HTML Report

An HTML report is the family where you write the page yourself. A pivot report is built by configuring the table; here you decide the layout, the formatting and how the figures are presented.

Two halves

The report is split in two on purpose, and the split is the point:

Half	Where it runs	What it does
Datasource	On the server	Prepares the data and returns it
HTML · CSS · JS	In the browser	Presentation only; receives the data ready-made

The datasource is a script and reaches the platform’s ins API. The page itself never calls back into the application — by the time the document starts rendering, the data is already embedded in it.

The page cannot call the API, and that is not a shortcoming

On a Custom HTML page you can read data from the browser; in an HTML report you cannot. The reason: the browser that prints the report has no identity of its own and no network access; it is handed a finished document.

The alternative — having the page fetch its own data while rendering — would need an authenticated channel into the application, a service account, a key, and a way to bind row access to that key. None of it is needed, because the fetch already happened: on the server, with the access of whoever triggered it.

Whose access the data is read under

The datasource reaches the ins API under the access of whoever triggered the render: the person looking at a preview, or the space’s automation identity on a scheduled run. Not the author’s, and not the owner’s.

That shows itself in addressed delivery: the report is produced separately for recipients with different row access, so each gets the data they are entitled to see.

The datasource script

Four values are already there for the script:

Binding	Type	Description
startDate · endDate	Epoch milliseconds	The range the report covers
isAuto	Boolean	true on a scheduled run, false when run by hand
lang	Text	The language chosen for the report; empty when none was given

What you return can be any shape you like — nothing downstream inspects it, it is carried to the page as it is.

// Datasource: prepare the slice's hourly averages
var rows = ins.getLoggedHourlyVariableValueStats(
    ['Temperature_C'], ins.getDate(startDate), ins.getDate(endDate));
return { title: 'Temperature summary', automatic: isAuto, rows: rows };

The script runs on the server, so the script engine’s rules apply here too — the syntax, the limits and the ins.* signatures.

How the data reaches the page

What the datasource returns is embedded in the document as window.__reportData. The page’s JavaScript reads it directly:

var d = window.__reportData;
document.getElementById('title').textContent = d.title;

Fields and scheduling

Field	Description
Name · Description	Name and description
HTML · CSS · JS	Presentation; up to 1 MB each
Datasource	The data script
Period · Report time	As on a pivot report — Hourly, Monthly, Minutes
Print · Mail	Separate day and time for printing and mail, each turned on independently

Scheduling, printing, mail and the output formats all work as they do on a pivot report, including the rule that recipients come from the report’s own Mail To field. Detail: Pivot Report.

The output is HTML and PDF, the PDF printed by a headless browser on the server.

Related

• HTML Report Editor — the screen where this report is written
• Pivot Report — the other family, built by configuration
• Script Engine — the rules the datasource follows