1 | ###########################################################################
|
---|
2 | #
|
---|
3 | # dbutil::jdbm -- utility functions for writing to jdbm 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::jdbm;
|
---|
28 |
|
---|
29 | use strict;
|
---|
30 |
|
---|
31 |
|
---|
32 | # -----------------------------------------------------------------------------
|
---|
33 | # JDBM IMPLEMENTATION
|
---|
34 | # -----------------------------------------------------------------------------
|
---|
35 |
|
---|
36 | # When DBUtil::* is properly structured with inheritence, then
|
---|
37 | # much of this code (along with GDBM and GDBM-TXT-GZ) can be grouped into
|
---|
38 | # a shared base class. Really it is only the the command that needs to
|
---|
39 | # be constructed that changes between much of the code that is used
|
---|
40 |
|
---|
41 |
|
---|
42 | sub open_infodb_write_handle
|
---|
43 | {
|
---|
44 | my $infodb_file_path = shift(@_);
|
---|
45 | my $opt_append = shift(@_);
|
---|
46 |
|
---|
47 | my $jdbmwrap_jar = &util::filename_cat($ENV{'GSDLHOME'},"bin","java", "JDBMWrapper.jar");
|
---|
48 | my $jdbm_jar = &util::filename_cat($ENV{'GSDLHOME'},"lib","java", "jdbm.jar");
|
---|
49 |
|
---|
50 | my $classpath = &util::pathname_cat($jdbmwrap_jar,$jdbm_jar);
|
---|
51 |
|
---|
52 | if ($^O eq "cygwin") {
|
---|
53 | # Away to run a java program, using a binary that is native to Windows, so need
|
---|
54 | # Windows directory and path separators
|
---|
55 |
|
---|
56 | $classpath = `cygpath -wp "$classpath"`;
|
---|
57 | chomp($classpath);
|
---|
58 | $classpath =~ s%\\%\\\\%g;
|
---|
59 | }
|
---|
60 |
|
---|
61 | my $infodb_file_handle = undef;
|
---|
62 | my $txt2jdb_cmd = "java -cp \"$classpath\" Txt2Jdb";
|
---|
63 |
|
---|
64 | if ((defined $opt_append) && ($opt_append eq "append")) {
|
---|
65 | $txt2jdb_cmd .= " -append";
|
---|
66 | print STDERR "Append operation to $infodb_file_path\n";
|
---|
67 | }
|
---|
68 | else {
|
---|
69 | print STDERR "Create database $infodb_file_path\n";
|
---|
70 | }
|
---|
71 |
|
---|
72 | # Lop off file extension, as JDBM does not expect this to be present
|
---|
73 | $infodb_file_path =~ s/\.jdb$//;
|
---|
74 |
|
---|
75 | if ($^O eq "cygwin") {
|
---|
76 | $infodb_file_path = `cygpath -w "$infodb_file_path"`;
|
---|
77 | chomp($infodb_file_path);
|
---|
78 | $infodb_file_path =~ s%\\%\\\\%g;
|
---|
79 | }
|
---|
80 |
|
---|
81 | $txt2jdb_cmd .= " \"$infodb_file_path\"";
|
---|
82 |
|
---|
83 | if (!open($infodb_file_handle, "| $txt2jdb_cmd"))
|
---|
84 | {
|
---|
85 | print STDERR "Error: Failed to open pipe to $txt2jdb_cmd";
|
---|
86 | print STDERR " $!\n";
|
---|
87 | return undef;
|
---|
88 | }
|
---|
89 |
|
---|
90 | binmode($infodb_file_handle,":utf8");
|
---|
91 | return $infodb_file_handle;
