README

GCJ Commandline submit tool, v1.0-beta4
Copyright 2010 Google Inc. All Rights Reserved.

Download and submit tool for the Google Code Jam
-------------------------------------------------------------------------------
TABLE OF CONTENTS:

0. Disclaimer
1. Prerequisites
2. Basic usage
  a. Configuring the system
  b. Initializing the contest
  c. Downloading an input
  d. Submitting a solution
  e. Checking your status
  f. Cleaning the contest
3. Configuration files
4. Executable files
  a. Contest crawler
  b. Login agent
  c. Input downloader
  d. Solution submitter
  e. Contest cleaner

-------------------------------------------------------------------------------
0. DISCLAIMER

GCJ Commandline submit tool, v1.0-beta3
Copyright 2010 Google Inc. All Rights Reserved.

This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program.  If not, see <http://www.gnu.org/licenses/>.

-------------------------------------------------------------------------------
1. PREREQUISITES

To install this software, just copy these files into any folder and execute the
python scripts from the main folder (the one that contains this file).

To run the scripts, you will need Python 2.6, which can be downloaded from
<http://www.python.org/download/releases/>. This was developed and tested using
Python 2.6.5, so we recommend using 2.6.5 or later.

-------------------------------------------------------------------------------
2. BASIC USAGE

To use this tool, open a terminal in the main folder (the one that contains
this file). The scripts in this directory expect to be run from this directory,
otherwise, they will not find the configuration files in the config/ directory.


2a. Configuring the system

Before attempting to run a script included with this software, you should
configure the system by writing some needed information in the user
configuration file, located at config/user_config.py. This file is formatted as
a Python dict structure, but a basic one is be included with this software.
Three values are needed to run the software:

- host: The name of the website hosting the contest. For normal contests and
  for the remaining of this example, the host is 'code.google.com'.

- user: Your google account, for example 'myaccount@gmail.com'.

- source_names_format: This is a list of strings that specify the default
  sources to be included with an output file. In this example, we will assume
  we use C++ and set it to ['{problem}-{input}-{id}.cpp']. Remember to change it
  before a contest or you might end submitting the included source code by
  accident!

Look at Section 3 for more information about these and many more configuration
parameters.


2b. Initializing the contest

The first step to use this software is initializing the contest. To do this
you have to specify the contest id, which can be found in the dashboard URL
for that contest. For example, for the GCJ 2010 Qualification Round the
dashboard URL is:

http://code.google.com/codejam/contest/dashboard?c=433101

so you can initialize the contest executing:

$ ./gcj_init_contest.py 433101

During this step, the software logs into Code Jam servers (it may ask for your
password) and extracts information from the contest, including problem ids,
names and the middleware tokens required to transfer information.

Then, all this information, including the login cookie and the information in
the user configuration file, is stored in the current configuration file (found
at config/current_config.py).

If you don't specify a contest id for this script, it will attempt to initialize
the tool for the current contest. If no contest is running it will print an
error message.

NOTE: Your password is not stored in the current configuration file if you did
not entered it in the user configuration file.


2c. Downloading an input

Once you have implemented a solution and think you got it right, you may
download an input file using the gcj_download_input.py script. This script
receives the problem letter (A, B or C for the GCJ10 Qual), the I/O set name and
an id to differentiate it from other attempts.

If a problem has only one small I/O set it will be called "small", but if it has
multiple small I/O sets they will be called "small1", "small2", "small3", and so
on (all quotes for clarity). Large I/O sets are named in the same way.

For example, to download a small input for the problem A, Snapper Chain, you
can execute:

$ ./gcj_download_input.py A small 0

which will download the input file and store it in source/A-small-0.in. Now
it is time to run your solution and generate the output file.

NOTE: Downloading an input file twice in the same attempt will just redownload
the same input file if its timer has *NOT* expired. Otherwise, it will download
another input and start another attempt for small I/O sets. However, if this is
the case it will ask for confirmation from the user.


