Closes https://github.com/janl/mustache.js/pull/738

In the spirit of keeping the public API of mustache.js as small as
possible for maintainence reasons, the undocumented and un-tested
`.to_html()` method is removed.

The functionality involved, where it can rather invoke an optional
function provided with the result of `.render()`, instead of returning
it as a string like `.render()` does, is something that using projects
very easily can do themselfs -- it does not have to be provided by
mustache.js.

These changes allows the internal template cache to
be customised, either by disabling it complete or
providing a custom implementation of how templates
are cached.

Minor adjustments needed to make the TypeScript compiler
that is built into Deno, be happy with how mustache.js'
ES module source looks in terms of function parameters
passed and object mutability.

Refs https://github.com/phillipj/mustache.js/pull/1

As part of transition the source code to be a proper plain ES module,
we're removing the handcrafted UMD wrapper around the source code.

This means the UMD wrapper will have to come from elsewhere; a build
step introduced later.

In an effort of moving with recent changes in Node.js and its built-in
support for ES modules, we're making the main source code reside inside
`mustache.mjs`. This will also allow Deno users to use this package as
an ES module directly by importing `mustache.mjs`.

An ES3 / UMD version of the source code is still planned to exist in
`mustache.js`, but will involve a build step to transform the ES module
source into a plain .js version.

By renaming the file *before* changing any contents, git will
recognise the file has been renamed and therefore allows us to see all
historical changes to the old .js source code with the `--follow` argument:

```
$ git log --follow -- mustache.mjs
```

Without ensuring git sees this is a file rename, we will in practise
loose the old source code's history, or at least make it quite weird
and less intuitive to find. That's not fair to the historical
contributors and in general getting hold of a well documented log of
changes to a file in git, is extremely valuable in those few scenarios
where it's *really* needed.

These changes fixes an indentation bug introduced when partials was
made to also be indented like its parent template. When partials are
rendered inline, it should not be indented with the amount of indentation
that the line already has.

Fixes https://github.com/janl/mustache.js/issues/715
Refs https://github.com/janl/mustache.js/pull/705

This small change fixes the output of functions used in partials with
indentation. Bug reports has shown that the functions output is shifted
with the amount of indentation the partial has.

The bug itself is best illustrated in the tests added in 621ae80652.

Closes https://github.com/janl/mustache.js/issues/712

As referenced in this issue:
https://github.com/janl/mustache.js/issues/562

Partials aren't indented as per the spec:
https://github.com/mustache/spec/blob/master/specs/partials.yml#L13-L15

This adds indentation tracking for partials and applies them to the
first instance of a partial on a line (to not indent inline partials
which would violate a different part of the spec).

* prevent value from being returned by Context.prototype.lookup if lookupHit is false

* add test for renderability of Array.length via dot notation

* Remove `typeof obj === 'object'` constraint in prop lookup

Allows rendering properties of primitive types that are not objects, such as a string.

* pop lookup needs to use hasOwnProperty for non-objs

* re-add  constraint in prop lookup, but make property lookups for primitives possible through dot notation

* add test to address #589 specifically

* enhance readability of primitiveHasOwnProperty and add comments to explain why it is used in one case but not the other

parse method

Pull requests #643 and #664 together fix the issue reported in issue #617, but the change in behavior is causing semvers concerns. See issue #669. Therefore, roll back #643 and #664 and later reintroduce in next planned major release (v3.0).

Fixes #669

Resolve issue #663 and add test.

When using amdclean to create a footprint less javascript file, the object needs to be returned to let it properly be assigned to the created script variable.

This closes a couple of potential exploit scenarios.
Backtick (`) for older IEs and equals (=) for unquoted attributes.

Refs 83b8e846a3
Closes https://github.com/janl/mustache.js/pull/388

Please see https://github.com/umdjs/umd/pull/63 and https://github.com/umdjs/umd/issues/35 for more information, but
this change prevents an HTML element with an id of `exports` from causing the UMD shim to incorrectly detect the environment as node.

As we require the template to be a `string` we now throw an error when given
any other data type. This should provide the developers with a meaningful
error rather than a cryptic TypeError from the murky depths of our source code.

Fixes #464

Fixes #445

Fixes #442

Made it eslint compatible.

Added `-v` option to CLI, bugfix for rendering `Number` when it was the
context, and specified files in package.json for cleaner `npm install`.