1 | ###########################################################################
|
---|
2 | #
|
---|
3 | # GreenstoneXMLPlugout.pm -- the plugout module for Greenstone Archives
|
---|
4 | # A component of the Greenstone digital library software
|
---|
5 | # from the New Zealand Digital Library Project at the
|
---|
6 | # University of Waikato, New Zealand.
|
---|
7 | #
|
---|
8 | # Copyright (C) 2006 New Zealand Digital Library Project
|
---|
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., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
23 | #
|
---|
24 | ###########################################################################
|
---|
25 |
|
---|
26 | package MySQLPlugout;
|
---|
27 |
|
---|
28 | use strict;
|
---|
29 | no strict 'refs';
|
---|
30 | no strict 'subs';
|
---|
31 |
|
---|
32 | use util;
|
---|
33 | use FileUtils;
|
---|
34 | #use BasePlugout;
|
---|
35 | use GreenstoneXMLPlugout;
|
---|
36 | use docprint;
|
---|
37 |
|
---|
38 | use IPC::Open2;
|
---|
39 | use POSIX ":sys_wait_h"; # for waitpid, http://perldoc.perl.org/functions/waitpid.html
|
---|
40 |
|
---|
41 | # this plugout does not output xml to a file, but outputs rows into a mysql table
|
---|
42 | sub BEGIN {
|
---|
43 | @MySQLPlugout::ISA = ('GreenstoneXMLPlugout');
|
---|
44 | }
|
---|
45 |
|
---|
46 |
|
---|
47 | # TODO: deal with -removeold and everything? Or type out instructions for user
|
---|
48 |
|
---|
49 | # TODO Q: what is "group" in GreenstoneXMLPlugout?
|
---|
50 | # TODO Q: site_name only exists for GS3. What about GS2?
|
---|
51 |
|
---|
52 | my $process_mode_list =
|
---|
53 | [ { 'name' => "meta_only",
|
---|
54 | 'desc' => "{MySQLPlugout.process_mode.meta_only}" },
|
---|
55 | { 'name' => "text_only",
|
---|
56 | 'desc' => "{MySQLPlugout.process_mode.text_only}" },
|
---|
57 | { 'name' => "all",
|
---|
58 | 'desc' => "{MySQLPlugout.process_mode.all}" } ];
|
---|
59 |
|
---|
60 | my $arguments = [
|
---|
61 | { 'name' => "process_mode",
|
---|
62 | 'desc' => "{MySQLPlugout.process_mode}",
|
---|
63 | 'type' => "enum",
|
---|
64 | 'list' => $process_mode_list,
|
---|
65 | 'deft' => "all",
|
---|
66 | 'reqd' => "no",
|
---|
67 | 'hiddengli' => "no"} ];
|
---|
68 |
|
---|
69 | my $options = { 'name' => "MySQLPlugout",
|
---|
70 | 'desc' => "{MySQLPlugout.desc}",
|
---|
71 | 'abstract' => "no",
|
---|
72 | 'inherits' => "yes",
|
---|
73 | 'args' => $arguments };
|
---|
74 |
|
---|
75 | sub new {
|
---|
76 | my ($class) = shift (@_);
|
---|
77 | my ($plugoutlist, $inputargs,$hashArgOptLists) = @_;
|
---|
78 | push(@$plugoutlist, $class);
|
---|
79 |
|
---|
80 | push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
|
---|
81 | push(@{$hashArgOptLists->{"OptList"}},$options);
|
---|
82 |
|
---|
83 | my $self = new GreenstoneXMLPlugout($plugoutlist,$inputargs,$hashArgOptLists);
|
---|
84 |
|
---|
85 | if ($self->{'info_only'}) {
|
---|
86 | # don't worry about any options etc
|
---|
87 | return bless $self, $class;
|
---|
88 | }
|
---|
89 | print STDERR "***** MySQLPlugout process mode = \"", $self->{'process_mode'}, "\"\n";
|
---|
90 |
|
---|
91 | return bless $self, $class;
|
---|
92 | }
|
---|
93 |
|
---|
94 | # connect here and ensure all tables and databases exist
|
---|
95 | sub begin {
|
---|
96 |
|
---|
97 | my $self= shift (@_);
|
---|
98 |
|
---|
99 | ########### TODO: these should be set from cmdline/GLI options to plugout #########
|
---|
100 | $self->{'db_driver'} = "mysql";
|
---|
101 | $self->{'site_name'} = "localsite";
|
---|
102 | $self->{'client_user'} = "root";
|
---|
103 | $self->{'client_pwd'} = "6reenstone3";
|
---|
104 | #$self->{'db_host'} = "127.0.0.1";
|
---|
105 | #$self->{'db_encoding'} = "utf8";
|
---|
106 |
|
---|
107 | ############ LOAD NECESSARY OPTIONS ###########
|
---|
108 | print STDERR "########## COLLECTION: ". $ENV{'GSDLCOLLECTION'}."\n";
|
---|
109 | $self->{'collection_name'} = $ENV{'GSDLCOLLECTION'};
|
---|
110 |
|
---|
111 | if(!$self->connect_to_db()) {
|
---|
112 | # This is fatal for the plugout, let's terminate here
|
---|
113 | # PrintError would already have displayed the warning message on connection fail
|
---|
114 | die("Could not connect to db. Can't proceed.\n");
|
---|
115 | }
|
---|
116 | if(!$self->load_db_and_tables()) {
|
---|
