arch.py

# -*- coding: utf-8 -*-
import subprocess
import time
import datetime
import os
from .utils import helper
from .utils.db import downstream
from .utils.types import Repo, DownstreamRelease

MIRROR = "ftp.archlinux.org"
HTTP_START_DIR = None
FTP_START_DIR = None

ARCHES = ["i686","x86_64"]

distro_id = downstream.distro("arch", "", "A rolling release binary distribution.", "http://www.archlinx.org")

# return a list of ["ubuntu", branch, codename, component, arch, None, None]
def get_repos(test):
	repos = []
	for comp in ["core","extra"]:
		for a in ARCHES:
			repo = Repo()
			repo.distro_id = distro_id
			repo.component = comp
			repo.architecture = a
			repo.codename = ""
			repos.append(repo)
			downstream.repo(repo, test)
			downstream.add_branch(repo, "current", test)
	
	for comp in ["testing"]:
		for a in ARCHES:
			repo = Repo()
			repo.distro_id = distro_id
			repo.component = comp
			repo.architecture = a
			repo.codename = ""
			repos.append(repo)
			downstream.repo(repo, test)
			downstream.add_branch(repo, "future", test)
	
	for comp in ["community"]:
		for a in ARCHES:
			repo = Repo()
			repo.distro_id = distro_id
			repo.component = comp
			repo.architecture = a
			repo.codename = ""
			repos.append(repo)
			downstream.repo(repo, test)
			downstream.add_branch(repo, "experimental", test)
	
	return repos

# return a list of [name, version, revision, epoch, time, extra]
def crawl_repo(repo):
	rels = []
	fn = "".join(("files/arch/",repo.component,"-",str(time.time()),".db.tar.gz"))
	url = "".join(("http://",MIRROR,"/",repo.component,"/os/",repo.architecture,"/",repo.component,".db.tar.gz"))
	new_dir = "".join(("files/arch/",repo.component,"-",str(time.time()),"/"))
	os.mkdir(new_dir)
	t = helper.open_url(url,fn,repo.last_crawl)
	if t:
		try:
			#print " ".join(("/bin/tar","-xzf ",fn,"-C",new_dir))
			p = subprocess.Popen(("/bin/tar","-xzf",fn,"-C",new_dir),stdout=None)
			x = p.wait()
		except OSError, e:
			print e
			x=-1
		
		for d in os.listdir(new_dir):
			fn = "".join((new_dir,d,"/desc"))
			try:
				last = os.stat(fn)
			except OSError:
				print "no cache"
				last = None
			
			if last:
				last = last.st_mtime
				last = datetime.datetime.fromtimestamp(last)
			
			# ignore if its not new
			if last and repo.last_crawl!=None and last<repo.last_crawl:
				continue
			
			pkg = {}
			key = None
			for line in open(fn):
				if line[0]=="%":
					key = line.strip("\n%")
					pkg[key] = []
				elif line=="\n":
					if key:
						pkg[key] = "".join(pkg[key]).strip()
					key = None
				else:
					pkg[key].append(line)
			
			version, revision = pkg["VERSION"].rsplit("-",1)
			if " " in pkg["BUILDDATE"]:
				try:
					released = datetime.datetime.strptime(pkg["BUILDDATE"],"%a %b %d %H:%M:%S %Y")
				except:
					#print "ERROR: cannot parse",pkg["BUILDDATE"]
					pass
			else:
				released = datetime.datetime.fromtimestamp(long(pkg["BUILDDATE"]))
			
			rel = DownstreamRelease()
			rel.repo_id = repo.id
			rel.package = pkg["NAME"]
			rel.version = version
			rel.revision = revision
			rel.released = released
			rels.append(rel)
	if t == None:
		t = repo.last_crawl
	return (t, rels)