沒有人會為弱者流眼淚

The title of my blog is actually a quote from one of my favorite comics: Gunnm (Battle Angel Alita).

C++ lambda inheritance

I was thinking about if we are able to inherit from a C++ lambda (say, anonymous class)

My first try was like this, use a helper function to forward the lambda body and store that as an private member of a class. But this is a HAS-A relationship, not really inherited from the lambda.

#include <iostream>
#include <utility>

template<typename lambda>
class Foo {
  lambda _t;
  int x = 5566;

public:
  Foo(lambda &&t) : _t(std::forward<lambda>(t)) {
    std::cout << "Foo()" << std::endl;
  }

  ~Foo() {
    std::cout << "~Foo()" << std::endl;
  }

  void operator()(void) {
    _t(x);
  }

  void operator()(int x) {
    std::cout << "Foo(int): " << x << std::endl;
  }
};

template<typename lambda>
Foo<lambda> make_class(lambda &&t) {
  return { std::forward<lambda>(t) };
}


int main (void) {
  auto qq = make_class([](int x) -> void {
    std::cout << "operator(int): " << x << std::endl;
  });

  qq();
  qq(7788);

  return 0;
}

Foo()
operator(int): 5566
Foo(int): 7788
~Foo()



suhorngT improved this as a real inheritance:

#include <iostream>
#include <utility>

template<typename lambda>
class Foo : public lambda {

public:
    Foo(lambda &&t) : lambda(t) {
      std::cout << "Foo()" << std::endl;
    }

    ~Foo() {
      std::cout << "~Foo()" << std::endl;
    }
};

template<typename lambda>
Foo<lambda> make_class(lambda &&t) {
  return { std::forward<lambda>(t) };
}


int main (void) {
  auto qq = make_class([](int x) -> void {
    std::cout << "operator(int): " << x << std::endl;
  });

  qq(123);
  return 0;
}

Foo()
operator(int): 123
~Foo()

After that I took out the make_class helper, and use decltype instead, make the syntax much better.

#include <iostream>

template<typename... lambda>
class Bar : public lambda... {

public:
  Bar(lambda... l) : lambda(l)... {
    std::cout << "Bar()" << std::endl;
  }

  void foo(void) {
    std::cout << "foo()" << std::endl;
  }
};


int main(void)  {
  int a = 55;

  auto gg = [a](int b) -> void {
    std::cout << "operator(): " << a << b << std::endl;
  };

  auto qq = Bar<decltype(gg)>(gg);

  qq(66);
  qq.foo();

  return 0;
}

Bar()
operator(): 5566
foo()


Reference:
http://cpptruths.blogspot.tw/2014/05/fun-with-lambdas-c14-style-part-2.html
http://cpptruths.blogspot.com/2014/03/fun-with-lambdas-c14-style-part-1.html

gdb pretty-print python path

On some of my dev machines (freaking Ubuntu 14.04), gdb can't find the pretty print python extension.

(^q^) r
Starting program: /home/xatier/xxxxx
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
Traceback (most recent call last):
  File "/usr/share/gdb/auto-load/usr/lib/x86_64-linux-gnu/libstdc++.so.6.0.19-gdb.py", line 63, in <module>
    from libstdcxx.v6.printers import register_libstdcxx_printers
ImportError: No module named 'libstdcxx'

Let's print python path:


On Ubuntu 14.04
(^q^) python print(sys.path)
['/usr/share/gdb/python', '/usr/lib/python3.4', '/usr/lib/python3.4/plat-x86_64-linux-gnu', '/usr/lib/python3.4/lib-dynload', '/usr/local/lib/python3.4/dist-packages', '/usr/lib/python3/dist-packages']

On arch linux
(^q^) python print(sys.path)
['/usr/lib/../share/gcc-5.2.0/python', '/usr/share/gdb/python', '/usr/lib/python35.zip', '/usr/lib/python3.5', '/usr/lib/python3.5/plat-linux', '/usr/lib/python3.5/lib-dynload', '/usr/lib/python3.5/site-packages']


See? So basically the python scripts are under /usr/share/<gcc-ver>/python

$ ls /usr/share | grep gcc
gcc-4.8


Just add this line to your ~/.gdbinit :

python sys.path.append("/usr/share/gcc-4.8/python")


Now you are good.

(^q^) python print(sys.path)
['/usr/share/gdb/python', '/usr/lib/python3.4', '/usr/lib/python3.4/plat-x86_64-linux-gnu', '/usr/lib/python3.4/lib-dynload', '/usr/local/lib/python3.4/dist-packages', '/usr/lib/python3/dist-packages', '/usr/share/gcc-4.8/python']

STL vector out-of-bound access is an undefined behavior

Found a bug today, which is a well-known pitfall in C++.

Consider the following code, std::vector::operator[] won't perform any boundary check on the index, a out-of-bound access is an undefined behavior.

#include <vector>                                                               
#include <iostream>                                                             
                                                                                
int main (void) {                                                               
  std::vector<bool> a(10);                                                      
                                                                                
  for (auto x : a)                                                              
      x = true;                                                                 
                                                                                
  if (a[11] == false)                                                           
      std::cout << "gg";                                                        
                                                                                
  return 0;                                                                     
}

gg

----

On the other hand, std::vector::at will yield an exception on out-of-bound access.