117 | # This is fatal for the plugout, let's terminate here
|
---|
118 | # PrintError would already have displayed the warning message on connection fail
|
---|
119 | die("Could not use db or prepare its tables. Can't proceed.\n");
|
---|
120 | }
|
---|
121 |
|
---|
122 | # prepare the shared/common HANDLES to SQL insert statements that contain placeholders
|
---|
123 | # and which we will reuse repeatedly when executing the actual insert statements
|
---|
124 | my $proc_mode = $self->{'process_mode'};
|
---|
125 | if($proc_mode eq "all" || $proc_mode eq "meta_only" ) {
|
---|
126 | $self->{'metadata_prepared_insert_statement_handle'} = $self->prepare_insert_metadata_row_stmthandle();
|
---|
127 | }
|
---|
128 | if($proc_mode eq "all" || $proc_mode eq "text_only" ) {
|
---|
129 | $self->{'fulltxt_prepared_insert_statement_handle'} = $self->prepare_insert_fulltxt_row_stmthandle();
|
---|
130 | }
|
---|
131 |
|
---|
132 | # finally, call begin on super
|
---|
133 | $self->GreenstoneXMLPlugout::begin(@_);
|
---|
134 | }
|
---|
135 |
|
---|
136 | # disconnect from database here, see inexport.pm
|
---|
137 | sub end
|
---|
138 | {
|
---|
139 | my $self = shift(@_);
|
---|
140 |
|
---|
141 | # do the superclass stuff
|
---|
142 | $self->GreenstoneXMLPlugout::end(@_);
|
---|
143 | $self->disconnect_from_db() || warn("Unable to disconnect from database " . $self->{'site_name'} . "\n");
|
---|
144 | #my $success = $self->GreenstoneXMLPlugout::end(@_);
|
---|
145 | #$success = $self->disconnect_from_db() && $success;
|
---|
146 | #return $success;
|
---|
147 | }
|
---|
148 |
|
---|
149 | # TODO: check arc-inf.db for whether each entry is to be deleted/indexed/reindexed/been indexed
|
---|
150 | sub saveas {
|
---|
151 | my $self = shift (@_);
|
---|
152 | my ($doc_obj, $doc_dir) = @_;
|
---|
153 |
|
---|
154 | # pre save out
|
---|
155 | my ($docxml_outhandler, $output_file) = $self->GreenstoneXMLPlugout::pre_saveas(@_);
|
---|
156 |
|
---|
157 |
|
---|
158 | # saving customisation
|
---|
159 | $self->{'debug_outhandle'} = $docxml_outhandler if ($self->{'debug'}); # STDOUT if debug
|
---|
160 |
|
---|
161 | # write the INVERSE into doc.xml as to what is written to the db
|
---|
162 | my $proc_mode = $self->{'process_mode'};
|
---|
163 | my $docxml_output_options = { 'output' => docprint::OUTPUT_NONE };
|
---|
164 | if($proc_mode eq "meta_only" ) { # since only meta to go into MySQL db, text will go into docxml
|
---|
165 | $docxml_output_options->{'output'} = docprint::OUTPUT_TEXT_ONLY;
|
---|
166 | } elsif($proc_mode eq "text_only" ) { # since only full text to go into MySQL db, meta will go into docxml
|
---|
167 | $docxml_output_options->{'output'} = docprint::OUTPUT_META_ONLY;
|
---|
168 | }
|
---|
169 |
|
---|
170 | # now we've prepared to write out whatever is meant to go into docxml
|
---|
171 | # and can do actual the steps superclass GreenstoneXMLPlugout carries out to write out docxml
|
---|
172 | # So: write out the doc xml file for the current document
|
---|
173 | my $section_text = &docprint::get_section_xml($doc_obj, $docxml_output_options);
|
---|
174 | print $docxml_outhandler $section_text;
|
---|
175 |
|
---|
176 | # We also write out whatever needs to go into the MySQL database
|
---|
177 | $self->write_meta_and_text($doc_obj);
|
---|
178 |
|
---|
179 |
|
---|
180 | # post save out
|
---|
181 | $self->GreenstoneXMLPlugout::post_saveas(@_);
|
---|
182 |
|
---|
183 | # TODO: close database connection here? Or do something like with groups
|
---|
184 | # so we don't open and close over and over during a single build?
|
---|
185 | }
|
---|
186 |
|
---|
187 |
|
---|
188 | # write meta and/or text PER DOC out to DB
|
---|
189 | sub write_meta_and_text {
|
---|
190 | my $self = shift (@_);
|
---|
191 | my ($doc_obj) = @_;
|
---|
192 | my $root_section = $doc_obj->get_top_section();
|
---|
193 | my $doc_oid = $doc_obj->get_OID(); # we're processing a single doc at a time, so single OID
|
---|
194 |
|
---|
195 | ##binmode($db_handle,":utf8"); ## WRONG FOR DB, NEED TO CREATE IN UTF8 MODE
|
---|
196 |
|
---|
197 | # TODO if $self->debug is on
|
---|
198 |
|
---|
199 | # Do we want to open and close a connection per doc?
|
---|
200 | # Would we not rather want to open and close per collection rebuild?
