path: root/subplot.md

# Introduction

Subplot is software to help capture and communicate acceptance
criteria for software and systems, and how they are verified, in a way
that's understood by all project stakeholders. The current document
contains the acceptance criteria for Subplot itself, and its
architecture.

The acceptance criteria are expressed as _scenarios_, which roughly
correspond to use cases. The scenario as accompanied by explanatory
text to explain things to the reader. Scenarios use a given/when/then
sequence of steps, where each step is implemented by code provided by
the developers of the system under test. This is very similar to the
[Cucumber][] tool, but with more emphasis of producing a standalone
document.

[Cucumber]: https://en.wikipedia.org/wiki/Cucumber_(software)

## Subplot architecture

Subplot reads an input document, in Markdown, and generates a typeset
output document, as PDF or HTML, for all stakeholders to understand.
Subplot also generates a test program, in Python, that verifies the
acceptance criteria are met, for developers and testers and auditors
to verify the sustem under test meets its acceptance criteria. The
generated program uses code written by the Subplot user to implement
the verification steps. The graph below illustrates this and shows how
data flows through the system.

```dot
digraph "architecture" {
md [label="foo.md \n (document, Markdown)"];
md [shape=box];

bindings [label="foo.myaml \n (bindings, YAML)"];
bindings [shape=box];

impl [label="foo.py \n (step implemenations, Python)"]
impl [shape=box];

subplot [label="Subplot"];
subplot [shape=ellipse];

pdf [label="foo.pdf \n PDF (generated)"]
pdf [shape=note];

html [label="foo.html \n HTML (generated)"]
html [shape=note];

testprog [label="test.py \n test program\n(generated)"]
testprog [shape=note];

report [label="Test report \n (stdout of test.py)"]
report [shape=note];

md -> subplot;
bindings -> subplot;
impl -> subplot;
subplot -> pdf;
subplot -> html;
subplot -> testprog;
testprog -> report;
}
```

[Pandoc]: https://pandoc.org/

Subplot uses the [Pandoc][] software for generating PDF and HTML
output documents. In fact, any output format supported by Pandoc can
be requested by the user. Depending on the output format, Pandoc may
use, for example, LaTeX. Subplot interprets parts of the Markdown
input file itself.

Subplot actually consists mainly of two separate programs: 
**sp-docgen** for generating output documents, and **sp-codegen** for
generating the test program. There are a couple of additional tools
(**sp-meta** for reporting meta data about a Subplot document, and
**sp-filter** for doing the document generation as a Pandoc filter).

Thus a more detailed architecture view is shown below.

```dot
digraph "architecture2" {
md [label="foo.md \n (document, Markdown)"];
md [shape=box];

bindings [label="foo.myaml \n (bindings, YAML)"];
bindings [shape=box];

impl [label="foo.py \n (step implemenations, Python)"]
impl [shape=box];

docgen [label="sp-docgen"];
docgen [shape=ellipse];

codegen [label="sp-codegen"];
codegen [shape=ellipse];

pdf [label="foo.pdf \n PDF (generated)"]
pdf [shape=note];

html [label="foo.html \n HTML (generated)"]
html [shape=note];

testprog [label="test.py \n test program\n(generated)"]
testprog [shape=note];

report [label="Test report \n (stdout of test.py)"]
report [shape=note];

md -> docgen;
bindings -> docgen;
md -> codegen;
bindings -> codegen;
impl -> codegen;
docgen -> pdf;
docgen -> html;
codegen -> testprog;
testprog -> report;
}
```

## A fairy tale of acceptance testing

The king was upset. This naturally meant the whole court was in a
tizzy and chattering excitedly at each other, while trying to avoid
the royal wrath.

"Who will rid me of this troublesome chore?" shouted the king, and
quaffed a flagon of wine. "And no killing of priests, this time!"

The grand hall's doors were thrown open. The grand wizard stood in the
doorway, robe, hat, and staff everything, but quite still. After the
court became silent, the wizard strode confidently to stand before the
king.

"What ails you, my lord?"

The king looked upon the wizard, and took a deep breath. It does not
do to shout at wizards, for they control dragons, and even kings are
tasty morsels to the great beasts.

"I am tired of choosing what to wear every day. Can't you do
something?"

The wizard stoke his long, grey beard. He turned around, looked at the
magnificent outfits worn by members of the court. He turned back, and
looked at the king.

"I believe I can fix this. Just to be clear, your beef is with having
to choose clothing, yes?"

"Yes", said the king, "that's what I said. When will you be done?"

The wizard raised his staff and brought it back down again, with a
loud bang.

"Done" said the wizard, smugly.

The king was amazed and started smiling, until he noticed that
everyone, including himself, was wearing identical burlap sacks and
nothing on their feet. His voice was high, whiny, like that of a
little child.

"Oh no, that's not at all what I wanted! Change it back! Change it
back now!"

The morale of this story is to be clear and precise in your acceptance
criteria, or you might get something other than what you really, really
wanted.


## Motivation for Subplot

Keeping track of requirements and acceptance criteria is necessary for
all but the simplest of software projects. Having all stakeholders in
a project agree to them is crucial, as is that all agree how it is
verified that the software meets the acceptance criteria. Subplot
provides a way for documenting the shared understanding of what the
acceptance criteria are and how they can be checked automatically.

Stakeholders in a project may include:

* those who pay for the work to be done; this may be the employer of
  the developers for in-house projects ("*customer*")
* those who use the resulting systems, whether they pay for it or not
  ("*user*")
* those who install and configure the systems and keep them functional
  ("*sysadmin*")
* those who support the users ("*support*")
* those who test the project for acceptability ("*tester*")
* those who develop the system in the first place ("*developer*")

The above list is incomplete and simplistic, but suffices as an
example.

All stakeholders need to understand the acceptance criteria, and how
the system is evaluated against the criteria. In the simplest case,
the customer and the developer need to both understand and agree so
that the developer knows when the job is done, and the customer knows
when they need to pay their bill.

However, even when the various stakeholder roles all fall upon the
same person, or only on people who act as developers, the Subplot
tooling can be useful. A developer would understand acceptance
criteria expressed only in code, but doing so may take time and energy
that are not always available. The Subplot approach aims to encourage
hiding unnecessary detail and documenting things in a way that is easy
to understand with little effort.

Unfortunately, this does mean that for a Subplot output document to
be good and helpful, writing it will require effort and skill. No tool
can replace that.