#include <vector>                                                               
#include <iostream>                                                             
 
int main (void) {                                                               
  std::vector<bool> a(10);                                                      
 
  for (auto x : a)                                                              
      x = true;                                                                 
 
  if (a.at(11) == false)                                                           
      std::cout << "gg";                                                        
 
  return 0;                                                                     
}

terminate called after throwing an instance of 'std::out_of_range'
what(): vector::_M_range_check: __n (which is 11) >= this->size() (which is 10)

ffmpeg x264 lossless video encoding

I re-encoded my previous club class video clips for my Google drive quota.

The result of is extremely good.


$ ls -lh
-rw-r----- 1 xatier staff 2.8G Jun 29 07:13 Topic 20 - [Programming 1] Python 1.mov
-rw-r--r-- 1 xatier staff 1.6G Jul 3 00:24 Topic 20 - [Programming 1] Python 1.mp4
-rw-r----- 1 xatier staff 3.1G Jun 29 07:30 Topic 21 - [Programming 2] Python 2.mov
-rw-r--r-- 1 xatier staff 1.4G Jul 3 13:43 Topic 21 - [Programming 2] Python 2.mp4

For the x264 lossless encoding, you can use the following commands:

If you don't have time, use the 'ultrafast' preset:

# fastest encoding
$ ffmpeg -i input -c:v libx264 -preset ultrafast -qp 0 -c:a copy output

If you need a compressed encoding, use the 'veryslow' preset:

# best compression
$ ffmpeg -i input -c:v libx264 -preset veryslow -qp 0 -c:a copy output


Both examples will provide the same quality output. (-qp 0)


It took me over 2 hours for this video on my 2012-mid MacbookAir (1.8 GHz Intel Core i5 / 8 GB 1600 MHz DDR3).


$ ffmpeg -i Topic\ 20\ -\ \[Programming\ 1\]\ Python\ 1.mov  -c:v libx264 -preset veryslow -qp 0 -c:a copy py1.mp4
...
frame=459704 fps= 23 q=-1.0 Lsize= 1660358kB time=02:07:41.73 bitrate=1775.3kbits/s dup=1008 drop=0
video:1387489kB audio:264157kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.527454%


Update: il gnaggnoy remembered not only qp but also both qpmax and qpmin should be set to 0 in order to force a lossless encoding in x264.

Reference:
https://trac.ffmpeg.org/wiki/Encode/H.264
https://wiki.archlinux.org/index.php/FFmpeg

List Google Drive files by size

When you run out of the quota limit of your Google Drive service, you must want to look at which file is occupying your space.

For Google Drive, you can just use this link to see the list of uploaded files by size.

https://drive.google.com/#quota

Alternatively, you can hover over your storage usage in the bottom left corner and click the Google Drive icon, that will lead you to the same link as above.

References:

https://support.google.com/a/answer/3118661?hl=en

http://googlesystem.blogspot.com/2014/07/google-drives-quota-page.html

Currency conversions in Google Spreadsheet

Just learned a tip in Google Spreadsheet.

To get the currency conversion rate between TWD-USD:

=GoogleFinance("CURRENCY:TWDUSD")

That will give you the rate from Google Finance database, history is also available.

=GoogleFinance("currency:USDTWD", "price", today()-10, today())

Other API parameters:
https://support.google.com/docs/answer/3093281

Reference:
http://googledocstips.com/2011/03/09/how-to-calculate-foreign-exchange/
http://stackoverflow.com/questions/20607627/on-google-spreadsheet-how-can-you-query-googlefinance-for-a-past-exchange-rate

Graduation

Alright, I just finished a big milestone in my life -- graduated from the college with a degree in computer science. I'm not a college kid anymore.


Actually, I have no idea about this certificate. For me, the only purpose of this "paper" is matching the requirement of my H-1B visa next year.


Requirement 2 - Your job must qualify as a specialty occupation by meeting one of the following criteria:
A bachelor’s degree or higher degree or its equivalent is normally the minimum requirement for the particular position;






Anyway, happy graduation.

Thanks to my parents, my sister, Dr. Sung, jserv, ... all of my friends, ... and lots of people surrounding me.

Interactive debugging in Python

Just put this thing anywhere in your python code.

import code                                                                    
code.interact(local=locals())                                                  

import IPython                                                                 
IPython.embed()

Rip Audio CDs in command line

We basically use two tools: cdparanoia and lame.

cdparanoia: Compact Disc Digital Audio extraction tool https://www.archlinux.org/packages/extra/x86_64/cdparanoia/
lame: A high quality MPEG Audio Layer III (MP3) encoder https://www.archlinux.org/packages/extra/x86_64/lame/


# grab CD information
$ cdparanoia -vsQ
cdparanoia III release 10.2 (September 11, 2008)

Using cdda library version: 10.2
Using paranoia library version: 10.2
Checking /dev/cdrom for cdrom...
Testing /dev/cdrom for SCSI/MMC interface
SG_IO device: /dev/sr0

CDROM model sensed sensed: ASUS DRW-24D1ST 1.00

Checking for SCSI emulation...
Drive is ATAPI (using SG_IO host adaptor emulation)

Checking for MMC style command set...
Drive is MMC style
DMA scatter/gather table entries: 1
table entry size: 131072 bytes
maximum theoretical transfer: 55 sectors
Setting default read size to 27 sectors (63504 bytes).

