2009年10月30日 23:30

アドビのAcrobat用バイトサービングスクリプトが何をやっているか今ならわかる

スポンサードリンク

AcrobatでWebからページごとのデータを分割して取得できるよう、サーバがHTTP/1.1のRangeに対応していない場合代わりにUAからの要求範囲を受け取ってPerlでPDFの特定部位を切り出して返すリダイレクトCGIなんですね。

http://www.adobe.com/jp/products/acrobat4/library/scripts/bsunix.txt

（2013-05-25 10:26追記）

上記URLはリンク切れなので、Weyback Archiveのリンクと、転記をしておきます。それにしてもアドビの人は、後世に伝えることに関して無関心すぎる。

http://web.archive.org/web/20000615073408/http://www.adobe.com/support/techguides/acrobat/byteserve/bsunix.txt

#!/usr/bin/perl
#!/usr/local/bin/perl
# byteserver script
# byteserver.pl
# Last updated May 10, 1996
# This script implements the draft "Byte Range Retrieval Extension to HTTP" described at 
# "http://www.w3.org/Protocols/rfc2068/rfc2068". 
# place this script wherever you can run CGIs on your Web server.
# 1) Users should make sure that the script as placed in the cgi-bin
#directory is renamed to byteserver.pl, and has execute permission on 
#a UNIX machine. This is done  by typing: chmod 755 byteserver.pl
#2.) The first line of the byteserver.pl script contains a line of the form
#!<pathname>
#where <pathname> points to
#where perl is found on the system.
#Users should modify this line to suit their system-- on a UNIX machine
#you can find out where perl resides by typing: which perl
#if the above returns /usr/local/bin/perl
#then the first line of the byteserver.pl should read
#!/usr/local/bin/perl
#Note: if executing the command: which perl
#returns the message: perl command not found or something similar it
#means that perl is not installed on your system. You *need* perl to
#use byteserver.pl 
# query is of the form
# http://server/cgi-bin/byteserver.pl/path/name/here
#       with HTTP_RANGE specifying the ranges in the form
#       "bytes=m1-n1,m2-n2,...,mn-nn"
#       /path/name/here is relative to server root
#       m-n are ranges (0 based). If m is absent, means n bytes from EOF, if n
#       is absent, mean first m bytes of the file 
#       "bytes=..." syntax
#       STDOUT the ranges
#       The script issues HTTP1.0 206 Partial Content as a partial
#       content response.     
# Env. variables:
#       PATH_TRANSLATED is the file name
#       CONTENT_TYPE is what the server says is the mime type of file
#       REQUEST_METHOD is the request method i.e. GET, POST etc. (we only accept GET)
#       HTTP_RANGE the byte range request
# Return syntax:
# if no ;bytes= is specified, returns:
#       Accept-ranges: bytes
#       Content-type: <content-type>
#       Content-length: <file-length>
#       <the file itself>
# if ;bytes= specified, returns:
#       
#       HTTP/1.0 206 Partial Content
#       Accept-ranges: bytes
#       Content-type: multipart/x-byteranges; boundary=$boundary
#       < multipart-body >
   
sub badjob {
     
local($mesg) = @_;
     
print "Status: 403 \"BOGUS\"\r\n\r\n";
     
print "<HEAD><TITLE>403 BOGUS</TITLE></HEAD>"; print "<BODY BGCOLOR="#FFFFFF" LINK="#0033CC" VLINK="#800080" ALINK="#FF0000"><H1>403 BOGUS</H1>";
print "<pre>$mesg</pre>\r\n";
print "Bad Request\r\n\r\n";
exit 1;
}
     
     
$path = $ENV{'PATH_TRANSLATED'};
$ranges = $ENV{'HTTP_RANGE'};
     
($file, $query) = split(/;/,$path);
     
$contenttype = $ENV{'CONTENT_TYPE'};
if ($contenttype eq "") 
{ 
        $contenttype = "application/pdf";
}
     
if ((! -r $file) || (length $file <= 3)) 
{ 
        &badjob("file $file not found");
}
     
if ($ENV{'REQUEST_METHOD'} ne 'GET') 
{ # not URL-based query 
        &badjob("METHOD not GET");
}
     
$size = -s $file;
     
if (($query eq "") && ($ranges eq "")) 
{
        # respond that we can do ranges
        print "Accept-ranges: bytes\r\n";
        print "Content-type: $contenttype\r\n"; 
        printf "Content-Length: %d\r\n\r\n", $size; 
        open(FILE, "< $file");
        while(read(FILE, $buffer, 4096)) {
        print STDOUT $buffer;
}
exit 0;
     
} else {
    if ($ranges ne "")
    { 
        $query = $ranges;
        $query =~ s/bytes=//;
    }
    else
    {           
        $query =~ s/\s+//g; # squeeze all whitespace out 
        $query =~ s/bytes=//; 
    }
     
    #check that the ranges are properly formatted 
    @byterangeCheck = split(/[-,=]/,$query); 
    while (defined($fbyte = shift(@byterangeCheck))) {
        $lbyte = shift(byterangeCheck);
        if (($fbyte > 0) && ($lbyte < 0)) {
                &badjob("query range malformed");
        }
        if (($fbyte < 0) && ($lbyte > 0)) {
                &badjob("query range malformed"); 
        }
        if ($fbyte > $lbyte) {
                &badjob("query range malformed");
        }
     
    }
    @byterange = split(/[,]/,$query);
}
     
# print 206 only if called with official syntax   
if ( $ranges ne "" )
{
        print "Status: 206 Partial Content\r\n";
}
# print other header info
$boundary="multipart-boundary";
print "Accept-ranges: bytes\r\n";
print "Content-type: multipart/x-byteranges; boundary=$boundary\r\n\r\n";
     
# Serve up the bytes:
     
open(FILE, "< $file");
while (defined ($range = shift(@byterange))) {
     
    ($fbyte, $lbyte) = split('-', $range); 
    $i = index($range,"-");
    if ($i == 0) { $fbyte = -1; }
    if ($fbyte < 0) {
        $fbyte = $size - $lbyte;
        $lbyte = $size;
    }
    $nbytes = $lbyte - $fbyte + 1;
     
    print "\r\n--$boundary\r\n";
    print "Content-type: $contenttype\r\n";
    printf "Content-Range: bytes %d-%d/%d\r\n\r\n", $fbyte, $lbyte, $size;
     
    seek(FILE, $fbyte, 0);
    read(FILE, $buffer, $nbytes);
    print STDOUT $buffer;
}
     
print "\r\n--$boundary--\r\n";
     
exit 0;

2009.10.30 23:30 投稿　大野義貴 [Perl] | 固定リンク |