Web scraping

From Rosetta Code
Task
Web scraping
You are encouraged to solve this task according to the task description, using any language you may know.

Create a program that downloads the time from this URL: http://tycho.usno.navy.mil/cgi-bin/timer.pl and then prints the current UTC time by extracting just the UTC time from the web page's HTML.

If possible, only use libraries that come at no extra monetary cost with the programming language and that are widely available and popular such as CPAN for Perl or Boost for C++.

AutoHotkey

<lang AutoHotkey> UrlDownloadToFile, http://tycho.usno.navy.mil/cgi-bin/timer.pl, time.html FileRead, timefile, time.html pos := InStr(timefile, "UTC") msgbox % time := SubStr(timefile, pos - 9, 8) </lang>

AWK

This is inspired by GETURL example in the manual for gawk. 

#! /usr/bin/awk -f

BEGIN {
  purl = "/inet/tcp/0/tycho.usno.navy.mil/80"
  ORS = RS = "\r\n\r\n"
  print "GET /cgi-bin/timer.pl HTTP/1.0" |& purl
  purl |& getline header
  while ( (purl |& getline ) > 0 )
  {
     split($0, a, "\n")
     for(i=1; i <= length(a); i++)
     {
        if ( a[i] ~ /UTC/ )
        {
          sub(/^<BR>/, "", a[i])
          printf "%s\n", a[i]
        }
     }
  }
  close(purl)
}

C

Works with: POSIX version .1-2001
Library: libcurl

There's no any proper error handling.

<lang c>#include <stdio.h>

  1. include <string.h>
  2. include <curl/curl.h>
  3. include <sys/types.h>
  4. include <regex.h>
  1. define BUFSIZE 16384

size_t lr = 0;

size_t filterit(void *ptr, size_t size, size_t nmemb, void *stream) { 

 if ( (lr + size*nmemb) > BUFSIZE ) return BUFSIZE;
 memcpy(stream+lr, ptr, size*nmemb);
 lr += size*nmemb;
 return size*nmemb;

}

int main() { 

 CURL *curlHandle;
 char buffer[BUFSIZE];
 regmatch_t amatch;
 regex_t cregex;

 curlHandle = curl_easy_init();
 curl_easy_setopt(curlHandle, CURLOPT_URL, "http://tycho.usno.navy.mil/cgi-bin/timer.pl");
 curl_easy_setopt(curlHandle, CURLOPT_FOLLOWLOCATION, 1);
 curl_easy_setopt(curlHandle, CURLOPT_WRITEFUNCTION, filterit);
 curl_easy_setopt(curlHandle, CURLOPT_WRITEDATA, buffer);
 int success = curl_easy_perform(curlHandle);
 curl_easy_cleanup(curlHandle);

 buffer[lr] = 0;
 
 regcomp(&cregex, " UTC", REG_NEWLINE);
 regexec(&cregex, buffer, 1, &amatch, 0);
 int bi = amatch.rm_so;
 while ( bi-- > 0 )
   if ( memcmp(&buffer[bi], "
", 4) == 0 ) break;

 buffer[amatch.rm_eo] = 0;

 printf("%s\n", &buffer[bi+4]);

 regfree(&cregex);
 return 0;

}</lang>

Common Lisp

Library: cl-ppcre
Library: drakma

<lang lisp>BOA> (let* ((url "http://tycho.usno.navy.mil/cgi-bin/timer.pl") 

           (regexp (load-time-value
                    (cl-ppcre:create-scanner "(?m)^.{4}(.+? UTC)")))
           (data (drakma:http-request url)))
      (multiple-value-bind (start end start-regs end-regs)
          (cl-ppcre:scan regexp data)
        (declare (ignore end))
        (when start
          (subseq data (aref start-regs 0) (aref end-regs 0)))))

"Aug. 12, 04:29:51 UTC"</lang>

E

<lang e>interp.waitAtTop(when (def html := <http://tycho.usno.navy.mil/cgi-bin/timer.pl>.getText()) -> { 

   def rx`(?s).*>(@time.*? UTC).*` := html
   println(time)

})</lang>

Erlang

Using regular expressions: <lang erlang>-module(scraping). -export([main/0]). -define(Url, "http://tycho.usno.navy.mil/cgi-bin/timer.pl"). -define(Match, "
(.+ UTC)").

main() -> inets:start(), {ok, {_Status, _Header, HTML}} = http:request(?Url), {match, [Time]} = re:run(HTML, ?Match, [{capture, all_but_first, binary}]), io:format("~s~n",[Time]). </lang>

Forth

Works with: GNU Forth version 0.7.0

<lang forth> include unix/socket.fs

extract-time ( addr len type len -- time len )
 dup >r
 search 0= abort" that time not present!"
 dup >r
 begin -1 /string  over 1- c@ [char] > = until       \ seek back to 
 at start of line
 r> - r> + ;

s" tycho.usno.navy.mil" 80 open-socket dup s\" GET /cgi-bin/timer.pl HTTP/1.0\n\n" rot write-socket dup pad 4096 read-socket s\" \r\n\r\n" search 0= abort" can't find headers!" \ skip headers s" UTC" extract-time type cr close-socket </lang>