Verifying CDDA command set...
Expected command set reads OK.

Attempting to set cdrom to full speed...
drive returned OK.

Table of contents (audio tracks only):
track length begin copy pre ch
===========================================================
1. 11033 [02:27.08] 0 [00:00.00] no no 2
2. 18007 [04:00.07] 11033 [02:27.08] no no 2
3. 10100 [02:14.50] 29040 [06:27.15] no no 2
4. 13237 [02:56.37] 39140 [08:41.65] no no 2
5. 9050 [02:00.50] 52377 [11:38.27] no no 2
6. 11076 [02:27.51] 61427 [13:39.02] no no 2
7. 23342 [05:11.17] 72503 [16:06.53] no no 2
8. 13141 [02:55.16] 95845 [21:17.70] no no 2
9. 16703 [03:42.53] 108986 [24:13.11] no no 2
10. 12635 [02:48.35] 125689 [27:55.64] no no 2
TOTAL 138324 [30:44.24] (audio only)





# rip stuffs
$ cdparanoia -B
cdparanoia III release 10.2 (September 11, 2008)


Ripping from sector 0 (track 1 [0:00.00])
to sector 138323 (track 10 [2:48.34])

outputting to track01.cdda.wav

(== PROGRESS == [ | 011032 00 ] == :^D * ==)

outputting to track02.cdda.wav

(== PROGRESS == [ | 029039 00 ] == :^D * ==)

outputting to track03.cdda.wav

(== PROGRESS == [ | 039139 00 ] == :^D * ==)

outputting to track04.cdda.wav

(== PROGRESS == [ | 052376 00 ] == :^D * ==)

outputting to track05.cdda.wav

(== PROGRESS == [ | 061426 00 ] == :^D * ==)


...

outputting to track10.cdda.wav

(== PROGRESS == [ | 138323 00 ] == :^D * ==)

Done.





# convert to mp3 format
$ for i in `ls`; do lame $i; done
LAME 3.99.5 64bits (http://lame.sf.net)
Using polyphase lowpass filter, transition band: 16538 Hz - 17071 Hz
Encoding track01.cdda.wav to track01.cdda.mp3
Encoding as 44.1 kHz j-stereo MPEG-1 Layer III (11x) 128 kbps qval=3
Frame | CPU time/estim | REAL time/estim | play/CPU | ETA
5633/5633 (100%)| 0:04/ 0:04| 0:04/ 0:04| 35.102x| 0:00
-------------------------------------------------------------------------------------
kbps LR MS % long switch short %
128.0 1.8 98.2 99.9 0.0 0.0
Writing LAME Tag...done
ReplayGain: -0.3dB
LAME 3.99.5 64bits (http://lame.sf.net)
Using polyphase lowpass filter, transition band: 16538 Hz - 17071 Hz
Encoding track02.cdda.wav to track02.cdda.mp3
Encoding as 44.1 kHz j-stereo MPEG-1 Layer III (11x) 128 kbps qval=3
Frame | CPU time/estim | REAL time/estim | play/CPU | ETA
9193/9193 (100%)| 0:07/ 0:07| 0:07/ 0:07| 34.246x| 0:00
-------------------------------------------------------------------------------------
kbps LR MS % long switch short %
128.0 0.3 99.7 100.0 0.0 0.0
Writing LAME Tag...done
ReplayGain: +0.1dB

...


done!

Reference: http://www.cyberciti.biz/faq/linux-ripping-and-encoding-audio-files/

skicka: Google drive command line tool

Install go and skicka

$ sudo pacman -S go
$ mkdir ~/go
$ export GOPATH=~/go
$ export PATH=$PATH:~/go/bin
$ go get github.com/google/skicka

Initialize the configuration and client id/secret key pairs from  https://console.developers.google.com/project
Read: https://github.com/google/skicka/blob/master/README.md


$ skicka init
2015/05/14 21:44:37 created configuration file /home/xatier/.skicka.config.
$ vim ~/.skicka.config



Oauth authentication for the first time
Generate ~/.skicka.metadata.cache and ~/.skicka.tokencache.json

$ skicka ls -l /
Go to the following link in your browser:
https://accounts.google.com/o/oauth2/auth?***********************************
Enter verification code: ******************************************
Updating metadata cache: 
[========================================================================================] 99.99 % 37s


Support commands:

$ skicka
usage: skicka [skicka options] [command options]

Supported commands are:
cat Print the contents of the given file
download Download a file or folder hierarchy from Drive to the local disk
df Display free space on Drive
du Report disk usage for a folder hierarchy on Drive
fsck Check consistency of files in Drive and local metadata cache
genkey Generate a new encryption key
init Create an initial skicka configuration file
ls List the contents of a folder on Google Drive
mkdir Create a new folder or folder hierarchy on Drive
rm Remove a file or folder on Google Drive
upload Upload a local file or directory hierarchy to Drive