|
---|
92 | }
|
---|
93 |
|
---|
94 |
|
---|
95 |
|
---|
96 | sub close_infodb_write_handle
|
---|
97 | {
|
---|
98 | my $infodb_handle = shift(@_);
|
---|
99 |
|
---|
100 | close($infodb_handle);
|
---|
101 | }
|
---|
102 |
|
---|
103 |
|
---|
104 | sub get_infodb_file_path
|
---|
105 | {
|
---|
106 | my $collection_name = shift(@_);
|
---|
107 | my $infodb_directory_path = shift(@_);
|
---|
108 |
|
---|
109 | my $infodb_file_extension = ".jdb";
|
---|
110 | my $infodb_file_name = &util::get_dirsep_tail($collection_name) . $infodb_file_extension;
|
---|
111 | return &util::filename_cat($infodb_directory_path, $infodb_file_name);
|
---|
112 | }
|
---|
113 |
|
---|
114 |
|
---|
115 | sub read_infodb_file
|
---|
116 | {
|
---|
117 | my $infodb_file_path = shift(@_);
|
---|
118 | my $infodb_map = shift(@_);
|
---|
119 |
|
---|
120 | my $jdbmwrap_jar = &util::filename_cat($ENV{'GSDLHOME'},"bin","java", "JDBMWrapper.jar");
|
---|
121 | my $jdbm_jar = &util::filename_cat($ENV{'GSDLHOME'},"lib","java", "jdbm.jar");
|
---|
122 |
|
---|
123 | my $classpath = &util::pathname_cat($jdbmwrap_jar,$jdbm_jar);
|
---|
124 |
|
---|
125 | if ($^O eq "cygwin") {
|
---|
126 | # Away to run a java program, using a binary that is native to Windows, so need
|
---|
127 | # Windows directory and path separators
|
---|
128 |
|
---|
129 | $classpath = `cygpath -wp "$classpath"`;
|
---|
130 | chomp($classpath);
|
---|
131 | $classpath =~ s%\\%\\\\%g;
|
---|
132 |
|
---|
133 | $infodb_file_path = `cygpath -w "$infodb_file_path"`;
|
---|
134 | chomp($infodb_file_path);
|
---|
135 | $infodb_file_path =~ s%\\%\\\\%g;
|
---|
136 | }
|
---|
137 |
|
---|
138 | my $jdb2txt_cmd = "java -cp \"$classpath\" Jdb2Txt";
|
---|
139 |
|
---|
140 | open (PIPEIN, "$jdb2txt_cmd \"$infodb_file_path\" |") || die "couldn't open pipe from db2txt \$infodb_file_path\"\n";
|
---|
141 | binmode(PIPEIN,":utf8");
|
---|
142 | my $infodb_line = "";
|
---|
143 | my $infodb_key = "";
|
---|
144 | my $infodb_value = "";
|
---|
145 | while (defined ($infodb_line = <PIPEIN>))
|
---|
146 | {
|
---|
147 | $infodb_line =~ s/(\r\n)+$//; # more general than chomp
|
---|
148 |
|
---|
149 | if ($infodb_line =~ /^\[([^\]]+)\]$/)
|
---|
150 | {
|
---|
151 | $infodb_key = $1;
|
---|
152 | }
|
---|
153 | elsif ($infodb_line =~ /^-{70}$/)
|
---|
154 | {
|
---|
155 | $infodb_map->{$infodb_key} = $infodb_value;
|
---|
156 | $infodb_key = "";
|
---|
157 | $infodb_value = "";
|
---|
158 | }
|
---|
159 | else
|
---|
160 | {
|
---|
161 | $infodb_value .= $infodb_line;
|
---|
162 | }
|
---|
163 | }
|
---|
164 |
|
---|
165 | close (PIPEIN);
|
---|
166 | }
|
---|
167 |
|
---|
168 | sub read_infodb_keys
|
---|
169 | {
|
---|
170 | my $infodb_file_path = shift(@_);
|
---|
171 | my $infodb_map = shift(@_);
|
---|
172 |
|
---|
