Reports¶

Overview¶

The main function of the program is to aggregate and display the data in helpful html reports. For each new report type there is a specific report class that inherits from the base Report class.

Reports¶

class Report(sites, period, recipients, frequency, subject)

sites - A list of site names to run the report on
period - A StatsRange object that defines the period that report should run over
recipients - A list of emails the report will send to
frequency - Can be “MONTHLY”, “WEEKLY” or “WOW_DAILY”. Specifies how often it runs and previous period comparison, e.g. month on month comparison for monthly reports.
subject - Subject line of the email and heading of the report

get_subject(): Adds date period onto the end of the argument subject.

E.g. for daily it will append Mon 12 Sep 2016.
get_freq_label(): Returns the string to be used for previous period comparisons.

E.g. for monthly returns MoM
check_data_availability(): Calls through to analytics to check data availability for that period for the requested sites.

Returns True if data is available, false if not.

If override is set to True, then it returns the name of the sites that must be overridden.
send_email(): Function to send a html email, inlines all the styles.
generate_html(): Function that gathers the data and renders the template

Here are the reports that are provided:

class YoutubeReport(Report)

Init function sets up the template file and link to youtube analytics data class.

generate_html(): Gather data tables from youtube analytics data class:

- Summary table

- Statistics table

- Country table

- Top Video table

- Traffic Source table

Render data tables to the specified html template

class AnalyticsCoreReport(Report)

Init function sets up the template file and link to google analytics data class. Report uses graphs so in __init__ set self.imgdata = None

get_site(): Returns either the name of the site to be used in the template

If the report iterates over all sites, then use a all inclusive name set in config, e.g. Network
generate_html(): Gather data tables from google analytics data class:

- Summary table

- Site summary table

- Country table

- Top Article table

- Traffic Source table

- Referral table

- Device table

- Social Network table

- If it’s not monthly:

- “Month to date” summary table

- If it is monthly:

- Device graph over the last year

- If it’s not all sites:

- All sites summary table

- If not monthly:

- “Month to date” network summary table

Render data to html template
send_email(): Overwrite basic send email function to include images

class AnalyticsSocialReport(Report)

Initialise data class, template, image data

get_site()

Returns either the name of the site to be used in the template

If the report iterates over all sites, then use a all inclusive name set in config, e.g. Network

generate_html()

Gathers data tables from google analytics data class:

- Summary table

- All sites summary table

- Social network table

- Social network graph

Render data to html template

send_email(): Overwrite basic send email function to include images

class AnalyticsYearSocialReport(Report)

Initialise data class, template etc

get_subject(): Overwrite basic subject function

Returns subject appended with date period from 3years ago to now
get_site(): Returns either the name of the site to be used in the template

If the report iterates over all sites, then use a all inclusive name set in config, e.g. Network
_get_social_data(): Gets social network data table for every month over the last 3 years
_get_top_networks(): Get social network data table over the last year and find top social networks

Other than Facebook, Twitter and reddit
generate_html(): Using the top social networks, find the data for each of these networks for each month over the last 3 years

using the data found in _get_social_data()

Render data to html template

class AnalyticsSocialExport(Report)

Initialise template, data class etc

generate_html(): Get social network table for each month over the last year

Render data to a csv file
send_email(): overwrites basic email function

sends email with a csv attachment

Templates¶

Templates use the Jinja 2 templating language.

The layout is set up so that each table function in the aggregating data classes corresponds to a template file, within the relevant folder.

For instance, within the templates folder there is GA folder in which each google analytics data table has a separate template file.

The file Templates/GA/base.html extends the main base, and it includes each table only if that table value exists. This is the file that the Google Analytics reports use as a template.

The youtube reports use the Templates/Youtube/base.html file as a template, and has similar structure where it includes the table template files, only if the table exists in that report.

All template files should inherit from Templates/base.html. In this file the styles are set, as is the subject and some confidentiality lines.

In Statsdash/render.py you can define filter functions to be used in the templates that aren’t already defined in Jinja.

Already defined filter functions:

int_comma(value): Adds commas to long numbers to make them more readable

e.g. when value = 1000, returns the string 1,000
cut(word, part=”www.”): Removes the substring defined by part from the string word

Substring to remove defaults as “www.” for site names, but can be any string

Returns the stripped word

Use

{{ value|int_comma }}
{{ site_name|cut("www.") }}

Add a new report¶

To set up a new report, you’ll need to create a new class that inherits from “Report”.

Basic example that returns just a google analytics summary table:

class MyReport(Report):

    def __init__(self, sites, period, recipients, frequency, subject):
        super(MyReport, self).__init__(sites, period, recipients, frequency, subject)
        self.sites = sites
        self.period = period
        self.recipients = recipients
        self.frequency = frequency
        self.subject = subject
        self.warning_sites = []
        self.template = self.env.get_template("template.html")
        self.date = AnalyticsData(self.sites, self.period, self.frequency)

    def generate_html(self):
        summary = self.data.summary_talbe()
        html = self.template.render(
            subject = self.get_subject(),
            change = self.get_freq_label(),
            report_span = self.frequency,
            warning_sites = self.warning_sites,
            summary_table = summary
        )
        return html

Run a report¶

Preview¶

You can preview an individual report using the preview_report.py file.

To preview a report you can run the file from the command line with these paramaters:

argument	Optional	Default	Definition
reporttype	Required	No default	The name of the report class, e.g. AnalyticsCoreReport
–sitename	Optional	All sites	Name of the site, as it is in the config
–destination	Optional	”.”	Path to where to save the report out
–filename	Optional	report_preview	Name of the file to save

Example configuration might be:

python preview_report.py AnalyticsCoreReport -s site.net -d file/to/path -n my_report_test

The frequency and periods are manually set up within the file. To alter these you’ll have to alter the file.There are preset monthly, weekly and daily StatsRange objects setup but you may need to just change the date or frequency of the report.

Schedule¶

Set up a report in report_schedule.py and run the scheduler.

For more info see Report Schedule Settings and Using the Scheduler.