Testing with my hinet 100/40 home use plan (roughly around 1.5MB/s):


$ skicka upload GG.mp4 /
Files: 23.83 MB / 23.83 MB 
[========================================================================================] 100.00 % 15s
2015/05/14 22:13:11 Preparation time 1s, sync time 15s
2015/05/14 22:13:11 Updated 1 Drive files, 0 local files
2015/05/14 22:13:11 23.83 MiB read from disk, 0 B written to disk
2015/05/14 22:13:11 23.83 MiB uploaded (1.58 MiB/s), 0 B downloaded (0 B/s)
2015/05/14 22:13:11 4.72 MiB peak memory used

$ skicka upload GG.mp4 /
Files: 23.83 MB / 23.83 MB 
[========================================================================================] 100.00 % 13s
2015/05/14 22:14:19 Preparation time 1s, sync time 13s
2015/05/14 22:14:19 Updated 1 Drive files, 0 local files
2015/05/14 22:14:19 23.83 MiB read from disk, 0 B written to disk
2015/05/14 22:14:19 23.83 MiB uploaded (1.72 MiB/s), 0 B downloaded (0 B/s)
2015/05/14 22:14:19 5.49 MiB peak memory used

$ skicka upload GG.mp4 /
Files: 23.83 MB / 23.83 MB 
[========================================================================================] 100.00 % 13s
2015/05/14 22:14:41 Preparation time 1s, sync time 13s
2015/05/14 22:14:41 Updated 1 Drive files, 0 local files
2015/05/14 22:14:41 23.83 MiB read from disk, 0 B written to disk
2015/05/14 22:14:41 23.83 MiB uploaded (1.71 MiB/s), 0 B downloaded (0 B/s)
2015/05/14 22:14:41 5.49 MiB peak memory used

$ skicka upload GG.mp4 /
Files: 23.83 MB / 23.83 MB 
[========================================================================================] 100.00 % 15s
2015/05/14 22:15:01 Preparation time 1s, sync time 15s
2015/05/14 22:15:01 Updated 1 Drive files, 0 local files
2015/05/14 22:15:01 23.83 MiB read from disk, 0 B written to disk
2015/05/14 22:15:01 23.83 MiB uploaded (1.51 MiB/s), 0 B downloaded (0 B/s)
2015/05/14 22:15:01 5.49 MiB peak memory used

Large file test:

$ skicka upload movie.mkv
Files: 8.39 GB / 8.39 GB 
[========================================================================================] 100.00 % 2h43m59s
2015/05/15 01:18:24 Preparation time 1s, sync time 2h 43m 59s
2015/05/15 01:18:24 Updated 1 Drive files, 0 local files
2015/05/15 01:18:24 8.39 GiB read from disk, 0 B written to disk
2015/05/15 01:18:24 8.39 GiB uploaded (893.93 kiB/s), 0 B downloaded (0 B/s)
2015/05/15 01:18:24 16.07 MiB peak memory used



Encryption

$ SKICKA_PASSPHRASE=gg skicka genkey
; Add the following lines to the [encryption] section
; of your ~/.skicka.config file.
salt=************************************
passphrase-hash=************************************
encrypted-key=************************************
encrypted-key-iv=************************************

$ SKICKA_PASSPHRASE=gg skicka upload -encrypt GG2.mp4 /
Files: 23.83 MB / 23.83 MB 
[========================================================================================] 100.00 % 25s
2015/05/14 21:59:16 Preparation time 1s, sync time 26s
2015/05/14 21:59:16 Updated 1 Drive files, 0 local files
2015/05/14 21:59:16 23.83 MiB read from disk, 0 B written to disk
2015/05/14 21:59:16 23.83 MiB uploaded (928.17 kiB/s), 0 B downloaded (0 B/s)
2015/05/14 21:59:16 4.24 MiB peak memory used

Make QT projects without QtCreator

Just a note for comping Qt stuffs without the shitty IDE.

$ ls
foo.cc

$ qmake -project

$ ls
foo.cc foo.pro

# maybe add some shit here, for example QT += concurrent
$ vim foo.pro

$ qmake foo.pro

$ make
g++ -c -pipe -O2 -march=x86-64 -mtune=generic -O2 -pipe -fstack-protector-strong --param=ssp-buffer-size=4 -Wall -W -D_REENTRANT -fPIC -DQT_NO_DEBUG -DQT_GUI_LIB -DQT_CONCURRENT_LIB -DQT_CORE_LIB -I. -I. -isystem /usr/include/qt -isystem /usr/include/qt/QtGui -isystem /usr/include/qt/QtConcurrent -isystem /usr/include/qt/QtCore -I. -I/usr/lib/qt/mkspecs/linux-g++ -o foo.o foo.cc
g++ -Wl,-O1 -Wl,-O1,--sort-common,--as-needed,-z,relro -o foo foo.o -lQt5Gui -lQt5Concurrent -lQt5Core -lGL -lpthread

$ ./foo

Adjust swappiness to improve responsiveness

I bought 16GB RAM recently for my desktop ( http://xatierlike.blogspot.tw/2015/04/ramdisk-for-chromium.html ). But sometimes the system is still a bit slow for me.