|
---|
201 |
|
---|
202 | #$self->create_db_connection();
|
---|
203 |
|
---|
204 | # load the prepared INSERT statement handles for both tables (can be undef for any table depending on whether meta_only or txt_only are set)
|
---|
205 | my $metadata_table_sth = $self->{'metadata_prepared_insert_statement_handle'};
|
---|
206 | my $fulltxt_table_sth = $self->{'fulltxt_prepared_insert_statement_handle'};
|
---|
207 |
|
---|
208 | $self->recursive_write_meta_and_text($doc_obj, $root_section, $metadata_table_sth, $fulltxt_table_sth);
|
---|
209 |
|
---|
210 | #$self->close_db_connection();
|
---|
211 | }
|
---|
212 | # Perl: Reading or Writing to Another Program
|
---|
213 | # https://nnc3.com/mags/Perl3/cookbook/ch16_05.htm
|
---|
214 | sub recursive_write_meta_and_text {
|
---|
215 | my $self = shift (@_);
|
---|
216 | my ($doc_obj, $doc_oid, $section, $metadata_table_sth, $fulltxt_table_sth) = @_;
|
---|
217 |
|
---|
218 | # If section=ROOT, write "root" as section name into table
|
---|
219 | # doc->get_top_section() is the name of the doc root section, which is ""
|
---|
220 | my $section_name = ($section eq "") ? "root" : $section;
|
---|
221 |
|
---|
222 | my $section_ptr = $doc_obj->_lookup_section ($section);
|
---|
223 | return "" unless defined $section_ptr;
|
---|
224 |
|
---|
225 | my $debug_out = $self->{'debug_outhandle'};
|
---|
226 |
|
---|
227 | #my $proc_mode = $self->{'process_mode'};
|
---|
228 | #if($proc_mode eq "all" || $proc_mode eq "meta_only" ) {
|
---|
229 | if($metadata_table_sth) { # meta insert statement handle will be undef if not writing meta
|
---|
230 |
|
---|
231 | foreach my $data (@{$section_ptr->{'metadata'}}) {
|
---|
232 | my $meta_name = $data->[0];
|
---|
233 | my $escaped_meta_value = &escape_text($data->[1]);
|
---|
234 |
|
---|
235 | # Write out the current section's meta to collection db's METADATA table
|
---|
236 |
|
---|
237 | # for each set of values to write to meta table, execute the prepared statement, filling in the values
|
---|
238 |
|
---|
239 | if($self->{'debug'}) {
|
---|
240 | # just print the statement we were going to execute
|
---|
241 |
|
---|
242 | print $debug_out $metadata_table_sth->Statement . "($doc_oid, $section_name, $meta_name, $escaped_meta_value)\n";
|
---|
243 | }
|
---|
244 | else {
|
---|
245 |
|
---|
246 | $metadata_table_sth->execute($doc_oid, $section_name, $meta_name, $escaped_meta_value);
|
---|
247 | #|| warn ("Unable to write metadata row to db:\n\tOID $doc_oid, section $section_name,\n\tmeta name: $meta_name, val: $escaped_meta_value");
|
---|
248 | # Execution failure will print out info anyway: since db connection sets PrintError
|
---|
249 | }
|
---|
250 | }
|
---|
251 | }
|
---|
252 |
|
---|
253 | #if($proc_mode eq "all" || $proc_mode eq "text_only" ) {
|
---|
254 | if($fulltxt_table_sth) { # fulltxt insert statement handle will be undef if not writing fulltxt
|
---|
255 |
|
---|
256 | if($self->{'debug'}) {
|
---|
257 | # just print the statement we were going to execute, minus the fulltxt value
|
---|
258 | my $txt_repr = $section_ptr->{'text'} ? "<TXT>" : "NULL";
|
---|
259 | print $debug_out $fulltxt_table_sth->Statement . "($doc_oid, $section_name, $txt_repr)\n";
|
---|
260 | } else {
|
---|
261 | my $section_text = &escape_text($section_ptr->{'text'});
|
---|
262 |
|
---|
263 | # fulltxt column can be SQL NULL. undef value gets written out as NULL:
|
---|
264 | # https://stackoverflow.com/questions/12708633/which-one-represents-null-undef-or-empty-string
|
---|
265 |
|
---|
266 | # Write out the current section's text to collection db's FULLTeXT table
|
---|
267 | $fulltxt_table_sth->execute($doc_oid, $section_name, $section_text);
|
---|
268 | #|| warn ("Unable to write fulltxt row to db for row:\n\tOID $doc_oid, section $section_name");
|
---|
269 | # Execution failure will print out info anyway: since db connection sets PrintError
|
---|
270 | }
|
---|
271 | }
|
---|
272 |
|
---|
273 | # output all subsections: RECURSIVE CALL
|
---|
274 | foreach my $subsection (@{$section_ptr->{'subsection_order'}}) {
|
---|
275 | &recursive_write_meta_and_text($doc_obj, $doc_oid, "$section.$subsection", $metadata_table_sth, $fulltxt_table_sth);
|
---|
276 | }
|
---|
277 | }
|
---|
278 |
|
---|
279 | #################################
|
---|
280 |
|
---|
281 | # Database access related functions
|
---|
282 | # http://g2pc1.bu.edu/~qzpeng/manual/MySQL%20Commands.htm
|
---|
283 | # https://www.guru99.com/insert-into.html
|
---|
284 |
|
---|
285 | # TODO Q: What on cancelling a build: delete table? But what if it was a rebuild and the rebuild is cancelled (not the original build)?
|
---|
286 | # Do we create a copy of the orig database as backup, then start populating current db, and if cancelled, delete current db and RENAME backup table to current?
|
---|
287 | # https://stackoverflow.com/questions/3280006/duplicating-a-mysql-table-indexes-and-data
|
---|
288 | # BUT what if the table is HUGE? (Think of a collection with millions of docs.) Huge overhead in copying?
|
---|
289 | # The alternative is we just quit on cancel, but then: cancel could leave the table in a partial committed state, with no way of rolling back.
|
---|
290 | # Unless they do a full rebuild, which will recreate the table from scratch?
|
---|
291 | # SOLUTION-> rollback transaction on error, see https://www.effectiveperlprogramming.com/2010/07/set-custom-dbi-error-handlers/
|
---|
292 |
|
---|
293 | # I'm using perl's open2 like in Z3950Download, as opposed to open3 like in WgetDownload
|
---|
294 | # since I'm assuming each insert statement is atomic: it either does the insertion or fails
|
---|
295 | # and we (may) get some message back. That means we should hopefully be able to terminate
|
---|
296 | # as well if we get SIGTERM/SIGKILL: we're not blocking, but are responsive after every INSERT stmt.