# Requirements

This chapter lists requirements for Subplot. These requirements are
not meant to be automatically verifiable. For specific, automatically
testable acceptance criteria, see the later [chapter with acceptance
tests for Subplot](#acceptance).

Each requirement here is given a unique mnemnoic id for easier
reference in discussions.

**UnderstandableTests**

:   Acceptance tests should be possible to express in a way that's
    easily understood by all stakeholders, includcing those who are
    not software developers.

    _Done_ but requires the Subplot document to be written with care.

**EasyToWriteDocs**

:   The markup language for writing documentation should be easy to
    write.

    _Done_ by using Markdown.

**AidsComprehension**

:   The formatted human-readable documentation should use good layout
    and typography to enhance comprension.

    _In progress_ — typesetting via Pandoc works, but may need
    review and improvement.

**CodeSeparately**

:   The code to implement the acceptance criteria should not be
    embedded in the documentation source, but be in separate files.
    This makes it easier to edit without specialised tooling.

    _Done_ by keeping scenario step implementations in a separate
    file.

**AnyProgammingLanguage**

:   The developers implementing the acceptance tests should be free to
    use a language they're familiar and comfortable with. Subplot
    should not require them to use a specific language.

    _Not done_ — only Python supported at the moment.

**FastTestExecution**

:   Executing the acceptance tests should be fast.

    _Not done_ &mash; the generated Python test program is simplistic
    and linear.

**NoDeployment**

:   The acceptance test tooling should assume the system under test is
    already deployed and available. Deploying is too big of a problem
    space to bring into the scope of acceptance testing, and there are
    already good tools for deployment.

    _Done_ by virtue of letting those who implement the scenario steps
    worry about it.

**MachineParseableResults**

:   The tests should produce a machine parseable result that can be
    archived, post-processed, and analyzed in ways that are of
    interest to the project using Subplot. For example, to see trends
    in how long tests take, how often tests fail, to find regressions,
    and to find tests that don't provide value.

    _Not done_ — the generated test program is simplistic.


# Subplot input language

Subplot reads three input files, each in a different format:

* The document file, which uses the Markdown dialect understood by
  Pandoc.
* The bindings file, in YAML.
* The functions file, in Python.

Subplot interprets specially marked parts of the input document
specially. It does this via the Pandoc abstract syntax tree, rather
than text manipulation, and thus anything that Pandoc understands is
understood by Subplot. We will not specify Pandoc's dialect of
Markdown here, only the parts Subplot pays attention to.


## Document metadata

Pandoc supports, and Subplot makes use of, a [YAML metadata block][] in a
Markdown document. This can and should be used to set the document
title, authors, date (version), and can be used to control some of the
typesetting. Crucially for Subplot, the bindings and functions files
are named in the metadata block, rather than Subplot deriving them
from the input file name.

[YAML metadata block]: https://pandoc.org/MANUAL.html#extension-yaml_metadata_block

As an example, the metadata block for the Subplot document might look
as follows. The `---` before and `...` after the block are mandatory:
they are how Pandoc recongizes the block.

~~~{.yaml .numberLines}
---
title: "Subplot"
author: The Subplot project
date: work in progress
bindings: subplot.yaml
functions: subplot.py
...
~~~


## Document markup

[fenced code blocks]: https://pandoc.org/MANUAL.html#fenced-code-blocks

Subplot understands certain tags for [fenced code blocks][] specially.
A scenario, for example, would look like this:

~~~~~~{.markdown .numberLines}
```scenario
given a standard setup
when peace happens
then everything is OK
```
~~~~~~

The `scenario` tag on the code block is recognized by Subplot, which
will typeset the scenario (in output documents) or generate code (for
the test program) accordingly. Scenario blocks do not need to be
complete scenario. Subplot will collect all the snippets into one
block for the test program. Snippets under the same heading belong
together; the next heading of the same or a higher level ends the
scenario.

For embedding test data files in the Markdown document, Subplot
understands the `file` tag:

~~~~~~~~markdown
~~~{#filename .file}
This data is accessible to the test program as 'filename'.
~~~
~~~~~~~~

The `.file` attribute is necessary, as is the identifier, here
`#filename`. The generated test program can access the data using the
identifier (without the #). The mechanism used is generic to Pandoc,
and can be used to affect the typesetting by adding more attributes.
For example, Pandoc can typeset the data in the code block using
syntax highlighting, if the language is specified: `.markdown`,
`.yaml`, or `.python`, for example.

Subplot also understands the `dot` and `roadmap` tags, and can use the
Graphviz dot program, or the [roadmap][] Rust crate, to produce
graphs. These can useful for describing things visually.

[roadmap]: https://crates.io/search?q=roadmap


## Bindings file

The bindings file binds scenario steps to Python functions that
implement the steps. The YAML file is a list of objects (also known as
dicts or hashmaps or key/value pairs), specifying a step kind (given,
when, then), a regular expression matching the text of the step and
optionally capturing interesting parts of the text, and the name of a
function that implements the step.

~~~{.yaml .numberLines}
- given: a standard setup
  function: create_standard_setup
- when: (?<thing>\S+) happens
  function: make_thing_happen
- then: everything is OK
  function: check_everything_is_ok
~~~

In the example above, there are three bindings:

* A binding for a "given a standard setup" step. The binding captures
  no part of the text, and causes the `create_standard_setup` function
  to be called.
* A binding for a "when" step consisting of one word followed by
  "happens". For example, "peace", as in "then peace happens". The
  word is captured as "thing", and given to the `make_thing_happen`
  function as an argument when it is called.
* A binding for a "then everthing is OK" step, which captures nothing,
  and calls the `check_everything_is_ok` function.

The regular expressions use [PCRE][] syntax as implemented by the Rust
[regex][] crate. The `(?P<name>pattern)` syntax is used to capture
parts of the step. The captured parts are given to the bound function
as arguments, when it's called.

[PCRE]: https://en.wikipedia.org/wiki/Perl_Compatible_Regular_Expressions
[regex]: https://crates.io/crates/regex



## Functions file

The functions file is not parsed by Subplot at all. Subplot merely
copies it to the output. All parsing and validation of the file is
done by the Python implementation.

The Python functions must accept a "context" argument, and a keyword
argument for each part of the step the corresponding regular
expression captures. The capture name and the keyword argument name
must be the same.

The context argument is a dict-like object, which the generated
program creates automatically. The context is carried from function
call to function call, to allow functions to manage state between
themselves. Typically, one step might do something, and record the
results into the context, and another step might check the results by
inspecting the context. This decouples functions from each other, and
avoids having them use global variables for state.



