Skip to content

pyrustic/backstage

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

40 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Demo

By © Jorge Royan / http://www.royan.com.ar, CC BY-SA 3.0, Link

Pyrustic Backstage

Three-speed scripting language and task automation tool

This project is part of the Pyrustic Open Ecosystem.

Installation     Demo     Latest     Modules

Table of contents

Overview

Backstage is a cross-platform automation tool that looks for a backstage.tasks file in the current working directory to run a specific task defined in that file on demand. A task can be a sequence or pipeline of processes to be spawned, instructions for performing file and directory manipulation, or something more sophisticated.

The backstage.tasks file uses the Jesth (Just Extract Sections Then Hack) file format that acts like a broken INI file parser that only extract sections each made of a header and a body which is just a list of lines.

Using an eponymous three-speed scripting language designed for the automation tool, the programmer can, inside the backstage.tasks file, define, coordinate and use the various resources at his disposal to automate things.

The three-speed scripting language concept is inspired from the three forward gear ratios of early automobiles transmission system. In the following subsections, we will explore each metaphorical gear of the Backstage scripting language, then we will briefly expose the automation tool itself.

First gear

In first gear, a backstage.tasks file is intended to store a list of tasks related to a specific project, each task exposing a list of commands or subtasks to be executed. Here, a command represents a process or pipeline of processes to be spawned. Environment variables can be used in commands via variable interpolation. No other logic is involved.

Example

[task1]
# three commands to run sequentially
$ git commit -m 'Update'
$ python -m my.package.module
$ program1 arg1 | program2 {HOME}
---
# run the subtask 'task2'
& task2
---
# run the subtask 'task3' in a new thread
~ task3

[task2]
$ program val1 {CWD}
$ git push origin master

[task3]
# some heavy computation
$ engine -x 5000
$ engine --cleanup

[_task4]
# This is a private task (with an underscore as prefix)
$ clean dir

Second gear

In second gear, the backstage.tasks file not only stores the tasks like in first gear, but here logic intervenes, variables are defined, control flow is used, built-in commands are called, et cetera. Basically, in second gear, Backstage unleashes its power and allows the programmer to anticipate problems, make sophisticated combinations of subtasks, in short to write a real script to automate things.

Example

[task1]
# commit changes
$ git commit -m 'Update'
---
# tell user if 'Commit' has been success
if R == 0
    # print 'Success !'
    : Success !
else
    : Failed to commit changes
---
# say hello ten times
set age = 42
set name = `John Doe`
from 1 to 10
    $ python -m say.hello {name} {age}
---
# create a file in user home
set pathname = {HOME}/iliad.txt
create file pathname
---
# append some data to a file
set data = `\nHello World !`
append data to pathname
---
# browse current working directory
browse files and dirs in CWD
    : Directory -> {R}
    : Files ->
    for item in files
        : - {item}
    : Dirs ->
    for item in dirs
        : - {item}

Third gear

In third gear, in addition with whatever can be done with previous gears, the programmer can directly from the backstage.tasks file, call Python functions with arguments and get the return ! Thanks to this third gear, any too complex or overly verbose calculation can be written in Python and called from Backstage. This functionality alone proves that Backstage is all about making the programmer's life better, not pretending to replace existing mature solutions that actually work.

Example

[task1]
interface with package.coffeemaker alias coffeebro

set sugar_cubes (int) = 1 + 1 + 1
set extra = `milk`

call coffeebro.make(sugar_cubes, extra, 42)

if R == 1
    : Coffee successfully made !
else
    : Oops, failed to make coffee...
    : Exception -> {EXCEPTION}
    : Traceback -> {TRACEBACK}

Automation tool

The scripting language help to define tasks in the backstage.tasks file that is intended to be consumed by the automation tool. As an automation tool, Backstage exposes a command line interface that allows the user to discover available tasks, run a task with arguments, read a task documentation, use a glob-like syntax to search for a task by its name or by a keyword (case-insensitive mode) that is part of the documentation of the task, et cetera.

Example 1

Let's assume that there is a backstage.tasks file in the directory /home/alex/project. This backstage.tasks file contains three public tasks and one private tasks (prefixed with an underscore).

This example shows how one could use the automation tool to run a task defined in the backstage.tasks file:

$ cd /home/alex/project

$ backstage -c
Available tasks (3):
    make_coffee  task1  task2

$ backstage make_coffee
Making coffee...

$ backstage mak*
Making coffee...

$ backstage make_coffee sugar=3
Making coffee with 3 sugar cubes...