|
---|
297 |
|
---|
298 | # Just runs the command without displaying it
|
---|
299 | sub run_command_basic
|
---|
300 | {
|
---|
301 | my ($self,$strCMD) = @_;
|
---|
302 |
|
---|
303 |
|
---|
304 | my $process_instr = $self->{'MYSQL_IN'};
|
---|
305 |
|
---|
306 | print $process_instr "$strCMD\n";
|
---|
307 | }
|
---|
308 |
|
---|
309 | sub run_command
|
---|
310 | {
|
---|
311 | my ($self,$strCMD) = @_;
|
---|
312 |
|
---|
313 | print STDERR "Running mysql command: $strCMD\n";
|
---|
314 | $self->run_command_basic($strCMD);
|
---|
315 |
|
---|
316 | }
|
---|
317 |
|
---|
318 | sub response_line_contains {
|
---|
319 | my ($self,$expected) = @_;
|
---|
320 |
|
---|
321 | my $out = $self->{'MYSQL_OUT'};
|
---|
322 | my $opening_line = <$out>;
|
---|
323 | if ($opening_line =~ m/$expected/i) {
|
---|
324 | return 1;
|
---|
325 | }
|
---|
326 | return 0;
|
---|
327 | }
|
---|
328 |
|
---|
329 | sub response_lines_contain {
|
---|
330 | my ($self,$expected) = @_;
|
---|
331 |
|
---|
332 | my $found_expected = 0;
|
---|
333 | my $out = $self->{'MYSQL_OUT'};
|
---|
334 | while (my $line = <$out>) {
|
---|
335 | print STDERR "$line\n";
|
---|
336 | if($line =~ m/$expected/){
|
---|
337 | #return 1;
|
---|
338 | $found_expected = 1; #won't break out of loop: loop will consume all on child out at present
|
---|
339 | }
|
---|
340 | }
|
---|
341 | return $found_expected;
|
---|
342 | }
|
---|
343 |
|
---|
344 | # based on Z3950Download.pm::start_yaz()
|
---|
345 | # ./PATH/TO/mysql -u root -p
|
---|
346 | # Returns:
|
---|
347 | # - 1 if client already quit after pwd fail (so can't send quit message to stopped client)
|
---|
348 | # - 0 if load_db failed (needs to be quit),
|
---|
349 | # - 1 if load_db succeeded (mysql client still running)
|
---|
350 | sub create_db_connection {
|
---|
351 | my $self = shift (@_);
|
---|
352 | print STDERR "Opening connection to MySQL db\n";
|
---|
353 |
|
---|
354 | my $mysql_client = $self->{'client_path'};
|
---|
355 | my $client_user = $self->{'client_user'} || "root";
|
---|
356 |
|
---|
357 |
|
---|
358 | my $launch_cmd = "\"./$mysql_client\" -u $client_user -p";
|
---|
359 | my $childpid = open2(*MYSQL_OUT, *MYSQL_IN, $launch_cmd)
|
---|
360 | or (print STDERR "Done\n" and die "can't open2 pipe to mysql client: $!");
|
---|
361 |
|
---|
362 | $self->{'pid'} = $childpid;
|
---|
363 | $self->{'MYSQL_OUT'} = *MYSQL_OUT;
|
---|
364 | $self->{'MYSQL_IN'} = *MYSQL_IN;
|
---|
365 |
|
---|
366 | # connect with pwd and load the database for this site
|
---|
367 | #my $conn_open = $self->open_connection();
|
---|
368 | #
|
---|
369 | #if (!$conn_open) {
|
---|
370 | # print STDERR "Cannot connect to mysql db with $launch_cmd\n";
|
---|
371 | # print STDERR "Done\n";
|
---|
372 | # return 0;
|
---|
373 | #}
|
---|
374 |
|
---|
375 | # connect with pwd
|
---|
376 | my $conn_success = $self->send_pwd();
|
---|
377 | if(!$conn_success) {
|
---|
378 | return -1; # if pwd failed, then the program already exited by itself
|
---|
379 | # (so don't send quit command after process terminated)
|
---|
380 | }
|
---|
381 | else {
|
---|
382 | return $self->load_db();
|
---|
383 | }
|
---|
384 |
|
---|
385 | # return $conn_open;
|
---|
386 | return $conn_success; # 1 if client already quit after pwd fail, 0 if load_db failed (needs to be quit), 1 if load_db succeeded (mysql client still running)
|
---|
387 | }
|
---|
388 |
|
---|
389 |
|
---|
390 | # Copied from Z3950Download.pm::quit_yaz()
|
---|
391 | sub close_db_connection {
|
---|
392 | my $self = shift (@_);
|
---|
393 |
|
---|
394 | $self->run_command("quit");
|
---|
395 | close($self->{'MYSQL_IN'}); # close the input to yaz. It also flushes quit command to mysql client
|
---|
396 |
|
---|
397 | # make sure nothing is being output by mysql client
|
---|
398 | # flush the mysql-client process' outputstream, else we'll be stuck in an infinite
|
---|
399 | # loop waiting for the process to quit.
|
---|
400 | my $output = $self->{'MYSQL_OUT'};
|
---|
401 | my $line;
|
---|
402 | while (defined ($line = <$output>)) {
|
---|
403 | if($line !~ m/\w/s) { # print anything other than plain whitespace in case it is important
|
---|
404 | print STDERR "***### $line";
|
---|
405 | }
|
---|
406 | }
|
---|
407 |
|
---|
408 | close($self->{'MYSQL_OUT'});
|
---|
409 |
|
---|
410 | # Is the following necessary? The PerlDoc on open2 (http://perldoc.perl.org/IPC/Open2.html)
|
---|
411 | # says that waitpid must be called to "reap the child process", or otherwise it will hang
|
---|
412 | # around like a zombie process in the background. Adding it here makes the code work as
|
---|
413 | # before, but it is certainly necessary to call waitpid on wget (see WgetDownload.pm).
