 |
Submit Your Code | |
 |
SEO Monitor | |
 |
UpTime Monitor | |
 |
PHP Code Contest | |
 |
My Favorite Examples | |
|
My Favorite Articles | |
 |
Update Your Profile | |
|
|
| More Weber Sites | ||
| PHP Code Search | ||
| Web Development Forums | ||
| Learn MySQL Playing Trivia | ||
| PHPBB2 Templates | ||
| Web Development Index | ||
| PHP Web Logs (BLogs) | ||
| Web Development Resources | ||
| Web Development Content | ||
| Recommended Links | ||
| PHPClasses | ||
| PHP Editor | ||
| PHP Jobs | ||
| Vision.To Design | ||
| Ajax Tutorials | ||
| PHP Programming Help | ||
| PHP/MySQL Programming | ||
| Webmaster Resources | ||
| Webmaster Forum | ||
| XML meta language | ||
| website builder | ||
| Recommended | ||
| Submit Site | ||
| Forex Trading Online forex trading platform | ||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Code | |||||
This function searches text or HTML files for URLs embedded in the text and returns them in an array. CAUTION: If used as a web spider to collect links on sites other than your own, use sleep()/usleep() to slow it down and avoid overloading other servers! Development was done under WindowsNT, but I can't see why it won't work under Unix. <?php # NOTES: # # Beware of setting the $depth parameter too high; this can cause lengthy # searches and return very large result arrays! (Once again, CAUTION: If used # as a web spider to collect links on sites other than your own, MODIFY THIS # CLASS to use sleep()/usleep() to slow it down and avoid overloading other # servers!) # # My main use for this function is to periodically scan web forum postings and # summarize any URL mentions buried deep in the message threads. On Unix, this # can be set up as a cron process; under WinNT, the command-line "AT" command # or WinAT can be used for scheduling. # # Please excuse the formatting; I collapsed all tabs to single spaces, so some # things aren't aligned nicely. # # If you have any suggestions/comments, email me: sbedberg@ucdavis.edu # # AND NOW, THE ACTUAL CODE: ?> <?php #------------------------------------------------------------------------------ # CLASS: # link_harvester # Scans text or HTML documents for URLs, and returns a list. #------------------------------------------------------------------------------ # Notes: # If the given filename is an actual file or a URL, it will scan that file; # if it is a directory, it will look for and scan ALL files in that # directory. # # If scanning a text file, will only recognize URLs with the "schema://..." # form; newlines, tabs, whitespace and some punctuation (',",<,>) terminate # it. Any links that are followed (depth > 0) will be followed as HTML, not # as text files. The doc title array will contain the the first line of the # file, up to 80 chars. Lastly, all links are assumed to be non-local (too # hard to determine otherwise). # # If scanning HTML, the routine only looks at HREFs. Anything without an # explicit scheme is - as is specified by the W3 standard - assumed to be # a reference to a local page. # # Public Methods: # harvest($filename) Wrapper for _harvest() # $filename This is the initial filename/URL to start scanning # # Private Methods: # _harvest($filename, $depth, $as_text) # Actual search routine # $filename As above. # $depth Depth of documents to search (see public vars, below) # $as_text Determines how to search $filename (see public vars) # # _is_schema($string) Returns TRUE if a valid schema is found at the head # of string (whitespace characters not permitted). # #------------------------------------------------------------------------------ # (version 0.9) written by S. Edberg (sbedberg@ucdavis.edu), March 1998. # # Change History: #------------------------------------------------------------------------------ # TODO? # BUGFIX: If (depth > 0 and local searching enabled), the base URL needs to # be prepended to the $url agrument to _harvest($url, $depth-1); current- # ly, it returns a fail-to-open error. # Have optional list of URLs to include or exclude # Duplicate URL removal option # Add timeout options (for entire script? for attempts to open URLs found?) # Allow wildcards in file specification # Optionally traverse subdirectories if a directory name is specified #------------------------------------------------------------------------------ class link_harvester { # Public variables: Input var $depth = 0; # Link depth to search (0= $filename only, 1=1 level deep, ...) var $as_text = 1; # If 1 (TRUE) scans docs as text; 0 (FALSE) - assumes HTML var $local_sw = 1; # If TRUE, collect local links; ignore local otherwise var $remote_sw = 1; # If TRUE, collect nonlocal link info; ignore otherwise # Public variables: Output var $attempts = 0; # No. of files attempted var $failures = 0; # No. of files unreachable var $lasterror = ''; # Last error message generated (excluding unreachable files) var $links; # Array of links harvested var $titles; # Array of titles ($title[x] is for $link[x]) var $source; # Array of source URLs of links($source[x] for $link[x]) # Private variables: var $MAXLINELEN = 2048; # Maximum length of input line. # Public methods function harvest($filename) { $ftype = filetype($filename); if ($this->_is_schema($filename) || $ftype == 'file') { $this->_harvest($filename, $this->depth, $this->as_text); } elseif ($ftype == 'dir') { if ($dh = opendir($filename)) { # First, add a trailing slash to directory name if there is none $dname = $filename.(ereg( '[\\/]$', $filename) ? "" : "/"); while ($fname = readdir($dh)) { if (filetype($dname.$fname) == 'file') { $this->_harvest($dname.$fname, $this->depth, $this->as_text); } } closedir($dh); } else { $this->lasterror = 'Unable to open directory: '.$filename; } } else { $this->lasterror = 'Unknown filetype for: '.$filename; } } # Private methods function _harvest($filename,$depth,$as_text) { $this->attempts++; if ($fp = @fopen($filename, "r")) { $in_url = 0; # state variable: are we currently in a URL? $in_link = 0; # state variable; are we in linked text? $first_time = 1; $getfunc = ($as_text ? "fgetss" : "fgets"); # ...strip HTML, PHP tags if reading as text file. $prev_line_fragment = ''; while($line = $getfunc($fp, $this->MAXLINELEN)) { $remaining_line = $prev_line_fragment.$line; $prev_line_fragment = ''; # ...prepend end of previous line in case there was a read-line # break in the middle of a tag if ($as_text && $first_time) { $title = substr($line,0,80); } while ($remaining_line != '') { if ($as_text) { $temp = split( '://', $remaining_line, 2); if (count($temp) > 1) { # '://' found; now extract schema $remaining_line = $temp[1]; $temp2 = split( '[^[:alpha:]]', strrev($temp[0]), 2); $url_pre = strrev($temp2[0]); # this gets schema if ($url_pre != '') { # this ensures a non-null schema $temp3 = split( "[[:space:]'\"<>]", $remaining_line, 2); $url_post = $temp3[0]; $remaining_line = $temp3[1]; $url = $url_pre. '://'.$url_post; $this->links[] = $url; $this->titles[] = "FROM: $title"; $this->source[] = $filename; if ($depth > 0) { $this->_harvest($url, $depth-1, 0); } # recurse? } } else { # '://' not found; go to next line $prev_line_fragment = $remaining_line; $remaining_line = ''; } } else { # as HTML if (!$in_url) { # look for start tag $temp = split( '<[[:space:]]*[Aa][[:space:]]*[Hh][Rr][Ee][Ff]'. '[[:space:]]*=[[:space:]]*"[[:space:]]*', $remaining_line, 2); if (count($temp) > 1) { # Found; $temp[1] contains rest of line, starting at schema $in_url = 1; $in_link = 0; # state varb: are we in linked text? (else, in URL) $url = ''; $title = ''; $remaining_line = $temp[1]; # start search over with new, truncated line. } else { # not found; go on to next line $prev_line_fragment = $remaining_line; $remaining_line = ''; } } elseif ($in_url && !$in_link) { # look for doublequote & '>' at end of HREF tag $temp = split( '[[:space:]]*"[[:space:]]*>[[:space:]]*', $remaining_line, 2); $url .= $temp[0]; if (count($temp) > 1) { # Found; $temp[1] contains rest of line after HREF="..." # Now, check if schema is found (determines local/nonlocal reference) $is_schema = $this->_is_schema($url); if (($is_schema && $this->remote_sw) || (!$is_schema && $this->local_sw)) { $in_link = 1; # OK...continue with accumulation. $remaining_line = $temp[1]; $title = ''; } else { # Ignore this; reset flags & continue. $in_link = 0; $in_url = 0; $remaining_line = $temp[1]; } } else { # end of URL not found on this line; go on to next $prev_line_fragment = $remaining_line; $remaining_line = ''; } } else { # $in_url AND $in_link: look for </A> tag after linked text $temp = split( '[[:space:]]*</[Aa]>[[:space:]]*', $remaining_line, 2); $title .= $temp[0]; if (count($temp) > 1) { # found end tag $this->links[] = $url; $this->titles[] = $title; $this->source[] = $filename; $in_url = 0; $in_link = 0; $remaining_line = $temp[1]; if ($depth > 0) { $this->_harvest($url, $depth-1, 0); } # recurse? } else { # end of title not found on this line; go on to next $prev_line_fragment = $remaining_line; $remaining_line = ''; } } } # end if ($as_text) } # end while ($remaining_line != '') $first_time = 0; } fclose($fp); if (!$as_text && $in_url) { $this->links[] = $url; $this->titles[] = ( $in_link ? $title : '' ); $this->source[] = $filename; } } else { $this->failures++; } } function _is_schema($string) { $x = eregi( '^[[:alpha:]]+://', $string); return($x); } }; ?> | |||||
| Related Code Examples | |||||
| User Comments | |||||
|
|||||
