Dependencies

The program is mainly based on the C++ standard library, except for:

Catch (http://catch-lib.net/) - a unit testing framework for C++
CLI11 (https://github.com/CLIUtils/CLI11) - framework to parse CLI parameters

The two dependencies are included in 3rdparty/ and are to no concern of a user.

Usage

Follows the task rather closely.

Use --db=path/to/dict/ (don't forget a trailing directory separator) to pass the path to your database. Otherwise the current working directory will be assumed to hold the database.

--help can be used to gain information about possible parameters:

Usage: ./tt1_wordnet [OPTIONS] [SUBCOMMAND]

Options:
  -h,--help                   Print this help message and exit
  --db TEXT=./                Path to the dict/ directory, where the WordNet database is located.

Subcommands:
  query                       
  semrels                     
  shortestpath

Task 1 - query

Usage: ./tt1_wordnet query [OPTIONS]

Options:
  -h,--help                   Print this help message and exit
  --pos TEXT in {a,n,r,v}     Part of speech
  --iw TEXT                   Index word
  --full                      Show index words for all synsets

Example invocation: ./tt1_wordnet --db=/home/user/WordNet/dict/ query --pos=n --iw=calculator --full

Task 2 - semrels

Usage: ./tt1_wordnet semrels [OPTIONS]

Options:
  -h,--help                   Print this help message and exit
  --pos TEXT in {a,n,r,v}     Part of speech
  --offset UINT               Offset
  --dir TEXT in {any,in,out}  Relation direction
  --full                      Show index words for all connected synsets

Task 3 - shortestpath

Usage: ./tt1_wordnet shortestpath [OPTIONS]

Options:
  -h,--help                   Print this help message and exit
  --pos1 TEXT in {a,n,r,v}    Part of speech
  --offset1 UINT              Offset
  --pos2 TEXT in {a,n,r,v}    Target part of speech
  --offset2 UINT              Target offset
  --directed                  Only search directed
  --full                      Show index words for all connected synsets

Task 4 - eccentricity

Usage: ./tt1_wordnet eccentricity [OPTIONS]

Options:
  -h,--help                   Print this help message and exit
  --pos TEXT in {a,n,r,v}     Part of speech
  --offset UINT               Offset
  --directed                  Only search directed

Code Organisation

`main.cpp`

command line parsing
query of Database for the necessary data
printing of the data

`Database.{h,cpp}`

implementation of algorithms on the data
loading of all data on construction
maintenance of indexes of the data

`FileAccess.{h,cpp}`

low level cache of input streams
mapping of requests for certain data to files

`Input.{h,cpp}`

implementation of operator>> for the different data structures

Reading of data is based on std::istringstream. It automatically parses text data into C++ data types like int, std:string (up to the next whitespace), and custom structures. Custom structures are read using operator>> functions, which are implemented in Input.cpp.

`Output.{h,cpp}`

implementation of output<< for visualization of data

Writing data on std::cout also works via operator overloads (of operator<<). The necessary operators are implement in Output.cpp.

`wordnet.h`

Definition of data structures. They are closely modelled after the file based database, so that they can be read by the operators in Input.{h,cpp}.

Development Environment

This was developed using CLion (https://www.jetbrains.com/clion/), but every other C++-IDE with CMake support should be enough:

QtCreator
Visual Studio
KDevelop
CLion

Of course, an editor can be enough, too.

Prerequisites:

C++14 compiler (Clang, GCC, Visual Studio)
CMake

Steps to compile:

tt1_wordnet$ mkdir build
build$ cmake ..
build$ cmake --build .
build$ ./tt1_wordnet --db=path/to/WordNet/ --help

A few common structures

PartOfSpeech (pos)

Is defined as char. Unfortunately, the CLI parameter parser only knows strings (std::string) as option values. These strings are converted to char like so: char partOfSpeech = partOfSpeechOptionValue[0].

LemmaIndexItem

One line in the index files.

SynsetPointer

Structure modelling a pointer between two synsets.

SynsetConnection

A SynsetConnection is relative to an "origin" element: SynsetConnection::pointer is the pointer between the two synsets. It is important that "origin" does not own the pointer, i.e. the pointer can also be pointing at origin. It depends on the context of where SynsetConnection is used.

SynsetIdentifier

A Synset can be identified by the database it is saved in (noun, verb, adjective, adverb) and its offset in the database. The SynsetIdentifier simply allows to handle these two data points in one object (first is the one-character-name for the database, second the offset in the database).

jangmarker/tt1_wordnet