2d. Submitting a solution

After running your solution and generating the output, it is time to submit the
output file with the source files using the gcj_submit_solution.py script. This
script receives the problem letter, the I/O set name and the id used when the
input was downloaded. See section 2c for I/O set naming.

For example, to submit the solution for the small I/O set of problem A, Snapper
Chain, you can execute:

$ ./gcj_submit_solution.py A small 0

which will submit the file source/A-small-0.out as the answer and
source/A-small-0.cpp as the only source file. If you want to add extra source
files, you may use the -a flag as many times you need. For example, if you want
to submit two extra files called library.cpp and template.txt you can execute:

$ ./gcj_submit_solution.py -a library.cpp -a template.txt A small 0

which will submit the file source/A-small-0.out as the answer and
source/A-small-0.cpp, library.cpp and template.txt as the source files. If you
add a directory using the -a option, the program will compress it into a zip
file before submitting it.

After submitting the file, the Code Jam server will answer with a message
string. The most important values are:

- 'Incorrect: Wrong Answer': Your solution for the small input was wrong, try
  again.
- 'Correct: Good Job!': Your solution for the small input was right,
  congratulations!
- 'Submitted: Your answer has been received and will be judged at the end of
  the contest.': This is received after submitting the output for a large input.
- 'Rejected: <reason>': Your submission was rejected for some reason, please
  try again. This attempt did not count as a wrong answer.


2e. Checking your status

During the contest you can check your status by using the gcj_get_status.py
script, which will show you your current rank, number of points, time at which
each input was solved, number of wrong attempts per input and remaining time for
current attempts.

To run this script you should execute:

$ ./gcj_get_status.py

which will show an output like this one:

Rank: 194 (66 points)
Problem: Snapper Chain
  small: 00:15:16 (2 wrong tries)
  large: 00:19:11
Problem: Fair Warning
  small: --:--:-- (1 wrong try) << 2:31 left
  large: --:--:--
Problem: Theme Park
  small: 01:42:45
  large: 01:47:18 << 6:38 left


2f. Cleaning the contest

After the contest is over, you can clean all contest-related information using
the ./gcj_clear_contest.py script, which receives no parameters. This script
will not delete your source code, input or output files.

-------------------------------------------------------------------------------
3. CONFIGURATION FILES

This tool needs one configuration file to run, which can be located at
config/user_config.py. This file is formatted as a python dict structure and
may have the following keys:

- 'host' : The name of the website hosting the contest. For normal contests,
  it is 'code.google.com'.

- 'user' : Your google account, the '@gmail.com' might be omitted for contests
  in 'code.google.com'.

- 'password' : The password for your google account. This value is optional, if
  it is not found the executables will ask for it when necessary.

- 'data_directory' : Directory where input files are written and the output and
  main source file is read from. This value is optional, if not specified it
  will default to './source'.

- 'input_name_format' : Format used to generate filenames for input files,
  where '{problem}' will be substituted for the problem letter, '{input}' for
  the input type (small or large) and '{id}' for the submit id. This value is
  optional, if not specified it will default to '{problem}-{input}-{id}.in'.

- 'output_name_format' : Format used to generate filenames for output files,
  whose format is similar to the input format. This value is optional, if not
  specified it will default to '{problem}-{input}-{id}.out'.

- 'source_names_format' : List with filename formats for the source files
  included by default; each list value must be formatted in the same way as the
  input name format. This value is *NOT* optional, to prevent accidental
  submits of outputs with no source files.

All these values (with the exception of 'host', 'user' and
'source_names_format') can be overriden in the scripts by using command line
options. When these values are specified on the command line, the values in the
configuration files and the default values are ignored.

For the input, output and source format the final filename is generated using
the Python str.format() function.

If you want to change the path of the configuration file, look at the
lib/constants.py file and change the respective variables. Remember that those
paths are relative to the current directory, not to the lib directory. You could
also use absolute paths or use the placeholder '{base_dir}' (quotes for
clarity); this placeholder will be substituted by the directory where the tool
is in the filesystem.