Example 2

This is the contents of a backstage.tasks file located at /home/alex/project:

[task]
# define 'x' as a variable with an 'int' assignment tag
set x (int) = 1 + 1 + 1

# default name (by default, the assignment tag is 'str')
set default_name = `John Doe`

# print some text
: Hi and Welcome !
: I can print your name {x} times in a row !

# take user input (always a 'str')
> name : `What is your name ? `

# control flow
if name == EMPTY
    set name = {default_name}

# iteration
from 1 to x
    : {R} - Ave {name} !

# branch subtask '_task2' and pass it an argument
& _task2 {name}


[_task2]
# define the variable 'name' (first argument passed to this task)
set name = {ARGS[0]}

# Just say Goodbye !
: Goodbye {name} !

Let's run the task named task from the command line:

$ cd /home/alex/project

$ backstage task
Hi and Welcome !
I can print your name 3 times in a row !
What is your name ? Alex
1 - Ave Alex !
2 - Ave Alex !
3 - Ave Alex !
Goodbye Alex !

Note: You can reproduce this example as it. Just install Backstage, copy-paste the script in a backstage.tasks file, then run backstage task in the command line.

And more

There is more to talk about Backstage, like the ability to embed documentation and tests in the backstage.tasks file and access them from the command line. Backstage is enough versatile to do the job of a trivial task runner or to automate things with its scripting language that has a built-in bridge to the powerful Python ecosystem.

In the following sections, we will explore this project in depth. You can also jump to the demo to start playing with Backstage !

Structure of the script file

As stated in the Overview section, a backstage.tasks file is basically a JesthFile.

In a backstage.tasks file, a section represents a task. The section title is the name of the task and the section body is made of commands to run and the constructs of the Backstage scripting language. A valid task name is an alphanumeric string that can contains an underscore.

You can write here at the top of the script,
a description of the script,
the date of its creation,
or any useful information.

[task1]
# body of task1
...

[task2]
# body of task2
...

[_private]
# prefix a task name with an underscore
# to turn it into a private task that
# won't appear in the list of available tasks
# when you will type 'backstage --check' in the command line

As you can guess, a line starting with # is a comment. But this is only true inside a task body, because in fact not all sections are tasks as described before: a section can also be an embedded test or documentation.

Embedded documentation

You can embed documentation inside the backstage.tasks file. To create a documentation for a task, create a section which name is postfixed with .help:

[task1]
pass


[task1.help]
This is the description line.

Usage:
    backstage task <option> <path>

Options:
    -m, --msg       Show message
    -x, --exit      Exit blah blah

From the command line, you can read the documentation of an arbitrary task:

$ backstage -h task1
This is the description line.

Usage:
    backstage task <option> <path>

Options:
    -m, --msg       Show message
    -x, --exit      Exit blah blah

Embedded tests

You can embed tests inside the backstage.tasks file. To create a test for a task, create a section which name is postfixed with .test:

[task1.test]
# perform some test here

# ...

assert some_var == some_var

From the command line, you can run the test of an arbitrary task:

$ backstage --test task1

or the tests of a bunch of tasks:

$ backstage --test task1 task2 task3

or run all tests defined in the backstage.tasks file:

$ backstage --test

Spawning processes and branching subtasks

You can write commands to spawn a process or a pipeline of processes:

[task]
# spawn Git to perform a 'commit'
$ git commit -m "Update"

# spawn a pipeline of three processes
$ program1 arg1 | program2 arg2 | program3

Commands to spawn processes support variable interpolation:

[task]
# use HOME environment variable
$ ls {HOME}

# access the first index of the ARGS list
$ program {ARGS[0]}

You can push data to the input of a process:

[task]
# define variables name and age
set name = `John Doe`
set age (int) = 40 + 2

# name and age will be pushed to the input of the next spawned process
push name age
$ program1 {HOME}

# from now you can access via the environment variable R,
# the return code (exit status code)
: The return code is {R}

You can capture a process:

[task]
# you can capture a process,
# so you will get a direct access to the output and error
($) program2

# print the content of OUTPUT and ERROR
: Output -> {OUTPUT}
: Error -> {ERROR}

You can redirect STDOUT and STDERR:

[task]

# redirect STDOUT and STDERR
set STDOUT = `/path/to/file_out`
set STDERR = `/path/to/file_err`

$ program1

---

# cross platform DEVNULL:
set STDOUT = /dev/null

$ program2

Branching subtasks

You can branch a subtask and pass arguments to it:

[task1]
# branch 'task2'
& task2 {HOME}

[task2]
# from task2, we can access arguments passed to it.
# ARGS (environment variable) is a list of arguments.
: Arguments -> {ARGS}

Whenever you branch a subtask, a new instance of this subtask is created, with its own variables. You can share data with a subtask with one of these three ways:

  • pass arguments to the subtask while branching it;
  • use the global namespace;
  • use the database namespace.

