contentdb/app/tasks/importtasks.py

392 lines
10 KiB
Python
Raw Normal View History

2018-05-17 16:18:20 +02:00
# Content DB
# Copyright (C) 2018 rubenwardy
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <https://www.gnu.org/licenses/>.
import flask, json, os, git, tempfile, shutil, gitdb
from git import GitCommandError
from git_archive_all import GitArchiver
2018-12-23 00:03:38 +01:00
from flask_sqlalchemy import SQLAlchemy
2018-05-15 20:35:59 +02:00
from urllib.error import HTTPError
import urllib.request
from urllib.parse import urlparse, quote_plus, urlsplit
from app import app
from app.models import *
from app.tasks import celery, TaskError
2018-05-15 20:35:59 +02:00
from app.utils import randomString
2020-01-19 02:22:33 +01:00
from .minetestcheck import build_tree, MinetestCheckError, ContentType
from .minetestcheck.config import parse_conf
2018-06-05 20:59:07 +02:00
class GithubURLMaker:
def __init__(self, url):
2019-08-09 12:17:39 +02:00
self.baseUrl = None
self.user = None
self.repo = None
2018-06-05 20:59:07 +02:00
# Rewrite path
import re
m = re.search("^\/([^\/]+)\/([^\/]+)\/?$", url.path)
if m is None:
return
user = m.group(1)
repo = m.group(2).replace(".git", "")
self.baseUrl = "https://raw.githubusercontent.com/{}/{}/master" \
.format(user, repo)
self.user = user
self.repo = repo
def isValid(self):
return self.baseUrl is not None
def getRepoURL(self):
return "https://github.com/{}/{}".format(self.user, self.repo)
def getScreenshotURL(self):
return self.baseUrl + "/screenshot.png"
def getModConfURL(self):
return self.baseUrl + "/mod.conf"
2018-06-05 20:59:07 +02:00
def getCommitsURL(self, branch):
return "https://api.github.com/repos/{}/{}/commits?sha={}" \
.format(self.user, self.repo, urllib.parse.quote_plus(branch))
def getCommitDownload(self, commit):
return "https://github.com/{}/{}/archive/{}.zip" \
.format(self.user, self.repo, commit)
krock_list_cache = None
krock_list_cache_by_name = None
def getKrockList():
global krock_list_cache
global krock_list_cache_by_name
if krock_list_cache is None:
2018-12-21 15:02:57 +01:00
contents = urllib.request.urlopen("https://krock-works.uk.to/minetest/modList.php").read().decode("utf-8")
list = json.loads(contents)
def h(x):
if not ("title" in x and "author" in x and \
"topicId" in x and "link" in x and x["link"] != ""):
return False
import re
m = re.search("\[([A-Za-z0-9_]+)\]", x["title"])
if m is None:
return False
x["name"] = m.group(1)
return True
def g(x):
return {
"title": x["title"],
"author": x["author"],
2018-06-05 01:10:47 +02:00
"name": x["name"],
"topicId": x["topicId"],
2018-06-05 01:10:47 +02:00
"link": x["link"],
}
krock_list_cache = [g(x) for x in list if h(x)]
krock_list_cache_by_name = {}
for x in krock_list_cache:
if not x["name"] in krock_list_cache_by_name:
krock_list_cache_by_name[x["name"]] = []
krock_list_cache_by_name[x["name"]].append(x)
return krock_list_cache, krock_list_cache_by_name
def findModInfo(author, name, link):
2018-05-14 14:58:31 +02:00
list, lookup = getKrockList()
2018-05-14 14:58:31 +02:00
if name is not None and name in lookup:
if len(lookup[name]) == 1:
return lookup[name][0]
for x in lookup[name]:
if x["author"] == author:
return x
2018-05-14 14:58:31 +02:00
if link is not None and len(link) > 15:
for x in list:
if link in x["link"]:
return x
return None
def generateGitURL(urlstr):
scheme, netloc, path, query, frag = urlsplit(urlstr)
