headers
Markus Hoenicka | 2 Sep 2006 01:42
Picon

need help to reproduce bug #1550514

Hi all,

did anyone of you encounter a problem like this:

http://sourceforge.net/tracker/index.php?func=detail&aid=1550514&group_id=26091&atid=385991

In brief, chapter and book titles imported from RIS documents
sometimes get truncated, leading to an incorrectly assembled SQL query
which throws an error.

I can't reproduce this problem in the following environments:

- FreeBSD 5.4, 6.1: MySQL, PostgreSQL, SQLite, SQLite3
- WinXP, Cygwin: MySQL
- Debian 3.1: SQLite

If anyone can reproduce it with the test data provided in the bug
report, please get back to the list for further analysis.

thanks,
Markus

--

-- 
Markus Hoenicka
markus.hoenicka <at> cats.de
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
(Continue reading)

Permalink | Reply | 0 comments |
headers
Olivier Berger | 4 Sep 2006 12:46
Picon

Re: latin1 characters in RIS entries not converted to UTF-8 by runbib when creating .bib.xml bibliography file

Le jeudi 31 août 2006 à 15:27 +0200, Markus Hoenicka a écrit :
> Hi,
> 
> Olivier Berger <olivier.berger <at> int-evry.fr> was heard to say:
> 
> > The .bib.xml output from runbib contains latin1 accentuated characters
> > instead of UTF-8, for the titles in the bibliography elements, hence
> > breaking the xsltproc done on the DocBook XML file afterwards, although
> > I invoke runbib with option "-E utf8"...
> >
> 
> Did you try "-E utf-8"? libdbi (the database abstraction layer used by RefDB)
> uses the MIME names of the encodings. These names are case-insensitive, but
> "utf8" is different from "utf-8". 

Well, that hasn't changed anything to the output with -E utf-8 (or -E
UTF-8).

My RIS file contains latin1 (so I suppose the same in the database of
refdb ?), the .bib file too, although it will be included in a utf8 XML
doc....

> If you are in doubt, check the file
> README.encodings shipped with libdbi-drivers.
> 

Just for future reference :
http://libdbi-drivers.cvs.sourceforge.net/*checkout*/libdbi-drivers/libdbi-drivers/README.encodings

If you have more hints on what could be wrong...
(Continue reading)

Permalink | Reply | 2 comments |
headers
Olivier Berger | 4 Sep 2006 13:02
Picon

Re: latin1 characters in RIS entries not converted to UTF-8 by runbib when creating .bib.xml bibliography file

Allright, sorry for previous message... in the meantime, I've been
thinking of how it works, and checked... and apparently the issue is
that I was editing the RIS entries as latin-1, whereas
in /etc/refdb/refdbcrc, I have :
fromencoding UTF-8

So it seems my database records where containing latin-1 accentuations,
which would be generated as such in the UTF8 XML...

I updated the RIS entry to use UTF8 accents, and now I get everything
clean and compiling...

Maybe I should change my /etc/refdb/refdbcrc preferences, then (or see
how to have emacs RIS mode handle everything as UTF8 by default ?)

Sorry for bothering.

Best regards,

Le lundi 04 septembre 2006 à 12:46 +0200, Olivier Berger a écrit :
> Le jeudi 31 août 2006 à 15:27 +0200, Markus Hoenicka a écrit :
> > Hi,
> > 
> > Olivier Berger <olivier.berger <at> int-evry.fr> was heard to say:
> > 
> > > The .bib.xml output from runbib contains latin1 accentuated characters
> > > instead of UTF-8, for the titles in the bibliography elements, hence
> > > breaking the xsltproc done on the DocBook XML file afterwards, although
> > > I invoke runbib with option "-E utf8"...
> > >
(Continue reading)

Permalink | Reply | 1 comment |
headers
Markus Hoenicka | 4 Sep 2006 13:23
Picon

Re: latin1 characters in RIS entries not converted to UTF-8 by runbib when creating .bib.xml bibliography file

Hi Olivier,

the latest prerelease and the current SVN version use UTF-8 as the default
throughout (input/storage/output). On the long run you'll be better off if you
set up your editor to use UTF-8 as well. I've got this in my .emacs:

