You can subscribe to this list here.
| 2007 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
(44) |
Aug
(143) |
Sep
(50) |
Oct
(46) |
Nov
(94) |
Dec
(106) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2008 |
Jan
(118) |
Feb
(77) |
Mar
(163) |
Apr
(203) |
May
(74) |
Jun
(90) |
Jul
(111) |
Aug
(23) |
Sep
(99) |
Oct
(46) |
Nov
(42) |
Dec
(38) |
| 2009 |
Jan
(55) |
Feb
(36) |
Mar
(60) |
Apr
(20) |
May
(75) |
Jun
(102) |
Jul
(117) |
Aug
(104) |
Sep
(188) |
Oct
(105) |
Nov
(45) |
Dec
(59) |
| 2010 |
Jan
(88) |
Feb
(147) |
Mar
(53) |
Apr
(54) |
May
(74) |
Jun
(76) |
Jul
(78) |
Aug
(99) |
Sep
(87) |
Oct
(133) |
Nov
(56) |
Dec
(61) |
| 2011 |
Jan
(61) |
Feb
(22) |
Mar
(57) |
Apr
(79) |
May
(128) |
Jun
(143) |
Jul
(64) |
Aug
(90) |
Sep
(90) |
Oct
(202) |
Nov
(91) |
Dec
(127) |
| 2012 |
Jan
(92) |
Feb
(122) |
Mar
(221) |
Apr
(248) |
May
(244) |
Jun
(228) |
Jul
(213) |
Aug
(189) |
Sep
(125) |
Oct
(140) |
Nov
(137) |
Dec
(65) |
| 2013 |
Jan
(124) |
Feb
(160) |
Mar
(168) |
Apr
(276) |
May
(154) |
Jun
(106) |
Jul
(117) |
Aug
(124) |
Sep
(113) |
Oct
(95) |
Nov
(19) |
Dec
(43) |
| 2014 |
Jan
(93) |
Feb
(39) |
Mar
(103) |
Apr
(99) |
May
(66) |
Jun
(78) |
Jul
(98) |
Aug
(32) |
Sep
(69) |
Oct
(71) |
Nov
(60) |
Dec
(64) |
| 2015 |
Jan
(32) |
Feb
(24) |
Mar
(36) |
Apr
(47) |
May
(86) |
Jun
(53) |
Jul
(65) |
Aug
(95) |
Sep
(67) |
Oct
(87) |
Nov
(37) |
Dec
(59) |
| 2016 |
Jan
(86) |
Feb
(80) |
Mar
(47) |
Apr
(74) |
May
(47) |
Jun
(87) |
Jul
(84) |
Aug
(72) |
Sep
(49) |
Oct
(43) |
Nov
(65) |
Dec
(40) |
| 2017 |
Jan
(32) |
Feb
(31) |
Mar
(52) |
Apr
(51) |
May
(48) |
Jun
(35) |
Jul
(51) |
Aug
(28) |
Sep
(54) |
Oct
(44) |
Nov
(48) |
Dec
(19) |
| 2018 |
Jan
(41) |
Feb
(34) |
Mar
(30) |
Apr
(64) |
May
(53) |
Jun
(49) |
Jul
(49) |
Aug
(36) |
Sep
(44) |
Oct
(34) |
Nov
(50) |
Dec
(16) |
| 2019 |
Jan
(40) |
Feb
(21) |
Mar
(28) |
Apr
(19) |
May
(32) |
Jun
(26) |
Jul
(11) |
Aug
(39) |
Sep
(41) |
Oct
(79) |
Nov
(42) |
Dec
(31) |
| 2020 |
Jan
(19) |
Feb
(37) |
Mar
(17) |
Apr
(9) |
May
(10) |
Jun
(14) |
Jul
(52) |
Aug
(41) |
Sep
(38) |
Oct
(24) |
Nov
(9) |
Dec
(13) |
| 2021 |
Jan
(29) |
Feb
(12) |
Mar
(21) |
Apr
(19) |
May
(28) |
Jun
(21) |
Jul
(26) |
Aug
(31) |
Sep
(26) |
Oct
(84) |
Nov
(18) |
Dec
(7) |
| 2022 |
Jan
(25) |
Feb
(11) |
Mar
(35) |
Apr
(25) |
May
(17) |
Jun
(7) |
Jul
(7) |
Aug
(22) |
Sep
(17) |
Oct
(40) |
Nov
(19) |
Dec
(27) |
| 2023 |
Jan
(14) |
Feb
(21) |
Mar
(60) |
Apr
(16) |
May
(26) |
Jun
(18) |
Jul
(14) |
Aug
(19) |
Sep
(13) |
Oct
(19) |
Nov
(23) |
Dec
(13) |
| 2024 |
Jan
(14) |
Feb
(25) |
Mar
(42) |
Apr
(31) |
May
(20) |
Jun
(17) |
Jul
(12) |
Aug
(49) |
Sep
(8) |
Oct
(23) |
Nov
(36) |
Dec
(13) |
| 2025 |
Jan
(35) |
Feb
(27) |
Mar
(56) |
Apr
(20) |
May
(15) |
Jun
(21) |
Jul
(20) |
Aug
(12) |
Sep
(14) |
Oct
(5) |
Nov
(4) |
Dec
(7) |
| 2026 |
Jan
(7) |
Feb
(6) |
Mar
(22) |
Apr
(9) |
May
(2) |
Jun
(6) |
Jul
(7) |
Aug
(4) |
Sep
(9) |
Oct
(3) |
Nov
(2) |
Dec
|
| S | M | T | W | T | F | S |
|---|---|---|---|---|---|---|
|
|
|
1
(2) |
2
(4) |
3
|
4
|
5
|
|
6
|
7
(1) |
8
|
9
|
10
|
11
(1) |
12
|
|
13
|
14
(3) |
15
(2) |
16
|
17
|
18
|
19
|
|
20
|
21
|
22
|
23
|
24
|
25
|
26
|
|
27
|
28
(5) |
29
(2) |
30
|
|
|
|
|
From: Demian K. <dem...@vi...> - 2025-04-29 13:17:37
|
Hello, everyone! The April, 2025 VuFind® newsletter is now available: https://vufind.org/wiki/community:newsletter:2025-04 Additionally, the next Community Call will be held on Tuesday, May 6, 2025 at 9am Eastern Daylight Time (13:00 GMT). Here is the full agenda: 1. Development Planning * Newsletter Highlights * Release 10.1.4 * Discussion: Date Range Handling * Discussion: Non-In-Place Configuration Upgrading * Follow-Up to Past Discussions i. Bootstrap 5 Button Styling * Pull Request / JIRA Ticket Review (release 10.2) * Pull Request / JIRA Ticket Review (release 11.0) 1. Project Tools * Future of Slack 2. Upcoming Conferences 3. Open Q&A / Other Topics? All are welcome to join the free online call, or to suggest additional agenda items. You can find information on joining here: https://vufind.org/wiki/community_call See you next week! - Demian |
|
From: Demian K. <dem...@vi...> - 2025-04-29 12:59:13
|
Hello, everyone - Just a heads-up that I will be traveling to another conference this week and will be out of office from tomorrow, Wednesday, April 30 until Sunday, May 4. I'll return to work on Monday, May 5. As usual, I will be slower to respond to emails and requests during the conference; if others can help out with support while I'm away, I would be grateful. - Demian |
|
From: Demian K. <dem...@vi...> - 2025-04-28 12:33:18
|
