view gpp/legacy/management/commands/import_old_downloads.py @ 467:b910cc1460c8

Add the ability to conditionally add model instances to the search index on update. This is not perfect, as some instances should be deleted from the index if they are updated such that they should not be in the index anymore. Will think about and address that later.
author Brian Neal <bgneal@gmail.com>
date Sun, 24 Jul 2011 18:12:20 +0000
parents 639cfdf59167
children
line wrap: on
line source
"""
import_old_downloads.py - For importing downloads from SG101 1.0 as csv files.
"""
from __future__ import with_statement
import csv
import datetime

from django.core.management.base import LabelCommand, CommandError
from django.contrib.auth.models import User

from downloads.models import Download, Category
from legacy.html2md import MarkdownWriter


# downloads with these lid's will be excluded
EXCLUDE_SET = set([1, 2, 3, 4, 277])

# Mapping of old category IDs to new; None means we don't plan on importing
CAT_MAP = {
  4: None,      # Misc
  3: None,      # Music
  1: None,      # Demos
  6: 2,         # Gear Samples
  8: 4,         # Ringtones
  9: 8,         # Tablature
  10: 6,        # Interviews
  11: None,     # 2008 MP3 Comp
  12: 1,        # Backing Tracks
  13: None,     # 2009 MP3 Comp
}

SG101_PREFIX = 'http://surfguitar101.com/'


class Command(LabelCommand):
    args = '<filename filename ...>'
    help = 'Imports downloads from the old database in CSV format'
    md_writer = MarkdownWriter()

    def handle_label(self, filename, **options):
        """
        Process each line in the CSV file given by filename by
        creating a new object and saving it to the database.

        """
        self.cats = {}
        try:
            self.default_user = User.objects.get(pk=2)
        except User.DoesNotExist:
            raise CommandError("Need a default user with pk=2")

        try:
            with open(filename, "rb") as f:
                self.reader = csv.DictReader(f)
                try:
                    for row in self.reader:
                        self.process_row(row)
                except csv.Error, e:
                    raise CommandError("CSV error: %s %s %s" % (
                        filename, self.reader.line_num, e))

        except IOError:
            raise CommandError("Could not open file: %s" % filename)

    def get_category(self, old_cat_id):
        """
        Return the Category object for the row.

        """
        cat_id = CAT_MAP[old_cat_id]
        if cat_id not in self.cats:
            try:
                cat = Category.objects.get(pk=cat_id)
            except Category.DoesNotExist:
                raise CommandError("Category does not exist: %s on line %s" % (
                    cat_id, self.reader.line_num))
            else:
                self.cats[cat_id] = cat
        return self.cats[cat_id]

    def get_user(self, username):
        """
        Return the user object for the given username.
        If the user cannot be found, self.default_user is returned.

        """
        try:
            return User.objects.get(username=username)
        except User.DoesNotExist:
            return self.default_user

    def process_row(self, row):
        """
        Process one row from the CSV file: create an object for the row
        and save it in the database.

        """
        lid = int(row['lid'])
        if lid in EXCLUDE_SET:
            return  # skip

        cat = int(row['cid'])
        if CAT_MAP.get(cat) is None:
            return  # skip this one; we aren't carrying these over

        dl_date = datetime.datetime.strptime(row['date'], "%Y-%m-%d %H:%M:%S")
        old_url = row['url'].decode('latin-1')
        if old_url.startswith(SG101_PREFIX):
            old_url = old_url[len(SG101_PREFIX):]
        if old_url.startswith('dls/'):
            old_url = old_url[4:]
        new_url = u'downloads/1.0/%s' % old_url

        dl = Download(
            id=lid,
            title=row['title'].decode('latin-1'),
            category=self.get_category(cat),
            description=self.to_markdown(row['description'].decode('latin-1')),
            file=new_url,
            user=self.get_user(row['submitter']),
            date_added=dl_date,
            ip_address='127.0.0.1',  # not available
            hits=int(row['hits']),
            average_score=float(row['downloadratingsummary']) / 2.0,
            total_votes=int(row['totalvotes']),
            is_public=True)
        dl.save()
        #print "cp %s %s" % (old_url, '/home/var/django-sites/sg101/sg101-trunk/media/' + new_url)

    def to_markdown(self, s):
        self.md_writer.reset()
        self.md_writer.feed(s)
        return self.md_writer.markdown()