akcgjc007/makDa

A multithreaded web crawler in Cpp using libCurl. The single-threaded version of a crawler is leveraged through using multiple threads and processing requests in different threads. The crawler is coded in C++.

Forked from Web-Crawler. Visit the above link for detailed information about the project.

makDa - A multithreaded web crawler with libcURL & thread-safe data structures

Added features

• Dowloading files with libcURL: One of the bugs in previous projects was crawler stopping at the end of its execution. In somecases https requests would not get any response and they would wait forever(very long time). Because of the use of OpenSSL, it was not possible to implement a time out functionality. By using libcURL, we can add the time out feature. Hurray! now crawler won't stop.
• Saving and resume working state: Once crawler runs adn finishes its execution, it was dumping the data into output files. I have added the saving feature, now the state of the crawler gets saved in a .zip file. And at the time of initialization, it recovers its state.

How to run makDa

Dependencies

• sudo apt install libcurl4-openssl-dev

Commands

• use make to compile and run makDa, it would run with its default arguments.
• supported arguments are:
  • maxlinks : maximum links extracted from a single website
  • pagelimit : maximum pages to process
  • threads : maximum concurrent threads to create
  • maxfilesize : maximum file size to download
  • timeout : maximum time waiting for website response
  • restore_data : restoring crawler state from previous run
  • save_data : saving current crawler state for next run