From 41886adc4962a2aea572be1f71297c8a01bf99fb Mon Sep 17 00:00:00 2001 From: Radek Czajka Date: Thu, 15 Dec 2022 11:47:56 +0100 Subject: [PATCH] stats --- src/catalogue/admin.py | 3 + .../migrations/0042_bookmonthlystats.py | 24 +++++++ ...ly_views_page_book_monthly_views_reader.py | 23 +++++++ src/catalogue/models.py | 68 +++++++++++++++++++ 4 files changed, 118 insertions(+) create mode 100644 src/catalogue/migrations/0042_bookmonthlystats.py create mode 100644 src/catalogue/migrations/0043_book_monthly_views_page_book_monthly_views_reader.py diff --git a/src/catalogue/admin.py b/src/catalogue/admin.py index 2cffc743..e0c4242b 100644 --- a/src/catalogue/admin.py +++ b/src/catalogue/admin.py @@ -190,6 +190,8 @@ class BookAdmin(WikidataAdminMixin, NumericFilterModelAdmin): "estimated_costs", "documents_book_link", "scans_source_link", + "monthly_views_page", + "monthly_views_reader", ] actions = [export_as_csv_action( fields=[ @@ -253,6 +255,7 @@ class BookAdmin(WikidataAdminMixin, NumericFilterModelAdmin): "notes", ("estimated_chars", "estimated_verses", "estimate_source"), "estimated_costs", + ("monthly_views_page", "monthly_views_reader"), ] }, ), diff --git a/src/catalogue/migrations/0042_bookmonthlystats.py b/src/catalogue/migrations/0042_bookmonthlystats.py new file mode 100644 index 00000000..25030b86 --- /dev/null +++ b/src/catalogue/migrations/0042_bookmonthlystats.py @@ -0,0 +1,24 @@ +# Generated by Django 4.0.6 on 2022-12-15 11:05 + +from django.db import migrations, models +import django.db.models.deletion + + +class Migration(migrations.Migration): + + dependencies = [ + ('catalogue', '0041_author_genitive_epoch_adjective_feminine_singular_and_more'), + ] + + operations = [ + migrations.CreateModel( + name='BookMonthlyStats', + fields=[ + ('id', models.AutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')), + ('date', models.DateField()), + ('views_reader', models.IntegerField(default=0)), + ('views_page', models.IntegerField(default=0)), + ('book', models.ForeignKey(on_delete=django.db.models.deletion.CASCADE, to='catalogue.book')), + ], + ), + ] diff --git a/src/catalogue/migrations/0043_book_monthly_views_page_book_monthly_views_reader.py b/src/catalogue/migrations/0043_book_monthly_views_page_book_monthly_views_reader.py new file mode 100644 index 00000000..f8ab704b --- /dev/null +++ b/src/catalogue/migrations/0043_book_monthly_views_page_book_monthly_views_reader.py @@ -0,0 +1,23 @@ +# Generated by Django 4.0.6 on 2022-12-15 11:43 + +from django.db import migrations, models + + +class Migration(migrations.Migration): + + dependencies = [ + ('catalogue', '0042_bookmonthlystats'), + ] + + operations = [ + migrations.AddField( + model_name='book', + name='monthly_views_page', + field=models.IntegerField(default=0), + ), + migrations.AddField( + model_name='book', + name='monthly_views_reader', + field=models.IntegerField(default=0), + ), + ] diff --git a/src/catalogue/models.py b/src/catalogue/models.py index 97454775..1eb497f0 100644 --- a/src/catalogue/models.py +++ b/src/catalogue/models.py @@ -1,6 +1,10 @@ from collections import Counter +from datetime import date, timedelta import decimal +import re +from urllib.request import urlopen from django.apps import apps +from django.conf import settings from django.db import models from django.template.loader import render_to_string from django.urls import reverse @@ -221,6 +225,9 @@ class Book(WikidataModel): free_license = models.BooleanField(_('free license'), default=False) polona_missing = models.BooleanField(_('missing on Polona'), default=False) + monthly_views_reader = models.IntegerField(default=0) + monthly_views_page = models.IntegerField(default=0) + class Meta: ordering = ("title",) verbose_name = _('book') @@ -286,6 +293,28 @@ class Book(WikidataModel): for work_type in WorkType.objects.all() } + def update_monthly_stats(self): + # Find publication date. + # By default, get previous 12 months. + this_month = date.today().replace(day=1) + cutoff = this_month.replace(year=this_month.year - 1) + months = 12 + + # If the book was published later, + # find out the denominator. + pbr = apps.get_model('documents', 'BookPublishRecord').objects.filter( + book__catalogue_book=self).order_by('timestamp').first() + if pbr is not None and pbr.timestamp.date() > cutoff: + months = (this_month - pbr.timestamp.date()).days / 365 * 12 + + stats = self.bookmonthlystats_set.filter(date__gte=cutoff).aggregate( + views_page=models.Sum('views_page'), + views_reader=models.Sum('views_reader') + ) + self.monthly_views_page = stats['views_page'] / months + self.monthly_views_reader = stats['views_reader'] / months + self.save(update_fields=['monthly_views_page', 'monthly_views_reader']) + class CollectionCategory(models.Model): name = models.CharField(_("name"), max_length=255) @@ -404,3 +433,42 @@ class Place(WikidataModel): def __str__(self): return self.name + + +class BookMonthlyStats(models.Model): + book = models.ForeignKey('catalogue.Book', models.CASCADE) + date = models.DateField() + views_reader = models.IntegerField(default=0) + views_page = models.IntegerField(default=0) + + @classmethod + def build_for_month(cls, date): + date = date.replace(day=1) + period = 'month' + + date = date.isoformat() + url = f'{settings.PIWIK_URL}?date={date}&filter_limit=-1&format=CSV&idSite={settings.PIWIK_WL_SITE_ID}&language=pl&method=Actions.getPageUrls&module=API&period={period}&segment=&token_auth={settings.PIWIK_TOKEN}&flat=1' + data = urlopen(url).read().decode('utf-16') + lines = data.split('\n')[1:] + for line in lines: + m = re.match('^/katalog/lektura/([^,./]+)\.html,', line) + if m is not None: + which = 'views_reader' + else: + m = re.match('^/katalog/lektura/([^,./]+)/,', line) + if m is not None: + which = 'views_page' + if m is not None: + slug = m.group(1) + _url, _uviews, views, _rest = line.split(',', 3) + views = int(views) + try: + book = Book.objects.get(slug=slug) + except Book.DoesNotExist: + continue + else: + cls.objects.update_or_create( + book=book, date=date, + defaults={which: views} + ) + book.update_monthly_stats() -- 2.20.1