#30 — Wrapped bibtex entries have too many blanks

by Bernd Rellermeyer — last modified Jan 08, 2009 02:37 PM

State	Unconfirmed
Version:	0.9.0
Area	Functionality
Issue type	Bug
Severity	Medium
Submitted by	Bernd Rellermeyer
Submitted on	May 12, 2007
Responsible	—
Target release:	—

Return to tracker

Tags:

import handling

When importing a bibtex entry with a value wrapped into more than one line, the result has too many blanks. The following patch in bibtex.py seems to help:

def parseEntry(self, entry):
    """
    parses a single entry

    returns a dictionary to be passed to
    BibliographyEntry's edit method
    """

    ...

    # remove newlines and <CR>s, and remove the last '}'
    entry = entry.replace('\r', '').rstrip().rstrip('}')
    entry = re.sub('\s+', ' ', entry)
    tokens = self.pattern.split(entry)

    ...

Steps to reproduce:: Import the following entry and look at the publisher entry:

@Book{stammen92:_das_schwere_erbe,
  editor = {Theo Stammen},
  title = {Das schwere Erbe: 1918{\textendash}1923},
  publisher = {Bayerische Landeszentrale f{\"u}r politische
                  Bildungsarbeit},
  year = 1992,
  volume = 1,
  series = {Die Weimarer Republik},
  address = {M{\"u}nchen},
  edition = {Zweite}
}

Added by Tim Hicks on May 20, 2007 11:33 AM

Can you provide a test that fails before your patch is applied, and passes after?

Added by Bernd Rellermeyer on May 28, 2007 11:12 AM

The example above is a test. I used emacs to write my bibtex file. The emacs mode has the ability to wrap long lines with an indent to match the previous line (pressing meta-q). That means, for a long title I get two lines with the second line indented with several blanks. The parser simply concatenated these lines without eliminating the blanks at the beginning of the second line.

Added by Tim Hicks on Jun 04, 2007 04:10 PM

Sorry, I'm still not totally clear on what you say the correct behaviour should be. That's why I asked if you could provide a test.

By a test, I mean a "unit test" of the sort that we have already for bibtex parsing. See http://dev.plone.org/[…]/testBibtexParsing.py. If you can provide a method that tests the specific functionality that you want - one that fails with current CMFBib code and passes with your proposed patch - then we can probably apply your change straight away.

No responses can be added.