[ANN] a lot

[ANN] a lot

[tantalum]

several gpl3 licensed guile modules:

sph-lib - a collection of libraries and example implementations
  cli: quickly create a command-line interface
  test: write and execute code tests
  test-performance: compare performance between arbitrary procedures
  lang-plcss: language for creating cascading stylesheets
  lang-scm-format: scheme code formatter
  scgi: interface for the creation of long-running web applications
  web-html: includes a multipart-form-data parser that can do stream
parsing and supports unlimited part-nesting
  tree: processing tree-like lists
  string: includes string-replace-string, a fast replacer procedure
  type-signature: reader/writer for a type-signature syntax
  process: includes procedures for process chaining like with bash and pipes
  ... and more, about 77 altogether
  http://sph.io/content/187

sescript
  scheme datum to ecmascript transcompiler
  so you can write javascript using scheme syntax. supports all
ecmascript and the r6rs library-form
  http://sph.io/content/413

sc
  a scheme datum to c transcompiler
  like sescript, but for c. supports all c.
  this thing works, already wrote programs with it
  http://sph.io/content/3d3

guile-fuse
  bindings to FUSE - filesystems in userspace
  probably linux specific because i do not know how to set up autotools
(and how i could do this without cluttering the project directory)
  http://sph.io/content/467

web-app
  a framework for web applications - scgi support, a design based on
fundamental scheme data types and short response times (2ms for plain
text responses on my computer)
  http://sph.io/content/66b

because this all is new, there are likely many things that need fix and
improvement. i am happy to make things work.
the website employs most of the aforementioned

Re: [ANN] a lot

[David Thompson]

tantalum <sph@posteo.eu> writes:

> several gpl3 licensed guile modules:
>
> sph-lib - a collection of libraries and example implementations
>   cli: quickly create a command-line interface
>   test: write and execute code tests
>   test-performance: compare performance between arbitrary procedures
>   lang-plcss: language for creating cascading stylesheets
>   lang-scm-format: scheme code formatter
>   scgi: interface for the creation of long-running web applications
>   web-html: includes a multipart-form-data parser that can do stream
> parsing and supports unlimited part-nesting
>   tree: processing tree-like lists
>   string: includes string-replace-string, a fast replacer procedure
>   type-signature: reader/writer for a type-signature syntax
>   process: includes procedures for process chaining like with bash and pipes
>   ... and more, about 77 altogether
>   http://sph.io/content/187
>
> sescript
>   scheme datum to ecmascript transcompiler
>   so you can write javascript using scheme syntax. supports all
> ecmascript and the r6rs library-form
>   http://sph.io/content/413
>
> sc
>   a scheme datum to c transcompiler
>   like sescript, but for c. supports all c.
>   this thing works, already wrote programs with it
>   http://sph.io/content/3d3
>
> guile-fuse
>   bindings to FUSE - filesystems in userspace
>   probably linux specific because i do not know how to set up autotools
> (and how i could do this without cluttering the project directory)
>   http://sph.io/content/467
>
> web-app
>   a framework for web applications - scgi support, a design based on
> fundamental scheme data types and short response times (2ms for plain
> text responses on my computer)
>   http://sph.io/content/66b
>
> because this all is new, there are likely many things that need fix and
> improvement. i am happy to make things work.
> the website employs most of the aforementioned
>

Wow!  Thank you for all of these neat projects!  web-app and sescript
are particularly interesting.  I haven't had a chance to look at any of
the code yet, but I hope to get the chance sometime soon.

To make building the software easier for myself and others, I recommend
taking the time to learn how to write automake and autoconf files.  I
don't know the autotools very well myself, but reading the configure.ac
and Makefile.am files of other Guile projects helped me a lot.

Thanks and happy hacking!

-- 
David Thompson
Web Developer - Free Software Foundation - http://fsf.org
GPG Key: 0FF1D807
Support the FSF: https://fsf.org/donate

Re: [ANN] a lot

[Max]

Excellent work.

The most interesting for me are process and record. Are there some code examples?
For instance I'd like to execute binary at /tmp/mybin with environment variable
XX2=LOL - do I call it like:


      (execute-with-environment 'XX2=LOL' /tmp/mybin)?

That's a long running process so I'd like to be able to send signals to it - will
this function return pid which I can use?
How do I store that into record alongside with LOL (value of variable I've set earlier)?

