
            return None
        for u in result_list:
            result.extend(u.split('\n'))
        return result

    def get_comment_of_file(self, f):
        #content = self.repo.get_contents(f.replace(PADDLE_ROOT, ''), 'pull/').decoded_content
        with open(f) as fd:
            lines = fd.readlines()
        lineno = 1
        inputs = ''
        for line in lines:
            #for line in content.split('\n'):
            #input += str(lineno) + '|' + line + '\n'
            inputs += str(lineno) + '|' + line
            lineno += 1
        fietype = ''
        if f.endswith('.h') or f.endswith('.cc') or f.endswith('.cu'):
            filetype = 'cc'
        if f.endswith('.py'):
            filetype = 'py'
        else:
            return None
        return self.__get_comment_by_filetype(inputs, filetype)

    def get_pr_diff_lines(self):
        file_to_diff_lines = {}
        r = requests.get(self.pr.diff_url)
        data = r.text
        data = data.split('\n')
        ix = 0
        while ix < len(data):
            if data[ix].startswith('+++'):
                if data[ix].rstrip('\r\n') == '+++ /dev/null':
                    ix += 1
                    continue
                filename = data[ix][6:]
                ix += 1
                while ix < len(data):
                    result = self.lineno_prog.match(data[ix])
                    if not result:
                        break
                    lineno = int(result.group(1))
                    length = int(result.group(2))
                    ix += 1
                    end = ix + length
                    while ix < end:
                        if data[ix][0] == '-':
                            end += 1
                        if data[ix][0] == '+':
                            line_list = file_to_diff_lines.get(filename)
                            line = '{}{}'.format(lineno, data[ix].replace('+',
                                                                          '|'))
                            if line_list:
                                line_list.append(line)
                            else:
                                file_to_diff_lines[filename] = [line, ]
                        if data[ix][0] != '-':
                            lineno += 1
                        ix += 1
            ix += 1
        return file_to_diff_lines

    def is_only_comment(self, f):
        file_to_diff_lines = self.get_pr_diff_lines()
        comment_lines = self.get_comment_of_file(f)
        #for l in comment_lines:
        #    print(l)
        diff_lines = file_to_diff_lines.get(f.replace(PADDLE_ROOT, ''))
        for l in diff_lines:
            if l not in comment_lines:
                return False
        return True

C
chalsliu 已提交
169 170
    def get_pr_ut(self):
        """ Get unit tests in pull request. """
C
chalsliu 已提交
171 172
        if self.full_case:
            return ''
C
chalsliu 已提交
173
        check_added_ut = False
C
chalsliu 已提交
174 175
        ut_list = []
        file_ut_map = None
176
        cmd = 'wget -q --no-check-certificate https://sys-p0.bj.bcebos.com/prec/file_ut.json' + self.suffix
C
chalsliu 已提交
177
        os.system(cmd)
178
        with open('file_ut.json' + self.suffix) as jsonfile:
C
chalsliu 已提交
179 180
            file_ut_map = json.load(jsonfile)
        for f in self.get_pr_files():
C
chalsliu 已提交
181
            if f not in file_ut_map:
182 183 184 185 186 187 188 189 190 191 192 193 194 195
                if f.endswith('.md'):
                    ut_list.append('md_placeholder')
                elif f.endswith('.h') or f.endswith('.cu'):
                    if self.is_only_comment(f):
                        ut_list.append('h_cu_comment_placeholder')
                    else:
                        return ''
                elif f.endswith('.cc'):
                    if f.find('test_') != -1 or f.find('_test') != -1:
                        check_added_ut = True
                    elif self.is_only_comment(f):
                        ut_list.append('cc_comment_placeholder')
                    else:
                        return ''
C
chalsliu 已提交
196 197
                else:
                    return ''
C
chalsliu 已提交
198
            else:
199 200 201 202
                if self.is_only_comment(f):
                    ut_list.append('cc_comment_placeholder')
                else:
                    ut_list.extend(file_ut_map.get(f))
C
chalsliu 已提交
203
        ut_list = list(set(ut_list))
204
        cmd = 'wget -q --no-check-certificate https://sys-p0.bj.bcebos.com/prec/prec_delta' + self.suffix
C
chalsliu 已提交
205
        os.system(cmd)
206
        with open('prec_delta' + self.suffix) as delta:
C
chalsliu 已提交
207 208 209
            for ut in delta:
                ut_list.append(ut.rstrip('\r\n'))

C
chalsliu 已提交
210
        if check_added_ut:
C
chalsliu 已提交
211 212
            cmd = 'bash {}/tools/check_added_ut.sh >/tmp/pre_ut 2>&1'.format(
                PADDLE_ROOT)
C
chalsliu 已提交
213
            os.system(cmd)
C
chalsliu 已提交
214 215 216
            with open('{}/added_ut'.format(PADDLE_ROOT)) as utfile:
                for ut in utfile:
                    ut_list.append(ut.rstrip('\r\n'))
C
chalsliu 已提交
217

C
chalsliu 已提交
218 219 220 221 222 223 224
        return ' '.join(ut_list)


if __name__ == '__main__':
    pr_checker = PRChecker()
    pr_checker.init()
    print(pr_checker.get_pr_ut())