A subtask can also return data that is cached in the R environment variable:

[task1]
& task2 "John Doe"
# from now, R contains 'Hello John Doe !'

[task2]
set result = `Hello {ARGS} !`
return result

Multithreading

Branching a subtask is done in the main thread. But one can create a new thread for a subtask:

[task1]
# run an instance of task2 in a new thread
~ task2

# the next command won't wait 'task2' to complete
$ git commit -m "Update"


[task2]
# this task sleeps for 5 seconds
sleep 5

Data types and control flow

In the next subsections we will talk about data types then control flow.

Data types

Backstage supports variables and let the programmer set, use, clear, and drop variables.

Under the hood, Backstage works with five Python data types: str, list (one-dimensional), dict (one-dimensional), int, and float. But these data types aren't intended to be directly used by the programmer. Instead, assignment tags are used to tell the interpreter how to treat a variable.

These assignment tags are: raw, str, list, dict, int, float, date, time, dtime, and tstamp.

[task]
# this is a string
set var = 42

# this is another string
set var (str) = `Hello World`

# this is a list
set var (list) = `reg green blue`

# this is the same list but edited
set var[0] = `yellow`
# var -> yellow green blue

# this is a dict
set var (dict) = `user="John Doe" age=42 location=Kernel`

# this is the same dict but edited
set var.user = `Jane Doe`
# var -> user='Jane Doe' age=42 location=Kernel

# this is an integer
set var (int) = 40 + 2
# var -> 42

# this is a raw string
set regex (raw) = `[\S\s]*?`

# get the current timestamp
set now (int) = {NOW}
# now -> 1662569326

# convert it into datetime
set var (dtime) = {now}
# var -> 2022-09-07 17:48:46

# go from a datetime to timestamp
set var (tstamp) = `2022-09-07 17:48:46`
# var -> 1662569326

# extract the time part of a timestamp
set var (time) = 1662569326
# var -> 17:48:46

Note that all variables have a string representation and backticks are used optionally as delimiters that will be ignored by the interpreter. So you can put backticks around an integer, and you can insert a list in a string.

Here document

Backstage supports here document for strings inside the backstage.tasks file:

[task]
# this is a text with two lines
set text (str) = `First line\nSecond line`

# this is another text with three lines
set text = `January\nFebruary\nMarch`

# this is not a text with two lines
set var (raw) = `Hello\nWorld`

# this is not a text with two lines
set var (str) = `Hello\\nWorld`

Variable interpolation

Variable interpolation is supported with the ability to access from a list or a string, the value at an arbitrary index, or from a dictionary, the value of a key.

[task]
# let's play with 'str' variables
set x = `red`
set var = `{x} green blue`
# var -> red green blue

# get the value of element at index 0
set value = {var[0]}
# value -> r

---

# let's play with a list
set x = `red`
set var (list) = `{x} green blue`

# get the value of element at index 0
set value = {var[0]}
# value -> red

# get the value of elements from index 1 to the end
set value = {var[1:]}
# value -> green blue

---

# let's play with a dict
set x (int) = 40 + 2
set var (dict) = `name="John Doe" age={x}`

# get the value of key 'name'
set value = {var.name}
# value -> John Doe

# get the value of key 'age'
set value = {var.age}
# value -> 42

---

# cancel the variable interpolation
set var1 = `Hello` 
set var2 = `{{var1}} World`
# var2 -> {var1} World

Control flow

Backstage implements conditionals and loops. A wide range of operators are available to compare values.

Conditionals

[task]
set var1 = 1
set var2 = 1
set x = 2
set regex (raw) = `[\S\s]*?`
set text = `Hello world`
set y = `Hello`

# conditionals support the classic
# operators: == != > < >= <= 
if var1 == 1
    $ program1
elif var2 == x
    $ program2
else
    $ program3
   
# Backstage supports logical and/or 
if var1 == var2 and var1 >= 3
    $ program1

# use the 'matches' operator
# or the negated one: !matches
if regex matches text
    : Matched !
elif regex !matches text
    : Mismatched !
    
