app/tasks/importtasks.py

   1 import flask, json
   2 from flask.ext.sqlalchemy import SQLAlchemy
   3 import urllib.request
   4 from urllib.parse import urlparse, quote_plus
   5 from app import app
   6 from app.models import *
   7 from app.tasks import celery, TaskError
   8
   9 class GithubURLMaker:
  10         def __init__(self, url):
  11                 # Rewrite path
  12                 import re
  13                 m = re.search("^\/([^\/]+)\/([^\/]+)\/?$", url.path)
  14                 if m is None:
  15                         return
  16
  17                 user = m.group(1)
  18                 repo = m.group(2)
  19                 self.baseUrl = "https://raw.githubusercontent.com/{}/{}/master" \
  20                                 .format(user, repo.replace(".git", ""))
  21                 self.user = user
  22                 self.repo = repo
  23
  24         def isValid(self):
  25                 return self.baseUrl is not None
  26
  27         def getRepoURL(self):
  28                 return "https://github.com/{}/{}".format(self.user, self.repo)
  29
  30         def getIssueTrackerURL(self):
  31                 return "https://github.com/{}/{}/issues/".format(self.user, self.repo)
  32
  33         def getModConfURL(self):
  34                 return self.baseUrl + "/mod.conf"
  35
  36         def getDescURL(self):
  37                 return self.baseUrl + "/description.txt"
  38
  39         def getScreenshotURL(self):
  40                 return self.baseUrl + "/placeholder.png"
  41
  42         def getCommitsURL(self, branch):
  43                 return "https://api.github.com/repos/{}/{}/commits?sha={}" \
  44                                 .format(self.user, self.repo, urllib.parse.quote_plus(branch))
  45
  46         def getCommitDownload(self, commit):
  47                 return "https://github.com/{}/{}/archive/{}.zip" \
  48                                 .format(self.user, self.repo, commit)
  49
  50
  51 krock_list_cache = None
  52 krock_list_cache_by_name = None
  53 def getKrockList():
  54         global krock_list_cache
  55         global krock_list_cache_by_name
  56
  57         if krock_list_cache is None:
  58                 contents = urllib.request.urlopen("http://krock-works.16mb.com/MTstuff/modList.php").read().decode("utf-8")
  59                 list = json.loads(contents)
  60
  61                 def h(x):
  62                         if not ("title"   in x and "author" in x and \
  63                                         "topicId" in x and "link"   in x and x["link"] != ""):
  64                                 return False
  65
  66                         import re
  67                         m = re.search("\[([A-Za-z0-9_]+)\]", x["title"])
  68                         if m is None:
  69                                 return False
  70
  71                         x["name"] = m.group(1)
  72                         return True
  73
  74                 def g(x):
  75                         return {
  76                                 "title":   x["title"],
  77                                 "author":  x["author"],
  78                                 "name":    x["name"],
  79                                 "topicId": x["topicId"],
  80                                 "link":    x["link"],
  81                         }
  82
  83                 krock_list_cache = [g(x) for x in list if h(x)]
  84                 krock_list_cache_by_name = {}
  85                 for x in krock_list_cache:
  86                         if not x["name"] in krock_list_cache_by_name:
  87                                 krock_list_cache_by_name[x["name"]] = []
  88
  89                         krock_list_cache_by_name[x["name"]].append(x)
  90
  91         return krock_list_cache, krock_list_cache_by_name
  92
  93 def findModInfo(author, name, link):
  94         list, lookup = getKrockList()
  95
  96         if name is not None and name in lookup:
  97                 if len(lookup[name]) == 1:
  98                         return lookup[name][0]
  99
 100                 for x in lookup[name]:
 101                         if x["author"] == author:
 102                                 return x
 103
 104         if link is not None and len(link) > 15:
 105                 for x in list:
 106                         if link in x["link"]:
 107                                 return x
 108
 109         return None
 110
 111
 112 def parseConf(string):
 113         retval = {}
 114         for line in string.split("\n"):
 115                 idx = line.find("=")
 116                 if idx > 0:
 117                         key   = line[:idx-1].strip()
 118                         value = line[idx+1:].strip()
 119                         retval[key] = value
 120
 121         return retval
 122
 123
 124 @celery.task()
 125 def getMeta(urlstr, author):
 126         url = urlparse(urlstr)
 127
 128         urlmaker = None
 129         if url.netloc == "github.com":
 130                 urlmaker = GithubURLMaker(url)
 131         else:
 132                 raise TaskError("Unsupported repo")
 133
 134         if not urlmaker.isValid():
 135                 raise TaskError("Error! Url maker not valid")
 136
 137         result = {}
 138
 139         result["repo"] = urlmaker.getRepoURL()
 140         result["issueTracker"] = urlmaker.getIssueTrackerURL()
 141
 142         try:
 143                 contents = urllib.request.urlopen(urlmaker.getModConfURL()).read().decode("utf-8")
 144                 conf = parseConf(contents)
 145                 for key in ["name", "description", "title"]:
 146                         try:
 147                                 result[key] = conf[key]
 148                         except KeyError:
 149                                 pass
 150         except OSError:
 151                 print("mod.conf does not exist")
 152
 153         if "name" in result:
 154                 result["title"] = result["name"].replace("_", " ").title()
 155
 156         if not "description" in result:
 157                 try:
 158                         contents = urllib.request.urlopen(urlmaker.getDescURL()).read().decode("utf-8")
 159                         result["description"] = contents.strip()
 160                 except OSError:
 161                         print("description.txt does not exist!")
 162
 163         if "description" in result:
 164                 desc = result["description"]
 165                 idx = desc.find(".") + 1
 166                 cutIdx = min(len(desc), 200 if idx < 5 else idx)
 167                 result["short_description"] = desc[:cutIdx]
 168
 169         info = findModInfo(author, result.get("name"), result["repo"])
 170         if info is not None:
 171                 result["forumId"] = info.get("topicId")
 172
 173         return result
 174
 175 @celery.task()
 176 def makeVCSRelease(id, branch):
 177         release = PackageRelease.query.get(id)
 178
 179         if release is None:
 180                 raise TaskError("No such release!")
 181
 182         if release.package is None:
 183                 raise TaskError("No package attached to release")
 184
 185         url = urlparse(release.package.repo)
 186
 187         urlmaker = None
 188         if url.netloc == "github.com":
 189                 urlmaker = GithubURLMaker(url)
 190         else:
 191                 raise TaskError("Unsupported repo")
 192
 193         if not urlmaker.isValid():
 194                 raise TaskError("Invalid github repo URL")
 195
 196         contents = urllib.request.urlopen(urlmaker.getCommitsURL(branch)).read().decode("utf-8")
 197         commits = json.loads(contents)
 198
 199         if len(commits) == 0:
 200                 raise TaskError("No commits found")
 201
 202         release.url = urlmaker.getCommitDownload(commits[0]["sha"])
 203         release.task_id = None
 204         db.session.commit()
 205
 206         return release.url