IMPORTANT: Scripts stores information from the current contest along with the
user information in the current configuration file (config/current_config.py).
However, any change you make in the user configuration file will be recognized
by the scripts, even if the current configuration file is not updated. This
implies that the user should update their information in the user configuration
file, otherwise the scripts will not see the change.

-------------------------------------------------------------------------------
4. EXECUTABLE FILES

This tool is composed by four executable scripts, one to initialize the tool
for a contest, one to download inputs, one to submit solutions and finally one
to login in the Google servers, just in case the authentication cookie becomes
invalid.

All scripts provide a --version and a --help options. These options will show
the current tool version and a list with all the possible options for the
script, including a brief description of each one.

All scripts might require your Google password if you do not specify it in the
configuration file. Another option is to pass the password using the option
--passwd, just be careful that specifying the password this way will be visible
in your screen and will be stored in the commands history.

The scripts to download input and submit outputs support a --login option; if
this option is specified the current login cookie will be ignored and a new one
will be requested. Also, if the cookie has expired the scripts will
automatically try to renew it. Beware that this might cause the executables to
ask for your password. Also, these scripts ask confirmation from the user before
downloading an input that will start a new timer and before resubmitting a
solution for a large input. These confirmations can be skipped by passing the
--force option.


4a. Contest crawler

This script is the gcj_init_contest.py script and receives one optional
positional argument which is the contest id (see Section 2b for instructions on
how to retrieve a contest id), that is used to retrieve the important
information of the contest, i.e., number of problems, middleware authentication
tokens, problem ids, etc.

If no contest id is specified in the arguments, this script will retrieve the
id of the current contest and initialize the tool for it. If there is no running
contest an error message will be shown to the user.

This script should be run once before each contest, so the correct information
is retrieved. Otherwise, all other scripts might fail because no contest data
will be found.


4b. Login agent

This script is the gcj_renew_login.py script and receives no positional
arguments, and will renew the login cookie in the current configuration file
with a new one.

This script should not be used under regular operation, this is just included
in case your login cookie expires or becomes invalid during a contest.


4c. Input downloader

This script is the gcj_download_input.py script and receives three positional
arguments, the problem letter, the input size ('small' or 'large') and a submit
id. The problem letter and the input size are always converted to uppercase and
lowercase before using them, respectively, but the id is case-sensitive.

This script downloads the specified input and stores it in a file inside the
data directory. However, the flags --data-directory and --input-name are
provided to override the default values and the ones in the configuration files.


4d. Solution submitter

This script is the gcj_submit_solution.py script and receives three positional
arguments, the problem letter, the input size ('small' or 'large') and a submit
id. Similar to the input downloader, both the problem letter and the input size
are converted to uppercase and lowercase before using them, respectively, but
the id is case-sensitive.

If a directory is specified as an output file, this executable will zip it into
a file before submitting, so the whole folder can be sent to the server.

By default this script submits the specified output files along with the
additional default source files specified in the configuration file. However,
this script provides many options to alter its default behaviour:

- Similar to the input downloader, the --data-directory option can be used to
  choose another data directory, and the --output-name option can be used to
  change the default name of the output file.

- If additional source files are needed, the --add-source option can be used to
  submit them along with the output files.

- If the source files being uploaded are too big, the --zip-sources option can
  be used to zip all the source files into a single file before submitting. This
  options is smart and will not zip already zipped files.

- The --ignore-zip option will indicate that all zip files inside a included
  directory should be ignored, otherwise they will be added. However, zip files
  specified directly in the source configuration or using the --add-source will
  be included.

- The --ignore-default-source option will indicate that all default source files
  in the configuration file should be ignored, and only source files specified
  through the --add-source option should be sent.

- There are two different encondings to send the output to the server. If the
  --gzip-content flag is specified the output and the source files will be
  sent using gzip; if the --nogzip-content flag is specified the output and the
  source files will be sent using text/plain. If no flag is specified, the data
  will be sent using gzip.


4e. Contest cleaner

This script is the gcj_clear_contest.py script and receives no positional
arguments. When this script is executed all contest-related information will be
removed from the system, which does not include your source code, input or
output files.