Spellchecker - our 2nd year C project

Context behind the project

In the LEPL1503 UCLouvain class, we had to code a C program called spellchecker which given an input file and dictionaries, can detect and can optionally correct mistakes generating binary output files
In order to achieve that, we had to program an efficient, fast and clean program capable of using multiple threads to speed up the process

Our group members - Group S1 :

• Younes - Yoouss
• Sofia - SCY-Sofia
• Maryam - MaryamAkraiche
• Adjovi - adjovimalaika2006
• Maurice - SuperMorisl

Caution

If a link is broken, please refer to the Contributors page

Please note that my GitHub contribution stats (additions and deletions) are misleading because I had to delete and re-upload heavy files (such as those in the benchmark folder)
My actual stats are much closer to those of Sofia - SCY-Sofia

Documentation

• The architecture & modules
• How to run tests & Valgrind
• Benchmarks & how to use them
• The LEPL1503 instructions (in French)

Installation

Requirements

Debian based users (WSL, Ubuntu, Mint, Pop_OS!, ...)

sudo apt install build-essential cmake libcunit-dev valgrind

Red Hat based users (Fedora, ...)

sudo dnf install gcc gcc-c++ make cmake CUnit-devel valgrind

Arch based users (Arch, EndeavourOS, ...)

sudo pacman -S gcc make cmake cunit valgrind

MacOS users

brew install gcc cmake cunit valgrind

Cloning the repo

• Via https :

git clone https://github.com/Yoouss/b2-spellchecker.git

• Or ssh :

git clone git@github.com:Yoouss/b2-spellchecker.git

Compiling and executing the spellchecker program

1) Compilation

make

Compiles the program's file (spellchecker)

---

2) Execution

./spellchecker --dicts <dictionary_path> --input <input_path> [--output <output_path>] [--mode <mode>] [--threads <number>] [--verbose]

Here's the specification of each arguments, the ones in brackets being optionals :

1. --dicts <dictionary_path> : Specifies the dictionaries' path

   The user can also choose a single dictionary or use all as an argument to specify all the dictionaries from the dicts directory

   -> examples : dicts, all, dicts/fr.dict

2. --input <input_path> : Specifies the input file's path

   -> example : inputs/input_10l.txt

3. --output_path <output_path> : Specifies the output files' name, without extension

   -> example : output

   If this argument isn't specified, the terminal will be used by default, the user then has 2 options :

   ./spellchecker ... > <output_path>

   Use a chevron > to get the binary output in the file <output_path> specified

   ./spellchecker ... | hexdump

   Or use | hexdump to see the binary output on the terminal

4. --mode <mode> : Specifies the program mode between :

   • detection : detects errors and writes them in stdout or in a new file <output_path>.err if <output_path> is specified

   • correction : detects errors, corrects them and writes them in stdout or in new files <output_path>.err for the detection and <output_path>.fix for the correction if <output_path> is specified

   If the mode isn't specified, the detection mode will be used by default

5. --threads <number> : Specifies the number of threads used

   • If <number> is smaller or equal to 1, the program will be executed in a mono-threaded way (sequentialy)

   • If <number> is greater than 1, the program will be executed in a multi-threaded way using <number> threads maximum per functions with configurable parallelism

6. --verbose : Displays more details in real time

3) Example of usage

make

./spellchecker --dicts dicts --input inputs/input_50l.txt --output my_output --mode correction --threads 2

Conclusion

This project has been a good opportunity to work together in groups despite the lack of time and motivation from some classmates (of course not those mentionned in the group members section !)

Overall it forced us to follow similar practices, to review others' code via forge (UCLouvain's Gitlab), to manage merge requests, to use tickets in order to stay focused on our own tasks and develop our programming skills