Tom, If you're running out of space on your database partition, is it possible this means that you are not running the expire_searches utility to clean up the search table? That can cause significant database growth under heavy traffic. Please let me know if you need more information! - Demian -----Original Message----- From: thomas schwaerzler via VuFind-General <vuf...@li...> Sent: Monday, April 28, 2025 8:28 AM To: vuf...@li... Subject: [EXTERNAL] Re: [VuFind-General] crititcal excepton at laminas-db i should have taken a closer look first. it was a problem with the database - as you already guessed - the partition was full. after a restart of the database everything runs as expected again thanks ere, tom. On 28/04/2025 13:41, Ere Maijala wrote: > Sounds like the database server (MySQL, MariaDB or Postgresql) isn't > running. I'd start by checking the service status. > > --Ere > > On 28.4.2025 13.41, thomas schwaerzler via VuFind-General wrote: >> this morning i found my vufind solr index broken down and in my >> vufind log file lots of messages like this: >> >> >> CRIT (2): Laminas\Db\Adapter\Exception\RuntimeException : Connection >> error ; Laminas\Db\Adapter\Exception\ErrorException : No such file or >> directory >> >> >> i am a bit confused, because i am not aware that anything changed in >> the vufind directory. >> >> >> can anyone give me a hint what goes wrong or where to look? >> >> >> best >> >> tom >> -- Thomas Schwaerzler Digital Services Department University Innsbruck Library 6020 Innsbruck - Innrain 52 - Austria Phone: ++43-(0)512-507-25406 Fax: ++43-(0)512-507-25449 Email: <Tho...@ui...> URL: http://www.uibk.ac.at/ulb/ds OpenPGP key: 0x2BD592C2, Key fingerprint = A17C 26AE FB4B BED6 7907 5DFC 5840 AB43 2BD5 92C2 _______________________________________________ VuFind-General mailing list VuF...@li... https://lists.sourceforge.net/lists/listinfo/vufind-general |
|
From: thomas s. <tho...@ui...> - 2025-04-28 12:28:37
|
i should have taken a closer look first. it was a problem with the database - as you already guessed - the partition was full. after a restart of the database everything runs as expected again thanks ere, tom. On 28/04/2025 13:41, Ere Maijala wrote: > Sounds like the database server (MySQL, MariaDB or Postgresql) isn't > running. I'd start by checking the service status. > > --Ere > > On 28.4.2025 13.41, thomas schwaerzler via VuFind-General wrote: >> this morning i found my vufind solr index broken down and in my vufind >> log file lots of messages like this: >> >> >> CRIT (2): Laminas\Db\Adapter\Exception\RuntimeException : Connection >> error ; Laminas\Db\Adapter\Exception\ErrorException : No such file or >> directory >> >> >> i am a bit confused, because i am not aware that anything changed in >> the vufind directory. >> >> >> can anyone give me a hint what goes wrong or where to look? >> >> >> best >> >> tom >> -- Thomas Schwaerzler Digital Services Department University Innsbruck Library 6020 Innsbruck - Innrain 52 - Austria Phone: ++43-(0)512-507-25406 Fax: ++43-(0)512-507-25449 Email: <Tho...@ui...> URL: http://www.uibk.ac.at/ulb/ds OpenPGP key: 0x2BD592C2, Key fingerprint = A17C 26AE FB4B BED6 7907 5DFC 5840 AB43 2BD5 92C2 |
|
From: Demian K. <dem...@vi...> - 2025-04-28 12:11:33
|
Just a reminder that the WOLFcon call for proposals has been extended to May 16. The conference will be held September 22-25 in Kansas City, Missouri (with virtual participation possible). If anyone wants to present on VuFind(r), please consider submitting. I'll be on-site, so if you want to do a remote presentation, just let me know; I'm willing to help coordinate the room for you. Here's a link to the submission form: https://cfp.sched.com/speaker/6WCTz8koYs/event thanks, Demian |
|
From: Ere M. <ere...@he...> - 2025-04-28 11:41:57
|
Sounds like the database server (MySQL, MariaDB or Postgresql) isn't running. I'd start by checking the service status. --Ere On 28.4.2025 13.41, thomas schwaerzler via VuFind-General wrote: > this morning i found my vufind solr index broken down and in my vufind > log file lots of messages like this: > > > CRIT (2): Laminas\Db\Adapter\Exception\RuntimeException : Connection > error ; Laminas\Db\Adapter\Exception\ErrorException : No such file or > directory > > > i am a bit confused, because i am not aware that anything changed in > the vufind directory. > > > can anyone give me a hint what goes wrong or where to look? > > > best > > tom > -- Ere Maijala Kansalliskirjasto / The National Library of Finland |
|
From: thomas s. <tho...@ui...> - 2025-04-28 11:06:41
|
this morning i found my vufind solr index broken down and in my vufind log file lots of messages like this: CRIT (2): Laminas\Db\Adapter\Exception\RuntimeException : Connection error ; Laminas\Db\Adapter\Exception\ErrorException : No such file or directory i am a bit confused, because i am not aware that anything changed in the vufind directory. can anyone give me a hint what goes wrong or where to look? best tom -- Thomas Schwaerzler Digital Services Department University Innsbruck Library 6020 Innsbruck - Innrain 52 - Austria Phone: ++43-(0)512-507-25406 Fax: ++43-(0)512-507-25449 Email: <Tho...@ui...> URL: http://www.uibk.ac.at/ulb/ds OpenPGP key: 0x2BD592C2, Key fingerprint = A17C 26AE FB4B BED6 7907 5DFC 5840 AB43 2BD5 92C2 |
|
From: Demian K. <dem...@vi...> - 2025-04-15 11:23:59
|
Great, I'm glad there was a solution to the mystery! 😊
- Demian
-----Original Message-----
From: Uwe Steinmann <vu...@st...>
Sent: Tuesday, April 15, 2025 1:00 AM
To: Demian Katz <dem...@vi...>
Cc: vuf...@li...