return "http://:@" + netloc + path + query
2018-06-05 01:10:47 +02:00
# Clones a repo from an unvalidated URL.
# Returns a tuple of path and repo on sucess.
# Throws `TaskError` on failure.
# Caller is responsible for deleting returned directory.
def cloneRepo(urlstr, ref=None, recursive=False):
2018-06-05 01:10:47 +02:00
gitDir = tempfile.gettempdir() + "/" + randomString(10)
err = None
try:
gitUrl = generateGitURL(urlstr)
print("Cloning from " + gitUrl)
if ref is None:
repo = git.Repo.clone_from(gitUrl, gitDir, \
progress=None, env=None, depth=1, recursive=recursive, kill_after_timeout=15)
else:
repo = git.Repo.clone_from(gitUrl, gitDir, \
progress=None, env=None, depth=1, recursive=recursive, kill_after_timeout=15, b=ref)
return gitDir, repo
except GitCommandError as e:
2018-06-05 23:34:57 +02:00
# This is needed to stop the backtrace being weird
err = e.stderr
except gitdb.exc.BadName as e:
err = "Unable to find the reference " + (ref or "?") + "\n" + e.stderr
raise TaskError(err.replace("stderr: ", "") \
.replace("Cloning into '" + gitDir + "'...", "") \
.strip())
2018-06-05 01:10:47 +02:00
2018-06-05 23:34:57 +02:00
@celery.task()
def getMeta(urlstr, author):
gitDir, _ = cloneRepo(urlstr, recursive=True)
2020-01-19 02:22:33 +01:00
try:
tree = build_tree(gitDir, author=author, repo=urlstr)
except MinetestCheckError as err:
raise TaskError(str(err))
shutil.rmtree(gitDir)
2018-06-05 01:10:47 +02:00
result = {}
result["name"] = tree.name
result["provides"] = tree.fold("name")
result["type"] = tree.type.name
2018-05-27 23:03:54 +02:00
2018-06-05 01:10:47 +02:00
for key in ["depends", "optional_depends"]:
result[key] = tree.fold("meta", key)
2018-05-27 23:03:54 +02:00
2018-06-05 01:10:47 +02:00
for key in ["title", "repo", "issueTracker", "forumId", "description", "short_description"]:
result[key] = tree.get(key)
2018-05-12 18:28:04 +02:00
2018-06-05 01:10:47 +02:00
for mod in result["provides"]:
result["depends"].discard(mod)
result["optional_depends"].discard(mod)
2018-05-27 23:03:54 +02:00
2018-06-05 01:10:47 +02:00
for key, value in result.items():
if isinstance(value, set):
result[key] = list(value)
return result
2018-05-11 16:04:17 +02:00
2018-05-15 20:35:59 +02:00
def makeVCSReleaseFromGithub(id, branch, release, url):
urlmaker = GithubURLMaker(url)
2018-05-11 16:04:17 +02:00
if not urlmaker.isValid():
raise TaskError("Invalid github repo URL")
2018-05-23 19:49:23 +02:00
commitsURL = urlmaker.getCommitsURL(branch)
try:
contents = urllib.request.urlopen(commitsURL).read().decode("utf-8")
commits = json.loads(contents)
except HTTPError:
raise TaskError("Unable to get commits for Github repository. Either the repository or reference doesn't exist.")
2018-05-11 16:04:17 +02:00
2018-05-23 19:49:23 +02:00
if len(commits) == 0 or not "sha" in commits[0]:
2018-05-11 16:04:17 +02:00
raise TaskError("No commits found")
2018-07-28 15:48:03 +02:00
release.url = urlmaker.getCommitDownload(commits[0]["sha"])
release.task_id = None
release.commit_hash = commits[0]["sha"]
2019-03-29 21:47:48 +01:00
release.approve(release.package.author)
2018-05-11 16:04:17 +02:00
db.session.commit()
return release.url
2018-05-15 20:35:59 +02:00
@celery.task()
def makeVCSRelease(id, branch):
release = PackageRelease.query.get(id)
if release is None:
raise TaskError("No such release!")
