root/main/trunk/greenstone2/perllib/plugouts/GreenstoneSQLPlugout.pm @ 32584

Revision 32584, 14.4 KB (checked in by ak19, 7 months ago)

Some more tidying up of the code.

Line 
1###########################################################################
2#
3# GreenstoneSQLPlugout.pm -- plugout module for writing all or some the
4# Greenstone document format (metadata and/or fulltext) into a (My)SQL db.
5# The rest is then still written out by GreenstoneXMLPlugout as usual.
6# A component of the Greenstone digital library software
7# from the New Zealand Digital Library Project at the
8# University of Waikato, New Zealand.
9#
10# Copyright (C) 2006 New Zealand Digital Library Project
11#
12# This program is free software; you can redistribute it and/or modify
13# it under the terms of the GNU General Public License as published by
14# the Free Software Foundation; either version 2 of the License, or
15# (at your option) any later version.
16#
17# This program is distributed in the hope that it will be useful,
18# but WITHOUT ANY WARRANTY; without even the implied warranty of
19# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
20# GNU General Public License for more details.
21#
22# You should have received a copy of the GNU General Public License
23# along with this program; if not, write to the Free Software
24# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
25#
26###########################################################################
27
28package GreenstoneSQLPlugout;
29
30use strict;
31no strict 'refs';
32no strict 'subs';
33
34use GreenstoneXMLPlugout;
35use docprint;
36use gssql;
37
38use DBI; # the central package for this plugout
39
40
41# + TODO: SIGTERM rollback and disconnect?
42# + TODO Q: what about verbosity for debugging, instead of current situation of printing out upon debug set at the expense of writing to db
43# TODO Q: introduced site_name param to plugins and plugouts. Did I do it right? And should they have hiddengli = "yes"
44# Did I do the pass by ref in docprint's escape and unescape textref functions correctly, and how they're called here?
45#   Any more optimisation I can do around this?
46
47# this plugout does not output the metadata and/or fulltxt xml to a file,
48# but outputs rows into a mysql table for metadata and/or a table for fulltxt
49sub BEGIN {
50    @GreenstoneSQLPlugout::ISA = ('GreenstoneXMLPlugout');
51}
52
53# + NOTTODO: die() statements need to be replaced with premature_termination
54# which should ensure the GreenstoneXMLPlugin (group)'s stuff is closed and cleaned up SOMEHOW
55# It's fine: the die() stmts all take place before setting up the super class' begin
56
57# + TODO Q: about build_mode: how to detect removeold. Now handled by
58#   GreenstoneSQLPlugout::remove_all(), which is inherited from a base plugin.
59# + TODO: deal with -removeold and everything? Or type out instructions for user
60
61# + TODO Q: what is "group" in GreenstoneXMLPlugout?
62
63my $process_mode_list =
64    [ { 'name' => "meta_only",
65        'desc' => "{GreenstoneSQLPlug.process_mode.meta_only}" },     
66      { 'name' => "text_only",
67        'desc' => "{GreenstoneSQLPlug.process_mode.text_only}" },
68      { 'name' => "all",
69        'desc' => "{GreenstoneSQLPlug.process_mode.all}" } ];
70
71my $rollback_on_cancel_list =
72    [ { 'name' => "true",
73        'desc' => "{GreenstoneSQLPlug.rollback_on_cancel}" },     
74      { 'name' => "false",
75        'desc' => "{GreenstoneSQLPlug.rollbacl_on_cancel}" } ];
76
77# The following are the saveas.options:
78my $arguments = [
79    { 'name' => "process_mode",
80      'desc' => "{GreenstoneSQLPlug.process_mode}",
81      'type' => "enum",
82      'list' => $process_mode_list,
83      'deft' => "all",
84      'reqd' => "no",
85      'hiddengli' => "no"},
86    { 'name' => "rollback_on_cancel",
87      'desc' => "{GreenstoneSQLPlug.rollback_on_cancel}",
88      'type' => "enum",
89      'list' => $rollback_on_cancel_list,
90      'deft' => "true", # TODO Q: what's the better default? If "true", any memory concerns?
