blue collar dev

Wednesday, July 13, 2016

execute ES6 scripts with ease

with modules installed privately

mkdir my_tests

cd my_tests

# you need an npm project to install npm modules locally
npm init -y

npm install --save-dev babel-core babel-cli babel-preset-es2015

# let babel know what version you want to use
echo '{ "presets":["es2015"] } ' > .babelrc

# use the new "import" syntax; here I imported something from redux just to get it to fail
echo 'import { createStore } from "redux";' > t1.js

# this will fail 
node t1.js 

# this will not fail 
./node_modules/.bin/babel-node t1.js

with modules installed globally you don't need a npm project file

mkdir my_other_tests

cd my_other_tests

npm install -g babel-core babel-cli babel-preset-es2015

# specify the ES version in the command line or create a .babelrc file like above
babel-node --presets es2015 t1.js

Monday, July 11, 2016

linode migration and downgrade

Finally began to understand Webpack. No clue as to how Redux works; I can follow the instructions but they don't make sense yet. Will persist.

Went through all the sites I had on my public servers and removed all those I did not use any more. Then noticed I was offered to migrate my Linode VPS from Xen to KVM and did it, upgraded to 4GB of RAM for free, resized my disks without losing data and then downgraded the account to a 2GB of RAM account ... all in less than 60 minutes and without losing data. I am thinking of consolidating the servers and I know which one will stay.

Also found a MongoDB database with data for an old project I might revive. Upgraded MongoDB on my workstation, upgraded the Perl bindings (required a server upgrade, else testing was failing).

Tuesday, June 21, 2016

sabbatical

After a few intense months I am taking a sabbatical.

Started looking in depth into React and the node.js ecosystem, mostly because those months would have been less intense if I had more confidence I understand what React.js is about.

Also trying to find out why f...g Agile is a pain in the behind.

Sunday, June 5, 2016

tracking page loads on this blog

The blogger.com stats do not seem to make any sense lately, so I am attempting to so some tracking of my own. I uploaded a transparent one pixel image on one of my servers. To force the browsers to reload my transparent pixel I added this snippet right in the template after the <head> tag.

<script type="text/javascript">
var tracker = new Image();
tracker.src = "http://lunch-break.ro/000000-0.png?data=" + Date.now();
</script>

Seems to be working. Maybe I'll even discover that the information in the "Stats" section is correct.

Thursday, September 17, 2015

double encoding of UTF8 strings

Turns out I was focusing on the wrong leads. These strings became values in a hash which was run through JSON->encode before saving to db. What prompted the re-encoding was not the values of the hash, but the keys. These were string literals in my program which were not marked as UTF8. Perl looked at my keys, looked at my values, saw a mixed bag, and decided to run the whole thing through the shredder again, just to be on the safe side.

The solution was simple:
use utf8

Because my keys were source-code string literals, I want my source code to be UTF8. use utf8 assures that.

thanks to: http://perlgerl.com/2011/04/08/utf8-follies/

Monday, August 10, 2015

how to debug anonymous functions and coderefs in Perl 5

http://perldoc.perl.org/B/Deparse.html

use strict;
use warnings;
use B::Deparse;

sub debug_coderef {
    my $possibly_code = shift; 
    my $deparse = B::Deparse->new();
    if (ref $possibly_code && ref $possibly_code eq 'CODE') {
        print STDERR "CODEREF: ", $deparse->coderef2text($possibly_code), "\n";
    }  else {
        print STDERR "NOT A CODEREF: ", Dumper($possibly_code), "\n";
    }
}

debug_coderef(sub {print "a";});

Wednesday, June 3, 2015

perl XS adventures, episode 1

Perl XS is easy if you read the docs in the proper order. What is the proper order ? I have no idea, I kept at it until it clicked. Well, not all of it clicked, just enough to get some results.

Summary so far: managed to persuade perl to run my code from the .xs file, to compile my .c files and to call code in the .c files from the .xs file.

To have it work I kept it simple. I tried the not-simple version and you don't want to know the story.

Hints for the inpatient

Quick hints

man 2 and man 3 are your best friends, if you don't know this already; for example
```
man 2 readdir
```
or
```
man 3 strcat
```
read perlguts before perlapi; better, do not attempt to read perlapi end-to-end, just consult it when you need
normal C code in custom .c and .h files, which all got to the root of the project (where Makefile.PL is); it is possible to put them some other place but I have not found out how to change Makefile.PL to deal with it.
include .h files in the .xs file
the code that uses Perl types and functions stays in the .xs file; Perl-independent C code can go before the
```
MODULE = My::Module  PACKAGE = My::Module
```
line

extra libraries added in the Makefile as obiect files in here

OBJECT            => '$(BASEEXT)$(OBJ_EXT) blah.o',

or even better

OBJECT            => '$(BASEEXT)$(OBJ_EXT) blah$(OBJ_EXT)'

to raise exceptions use
```
croak(const char * blah)
```
instead of exit(1) in the code in the .xs files

How to change the generated Makefile

based on http://www.perlmonks.org/?node_id=664066

Add a new compilation target

sub postamble {
    # this should be empty 
    my $inherited = shift->SUPER::postamble(@_);

    my $extra = q{
clean_temp_files :
 if [ -d /tmp/blah ]; then rm -rf /tmp/blah; fi
};
    return $inherited . "\n" . $extra;
}

Modify an existing compilation target

I think all the sections in the generated Makefile can be intercepted and modified, but so far I have tried only with clean:

sub clean {
  my $inherited = shift->SUPER::clean(@_);
  $inherited =~ s/clean :: clean_subdirs/clean :: clean_subdirs clean_temp_files/;
  return $inherited;
}

Step by step instructions for the very patient XS Mechanics or XS Fun.