;; xml files, planner files, as well as files with "utf8" somewhere
;; in the path are opened as Unicode
(setq file-coding-system-alist
      (append (list
	       (cons "\\.xml$" 'utf-8)
	       (cons "utf8" 'utf-8)
	       (cons "Plans" 'utf-8))
	      file-coding-system-alist))

This covers risx files as long as they use the suffix ".xml", and RIS files as
long as you put "utf8" somewhere in the filename, like in "test.utf8.ris".

regards,
Markus

Olivier Berger <olivier.berger <at> int-evry.fr> was heard to say:

> Allright, sorry for previous message... in the meantime, I've been
> thinking of how it works, and checked... and apparently the issue is
> that I was editing the RIS entries as latin-1, whereas
> in /etc/refdb/refdbcrc, I have :
> fromencoding UTF-8
>
> So it seems my database records where containing latin-1 accentuations,
(Continue reading)

Permalink | Reply | 0 comments |
headers
Dominik Reusser | 7 Sep 2006 13:45
Picon
Favicon

make error: /bin/sh: dtdparse: command not found

I just tried to compile the latest svn-version (165) of refdb on a debian 
testing system:
./autogen.sh
./configure
...
checking for dtdparse... no
checking for dtdformat... no
....

./make

I get the following error:
SGML_CATALOG_FILES=""  dtdparse --title "citationlistx XML DTD" --output 
citationlistx.xml --declaration ../declarations/xml.dcl --system-id
"http://refdb.sourceforge.net/dtd/citationlistx.dtd" ../dtd/citationlistx.dtd
/bin/sh: dtdparse: command not found
make[1]: *** [citationlistx.xml] Error 127
make[1]: Leaving directory `/home/admin/refdb/doc'
make: *** [all-recursive] Error 1

This can be omitted using the --disable-docs option of configure or installing 
dtdparse. However, I'm wondering wether configure shouldn't avoid this 
problem and not build the docs if dtdparse is not present?

But maybe, I'm just missing something
Thanks for your help.
Dominik

--

-- 
Dominik Reusser
(Continue reading)

Permalink | Reply | 1 comment |
headers
Markus Hoenicka | 7 Sep 2006 14:06
Picon

Re: make error: /bin/sh: dtdparse: command not found

Dominik Reusser <dreusser <at> uni-potsdam.de> was heard to say:

> This can be omitted using the --disable-docs option of configure or
> installing
> dtdparse. However, I'm wondering wether configure shouldn't avoid this
> problem and not build the docs if dtdparse is not present?
>

This is sort of a catch22. The release and prerelease archives contain the
prebuilt docs. make is not supposed to do anything in the /doc subdir, but in
order to install the prebuilt docs, make install should run in /doc even if
dtdparse and dtdformat are not present. This allows us to use the convenient
doc installation routines of the autotools, but it means that the lack of these
tools must not turn off the processing of /doc.

As you have noticed, this means for the svn version that make will attempt to
build the docs even if the necessary tools are not around. I consider this the
lesser of two evils, as someone building from svn can be expected to read the
manual which lists the prerequisites. The configure checks for these tools
should therefore be treated as a warning, not an error.

regards,
Markus

--

-- 
Markus Hoenicka
markus.hoenicka <at> cats.de
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de
(Continue reading)

Permalink | Reply | 0 comments |
headers
Jeremy Malcolm | 16 Sep 2006 11:59
Picon
Favicon

Id- in citation keys

Hello.  I don't know if this deserves to be called a bug, but I had a 
citation key that said "Johnson2001The-Idea-of-ICA" (I have configured 
my bibliography manager to put as much information in the citation keys 
as it can, so they're easy to recognise.  Because this has "-Id" in it, 
it caused runbib to fail with this error:

1146: Table 'idjohnson2001the.t_refdb' doesn't exist
SELECT refdb_id, refdb_type FROM idjohnson2001the.t_refdb WHERE 
refdb_citekey='EA-OF-ICA' AND refdb_type!='DUMMY'
select failed

Anyway, if it's worth reporting as a bug, consider it reported.  No need 
to fix it on my account, as it's easy to workaround now that I know what 
caused it.  Here's the full reference in RIS format:

TY  - ELEC
AU  - Johnson, David R
AU  - Crawford, Susan P
PY  - 2001//
TI  - The Idea of ICANN
UR  - http://www.icannwatch.org/archive/the_idea_of_icann.htm
ID  - JOHNSON2001THE-IDEA-OF-ICA
ER  -

Thanks.

