import os
import re
import shutil
import urllib2
import urlparse
from datetime import datetime
from download_content_yes_to_lic import LicenseProtectedFileFetcher

# Snapshots base URL
snapshots_url = "http://snapshots.linaro.org"
# Current default distribution
distribution = "ubuntu"

def geturl(path):
    try:
        fd = urllib2.urlopen(path)
    except urllib2.HTTPError:
        return ""
    data = fd.read()
    fd.close()
    return data

def download(url, path=""):
    urlpath = urlparse.urlsplit(url).path
    filename = os.path.basename(urlpath)
    if path:
        filename = os.path.join(path,filename)
    fd = open(filename, "w")
    try:
        response = urllib2.urlopen(urllib2.quote(url, safe=":/"))
        fd = open(filename, 'wb')
        shutil.copyfileobj(response,fd,0x10000)
        fd.close()
        response.close()
    except:
        raise RuntimeError("Could not retrieve %s" % url)
    return filename


def find_builds(data):
    found = []
    pattern = re.compile(
        r'<td>\s*<a href="(/.*/)?\d+/?"[\n ]*>(?P<buildstamp>\d+)/?</a>',
         flags=re.MULTILINE)
    for search in pattern.finditer(data):
        found.append(search.group('buildstamp'))
    return found

def _get_filename(filelist):
    match = None
    tgzpattern = re.compile(
        r'<a href=".*"[^>]*>(?P<tgzfilename>[\w\-]+(?!\.config)\.tar\.gz)</a>',
        flags=re.MULTILINE)
    for match in tgzpattern.finditer(filelist):
        pass
    if match is None:
        # A file may not be found if the build failed
        return None
    return match.group('tgzfilename')

def find_latest(url):
    """Find the latest image at the specified url

    :param url: The base url to search
    :param extra: The extra path needed to complete the url
    """
    fetcher = LicenseProtectedFileFetcher()
    builddates = fetcher.get(url)
    dates = find_builds(builddates)
    dates = [int(date) for date in dates]
    dates.sort()
    if not dates:
        raise ValueError("No builds found at index url: " + url)
    lastdate = dates.pop()
    lastdate = str(lastdate) + "/"
    latesturl = urlparse.urljoin(url, lastdate)
    content = fetcher.get(latesturl)
    filename = _get_filename(content)
    tgz_url = os.path.join(latesturl, filename)
    fetcher.close()
    return tgz_url

def find_latest_rootfs(rootfs):
    """Find the latest root fs image and return the build_id and url

    :param rootfs: The short name of the rootfs type.
    """
    rootfs_url= "%s/%s/images/%s/" % (snapshots_url, distribution, rootfs)
    filename = find_latest(rootfs_url)
    return filename

def find_latest_hwpack(hwpack):
    """Find the latest hardware pack and return the build_id and url

    :param hwpack: The name of the hardware pack to look for.
    """
    hwpack_url="%s/%s/hwpacks/%s/" % (snapshots_url, distribution, hwpack)
    filename = find_latest(hwpack_url)
    return filename

def find_ci_builds(data, latest_hwpack):

    found = False
    for line in data.splitlines():
        if latest_hwpack in line:
            found = True
            break
    return found

def find_ci_latest(url, latest_hwpack):
    """Find the latest image at the specified url

    :param url: The base url to search
    :param extra: The extra path needed to complete the url
    """
    filename = None
    fetcher = LicenseProtectedFileFetcher()
    builddates = fetcher.get(url)
    found = find_ci_builds(builddates, latest_hwpack)

    try:
        if found:
            filename = os.path.join(url, latest_hwpack)
    except Exception:
        pass
    fetcher.close()
    return filename