# you can use the in operator
# and also the negated one: !in
if y in text
    pass
elif y !in text
    pass

Loops

[task]

# From To loop
from 10 to 0
    : {R}
    
# For loop - iterate over a string
set text = `Hello World`
for char in text
    # N is an environment variable that serves as counter
    # for all loops
    : {N}- {char}

# For loop - iterate over a list
set data (list) = `red green blue`
for item in data
    : {item}
    
# For loop - iterate over a dict
set data (dict) = `name="John Doe" age=42`
for item in data
    : key -> {item[0]}  value -> {item[1]}
    
# For loop - iterate over a file
set path = `/home/alex/iliad.txt`
for line in path (file)
    : Line {N}
    : {line}
    :
    
# while loop
set var = 1
while var == 1
    : One Time Hello
    break
    
# browse loop
set path = `/home/alex`
browse files and dirs in path
    : Directory -> {R}
    for item in files
        : {item}
    for item in dirs
        : {item} 

Namespaces and persistence

Namespaces are implemented in Backstage to provide convenient management of variables by defining three scopes:

  • L for Local scope;
  • G for Global scope;
  • D for Database scope.

By default, variables exist in the Local namespace and are only accessible to the running task.

To share data with a subtask, one can expose arbitrary variables that will be copied into the Global namespace which is readable and writable (thread-safe) by all subtasks.

[task]
# by default, variables are defined in Local,
# i.e. they are only visible in the scope of the current task
set var = 42
: Var contains {var}
: Var still contains {L:var}

# you can make local variables public
expose var
# from now, you can get a thread-safe access to var 
# from any running task:
: Global var contains {G:var}

# branch _task2
& _task2

[_task2]
: I got {G:var} !

Data can also be persisted:

[task]
set var = 42
store var

# from now, 'var' can be accessed by all tasks
# in this runtime but also in future runtimes
: Var contains {D:var}

# if you aren't sure about the existence
# of a variable, just do this:
default var

# it also works with a bunch of variables:
default var1 L:var2 D:var3 G:var4

# from now, if 'var1' hasn't been manually defined
# by the programmer, it will be
# automatically initialized and
# its value will be an empty string

Persisted variables are stored in .backstage/database.json.

Variable interpolation and escaping

During the string interpolation of a command that spawn processes or branch a subtask, variables that are of the str type are automatically shell-escaped.

[task]
# This is a 'str' variable (backticks aren't quotes, by the way!)
set name = `John Doe` 
set colors (list) = `red green blue`

: Welcome {name} !
# Welcome John Doe

$ program name={name} -c {colors}
# program name='John Doe' -c red green blue

# Notice the quotes automatically added around the name John Doe

Environment variables and language syntax

Environment variables are local to each instance of task. They are defined as uppercase strings. One can edit their contents but can't create new environment variables.

This is the exhaustive list of environment variables:

Variables Description
ARGS List of arguments passed to this task from the command line
CWD Current working directory
DATE The current date in the YYYY-MM-DD format
EMPTY Just an empty string
ERROR Error string from a process previously spawned
EXCEPTION Name of the last exception raised
FALSE The integer 0
HOME The path to $HOME. Example: /home/alex
LINE The current line (1-based numbering) of execution in the task body
N Counter for while, for, from, and browse loops
NOW Current timestamp in seconds
ONE The integer 1
OS The running operating system: aix, linux, win32, cygwin, darwin
OUTPUT Output string from a process previously spawned
R The return of Python functions, built-in commands, statements, constructs, or process exit status codes
RANDOM Random integer between 0 and 255 (closed interval)
SPACE One space character
STDERR Use this variable to perform STDERR redirection
STDIN Use this variable to perform STDIN redirection
STDOUT Use this variable to perform STDOUT redirection
TASK The name of the currently running task
TIME The current time in the HH:MM:SS format
TIMEOUT Timeout in seconds for commands that spawn processes. Default value: 30 seconds
TMP Temporary directory. Attention, this directory will automatically disappear at the end of the runtime ! So think twice before moving files inside
TRACEBACK Traceback of the last exception raised
TRASH Path to the trash: $HOME/PyrusticData/trash
TRUE The integer 1
ZERO The integer 0

Note that the TRACEBACK and EXCEPTION variables are cleared after the next successful command. Backstage also generates for convenience, ARG0, ARG1, ARGx, according to the contents of ARGS. For example, if ARGScontains two arguments, you can expect that ARG0 and ARG1 exist.

Language syntax

In this section we will explore the built-in commands, statements, keywords, symbols, and language constructs that make Backstage.