Subject: Re: [EXTERNAL] [VuFind-General] Setting VUFIND_LOCAL_DIR for util/expire_searches
Thanks Demian,
for sending me in the right direction. I could solve it. After I put some echos into the php code, it was clear that it was not just looking for a config.ini in $VUFIND_LOCAL_DIR but also for /config/vufind/config.ini That second config.ini came from a Parent_Config in the first config.ini and it uses a path ${VUFIND_APPLICATION_PATH}/config/vufind/config.ini
I propably choose $VUFIND_APPLICATION_PATH instead of $VUFIND_HOME, because it is set in the apache configiration.
Since I didn't set $VUFIND_APPLICATION_PATH in the shell it couldn't find that parent config.ini.
Uwe
Am Mon, Apr 14, 2025 at 03:56:00PM +0000 schrieb Demian Katz:
> Uwe,
>
> I can't think of a reason why 9.1.1 and 10.0 would behave differently in this regard. It is normal behavior for the configuration loader to look in $VUFIND_LOCAL_DIR first, then fall back to $VUFIND_HOME if the variable is unset or the local file does not exist. But I'm not sure why the local configuration would fail to load if the variable is set correctly. If the message is literally saying "'/config/vufind/config.ini' doesn't exist" that makes it look like $VUFIND_LOCAL_DIR is somehow set to an empty string, since it appears to be looking in the root directory for the configuration. Is it possible there's something unusual about your setup that is preventing the environment variable from being picked up correctly?
>
> - Demian
>
> -----Original Message-----
> From: Uwe Steinmann <vu...@st...>
> Sent: Monday, April 14, 2025 11:30 AM
> To: vuf...@li...
> Subject: [EXTERNAL] [VuFind-General] Setting VUFIND_LOCAL_DIR for
> util/expire_searches
>
> Hi,
>
> I run a vufind 9.1.1 installation and got stuck with the command
>
> php public/index.php util/expire_searches
>
> If I don't set VUFIND_LOCAL_DIR it tries to read config.ini from
> $VUFIND_HOME/config/vufind/config.ini (I can tell because it uses the
> database configuration from there). Hence, I need to set
> VUFIND_LOCAL_DIR to $VUFIND_HOME/local, but know public/index.php
> complains about a missing config.ini
>
> Next Laminas\ServiceManager\Exception\ServiceNotCreatedException:
> Service with name "config" could not be created. Reason: File
> '/config/vufind/config.ini' doesn't exist or not readable
>
> I wonder why it tries to read config.ini from '/config/vufind/config.ini'
> It appears that setting VUFIND_LOCAL_DIR has some effect, but not how expected.
>
> The same approach in a vufind 10.0 installation works after setting VUFIND_LOCAL_DIR. Is this a know bug?
>
> Uwe
>
>
>
>
> _______________________________________________
> VuFind-General mailing list
> VuF...@li...
> https://lists.sourceforge.net/lists/listinfo/vufind-general
--
MMK GmbH, Fleyer Str. 196, 58097 Hagen
Uwe...@mm...
Tel: 02331 840446 Fax: 02331 843920
|
|
From: Uwe S. <vu...@st...> - 2025-04-15 05:00:44
|
Thanks Demian,
for sending me in the right direction. I could solve it. After
I put some echos into the php code, it was clear that it was not
just looking for a config.ini in $VUFIND_LOCAL_DIR but also for
/config/vufind/config.ini That second config.ini came from
a Parent_Config in the first config.ini and it uses a path
${VUFIND_APPLICATION_PATH}/config/vufind/config.ini
I propably choose $VUFIND_APPLICATION_PATH instead of
$VUFIND_HOME, because it is set in the apache configiration.
Since I didn't set $VUFIND_APPLICATION_PATH in the shell it
couldn't find that parent config.ini.
Uwe
Am Mon, Apr 14, 2025 at 03:56:00PM +0000 schrieb Demian Katz:
> Uwe,
>
> I can't think of a reason why 9.1.1 and 10.0 would behave differently in this regard. It is normal behavior for the configuration loader to look in $VUFIND_LOCAL_DIR first, then fall back to $VUFIND_HOME if the variable is unset or the local file does not exist. But I'm not sure why the local configuration would fail to load if the variable is set correctly. If the message is literally saying "'/config/vufind/config.ini' doesn't exist" that makes it look like $VUFIND_LOCAL_DIR is somehow set to an empty string, since it appears to be looking in the root directory for the configuration. Is it possible there's something unusual about your setup that is preventing the environment variable from being picked up correctly?
>
> - Demian
>
> -----Original Message-----
> From: Uwe Steinmann <vu...@st...>
> Sent: Monday, April 14, 2025 11:30 AM
> To: vuf...@li...
> Subject: [EXTERNAL] [VuFind-General] Setting VUFIND_LOCAL_DIR for util/expire_searches
>
> Hi,
>
> I run a vufind 9.1.1 installation and got stuck with the command
>
> php public/index.php util/expire_searches
>
> If I don't set VUFIND_LOCAL_DIR it tries to read config.ini from $VUFIND_HOME/config/vufind/config.ini (I can tell because it uses the database configuration from there). Hence, I need to set VUFIND_LOCAL_DIR to $VUFIND_HOME/local, but know public/index.php complains about a missing config.ini
>
> Next Laminas\ServiceManager\Exception\ServiceNotCreatedException: Service with name "config" could not be created. Reason: File '/config/vufind/config.ini' doesn't exist or not readable
>
> I wonder why it tries to read config.ini from '/config/vufind/config.ini'
> It appears that setting VUFIND_LOCAL_DIR has some effect, but not how expected.
>
> The same approach in a vufind 10.0 installation works after setting VUFIND_LOCAL_DIR. Is this a know bug?
>
> Uwe
>
>
>
>
> _______________________________________________
> VuFind-General mailing list
> VuF...@li...
> https://lists.sourceforge.net/lists/listinfo/vufind-general
--
MMK GmbH, Fleyer Str. 196, 58097 Hagen
Uwe...@mm...