|
---|
414 | # http://perldoc.perl.org/functions/waitpid.html
|
---|
415 | my $kidpid;
|
---|
416 | do {
|
---|
417 | $kidpid = waitpid($self->{'pid'}, WNOHANG);
|
---|
418 | } while $kidpid > 0; # waiting for pid to become -1
|
---|
419 | }
|
---|
420 |
|
---|
421 | # should be called only once per site
|
---|
422 | sub create_database {
|
---|
423 | my $self = shift (@_);
|
---|
424 | #my $sitename = shift(@_);
|
---|
425 | my $sitename = $self->{'site_name'};
|
---|
426 | my $cmd = "CREATE DATABASE $sitename;";
|
---|
427 | $self->run_command($cmd);
|
---|
428 | }
|
---|
429 |
|
---|
430 |
|
---|
431 | sub send_pwd {
|
---|
432 | my $self = shift (@_);
|
---|
433 |
|
---|
434 | my $client_pwd = $self->{'client_pwd'};
|
---|
435 | my $out = $self->{'MYSQL_OUT'};
|
---|
436 |
|
---|
437 | # if connected, it's prompting for pwd. Write the pwd to the mysql client process:
|
---|
438 | $self->run_command_basic($client_pwd);
|
---|
439 |
|
---|
440 | #my $opening_line = <$out>;
|
---|
441 | #if ($opening_line =~ m/Access denied/i) {
|
---|
442 | #print STDERR "Password not recognised. Got: $opening_line\n";
|
---|
443 | #return 0;
|
---|
444 | #}
|
---|
445 |
|
---|
446 | if($self->response_line_contains("Access denied")) {
|
---|
447 | print STDERR "Password not recognised. Got: Access denied.\n";
|
---|
448 | return 0;
|
---|
449 | }
|
---|
450 |
|
---|
451 | return 1;
|
---|
452 | }
|
---|
453 |
|
---|
454 |
|
---|
455 | sub load_db {
|
---|
456 | my $self = shift (@_);
|
---|
457 |
|
---|
458 | # attempt to load the db
|
---|
459 | # use the database
|
---|
460 | my $db_name = $self->{'site_name'}; # TODO Q: site_name only exists for GS3. What about GS2?
|
---|
461 | $self->run_command("use " . $db_name . ";");
|
---|
462 |
|
---|
463 | my $db_found = 0;
|
---|
464 | my $out = $self->{'MYSQL_OUT'};
|
---|
465 | while (my $line = <$out>) {
|
---|
466 | print STDERR "$line\n";
|
---|
467 | if($line =~ m/Database changed/){
|
---|
468 | # return 1; # TODO Q: consume all output of running command
|
---|
469 | $db_found = 1;
|
---|
470 | }
|
---|
471 | elsif($line =~ m/Unknown database/){
|
---|
472 | $db_found = -1;
|
---|
473 | }
|
---|
474 | }
|
---|
475 |
|
---|
476 | if($db_found == 1) {
|
---|
477 | return $db_found;
|
---|
478 | }
|
---|
479 | elsif ($db_found == -1) { # a db for the current sitename didn't exist, create it
|
---|
480 | $self->create_database();
|
---|
481 |
|
---|
482 | # attempt to load the newly created db
|
---|
483 | if($self->_load_db()) { # recursive call!
|
---|
484 | return 1;
|
---|
485 | }
|
---|
486 |
|
---|
487 |
|
---|
488 | #my $opening_line = <$out>;
|
---|
489 | #if ($opening_line !~ m/Query OK/) {
|
---|
490 | if(!$self->response_line_contains("Query OK")) {
|
---|
491 | print STDERR "Could not create db\n";
|
---|
492 | return 0; # couldn't even create the db
|
---|
493 | } else { # success creating db
|
---|
494 | # so let's create the metadata and fulltxt tables for the current coll while we're at it
|
---|
495 | if($self->create_meta_table()) {
|
---|
496 | return $self->create_fulltxt_table();
|
---|
497 | }
|
---|
498 | }
|
---|
499 | }
|
---|
500 | else { # unknown error trying to load db, bail
|
---|
501 | return 0;
|
---|
502 | }
|
---|
503 | }
|
---|
504 |
|
---|
505 |
|
---|
506 | sub create_meta_table {
|
---|
507 | my $self = shift (@_);
|
---|
508 | my $table_name = $self->{'collection_name'} . "metadata";
|
---|
509 | #my $cmd = "CREATE TABLE $table_name (id VARCHAR(255) NOT NULL UNIQUE, did VARCHAR(63) NOT NULL, sid VARCHAR(63) NOT NULL, metaname VARCHAR(127) NOT NULL, metavalue VARCHAR(1023) NOT NULL, PRIMARY KEY(id));";
|
---|
510 |
|
---|
511 | # If using an auto incremented primary key:
|
---|
512 | my $cmd = "CREATE TABLE $table_name (id INT NOT NULL AUTO_INCREMENT, did VARCHAR(63) NOT NULL, sid VARCHAR(63) NOT NULL, metaname VARCHAR(127) NOT NULL, metavalue VARCHAR(1023) NOT NULL, PRIMARY KEY(id));";
|
---|
513 |
|
---|
514 | $self->run_command($cmd);
|
---|
515 |
|
---|
516 | if(!$self->response_lines_contain("Query OK")) {
|
---|
517 | print STDERR "Could not create metadata table\n";
|
---|
518 | return 0;
|
---|
519 | } else {
|
---|
520 | return 1;
|
---|
521 | }
|
---|
522 | }
|
---|
523 |
|
---|
524 | sub create_fulltxt_table {
|
---|
525 | my $self = shift (@_);
|
---|
526 | my $table_name = $self->{'collection_name'} . "fulltxt";
|
---|
527 | #my $cmd = "CREATE TABLE $table_name (id VARCHAR(255) NOT NULL UNIQUE, did VARCHAR(63) NOT NULL, sid VARCHAR(63) NOT NULL, fulltxt LONGTEXT, PRIMARY KEY(id));";
|
---|
528 |
|
---|
529 | # If using an auto incremented primary key:
|
---|
530 | my $cmd = "CREATE TABLE $table_name (id INT NOT NULL AUTO_INCREMENT, did VARCHAR(63) NOT NULL, sid VARCHAR(63) NOT NULL, fulltxt LONGTEXT, PRIMARY KEY(id));";