Note that wherever a built-in command or statement expects a variable that will be read, for convenience you can instead of supplying a variable name, define an inline int or float literal.

Also, consider that the R environment variable is your friend, since it is used to cache the data returned by a statement, a construct, or a command.

APPEND

Append data to a file.

Usage: append <var> to <filename_var>

ASSERT

Test is a condition is true.

Usage: assert (<var1>|<regex_var>) (==|!=|<=|>=|<|>|in|!in|rin|!rin|matches|!matches) <var2> [and|or] ...

Example: assert regex_var matches text_var and var1 in list

Note that ! is used to express negation and rin is a Regex-based in. A regexly-in operator ;)

BRANCH

Branch a subtask. The syntax is similar to the one to spawn processes, i.e., a string of words. The syntax supports variable interpolation.

Usage: & <subtask> [<argument> ...]

Example: & subtask1 name="John Doe" age=42 city={city}

In this example, the city variable will be automatically shell-escaped during its interpolation.

BREAK

Break a loop.

Usage: break

BROWSE

Loop construct to browse a directory.

Usage: browse [files] [and] [dirs] in <dirname_var>

Example:

[task]
browse files and dirs in dirname
    : Root -> {R}
    for item in files
        : {item}
    for item in dirs
        : {item}
        
browse files in dirname
    pass
    
browse dirs in dirname
    pass

CALL

Call a Python function from Backstage with arguments, then get the return !

Usage: call <module>.<function>[(<argument_var>, ...)]

Example:

[task]
# interface with the Python module
interface with package.coffee_module alias coffeemaker

# call the 'make' function with arguments then get the return
call coffeemaker.make(sugar_cube, extra, 42)
: Result -> {R}

CD

Change directory.

Usage: cd <dirname_var>

CHECK

Return the data type (str, list, dict, int, float) of a variable if it exists, else return an empty string.

Usage: check <var>

Example:

[task]
check myvar
if R == EMPTY
    : This variable doesn't exist at all !
else
    : 'myvar' exists, its data type is {R}

CLEAR

Clear the content of a variable or a list of variables.

Usage: clear <var> ...

Example: clear var1 var2 var3

COMMENT

Comment.

Usage: # <comment>

CONFIG

Read and write configuration options (FailFast, ReportException, ShowTraceback, TestMode, AutoLineBreak).

Usage: config <option> ...

Example:

[task]
config FailFast=1 AutoLineBreak=0
config TestMode
if R == 1
    : Test Mode On
elif R == 0
    : Test Mode Off

COPY

Copy a file or a directory tree to a new destination.

Usage: copy <src_path_var> to <dest_path_var>

COUNT

Count chars, items, and lines in the content of a variable or inside a file (if the (file) tag is applied).

Usage: count (chars|items|lines) in (<var>|<filename_var>) [(file)]

Example:

[task]
set path = /home/alex/iliad.txt
count chars in path (file)
if R == 0
    : The file is empty !

CREATE

Create a new file or directory.

Usage: create (dir|file) <path_var>

DEFAULT

Define an empty variable (or a bunch of variables) if it doesn't exist yet in the namespace.

Usage: default <var> ...

Example:

[task]
# default two variables in the Local namespace
default var1 L:var2

# default one variable in the Database namespace
default D:name

# from now, 'D:name' can be safely accessed
# even though the 'database.json' file supposed
# to contain the 'name' value was inadvertently deleted.

DROP

Destroy a variable (or a bunch of variables).

Usage: drop <var> ...

ELIF

Part of the if conditional construct.

Usage: elif (<var1>|<regex_var>) (==|!=|<=|>=|<|>|in|!in|rin|!rin|matches|!matches) <var2> [and|or] ...

ELSE

Part of the if conditional construct.

Usage: else

ENTER

Invite user to submit data.

Usage: > [<var> [: <text>]]

Example:

[task]

> name : Please enter your name 
# have you spotted the space at the end the line above ?

# the same line can be rewritten like this:
> name : `Please enter your name `
# backticks serve as delimiters that will be ignored

# this one is also possible:
set msg = `Please enter your name `
> name : {msg}

# even this:
set info (dict) = name="John Doe" age=42
> info.name : `Please enter your name`

EXIT

Exit.

Usage: exit

EXPOSE

Copy a variable (or a bunch of variables) into the Global namespace.

Usage: expose <var> ...

FAIL

Deliberately fail. It breaks the running task and mark it as a failure.

Usage: fail

FIND

Find files and or directories paths.

