pazpar2-moved-to-github.git
13 years agoFlush HTTP recording file
Adam Dickmeiss [Wed, 19 Oct 2011 09:31:38 +0000 (11:31 +0200)]
Flush HTTP recording file

13 years agoHTTP recording saves both requests and responses
Adam Dickmeiss [Wed, 19 Oct 2011 08:32:24 +0000 (10:32 +0200)]
HTTP recording saves both requests and responses

13 years agoAdd support for CF data element "id".
Mike Taylor [Tue, 18 Oct 2011 08:26:43 +0000 (09:26 +0100)]
Add support for CF data element "id".
Gets translated into pz:metadata[@type="id"].
May be needed for bug #4713, persistent identifiers.
(And will be useful even if we don't do that for EBSCO.)

13 years agoAvoid work on bad metadata (0 prt ref)
Adam Dickmeiss [Fri, 14 Oct 2011 10:30:21 +0000 (12:30 +0200)]
Avoid work on bad metadata (0 prt ref)

Problem introduced in commit 9d9593.

13 years agoRecord id is leaded "content: " or "position: "
Adam Dickmeiss [Wed, 12 Oct 2011 14:18:45 +0000 (16:18 +0200)]
Record id is leaded "content: " or "position: "

This is to indicate whether a record ID is based of the generated
mergekey (typically title and other) or depending on the result set
position.

13 years agoChanges
Adam Dickmeiss [Wed, 12 Oct 2011 14:18:38 +0000 (16:18 +0200)]
Changes

13 years agosort: only skip "identical" records from same target
Adam Dickmeiss [Wed, 12 Oct 2011 11:15:47 +0000 (13:15 +0200)]
sort: only skip "identical" records from same target

13 years agoDocument pz:sortmap
Adam Dickmeiss [Wed, 12 Oct 2011 11:11:33 +0000 (13:11 +0200)]
Document pz:sortmap

13 years agoHonor block=1 for show when re-searching (sort)
Adam Dickmeiss [Wed, 12 Oct 2011 10:40:21 +0000 (12:40 +0200)]
Honor block=1 for show when re-searching (sort)

Note that the condition for block=1 has changed. It will block if
any client has more records to be fetched. Previously: block if
no records at all have been fetched.

13 years agoUpdate test results for sort test
Adam Dickmeiss [Wed, 12 Oct 2011 09:34:23 +0000 (11:34 +0200)]
Update test results for sort test

13 years agoInitial search may perform (relevance) sort
Adam Dickmeiss [Wed, 12 Oct 2011 09:30:08 +0000 (11:30 +0200)]
Initial search may perform (relevance) sort

The increasing flag is now taken into consideration. The pz:sortmap
does not include sort spec flags anymore: just strategy:field .
More test results added. It is clear that the block flag is not
honored for show as it should: would like block=1 to wait until
first record comes in - after (re)search with different sort criteria.

13 years agoMore work on sorting
Adam Dickmeiss [Tue, 11 Oct 2011 13:25:40 +0000 (15:25 +0200)]
More work on sorting

Session only searches once for each sort criteria. Each record is
matched against all records in cluster, to avoid duplicates.

13 years agoRefactor how total hits are calculated
Adam Dickmeiss [Mon, 10 Oct 2011 14:18:24 +0000 (16:18 +0200)]
Refactor how total hits are calculated

This is to ensure it doesn't get increased when we extend a search.

13 years agoShow command may re-search for targets that support it
Adam Dickmeiss [Fri, 7 Oct 2011 13:12:16 +0000 (15:12 +0200)]
Show command may re-search for targets that support it

13 years agoRename client_get_url to client_get_id
Adam Dickmeiss [Thu, 6 Oct 2011 10:45:01 +0000 (12:45 +0200)]
Rename client_get_url to client_get_id

13 years agostruct database: rename member url to id
Adam Dickmeiss [Thu, 6 Oct 2011 10:40:04 +0000 (12:40 +0200)]
struct database: rename member url to id

13 years agoUse client_get_url rather than database->url
Adam Dickmeiss [Thu, 6 Oct 2011 10:32:04 +0000 (12:32 +0200)]
Use client_get_url rather than database->url

client_get_url is safer and more appropriate in client.c.