91      'reqd' => "no",
92      'hiddengli' => "no"},
93    { 'name' => "db_driver",
94      'desc' => "{GreenstoneSQLPlug.db_driver}",
95      'type' => "string",   
96      'deft' => "mysql",
97      'reqd' => "yes"},
98    { 'name' => "db_client_user",
99      'desc' => "{GreenstoneSQLPlug.db_client_user}",
100      'type' => "string",   
101      'deft' => "root",
102      'reqd' => "yes"},
103    { 'name' => "db_client_pwd",
104      'desc' => "{GreenstoneSQLPlug.db_client_pwd}",
105      'type' => "string",
106      'deft' => "",
107      'reqd' => "yes"}, # pwd required?
108    { 'name' => "db_host",
109      'desc' => "{GreenstoneSQLPlug.db_host}",
110      'type' => "string",
111      'deft' => "127.0.0.1",
112      'reqd' => "yes"}
113    ];
114
115my $options = { 'name'     => "GreenstoneSQLPlugout",
116        'desc'     => "{GreenstoneSQLPlugout.desc}",
117        'abstract' => "no",
118        'inherits' => "yes",
119        'args'     => $arguments };
120
121##### This entire class is called only during import.pl #####
122
123##### Overridden methods #####
124
125sub new {
126    my ($class) = shift (@_);
127    my ($plugoutlist, $inputargs,$hashArgOptLists) = @_;
128    push(@$plugoutlist, $class);
129
130    push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
131    push(@{$hashArgOptLists->{"OptList"}},$options);
132
133    my $self = new GreenstoneXMLPlugout($plugoutlist,$inputargs,$hashArgOptLists);
134   
135    if ($self->{'info_only'}) {
136        # don't worry about any options etc
137        return bless $self, $class;
138    }
139    #print STDERR "***** GreenstoneSQLPlugout process mode = \"", $self->{'process_mode'}, "\"\n";
140   
141    return bless $self, $class;
142}
143
144# connect here and ensure all tables and databases exist
145sub begin {
146
147    my $self= shift (@_);
148   
149    # The saveas.options
150    #print STDERR "@@@@ PLUGOUT db_pwd: " . $self->{'db_client_pwd'} . "\n";
151    #print STDERR "@@@@ user: " . $self->{'db_client_user'} . "\n";
152    #print STDERR "@@@@ db_host: " . $self->{'db_host'} . "\n";
153    #print STDERR "@@@@ db_driver: " . $self->{'db_driver'} . "\n";
154    #print STDERR "@@@@ proc_mode: " . $self->{'process_mode'} . "\n";
155
156    ############ LOAD NECESSARY OPTIONS ###########
157    #print "@@@ plugout SITE NAME: ". $self->{'site_name'} . "\n" if defined $self->{'site_name'};
158    #print STDERR "########## COLLECTION: ". $ENV{'GSDLCOLLECTION'}."\n";
159
160    print STDERR "***** GreenstoneSQLPlugout process mode = \"", $self->{'process_mode'}, "\"\n";
161
162    my $db_params = {
163    'collection_name' => $ENV{'GSDLCOLLECTION'},
164    'verbosity' => $self->{'verbosity'} || 0
165   
166    };
167
168    my $gs_sql = new gssql($db_params);
169
170    # if autocommit is set, there's no rollback support
171    my $autocommit = ($self->{'rollback_on_cancel'} eq "false") ? 1 : 0;
172   
173    # try connecting to the mysql db, die if that fails
174    # So don't bother preparing GreenstoneXMLPlugout by calling superclass' begin() yet
175    if(!$gs_sql->connect_to_db({
176    'db_driver' => $self->{'db_driver'},
177    'db_client_user' => $self->{'db_client_user'},
178    'db_client_pwd' => $self->{'db_client_pwd'},
179    'db_host' => $self->{'db_host'},
180    'autocommit' => $autocommit
181                   })
182    )
183    {
184    # This is fatal for the plugout, let's terminate here
185    # PrintError would already have displayed the warning message on connection fail   
186    die("Could not connect to db. Can't proceed.\n");
187    }
188
189    #die("@@@@ TEST. Connected successfully. Testing gssql::destructor.\n"); # WORKS
190   
191    my $db_name = $self->{'site_name'} || "greenstone2"; # one database per GS3 site, for GS2 the db is called greenstone2
192    my $proc_mode = $self->{'process_mode'};
193   
194
195    my $success = $gs_sql->use_db($db_name);
196   
197    if($success && $proc_mode ne "text_only") {