Usage 1: find [all] (paths|files|dirs) in <dirname_var>

Usage 2: find ... matching <regex_var>

Usage 3: find ... [and] (accessed|modified|created) (at|after|before|between) <timestamp_var> [and <timestamp_var>]

Example: find files in dirname matching regex and accessed between timestamp1 and timestamp2

FOR

A for loop to iterate the content of a variable or the content of a file (if you apply the (file) tag).

Usage: for (char|item|line) in (<var>|<filename_var>) [(file)]

Example:

[task]
# this code iterates over each character of the Iliad,
# and outputs it as it,
# with one twist: each line starts with its index (0-based)

set path = `/home/alex/iliad.txt`

# the print statement (:) won't anymore
# automatically add a line break !
config AutoLineBreak=0

for line in path (file)
    : `{N} `
    for char in line
        : {char}
    : \n

FROM

A loop to go from an integer to another one. If the start integer is superior to the end integer, the count will decrease.

Usage: from <start> to <end>

Example:

[task]
from 10 to 0
    # here, N will go from 0 to 10
    # but R will go from 10 to 0
    # because N is a counter for all loops
    # while R is a cache for whatever is returned
    # by a command, a statement, or a construct
    : {N}\t{R}

# As you can see, I can add a Tab \t since
# Backstage supports natively here document ;)

# To get a simple backslash followed by a 't':  \\t

GET

Get a char, an item, or a line at index x (including negative index) from a target. The target can be the content of a variable or a file (if you apply the (file) tag).

Usage: get (char|item|line) <index> from (<var>|<filename_var>) [(file)]

IF

Conditional construct.

Usage: if (<var1>|<regex_var>) (==|!=|<=|>=|<|>|in|!in|rin|!rin|matches|!matches) <var2> [and|or] ...

Example:

[task]
default var1 var2 var3 var4
if var1 == var2 or var3 == var4
    pass
elif EMPTY == EMPTY
    pass
else
    pass

INTERFACE

Interface with a Python module.

Usage: interface with [<package>.]<module> [alias <name>]

Example:

[task]
interface with package.mymodule alias module
default var1 var2
call module.function(var1, var2)

LINE

Draw a line.

Usage: (=|-) ...

Example: ---------- or ==========

MOVE

Move a file or a directory tree to a new destination.

Usage: move <src_path_var> to <dest_path_var>

PASS

Placeholder for the code that you might write in the future. This statement does nothing. It is the same as the eponymous one in Python.

Usage: pass

POKE

Poke a file or directory to get access to a dict of properties if this path exists. Available properties: size mtime ctime atime nlink uid gid mode ino dev.

Usage: poke <path_var>

Example:

[task]
set path = /home/alex/iliad.txt
poke path
if R == EMPTY
    : Oops ! This file doesn't exist
else
    : File size -> {R.size}

PREPEND

Prepend data to a file

Usage: prepend <var> to <filename_var>

PRINT

Print data. You can use backquotes as delimiters. This statement supports variable interpolation.

Usage: : <text>

Example:

[task]
:  Hello World ! 
# Have you spotted the two extra spaces characters ?

: ` Hello World ! `
# hehehe, got you ! ;)

PUSH

Push variables into the input of the next spawned process.

Usage: push <var> ...

READ

Read all or a specific line index (including negative index) from a file.

Usage: read (*|<index>) from <filename_var>

REPLACE

Replace some pattern in a text with a replacement value.

Usage: replace <regex_var> in <text_var> with <replacement_var>

RETURN

Return from a task with a value.

Usage: return [<var>]

SET

Define a new variable or update the content of an existing variable. You don't can't specify a data type, but instead you can apply an assigment tag that is one of: (raw) (str) (list) (dict) (int) (float) (date) (time) (dtime) (tstamp). Note that backquotes can be used as delimiters for the value (right side of the equal sign). These delimiters will be ignored. Backticks aren't quotes. This statement supports variable interpolation.

Usage: set <var> [(raw)|(str)|(list)|(dict)|(int)|(float)|(date)|(time)|(dtime)|(tstamp))] = <value>

Example: set var (int) = 1 + 2

SLEEP

Sleep for x seconds.

Usage: sleep <seconds>

SPAWN

Spawn a new process.

Usage: $ <program> [<argument> ...]

Example: $ program1 arg {var} | program2

SPLIT

Split with a regex pattern a text into a list.

Usage: split <text_var> with <regex_var>

SPOT

Count the number of occurrences of a regex pattern inside a text.