The reason is the high usage of the swap filesystem. The size of swap will go to over 1GB sometimes.

(for example)
$ free -h
              total        used        free      shared  buff/cache   available
Mem:            15G        5.0G        158M        1.0G         10G        9.1G
Swap:          1.9G        1.2G        612M

$ lsblk
NAME   MAJ:MIN RM   SIZE RO TYPE MOUNTPOINT
sda      8:0    0 931.5G  0 disk
├─sda1   8:1    0    28G  0 part /
├─sda2   8:2    0   1.9G  0 part [SWAP]
└─sda3   8:3    0 901.7G  0 part /home

According to Arch wiki: https://wiki.archlinux.org/index.php/Swap

Swappiness
The swappiness sysctl parameter represents the kernel's preference (or avoidance) of swap space. Swappiness can have a value between 0 and 100, the default value is 60. Setting this parameter to a low value will reduce swapping from RAM, and is known to improve responsiveness on many systems.

Sample usage:

# current swappiness
$ cat /proc/sys/vm/swappiness
60

# same as above
$ sysctl vm.swappiness
vm.swappiness = 60

# set it to 10
$ sudo sysctl vm.swappiness=10
vm.swappiness = 10

# unload old swap
$ sudo swapoff -a && sudo swapon /dev/sda2

# set the swappiness value permanently,
/etc/sysctl.d/99-sysctl.conf
----------------
vm.swappiness=10


Okay, now my system is back to flying.

STL pretty print support in gdb

I program in C/C++ for a long time, it's really hateful to debug C++ programs with STL containers in gdb, but yeah, who doesn't use STL?

gdb always prints lots of useless stuffs from a container.

$ g++ -std=c++11 foo.cc -g

$ gdb -q ./a.out

Reading symbols from ./a.out...done.

(^q^) l

1 #include <vector>

2

3 int main (void) {

4    std::vector<int> v = {1, 2, 3, 4, 5};

5    return 0;

6 }

(^q^) b 5

Breakpoint 1 at 0x40082c: file foo.cc, line 5.

(^q^) r

Starting program: /tmp/a.out 

Breakpoint 1, main () at foo.cc:5

5    return 0;

(^q^) p

$2 = {

  <std::_Vector_base<int, std::allocator<int> >> = {

    _M_impl = {

      <std::allocator<int>> = {

        <__gnu_cxx::new_allocator<int>> = {<No data fields>}, <No data fields>}, 

      members of std::_Vector_base<int, std::allocator<int> >::_Vector_impl: 

      _M_start = 0x602010, 

      _M_finish = 0x602024, 

      _M_end_of_storage = 0x602024

    }

  }, <No data fields>}

(^q^) 

And we're able to use a very tricky way to print that out.


(^q^) p *(v._M_impl._M_start)@(v._M_impl._M_finish - v._M_impl._M_start)

$22 =   {[0] = 1,

  [1] = 2,

  [2] = 3,

  [3] = 4,

  [4] = 5}

(^q^) 

A container is a "container", we don't care the implementation details, what we care about is the data.

With the pretty printer, we're able to print a container out like this:

(^q^) p v

$1 = std::vector of length 5, capacity 5 = {

  [0] = 1,

  [1] = 2,

  [2] = 3,

  [3] = 4,

  [4] = 5

}

According to the gdb wiki, the gdb python pretty printer is supported since 7.0. https://sourceware.org/gdb/wiki/STLSupport

Download the latest python script from gcc.gnu.org
---
mkdir ~/.gdb
cd ~/.gdb
svn co svn://gcc.gnu.org/svn/gcc/trunk/libstdc++-v3/python
---

put the following stuffs appended to your ~/.gdbinit
---
python
import sys
sys.path.insert(0, '/home/xatier/.gdb/python')
import libstdcxx.v6
end
---

Note, we only need to import the lib, let __init__.py do the magic ;-)

Done!


If you really want to dig up the implementation details, use " p /r "


(^q^) p /r v
$1 = {
  <std::_Vector_base<int, std::allocator<int> >> = {
    _M_impl = {
      <std::allocator<int>> = {
        <__gnu_cxx::new_allocator<int>> = {<No data fields>}, <No data fields>}, 
      members of std::_Vector_base<int, std::allocator<int> >::_Vector_impl: 
      _M_start = 0x602010, 
      _M_finish = 0x602024, 
      _M_end_of_storage = 0x602024
    }
  }, <No data fields>}

(^q^)


Actually this python script will be installed and automatically loaded in Archlinux once you install gcc/gcc-multilib.


