Revision 4165
Added by ben leinfelder over 16 years ago
src/perl/update_emldoc_data.pl | ||
---|---|---|
1 |
#!/usr/bin/perl |
|
2 |
# |
|
3 |
# '$RCSfile: transfer_emldocs_from_metacat1_to_metacat2.pl,v $' |
|
4 |
# Copyright: 2000 Regents of the University of California |
|
5 |
# |
|
6 |
# '$Author: sgarg $' |
|
7 |
# '$Date: 2004/04/09 23:12:46 $' |
|
8 |
# '$Revision: 1.1 $' |
|
9 |
# |
|
10 |
# This program is free software; you can redistribute it and/or modify |
|
11 |
# it under the terms of the GNU General Public License as published by |
|
12 |
# the Free Software Foundation; either version 2 of the License, or |
|
13 |
# (at your option) any later version. |
|
14 |
# |
|
15 |
# This program is distributed in the hope that it will be useful, |
|
16 |
# but WITHOUT ANY WARRANTY; without even the implied warranty of |
|
17 |
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
|
18 |
# GNU General Public License for more details. |
|
19 |
# |
|
20 |
# You should have received a copy of the GNU General Public License |
|
21 |
# along with this program; if not, write to the Free Software |
|
22 |
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA |
|
23 |
# |
|
24 |
# |
|
25 |
# This is a CGI application for inserting metadata documents into |
|
26 |
# the Metacat database. It utilizes the Metacat.pm module for most work. |
|
27 |
# You can specify two metacats and a list of documents. This script will |
|
28 |
# copy documents from one metacat to another. |
|
29 |
|
|
30 |
|
|
31 |
use Metacat; |
|
32 |
|
|
33 |
my $metacat1; |
|
34 |
my $docid; |
|
35 |
my $rev; |
|
36 |
my $error = 0; |
|
37 |
my $xmldoc; |
|
38 |
my $xa; |
|
39 |
my $response; |
|
40 |
|
|
41 |
my $listname = "kruger_docids"; |
|
42 |
|
|
43 |
my $metacat1_url = "http://dataknp.sanparks.org/sanparks/metacat"; |
|
44 |
my $username = "uid=judithk,o=SANParks,dc=ecoinformatics,dc=org"; |
|
45 |
my $password = "xxxxxx"; |
|
46 |
#my $metacat1_url = "http://fred.msi.ucsb.edu/knb/metacat"; |
|
47 |
#my $username = "uid=leinfelder,o=NCEAS,dc=ecoinformatics,dc=org"; |
|
48 |
#my $password = "xxxxxx"; |
|
49 |
|
|
50 |
#used for inserting xml in the appropriate location |
|
51 |
my $updatedcount = 0; |
|
52 |
my $insertionindex; |
|
53 |
#my $insertionmarker = "</abstract>"; |
|
54 |
my $insertionmarker = "</keywordSet>"; |
|
55 |
my $xmlsnippet = ""; |
|
56 |
$xmlsnippet .= "<keywordSet>"; |
|
57 |
#$xmlsnippet .= "<keywordThesaurus>location</keywordThesaurus>"; |
|
58 |
$xmlsnippet .= "<keyword>SANParks, South Africa</keyword>"; |
|
59 |
$xmlsnippet .= "<keyword>Kruger National Park, South Africa</keyword>"; |
|
60 |
$xmlsnippet .= "</keywordSet>"; |
|
61 |
|
|
62 |
$metacat1 = Metacat->new(); |
|
63 |
|
|
64 |
if ($metacat1) { |
|
65 |
$metacat1->set_options( metacatUrl => $metacat1_url); |
|
66 |
} else { |
|
67 |
#die "failed during metacat creation\n"; |
|
68 |
print "Failed during metacat1 creation."; |
|
69 |
$error = 1; |
|
70 |
} |
|
71 |
|
|
72 |
# Login to metacat |
|
73 |
print "Connecting to metacat1..........\n"; |
|
74 |
my $response1 = $metacat1->login($username, $password); |
|
75 |
if (! $response1) { |
|
76 |
print $metacat1->getMessage(); |
|
77 |
print "Failed during login: metacat1.\n"; |
|
78 |
$error = 2; |
|
79 |
} else { |
|
80 |
print "Connected to metacat1\n"; |
|
81 |
} |
|
82 |
|
|
83 |
if($error == 0){ |
|
84 |
open(file,$listname) || die ("Couldn't open the file"); |
|
85 |
while(<file>) { |
|
86 |
chomp(); |
|
87 |
$docid = $_; |
|
88 |
|
|
89 |
print "Original docid: $docid\n"; |
|
90 |
|
|
91 |
#get the document |
|
92 |
$xmldoc = $metacat1->read($docid); |
|
93 |
$xa = $xmldoc->content; |
|
94 |
|
|
95 |
#construct the next docid revision |
|
96 |
#three parts to the docid: scope.id.rev |
|
97 |
@docidparts = split(/\./, $docid); |
|
98 |
$rev = @docidparts[2]; |
|
99 |
$rev = $rev + 1; |
|
100 |
print "next rev: $rev\n"; |
|
101 |
@docidparts[2] = $rev; |
|
102 |
$docid = join('.', @docidparts); |
|
103 |
|
|
104 |
print "Revised docid: $docid\n"; |
|
105 |
|
|
106 |
#check for the inserted data (maybe we did it already) |
|
107 |
$insertionindex = index($xa, $xmlsnippet); |
|
108 |
if ($insertionindex > 0) { |
|
109 |
print "Already added the xml snippet to this doc\n"; |
|
110 |
next; |
|
111 |
} |
|
112 |
|
|
113 |
#insert the snippet |
|
114 |
$insertionindex = index($xa, $insertionmarker); |
|
115 |
if ($insertionindex > 0) { |
|
116 |
#we want it after the end of the marker |
|
117 |
$insertionindex += length($insertionmarker); |
|
118 |
|
|
119 |
#construct the new xml with inserted elements |
|
120 |
$xa = substr($xa, 0, $insertionindex) . $xmlsnippet . substr($xa, $insertionindex); |
|
121 |
|
|
122 |
print "\n----------\n"; |
|
123 |
#print $xa; |
|
124 |
#print "\n----------\n"; |
|
125 |
|
|
126 |
#send the updated document |
|
127 |
$response = $metacat1->update($docid, $xa); |
|
128 |
print $metacat1->getMessage(); |
|
129 |
|
|
130 |
$updatedcount++; |
|
131 |
} |
|
132 |
} |
|
133 |
print "Updated $updatedcount documents\n"; |
|
134 |
} else { |
|
135 |
print $error; |
|
136 |
} |
|
0 | 137 |
Also available in: Unified diff
script used to update keywords for existing kruger documents when upgrade to new SANParks/SAEON skin. committing before additional work to update the package ids for these files (incorrectly set to the wrong rev number)