[21411] | 1 | ###########################################################################
|
---|
| 2 | #
|
---|
| 3 | # dbutil::mssql -- utility functions for writing to mssql databases
|
---|
| 4 | #
|
---|
| 5 | # A component of the Greenstone digital library software
|
---|
| 6 | # from the New Zealand Digital Library Project at the
|
---|
| 7 | # University of Waikato, New Zealand.
|
---|
| 8 | #
|
---|
| 9 | # Copyright (C) 2009
|
---|
| 10 | #
|
---|
| 11 | # This program is free software; you can redistribute it and/or modify
|
---|
| 12 | # it under the terms of the GNU General Public License as published by
|
---|
| 13 | # the Free Software Foundation; either version 2 of the License, or
|
---|
| 14 | # (at your option) any later version.
|
---|
| 15 | #
|
---|
| 16 | # This program is distributed in the hope that it will be useful,
|
---|
| 17 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
| 18 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
| 19 | # GNU General Public License for more details.
|
---|
| 20 | #
|
---|
| 21 | # You should have received a copy of the GNU General Public License
|
---|
| 22 | # along with this program; if not, write to the Free Software
|
---|
| 23 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
| 24 | #
|
---|
| 25 | ###########################################################################
|
---|
| 26 |
|
---|
| 27 | package dbutil::mssql;
|
---|
| 28 |
|
---|
| 29 | use Encode;
|
---|
| 30 | use strict;
|
---|
| 31 |
|
---|
| 32 |
|
---|
| 33 |
|
---|
| 34 | # -----------------------------------------------------------------------------
|
---|
| 35 | # MSSQL IMPLEMENTATION
|
---|
| 36 | # -----------------------------------------------------------------------------
|
---|
| 37 |
|
---|
| 38 | my $mssql_collection_name = "";
|
---|
| 39 | my $mssql_data_table_name = "";
|
---|
| 40 | my $mssql_document_metadata_table_name = "";
|
---|
| 41 |
|
---|
| 42 |
|
---|
| 43 | sub open_infodb_write_handle
|
---|
| 44 | {
|
---|
| 45 | my $infodb_file_path = shift(@_);
|
---|
| 46 |
|
---|
| 47 | # You might have to install the DBD::ADO module from CPAN
|
---|
| 48 | #================================================================#
|
---|
| 49 | # Uncomment this if you want to use MSSQL!!!
|
---|
| 50 | # By the way, MSSQL only works on a Windows machine...
|
---|
| 51 | #================================================================#
|
---|
| 52 | #use DBI;
|
---|
| 53 | #use DBD::ADO;
|
---|
| 54 | #Win32::OLE->Option(CP => Win32::OLE::CP_UTF8);
|
---|
| 55 |
|
---|
| 56 | # The hard coded server connection thingy which should be placed
|
---|
| 57 | # in some configuration file.
|
---|
| 58 | # If you have problem connecting to your MS SQL server:
|
---|
| 59 | # 1. Check if your MSSQL server has been started.
|
---|
| 60 | # 2. Check if the TCP/IP connection has been enabled.
|
---|
| 61 | # 3. Use telnet to the server
|
---|
| 62 | # (don't forget to specify the port, which can be found in the configuration manager)
|
---|
| 63 | # If none of the above helped, the you need to start googling then.
|
---|
| 64 | my $host = "localhost,1660"; # Need to look up your SQL server and see what port is it using.