|
---|
531 |
|
---|
532 | $self->run_command($cmd);
|
---|
533 |
|
---|
534 | if(!$self->response_lines_contain("Query OK")) {
|
---|
535 | print STDERR "Could not create table\n";
|
---|
536 | return 0;
|
---|
537 | } else {
|
---|
538 | return 1;
|
---|
539 | }
|
---|
540 | }
|
---|
541 |
|
---|
542 | # https://www.guru99.com/insert-into.html
|
---|
543 | # and https://dev.mysql.com/doc/refman/8.0/en/example-auto-increment.html
|
---|
544 | # for inserting multiple rows at once
|
---|
545 | sub get_cmd_insert_new_row_into_meta_table {
|
---|
546 | my $self = shift (@_);
|
---|
547 | my ($did, $sid, $metaname, $metavalue) = @_;
|
---|
548 | my $tablename = $self->{'colname'}."_metadata";
|
---|
549 |
|
---|
550 | my $cmd = "INSERT INTO $tablename (did, sid, metaname, metavalue) VALUES\n";
|
---|
551 | $cmd .= "('$did', '$sid', '$metaname', '$metavalue');\n";
|
---|
552 | return $cmd;
|
---|
553 | }
|
---|
554 |
|
---|
555 | sub get_cmd_insert_new_row_into_txt_table {
|
---|
556 | my $self = shift (@_);
|
---|
557 | my ($did, $sid, $fulltext) = @_;
|
---|
558 | my $tablename = $self->{'colname'}."_fulltxt";
|
---|
559 |
|
---|
560 | my $cmd = "INSERT INTO $tablename (did, sid, fulltxt) VALUES\n";
|
---|
561 | $cmd .= "('$did', '$sid', '$fulltext');\n";
|
---|
562 | return $cmd;
|
---|
563 | }
|
---|
564 |
|
---|
565 | # TODO: later add edit and delete (and nothing for "been indexed" status?)
|
---|
566 |
|
---|
567 | ##################
|
---|
568 |
|
---|
569 | ## UNUSED
|
---|
570 | sub X_open_connection()
|
---|
571 | {
|
---|
572 | my $self = shift (@_);
|
---|
573 | # connect with pwd
|
---|
574 | my $conn_success = $self->send_pwd();
|
---|
575 | if(!$conn_success) {
|
---|
576 | return -1; # if pwd failed, then the program already exists by itself (so don't send quit command after process terminated)
|
---|
577 | }
|
---|
578 |
|
---|
579 | if($conn_success) {
|
---|
580 | return $self->load_db();
|
---|
581 | } else {
|
---|
582 | return $conn_success;
|
---|
583 | }
|
---|
584 | }
|
---|
585 |
|
---|
586 | #################
|
---|
587 | # Database functions that use the perl DBI module (with the DBD driver module for mysql)
|
---|
588 | #################
|
---|
589 |
|
---|
590 | # THE NEW DB FUNCTIONS
|
---|
591 | # NOTE: FULLTEXT is a reserved keyword in (My)SQL. So we can't name a table or any of its columns "fulltext".
|
---|
592 | # https://dev.mysql.com/doc/refman/5.5/en/keywords.html
|
---|
593 |
|
---|
594 | # TODO: Consider AutoCommit status (and Autocommit off allowing commit or rollback for GS coll build cancel) later
|
---|
595 |
|
---|
596 | sub connect_to_db {
|
---|
597 | my $self= shift (@_);
|
---|
598 |
|
---|
599 | my $db_driver = $self->{'db_driver'};
|
---|
600 | my $db_user = $self->{'client_user'} || "root";
|
---|
601 | my $db_pwd = $self->{'client_pwd'};
|
---|
602 | my $db_host = $self->{'db_host'} || "127.0.0.1";
|
---|
603 | my $db_enc = $self->{'db_encoding'} || "utf8";
|
---|
604 |
|
---|
605 | #my $db_name = $self->{'site_name'};
|
---|
606 |
|
---|
607 | # try connecting to the mysql db, if that fails it will die
|
---|
608 | # so don't bother preparing GreenstoneXMLPlugout by calling superclass' begin()
|
---|
609 |
|
---|
610 | # localhost doesn't work for us, but 127.0.0.1 works
|
---|
611 | # https://metacpan.org/pod/DBD::mysql
|
---|
612 | # "The hostname, if not specified or specified as '' or 'localhost', will default to a MySQL server
|
---|
613 | # running on the local machine using the default for the UNIX socket. To connect to a MySQL server
|
---|
614 | # on the local machine via TCP, you must specify the loopback IP address (127.0.0.1) as the host."
|
---|
615 | #my $connect_str = "dbi:$db_driver:database=$db_name;host=$db_host";
|
---|
616 | my $connect_str = "dbi:$db_driver:host=$db_host"; # don't provide db, so we can check the db is there
|
---|
617 | my $dbh = DBI->connect("$connect_str", $db_user, $db_pwd,
|
---|
618 | {
|
---|
619 | ShowErrorStatement => 1, # more informative as DBI will append failed SQL stmt to error message
|
---|
620 | PrintError => 1, # on by default, but being explicit
|
---|
621 | RaiseError => 0, # off by default, but being explicit
|
---|
622 | AutoCommit => 1, # on by default, but being explicit
|
---|
623 | });
|
---|
624 |
|
---|
625 | if(!$dbh) {
|
---|
626 | # NOTE, despite handle dbh being undefined, error code will be in DBI->err
|
---|
627 | return 0;
|
---|
628 | }
|
---|
629 |
|
---|
630 | # set encoding https://metacpan.org/pod/DBD::mysql
|
---|
631 | # https://dev.mysql.com/doc/refman/5.7/en/charset.html
|
---|
632 | # https://dev.mysql.com/doc/refman/5.7/en/charset-conversion.html
|
---|
633 | # Setting the encoding at db server level.