13 years agoRemove prototype for undefined load_targets
Adam Dickmeiss [Thu, 6 Oct 2011 10:27:32 +0000 (12:27 +0200)]
Remove prototype for undefined load_targets

13 years agoMinor refactor WRT databases
Adam Dickmeiss [Thu, 6 Oct 2011 10:21:13 +0000 (12:21 +0200)]
Minor refactor WRT databases

13 years agoRemove useless memset for database creation
Adam Dickmeiss [Thu, 6 Oct 2011 10:08:26 +0000 (12:08 +0200)]
Remove useless memset for database creation

13 years agoAdd test for chaning pz:url for target within a session
Adam Dickmeiss [Thu, 6 Oct 2011 10:03:10 +0000 (12:03 +0200)]
Add test for chaning pz:url for target within a session

13 years agoRefactor: host entries no longer part of database
Adam Dickmeiss [Thu, 6 Oct 2011 10:01:36 +0000 (12:01 +0200)]
Refactor: host entries no longer part of database

Simplifies code and allows pz:url to be modified within a session for
some target.

13 years agoPazpar2 honors pz:url setting
Adam Dickmeiss [Wed, 5 Oct 2011 13:18:35 +0000 (15:18 +0200)]
Pazpar2 honors pz:url setting

13 years agoRemove useless headers for database utilities
Adam Dickmeiss [Wed, 5 Oct 2011 11:46:27 +0000 (13:46 +0200)]
Remove useless headers for database utilities

13 years agoRemove useless comments
Adam Dickmeiss [Wed, 5 Oct 2011 11:41:13 +0000 (13:41 +0200)]
Remove useless comments

13 years agoRemove unused setting pz:ip
Adam Dickmeiss [Wed, 5 Oct 2011 11:40:01 +0000 (13:40 +0200)]
Remove unused setting pz:ip

13 years agoRemove unused pz:option_recordfilter
Adam Dickmeiss [Wed, 5 Oct 2011 11:38:35 +0000 (13:38 +0200)]
Remove unused pz:option_recordfilter

13 years agoResolve host for database just before search
Adam Dickmeiss [Wed, 5 Oct 2011 11:29:58 +0000 (13:29 +0200)]
Resolve host for database just before search

Postpone resolving of host for database till just befor search so
that we can take care of a setting to set the actual URL for
target rather than use database ID as before.

13 years agoSimplify ZURL resolving and use regular getaddrinfo
Adam Dickmeiss [Tue, 4 Oct 2011 13:12:56 +0000 (15:12 +0200)]
Simplify ZURL resolving and use regular getaddrinfo

Caching of DNS entries is done better by a local caching DNS or by libc.

13 years agoReformat
Adam Dickmeiss [Tue, 4 Oct 2011 07:40:19 +0000 (09:40 +0200)]
Reformat

13 years agoMinor refactor WRT settings parsing
Adam Dickmeiss [Mon, 3 Oct 2011 13:20:32 +0000 (15:20 +0200)]
Minor refactor WRT settings parsing

Shouldn't change functionality.

13 years agoRemove settings member from conf_service
Adam Dickmeiss [Mon, 3 Oct 2011 12:50:32 +0000 (14:50 +0200)]
Remove settings member from conf_service

Was not in use anywhere.

13 years agochdir may be defined in unistd.h
Adam Dickmeiss [Mon, 3 Oct 2011 12:49:52 +0000 (14:49 +0200)]
chdir may be defined in unistd.h

13 years agoUpdate test_facets results due to different output
Adam Dickmeiss [Fri, 30 Sep 2011 18:35:08 +0000 (20:35 +0200)]
Update test_facets results due to different output

13 years agoMerge branch 'master' of ssh://git.indexdata.com/home/git/pub/pazpar2
Adam Dickmeiss [Fri, 30 Sep 2011 12:49:39 +0000 (14:49 +0200)]
Merge branch 'master' of ssh://git.indexdata.com/home/git/pub/pazpar2

13 years agoRefactor Pazpar2 HTTP response writing code
Adam Dickmeiss [Fri, 30 Sep 2011 12:47:54 +0000 (14:47 +0200)]
Refactor Pazpar2 HTTP response writing code

Header and footer for successful responses is handled by a
command handler. This makes the formatting slightly different.
It also fixes the termlist response which did not include the
XML header before.