|
---|
| 65 | my $user = "sa";
|
---|
| 66 | my $pwd = "[When installing the MSSQL, you will be asked to input a password for the sa user, use that password]";
|
---|
| 67 | my $database = "[Create a database in MSSQL and use it here]";
|
---|
| 68 |
|
---|
| 69 | # Create the unique name for the table
|
---|
| 70 | # We do not want to change the database for the current running index
|
---|
| 71 | # Therefore we use timestamp and collection short name to create an unqiue name
|
---|
| 72 | my $cur_time = time();
|
---|
| 73 | my $unique_key = $mssql_collection_name . "_" . $cur_time;
|
---|
| 74 | $mssql_data_table_name = "data_" . $unique_key;
|
---|
| 75 | $mssql_document_metadata_table_name = "document_metadata_" . $unique_key;
|
---|
| 76 | print STDERR "MSSQL: Creating unique table name. Unique ID:[" . $unique_key . "]\n";
|
---|
| 77 |
|
---|
| 78 | # Store these information into the infodbfile
|
---|
| 79 | open(FH, ">" . $infodb_file_path);
|
---|
| 80 | print FH "mss-host\t" . $host . "\n";
|
---|
| 81 | print FH "username\t" . $user . "\n";
|
---|
| 82 | print FH "password\t" . $pwd . "\n";
|
---|
| 83 | print FH "database\t" . $database . "\n";
|
---|
| 84 | print FH "tableid\t" . $unique_key . "\n";
|
---|
| 85 | close(FH);
|
---|
| 86 | print STDERR "MSSQL: Saving db info into :[" . $infodb_file_path . "]\n";
|
---|
| 87 |
|
---|
| 88 | # Make the connection
|
---|
| 89 | my $dsn = "Provider=SQLNCLI;Server=$host;Database=$database";
|
---|
| 90 | my $infodb_handle = DBI->connect("dbi:ADO:$dsn", $user, $pwd, { RaiseError => 1, AutoCommit => 1}) || return undef;
|
---|
| 91 | print STDERR "MSSQL: Connect to MS SQL database. DSN:[" . $dsn . "]\n";
|
---|
| 92 |
|
---|
| 93 | # Make sure the data table has been created.
|
---|
| 94 | my $data_table_checker_array = dbgetarray($infodb_handle, "SELECT name FROM sysobjects WHERE name = '" . $mssql_data_table_name . "' AND OBJECTPROPERTY(id,'IsUserTable') = 1");
|
---|
| 95 | if (scalar(@{$data_table_checker_array}) == 0)
|
---|
| 96 | {
|
---|
| 97 | dbquery($infodb_handle, "CREATE TABLE " . $mssql_data_table_name . " (one_key NVARCHAR(50) UNIQUE, one_value NVARCHAR(MAX))");
|
---|
| 98 | }
|
---|
| 99 | print STDERR "MSSQL: Making sure the data table(" . $mssql_data_table_name . ") exists\n";
|
---|
| 100 |
|
---|
| 101 | # Make sure the document_metadata table has been created.
|
---|
| 102 | my $document_metadata_table_checker_array = dbgetarray($infodb_handle, "SELECT name FROM sysobjects WHERE name = '" . $mssql_document_metadata_table_name . "' AND OBJECTPROPERTY(id,'IsUserTable') = 1");
|
---|
| 103 | if (scalar(@{$document_metadata_table_checker_array}) == 0)
|
---|
| 104 | {
|
---|
| 105 | dbquery($infodb_handle, "CREATE TABLE " . $mssql_document_metadata_table_name . " (id INTEGER IDENTITY(1,1) PRIMARY KEY, docOID NVARCHAR(50), element NVARCHAR(MAX), value NVARCHAR(MAX))");
|
---|
| 106 | dbquery($infodb_handle, "CREATE INDEX dmd ON " . $mssql_document_metadata_table_name . "(docOID)");
|
---|
| 107 | }
|
---|
| 108 | print STDERR "MSSQL: Making sure the document_metadata table(" . $mssql_data_table_name . ") exists.\n";
|
---|
| 109 |
|
---|
| 110 | return $infodb_handle;
|
---|
| 111 | }
|
---|
| 112 |
|
---|
| 113 |
|
---|
| 114 | sub close_infodb_write_handle
|
---|
| 115 | {
|
---|
| 116 | my $infodb_handle = shift(@_);
|
---|
| 117 |
|
---|