Java

<lang java>import java.io.BufferedReader; import java.io.IOException; import java.io.InputStreamReader; import java.net.URL; import java.net.URLConnection;


public class WebTime{ public static void main(String[] args){ try{ URL address = new URL( "http://tycho.usno.navy.mil/cgi-bin/timer.pl"); URLConnection conn = address.openConnection(); BufferedReader in = new BufferedReader( new InputStreamReader(conn.getInputStream())); String line; while(!(line = in.readLine()).contains("UTC")); System.out.println(line.substring(4)); }catch(IOException e){ System.err.println("error connecting to server."); e.printStackTrace(); } } } </lang>

OCaml

<lang ocaml>let () = 

 let _,_, page_content = make_request ~url:Sys.argv.(1) ~kind:GET () in

 let lines = Str.split (Str.regexp "\n") page_content in
 let str =
   List.find
     (fun line ->
       try ignore(Str.search_forward (Str.regexp "UTC") line 0); true
       with Not_found -> false)
     lines
 in
 let str = Str.global_replace (Str.regexp "
") "" str in
 print_endline str;
</lang>

There are libraries for this, but it's rather interesting to see how to use a socket to achieve this, so see the implementation of the above function make_request on this page.

Perl

<lang perl>use LWP::Simple;

my $url = 'http://tycho.usno.navy.mil/cgi-bin/timer.pl'; get($url) =~ /
(.+? UTC)/ 

   and print "$1\n";</lang>

Python

<lang python>import urllib

page = urllib.urlopen('http://tycho.usno.navy.mil/cgi-bin/timer.pl') for line in page: 

   if ' UTC' in line:
       print line.strip()[4:]
       break

page.close()</lang> Sample output: 

Aug. 12, 15:22:08 UTC           Universal Time

R

Library: RCurl
Library: XML

First, retrieve the web page. See HTTP_Request for more options with this. <lang R> library(RCurl) webpage <- getURL("http://tycho.usno.navy.mil/cgi-bin/timer.pl") </lang> Now parse the html code into a tree and retrieve the interesting bit <lang R> library(XML) pagetree <- htmlTreeParse(webpage ) timesnode <- pagetree$children$html$children$body$children$h3$children$pre$children timesnode <- timesnode[names(timesnode)=="text"] </lang> Finally, find the line with universal time and parse it <lang R> timestrings <- sapply(timesnode, function(x) x$value) index <- grep("Universal Time", timestrings) utctimestr <- strsplit(timestrings[index], "\t")$text[1] utctime <- strptime(utctimestr, "%b. %d, %H:%M:%S UTC")

  1. Print the date in any format you desire.

strftime(utctime, "%A, %d %B %Y, %H:%M:%S") </lang> 

Monday, 03 August 2009, 16:15:37

Ruby

<lang ruby>require "open-uri"

open('http://tycho.usno.navy.mil/cgi-bin/timer.pl') do |p| 

 p.each_line do |line|
   if line =~ / UTC/
     puts line[4..-1]
     break
   end
 end

end</lang>

Tcl

<lang tcl>package require http

set request [http::geturl "http://tycho.usno.navy.mil/cgi-bin/timer.pl"] if {[regexp -line {
(.* UTC)} [http::data $request] --> utc]} { 

   puts $utc

}</lang>

UNIX Shell

Works with: Bourne Again SHell

This solution uses curl, which can be downloaded for free (and very easily on GNU/Linux systems), and popular (at list in the GNU and *n*x world) utilities programs like grep and sed. 

#! /bin/bash
curl -s http://tycho.usno.navy.mil/cgi-bin/timer.pl |
   grep ' UTC' |
   sed -e 's/^<BR>//'

Ursala

This works by launching the wget command in a separate process and capturing its output. The program is compiled to an executable command. <lang Ursala>#import std

  1. import cli
  1. executable ('parameterized',)

whatime =

<.file$[contents: --<>]>+ -+ 

  @hm skip/*4+ ~=(9%cOi&)-~l*+ *~ ~&K3/'UTC',
  (ask bash)/0+ -[wget -O - http://tycho.usno.navy.mil/cgi-bin/timer.pl]-!+-</lang>

Here is a bash session. 

$ whatime
Jun. 26, 20:49:52 UTC

Visual Basic .NET

New, .NET way with StringReader: <lang vbnet>Imports System.Net Imports System.IO Dim client As WebClient = New WebClient() 

       Dim content As String = client.DownloadString("http://tycho.usno.navy.mil/cgi-bin/timer.pl")
       Dim sr As New StringReader(content)
       While sr.peek <> -1
           Dim s As String = sr.ReadLine
           If s.Contains("UTC") Then
               Console.WriteLine(s)
           End If
       End While

</lang>

Alternative, old fashioned way using VB "Split" function: <lang vbnet>Imports System.Net 

       Dim client As WebClient = New WebClient()
       Dim content As String = client.DownloadString("http://tycho.usno.navy.mil/cgi-bin/timer.pl")
       Dim lines() As String = Split(content, vbLf) 'may need vbCrLf 
       For Each line In lines
           If line.Contains("UTC") Then
               Console.WriteLine(line)
           End If
       Next

</lang>

Retrieved from "https://rosettacode.org/wiki/Web_scraping?oldid=79720"