13 years agoMerge branch 'master' of ssh://git.indexdata.com/home/git/pub/pazpar2
Dennis Schafroth [Fri, 30 Sep 2011 12:34:11 +0000 (14:34 +0200)]
Merge branch 'master' of ssh://git.indexdata.com/home/git/pub/pazpar2

13 years agoCheck for valid zquery
Dennis Schafroth [Fri, 30 Sep 2011 12:34:07 +0000 (14:34 +0200)]
Check for valid zquery

13 years agocommand=termlist without name parameter returns all termlists
Adam Dickmeiss [Fri, 30 Sep 2011 12:00:56 +0000 (14:00 +0200)]
command=termlist without name parameter returns all termlists

Previously if name parameter was omitted, only "subject" was returned.

13 years agoDocument the OR operation for limit parameter
Adam Dickmeiss [Wed, 28 Sep 2011 10:52:40 +0000 (12:52 +0200)]
Document the OR operation for limit parameter

13 years agoDepend on YAZ 4.2.17
Adam Dickmeiss [Wed, 28 Sep 2011 10:42:34 +0000 (12:42 +0200)]
Depend on YAZ 4.2.17

Because of new function nmem_strsplit_escape2.

13 years agoImplement limitmap OR. Bug #4675
Adam Dickmeiss [Tue, 27 Sep 2011 14:12:33 +0000 (16:12 +0200)]
Implement limitmap OR. Bug #4675

This requires nmem_strsplit_escape2 from YAZ.

13 years agopazpar2 configure: needs YAZ 4.2.16
Adam Dickmeiss [Tue, 27 Sep 2011 14:12:27 +0000 (16:12 +0200)]
pazpar2 configure: needs YAZ 4.2.16

13 years agoVersion 1.6.3 v1.6.3
Adam Dickmeiss [Thu, 22 Sep 2011 10:51:18 +0000 (12:51 +0200)]
Version 1.6.3

13 years agoNews
Adam Dickmeiss [Thu, 22 Sep 2011 10:51:06 +0000 (12:51 +0200)]
News

13 years agoMake termlist sorting stable
Adam Dickmeiss [Wed, 21 Sep 2011 12:48:49 +0000 (14:48 +0200)]
Make termlist sorting stable

Terms with same frequency are now sorted by their display name.
This makes a pretty display and improves our regression test because
qsort is not a stable sort.

13 years agosolr-pz2.xsl part of dist; used by test_solr
Adam Dickmeiss [Wed, 21 Sep 2011 10:01:40 +0000 (12:01 +0200)]
solr-pz2.xsl part of dist; used by test_solr

13 years agoReference to limitmap; spell fixes
Adam Dickmeiss [Wed, 21 Sep 2011 09:59:01 +0000 (11:59 +0200)]
Reference to limitmap; spell fixes

13 years agoMention pz:limitmap ; not pz:facetmap
Adam Dickmeiss [Wed, 21 Sep 2011 09:15:53 +0000 (11:15 +0200)]
Mention pz:limitmap ; not pz:facetmap

13 years agotest_solr: update results for other SOLR server
Adam Dickmeiss [Tue, 20 Sep 2011 11:26:41 +0000 (13:26 +0200)]
test_solr: update results for other SOLR server

13 years agomake dist: include test_solr.{cfg,urls}
Adam Dickmeiss [Tue, 20 Sep 2011 11:14:15 +0000 (13:14 +0200)]
make dist: include test_solr.{cfg,urls}

13 years agotest_solr.sh: use http://ocs-test.indexdata.com/solr/
Adam Dickmeiss [Tue, 20 Sep 2011 11:07:06 +0000 (13:07 +0200)]
test_solr.sh: use ocs-test.indexdata.com/solr/

Is a public address.