Tel: 02331 840446 Fax: 02331 843920
|
|
From: Demian K. <dem...@vi...> - 2025-04-14 15:56:28
|
Uwe, I can't think of a reason why 9.1.1 and 10.0 would behave differently in this regard. It is normal behavior for the configuration loader to look in $VUFIND_LOCAL_DIR first, then fall back to $VUFIND_HOME if the variable is unset or the local file does not exist. But I'm not sure why the local configuration would fail to load if the variable is set correctly. If the message is literally saying "'/config/vufind/config.ini' doesn't exist" that makes it look like $VUFIND_LOCAL_DIR is somehow set to an empty string, since it appears to be looking in the root directory for the configuration. Is it possible there's something unusual about your setup that is preventing the environment variable from being picked up correctly? - Demian -----Original Message----- From: Uwe Steinmann <vu...@st...> Sent: Monday, April 14, 2025 11:30 AM To: vuf...@li... Subject: [EXTERNAL] [VuFind-General] Setting VUFIND_LOCAL_DIR for util/expire_searches Hi, I run a vufind 9.1.1 installation and got stuck with the command php public/index.php util/expire_searches If I don't set VUFIND_LOCAL_DIR it tries to read config.ini from $VUFIND_HOME/config/vufind/config.ini (I can tell because it uses the database configuration from there). Hence, I need to set VUFIND_LOCAL_DIR to $VUFIND_HOME/local, but know public/index.php complains about a missing config.ini Next Laminas\ServiceManager\Exception\ServiceNotCreatedException: Service with name "config" could not be created. Reason: File '/config/vufind/config.ini' doesn't exist or not readable I wonder why it tries to read config.ini from '/config/vufind/config.ini' It appears that setting VUFIND_LOCAL_DIR has some effect, but not how expected. The same approach in a vufind 10.0 installation works after setting VUFIND_LOCAL_DIR. Is this a know bug? Uwe _______________________________________________ VuFind-General mailing list VuF...@li... https://lists.sourceforge.net/lists/listinfo/vufind-general |
|
From: Uwe S. <vu...@st...> - 2025-04-14 15:48:27
|
Hi, I run a vufind 9.1.1 installation and got stuck with the command php public/index.php util/expire_searches If I don't set VUFIND_LOCAL_DIR it tries to read config.ini from $VUFIND_HOME/config/vufind/config.ini (I can tell because it uses the database configuration from there). Hence, I need to set VUFIND_LOCAL_DIR to $VUFIND_HOME/local, but know public/index.php complains about a missing config.ini Next Laminas\ServiceManager\Exception\ServiceNotCreatedException: Service with name "config" could not be created. Reason: File '/config/vufind/config.ini' doesn't exist or not readable I wonder why it tries to read config.ini from '/config/vufind/config.ini' It appears that setting VUFIND_LOCAL_DIR has some effect, but not how expected. The same approach in a vufind 10.0 installation works after setting VUFIND_LOCAL_DIR. Is this a know bug? Uwe |
|
From: Demian K. <dem...@vi...> - 2025-04-14 12:39:15
|
Hello, everyone! Just a heads-up that I will be out of the office starting April 15 for a conference. I will return to work next week, on Tuesday, April 22. As always, I will try to monitor email while away, but my response times may be slower than normal. If others can help with support requests while I am away, I would be grateful for any assistance you can offer! thanks, Demian |
|
From: Ere M. <ere...@he...> - 2025-04-11 18:21:16
|
True, I missed those. So it's not a green field, but it could use some improvement. And more obvious logic so that I could find it. :) I'm not suggesting this needs to be changed right now, but it could at least be something to keep in mind for the future. --Ere On 31.3.2025 15.59, Demian Katz wrote: > Ere, > > You make a good point, but I'm not sure if it's true that the code doesn't use author_role/author2_role. There's some dynamic method-name abstraction in getAuthorDataFields (see https://github.com/vufind-org/vufind/blob/dev/module/VuFind/src/VuFind/RecordDriver/DefaultRecord.php#L180) which obscures things, but I think those fields are being used to retrieve data for display there via the getPrimaryAuthorsRoles and getSecondaryAuthorsRoles methods. I can't find any equivalent MARC-based logic doing the same thing, so my conclusion is that the Solr data is in fact the source of the role displays (which come out through getDeduplicatedAuthors in the record driver). It's possible I'm missing something, though -- I didn't write this code, and it's been years since I reviewed it, so my understanding is fairly shallow. > > > I'm certainly open to making this more reliable and easier to understand, since the current implementation is confusing and apparently has bugs somewhere. But I don't think we have a totally green field to build something on -- we'll need to think about a clean migration path if we change things. > > - Demian > > -----Original Message----- > From: Ere Maijala <ere...@he...> > Sent: Monday, March 31, 2025 6:34 AM > To: vuf...@li... > Subject: Re: [VuFind-General] [EXTERNAL] Multiple subfield 'e' relator codes in one MARC fieled > > Hi, > > May I also use this opportunity to start a discussion of whether the author_role or author2_role fields makes sense at all? :) Since there's no way to search for author's name and the associated role using two separate Solr fields, it feels to me like we'd be better of having a field such as author_with_role. Also, while the current index structure maintains 1:1 relationship between the author and author_role fields so that they can be used for display, VuFind already gets author names from MARC (and doesn't use author_role or author2_role). And if we were to use docvalues in Solr, they wouldn't maintain the order or duplicate values for a field. > > --Ere > > On 31.3.2025 10.40, Uwe Reh wrote: >> Hi, >> >> This is on of the problems that made us to use SOLR only for searching. >> The title representation is generated from MARC (fullrecord) only. >> An other advantage is that we can modify the title representation >> anytime (w/o reindexing). >> >> Uwe >> >> >> Am 28.03.25 um 18:37 schrieb Demian Katz via VuFind-General: >>> Peter, >>> >>> Yes, your diagnosis sounds correct to me. >>> >>> It's unfortunate that the limitations of Solr make "parallel arrays" >>> the most practical way to associate data (short of creating our own >>> custom delimiters and extra processing). It seems likely that there >>> is some problem with the logic in >>> https://git/ >>> hub.com%2Fvufind-org%2Fvufind%2Fblob%2Fdev%2F&data=05%7C02%7Cdemian.k >>> atz%40villanova.edu%7Cdfa892179a2645ff5de708dd703f9478%7C765a8de5cf94 >>> 44f09cafae5bf8cfa366%7C0%7C0%7C638790140579795202%7CUnknown%7CTWFpbGZ >>> sb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFO >>> IjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=%2FtJ9sdNw9UQYiQBLIuG >>> %2FhhxJh7H9J8Wc2barlfk9BqI%3D&reserved=0 >>> import/index_java/src/org/vufind/index/CreatorTools.java <https:// >>> github.com/vufind-org/vufind/blob/dev/import/index_java/src/org/vufin >>> d/ index/CreatorTools.java>. That being said, though, a quick browse >>> through the Java code suggests that it's at least /supposed/ to >>> duplicate author names when they have multiple associated relators, >>> in order to maintain the parallel association. So it's not a total >>> oversight - maybe an edge case that's going wrong for some unknown >>> reason. >>> >>> Have you looked at the raw Solr data? It might be helpful to confirm >>> that the problem is related to the author and author_role fields >>> being out of alignment with each other, or if it's something else - >>> then we can confirm whether it's an indexing problem, or something >>> that's happening later in the presentation. >>> >>> - Demian >>> >>> *From:*Peter Murray <pe...@in...> >>> *Sent:* Friday, March 28, 2025 12:25 PM >>> *To:* VuFind-General <vuf...@li...> >>> *Subject:* [EXTERNAL] [VuFind-General] Multiple subfield 'e' relator >>> codes in one MARC fieled >>> >>> I'm seeing an issue with the reconstruction of second author fields >>> with second author relator fields. VuFind doesn't seem to handle the >>> case where there are two subfield 'e' relator values on a 700 record: >>> >>> <marc:datafield tag="700" ind1="1" ind2=" "> >>> <marc:subfield code="a">Glazer, Jonathan,</marc:subfield> >>> <marc:subfield code="e">screenwriter,</marc:subfield> >>> <marc:subfield code="e">film director.</marc:subfield> >>> </marc:datafield> >>> >>> The problem seems to be when the record is broken up into its >>> subfield 'a' and subfield 'e' fields in the Solr index, it loses >>> track that two of the array entries in the author2_role Solr field >>> belong to one of the entries in the author2 Solr field. The >>> |author2| and |author2_role| fields in the Solr schema are >>> populated as separate arrays, and there is no explicit mechanism to >>> maintain the association between entries in these arrays. >>> >>> If I take the 700 fields with the two subfield 'e' values and break >>> them up into separate 700 fields, the display in VuFind reflects the >>> MARC record. >>> >>> <marc:datafield tag="700" ind1="1" ind2=" "> >>> <marc:subfield code="a">Glazer, Jonathan,</marc:subfield> >>> <marc:subfield code="e">screenwriter,</marc:subfield> >>> </marc:datafield> >>> <marc:datafield tag="700" ind1="1" ind2=" "> >>> <marc:subfield code="a">Glazer, Jonathan,</marc:subfield> >>> <marc:subfield code="e">film director.</marc:subfield> >>> </marc:datafield> >>> >>> Have I analyzed this correctly? >>> >>> Peter >>> >>> -- >>> >>> Peter Murray >>> >>> Open Source Community Advocate >>> >>> Index Data, LLC >>> >>> >>> >>> _______________________________________________ >>> VuFind-General mailing list >>> VuF...@li... >>> https://lis/ >>> ts.sourceforge.net%2Flists%2Flistinfo%2Fvufind-general&data=05%7C02%7 >>> Cdemian.katz%40villanova.edu%7Cdfa892179a2645ff5de708dd703f9478%7C765 >>> a8de5cf9444f09cafae5bf8cfa366%7C0%7C0%7C638790140579816080%7CUnknown% >>> 7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4 >>> zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=ECguFy535TZJ >>> XphYHK1qdgGdvA86XHoJOXG9aNdkZeE%3D&reserved=0 >> >> >> _______________________________________________ >> VuFind-General mailing list >> VuF...@li... >> https://list/ >> s.sourceforge.net%2Flists%2Flistinfo%2Fvufind-general&data=05%7C02%7Cd >> emian.katz%40villanova.edu%7Cdfa892179a2645ff5de708dd703f9478%7C765a8d >> e5cf9444f09cafae5bf8cfa366%7C0%7C0%7C638790140579832154%7CUnknown%7CTW >> FpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIs >> IkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=QTE5JcMiail6mBccP% >> 2BGmM2SS7XVLcBAogu6AHtIbcmY%3D&reserved=0 > -- > Ere Maijala > Kansalliskirjasto / The National Library of Finland > > > > _______________________________________________ > VuFind-General mailing list > VuF...@li... > https://lists.sourceforge.net/lists/listinfo/vufind-general -- Ere Maijala Kansalliskirjasto / The National Library of Finland |
|
From: Demian K. <dem...@vi...> - 2025-04-07 13:31:49
|
VuFind® 10.1.4 was released today. This is a minor release fixing a small number of bugs; the most significant issue was a bug introduced by release 10.1.2 which could cause warnings to accumulate in Apache or PHP logs. All users of earlier 10.1.x releases are encouraged to upgrade. It is likely there will be no further releases in the 10.1 series (barring discovery of a serious bug), as work on 10.2.0 is now underway. You can download the release here: https://vufind.org/vufind/downloads.html Please let me know if you have questions, problems, or concerns! thanks, Demian |
|
From: Demian K. <dem...@vi...> - 2025-04-02 16:14:01
|
Peter,
What class is your local custom class extending? My guess would be \VuFind\RecordDriver\SolrMarc. Note that MarcBasicTrait is NOT included in that class – MarcBasicTrait was designed to pull fields from MARC records that are normally pulled from the Solr index (that’s what’s “basic” about it – MarcAdvancedTrait is designed to augment a driver with MARC data that can’t be obtained from Solr, so that’s the only one included in the SolrMarc driver). MarcBasicTrait is only currently used in non-Solr-based drivers like ProQuestFSG. If you want to use the author methods from MarcBasicTrait for any reason, it should be possible to just import the ones you want into your local subclass – but in theory, the Solr-based way should be the intended option as the code is currently written.
- Demian
From: Peter Murray <pe...@in...>
Sent: Wednesday, April 2, 2025 11:14 AM
To: VuFind-General <vuf...@li...>
Subject: [EXTERNAL] Re: [VuFind-General] Multiple subfield 'e' relator codes in one MARC fieled
Okay, I dug into this further, and I have a reality check that I want to propose:
* For MARC records, the current RecordDataFormatter helper functions pull information from the Solr core for display. This is unlike other fields, which read information from the MARC record for display.