# Acceptance criteria for Subplot {#acceptance}


Add the acceptance criteria test scenarios for Subplot here.


## Test data shared between scenarios

The scenarios below test Subplot by running it against specific input
files. This section specifies the bindings and functions files.
They're separate from the scenarios so that the scenarios are shorter
and clearer, but also so that the input files do no need to be
duplicated for each scenario.

~~~~{#simple.md .file .markdown .numberLines}
---
title: Test scenario
bindings: b.yaml
functions: f.py
...

# Simple
This is the simplest possible test scenario

```scenario
given precondition foo
when I do bar
then bar was done
```
~~~~

~~~{#b.yaml .file .yaml .numberLines}
- given: precondition foo
  function: precond_foo
- when: I do bar
  function: do_bar
- when: I do foobar
  function: do_foobar
- then: bar was done
  function: bar_was_done
- then: foobar was done
  function: foobar_was_done
~~~


~~~{#f.py .file .python .numberLines}
def precond_foo(ctx):
    ctx['bar_done'] = False
    ctx['foobar_done'] = False
def do_bar(ctx):
    ctx['bar_done'] = True
def bar_was_done(ctx):
    assert_eq(ctx['bar_done'], True)
def do_foobar(ctx):
    ctx['foobar_done'] = True
def foobar_was_done(ctx):
    assert_eq(ctx['foobar_done'], True)
~~~


## Smoke test

This tests that Subplot can build a PDF and an HTML document, and
execute a simple scenario successfully. The test is based on
generating the test program from an input file, running the test
program, and examining the output.

~~~scenario
given file simple.md
and file b.yaml
and file f.py
when I run sp-docgen simple.md -o simple.pdf
then file simple.pdf exists
when I run sp-docgen simple.md -o simple.html
then file simple.html exists
when I run sp-codegen --run simple.md -o test.py
then scenario "Simple" was run
and step "given precondition foo" was run
and step "when I do bar" was run
and step "then bar was done" was run
and program finished successfully
~~~


## Keywords

Subplot supports the keywords **given**, **when**, and **then**, and
the aliases **and** and **but**. The aliases stand for the same
(effective) keyword as the previous step in the scenario. This chapter
has scenarios to check the keywords and aliases in various
combinations.

### All the keywords

~~~scenario
given file allkeywords.md
and file b.yaml
and file f.py
when I run sp-docgen allkeywords.md -o foo.pdf
then file foo.pdf exists
when I run sp-codegen --run allkeywords.md -o test.py
then scenario "All keywords" was run
and step "given precondition foo" was run
and step "when I do bar" was run
and step "then bar was done" was run
and program finished successfully
~~~

~~~{#allkeywords.md .file .markdown .numberLines}
---
title: All the keywords scenario
bindings: b.yaml
functions: f.py
...

# All keywords

This uses all the keywords.

```scenario
given precondition foo
when I do bar
and I do foobar
then bar was done
but foobar was done
```
~~~

### Keyword aliases

~~~scenario
given file aliases.md
and file b.yaml
and file f.py
when I run sp-docgen aliases.md -o aliases.html
then file aliases.html matches /given<[^>]*> precondition foo/
and file aliases.html matches /when<[^>]*> I do bar/
and file aliases.html matches /and<[^>]*> I do foobar/
and file aliases.html matches /then<[^>]*> bar was done/
and file aliases.html matches /but<[^>]*> foobar was done/
and program finished successfully
~~~

~~~{#aliases.md .file .markdown .numberLines}
---
title: Keyword aliasesG
bindings: b.yaml
functions: f.py
...

# Aliases

```scenario
given precondition foo
when I do bar
and I do foobar
then bar was done
but foobar was done
```
~~~

### Misuse of continuation keywords

When continuation keywords (`and` and `but`) are used, they have to not be
the first keyword in a scenario.  Any such scenario will fail to parse because
subplot will be unable to determine what kind of keyword they are meant to
be continuing.

~~~scenario
given file continuationmisuse.md
and file b.yaml
and file f.py
when I run sp-docgen continuationmisuse.md -o foo.pdf
then file foo.pdf exists
when I try to run sp-codegen --run continuationmisuse.md -o test.py
then exit code is non-zero
~~~

~~~{#continuationmisuse.md .file .markdown .numberLines}
---
title: Continuation keyword misuse
bindings: b.yaml
functions: f.py
...

# Continuation keyword misuse

This scenario should fail to parse because we misuse a
continuation keyword at the start.

```scenario
and precondition foo
when I do bar
then bar was done
```
~~~


## Empty lines in scenarios

This scenario verifies that empty lines in scenarios are ignored.

~~~scenario
given file emptylines.md
and file b.yaml
and file f.py
when I run sp-docgen emptylines.md -o emptylines.pdf
then file emptylines.pdf exists
when I run sp-docgen emptylines.md -o emptylines.html
then file emptylines.html exists
when I run sp-codegen --run emptylines.md -o test.py
then scenario "Simple" was run
and step "given precondition foo" was run
and step "when I do bar" was run
and step "then bar was done" was run
and program finished successfully
~~~

~~~~{#emptylines.md .file .markdown .numberLines}
---
title: Test scenario
bindings: b.yaml
functions: f.py
...

# Simple
This is the simplest possible test scenario

```scenario
given precondition foo

when I do bar

then bar was done

```
~~~~


## Automatic cleanup in scenarios

A binding can define a cleanup function, which gets called at the end
of the scenario in reverse order for the successful steps. If a step
fails, all the cleanups for the successful steps are still called. We
test this for every language templat we support.

~~~{#cleanup.yaml .file .yaml .numberLines}
- given: foo
  function: foo
  cleanup: foo_cleanup
- given: bar
  function: bar
  cleanup: bar_cleanup
- given: failure
  function: failure
  cleanup: failure_cleanup
~~~

~~~{#cleanup.py .file .python .numberLines}
def foo(ctx):
   pass
def foo_cleanup(ctx):
   pass
def bar(ctx):
   pass
def bar_cleanup(ctx):
   pass
def failure(ctx):
   assert 0
def failure_cleanup(ctx):
   pass
~~~

~~~{#cleanup.sh .file .bash .numberLines}
foo() {
    true
}
foo_cleanup() {
    true
}
bar() {
    true
}
bar_cleanup() {
    true
}
failure() {
   return 1
}
failure_cleanup() {
   true
}
~~~