13 years agotest_solr.sh uses stable SOLR test server
Adam Dickmeiss [Tue, 20 Sep 2011 11:04:27 +0000 (13:04 +0200)]
test_solr.sh uses stable SOLR test server

13 years agotest: proper header about what sh is using each cfg
Adam Dickmeiss [Tue, 20 Sep 2011 10:46:28 +0000 (12:46 +0200)]
test: proper header about what sh is using each cfg

13 years agolocal clean removes *.log *.dif
Adam Dickmeiss [Tue, 20 Sep 2011 10:35:09 +0000 (12:35 +0200)]
local clean removes *.log *.dif

13 years agoAdd target url to missing limitmap warning
Dennis Schafroth [Tue, 20 Sep 2011 10:13:19 +0000 (12:13 +0200)]
Add target url to missing limitmap warning

13 years agoMerge branch 'master' of ssh://git.indexdata.com/home/git/pub/pazpar2
Dennis Schafroth [Tue, 20 Sep 2011 09:23:06 +0000 (11:23 +0200)]
Merge branch 'master' of ssh://git.indexdata.com/home/git/pub/pazpar2

13 years agoNew special sort parameter value 'position'.
Adam Dickmeiss [Mon, 19 Sep 2011 13:21:28 +0000 (15:21 +0200)]
New special sort parameter value 'position'.

The 'position' sorts merged records by their original position
from the remote target. This is primarily useful for debugging and
may be used for targets that already perform some kind of relevance
ranking. Note that sort by default is decreasing; so to get records
in their original order sort=position:1 must be used.

13 years agoMove declaration
Adam Dickmeiss [Mon, 19 Sep 2011 13:20:55 +0000 (15:20 +0200)]
Move declaration

13 years agoRefactor code that extracts terms for relevance-ranking
Adam Dickmeiss [Mon, 19 Sep 2011 12:03:16 +0000 (14:03 +0200)]
Refactor code that extracts terms for relevance-ranking

Function pull_terms checks for max number of terms in termlist -
to prevent SEGV/exploit.

13 years agoRemove dead code: TXML is not a transfer syntax
Adam Dickmeiss [Mon, 19 Sep 2011 10:49:58 +0000 (12:49 +0200)]
Remove dead code: TXML is not a transfer syntax

13 years agotest facets uses version reported by yaz-client
Adam Dickmeiss [Thu, 15 Sep 2011 13:08:55 +0000 (15:08 +0200)]
test facets uses version reported by yaz-client

rather than yaz-config which is only available in the YAZ
development package.

13 years agoVersion 1.6.2 v1.6.2
Adam Dickmeiss [Thu, 15 Sep 2011 10:30:27 +0000 (12:30 +0200)]
Version 1.6.2

13 years agodoc: use xref id icuchain instead of icu_chain
Adam Dickmeiss [Thu, 15 Sep 2011 10:30:14 +0000 (12:30 +0200)]
doc: use xref id icuchain instead of icu_chain

The underscore in xref id breaks dblatex. Introducing the xreflabel
helps. See http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=492959

13 years agoFacility to change working directory for daemon
Adam Dickmeiss [Thu, 15 Sep 2011 09:12:15 +0000 (11:12 +0200)]
Facility to change working directory for daemon

Option -wdir sets working directory to dir.

13 years agotmarc.xsl: all 773$g to journal-subpart
Adam Dickmeiss [Wed, 14 Sep 2011 17:31:19 +0000 (19:31 +0200)]
tmarc.xsl: all 773$g to journal-subpart

13 years agoFix copyright sign
Adam Dickmeiss [Wed, 14 Sep 2011 12:54:48 +0000 (14:54 +0200)]
Fix copyright sign

13 years agoVersion 1.6.1 v1.6.1
Adam Dickmeiss [Wed, 14 Sep 2011 12:22:23 +0000 (14:22 +0200)]
Version 1.6.1

13 years agoRemove a few log stms
Adam Dickmeiss [Wed, 14 Sep 2011 11:52:37 +0000 (13:52 +0200)]
Remove a few log stms

13 years agoDocument metadata attribute 'facetrule'
Adam Dickmeiss [Wed, 14 Sep 2011 11:08:48 +0000 (13:08 +0200)]
Document metadata attribute 'facetrule'

13 years agoRename metadata attribute icu_chain to facetrule
Adam Dickmeiss [Wed, 14 Sep 2011 11:06:48 +0000 (13:06 +0200)]
Rename metadata attribute icu_chain to facetrule

This is have the possiblity to override other than just facet
normalization per metadata field in the future.