|
---|
634 | # Not sure if this command is mysql specific:
|
---|
635 | my $stmt = "set NAMES '" . $db_enc . "'";
|
---|
636 | $dbh->do($stmt) || warn("Unable to set charset encoding at db server level to: " . $db_enc . "\n");
|
---|
637 |
|
---|
638 | # if we're here, then connection succeeded, store handle
|
---|
639 | $self->{'db_handle'} = $dbh;
|
---|
640 | return 1;
|
---|
641 | }
|
---|
642 |
|
---|
643 | sub load_db_and_tables {
|
---|
644 | my $self= shift (@_);
|
---|
645 | my $db_name = $self->{'site_name'}; # one database per GS3 site
|
---|
646 | my $dbh = $self->{'db_handle'};
|
---|
647 |
|
---|
648 | # perl DBI switch database: https://www.perlmonks.org/?node_id=995434
|
---|
649 | # do() returns undef on error.
|
---|
650 | # connection succeeded, try to load our database. If that didn't work, attempt to create db
|
---|
651 | my $success = $dbh->do("use $db_name");
|
---|
652 |
|
---|
653 | if(!$success && $dbh->err == 1049) { # "Unknown database" error has code 1049 (mysql only?) meaning db doesn't exist yet
|
---|
654 | # attempt to create the db and its tables
|
---|
655 | $self->create_db($db_name) || return 0;
|
---|
656 |
|
---|
657 | # once more attempt to use db, now that it exists
|
---|
658 | $dbh->do("use $db_name") || return 0;
|
---|
659 | #$dbh->do("use localsite") or die "Error (code" . $dbh->err ."): " . $dbh->errstr . "\n";
|
---|
660 |
|
---|
661 | # attempt to create tables in current db
|
---|
662 | $self->create_metadata_table() || return 0;
|
---|
663 | $self->create_fulltext_table() || return 0;
|
---|
664 |
|
---|
665 | $success = 1;
|
---|
666 | }
|
---|
667 | elsif($success) { # database existed and loaded successfully, but
|
---|
668 | # before proceeding check that the current collection's tables exist
|
---|
669 |
|
---|
670 | # attempt to create tables in current db
|
---|
671 | if(!$self->table_exists($self->{'collection_name'} . "metadata")) {
|
---|
672 | $self->create_metadata_table() || return 0;
|
---|
673 | }
|
---|
674 | if(!$self->table_exists($self->{'collection_name'} . "fulltxt")) {
|
---|
675 | $self->create_fulltext_table() || return 0;
|
---|
676 | }
|
---|
677 | }
|
---|
678 |
|
---|
679 | return $success; # could still return 0, if database failed to load with an error code != 1049
|
---|
680 | }
|
---|
681 |
|
---|
682 | # disconnect from db - https://metacpan.org/pod/DBI#disconnect
|
---|
683 | # TODO: make sure to have committed or rolled back before disconnect
|
---|
684 | # and that you've call finish() on statement handles if any fetch remnants remain
|
---|
685 | sub disconnect_from_db {
|
---|
686 | my $self= shift (@_);
|
---|
687 | my $dbh = $self->{'db_handle'};
|
---|
688 |
|
---|
689 | # make sure any active stmt handles are finished
|
---|
690 | # NO: "When all the data has been fetched from a SELECT statement, the driver will automatically call finish for you. So you should not call it explicitly except when you know that you've not fetched all the data from a statement handle and the handle won't be destroyed soon."
|
---|
691 |
|
---|
692 | #$meta_sth = $self->{'metadata_prepared_insert_statement_handle'};
|
---|
693 | #$txt_sth = $self->{'fulltxt_prepared_insert_statement_handle'};
|
---|
694 | #$meta_sth->finish() if($meta_sth);
|
---|
695 | #$txt_sth->finish() if($txt_sth);
|
---|
696 |
|
---|
697 | my $rc = $dbh->disconnect or warn $dbh->errstr; # The handle is of little use after disconnecting. Possibly PrintError already prints a warning and this duplicates it?