173 | my $jdbmwrap_jar = &util::filename_cat($ENV{'GSDLHOME'},"bin","java", "JDBMWrapper.jar");
|
---|
174 | my $jdbm_jar = &util::filename_cat($ENV{'GSDLHOME'},"lib","java", "jdbm.jar");
|
---|
175 |
|
---|
176 | my $classpath = &util::pathname_cat($jdbmwrap_jar,$jdbm_jar);
|
---|
177 |
|
---|
178 | my $jdbkeys_cmd = "java -cp \"$classpath\" JdbKeys";
|
---|
179 |
|
---|
180 | open (PIPEIN, "$jdbkeys_cmd \"$infodb_file_path\" |") || die "couldn't open pipe from jdbmkeys \$infodb_file_path\"\n";
|
---|
181 | binmode(PIPEIN,":utf8");
|
---|
182 | my $infodb_line = "";
|
---|
183 | my $infodb_key = "";
|
---|
184 | my $infodb_value = "";
|
---|
185 | while (defined ($infodb_line = <PIPEIN>))
|
---|
186 | {
|
---|
187 | # chomp $infodb_line; # remove end of line
|
---|
188 | $infodb_line =~ s/(\r\n)+$//; # more general than chomp
|
---|
189 |
|
---|
190 | $infodb_map->{$infodb_line} = 1;
|
---|
191 | }
|
---|
192 |
|
---|
193 | close (PIPEIN);
|
---|
194 | }
|
---|
195 |
|
---|
196 |
|
---|
197 |
|
---|
198 | sub write_infodb_entry
|
---|
199 | {
|
---|
200 |
|
---|
201 | my $infodb_handle = shift(@_);
|
---|
202 | my $infodb_key = shift(@_);
|
---|
203 | my $infodb_map = shift(@_);
|
---|
204 |
|
---|
205 | print $infodb_handle "[$infodb_key]\n";
|
---|
206 | foreach my $infodb_value_key (keys(%$infodb_map))
|
---|
207 | {
|
---|
208 | foreach my $infodb_value (@{$infodb_map->{$infodb_value_key}})
|
---|
209 | {
|
---|
210 | if ($infodb_value =~ /-{70,}/)
|
---|
211 | {
|
---|
212 | # if value contains 70 or more hyphens in a row we need to escape them
|
---|
213 | # to prevent txt2db from treating them as a separator
|
---|
214 | $infodb_value =~ s/-/&\#045;/gi;
|
---|
215 | }
|
---|
216 | print $infodb_handle "<$infodb_value_key>" . $infodb_value . "\n";
|
---|
217 | }
|
---|
218 | }
|
---|
219 | print $infodb_handle '-' x 70, "\n";
|
---|
220 | }
|
---|
221 |
|
---|
222 |
|
---|
223 | sub write_infodb_rawentry
|
---|
224 | {
|
---|
225 |
|
---|
226 | my $infodb_handle = shift(@_);
|
---|
227 | my $infodb_key = shift(@_);
|
---|
228 | my $infodb_val = shift(@_);
|
---|
229 |
|
---|
230 | print $infodb_handle "[$infodb_key]\n";
|
---|
231 | print $infodb_handle "$infodb_val\n";
|
---|
232 | print $infodb_handle '-' x 70, "\n";
|
---|
233 | }
|
---|
234 |
|
---|
235 | sub set_infodb_entry
|
---|
236 | {
|
---|
237 | my $infodb_file_path = shift(@_);
|
---|
238 | my $infodb_key = shift(@_);
|
---|
239 | my $infodb_map = shift(@_);
|
---|
240 |
|
---|
241 | # HTML escape anything that is not part of the "contains" metadata value
|
---|
242 | foreach my $k (keys %$infodb_map) {
|
---|
243 | my @escaped_v = ();
|
---|
244 | foreach my $v (@{$infodb_map->{$k}}) {
|
---|
245 | if ($k eq "contains") {
|
---|
246 | push(@escaped_v, $v);
|
---|
247 | }
|
---|
248 | else {
|