### Cleanup functions gets called on success (Python)

~~~scenario
given file cleanup-success-python.md
and file cleanup.yaml
and file cleanup.py
when I run sp-codegen --run cleanup-success-python.md -o test.py
then scenario "Cleanup" was run
and step "given foo" was run, and then step "given bar"
and cleanup for "given bar" was run, and then for "given foo"
and program finished successfully
~~~


~~~~~{#cleanup-success-python.md .file .markdown .numberLines}
---
title: Cleanup
bindings: cleanup.yaml
functions: cleanup.py
template: python
...

# Cleanup

~~~scenario
given foo
given bar
~~~
~~~~~


### Cleanup functions get called on failure (Python)

~~~scenario
given file cleanup-fail-python.md
and file cleanup.yaml
and file cleanup.py
when I try to run sp-codegen --run cleanup-fail-python.md -o test.py
then scenario "Cleanup" was run
and step "given foo" was run, and then step "given bar"
and cleanup for "given bar" was run, and then for "given foo"
and cleanup for "given failure" was not run
and exit code is non-zero
~~~

~~~~~{#cleanup-fail-python.md .file .markdown .numberLines}
---
title: Cleanup
bindings: cleanup.yaml
functions: cleanup.py
template: python
...

# Cleanup

~~~scenario
given foo
given bar
given failure
~~~
~~~~~


### Cleanup functions gets called on success (Bash)

~~~scenario
given file cleanup-success-bash.md
and file cleanup.yaml
and file cleanup.sh
when I run sp-codegen --run cleanup-success-bash.md -o test.sh
then scenario "Cleanup" was run
and step "given foo" was run, and then step "given bar"
and cleanup for "given bar" was run, and then for "given foo"
and program finished successfully
~~~

~~~~~{#cleanup-success-bash.md .file .markdown .numberLines}
---
title: Cleanup
bindings: cleanup.yaml
functions: cleanup.sh
template: bash
...

# Cleanup

~~~scenario
given foo
given bar
~~~
~~~~~


### Cleanup functions get called on failure (Bash)

If a step fails, all the cleanups for the preceding steps are still
called, in reverse order.

~~~scenario
given file cleanup-fail-bash.md
and file cleanup.yaml
and file cleanup.sh
when I try to run sp-codegen --run cleanup-fail-bash.md -o test.sh
then scenario "Cleanup" was run
and step "given foo" was run, and then step "given bar"
and cleanup for "given bar" was run, and then for "given foo"
and cleanup for "given failure" was not run
and exit code is non-zero
~~~

~~~~~{#cleanup-fail-bash.md .file .markdown .numberLines}
---
title: Cleanup
bindings: cleanup.yaml
functions: cleanup.sh
template: bash
...

# Cleanup

~~~scenario
given foo
given bar
given failure
~~~
~~~~~



## Capturing parts of steps for functions

A scenario step binding can capture parts of a scenario step, to be
passed to the function implementing the step as an argument. Captures
can be done using regular expressions.

### Capture using simple patterns

~~~scenario
given file simplepattern.md
and file simplepattern.yaml
and file capture.py
when I run sp-codegen --run simplepattern.md -o test.py
then scenario "Simple pattern" was run
and step "given I am Tomjon" was run
and output matches /function got argument name as Tomjon/
and program finished successfully
~~~

~~~~{#simplepattern.md .file .markdown .numberLines}
---
title: Simple pattern capture
bindings: simplepattern.yaml
functions: capture.py
...

# Simple pattern

~~~scenario
given I am Tomjon
~~~
~~~~

~~~{#simplepattern.yaml .file .yaml .numberLines}
- given: I am {name}
  function: func
~~~

~~~{#capture.py .file .python .numberLines}
def func(ctx, name=None):
    print('function got argument name as', name)
~~~

### Simple patterns with regex metacharacters: forbidden case

Help use to avoid accidental regular expression versus simple pattern
confusion. The rule is that a simple pattern mustn't contain regular
expression meta characters unless the rule is explicitly marked as not
being a regular expression pattern.

~~~scenario
given file confusedpattern.md
and file confusedpattern.yaml
and file capture.py
when I try to run sp-codegen --run confusedpattern.md -o test.py
then exit code is non-zero
and stderr matches /simple pattern contains regex/
~~~

~~~~{#confusedpattern.md .file .markdown .numberLines}
---
title: Simple pattern capture
bindings: confusedpattern.yaml
functions: capture.py
...

# Simple pattern

~~~scenario
given I* am Tomjon
~~~
~~~~

~~~{#confusedpattern.yaml .file .yaml .numberLines}
- given: I* am {name}
  function: func
~~~

### Simple patterns with regex metacharacters: allowed case

~~~scenario
given file confusedbutok.md
and file confusedbutok.yaml
and file capture.py
when I run sp-codegen --run confusedbutok.md -o test.py
then program finished successfully
~~~

~~~~{#confusedbutok.md .file .markdown .numberLines}
---
title: Simple pattern capture
bindings: confusedbutok.yaml
functions: capture.py
...

# Simple pattern

~~~scenario
given I* am Tomjon
~~~
~~~~

~~~{#confusedbutok.yaml .file .yaml .numberLines}
- given: I* am {name}
  function: func
  regex: false
~~~

### Capture using regular expressions

~~~scenario
given file regex.md
and file regex.yaml
and file capture.py
when I run sp-codegen --run regex.md -o test.py
then scenario "Regex" was run
and step "given I am Tomjon" was run
and output matches /function got argument name as Tomjon/
and program finished successfully
~~~

~~~~{#regex.md .file .markdown .numberLines}
---
title: Regex capture
bindings: regex.yaml
functions: capture.py
...

# Regex

~~~scenario
given I am Tomjon
~~~
~~~~

~~~{#regex.yaml .file .yaml .numberLines}
- given: I am (?P<name>\S+)
  function: func
  regex: true
~~~


## Avoid changing typesetting output file needlessly

### Avoid typesetting if output is newer than source files

This scenario make sure that if docgen generates the bitwise identical
output to the existing output file, it doesn't actually write it to
the output file, including its timestamp. This avoids triggering
programs that monitor the output file for changes.

~~~scenario
given file simple.md
and file b.yaml
and file f.py
when I run sp-docgen simple.md -o simple.pdf
then file simple.pdf exists
when I remember the metadata for simple.pdf
and I run sp-docgen simple.md -o simple.pdf
then simple.pdf has the same metadata as before
and only files simple.md, b.yaml, f.py, simple.pdf exist
~~~

