iwla/plugins/display/referers.py

# -*- coding: utf-8 -*-
#
# Copyright Grégory Soutadé 2015

# This file is part of iwla

# iwla is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# iwla is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with iwla.  If not, see <http://www.gnu.org/licenses/>.
#

from iwla import IWLA
from iplugin import IPlugin
from display import *

#
# Display hook
# 
# Create Referers page
#
# Plugin requirements :
#     post_analysis/referers
#
# Conf values needed :
#     max_referers_displayed*
#     create_all_referers_page*
#     max_key_phrases_displayed*
#     create_all_key_phrases_page*
#
# Output files :
#     OUTPUT_ROOT/year/month/referers.html
#     OUTPUT_ROOT/year/month/key_phrases.html
#     OUTPUT_ROOT/year/month/index.html
#
# Statistics creation :
#     None
#
# Statistics update :
#     None
#
# Statistics deletion :
#     None
#

class IWLADisplayReferers(IPlugin):
    def __init__(self, iwla):
        super(IWLADisplayReferers, self).__init__(iwla)
        self.API_VERSION = 1
        self.requires = ['IWLAPostAnalysisReferers']
        self.max_referers = self.iwla.getConfValue('max_referers_displayed', 0)
        self.create_all_referers = self.iwla.getConfValue('create_all_referers_page', True)
        self.max_key_phrases = self.iwla.getConfValue('max_key_phrases_displayed', 0)
        self.create_all_key_phrases = self.iwla.getConfValue('create_all_key_phrases_page', True)

    def hook(self):
        display = self.iwla.getDisplay()
        month_stats = self.iwla.getMonthStats()
        referers = month_stats.get('referers', {})
        robots_referers = month_stats.get('robots_referers', {})
        search_engine_referers = month_stats.get('search_engine_referers', {})
        key_phrases = month_stats.get('key_phrases', {})

        top_referers = [(k, referers[k]['pages']) for k in referers.keys()]
        top_referers = sorted(top_referers, key=lambda t: t[1], reverse=True)

        top_robots_referers = [(k, robots_referers[k]['pages']) for k in robots_referers.keys()]
        top_robots_referers = sorted(top_robots_referers, key=lambda t: t[1], reverse=True)

        top_search_engine_referers = [(k, search_engine_referers[k]['pages']) for k in search_engine_referers.keys()]
        top_search_engine_referers = sorted(top_search_engine_referers, key=lambda t: t[1], reverse=True)

        top_key_phrases = key_phrases.items()
        top_key_phrases = sorted(top_key_phrases, key=lambda t: t[1], reverse=True)

        cur_time = self.iwla.getCurTime()
        index = self.iwla.getDisplayIndex()

        # All referers in a file
        if self.create_all_referers:
            title = createCurTitle(self.iwla, u'Connexion from')

            filename = 'referers.html'
            path = self.iwla.getCurDisplayPath(filename)

            page = display.createPage(title, path, self.iwla.getConfValue('css_path', []))
            table = display.createBlock(DisplayHTMLBlockTable, self.iwla._(u'Connexion from'), [self.iwla._(u'Origin'), self.iwla._(u'Pages'), self.iwla._(u'Hits')])
            table.setColsCSSClass(['', 'iwla_page', 'iwla_hit'])

            total_search = [0]*3
            table.appendRow(['<b>%s</b>' % (self.iwla._(u'Search Engine')), '', ''])
            new_list = self.max_referers and top_search_engine_referers[:self.max_referers] or top_search_engine_referers
            for r,_ in new_list:
                row = [r, search_engine_referers[r]['pages'], search_engine_referers[r]['hits']]
                total_search[1] += search_engine_referers[r]['pages']
                total_search[2] += search_engine_referers[r]['hits']
                table.appendRow(row)
            if self.max_referers:
                others = 0
                for (uri, entrance) in top_referers[self.max_referers:]:
                    others += entrance
                table.appendRow([self.iwla._(u'Others'), others])
                table.setCellCSSClass(table.getNbRows()-1, 0, 'iwla_others')

            total_external = [0]*3
            table.appendRow(['<b>%s</b>' % (self.iwla._(u'External URL')), '', ''])
            new_list = self.max_referers and top_referers[:self.max_referers] or top_referers
            for r,_ in new_list:
                row = [generateHTMLLink(r), referers[r]['pages'], referers[r]['hits']]
                total_external[1] += referers[r]['pages']
                total_external[2] += referers[r]['hits']
                table.appendRow(row)
            if self.max_referers:
                others = 0
                for (uri, entrance) in top_referers[self.max_referers:]:
                    others += entrance
                table.appendRow([self.iwla._(u'Others'), others])
                table.setCellCSSClass(table.getNbRows()-1, 0, 'iwla_others')

