Skip to content

MarquisdeGeek/ultra

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

.----------------. .----------------. .----------------. .----------------. .----------------. 
| .--------------. | .--------------. | .--------------. | .--------------. | .--------------. |
| | _____  _____ | | |   _____      | | |  _________   | | |  _______     | | |      __      | |
| ||_   _||_   _|| | |  |_   _|     | | | |  _   _  |  | | | |_   __ \    | | |     /  \     | |
| |  | |    | |  | | |    | |       | | | |_/ | | \_|  | | |   | |__) |   | | |    / /\ \    | |
| |  | '    ' |  | | |    | |   _   | | |     | |      | | |   |  __ /    | | |   / ____ \   | |
| |   \ `--' /   | | |   _| |__/ |  | | |    _| |_     | | |  _| |  \ \_  | | | _/ /    \ \_ | |
| |    `.__.'    | | |  |________|  | | |   |_____|    | | | |____| |___| | | ||____|  |____|| |
| |              | | |              | | |              | | |              | | |              | |
| '--------------' | '--------------' | '--------------' | '--------------' | '--------------' |
 '----------------' '----------------' '----------------' '----------------' '----------------' 

Ultra - Copyright 2014 Steven Goodwin

Released under the GNU GPL, version 3

Version 0.1.1 - 15th April 2014

Ultra is an ultra small (<64k binary), ultra fast, all-in-memory web server incorporating an in-built NoSQL datastore, a data processing language, SSI, multiple configurations, and logging.

It is intended for IoT and BigData applications.

Compile:

cd Release
make
cd ..

Usage:

Release/Ultra site_dir
Release/Ultra site_dir runmode
Release/Ultra site_dir runmode test

site_dir defaults to 'site' if omitted, which contains a sample website.

runmode determines the type of configuration to use, e.g. develop/live, and 'test' is a simple test output for internal use only.

It outputs its PID (for later use) to stdout, along with info/warning messages. Error messages are sent to stderr.

Then visit:

http://localhost:8088

Performance

On my desktop PC, siege gives me:

Transactions:                   16654 hits
Availability:                 100.00 %
Elapsed time:                   6.40 secs
Data transferred:              34.80 MB
Response time:                  0.00 secs
Transaction rate:            2602.19 trans/sec
Throughput:                     5.44 MB/sec
Concurrency:                    4.93
Successful transactions:       16654
Failed transactions:               0
Longest transaction:            0.03
Shortest transaction:           0.00

Whilst running:

kill -SIGHUP pid  	; reload the config and data files
kill -SIGUSR1 pid	; serialize/flush all data to disc

Conventions:

site/db/table_name		; name=value pairs, one on each line
site/docs/view_files.htm	; all html/ascii content
site/docs/assets/image.png	; all binary content
site/config/ultra.conf		; name=value pairs. e.g. port=8088 also dev.port=8089 for 'dev' runmode
site/config/mime.conf		; name=value pairs holding the MIME types

Configuration: (inside ultra.conf)

port=8088			; >1024 for normal users
usefork=true			; fork at each http request
maxrequestsize=32768		; max size of request packet

To have different parameters for different runmodes simply prefix the name with the runmode, e.g.

live.port=8080
develop.port=8081

You can then invoke this configuration automatically by selecting the 'runmode' with,

ultra site_dir live

as mentioned above.

Database:

It's a simple heirarchical name-value pair. Any text page (as found in site/docs) can use the format {(db:users.3.name)} to replace the {(meta data)} with the information from table 'users', where ID=3, and show the 'name' field.

You may have any number of fields, tables,subfields, and IDs as you wish. They may also be of any type.

You can amend the data in the DB with the format: {(db:users.3.name=Name goes here)} It will take place whenever a page is loaded containing such a string.

Meta commands:

All {(commands)} are processed when a page is requested. You can also nest them.

Accessing and modifying the database:

{(db:table.id.field)}	; retrieve the field, and write to the stream
{(db!:table.id.field)}	; retrieve the field, but don't echo it. Used for nested expressions
{(db:table.id.field=n)}	; assign the number 'n' to the DB field
{(db:table.id.field?n)}	; assign the number 'n' to the DB field, provided n is not empty
{(db:table.id.field+n)}	; add the number 'n' to the DB field
{(db:table.id.field-n)}	; substract the number 'n' to the DB field

Note: All =+-? can be used with the 'db' or 'db!' versions.

By convention, the 'var' table is used to store local variables for processing. This allows us take a specific field by nesting the var in a DB request, like this:

  {(db:users.{(db:var.id)}.name)}

Data processing:

{(op.range:value minimum maximum)} ; writes the value, clamped to the given range {(op.==:value1 value2 value3)} ; writes either 1 or 0, according to whether all valueX arguments are identical {{op.if:condition true_case false_case_optional}} ; writes either true or false case, depending on condition

Other commands include:

{(day)}  {(month)}  {(year)}  {(hours)}  {(minutes)}  {(seconds)}
{(home)}			; a link to the home page. i.e. /
{(get:arg)}			; take the argument from the GET request, i.e. url?arg=1
{(ssi:filename)}		; include the filename directly into the stream
{(link:file.htm)}		; a link
{(link:file.htm The file)} 	; a link with nicer anchor text
{(redirect:url)}		; issues a command to the client to redirect
{(exec:command args)}		; shell out to arbitrary command
{(config.dump:all)}		; write the config to the stream
{(db.dump:all)}			; write the current DB to the stream
{(mime.dump:all)}		; write the current MIME types table to the stream
{(stats.dump:all)}		; write the current logging stats (i.e. page accesses) to the stream
{(rem:ignore this)}

Notes to developers:

This is a 0.1 release for good reason - I started hacking around with the idea of a web server and got side-tracked by "real life"! My first code was 47 lines of C, which grew to 200 of C, which then grew to 1500 of C++. And then to this...

The language change happened when I looked at the amount of OO and utility libraries I was using. I wasn't going to re-invent/re-learn an old C library when I knew C++ would handle it well.

Consequently, half the code has a C-oriented mindset, while the rest thinks it's C++.

I'll unify it one day...

About

An ultra-small, ultra-fast, web server.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published