### Do typeset if output is older than markdown

~~~scenario
given file simple.md
and file b.yaml
and file f.py
when I run sp-docgen simple.md -o simple.pdf
then file simple.pdf exists
when I remember the metadata for simple.pdf
and I touch simple.md
and I run sp-docgen simple.md -o simple.pdf
then simple.pdf has changed from before
~~~

### Do typeset if output is older than functions

~~~scenario
given file simple.md
and file b.yaml
and file f.py
when I run sp-docgen simple.md -o simple.pdf
then file simple.pdf exists
when I remember the metadata for simple.pdf
and I touch f.py
and I run sp-docgen simple.md -o simple.pdf
then simple.pdf has changed from before
~~~

### Do typeset if output is older than bindings

~~~scenario
given file simple.md
and file b.yaml
and file f.py
when I run sp-docgen simple.md -o simple.pdf
then file simple.pdf exists
when I remember the metadata for simple.pdf
and I touch b.yaml
and I run sp-docgen simple.md -o simple.pdf
then simple.pdf has changed from before
~~~

## Document structure

Subplot uses chapters and sections to keep together scenario snippets
that form a complete scenario. The lowest level heading before a
snippet starts a scenario and is the name of the scenario. If there's
subheadings, they divide the description of the scenario into parts,
but don't start a new scenario. The next heading at the same or a
higher level starts a new scenario.

### Lowest level heading is name of scenario

~~~scenario
given file scenarioislowest.md
and file b.yaml
and file f.py
when I run sp-codegen --run scenarioislowest.md -o test.py
then scenario "heading 1.1.1" was run
and program finished successfully
~~~

~~~~{#scenarioislowest.md .file .markdown .numberLines}

---
title: Test scenario
bindings: b.yaml
functions: f.py
...

# heading 1
## heading 1.1
### heading 1.1.1

```scenario
given precondition foo
```
~~~~

### Subheadings don't start new scenario

~~~scenario
given file subisnotnewscenario.md
and file b.yaml
and file f.py
when I run sp-codegen --run subisnotnewscenario.md -o test.py
then scenario "heading 1.1a" was run
and program finished successfully
~~~

~~~~{#subisnotnewscenario.md .file .markdown .numberLines}

---
title: Test scenario
bindings: b.yaml
functions: f.py
...

# heading 1
## heading 1.1a

```scenario
given precondition foo
```

### heading 1.1.1
### heading 1.1.2

~~~~

### Next heading at same level starts new scenario

~~~scenario
given file samelevelisnewscenario.md
and file b.yaml
and file f.py
when I run sp-codegen --run samelevelisnewscenario.md -o test.py
then scenario "heading 1.1.1" was run
and scenario "heading 1.1.2" was run
and program finished successfully
~~~

~~~~{#samelevelisnewscenario.md .file .markdown .numberLines}

---
title: Test scenario
bindings: b.yaml
functions: f.py
...

# heading 1
## heading 1.1
### heading 1.1.1

```scenario
given precondition foo
```
### heading 1.1.2

```scenario
given precondition foo
```
~~~~


### Next heading at higher level starts new scenario

~~~scenario
given file higherisnewscenario.md
and file b.yaml
and file f.py
when I run sp-codegen --run higherisnewscenario.md -o test.py
then scenario "heading 1.1.1" was run
and scenario "heading 1.2" was run
and program finished successfully
~~~

~~~~{#higherisnewscenario.md .file .markdown .numberLines}

---
title: Test scenario
bindings: b.yaml
functions: f.py
...

# heading 1
## heading 1.1
### heading 1.1.1

```scenario
given precondition foo
```
## heading 1.2

```scenario
given precondition foo
```
~~~~


### Document titles

The document and code generators require a document title, because
it's a common user error to not have one, and Subplot should help make
good documents. The Pandoc filter, however, mustn't require a document
title, because it's used for things like formatting websites using
ikiwiki, and ikiwiki has a different way of specifying page titles.

#### Document generator gives an error if input document lacks title

~~~scenario
given file notitle.md
when I try to run sp-docgen notitle.md -o foo.md
then exit code is non-zero
~~~

~~~{#notitle.md .file .markdown .numberLines}
---
bindings: b.yaml
functions: f.py
...


# Introduction

This is a very simple Markdown file without a YAML metadata block,
and thus also no document title.

```scenario
given precondition foo
when I do bar
then bar was done
~~~

#### Code generator gives an error if input document lacks title

~~~scenario
given file notitle.md
when I try to run sp-codegen --run notitle.md -o test.py
then exit code is non-zero
~~~


## Running only chosen scenarios

To make the edit-test loop more convenient for the test programs
generated by Subplot, we allow the user to specify patterns for
scenarios to run. Default is to run all scenarios.

### Running only chosen scenarios with Python

This verifies that the generated Python test program can run only
chosen scenarios.

~~~scenario
given file twoscenarios-python.md
and file b.yaml
and file f.py
when I run sp-codegen twoscenarios-python.md -o test.py
and I run python3 test.py on
then scenario "One" was run
and scenario "Two" was not run
and program finished successfully
~~~

~~~{#twoscenarios-python.md .file .markdown .numberLines}
---
title: Test scenario
bindings: b.yaml
functions: f.py
...

# One

```scenario
given precondition foo
when I do bar
then bar was done
```

# Two

```scenario
given precondition foo
when I do bar
then bar was done
```
~~~

### Running only chosen scenarios with Bash

This verifies that the generated Bash test program can run only
chosen scenarios.

~~~scenario
given file twoscenarios-bash.md
and file b.yaml
and file f.sh
when I run sp-codegen twoscenarios-bash.md -o test.sh
and I run bash test.sh on
then scenario "One" was run
and scenario "Two" was not run
and program finished successfully
~~~


~~~{#twoscenarios-bash.md .file .markdown .numberLines}
---
title: Test scenario
bindings: b.yaml
functions: f.sh
template: bash
...