13 years agoOverride id in icu_chain for relevance, sort, ..
Adam Dickmeiss [Wed, 14 Sep 2011 11:05:32 +0000 (13:05 +0200)]
Override id in icu_chain for relevance, sort, ..

Ignore the id attribute for the obsolete rules in relevance, sort,
mergekey, facet.

13 years agoReformat
Adam Dickmeiss [Wed, 14 Sep 2011 10:46:28 +0000 (12:46 +0200)]
Reformat

13 years agoDocument new setting icu_chain for service/server
Adam Dickmeiss [Wed, 14 Sep 2011 10:40:33 +0000 (12:40 +0200)]
Document new setting icu_chain for service/server

13 years agoReformat documentation for nxml mode.
Adam Dickmeiss [Wed, 14 Sep 2011 09:27:25 +0000 (11:27 +0200)]
Reformat documentation for nxml mode.

13 years agoRename some charset functions
Adam Dickmeiss [Wed, 14 Sep 2011 07:37:54 +0000 (09:37 +0200)]
Rename some charset functions

The _relevance_-name was relevant when charset system was only used
to normalize relevance terms, but is inappropriate when it's used to
normalize 4 different types of terms.

13 years agoPreserve order for merged metadata
Adam Dickmeiss [Tue, 13 Sep 2011 20:32:08 +0000 (22:32 +0200)]
Preserve order for merged metadata

Fixes issue mentioned by Sven Porst
http://lists.indexdata.dk/pipermail/yazlist/2011-July/003230.html

13 years agoLog if attribute id for icu_chain is missing
Adam Dickmeiss [Tue, 13 Sep 2011 19:01:08 +0000 (21:01 +0200)]
Log if attribute id for icu_chain is missing

13 years agoInit member database_hosts of conf_config
Adam Dickmeiss [Tue, 13 Sep 2011 19:00:10 +0000 (21:00 +0200)]
Init member database_hosts of conf_config

13 years agotmarc.xsl: set journal-subpart to 773$g only
Adam Dickmeiss [Tue, 13 Sep 2011 16:53:46 +0000 (18:53 +0200)]
tmarc.xsl: set journal-subpart to 773$g only

13 years agoInstalled service/server uses new ICU chain definitions
Adam Dickmeiss [Tue, 13 Sep 2011 13:50:50 +0000 (15:50 +0200)]
Installed service/server uses new ICU chain definitions

13 years agoAllow multiple ICU chains for facets
Adam Dickmeiss [Tue, 13 Sep 2011 13:25:52 +0000 (15:25 +0200)]
Allow multiple ICU chains for facets

The charsets ICU interface can keep any number of identified
ICU chains where "relevance", "sort", "mergekey", and "facet" are
just the existing ones. The elements in server/service for
defining ICU chains relevance, sort, mergekey and facet are deprecated
and may be replaced by <icu_chain id="id" >..</icu_chain>.

13 years agoICU per metadata element (icu_chain attribute)
Adam Dickmeiss [Mon, 12 Sep 2011 19:16:53 +0000 (21:16 +0200)]
ICU per metadata element (icu_chain attribute)

13 years agoRefactor metadata config parsing
Adam Dickmeiss [Mon, 12 Sep 2011 18:15:02 +0000 (20:15 +0200)]
Refactor metadata config parsing

Avoid memory leaks and throw diagnostic for unknown metadata attribute.