--

-- 
Jeremy Malcolm LLB (Hons) B Com
Internet and Open Source lawyer, IT consultant, actor
host -t NAPTR 1.0.8.0.3.1.2.9.8.1.6.e164.org|awk -F! '{print $3}'
(Continue reading)

Permalink | Reply | 2 comments |
headers
Markus Hoenicka | 16 Sep 2006 23:34
Picon

Id- in citation keys

Hi,

I consider it a bug because refdbd should not accept or construct
citation keys which cause this error. I'll look into this. Thanks for
reporting the problem.

regards,
Markus 

Jeremy Malcolm writes:
 > Hello.  I don't know if this deserves to be called a bug, but I had a 
 > citation key that said "Johnson2001The-Idea-of-ICA" (I have configured 
 > my bibliography manager to put as much information in the citation keys 
 > as it can, so they're easy to recognise.  Because this has "-Id" in it, 
 > it caused runbib to fail with this error:
 > 
 > 1146: Table 'idjohnson2001the.t_refdb' doesn't exist
 > SELECT refdb_id, refdb_type FROM idjohnson2001the.t_refdb WHERE 
 > refdb_citekey='EA-OF-ICA' AND refdb_type!='DUMMY'
 > select failed
 > 
 > Anyway, if it's worth reporting as a bug, consider it reported.  No need 
 > to fix it on my account, as it's easy to workaround now that I know what 
 > caused it.  Here's the full reference in RIS format:
 > 
 > TY  - ELEC
 > AU  - Johnson, David R
 > AU  - Crawford, Susan P
 > PY  - 2001//
 > TI  - The Idea of ICANN
(Continue reading)

Permalink | Reply | 0 comments |
headers
Markus Hoenicka | 19 Sep 2006 22:30
Picon

Id- in citation keys

Hi,

I've fixed this problem in svn. It will also be available shortly in
the upcoming 0.9.8-pre4 prerelease. refdbd will now change the "-ID"
to "_ID" when checking or creating the citation key.

regards,
Markus

Jeremy Malcolm writes:
 > Hello.  I don't know if this deserves to be called a bug, but I had a 
 > citation key that said "Johnson2001The-Idea-of-ICA" (I have configured 
 > my bibliography manager to put as much information in the citation keys 
 > as it can, so they're easy to recognise.  Because this has "-Id" in it, 
 > it caused runbib to fail with this error:
 > 

--

-- 
Markus Hoenicka
markus.hoenicka <at> cats.de
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Permalink | Reply | 0 comments |
headers
Markus Hoenicka | 19 Sep 2006 23:57
Picon

ANN: refdb-0.9.8-pre4 available for download

Hi,

time for another prerelease:

http://refdb.sourceforge.net/pre/refdb-latest.tar.gz

The focus of this prerelease is the main database upgrading
stuff. Users upgrading from 0.9.8-pre2 and earlier will have to
upgrade their main databases. refdbd can now handle this for
you. While this works for me (tested with all database engines), I'd
be interested to hear reports from others. Fear not, you can always
create your main database manually by following the procedure in the
manual. To see how to use this new feature, please check out the
UPGRADING file.

Other notable changes are listed below.

regards,
Markus

Bugfixes:

- exclusive queries (e.g. :UR:!~"http") now work as expected for links
  (UR, L1-L4), authors, and keywords

- refdbd now modifies citation keys containing the string "-ID" as
  this may cause problems when creating bibliographies. "-ID" will be
  replaced with "_ID" to avoid problems.

Improvements:
(Continue reading)

Permalink | Reply | 0 comments |

Gmane