# One

```scenario
given precondition foo
when I do bar
then bar was done
```

# Two

```scenario
given precondition foo
when I do bar
then bar was done
```
~~~

~~~{#f.sh .file .bash .numberLines}
precond_foo() {
    ctx_set bar_none 0
    ctx_set foobar_none 0
}

do_bar() {
    ctx_set bar_done 1
}

do_foobar() {
    ctx_set foobar_done 1
}

bar_was_done() {
    actual="$(ctx_get bar_done)"
    assert_eq "$actual" 1
}

foobar_was_done() {
    actual="$(ctx_get foobar_done)"
    assert_eq "$actual" 1
}
~~~

## Document metadata

Some document metadata should end up in the typeset document,
especially the title, authors. The document date is more complicated,
to cater to different use cases:

* a work-in-progress document needs a new date for each revision
  - maintaining the `date` metadata field manually is quite tedious,
    so Subplot provides it automatically using the document source
    file modification time
  - some people would prefer a `git describe` or similar method for
    indicating the document revision, so Subplot allows the date to be
    specified via the command line
* a finished, reviewed, officiall stamped document needs a fixed date
  - Subplot allows this to be written as the `date` metadata field

The rules for what Subplot uses as the date or document revision
information are, then:

* if there is `date` metadata field, that is used
* otherwise, if the user gives the `--date` command line option, that
  is used
* otherwise, the markdown file's modification time is used

### Date given in metadata

This scenario tests that the `date` field in metadata is used if
specified.

~~~scenario
given file metadate.md
when I run sp-docgen metadate.md -o metadate.html
then file metadate.html exists
and file metadate.html contains "<title>The Fabulous Title</title>"
and file metadate.html contains "Alfred Pennyworth"
and file metadate.html contains "Geoffrey Butler"
and file metadate.html contains "WIP"
~~~

~~~{#metadate.md .file .markdown .numberLines}
---
title: The Fabulous Title
author:
- Alfred Pennyworth
- Geoffrey Butler
date: WIP
...
# Introduction
This is a test document. That's all.
~~~

### Date given on command line

This scenario tests that the `--date` command line option is used.

~~~scenario
given file dateless.md
when I run sp-docgen dateless.md -o dateoption.html --date=FANCYDATE
then file dateoption.html exists
and file dateoption.html contains "<title>The Fabulous Title</title>" 
and file dateoption.html contains "Alfred Pennyworth"
and file dateoption.html contains "Geoffrey Butler"
and file dateoption.html contains "FANCYDATE"
~~~

~~~{#dateless.md .file .markdown .numberLines}
---
title: The Fabulous Title
author:
- Alfred Pennyworth
- Geoffrey Butler
...
# Introduction
This is a test document. It has no date metadata.
~~~

### No date anywhere

This scenario tests the case of no metadata `date` and no command line
option, either. The date in the typeset document shall come from the
modification time of the input file, and shall have the date in ISO
8601 format, with time to the minute.

~~~scenario
given file dateless.md
and file dateless.md has modification time 2020-02-26 07:53:17
when I run sp-docgen dateless.md -o mtime.html
then file mtime.html exists
and file mtime.html contains "<title>The Fabulous Title</title>"
and file mtime.html contains "Alfred Pennyworth"
and file mtime.html contains "Geoffrey Butler"
and file mtime.html contains "2020-02-26 07:53"
~~~

### Missing bindings file

If a bindings file is missing, the error message should name the
missing file.

~~~scenario
given file missing-binding.md
when I try to run sp-docgen missing-binding.md -o foo.htmlh
then exit code is non-zero
and stderr matches /: missing-binding.yaml:/
~~~

~~~{#missing-binding.md .file .markdown .numberLines}
---
title: Missing binding
bindings: missing-binding.yaml
...
~~~

### Missing functions file

If a functions file is missing, the error message should name the
missing file.

~~~scenario
given file missing-functions.md
and file b.yaml
when I try to run sp-codegen --run missing-functions.md -o foo.py
then exit code is non-zero
and stderr matches /: missing-functions.py:/
~~~

~~~{#missing-functions.md .file .markdown .numberLines}
---
title: Missing functions
bindings: b.yaml
functions: missing-functions.py
...
~~~

### Extracting metadata from a document

The **sp-meta** program extracts metadata from a document. It is
useful to see the scenarios, for example. For example, given a
document like this:

sp-meta would extract this information from the **simple.md** example:

~~~
title: Test scenario
bindings: b.yaml
functions: f.py
scenario Simple
~~~

This scenario check sp-meta works. Note that it requires the bindings
or functions files.

~~~scenario
given file images.md
and file b.yaml
and file other.yaml
and file f.py
and file other.py
and file foo.bib
and file bar.bib
and file expected.json
when I run sp-meta images.md
then output matches /source: images.md/
and output matches /source: b.yaml/
and output matches /source: other.yaml/
and output matches /source: f.py/
and output matches /source: other.py/
and output matches /source: foo.bib/
and output matches /source: bar.bib/
and output matches /source: image.gif/
and output matches /bindings: b.yaml/
and output matches /bindings: other.yaml/
and output matches /functions: f.py/
when I run sp-meta images.md -o json
then JSON output matches expected.json
~~~


~~~{#images.md .file .markdown .numberLines}
---
title: Document refers to external images
bindings: 
- b.yaml
- other.yaml
functions: 
- f.py
- other.py
bibliography: [foo.bib, bar.bib]
...

![alt text](image.gif)
~~~

~~~{#other.yaml .file .yaml .numberLines}
[]
~~~

~~~{#other.py .file .python .numberLines}
~~~

~~~{#foo.bib .file .numberLines}
@book{foo2020,
 author    = "James Random",
 title     = "The Foo book",
 publisher = "The Internet",
 year      =  2020,
 address   = "World Wide Web",
}
~~~

~~~{#bar.bib .file .numberLines}
@book{foo2020,
 author    = "James Random",
 title     = "The Bar book",
 publisher = "The Internet",
 year      =  2020,
 address   = "World Wide Web",
}
~~~

~~~{#expected.json .file .json}
{
	"title": "Document refers to external images",
    "sources": [
	  "images.md",
	  "b.yaml",
	  "other.yaml",
	  "f.py",
	  "other.py",
	  "foo.bib",
	  "bar.bib",
	  "image.gif"
    ],
	"binding_files": [
	  "b.yaml",
	  "other.yaml"
	],
	"function_files": [
	  "f.py",
	  "other.py"
	],
	"bibliographies": [
	  "foo.bib",
	  "bar.bib"
	],
	"files": [],
	"scenarios": []
}
~~~


## Embedded files

Subplot allows data files to be embedded in the input document. This
is handy for small test files and the like.

Handling of a newline character on the last line is tricky. Pandoc
doesn't include a newline on the last line. Sometimes one is
needed&mdash;but sometimes it's not wanted. A newline can be added by
having an empty line at the end, but that is subtle and easy to miss.
Subplot helps the situation by allowing a `add-newline=` class to be added
to the code blocks, with one of three allowed cases:

* no `add-newline` class&mdash;default handling: same as `add-newline=auto`
* `add-newline=auto`&mdash;add a newline, if one isn't there
* `add-newline=no`&mdash;never add a newline, but keep one if it's there
* `add-newline=yes`&mdash;always add a newline, even if one is already
  there

The scenarios below test the various cases. 

### Extract embedded file

This scenario checks that an embedded file can be extracted, and used
in a subplot.