198    ##print STDERR "@@@@ Ensuring meta table exists\n";
199    $success = $gs_sql->ensure_meta_table_exists();
200    }   
201    if($success && $proc_mode ne "meta_only") {
202    ##print STDERR "@@@@ Ensuring fulltxt table exists\n";
203    $success = $gs_sql->ensure_fulltxt_table_exists();
204    }
205   
206    if(!$success) {
207    # This is fatal for the plugout, let's terminate here after disconnecting again
208    # PrintError would already have displayed the warning message on load fail
209    # And on die() perl will call gssql destroy which will ensure a disconnect() from db
210    #$gs_sql->force_disconnect_from_db(); # disconnect_from_db() will issue a warning on error
211    die("Could not use db $db_name and/or prepare its tables. Can't proceed.\n");
212    }
213
214    # store the DBI wrapper instance
215    $self->{'gs_sql'} = $gs_sql;
216   
217   
218    # if setting up to work with sql db failed, we'd have terminated and wouldn't come up to here:
219    # won't bother preparing GreenstoneXMLPlugout by calling superclass' begin()
220    # finally, can call begin on super - important as doc.xml is opened as a group etc
221   
222    $self->SUPER::begin(@_);
223}
224
225# disconnect from database here, see inexport.pm
226sub end
227{
228    my $self = shift(@_);
229
230    # do the superclass stuff first, as any sql db failures should not prevent superclass cleanup
231    $self->SUPER::end(@_);   
232
233    # Important to call finished():
234    # it will disconnect from db if this is the last gssql instance,
235    # and it will commit to db before disconnecting if rollbback_on_cancel turned on
236    $self->{'gs_sql'}->finished();
237    delete $self->{'gs_sql'}; # key gs_sql no longer exists, not just the value being undef
238}
239
240# Produce files called docsql.xml instead of doc.xml
241sub get_doc_xml_filename {
242    my $self = shift (@_);
243    my ($doc_obj) = @_;
244   
245    return "docsql.xml";
246}
247
248# overriding to store doc OID as attribute of top level element: <Archive docoid="oid">
249sub output_xml_header {
250    my $self = shift (@_);
251    my ($outhandle, $doc_oid) = @_;
252
253    print $outhandle '<?xml version="1.0" encoding="utf-8" standalone="no"?>' . "\n";
254    print $outhandle "<!DOCTYPE Archive SYSTEM \"http://greenstone.org/dtd/Archive/1.0/Archive.dtd\">\n";
255    print $outhandle "<Archive docoid=\"$doc_oid\">\n";
256}
257 
258# + X TODO: check arc-inf.db for whether each entry is to be deleted/indexed/reindexed/been indexed?
259# That's only for indexing, not for this step which only generates the content in archives dir
260sub saveas {
261    my $self = shift (@_);
262    my ($doc_obj, $doc_dir) = @_;
263
264    my $proc_mode = $self->{'process_mode'};
265   
266    # 1. pre save out and saving debug handle
267
268    # must call superclass (pre/post) saveas methods, as they handle assoc_files too
269    my ($docxml_outhandler, $output_file) = $self->SUPER::pre_saveas(@_);
270
271    $self->{'debug_outhandle'} = $docxml_outhandler if ($self->{'debug'}); # STDOUT if debug
272
273    # TODO: also set debugging in begin()? Then stmts creating db and tables also sent to debug out and not executed
274   
275   
276    # 2. overriding saving behaviour to do what the superclass does (writing out doc.xml files,
277    # under new name of docsql.xml, with breadcrumbs pointing to sql db) PLUS saving to sql db   
278
279    # NOTE: if proc_mode == all, then "breadcrumbs" (statements pointing viewer to the sql db
280    # for contents) go into both meta and txt elements of doc.xml (docsql.xml specifically):
281   
282    # write the INVERSE into doc.xml as to what is written to the SQL db   
283    my $docxml_output_options = { 'output' => docprint::OUTPUT_NONE };
284    if($proc_mode eq "meta_only" ) { # since only meta to go into MySQL db, text will go into docxml
285    $docxml_output_options->{'output'} = docprint::OUTPUT_TEXT_ONLY;