elif release.package is None:
raise TaskError("No package attached to release")
# url = urlparse(release.package.repo)
# if url.netloc == "github.com":
# return makeVCSReleaseFromGithub(id, branch, release, url)
gitDir, repo = cloneRepo(release.package.repo, ref=branch, recursive=True)
2020-01-19 02:22:33 +01:00
try:
tree = build_tree(gitDir, expected_type=ContentType[release.package.type.name], \
author=release.package.author.username, name=release.package.name)
except MinetestCheckError as err:
raise TaskError(str(err))
try:
filename = randomString(10) + ".zip"
destPath = os.path.join(app.config["UPLOAD_DIR"], filename)
assert(not os.path.isfile(destPath))
archiver = GitArchiver(force_sub=True, main_repo_abspath=gitDir)
archiver.create(destPath)
assert(os.path.isfile(destPath))
release.url = "/uploads/" + filename
release.task_id = None
release.commit_hash = repo.head.object.hexsha
release.approve(release.package.author)
print(release.url)
db.session.commit()
return release.url
finally:
shutil.rmtree(gitDir)
2018-05-15 20:35:59 +02:00
@celery.task()
def importRepoScreenshot(id):
package = Package.query.get(id)
2018-05-25 19:28:24 +02:00
if package is None or package.soft_deleted:
2018-05-15 20:35:59 +02:00
raise Exception("Unexpected none package")
# Get URL Maker
try:
gitDir, _ = cloneRepo(package.repo)
except TaskError as e:
# ignore download errors
print(e)
return None
2018-05-15 20:35:59 +02:00
2018-06-05 23:34:57 +02:00
# Find and import screenshot
2018-05-15 20:35:59 +02:00
try:
2018-06-05 23:34:57 +02:00
for ext in ["png", "jpg", "jpeg"]:
sourcePath = gitDir + "/screenshot." + ext
if os.path.isfile(sourcePath):
filename = randomString(10) + "." + ext
2020-01-18 02:20:32 +01:00
destPath = os.path.join(app.config["UPLOAD_DIR"], filename)
2018-06-05 23:34:57 +02:00
shutil.copyfile(sourcePath, destPath)
ss = PackageScreenshot()
ss.approved = True
ss.package = package
ss.title = "screenshot.png"
ss.url = "/uploads/" + filename
db.session.add(ss)
db.session.commit()
return "/uploads/" + filename
finally:
shutil.rmtree(gitDir)
print("screenshot.png does not exist")
2018-05-15 20:35:59 +02:00
return None
def getDepends(package):
url = urlparse(package.repo)
urlmaker = None
if url.netloc == "github.com":
urlmaker = GithubURLMaker(url)
else:
return {}
result = {}
if not urlmaker.isValid():
return {}
#
# Try getting depends on mod.conf
#
try:
contents = urllib.request.urlopen(urlmaker.getModConfURL()).read().decode("utf-8")
2020-01-19 02:22:33 +01:00
conf = parse_conf(contents)
for key in ["depends", "optional_depends"]:
try:
result[key] = conf[key]
except KeyError:
pass
except HTTPError:
print("mod.conf does not exist")
if "depends" in result or "optional_depends" in result:
return result
#
# Try depends.txt
#
import re
pattern = re.compile("^([a-z0-9_]+)\??$")
try:
contents = urllib.request.urlopen(urlmaker.getDependsURL()).read().decode("utf-8")
soft = []
hard = []
for line in contents.split("\n"):
line = line.strip()
if pattern.match(line):
if line[len(line) - 1] == "?":
soft.append( line[:-1])
else:
hard.append(line)
result["depends"] = ",".join(hard)
result["optional_depends"] = ",".join(soft)
except HTTPError:
print("depends.txt does not exist")
return result
def importDependencies(package, mpackage_cache):
if Dependency.query.filter_by(depender=package).count() != 0:
return
result = getDepends(package)
if "depends" in result:
deps = Dependency.SpecToList(package, result["depends"], mpackage_cache)
print("{} hard: {}".format(len(deps), result["depends"]))
for dep in deps:
dep.optional = False
db.session.add(dep)
if "optional_depends" in result:
deps = Dependency.SpecToList(package, result["optional_depends"], mpackage_cache)
print("{} soft: {}".format(len(deps), result["optional_depends"]))
for dep in deps:
dep.optional = True
db.session.add(dep)
@celery.task()
def importAllDependencies():
Dependency.query.delete()
mpackage_cache = {}
packages = Package.query.filter_by(type=PackageType.MOD).all()
for i, p in enumerate(packages):
print("============= {} ({}/{}) =============".format(p.name, i, len(packages)))
importDependencies(p, mpackage_cache)
db.session.commit()