~~~scenario
given file embedded.md
when I run sp-docgen embedded.md -o foo.html
then file foo.html exists
and file foo.html matches /embedded\.txt/
~~~

~~~~~~~{#embedded.md .file .markdown .numberLines}
---
title: One embedded file
...

~~~{#embedded.txt .file}
This is the embedded file.
~~~
~~~~~~~


### Extract embedded file, by default add missing newline

This scenario checks the default handling: add a newline if one is
missing.

~~~scenario
given file default-without-newline.txt
then default-without-newline.txt ends in one newline
~~~

~~~{#default-without-newline.txt .file .numberLines}
This file does not end in a newline.
~~~


### Extract embedded file, by default do not add a second newline

This scenario checks the default handling: if content already ends in
a newline, do not add another newline.

~~~scenario
given file default-has-newline.txt
then default-has-newline.txt ends in one newline
~~~

~~~{#default-has-newline.txt .file .numberLines}
This file ends in a newline.

~~~

### Extract embedded file, automatically add missing newline

Explicitly request automatic newlines, when the file does not end in
one.

~~~scenario
given file auto-without-newline.txt
then auto-without-newline.txt ends in one newline
~~~

~~~{#auto-without-newline.txt .file add-newline=auto .numberLines}
This file does not end in a newline.
~~~


### Extract embedded file, do not automatically add second newline

Explicitly request automatic newlines, when the file already ends in
one.

~~~scenario
given file auto-has-newline.txt
then auto-has-newline.txt ends in one newline
~~~

~~~{#auto-has-newline.txt .file add-newline=auto .numberLines}
This file ends in a newline.

~~~


### Extract embedded file, explicitly add missing newline

Explicitly request automatic newlines, when the file doesn't end with
one.

~~~scenario
given file add-without-newline.txt
then add-without-newline.txt ends in one newline
~~~

~~~{#add-without-newline.txt .file add-newline=yes .numberLines}
This file does not end in a newline.
~~~


### Extract embedded file, explicitly add second newline

Explicitly request automatic newlines, when the file already ends with
one.

~~~scenario
given file add-has-newline.txt
then add-has-newline.txt ends in two newlines
~~~

~~~{#add-has-newline.txt .file add-newline=yes .numberLines}
This file ends in a newline.

~~~



### Extract embedded file, do not add missing newline

Explicitly ask for no newline to be added.

~~~scenario
given file no-adding-without-newline.txt
then no-adding-without-newline.txt does not end in a newline
~~~

~~~{#no-adding-without-newline.txt .file add-newline=no .numberLines}
This file does not end in a newline.
~~~


### Fail if the same filename is used twice

~~~scenario
given file onefiletwice.md
when I try to run sp-docgen onefiletwice.md -o onefiletwice.html
then exit code is non-zero
and file onefiletwice.html does not exist
~~~

~~~~{#onefiletwice.md .file .markdown .numberLines}
---
title: Two embedded files with the same name
...

```{#filename .file}
This is the embedded file.
```

```{#filename .file}
This is another embedded file, and has the same name.
```
~~~~


### Fail if two filenames only differ in case

~~~scenario
given file casediff.md
when I try to run sp-docgen casediff.md -o casediff.html
then exit code is non-zero
and file casediff.html does not exist
~~~

~~~~{#casediff.md .file .markdown .numberLines}
---
title: Two embedded files with names differing only in case
...

```{#filename .file}
This is the embedded file.
```

```{#FILENAME .file}
This is another embedded file, and has the same name in uppercase.
```
~~~~

## Steps must match bindings

Subplot permits the binding author to define arbitrarily complex regular
expressions for binding matches.  In order to ensure that associating steps
to bindings is both reliable and tractable, a step must match _exactly one_
binding.

```{#badbindings.yaml .file .yaml}
- given: a binding
  function: a_binding
- given: a (?:broken)? binding
  function: a_broken_binding
  regex: true
- given: a capitalised Binding
  function: os.getcwd
  case_sensitive: true
```

### Steps which do not match bindings do not work

~~~~{#nobinding.md .file .markdown}
---
title: No bindings available
bindings:
- badbindings.yaml
...
# Broken scenario because step has no binding

```scenario
given a missing binding
then nothing works
```
~~~~

```scenario
given file nobinding.md
and file badbindings.yaml
when I try to run sp-codegen --run nobinding.md -o test.py
then exit code is non-zero
```

### Steps which do not case-sensitively match sensitive bindings do not work

~~~~{#casemismatch.md .file .markdown}
---
title: Case sensitivity mismatch
bindings:
- badbindings.yaml
...
# Broken scenario because step has a case mismatch with sensitive binding

```scenario
given a capitalised binding
```
~~~~

```scenario
given file casemismatch.md
and file badbindings.yaml
when I try to run sp-codegen --run casemismatch.md -o test.py
then exit code is non-zero
```

### Steps which match more than one binding do not work

~~~~{#twobindings.md .file .markdown}
---
title: Two bindings match
bindings:
- badbindings.yaml
...
# Broken scenario because step has two possible bindings

```scenario
given a binding
```
~~~~

```scenario
given file twobindings.md
and file badbindings.yaml
when I try to run sp-codegen --run twobindings.md -o test.py
then exit code is non-zero
```


### List embedded files

The `sp-meta` command lists embedded files in its output.

~~~scenario
given file two-embedded.md
when I run sp-meta two-embedded.md
then output matches /foo.txt/
and output matches /bar.yaml/
~~~

~~~~~~{#two-embedded.md .file .markdown .numberLines}
---
title: Two embedded files
...

~~~{#foo.txt .file}
~~~

~~~{#bar.yaml. .file}
~~~
~~~~~~


## Embedded graphs

Subplot allows embedding markup to generate graphs into the Markdown document.

### Dot

[Graphviz]: http://www.graphviz.org/

Dot is a program from the [Graphviz][] suite to generate directed
graphs, such as this one.

~~~dot
digraph "example" {
thing -> other
}
~~~

The scenario checks that a graph is generated and embedded into the
HTML output, not referenced as an external image.