| 118 | $infodb_handle->disconnect();
|
---|
| 119 | }
|
---|
| 120 |
|
---|
| 121 |
|
---|
| 122 | sub get_infodb_file_path
|
---|
| 123 | {
|
---|
| 124 | my $collection_name = shift(@_);
|
---|
| 125 | my $infodb_directory_path = shift(@_);
|
---|
| 126 |
|
---|
| 127 | my $infodb_file_extension = ".mssqldbinfo";
|
---|
| 128 | my $infodb_file_name = &util::get_dirsep_tail($collection_name) . $infodb_file_extension;
|
---|
| 129 |
|
---|
| 130 | # This will be used in the open_infodb_write_handle function
|
---|
| 131 | $mssql_collection_name = $collection_name;
|
---|
| 132 |
|
---|
| 133 | return &util::filename_cat($infodb_directory_path, $infodb_file_name);
|
---|
| 134 | }
|
---|
| 135 |
|
---|
| 136 |
|
---|
| 137 | sub read_infodb_file
|
---|
| 138 | {
|
---|
| 139 | my $infodb_file_path = shift(@_);
|
---|
| 140 | my $infodb_map = shift(@_);
|
---|
| 141 |
|
---|
| 142 | # !! TO IMPLEMENT
|
---|
| 143 | }
|
---|
| 144 |
|
---|
| 145 |
|
---|
| 146 | sub write_infodb_entry
|
---|
| 147 | {
|
---|
| 148 | my $infodb_handle = shift(@_);
|
---|
| 149 | my $infodb_key = shift(@_);
|
---|
| 150 | my $infodb_map = shift(@_);
|
---|
| 151 |
|
---|
| 152 | # Add the key -> value mapping into the "data" table
|
---|
| 153 | my $infodb_entry_value = "";
|
---|
| 154 | foreach my $infodb_value_key (keys(%$infodb_map))
|
---|
| 155 | {
|
---|
| 156 | foreach my $infodb_value (@{$infodb_map->{$infodb_value_key}})
|
---|
| 157 | {
|
---|
| 158 | $infodb_entry_value .= "<$infodb_value_key>" . &Encode::decode_utf8($infodb_value) . "\n";
|
---|
| 159 | }
|
---|
| 160 | }
|
---|
| 161 |
|
---|
| 162 | # Prepare the query
|
---|
| 163 | my $safe_infodb_key = &mssql_safe($infodb_key);
|
---|
| 164 | my $query = "INSERT INTO " . $mssql_data_table_name . " (one_key, one_value) VALUES (N'" . $safe_infodb_key . "', N'" . &mssql_safe($infodb_entry_value) . "')";
|
---|
| 165 | dbquery($infodb_handle, $query);
|
---|
| 166 |
|
---|
| 167 | # If this infodb entry is for a document, add all the interesting document metadata to the
|
---|
| 168 | # "document_metadata" table (for use by the dynamic classifiers)
|
---|
| 169 | if ($infodb_key !~ /\./ && $infodb_entry_value =~ /\<doctype\>doc\n/)
|
---|
| 170 | {
|
---|
| 171 | dbquery($infodb_handle, "DELETE FROM " . $mssql_document_metadata_table_name . " WHERE docOID=N'" . $safe_infodb_key . "'");
|
---|
| 172 |
|
---|
| 173 | foreach my $infodb_value_key (keys(%$infodb_map))
|
---|
| 174 | {
|
---|
| 175 | # We're not interested in most of the automatically added document metadata
|
---|
| 176 | next if ($infodb_value_key eq "archivedir" ||
|
---|
| 177 | $infodb_value_key eq "assocfilepath" ||
|
---|
| 178 | $infodb_value_key eq "childtype" ||
|
---|
| 179 | $infodb_value_key eq "contains" ||
|
---|
| 180 | $infodb_value_key eq "docnum" ||
|
---|
| 181 | $infodb_value_key eq "doctype" ||
|
---|
| 182 | $infodb_value_key eq "Encoding" ||
|
---|
| 183 | $infodb_value_key eq "FileSize" ||
|
---|
| 184 | $infodb_value_key eq "hascover" ||
|
---|
| 185 | $infodb_value_key eq "hastxt" ||
|
---|
| 186 | $infodb_value_key eq "lastmodified" ||
|
---|
| 187 | $infodb_value_key eq "metadataset" ||
|
---|
| 188 | $infodb_value_key eq "thistype" ||
|
---|
| 189 | $infodb_value_key =~ /^metadatafreq\-/ ||
|
---|
| 190 | $infodb_value_key =~ /^metadatalist\-/);
|
---|
| 191 | foreach my $infodb_value (@{$infodb_map->{$infodb_value_key}})
