Monthly Archives: June 2016


fn is no fun


Installing Task Translation System for IOI competitions (Linguist) on Ubuntu Server 14.04LTS

First, install any missing (also for font) related dependencies:

sudo apt-get update;
sudo apt-get install git unzip build-essential chrpath libssl-dev libxft-dev libfreetype6 libfreetype6-dev libfontconfig1 libfontconfig1-dev fontconfig fontconfig-config fonts-dejavu-core fonts-droid fonts-freefont-ttf fonts-kacst fonts-kacst-one fonts-khmeros-core fonts-lao fonts-liberation fonts-lklug-sinhala fonts-nanum fonts-opensymbol fonts-sil-abyssinica fonts-sil-padauk fonts-takao-pgothic fonts-thai-tlwg fonts-tibetan-machine fonts-tlwg-garuda fonts-tlwg-kinnari fonts-tlwg-loma fonts-tlwg-mono fonts-tlwg-norasi fonts-tlwg-purisa fonts-tlwg-sawasdee fonts-tlwg-typewriter fonts-tlwg-typist fonts-tlwg-typo fonts-tlwg-umpush fonts-tlwg-waree -y;

Then, get a copy of the repository:

git clone https://github.com/ioi/translation.git

Switch to the newly created directory:

cd translation/

And, add the gpg keys needed to make the setup.
The gpg command in the setup will contact a public key server (hkp://keys.gnupg.net) and request the key associated with the RVM project key which is used to sign each RVM release. To get the key, we need to provide the ID that is related to the key, in this case the ID is 409B6B1796C275462A1703113804BB82D39DC0E3. Having the RVM project’s public key allows us to verify the legitimacy of the RVM release we will be downloading, which is signed with the matching private key.

gpg --keyserver hkp://keys.gnupg.net --recv-keys 409B6B1796C275462A1703113804BB82D39DC0E3

If the above command fails with the following error

$ gpg --keyserver hkp://keys.gnupg.net --recv-keys 409B6B1796C275462A1703113804BB82D39DC0E3
gpg: requesting key D39DC0E3 from hkp server keys.gnupg.net
?: [fd 4]: read error: Connection reset by peer
gpgkeys: HTTP fetch error 7: couldn't connect: eof
gpg: no valid OpenPGP data found.
gpg: Total number processed: 0

Try using the following alternative:

command curl -sSL https://rvm.io/mpapis.asc | gpg --import -

Afterwards, perform the installation (this might take some time, depending on your connection to the server where it will download the necessary packages and your CPU performance as it will perform some compilation locally):

./deploy.sh

While installing codemirror we got the following prompt:

replace public/codemirror-3.22/doc/upgrade_v3.html? [y]es, [n]o, [A]ll, [N]one, [r]ename:

We typed A and pressed Enter. We are not sure if this is what we are supposed to do but it worked later on properly.

Edit the ./config.yml file and set new values for the api_token and cookie_secret.

api_token: "4c0a6fe55f3d4aa9c5dbb9a59db7b20e"
cookie_secret: "SDfadadsf90u84oh23jnrwf"
  • api_token is a 32 characters long random key. It can contain only character a-f and numbers 0-9.
  • cookie_secret is a 23 characters long value. This example contains any character from a-z and A-Z and the numbers 0-9.

You can generate random passwords from here http://bytefreaks.net/random-password-generator

Then start the redis service using the following:

redis-server

 

In an new terminal, go to DbInit folder:

cd translation/DbInit/

and update the files users.json and tasks.json to prepare the initial data to be imported in redis:

  • users.json Be sure to update the passwords of the users.
  • tasks.json Set the names of your tasks and the .md filenames of the original content.

The original content of tasks.json is:

[
    { "id": "1",    "title": "notice",     "filename": "notice.md" },
    { "id": "2",    "title": "gondola",    "filename": "gondola.md" },
    { "id": "3",    "title": "friends",    "filename": "friends.md" },
    { "id": "4",    "title": "holiday",    "filename": "holiday.md" }
]

Place the original .md files in this folder. And then initialize redis using the following command:

ruby dbinit.rb

Finally, go to the previous folder and start the translation system:

cd ..;
shotgun -o 0.0.0.0 -p 8080;

Visit http://SERVER_NAME_OR_IP:8080 to view the translation system.

You can use the admin account to make changes to users, to send notifications (if you omit the to field, then the notification will be send to all users) and to check out all task with their translations.

Staff accounts allow you to check out all task with their translations.