$ pacman -Ql gcc-multilib | grep libstd
gcc-multilib /usr/lib/libstdc++.a
gcc-multilib /usr/lib32/libstdc++.a
gcc-multilib /usr/share/gcc-4.9.2/python/libstdcxx/
gcc-multilib /usr/share/gcc-4.9.2/python/libstdcxx/__init__.py
gcc-multilib /usr/share/gcc-4.9.2/python/libstdcxx/v6/
gcc-multilib /usr/share/gcc-4.9.2/python/libstdcxx/v6/__init__.py
gcc-multilib /usr/share/gcc-4.9.2/python/libstdcxx/v6/printers.py
gcc-multilib /usr/share/gdb/auto-load/usr/lib/libstdc++.so.6.0.20-gdb.py

If you're using other distros like Debian, that should be here, but not auto-loaded by default :(

$ apt-file list gcc | grep libstd
gcc-snapshot: /usr/lib/gcc-snapshot/share/gcc-4.9.0/python/libstdcxx/__init__.py
gcc-snapshot: /usr/lib/gcc-snapshot/share/gcc-4.9.0/python/libstdcxx/v6/__init__.py
gcc-snapshot: /usr/lib/gcc-snapshot/share/gcc-4.9.0/python/libstdcxx/v6/printers.py


Happy debugging C++! (^q^)

Reverseing reverse engineered tools: reverse on the pyc of Easy-Card tool

A hacker called  Zhi-Wei Cai did some reverse engineering on Taipei city passport Easy Card credit querying system.


https://github.com/x43x61x69/Easy-Card


But he only released the .pyc file.

I tool a glance on the binary file and smiled, he didn't apply any code obfuscation on that, that is, it would be pretty easy for those who want to read the source code.

With the help of this tool: https://github.com/wibiti/uncompyle2

uncompyle2 easycard.pyc > easycard.py

You can decompile that yourself.




Basically the API takes 4 parameters: verify, cardID, begin, end.


verify = md5((seed * const) + salt)
where
    seed = date.month + date.day + date.hour
    salt = 'L0CalKing'
    const = 8544

cardID = base64( des3( data, key, iv, mode=DEC3.MODE_CBC) )
where
    data = 'your card ID', like '1234567889'
    key = 'EasyCardToKingay23456789
    iv = '01234567'
   
begin / end = time period


I'm really curious about how did he get these constants , but I don't want to dig up the original app. :)


Lesson: don't release .pyc file without code obfuscation if you really don't want any people try to dig up your code.


Code listing (partially omitted):

#!/usr/bin/env python2
# -*- encoding: utf8 -*-

# 2015.05.04 17:41:47 CST
import sys
import datetime
import hashlib
import urllib
import urllib2
import json
from Crypto.Cipher import DES3
import pytz
version = '0.3'
copyright = 'Copyright (C) 2015 Zhi-Wei Cai.'
key = 'EasyCardToKingay23456789'
iv = '01234567'
salt = 'L0CalKing'
const = 8544

def getID(data, isEncrypt, key, iv, encode):
    size = len(data)
    # '\x06' is the padding of DES3
    if size % 16 != 0:
        data += '\x06' * (16 - size % 16)
    des3 = DES3.new(key, DES3.MODE_CBC, iv)
    if isEncrypt:
        result = des3.encrypt(data).encode(encode).rstrip()
    else:
        result = des3.decrypt(data.decode(encode))
    return result



def getVerify(const, seed, salt):
    hash = hashlib.md5()
    hash.update(str(seed * const) + salt)
    return hash.hexdigest().upper()



def proc(data):
    e = getID(data, 1, key, iv, 'base64')
    cardID = urllib.quote_plus(e)
    date = datetime.datetime.now(pytz.timezone('Asia/Taipei'))
    seed = date.month + date.day + date.hour
    begin = '{:%Y-%m-%d}'.format(date - datetime.timedelta(days=30))
    end = '{:%Y-%m-%d}'.format(date)
    verify = getVerify(const, seed, salt)
    url = '<Easy Card API URL>'.format(verify, cardID, begin, end)
    req = urllib2.Request(url)
    response = urllib2.urlopen(req)
    content = response.read()
    dict = json.loads(content)

   # the rest part of the code is omitted


if __name__ == '__main__':
    print '\n悠遊卡餘額明細查詢 v{}'.format(version)
    print '{}\n'.format(copyright)
    if len(sys.argv) > 1:
        try:
            print '\n{:=^90}\n'.format('[ 查詢開始 ]')
            proc(str(sys.argv[1]))
        except ValueError as err:
            pass
    else:
        while 1:
            try:
                data = raw_input('請輸入卡片號碼：').replace(' ', '')
                if len(data):
                    print '\n{:=^90}\n'.format('[ 查詢開始 ]')
                    proc(data)
                else:
                    break
            except ValueError as err:
                pass


#+++ okay decompyling easycard.pyc
# decompiled 1 files: 1 okay, 0 failed, 0 verify failed
# 2015.05.04 17:41:47 CST

Bill of Tainan trip

BJ4

https://gist.github.com/xatier/5b2a71233e40d12408e1

More notes for shadowsocks

I wrote a note for the usage of shadowsocks few days ago: http://xatierlike.blogspot.tw/2015/03/note-for-shadowsocks.html

I spent some time to dig into the source code of the project and came up with this note, just a note for what I've found. :D


License

shadowsocks is under Apache 2.0.

autoban

There's a script called autoban.py under shadowsocks/utils .

According to the official wiki, that is used for banning brute force crackers.
https://github.com/shadowsocks/shadowsocks/wiki/Ban-Brute-Force-Crackers

Actually, autoban.py is implemented by iptables , this script looks at the log and find something like this and grab the remote remote IP out.