Usage: spot <regex_var> in <text_var>

STORE

Store a variable (or a bunch of variables) in the Database namespace. A stored variable can be accessed like this: D:var

Usage: store <var> ...

THREAD

Branch a subtask... but in a new thread.

Usage: ~ <subtask> [<argument> ...]

WHILE

The while loop. Use break to break it, and check N if you need a counter.

Usage: while (<var1>|<regex_var>) (==|!=|<=|>=|<|>|in|!in|rin|!rin|matches|!matches) <var2> [and|or] ...

WRITE

Erase the content of a file to write some data inside.

Usage: write <var> to <filename_var>

File and directory manipulation

Let's explore how file and directory manipulatin is performed with Backstage.

Resource creation

Create a file:

[task]
# create a file
set path = /home/alex/iliad.txt
create file path

Create a directory:

[task]
# create a directory
set path = /home/alex/new/directory
create dir path

File edition

[task]
set path = /home/alex/iliad.txt
set var = Hello World

# write data
write var to path

# append data to a file
append var to path

# prepend data to a file
prepend var to path

Read the content of a file

[task]
set path = /home/alex/iliad.txt

# read all from 'iliad.txt'
read * from path
: {R}

# read the line at index 3
set index (int) = 1 + 1 + 1
read index from path
: {R}

# just want to read the last line ?
read -1 from path 
: {R}

Iterating the content of a file

[task]
set path = /home/alex/iliad.txt

# iterate over the characters in a file
for char in path (file)
    : Character -> char

# iterate over the lines in a file
for line in path (file)
    : {line}

Browse a folder

[task]
set folder = /home/alex

browse files and dirs in folder
    : Directory -> {R}
    for item in files
        : {item}
    for item in dirs
        : {item}

Find resources

The find statement is like Glob but on steroid:

[task]
set folder = /home/alex
set regex (raw) = `[\S\s]*?`
set timestamp1 = 1223322233

find files in folder matching regex and accessed between timestamp1 and NOW
: Results -> {R}

Read resource properties

You can get from Backstage the properties of an arbitrary resource, like its size:

[task]
set path = /home/alex/iliad.txt

# poke a file
poke path
: Creation timestamp -> {R.ctime}
: Size -> {R.size}

Interfacing with Python

Interfacing with Python is as simple as this:

[task]
interface with python.module as my_module
set name = `John Doe`
set age (int) = 40 + 2
call my_module.function(name, age)
: Return -> {R}

Allowed return data types: str, list (one-dimensional),tuple (one-dimensional), dict (one-dimensional), int, and float. Python functions can also return True, False, and None, which will be converted to 1, 0 and an empty string, respectively.

Exception handling and tests

Whenever an exception is raised, the variables EXCEPTION and TRACEBACK are updated and Backstage continues calmly its execution.

Note that the variables TRACEBACK and EXCEPTION are cleared after the next successful command.

If you want the execution to stop whenever an exception is raised, just set 1 to the FailFast configuration option.

[task]
config FailFast=1

If you want to read a report of an exception when it's raised, just set 1 to the ReportException configuration option.

[task]
config ReportException=1

If you want to read the verbose traceback of an exception when it's raised, just set 1 to the ShowTraceback configuration option.

[task]
config ShowTraceback=1

You can edit these configuration options in the same command:

[task]
config FailFast=1 ReportException=1 ShowTraceback=0

You can read the current value of an arbitrary configuration option:

[task]
config FailFast
: FailFast -> {R}

Debug mode

Instead of manually setting the ReportException configuration option to 1, you can simply run a task in debug mode:

$ backstage -d task arg

Tests

To create a test, just postfix .test to the name of a task. Then from the command line, just run the test backstage --test task.

Example

[task]
set val (int) = {ARGS[0]} + {ARGS[1]}
return val

[task.test]
# here we branch the task with the arguments 40 and 2
& task 40 2
# we expect 42 as return 
assert R == 42

Command line interface and developer experience

$ backstage --help
Welcome to Pyrustic Backstage !
Ultimate task automation tool for hackers.

Usage:
    backstage
    backstage <task> [<arg> ...]
    backstage <option> [<arg> ...]
    
Options:
    -i, --intro                     Show file introductory text
    -c, --check                     Show the list of tasks
    -C, --Check                     Show the descriptive list of tasks
    -d, --debug <task> [<arg> ...]  Run task in debug mode
    -t, --test [<task> ...]         Run tests
    -T, --Test [<task> ...]         Run tests in debug mode
    -s, --search <task>             Search for a task by its name
    -S, --Search <task>             Search for a task by keyword
    -h, --help [<task>]             Show help text

    The <task> string can use a glob-like syntax that allows 
    wildcards '*' and '?'. Therefore, 'task1' is identical to 'task*'.
    