Note on the architecture

The deploy.sh script assumes your architecture and OS is 64bit. To find what type of architecture you are using execute uname -i. If the result is not x86_64, then phantomjs will not work for you and you will not be able to generate the PDFs. To fix this issue you need to download the correct version from https://bitbucket.org/ariya/phantomjs/downloads. At the time this tutorial was written deploy.sh was installing version https://bitbucket.org/ariya/phantomjs/downloads/phantomjs-1.9.7-linux-x86_64.tar.bz2, so we installed https://bitbucket.org/ariya/phantomjs/downloads/phantomjs-1.9.7-linux-i686.tar.bz2 using the following commands:

wget https://bitbucket.org/ariya/phantomjs/downloads/phantomjs-1.9.7-linux-i686.tar.bz2;
tar xf phantomjs-1.9.7-linux-i686.tar.bz2;
sudo cp phantomjs-1.9.7-linux-i686/bin/phantomjs /usr/local/bin;
rm -rf phantomjs-1.9.7-linux-i686*;

The above commands follow the example in deploy.sh.

Updating logos and website look

To update the views of the website you can do the by editing the .erb files that are in the translation/views folder.

Examples:

  • login.erb You need to modify this file to change the login screen
  • tasks_index.erb This file holds the structure of the main index the user sees after login (/tasks)
  • _navbar.erb Modify this file to change the navigation bar on top of each page
  • .. and more

C/C++: How NOT to define multiple pointers

When defining pointers in C/C++ you should be careful on how you use the * characters. If you try to define multiple pointers on the same line and you do not add the * character in front of each variable, then the results will not be what you would expect. In the following examples we added the * to the data type definition, hoping that all variables would become pointers of that data type. Unfortunately, as the compiler points out later on while making the comparisons, only the first variable in each line becomes a pointer of the data type.

Wrong Examples

C Source Code:

#include <stdio.h>

int main()
{
    int* a, b;
    int *c, d;
    int * e, f;

    a == b;
    c == d;
    e == f;

    return 0;
}

Compiler Output:

$ gcc -o main *.c                                                                                          
main.c: In function 'main':                                                                                      
main.c:9:7: warning: comparison between pointer and integer                                                      
     a == b;                                                                                                     
       ^                                                                                                         
main.c:10:7: warning: comparison between pointer and integer                                                     
     c == d;                                                                                                     
       ^                                                                                                         
main.c:11:7: warning: comparison between pointer and integer                                                     
     e == f;                                                                                                     
       ^                                                                                                         

C++ Source Code:

#include <iostream>
using namespace std;

int main()
{
    int* a, b;
    int *c, d;
    int * e, f;

    a == b;
    c == d;
    e == f;

   return 0;
}

Compiler Output:

$ g++ -std=c++11 *.cpp -o main

main.cpp: In function ‘int main()’:
main.cpp:10:10: error: ISO C++ forbids comparison between pointer and integer [-fpermissive]
     a == b;
          ^
main.cpp:11:10: error: ISO C++ forbids comparison between pointer and integer [-fpermissive]
     c == d;
          ^
main.cpp:12:10: error: ISO C++ forbids comparison between pointer and integer [-fpermissive]
     e == f;
          ^

Correct Examples

#include <stdio.h>

int main()
{
    int* a, * b;
    int *c, *d;
    int * e, * f;

    a == b;
    c == d;
    e == f;

    return 0;
}
#include <iostream>
using namespace std;

int main()
{
    int* a, * b;
    int *c, *d;
    int * e, * f;

    a == b;
    c == d;
    e == f;

   return 0;
}

MySQL Workbench: A solution to Error Code 1175 2

While using MySQL Workbench, we tried to execute an UPDATE command that did not use the Key of the table in the WHERE clause. In our case, it did not matter and it was expected to perform that update without that restriction.

When we tried to execute the code though, we got the following error:

Error Code: 1175. You are using safe update mode and you tried to update a table without a WHERE that uses a KEY column

and the following solution from Workbench:

To disable safe mode, toggle the option in Preferences -> SQL Editor and reconnect.

We did not want to close the session for many reasons (we had data in other tabs, we would lose the undo history, we would have to revert the change later on, etc.).

To solve this issue without reconnecting nor changing the configuration of the tool, we appended the following code above the update statement and executed both lines together (we selected the two commands and pressed CTRL+Shift+Enter):

SET SQL_SAFE_UPDATES = 0;

This line of code, instructed the tool to ignore Error 1175 and allowed us to complete our task.