|
---|
698 | return $rc;
|
---|
699 | }
|
---|
700 |
|
---|
701 | sub create_db {
|
---|
702 | my $self= shift (@_);
|
---|
703 | my $db_name = $self->{'site_name'};
|
---|
704 | my $dbh = $self->{'db_handle'};
|
---|
705 |
|
---|
706 | # https://stackoverflow.com/questions/5025768/how-can-i-create-a-mysql-database-from-a-perl-script
|
---|
707 | return $dbh->do("create database $db_name"); # do() will return undef on fail, https://metacpan.org/pod/DBI#do
|
---|
708 | }
|
---|
709 |
|
---|
710 | sub create_metadata_table {
|
---|
711 | my $self= shift (@_);
|
---|
712 | my $dbh = $self->{'db_handle'};
|
---|
713 |
|
---|
714 | my $table_name = $self->{'collection_name'} . "metadata";
|
---|
715 |
|
---|
716 | # If using an auto incremented primary key:
|
---|
717 | my $stmt = "CREATE TABLE $table_name (id INT NOT NULL AUTO_INCREMENT, did VARCHAR(63) NOT NULL, sid VARCHAR(63) NOT NULL, metaname VARCHAR(127) NOT NULL, metavalue VARCHAR(1023) NOT NULL, PRIMARY KEY(id));";
|
---|
718 | return $dbh->do($stmt);
|
---|
719 | }
|
---|
720 |
|
---|
721 | # TODO: Investigate: https://dev.mysql.com/doc/search/?d=10&p=1&q=FULLTEXT
|
---|
722 | # 12.9.1 Natural Language Full-Text Searches
|
---|
723 | # to see whether we have to index the 'fulltxt' column of the 'fulltext' tables
|
---|
724 | # or let user edit this file, or add it as another option
|
---|
725 | sub create_fulltext_table {
|
---|
726 | my $self= shift (@_);
|
---|
727 | my $dbh = $self->{'db_handle'};
|
---|
728 |
|
---|
729 | my $table_name = $self->{'collection_name'} . "fulltxt";
|
---|
730 |
|
---|
731 | # If using an auto incremented primary key:
|
---|
732 | my $stmt = "CREATE TABLE $table_name (id INT NOT NULL AUTO_INCREMENT, did VARCHAR(63) NOT NULL, sid VARCHAR(63) NOT NULL, fulltxt LONGTEXT, PRIMARY KEY(id));";
|
---|
733 | return $dbh->do($stmt);
|
---|
734 |
|
---|
735 | }
|
---|
736 |
|
---|
737 |
|
---|
738 | # USEFUL: https://metacpan.org/pod/DBI
|
---|
739 | # "Many methods have an optional \%attr parameter which can be used to pass information to the driver implementing the method. Except where specifically documented, the \%attr parameter can only be used to pass driver specific hints. In general, you can ignore \%attr parameters or pass it as undef."
|
---|
740 |
|
---|
741 |
|
---|
742 | # https://www.guru99.com/insert-into.html
|
---|
743 | # and https://dev.mysql.com/doc/refman/8.0/en/example-auto-increment.html
|
---|
744 | # for inserting multiple rows at once
|
---|
745 | # https://www.perlmonks.org/bare/?node_id=316183
|
---|
746 | # https://metacpan.org/pod/DBI#do
|
---|
747 | # https://www.quora.com/What-is-the-difference-between-prepare-and-do-statements-in-Perl-while-we-make-a-connection-to-the-database-for-executing-the-query
|
---|
748 | # https://docstore.mik.ua/orelly/linux/dbi/ch05_05.htm
|
---|
749 |
|
---|
750 | # https://metacpan.org/pod/DBI#performance
|
---|
751 | # 'The q{...} style quoting used in this example avoids clashing with quotes that may be used in the SQL statement. Use the double-quote like qq{...} operator if you want to interpolate variables into the string. See "Quote and Quote-like Operators" in perlop for more details.'
|
---|
752 | sub prepare_insert_metadata_row_stmthandle {
|
---|
753 | my $self = shift (@_);
|
---|
754 | my ($did, $sid, $metaname, $metavalue) = @_;
|
---|
755 | my $dbh = $self->{'db_handle'};
|
---|
756 |
|
---|
757 | my $tablename = $self->{'colname'}."_metadata";
|
---|
758 |
|
---|
759 | #my $stmt = "INSERT INTO $tablename (did, sid, metaname, metavalue) VALUES ('$did', '$sid', '$metaname', '$metavalue');"; # ?, ?, ?, ?
|
---|
760 |
|
---|
761 | # using qq{} since we want $tablename placeholder to be filled in
|
---|
762 | # returns Statement Handle object!
|
---|
763 | my $sth = $dbh->prepare(qq{INSERT INTO $tablename (did, sid, metaname, metavalue) VALUES (?, ?, ?, ?)"}) || warn("Could not prepare insert statement for metadata table\n");
|
---|
764 |
|
---|
765 | return $sth;
|
---|
766 | }
|
---|
767 |
|
---|
768 | sub prepare_insert_fulltxt_row_stmthandle {
|
---|
769 | my $self = shift (@_);
|
---|
770 | my ($did, $sid, $fulltext) = @_;
|
---|
771 | my $dbh = $self->{'db_handle'};
|
---|
772 |
|
---|
773 | my $tablename = $self->{'colname'}."_fulltxt";
|
---|
774 |
|
---|
775 | #my $stmt = "INSERT INTO $tablename (did, sid, fulltxt) VALUES ('$did', '$sid', '$fulltext');"; ?, ?, ?
|
---|
776 |
|
---|
777 | # using qq{} since we want $tablename placeholder to be filled in
|
---|
778 | # returns Statement Handle object!
|
---|
779 | my $sth = $dbh->prepare(qq{INSERT INTO $tablename (did, sid, fulltxt) VALUES (?, ?, ?)"}) || warn("Could not prepare insert statement for fulltxt table\n");
|
---|
780 |
|
---|
781 | return $sth;
|
---|
782 | }
|
---|
783 |
|
---|
784 | # I can get my version of table_exists to work, but it's not so ideal
|
---|
785 | # Interesting that MySQL has non-standard command to CREATE TABLE IF NOT EXISTS and DROP TABLE IF EXISTS,
|
---|
786 | # see https://www.perlmonks.org/bare/?node=DBI%20Recipes
|
---|
787 | # The page further has a table_exists function that could work with proper comparison
|
---|
788 | # Couldn't get the first solution at https://www.perlmonks.org/bare/?node_id=500050 to work though
|
---|
789 | sub table_exists {
|
---|
790 | my ($dbh,$table_name) = @_;
|
---|
791 |
|
---|
792 | my @table_list = $dbh->tables;
|
---|
793 | #my $tables_str = @table_list[0];
|
---|
794 | foreach my $table (@table_list) {
|
---|
795 | return 1 if ($table =~ m/$table_name/);
|
---|
796 | }
|
---|
797 | return 0;
|
---|
798 | }
|
---|
799 |
|
---|
800 | 1;
|
---|