In this project three different, collaborative applications were implemented:
- a shell application that creates a set of websites on the disk.
- a web server that accepts HTTP requests for specific webpages from the websites.
- a web crawler, whose goal is to download all websites from the web server and serve search commands, as well.
These applications are similar (in a simplified form) to those used by popular search engines like Google.
The project encapsulates the whole Distributed-File-Management-Tool project.
It was developed for the class of "System Programing" in the Informatics Department, using C++ (without STL) and Object Oriented Programming practices.
It is a bash script that generates a set of websites, in a random way, using an input text file.
To run the script use the command:
$ ./webcreator.sh <root_directory> <text_file> w p
The webcreator.sh application works as follows:
It creates w directories, in the root_directory, that each represents a website. These directories will be in the form of site0, site1, ..., sitew-1. Then it creates p pages in each directory, that represent the webpages of each website. The websites will be in the form of pagei_j.html where i is the website and j is a random, unique for each website, number.
For example, if w = 3 and p = 2 the following directories and pages could be created:
root_dir/site0/
root_dir/site0/page0_2345.html
root_dir/site0/page0_2391.html
root_dir/site1/
root_dir/site1/page1_2345.html
root_dir/site1/page1_2401.html
root_dir/site2/
root_dir/site2/page2_6530.html
root_dir/site2/page2_16672.html
Each webpage will have the following format:
<!DOCTYPE html>
<html>
<body>
word1 word2 ... <a href=”link1”>link1_text</a>
wordn wordn+1 ... <a href=”link2”>link2_text</a> wordm wordm+1 ...
</body>
</html>
- root_directory: is a directory that must already exist and in which the websites will be generated. If the root_directory is not empty the script will purge it.
- text_file: is a text file, from which random words will be selected to create the webpages. One such file is J. Verne's "Twenty Thousand Leagues under the Sea", which is included. It must contain at least 10000 lines.
- w: the number of the websites that will be created.
- p: the number of the webpages, per website, that will be created.
Constraints: w >= 2, p >= 3
To run the application use the command:
$ ./myhttpd -p <serving_port> -c <command_port> -t <num_of_threads> -d <root_dir>
In the service port the server simply accepts HTTP/ 1.1 requests. In this project only a simple GET request is implemented. The server is listening to the specific port for requests of the following format:
GET /site0/page0_1244.html HTTP/1.1
User-Agent: Mozilla/4.0 (compatible; MSIE5.01; Windows NT)
Host: www.tutorialspoint.com
Accept-Language: en-us
Accept-Encoding: gzip, deflate
Connection: Keep-Alive
[blank line here]
When it receives such a request, the server assigns it to one of the threads from the thread pool. The thread's job is to return the root_dir/site0/page0_1244.html file back.
The operation of the server with the threads is as follows:
The server accepts a connection to the socket and places the corresponding file descriptor in a buffer from which the threads read. Each thread is responsible for a file descriptor, from which it will
read the GET request and return the answer, i.e. the webpage's content (or a different answer, in case the file does not exist, or the thread does not have the appropriate permissions, as depicted below).
-
If the requested file exists, and the server has the permission to read it, then a response of the following format is returned:
HTTP/1.1 200 OK Date: Mon, 27 May 2018 12:28:53 GMT Server: myhttpd/1.0.0 (Ubuntu64) Content-Length: 8873 Content-Type: text/html Connection: Closed [blank line] [content of the requested file here... e.g. <html>hello one two ...</html>]The length is in bytes, expressing only the size of the content, i.e. without including he header.
-
If the requested file does not exist, then the server return a response of the following format:
HTTP/1.1 404 Not Found Date: Mon, 27 May 2018 12:28:53 GMT Server: myhttpd/1.0.0 (Ubuntu64) Content-Length: 124 Content-Type: text/html Connection: Closed [blank line] <html>Sorry dude, couldn’t find this file.</html> -
If the server does not the have the appropriate permissions, then a response of the following format is returned:
HTTP/1.1 403 Forbidden Date: Mon, 27 May 2018 12:28:53 GMT Server: myhttpd/1.0.0 (Ubuntu64) Content-Length: 124 Content-Type: text/html Connection: Closed [blank line] <html>Trying to access this file but don’t think I can make it.</html>
- -p (serving_port): the port that the server is listening in order to return webpages.
- -c (command_port): the port that the server is listening in order to receive commands.
- -t ( num_of_threads): the number of threads that the server will create in order to manage incoming requests. These threads are all created together, at the beginning, in a thread pool so that the server can reuse them. In the event that a thread is terminated, the server creates a new one.
- -d (root_dir): the directory that contains all the websites created by the webcreator.sh.
Furthermore, at the command port, the server listens and accepts the following simple commands (1 word each), which are executed directly by the server, without having to be assigned to a thread:
- STATS: the server responds with specific statistics: the duration it runs, how many pages it has returned and the total number of bytes, as well. E.g.:
Server up for 02:03.45, served 124 pages, 345539 bytes - SHUTDOWN: the server stops serving additional requests, releases any memory (shared or not) has allocated and stops its execution.
The crawler's job is to download all the websites' webpages from the webserver, analyzing them and finding links within the pages that will follow retrospectively.
To run the application use the command:
$ ./mycrawler -h <host_or_IP> -p <port> -c <command_port> -t <num_of_threads> -d <save_dir> <starting_URL>
The crawler works as follows:
In the beginning, it creates a thread pool with the corresponding threads. The threads are being reused.
It also creates a queue in order to store the links that have been found so far and appends the starting_URL to this queue.
One of the threads takes the URL from the queue and requests it from the server. The URL is in the format of (for example) http://linux01.di.uoa.gr:8080/site1/page0_1234.html.
After downloading the file, it saves it to the corresponding directory/file in the save_dir.
It analyzes the file, that has just been downloaded, and finds additional links that appends to the queue.
It repeats this process with all the threads in parallel, until there are no other links in the queue.
- -h (host_or_IP): the name of the machine or the IP that the server is running.
- -p (port): the port on which the server is listening.
- -c (command_port): the port on which the crawler is listening in order to receive commands.
- -t (num_of_threads): the number of threads that the crawler creates and runs. In the event that a thread terminates, the crawler creates a new one.
- -d (save_dir): the directory in which the crawler will save the downloaded pages.
- starting_URL: the URL from which the crawler starts its operation.
Furthermore, at the command port the crawler listens and accepts the following simple commands (1 word each), which are executed directly by the crawler, without having to be assigned to a thread:
- STATS: the crawler responds with specific statistics: the duration it runs, how many pages it has downloaded and the total number of bytes, as well. E.g.:
Crawler up for 01:06.45, downloaded 124 pages, 345539 bytes - SEARCH e.g.
SEARCH word1 word2 word3 ... word10: If the crawler still has pages in the queue, then it returns a message indicating that the crawling is still in progress.
If the crawler has finished downloading all the webpages, then this command will utilize the Distributed-File-Management-Tool project's search command and will return the same results, but over the socket. - SHUTDOWN: the crawler stops downloading webpages, releases any memory (shared or not) has allocated and stops its execution.
A makefile is provided. You can use the following commands:
$ make: for easy compilation (both the web server and crawler will be compiled).$ make myhttpd: to only compile the web-server application.$ make mycrawler: to only compile the web-crawler application.$ make clean1: to delete all *.o auto generated files from the web-server application.$ make clean2: to delete all *.o auto generated files from the web-crawler application.$ make cleanallto delete all *.o auto generated files from both applications.$ make cleanfifos: to delete all named pipes.
For execution details and appropriate command line arguments see each application's chapter.