'2015-04-07 16:42:26 ERROR    can not parse header when handling connection from 61.157.96.193:27242'

if 'can not parse header when' in line:
    ip = line.split()[-1].split(':')[0
    ...
    cmd = 'iptables -A INPUT -s %s -j DROP' % ip
    print(cmd, file=sys.stderr)
    os.system(cmd)


Versions

shadowsocks uses a very strange way to determine it's running under python2 or python3.

if bytes == str

This is True in python2 but False in python3.

Also, here's a strange logic to check the version.

I will write if info[0] == 2 and info[1] < 6 rather than the author does.

def check_python():
    info = sys.version_info
    if info[0] == 2 and not info[1] >= 6:
        print('Python 2.6+ required')
        sys.exit(1)
    elif info[0] == 3 and not info[1] >= 3:
        print('Python 3.3+ required')
        sys.exit(1)
    elif info[0] not in [2, 3]:
        print('Python version not supported')
        sys.exit(1)
     
     
Argument parsing

The entry points of sslocal and ssserver commands are the main functions in local.py and server.py, respectively.

shell.py checks command line arguments and checks the configuration files, it's using getopt, I think that should should be rewrite with argparse .


Event loop and Relay:

Basically shadowsocks abstracts three kinds of polling system: epool, kqueue and system select, it will use them in order if available.

class EventLoop(object):
    def __init__(self):
        self._iterating = False
        if hasattr(select, 'epoll'):
            self._impl = EpollLoop()
            model = 'epoll'
        elif hasattr(select, 'kqueue'):
            self._impl = KqueueLoop()
            model = 'kqueue'
        elif hasattr(select, 'select'):
            self._impl = SelectLoop()
            model = 'select'
        else:
            raise Exception('can not find any available functions in select '
                            'package')


So basically shadowsocks has a local server connected to the SOCK5 proxy and send data to remote via TCP/UDP relays.

Both of TCP/UDP relays will encrypt the payload with specified algorithms.

Here's the diagram of the idea of shadowsocks.

browser <== SOCKS proxy ==> local <== TCP/UDP relays ==> remote => free world

browser <== plain text ==> local <= encrypted data => GFW <= encrypted data => remote => free world


Pretty similar to SSH tunnel, right?

browser <= socks proxy => ssh client <= tunnel => ssh server => free world

The feathers of SSH handshaking traffic is easily blocked by GFW, shadowsocks are just simple standard TCP/UDP traffic with unknown/encrypted payloads.


The following is from the comments of tcprelay.py and rdprelay.py .

TCP Relay

# for each opening port, we have a TCP Relay
# for each connection, we have a TCP Relay Handler to handle the connection
# for each handler, we have 2 sockets:
#    local:   connected to the client
#    remote:  connected to remote server

# as sslocal:
# stage 0 SOCKS hello received from local, send hello to local
# stage 1 addr received from local, query DNS for remote
# stage 2 UDP assoc
# stage 3 DNS resolved, connect to remote
# stage 4 still connecting, more data from local received
# stage 5 remote connected, piping local and remote

# as ssserver:
# stage 0 just jump to stage 1
# stage 1 addr received from local, query DNS for remote
# stage 3 DNS resolved, connect to remote
# stage 4 still connecting, more data from local received
# stage 5 remote connected, piping local and remote

UDP Relay

# HOW TO NAME THINGS
# ------------------
# `dest`    means destination server, which is from DST fields in the SOCKS5
#           request
# `local`   means local server of shadowsocks
# `remote`  means remote server of shadowsocks
# `client`  means UDP clients that connects to other servers
# `server`  means the UDP server that handles user requests

Misc

shadowsocks implements its own DNS query (asyncdns.py) and LRU caching  (lru_cache.py) system.

Reference:
http://vc2tea.com/whats-shadowsocks/
http://gpio.me/readcode-ShadowSocks.html

steal LINE stickers to telegram

1. pull all LINE stickers from you phone

adb pull /storage/sdcard0/Android/data/jp.naver.line.android/stickers/ .

2. find your purchased sticker pack (you can look at the "preview" file)

for example, the sticker pack of Puella Magi Madoka Magica is # 1101

Note, you also can find stickers sent by your friend.

3. convert your the files (from PNG) to the WebP format

I'm too lazy so I wrote a script to do the following stuffs:

mkdir madoka
cp ../stickers/1101/* madoka/
cd chocola/
rm *_key preview thumbnail *.tmp
for i in *; do cwebp $i -o $i.webp ; done
for i in `ls | grep -v webp`; do mv $i $i.png; done
cd ..

4. put them in your phone, done!



Test:

The first one is the file in PNG, the second one is in WebP.

Automatic script: https://gist.github.com/xatier/971e1abe16f3bcbc51d9

Reference
https://telegram.org/blog/stickers
https://developers.google.com/speed/webp/docs/using

netowrk speed test between two linux boxes

This trick with nc and dd can be used as speed testing between two linux boxes.