~~~scenario
given file dot.md
and file b.yaml
when I run pandoc --filter sp-filter dot.md -o dot.html
then file dot.html matches /img src="data:image/svg\+xml;base64,/
~~~

The sample input file **dot.md**:

~~~~~~~~{#dot.md .file .markdown .numberLines}
This is an example Markdown file, which embeds a graph using dot markup.

~~~dot
digraph "example" {
thing -> other
}
~~~
~~~~~~~~



### PlantUML

[PlantUML]: https://plantuml.com/

[PlantUML][] is a program to generate various kinds of graphs for
describing software, such as this one:

~~~plantuml
@startuml
Alice -> Bob: Authentication Request
Bob --> Alice: Authentication Response

Alice -> Bob: Another authentication Request
Alice <-- Bob: Another authentication Response
@enduml
~~~

The scenario below checks that a graph is generated and embedded into
the HTML output, not referenced as an external image.

~~~scenario
given file plantuml.md
and file b.yaml
when I run pandoc --filter sp-filter plantuml.md -o plantuml.html
then file plantuml.html matches /img src="data:image/svg\+xml;base64,/
~~~

The sample input file **plantuml.md**:

~~~~~~~~{#plantuml.md .file .markdown .numberLines}
This is an example Markdown file, which embeds a graph using
PlantUML markup.

~~~plantuml
@startuml
Alice -> Bob: Authentication Request
Bob --> Alice: Authentication Response

Alice -> Bob: Another authentication Request
Alice <-- Bob: Another authentication Response
@enduml
~~~
~~~~~~~~


### Roadmap

Subplot supports visual roadmaps using a YAML based markup language,
implemnted by the [roadmap][] Rust library. The library converts the
roadmap into dot, and that gets rendered as SVG and embedded in the
output document by Subplot.

An example:

~~~roadmap
goal:
  label: |
    This is the end goal:
    if we reach here, there
    is nothing more to be
    done in the project
  depends:
  - finished
  - blocked

finished:
  status: finished
  label: |
    This task is finished;
    the arrow indicates what
    follows this task (unless
    it's blocked)

ready:
  status: ready
  label: |
    This task is ready 
    to be done: it is not
    blocked by anything

next:
  status: next
  label: |
    This task is chosen 
    to be done next

blocked:
  status: blocked
  label: |
    This task is blocked
    and can't be done until
    something happens
  depends:
  - ready
  - next
~~~

This scenario checks that a graph is generated and embedded into the
HTML output, not referenced as an external image.

~~~scenario
given file roadmap.md
and file b.yaml
when I run pandoc --filter sp-filter roadmap.md -o roadmap.html
then file roadmap.html matches /img src="data:image/svg\+xml;base64,/
~~~

The sample input file **roadmap.md**:

~~~~~~~~{#roadmap.md .file .markdown .numberLines}
This is an example Markdown file, which embeds a roadmap.

~~~roadmap
goal:
  label: |
    This is the end goal:
    if we reach here, there
    is nothing more to be
    done in the project
  depends:
  - finished
  - blocked

finished:
  status: finished
  label: |
    This task is finished;
    the arrow indicates what
    follows this task (unless
    it's blocked)

ready:
  status: ready
  label: |
    This task is ready 
    to be done: it is not
    blocked by anything

next:
  status: next
  label: |
    This task is chosen 
    to be done next

blocked:
  status: blocked
  label: |
    This task is blocked
    and can't be done until
    something happens
  depends:
  - ready
  - next
~~~
~~~~~~~~


### Class name validation

When Subplot loads a document it will validate that the block classes
match a known set.  Subplot has a built-in set which it treats as special,
and it knows some pandoc-specific classes and a number of file type classes.

If the author of a document wishes to use additional class names then they can
include a `classes` list in the document metadata which subplot will treat
as valid.

~~~scenario
given file unknown-class-name.md
and file known-class-name.md
and file b.yaml
when I try to run sp-docgen unknown-class-name.md -o unknown-class-name.html
then exit code is non-zero
and file unknown-class-name.html does not exist
and stderr matches /Unknown classes found in the document: foobar/
when I run sp-docgen known-class-name.md -o known-class-name.html
then file known-class-name.html exists
~~~

~~~~~~~~{#unknown-class-name.md .file .markdown .numberLines}
---
title: A document with an unknown class name
...

```foobar
This content is foobarish
```

~~~~~~~~

~~~~~~~~{#known-class-name.md .file .markdown .numberLines}
---
title: A document with a previously unknown class name
classes:
- foobar
...

```foobar
This content is foobarish
```

~~~~~~~~

## Using as a Pandoc filter

Subplot can be used as a Pandoc _filter_, which means Pandoc can allow
Subplot to modify the document while it is being converted or typeset.
This can useful in a variety of ways, such as when using Pandoc to
improve Markdown processing in the [ikiwiki][] blog engine.

[ikiwiki]: http://ikiwiki.info/

The way filters work is that Pandoc parses the input document into an
abstract syntax tree, serializes that into JSON, gives that to the
filter (via the standard input), gets a modified abstract syntax tree
(again as JSON, via the filter's standard output).

Subplot supports this via the **sp-filter** executable. It is built
using the same internal logic as Subplot's docgen. The interface is
merely different to be usable as a Pandoc filter.

This scenarios verifies that the filter works at all. More
importantly, it does that by feeding the filter a Markdown file that
does not have a YAML metadata block. For the ikiwiki use case, that's
what the input files are like.

~~~scenario
given file justdata.md
when I run pandoc --filter sp-filter justdata.md -o justdata.html
then file justdata.html matches /does not have a YAML metadata/
~~~

The input file **justdata.md**:

~~~~~~~~{#justdata.md .file .markdown .numberLines}
This is an example Markdown file.
It does not have a YAML metadata block.
~~~~~~~~


# Extract embedded files

`sp-extract` extracts embedded files from a subplot file.

~~~scenario
given file embedded-file.md
and file expected.txt
when I run sp-extract embedded-file.md foo.txt -d .
then files foo.txt and expected.txt match
~~~

~~~~~~{#embedded-file.md .file .markdown .numberLines}
---
title: Embedded file
...

~~~{#foo.txt .file}
This is a test file.
~~~

~~~~~~

~~~{#expected.txt .file}
This is a test file.
~~~




---
title: "Subplot"
author: The Subplot project
bindings: 
- subplot.yaml
- runcmd.yaml
- files.yaml
functions: 
- subplot.py
- runcmd.py
- files.py
classes:
- json
...