286    } elsif($proc_mode eq "text_only" ) { # since only full text to go into MySQL db, meta will go into docxml
287    $docxml_output_options->{'output'} = docprint::OUTPUT_META_ONLY;
288    }
289   
290    # now we've prepared to write out whatever is meant to go into docxml
291    # and can do actual the steps superclass GreenstoneXMLPlugout carries out to write out docxml
292    # So: write out the doc xml file, "docsql.xml", for the current document
293    my $section_text = &docprint::get_section_xml($doc_obj, $docxml_output_options);
294    print $docxml_outhandler $section_text;   
295   
296   
297    # We also write out whatever needs to go into the MySQL database
298    $self->write_meta_and_text($doc_obj);
299
300   
301    # 3. post save out
302    $self->SUPER::post_saveas($doc_obj, $doc_dir, $docxml_outhandler, $output_file);
303   
304   
305    # database connection is closed in end() method
306    # so we don't open and close over and over for each doc during a single build
307}
308
309##### New methods, not inherited  #####
310
311# write meta and/or text PER DOC out to DB
312sub write_meta_and_text {
313    my $self = shift (@_);
314    my ($doc_obj) = @_;
315    my $doc_oid = $doc_obj->get_OID(); # this method processes a single doc at a time, so it uses the same OID throughout
316    my $root_section = $doc_obj->get_top_section();
317
318    $self->recursive_write_meta_and_text($doc_obj, $doc_oid, $root_section);
319}
320
321# Perl: Reading or Writing to Another Program
322# https://nnc3.com/mags/Perl3/cookbook/ch16_05.htm
323sub recursive_write_meta_and_text {
324    my $self = shift (@_);
325    my ($doc_obj, $doc_oid, $section) = @_;   
326
327    # If section=ROOT, write "root" as section name into table
328    # doc->get_top_section() is the name of the doc root section, which is ""
329    my $section_name = ($section eq "") ? "root" : $section;
330   
331    my $section_ptr = $doc_obj->_lookup_section ($section);
332    return "" unless defined $section_ptr;
333
334    my $debug_out = $self->{'debug_outhandle'};
335
336    my $gs_sql = $self->{'gs_sql'};
337    my $proc_mode = $self->{'process_mode'};
338    if($proc_mode eq "all" || $proc_mode eq "meta_only" ) {
339   
340    foreach my $data (@{$section_ptr->{'metadata'}}) {
341        my $meta_name = $data->[0];
342        # TODO: does it need to be stored escaped, as it requires unescaping when read back in
343        # from db (unlike for reading back in from doc.xml)
344        my $escaped_meta_value = &docprint::escape_text($data->[1]);
345       
346        # Write out the current section's meta to collection db's METADATA table       
347       
348        # For each set of values to write to meta table, this next method call will
349        # efficiently execute an insert SQL statement (using a prepared insert statement),
350        # filling in the values
351        # OR if debugging, then it will print the SQL insert statement but not execute it
352       
353        $gs_sql->insert_row_into_metadata_table($doc_oid, $section_name, $meta_name, $escaped_meta_value, $self->{'debug'});
354    }
355    }
356   
357   
358    if($proc_mode eq "all" || $proc_mode eq "text_only" ) {
359   
360    my $section_textref = &docprint::escape_textref(\$section_ptr->{'text'});
361   
362    # fulltxt column can be SQL NULL. undef value gets written out as NULL:
363    # https://stackoverflow.com/questions/12708633/which-one-represents-null-undef-or-empty-string
364    # The following will do the SQL insertion
365    # or if debug, the following will print the SQL insert stmt without executing it
366    $gs_sql->insert_row_into_fulltxt_table($doc_oid, $section_name, $section_textref, $self->{'debug'});
367   
368    }
369   
370    # output all subsections: RECURSIVE CALL
371    foreach my $subsection (@{$section_ptr->{'subsection_order'}}) {
372    $self->recursive_write_meta_and_text($doc_obj, $doc_oid, "$section.$subsection");
373    }
374}
375
376
3771;
Note: See TracBrowser for help on using the browser.