|
---|
| 192 | {
|
---|
| 193 | $infodb_handle->{LongReadLen} = 65535; # Added for the encoding issue
|
---|
| 194 | my $query = "INSERT INTO " . $mssql_document_metadata_table_name . " (docOID, element, value) VALUES (N'" . $safe_infodb_key . "', N'" . &mssql_safe(&Encode::decode_utf8($infodb_value_key)) . "', N'" . &mssql_safe(&Encode::decode_utf8($infodb_value)) . "')";
|
---|
| 195 | dbquery($infodb_handle, $query);
|
---|
| 196 | }
|
---|
| 197 | }
|
---|
| 198 | }
|
---|
| 199 | }
|
---|
| 200 |
|
---|
| 201 | sub delete_infodb_entry
|
---|
| 202 | {
|
---|
| 203 | my $infodb_handle = shift(@_);
|
---|
| 204 | my $infodb_key = shift(@_);
|
---|
| 205 |
|
---|
| 206 | # Delete the key from the "data" table
|
---|
| 207 |
|
---|
| 208 |
|
---|
| 209 | # Prepare the query
|
---|
| 210 | my $safe_infodb_key = &mssql_safe($infodb_key);
|
---|
| 211 | my $query = "DELETE FROM " . $mssql_data_table_name . " WHERE one_key=N'" . $safe_infodb_key . "'";
|
---|
| 212 | dbquery($infodb_handle, $query);
|
---|
| 213 |
|
---|
| 214 | # If this infodb entry is for a document, add all the interesting document metadata to the
|
---|
| 215 | # "document_metadata" table (for use by the dynamic classifiers)
|
---|
| 216 | if ($infodb_key !~ /\./)
|
---|
| 217 | {
|
---|
| 218 | # Possible for there not to be a docOID matching this infodb_key
|
---|
| 219 | # (entries are only made when <doctype> == doc
|
---|
| 220 | # Attempt to delete it, and don't complain if one isn't found
|
---|
| 221 |
|
---|
| 222 | dbquery($infodb_handle, "DELETE FROM " . $mssql_document_metadata_table_name . " WHERE docOID=N'" . $safe_infodb_key . "'");
|
---|
| 223 |
|
---|
| 224 | }
|
---|
| 225 | }
|
---|
| 226 |
|
---|
| 227 |
|
---|
| 228 |
|
---|
| 229 | sub mssql_safe
|
---|
| 230 | {
|
---|
| 231 | my $value = shift(@_);
|
---|
| 232 |
|
---|
| 233 | # Escape any single quotes in the value
|
---|
| 234 | $value =~ s/\'/\'\'/g;
|
---|
| 235 |
|
---|
| 236 | return $value;
|
---|
| 237 | }
|
---|
| 238 |
|
---|
| 239 |
|
---|
| 240 | sub dbquery
|
---|
| 241 | {
|
---|
| 242 | my $infodb_handle = shift(@_);
|
---|
| 243 | my $sql_query = shift(@_);
|
---|
| 244 |
|
---|
| 245 | # Execute the SQL statement
|
---|
| 246 | my $statement_handle = $infodb_handle->prepare($sql_query);
|
---|
| 247 | $statement_handle->execute();
|
---|
| 248 | if ($statement_handle->err)
|
---|
| 249 | {
|
---|
| 250 | print STDERR "Error:" . $statement_handle->errstr . "\n";
|
---|
| 251 | return undef;
|
---|
| 252 | }
|
---|
| 253 |
|
---|
| 254 | return $statement_handle;
|
---|
| 255 | }
|
---|
| 256 |
|
---|
| 257 |
|
---|
| 258 | sub dbgetarray
|
---|
| 259 | {
|
---|
| 260 | my $infodb_handle = shift(@_);
|
---|
| 261 | my $sql_query = shift(@_);
|
---|
| 262 |
|
---|
| 263 | my $statement_handle = dbquery($infodb_handle, $sql_query);
|
---|
| 264 | my $return_array = [];
|
---|
| 265 |
|
---|
| 266 | # Iterate through the results and push them into an array
|
---|
| 267 | if (!defined($statement_handle))
|
---|
| 268 | {
|
---|
| 269 | return [];
|
---|
| 270 | }
|
---|
| 271 |
|
---|
| 272 | while ((my $temp_hash = $statement_handle->fetchrow_hashref()))
|
---|
| 273 | {
|
---|
| 274 | push(@$return_array, $temp_hash);
|
---|
| 275 | }
|
---|
| 276 |
|
---|
| 277 | return $return_array;
|
---|
| 278 | }
|
---|
| 279 |
|
---|
| 280 |
|
---|
| 281 | 1;
|
---|