#102 L1 - L4 fields are not displayed in HTML output

refdbd (55)

Until recently I used pdfroot and AV field to store the location of my articles in pdf.
The command
refdbc -C getref '(:AU:~Mason) -t html -o my.html -S PY -d test1'
would produce my.html where for all citations there was a link to a pdf file

Now the article locations are stored in L1, L2, L3 , L4 fields.
The same command produces HTML file, where there is a line
but there are no links to the file.

But when I run the comman without HTML output, I see
refdbc -C getref '(:AU:~Mason) -S PY -d test1'

PDF: PATH:work/lit/ms/S/SteinerSaenger1991.pdf
FULLTEXT: PATH:work/lit/ms/S/Ste991.pdf
RELATED: PATH:work/lit/ms/S/Ste99aaaaa1.pdf
IMAGE: PATH:work/lit/ms/S/eaaaa991.pdf


  • akusmin

    Some more info on the version of RefDB I use:

    refdba: viewstat
    You are served by: refdb 0.9.9-1
    SVN revision: 531
    Client IP:
    Connected via mysql driver (dbd_mysql v0.8.3-1)
    to: 5.0.84
    db version: 3
    serverip: localhost
    timeout: 180
    dbs_port: 3306

  • Thanks for reporting this problem. There was indeed an inconsistency in evaluating the string which specifies optional components of the screen and html outputs. I've fixed this problem in SVN. For the time being, please use "-s UR" or an equivalent entry in your refdbc config file to explicitly request L1-L4 output from the html backend. This solution works at least with the previous svn revision. Please shout again if this workaround doesn't help in 0.9.9-1.

    • labels: --> refdbd
    • assigned_to: nobody --> mhoenicka
    • status: open --> closed-fixed
  • Ok, I forgot to mention that the proper way to request L1-L4 output using the current SVN revision (i.e. after fixing the bug) is to use "-s LX" or an appropriate entry in your refdbc config file.

  • akusmin

    • status: closed-fixed --> open-fixed
  • akusmin

    Thanks for a quick response.
    Yes, adding UR into refdbcrc did result in displaying the links in a HTML output.

    But there is one problem left (and I had to say about it before):
    the links (with or without HTML output, see the example in the original bug description)
    still contain the word PATH which is not substituted by the pdfroot .
    Thus, I can not open a pdf file simply by clicking on the link in a browser.
    Did you fix it as well ?
    Should this problem be reported as a separate bug ?

    best regards,
    A. Kusmin

  • If you continue to use data from previous RefDB releases, you'll have to migrate these data properly in order to get rid of the "PATH:" prefix. The exact procedure is described in the file UPGRADING in the top level directory of the RefDB sources. In brief, you'll have to substitute the string "PATH:" by a valid URI prefix like "file://" to make the links work on your box. UPGRADING contains an example of a sed command which can do this conversion on your exported data. After the substitution, re-import your data again.

  • akusmin

    Right. I am sorry, I read about it in the RefDB manual, but it slipped my mind. Thank you.