What is leading me to this conclusion is how the overriding getSecondaryAuthors() in a local module is messing up display of relator codes on the record. With the default display spec line for authors:
$spec->setMultiLine(
'Authors',
'getDeduplicatedAuthors',
$this->getAuthorFunction()
);
...what is getting called is the getSecondaryAuthors() from the DefaultRecord class, not the one from MarcBasicTrait trait. This is the stack trace when I'm in the debugger and pausing at the point where getSecondaryAuthors() is called:
VuFind\RecordDriver\DefaultRecord->getSecondaryAuthors (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/RecordDriver/DefaultRecord.php:1195)
VuFind\RecordDriver\AbstractBase->tryMethod (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/RecordDriver/AbstractBase.php:496)
VuFind\RecordDriver\DefaultRecord->getAuthorDataFields (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/RecordDriver/DefaultRecord.php:186)
VuFind\RecordDriver\DefaultRecord->getDeduplicatedAuthors (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/RecordDriver/DefaultRecord.php:460)
VuFind\RecordDriver\AbstractBase->tryMethod (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/RecordDriver/AbstractBase.php:496)
VuFind\View\Helper\Root\RecordDataFormatter->extractData (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/View/Helper/Root/RecordDataFormatter.php:341)
VuFind\View\Helper\Root\RecordDataFormatter->getData (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/View/Helper/Root/RecordDataFormatter.php:212)
include (/usr/local/vufind-lafayette/themes/bootstrap3/templates/RecordDriver/DefaultRecord/core-fields.phtml:3)
Laminas\View\Renderer\PhpRenderer->render (/usr/local/vufind-lafayette/vendor/laminas/laminas-view/src/Renderer/PhpRenderer.php:526)
VuFind\View\Helper\Root\Record->renderClassTemplate (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/View/Helper/Root/ClassBasedTemplateRendererTrait.php:136)
VuFind\View\Helper\Root\Record->renderTemplate (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/View/Helper/Root/Record.php:145)
include (/usr/local/vufind-lafayette/themes/lafayette-bootstrap3/templates/RecordDriver/DefaultRecord/core.phtml:92)
Laminas\View\Renderer\PhpRenderer->render (/usr/local/vufind-lafayette/vendor/laminas/laminas-view/src/Renderer/PhpRenderer.php:526)
VuFind\View\Helper\Root\Record->renderClassTemplate (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/View/Helper/Root/ClassBasedTemplateRendererTrait.php:136)
VuFind\View\Helper\Root\Record->renderTemplate (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/View/Helper/Root/Record.php:145)
VuFind\View\Helper\Root\Record->getCoreMetadata (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/View/Helper/Root/Record.php:178)
include (/usr/local/vufind-lafayette/themes/bootstrap3/templates/record/view.phtml:46)
Laminas\View\Renderer\PhpRenderer->render (/usr/local/vufind-lafayette/vendor/laminas/laminas-view/src/Renderer/PhpRenderer.php:526)
Laminas\View\View->render (/usr/local/vufind-lafayette/vendor/laminas/laminas-view/src/View.php:194)
Laminas\View\View->renderChildren (/usr/local/vufind-lafayette/vendor/laminas/laminas-view/src/View.php:222)
In the DefaultRecord class, getSecondaryAuthors() is pulling information from the Solr core. In the MarcBasicTrait, it is getting it from the MARC record. So when I override getSecondaryAuthors() in the local module to add additonal subfields, what gets returned to DefaultRecord::getAuthorDataFields() is not what that function is expecting. In order to get what I'd like—multiple relator codes displayed for a 7xx entry _plus_ additional subfields displayed—I think I'd need to provide my own getDeduplicatedAuthors() and getAuthorDataFields() in a SolrMarc class context that overrides what is happening in the DefaultRecord class.
Does this sound right?
Peter
--
Peter Murray
Open Source Community Advocate
Index Data, LLC
On Mar 31, 2025 at 12:42 PM -0400, Peter Murray <pe...@in...<mailto:pe...@in...>>, wrote:
Thanks for the reply, Demian. That gave me a path start looking, and — wow, oh wow — I ended up in a place I didn't expect. Somewhere along the way, for this library we had implemented a custom getSecondaryAuthors() to add additional subfields (the 700 ‡t, ‡m, and ‡n), and that change had caused the problem. I'm still not sure why reverting that code solved the issue, and I'm walking away from this issue understanding even less about how the primary, secondary, and corporate author displays are constructed.
I think Uwe is onto something about simplifying how all of this is calculated, but I'm even less sure now of what to recommend than when I started.
Peter
On Fri, Mar 28, 2025 at 1:37 PM Demian Katz <dem...@vi...<mailto:dem...@vi...>> wrote:
Peter,
Yes, your diagnosis sounds correct to me.
It’s unfortunate that the limitations of Solr make “parallel arrays” the most practical way to associate data (short of creating our own custom delimiters and extra processing). It seems likely that there is some problem with the logic in https://github.com/vufind-org/vufind/blob/dev/import/index_java/src/org/vufind/index/CreatorTools.java. That being said, though, a quick browse through the Java code suggests that it’s at least supposed to duplicate author names when they have multiple associated relators, in order to maintain the parallel association. So it’s not a total oversight – maybe an edge case that’s going wrong for some unknown reason.
Have you looked at the raw Solr data? It might be helpful to confirm that the problem is related to the author and author_role fields being out of alignment with each other, or if it’s something else – then we can confirm whether it’s an indexing problem, or something that’s happening later in the presentation.
- Demian
From: Peter Murray <pe...@in...<mailto:pe...@in...>>
Sent: Friday, March 28, 2025 12:25 PM
To: VuFind-General <vuf...@li...<mailto:vuf...@li...>>
Subject: [EXTERNAL] [VuFind-General] Multiple subfield 'e' relator codes in one MARC fieled
I'm seeing an issue with the reconstruction of second author fields with second author relator fields. VuFind doesn't seem to handle the case where there are two subfield 'e' relator values on a 700 record:
<marc:datafield tag="700" ind1="1" ind2=" ">
<marc:subfield code="a">Glazer, Jonathan,</marc:subfield>
<marc:subfield code="e">screenwriter,</marc:subfield>
<marc:subfield code="e">film director.</marc:subfield>
</marc:datafield>
The problem seems to be when the record is broken up into its subfield 'a' and subfield 'e' fields in the Solr index, it loses track that two of the array entries in the author2_role Solr field belong to one of the entries in the author2 Solr field. The author2 and author2_role fields in the Solr schema are populated as separate arrays, and there is no explicit mechanism to maintain the association between entries in these arrays.
If I take the 700 fields with the two subfield 'e' values and break them up into separate 700 fields, the display in VuFind reflects the MARC record.
<marc:datafield tag="700" ind1="1" ind2=" ">
<marc:subfield code="a">Glazer, Jonathan,</marc:subfield>
<marc:subfield code="e">screenwriter,</marc:subfield>
</marc:datafield>
<marc:datafield tag="700" ind1="1" ind2=" ">
<marc:subfield code="a">Glazer, Jonathan,</marc:subfield>
<marc:subfield code="e">film director.</marc:subfield>
</marc:datafield>
Have I analyzed this correctly?