Server:
nc -vvlnp 12345 > /dev/null

Client:
dd if=/dev/zero bs=1M count=1k | nc -vvn <server IP> 12345

ramdisk for chromium

I just bought two new ram for my desktop.
HyperX FURY black DDR3-1866 8GB(8GBx2) (HX318C10FBK2/16)
http://24h.pchome.com.tw/prod/DRAL1H-A90052AUN?q=/S/DRAL1I


/etc/fstab
tmpfs     /tmp/cache    tmpfs     nodev,nosuid,noatime,size=2G      0 0

$ chromium --disk-cache-dir=/tmp/cache

$ mount | grep cache
tmpfs on /tmp/cache type tmpfs (rw,nosuid,nodev,noatime,size=2097152k

manually create a ramdisk
$ mkdir -p /tmp/ram
$ sudo mount -t tmpfs -o size=1024M tmpfs /tmp/ram/


old post: http://xatierlike.blogspot.tw/2012/08/create-ramdisk-on-ubuntu.html

GnuPG notes

Some notes for GnuPG

# usually we can replace <key ID> with <user ID>
# keygen
gpg --full-gen-key (choose DSA & Elgamal here)


# editing
gpg --edit-key <key ID>


# key listing
gpg -k [ <user ID> or <key ID> ]
gpg -K [ <user ID> or <key ID> ]

-k: --list-public-keys / --list-keys
-k: --list-secret-keys


# fingerprint
gpg --fingerprint [ <key ID> ]


# import and export (backup & restore)
gpg --import filename

gpg --export <key id>
gpg --export-secret-keys <key id>

# --armor (-a): ASCII text format
gpg -a --export <key id>
gpg -a --export-secret-subkeys <key id>

gpg --enarmor filename.gpg
gpg --dearmor filename.asc


# keyserver
gpg --keyserver pgp.mit.edu --send-keys <key ID>
gpg --keyserver pgp.mit.edu --recv-keys <key ID>
gpg --keyserver pgp.mit.edu --search-key <key ID>


# encryption and decryption
gpg -e filename
gpg -r <key ID> -e filename
gpg -o filename -d filename.gpg

-e: --encrypt
-d: --decrypt
-o: --output
-r: --recipient


# signature
(in place signature)
gpg --sign filename
gpg --clearsign filename
gpg --verify filename.gpg

(saperated key)
gpg --detach-sign filename
gpg -a --detach-sign filename
gpg --verify filename.sig filename


# sign-key
gpg --sign-key <user id>

OpenVPN with VPN Gate

VPN Gate is a project by University of Tsukuba, Japan.
http://www.vpngate.net/en/

Basically there're thousands of relay servers hosted by volunteers around the world.

As a Linux user, the easiest way to connect to VPN Gate servers is OpenVPN.

Just install openvpn from the official repositories.

sudo pacman -S openvpn

and randomly grab a configuration file like this: vpngate_vpn197292320.opengw.net_udp_1786.ovpn

http://www.vpngate.net/en/do_openvpn.aspx?fqdn=vpn197292320.opengw.net&ip=223.223.103.92&tcp=1620&udp=1786&sid=1427186718705&hid=1019730

Then use openvpn client to read the config file and connect to the free internet:

sudo openvpn vpngate_vpn197292320.opengw.net_udp_1786.ovpn

I wrote simple Perl script to get OpenVPN configs from VPN Gate

https://gist.github.com/xatier/8911e8737089e9eaa236

That will show you a list of available VPNs and save the config file for you.

demo:

Reference:
http://www.vpngate.net/en/
https://wiki.archlinux.org/index.php/OpenVPN

Note for shadowsocks

Shadowsocks is a popular open sourced tunneling tool in China.

https://github.com/shadowsocks/shadowsocks
https://www.archlinux.org/packages/community/any/shadowsocks/

(both server and client)
sudo pacman -S shadowsocks

Salsa20 & Chacha20 support
sudo pacman -S libsodium python2-numpy python2-salsa20

/etc/shadowsocks/config.json
{
"server":"remote-shadowsocks-server-ip-addr",
"server_port":8888,
"local_address":"127.0.0.1",
"local_port":1080,
"password":"your-passwd",
"timeout":300,
"method":"aes-256-cfb",
"fast_open":false,
"workers":1
}


server
sudo ssserver -c /etc/shadowsocks/config.json --user nobody

run as daemon
sudo ssserver -c /etc/shadowsocks/config.json --user nobody -d start
sudo ssserver -d stop

client
sslocal -c /etc/shadowsocks/config.json

run as daemon
sudo sslocal -c /etc/shadowsocks/config.json -d start

sudo sslocal -c /etc/shadowsocks/config.json -d stop

Chromium: use Proxy SwitchyOmega and connect to a local socks5 proxy
https://chrome.google.com/webstore/detail/proxy-switchyomega/padekgcemlokbadohgkifijomclgjgif

Android Client
https://play.google.com/store/apps/details?id=com.github.shadowsocks

QR code for Android client
sudo pacman -S python2-qrcode
echo -n "ss://"`echo -n aes-256-cfb:password@1.2.3.4:8388 | base64` | qr


References:

https://github.com/shadowsocks/shadowsocks/wiki
https://wiki.archlinux.org/index.php/Shadowsocks_%28%E7%AE%80%E4%BD%93%E4%B8%AD%E6%96%87%29