Visit the webpage: https://github.com/pyrustic/backstage

Developer experience

Backstage will do its best to help you understand raised exceptions:

$ backstage task1
ZeroDivisionError at line 3 of [task1] !
division by zero

$ backstage task2
InterpretationError at line 7 of [task2] !
Usage: sleep <seconds>

When you run Backstage in the loop mode, you can enjoy the autocomplete functionality (use the Tab key to complete your input) and also the history functionality (use Up and Down arrows).

$ backstage
Welcome to Pyrustic Backstage !
Ultimate task automation tool for hackers.
Press 'Ctrl-c' or 'Ctrl-d' to quit.
Type '--help' or '-h' to show more information.

(backstage) task(Tab Tab)
task    task1   task2   task3

(backstage) --h(Tab)

...

Miscellaneous

In the following subsections, we will explore some miscellaneous information.

Dogfooding

Backstage itself as a project relies on a backstage.tasks file (check the root of this repository). You are reading a document about Backstage that has been updated with Backstage !

Dependencies

Backstage relies on these Python packages:

  • Subrun to spawn new processes;
  • Shared to store data;
  • Jesth to parse backstage.tasks files;
  • Oscan to extract tokens from the script.

Indentation

Four (4) spaces by indent. Period.

Python 3

Inside the script file, you don't have to type python3 in a command to spawn the Python interpreter. Just type python to spawn the same interpreter that is running Backstage:

[task]
$ python -m my.package.module

Shell

Backstage doesn't rely on any Shell. But you can still pipe commands !

Example, if you want to change the current working directory:

[task]
# instead of doing this
$ cd {HOME}

# use the built-in 'cd'
cd HOME

# you can still spawn programs commonly used in the shell
$ ls

# or make this complex stuff (successfully performed on Ubuntu)
$ python -m this | tail --lines=+3 | sort

Data cache

Backstage stores data in an automatically created directory .backstage located in the current working directory. Inside this directory you can find the execution.log and database.json files.

Automatic line break

If you don't want anymore an extra line break at the end of printed strings, you can turn off this functionality:

[task]
# turn off auto line break
config AutoLineBreak=0
: `Hello `
: `World`
# turn on auto line break
config AutoLineBreak=1
: Hello World

Lines

You can draw lines with the characters = or -. If you pick one, only this one is allowed to appear on the same line.

[task1]

$ program1
$ program2

-----------------

[task2]
pass

=================

Demo

The demo is a repository that contains a backstage.tasks file similar to the one used to build, package and publish my projects. Your mission, if you accept it, is to clone the demo repository and run the backstage.tasks which contains tasks to create a new Python Hello Friend ! project, build it, perform versioning, init Git , perform Git Commit and Git Push, and even push the latest built package to PyPI !

# 1- clone the repository
$ git clone https://github.com/pyrustic/project
$ cd project

# 2- install backstage
$ pip install backstage

# 3- install buildver
$ pip install buildver

# 4- install setupinit
$ pip install setupinit

# 5- list the tasks available in the `backstage.tasks` file
$ backstage -c
Available tasks (11):
    build  check  clean  gendoc  gitcommit  gitinit  gitpush  init
    release  test  upload2pypi

# 6- descriptive list of tasks
$ backstage -C

# 7- initialize the project
$ backstage init
Successfully initialized !

# 8- run the project
$ python3 -m project
Hello Friend !

# 9- build the project
$ backstage build
building v0.0.1 ...
Successfully built 'project' v0.0.1 !
VERSION file updated from 0.0.1 to 0.0.2

# 10- check the project
$ backstage check
project v0.0.2 (source)
.whl v0.0.1 (package) built 28 secs ago

# 11- initialize Git
$ backstage gitinit
Origin: https://github.com/pyrustic/project.git

# 12- perform a Git Commit
$ backstage gitcommit

# 13- perform a Git Push
$ backstage gitpush

# 14- upload to PyPI
$ backstage upload2pypi

Note: Commands 9, 12, 13, and 14 can be replaced with one command: backstage release

Installation

Backstage is cross platform and versions under 1.0.0 will be considered Beta at best. It is built on Ubuntu with Python 3.8 and should work on Python 3.5 or newer.

For the first time

$ pip install backstage

Upgrade

$ pip install backstage --upgrade --upgrade-strategy eager



Back to top