Peter
--
Peter Murray
Open Source Community Advocate
Index Data, LLC
--
Peter Murray
Open Source Community Advocate
Index Data, LLC
|
|
From: Demian K. <dem...@vi...> - 2025-04-02 16:10:07
|
Yes, I think so… though my understanding of getDeduplicatedAuthors is limited (I didn’t write it and haven’t recently reviewed it) so there may be nuance I’m overlooking. Ultimately I don’t think it matters if some data comes from MARC and some comes from Solr as long as the logic is in sync so that the values align properly. It’s probably easier to do this using a consistent method across the board, but it’s not necessarily REQUIRED. 😊 - Demian From: Peter Murray <pe...@in...> Sent: Wednesday, April 2, 2025 11:51 AM To: VuFind-General <vuf...@li...> Subject: [EXTERNAL] Re: [VuFind-General] Multiple subfield 'e' relator codes in one MARC fieled Oh! That's the architectural reasoning behind that naming scheme! Yes, extending \VuFind\RecordDriver\SolrMarc. And, I think your description confirms my theory—the design is working as intended to get author data from the Solr index rather than the MARC record. So if I want to add additional subfields from the MARC record to the display in VuFind (say, the 700 ‡r<https://www.loc.gov/marc/bibliographic/bd700.html>), I'll either need to: 1. Modify `marc_local.properties` to add the fields (although I'm not sure which field I'd use); or 2. Write a new getDeduplicatedAuthors() that uses information from the MARC record instead of the Solr index. Correct so far? Peter -- Peter Murray Open Source Community Advocate Index Data, LLC On Apr 2, 2025 at 11:40 AM -0400, Demian Katz <dem...@vi...<mailto:dem...@vi...>>, wrote: Peter, What class is your local custom class extending? My guess would be \VuFind\RecordDriver\SolrMarc. Note that MarcBasicTrait is NOT included in that class – MarcBasicTrait was designed to pull fields from MARC records that are normally pulled from the Solr index (that’s what’s “basic” about it – MarcAdvancedTrait is designed to augment a driver with MARC data that can’t be obtained from Solr, so that’s the only one included in the SolrMarc driver). MarcBasicTrait is only currently used in non-Solr-based drivers like ProQuestFSG. If you want to use the author methods from MarcBasicTrait for any reason, it should be possible to just import the ones you want into your local subclass – but in theory, the Solr-based way should be the intended option as the code is currently written. - Demian |
|
From: Peter M. <pe...@in...> - 2025-04-02 15:51:30
|
Oh! That's the architectural reasoning behind that naming scheme! Yes, extending \VuFind\RecordDriver\SolrMarc. And, I think your description confirms my theory—the design is working as intended to get author data from the Solr index rather than the MARC record. So if I want to add additional subfields from the MARC record to the display in VuFind (say, the 700 ‡r), I'll either need to: 1. Modify `marc_local.properties` to add the fields (although I'm not sure which field I'd use); or 2. Write a new getDeduplicatedAuthors() that uses information from the MARC record instead of the Solr index. Correct so far? Peter -- Peter Murray Open Source Community Advocate Index Data, LLC On Apr 2, 2025 at 11:40 AM -0400, Demian Katz <dem...@vi...>, wrote: Peter, What class is your local custom class extending? My guess would be \VuFind\RecordDriver\SolrMarc. Note that MarcBasicTrait is NOT included in that class – MarcBasicTrait was designed to pull fields from MARC records that are normally pulled from the Solr index (that’s what’s “basic” about it – MarcAdvancedTrait is designed to augment a driver with MARC data that can’t be obtained from Solr, so that’s the only one included in the SolrMarc driver). MarcBasicTrait is only currently used in non-Solr-based drivers like ProQuestFSG. If you want to use the author methods from MarcBasicTrait for any reason, it should be possible to just import the ones you want into your local subclass – but in theory, the Solr-based way should be the intended option as the code is currently written. - Demian |
|
From: Peter M. <pe...@in...> - 2025-04-02 15:14:06
|
Okay, I dug into this further, and I have a reality check that I want to propose: • For MARC records, the current RecordDataFormatter helper functions pull information from the Solr core for display. This is unlike other fields, which read information from the MARC record for display. What is leading me to this conclusion is how the overriding getSecondaryAuthors() in a local module is messing up display of relator codes on the record. With the default display spec line for authors: $spec->setMultiLine( 'Authors', 'getDeduplicatedAuthors', $this->getAuthorFunction() ); ...what is getting called is the getSecondaryAuthors() from the DefaultRecord class, not the one from MarcBasicTrait trait. This is the stack trace when I'm in the debugger and pausing at the point where getSecondaryAuthors() is called: VuFind\RecordDriver\DefaultRecord->getSecondaryAuthors (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/RecordDriver/DefaultRecord.php:1195) VuFind\RecordDriver\AbstractBase->tryMethod (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/RecordDriver/AbstractBase.php:496) VuFind\RecordDriver\DefaultRecord->getAuthorDataFields (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/RecordDriver/DefaultRecord.php:186) VuFind\RecordDriver\DefaultRecord->getDeduplicatedAuthors (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/RecordDriver/DefaultRecord.php:460) VuFind\RecordDriver\AbstractBase->tryMethod (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/RecordDriver/AbstractBase.php:496) VuFind\View\Helper\Root\RecordDataFormatter->extractData (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/View/Helper/Root/RecordDataFormatter.php:341) VuFind\View\Helper\Root\RecordDataFormatter->getData (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/View/Helper/Root/RecordDataFormatter.php:212) include (/usr/local/vufind-lafayette/themes/bootstrap3/templates/RecordDriver/DefaultRecord/core-fields.phtml:3) Laminas\View\Renderer\PhpRenderer->render (/usr/local/vufind-lafayette/vendor/laminas/laminas-view/src/Renderer/PhpRenderer.php:526) VuFind\View\Helper\Root\Record->renderClassTemplate (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/View/Helper/Root/ClassBasedTemplateRendererTrait.php:136) VuFind\View\Helper\Root\Record->renderTemplate (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/View/Helper/Root/Record.php:145) include (/usr/local/vufind-lafayette/themes/lafayette-bootstrap3/templates/RecordDriver/DefaultRecord/core.phtml:92) Laminas\View\Renderer\PhpRenderer->render (/usr/local/vufind-lafayette/vendor/laminas/laminas-view/src/Renderer/PhpRenderer.php:526) VuFind\View\Helper\Root\Record->renderClassTemplate (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/View/Helper/Root/ClassBasedTemplateRendererTrait.php:136) VuFind\View\Helper\Root\Record->renderTemplate (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/View/Helper/Root/Record.php:145) VuFind\View\Helper\Root\Record->getCoreMetadata (/usr/local/vufind-lafayette/module/VuFind/src/VuFind/View/Helper/Root/Record.php:178) include (/usr/local/vufind-lafayette/themes/bootstrap3/templates/record/view.phtml:46) Laminas\View\Renderer\PhpRenderer->render (/usr/local/vufind-lafayette/vendor/laminas/laminas-view/src/Renderer/PhpRenderer.php:526) Laminas\View\View->render (/usr/local/vufind-lafayette/vendor/laminas/laminas-view/src/View.php:194) Laminas\View\View->renderChildren (/usr/local/vufind-lafayette/vendor/laminas/laminas-view/src/View.php:222) In the DefaultRecord class, getSecondaryAuthors() is pulling information from the Solr core. In the MarcBasicTrait, it is getting it from the MARC record. So when I override getSecondaryAuthors() in the local module to add additonal subfields, what gets returned to DefaultRecord::getAuthorDataFields() is not what that function is expecting. In order to get what I'd like—multiple relator codes displayed for a 7xx entry _plus_ additional subfields displayed—I think I'd need to provide my own getDeduplicatedAuthors() and getAuthorDataFields() in a SolrMarc class context that overrides what is happening in the DefaultRecord class. Does this sound right? Peter -- Peter Murray Open Source Community Advocate Index Data, LLC On Mar 31, 2025 at 12:42 PM -0400, Peter Murray <pe...@in...>, wrote: > Thanks for the reply, Demian. That gave me a path start looking, and — wow, oh wow — I ended up in a place I didn't expect. Somewhere along the way, for this library we had implemented a custom getSecondaryAuthors() to add additional subfields (the 700 ‡t, ‡m, and ‡n), and that change had caused the problem. I'm still not sure why reverting that code solved the issue, and I'm walking away from this issue understanding even less about how the primary, secondary, and corporate author displays are constructed. > > I think Uwe is onto something about simplifying how all of this is calculated, but I'm even less sure now of what to recommend than when I started. > > > Peter > > On Fri, Mar 28, 2025 at 1:37 PM Demian Katz <dem...@vi...> wrote: > > Peter, > > > > Yes, your diagnosis sounds correct to me. > > > > It’s unfortunate that the limitations of Solr make “parallel arrays” the most practical way to associate data (short of creating our own custom delimiters and extra processing). It seems likely that there is some problem with the logic in https://github.com/vufind-org/vufind/blob/dev/import/index_java/src/org/vufind/index/CreatorTools.java. That being said, though, a quick browse through the Java code suggests that it’s at least supposed to duplicate author names when they have multiple associated relators, in order to maintain the parallel association. So it’s not a total oversight – maybe an edge case that’s going wrong for some unknown reason. > > > > Have you looked at the raw Solr data? It might be helpful to confirm that the problem is related to the author and author_role fields being out of alignment with each other, or if it’s something else – then we can confirm whether it’s an indexing problem, or something that’s happening later in the presentation. > > > > - Demian > > > > From: Peter Murray <pe...@in...> > > Sent: Friday, March 28, 2025 12:25 PM > > To: VuFind-General <vuf...@li...> > > Subject: [EXTERNAL] [VuFind-General] Multiple subfield 'e' relator codes in one MARC fieled > > > > I'm seeing an issue with the reconstruction of second author fields with second author relator fields. VuFind doesn't seem to handle the case where there are two subfield 'e' relator values on a 700 record: > > > > <marc:datafield tag="700" ind1="1" ind2=" "> > > <marc:subfield code="a">Glazer, Jonathan,</marc:subfield> > > <marc:subfield code="e">screenwriter,</marc:subfield> > > <marc:subfield code="e">film director.</marc:subfield> > > </marc:datafield> > > > > The problem seems to be when the record is broken up into its subfield 'a' and subfield 'e' fields in the Solr index, it loses track that two of the array entries in the author2_role Solr field belong to one of the entries in the author2 Solr field. The author2 and author2_role fields in the Solr schema are populated as separate arrays, and there is no explicit mechanism to maintain the association between entries in these arrays. > > > > If I take the 700 fields with the two subfield 'e' values and break them up into separate 700 fields, the display in VuFind reflects the MARC record. > > > > <marc:datafield tag="700" ind1="1" ind2=" "> > > <marc:subfield code="a">Glazer, Jonathan,</marc:subfield> > > <marc:subfield code="e">screenwriter,</marc:subfield> > > </marc:datafield> > > <marc:datafield tag="700" ind1="1" ind2=" "> > > <marc:subfield code="a">Glazer, Jonathan,</marc:subfield> > > <marc:subfield code="e">film director.</marc:subfield> > > </marc:datafield> > > > > Have I analyzed this correctly? > > > > Peter > > -- > > Peter Murray > > Open Source Community Advocate > > Index Data, LLC > > > -- > Peter Murray > Open Source Community Advocate > Index Data, LLC |
|
From: dago s. <dag...@gm...> - 2025-04-01 22:24:05
|
Hi Demian, Thanks, it's work. had a typo (disable instead of disabled) The print button is fine, it doesn't bother me. Thanks! On Mon, Mar 31, 2025 at 1:15 PM Demian Katz <dem...@vi...> wrote: > Hello, > > > > To get rid of “Text This,” set the “sms” setting to “disabled” in > config.ini. > > > > There is not currently a configuration setting to disable the print > button. However, you can customize the toolbar template in your local theme > to get rid of it. See: > > > > > https://github.com/vufind-org/vufind/blob/release-10.0/themes/bootstrap5/templates/RecordDriver/DefaultRecord/toolbar.phtml#L34-L39 > > > > I hope this helps! > > > > > > - Demian > > > > *From:* dago salas <dag...@gm...> > *Sent:* Monday, March 31, 2025 3:11 PM > *To:* vuf...@li... > *Subject:* [EXTERNAL] [VuFind-General] remove the "text this" and "print" > buttons? > > > > Hi friends, how can I remove the "text this" and "print" buttons? My > version is 10.0.1 > > > > -- > > Dagoberto Salas > dag...@gm... > -- Dagoberto Salas dag...@gm... |
|
From: Demian K. <dem...@vi...> - 2025-04-01 13:51:09
|
Here are the minutes from today's VuFind(r) Community Call: https://vufind.org/wiki/community_call:minutes20250401 The next call will be Tuesday, May 6, 2025 at 9am Eastern Daylight Time (13:00 GMT). - Demian |