13 years agoRemove section with unused declarations
Adam Dickmeiss [Mon, 12 Sep 2011 16:45:07 +0000 (18:45 +0200)]
Remove section with unused declarations

13 years agoICU chain "facet" honors YAC ICU element <display/>.
Adam Dickmeiss [Mon, 12 Sep 2011 16:29:48 +0000 (18:29 +0200)]
ICU chain "facet" honors YAC ICU element <display/>.

Display term and normalized term is now saved per facet.

13 years agoVersion 1.6.0 v1.6.0
Adam Dickmeiss [Mon, 12 Sep 2011 12:40:40 +0000 (14:40 +0200)]
Version 1.6.0

13 years agoDescribe pz:limitmap
Adam Dickmeiss [Mon, 12 Sep 2011 12:40:26 +0000 (14:40 +0200)]
Describe pz:limitmap

13 years agoAdd pz:metadata type pages-number
Adam Dickmeiss [Mon, 12 Sep 2011 12:03:28 +0000 (14:03 +0200)]
Add pz:metadata type pages-number

Because this is now generated in tmarc.xsl (773$g).

13 years agotmarc.xsl: get components out of 773$g.
Adam Dickmeiss [Mon, 12 Sep 2011 11:59:00 +0000 (13:59 +0200)]
tmarc.xsl: get components out of 773$g.

The following pz:metadata elements are generated: volume-number,
issue-number, pages-number. Bug #4572.

13 years agoRemove unused named templates
Adam Dickmeiss [Mon, 12 Sep 2011 10:32:33 +0000 (12:32 +0200)]
Remove unused named templates

13 years agoTrim whitespace and comma for ,-list for field 600
Adam Dickmeiss [Mon, 12 Sep 2011 09:09:18 +0000 (11:09 +0200)]
Trim whitespace and comma for ,-list for field 600

Fixes bug #4594.

13 years agopz:limitmap now in use
Adam Dickmeiss [Fri, 9 Sep 2011 13:01:39 +0000 (15:01 +0200)]
pz:limitmap now in use

pz:limitmap[name] is used as query translation when applying
limit parameter for search. Previously, pz:facetmap was used.
pz:limitmap[name] can have one of two forms (types). The leading
type is followed immedately by colon then by form specific name.
For type "ccl:" the value is an existing cclmap.  Example:
limitmap[author]="ccl:au", which could work for many existing Z39.50
targets. For type "rpn:" the value is PQF verbatim. Example:
limitmap[author]="rpn:@attr 1=1003 @attr 6=3" (author, complete field).

13 years agoMerge branch 'master' of ssh://git.indexdata.com/home/git/pub/pazpar2
Dennis Schafroth [Fri, 2 Sep 2011 13:26:26 +0000 (15:26 +0200)]
Merge branch 'master' of ssh://git.indexdata.com/home/git/pub/pazpar2

13 years agoFix deadlock on wrongly parsed limit parameter
Dennis Schafroth [Mon, 29 Aug 2011 13:52:12 +0000 (13:52 +0000)]
Fix deadlock on wrongly parsed limit parameter

13 years agoAttaches item id to loan period
Niels Erik G. Nielsen [Mon, 29 Aug 2011 06:29:58 +0000 (08:29 +0200)]
Attaches item id to loan period

13 years agoMove to mk2-config-dbc
Dennis Schafroth [Wed, 24 Aug 2011 11:08:42 +0000 (11:08 +0000)]
Move to mk2-config-dbc

13 years agoMerge branch 'master' of ssh://git.indexdata.com/home/git/pub/pazpar2
Dennis Schafroth [Wed, 24 Aug 2011 10:33:41 +0000 (12:33 +0200)]
Merge branch 'master' of ssh://git.indexdata.com/home/git/pub/pazpar2

13 years agoMention limit parameter
Adam Dickmeiss [Tue, 23 Aug 2011 12:03:56 +0000 (14:03 +0200)]
Mention limit parameter

13 years agoMerge branch 'master' of ssh://git.indexdata.com/home/git/pub/pazpar2 v1.5.7
Adam Dickmeiss [Fri, 19 Aug 2011 11:46:15 +0000 (13:46 +0200)]
Merge branch 'master' of ssh://git.indexdata.com/home/git/pub/pazpar2