---|
249 | my $ev = &ghtml::unescape_html($v);
|
---|
250 | push(@escaped_v, $ev);
|
---|
251 | }
|
---|
252 | }
|
---|
253 | $infodb_map->{$k} = \@escaped_v;
|
---|
254 | }
|
---|
255 |
|
---|
256 | # Generate the record string
|
---|
257 | my $serialized_infodb_map = &dbutil::convert_infodb_hash_to_string($infodb_map);
|
---|
258 | ### print STDERR "**** ser dr\n$serialized_infodb_map\n\n\n";
|
---|
259 |
|
---|
260 | # Store it into JDBM using 'Txt2Jdb .... -append' which despite its name
|
---|
261 | # actually replaces the record if it already exists
|
---|
262 |
|
---|
263 | my $jdbmwrap_jar = &util::filename_cat($ENV{'GSDLHOME'},"bin","java", "JDBMWrapper.jar");
|
---|
264 | my $jdbm_jar = &util::filename_cat($ENV{'GSDLHOME'},"lib","java", "jdbm.jar");
|
---|
265 |
|
---|
266 | my $classpath = &util::pathname_cat($jdbmwrap_jar,$jdbm_jar);
|
---|
267 |
|
---|
268 | # Lop off file extension, as JDBM does not expect this to be present
|
---|
269 | $infodb_file_path =~ s/\.jdb$//;
|
---|
270 |
|
---|
271 | if ($^O eq "cygwin") {
|
---|
272 | # Away to run a java program, using a binary that is native to Windows, so need
|
---|
273 | # Windows directory and path separators
|
---|
274 |
|
---|
275 | $classpath = `cygpath -wp "$classpath"`;
|
---|
276 | chomp($classpath);
|
---|
277 | $classpath =~ s%\\%\\\\%g;
|
---|
278 |
|
---|
279 | $infodb_file_path = `cygpath -w "$infodb_file_path"`;
|
---|
280 | chomp($infodb_file_path);
|
---|
281 | $infodb_file_path =~ s%\\%\\\\%g;
|
---|
282 | }
|
---|
283 |
|
---|
284 | my $cmd = "java -cp \"$classpath\" Txt2Jdb -append \"$infodb_file_path\"";
|
---|
285 |
|
---|
286 | my $status = undef;
|
---|
287 | if(!open(GOUT, "| $cmd"))
|
---|
288 | {
|
---|
289 | print STDERR "Error: jdbm::set_infodb_entry() failed to open pipe to: $cmd\n";
|
---|
290 | print STDERR " $!\n";
|
---|
291 | $status = -1;
|
---|
292 | }
|
---|
293 | else {
|
---|
294 | binmode(GOUT,":utf8");
|
---|
295 |
|
---|
296 | print GOUT "[$infodb_key]\n";
|
---|
297 | print GOUT "$serialized_infodb_map\n";
|
---|
298 |
|
---|
299 | close(GOUT);
|
---|
300 | $status = 0; # as in exit status of cmd OK
|
---|
301 | }
|
---|
302 |
|
---|
303 | return $status;
|
---|
304 | }
|
---|
305 |
|
---|
306 |
|
---|
307 |
|
---|
308 |
|
---|
309 | sub delete_infodb_entry
|
---|
310 | {
|
---|
311 | my $infodb_handle = shift(@_);
|
---|
312 | my $infodb_key = shift(@_);
|
---|
313 |
|
---|
314 | # A minus at the end of a key (after the ]) signifies 'delete'
|
---|
315 | print $infodb_handle "[$infodb_key]-\n";
|
---|
316 |
|
---|
317 | # The 70 minus signs are also needed, to help make the parsing by db2txt simple
|
---|
318 | print $infodb_handle '-' x 70, "\n";
|
---|
319 | }
|
---|
320 |
|
---|
321 |
|
---|
322 |
|
---|
323 |
|
---|
324 |
|
---|
325 | 1;
|
---|