            total_robot = [0]*3
            table.appendRow(['<b>%s</b>' % (self.iwla._(u'External URL (robot)')), '', ''])
            new_list = self.max_referers and top_robots_referers[:self.max_referers] or top_robots_referers
            for r,_ in new_list:
                row = [generateHTMLLink(r), robots_referers[r]['pages'], robots_referers[r]['hits']]
                total_robot[1] += robots_referers[r]['pages']
                total_robot[2] += robots_referers[r]['hits']
                table.appendRow(row)
            if self.max_referers:
                others = 0
                for (uri, entrance) in top_referers[self.max_referers:]:
                    others += entrance
                table.appendRow([self.iwla._(u'Others'), others])
                table.setCellCSSClass(table.getNbRows()-1, 0, 'iwla_others')

            page.appendBlock(table)

            display.addPage(page)

        title = self.iwla._(u'Top Referers')
        if self.create_all_referers:
            link = '<a href=\'%s\'>%s</a>' % (filename, self.iwla._(u'All Referers'))
            title = '%s - %s' % (title, link)

        # Top referers in index
        table = display.createBlock(DisplayHTMLBlockTable, title, [self.iwla._(u'Origin'), self.iwla._(u'Pages'), self.iwla._(u'Hits')])
        table.setColsCSSClass(['', 'iwla_page', 'iwla_hit'])

        table.appendRow(['<b>%s</b>' % (self.iwla._(u'Search Engine')), '', ''])
        for r,_ in top_search_engine_referers[:10]:
            row = [r, search_engine_referers[r]['pages'], search_engine_referers[r]['hits']]
            total_search[1] -= search_engine_referers[r]['pages']
            total_search[2] -= search_engine_referers[r]['hits']
            table.appendRow(row)
        if total_search[1] or total_search[2]:
            total_search[0] = self.iwla._(u'Others')
            table.appendRow(total_search)
            table.setCellCSSClass(table.getNbRows()-1, 0, 'iwla_others')

        table.appendRow(['<b>%s</b>' % (self.iwla._(u'External URL')), '', ''])
        for r,_ in top_referers[:10]:
            row = [generateHTMLLink(r), referers[r]['pages'], referers[r]['hits']]
            total_external[1] -= referers[r]['pages']
            total_external[2] -= referers[r]['hits']            
            table.appendRow(row)
        if total_external[1] or total_external[2]:
            total_external[0] = self.iwla._(u'Others')
            table.appendRow(total_external)
            table.setCellCSSClass(table.getNbRows()-1, 0, 'iwla_others')

        table.appendRow(['<b>%s</b>' % (self.iwla._(u'External URL (robot)')), '', ''])
        for r,_ in top_robots_referers[:10]:
            row = [generateHTMLLink(r), robots_referers[r]['pages'], robots_referers[r]['hits']]
            total_robot[1] -= robots_referers[r]['pages']
            total_robot[2] -= robots_referers[r]['hits']
            table.appendRow(row)
        if total_robot[1] or total_robot[2]:
            total_robot[0] = self.iwla._(u'Others')
            table.appendRow(total_robot)
            table.setCellCSSClass(table.getNbRows()-1, 0, 'iwla_others')

        index.appendBlock(table)

        # All key phrases in a file
        if self.create_all_key_phrases:
            title = createCurTitle(self.iwla, u'Key Phrases')

            filename = 'key_phrases.html'
            path = self.iwla.getCurDisplayPath(filename)

            total_search = [0]*2
            page = display.createPage(title, path, self.iwla.getConfValue('css_path', []))
            table = display.createBlock(DisplayHTMLBlockTable, self.iwla._(u'Top key phrases'), [self.iwla._(u'Key phrase'), self.iwla._(u'Search')])
            table.setColsCSSClass(['', 'iwla_search'])
            new_list = self.max_key_phrases and top_key_phrases[:self.max_key_phrases] or top_key_phrases
            for phrase in new_list:
                table.appendRow([phrase[0], phrase[1]])
                total_search[1] += phrase[1]
            page.appendBlock(table)

            display.addPage(page)

        title = self.iwla._(u'Top key phrases')
        if self.create_all_key_phrases:
            link = '<a href=\'%s\'>%s</a>' % (filename, self.iwla._(u'All key phrases'))
            title = '%s - %s' % (title, link)

        # Top key phrases in index
        table = display.createBlock(DisplayHTMLBlockTable, title, [self.iwla._(u'Key phrase'), self.iwla._(u'Search')])
        table.setColsCSSClass(['', 'iwla_search'])
        for phrase in top_key_phrases[:10]:
            table.appendRow([phrase[0], phrase[1]])
            total_search[1] -= phrase[1]
        if total_search[1]:
            total_search[0] = self.iwla._(u'Others')
            table.appendRow(total_search)
            table.setCellCSSClass(table.getNbRows()-1, 0, 'iwla_others')
        index.appendBlock(table)