(make-record myrec)
(record-append pid LOL)

The greatest possible improvement would be if every documentation entry would include
code sample but it's tremendous amount of rather boring work so I would be happy to
see any sort of examples.

thanks,
Max.

Re: [ANN] a lot

[tantalum]

> Are there some code examples?
except for process-chain-* probably not, but i put that on my todo list.

something like this should work:
  (execute-with-environment "XX2=LOL" "/tmp/mybin")

working example:
  guile -c '(import (sph process)) (execute-with-environment (list
"XX2=LOL") "/usr/bin/echo" "test")'

but this is using the system call "execve", which leads to the whole
guile process being replaced by the specified program. (execve syscall
reference: http://man7.org/linux/man-pages/man2/execve.2.html)

"process-create" creates a new parallel process
---
(define pid
  (process-create
    (lambda ()
      ;this will be executed in a completely cloned new process in the
background
      (execute-with-environment "XX2=LOL" "/tmp/mybin"))))

;now, still in the guile process, send a signal to the new process
(kill pid SIGINT)
---
now what happens is that a new process is created by cloning, and is
then immediately replaced by your program. as far as i know this is the
only way on linux or similar systems to archieve this.

for storing values in a record it would probably be best to use a record
setter procedure. the library supports setting fields by symbolic name,
but that is slower.
in general, as a sidenote, a record like this differs from an
association list for example, in that the keys are not stored inside the
data-structure, but separately: the "layout" stores the field names and
the positions of where the field values are stored in the records.
example:
---
(define myrec (make-record-layout (quote (pid env))))
(define-record-accessors myrec (myrec-pid (quote pid)) (myrec-env (quote
env)))
(define-record-setters myrec (myrec-pid-set! (quote pid))
(myrec-env-set! (quote env)))

;two different ways to create a record
(define a (record myrec pid "LOL"))
(define b (make-record myrec))

(myrec-pid a)
(myrec-pid-set! a 3)
(myrec-env-set! b 2)
---

EOM

Re: [ANN] a lot

[Max]

21.07.2014 19:14, tantalum пишет:
>> Are there some code examples?
> except for process-chain-* probably not, but i put that on my todo list.
>
> something like this should work:
>   (execute-with-environment "XX2=LOL" "/tmp/mybin")
>
> working example:
>   guile -c '(import (sph process)) (execute-with-environment (list
> "XX2=LOL") "/usr/bin/echo" "test")'
>
> but this is using the system call "execve", which leads to the whole
> guile process being replaced by the specified program. (execve syscall
> reference: http://man7.org/linux/man-pages/man2/execve.2.html)

Neat. Just to double-check - the environment variables of my current process are
preserved or /tmp/mybin would run with environment consisting only of variables which
I've explicitly put into execute-with-environment?

>
> "process-create" creates a new parallel process
> ---
> (define pid
>   (process-create
>     (lambda ()
>       ;this will be executed in a completely cloned new process in the
> background
>       (execute-with-environment "XX2=LOL" "/tmp/mybin"))))
>
> ;now, still in the guile process, send a signal to the new process
> (kill pid SIGINT)
> ---
> now what happens is that a new process is created by cloning, and is
> then immediately replaced by your program. as far as i know this is the
> only way on linux or similar systems to archieve this.

Yes, that's totally fine.

>
> for storing values in a record it would probably be best to use a record
> setter procedure. the library supports setting fields by symbolic name,
> but that is slower.

Could you illustrate that with some code too? I do not plan to have that many records
so I'm not concerned with the speed that much :)

> in general, as a sidenote, a record like this differs from an
> association list for example, in that the keys are not stored inside the
> data-structure, but separately: the "layout" stores the field names and
> the positions of where the field values are stored in the records.
> example:
> ---
> (define myrec (make-record-layout (quote (pid env))))
> (define-record-accessors myrec (myrec-pid (quote pid)) (myrec-env (quote
> env)))
> (define-record-setters myrec (myrec-pid-set! (quote pid))
> (myrec-env-set! (quote env)))

Hmm... looks like a boilerplate - I thought those are generated automatically.

> ;two different ways to create a record
> (define a (record myrec pid "LOL"))
> (define b (make-record myrec))
>
> (myrec-pid a)
> (myrec-pid-set! a 3)
> (myrec-env-set! b 2)
> ---

I can also have arbitrary number of fields in the record?

(define triad (make-record-layout (quote (pid env num))))

(define t (record triad pid "LOL" 5))


thanks,
Max.