two problems with BibTex import (middle initials and 8-bit)

[Massachusetts user]

I am importing a BibTex database of references (from http://people.umass.edu/kbj/homepage/Co ... graphy.bib), and I am encountering two problems.

I have no experience with BibTex until now, so perhaps I have made some obvious error.

The first problem concerns middle initials for authors. With references like the following:

@book{Fodor:1975,
Address = {Cambridge, Mass.},
Author = {Fodor, Jerry A.},
Keywords = {Cognition.},
Pages = {x, 214},
Publisher = {Harvard University Press},
Title = {The language of thought},
Year = {1975}}

-- Bookends is consistently rendering the author's name as follows: "A., Fodor, Jerry". It seems to be taking all middle initials and placing them with a comma in front of the last name.

Also, though I have turned on the 8-bit handling/importing feature on the BibTex tab in Preferences, it seems to be doing nothing at all. Thus, where the BibTex file has "Bo{\v{s}}kovi{\'c}, {\v{Z}}eljko" in the author field, this survives unchanged in the Bookends import.

[Jon]

Hi,

For the first problem, you have to change the import filter and tell it to expect the author's name as surname-first (or turn off author/editor parsing altogether).

As for the second, Bookends doesn't check for every possible escaped accented character. If you have characters you want to import that Bookends doesn't support, please send me the one-letter representation as well as the BibTeX equivalent you are trying to import and I'll add them.

Jon
Sonny Software

[Massachusetts user]

Thank you for your help -- but unfortunately I am still having problems. I have tried your suggestion: editing the import filter so that it expects the author's name as surname-first. This gives me some properly imported items, but also a big problem. Whenever an authors name includes a middle initial, the import is fine. When the middle initial is missing, I get most of the entry stuck in the author field.

I have done a lot of playing around, both random and purposeful, with the filter description, but I have been unable to get this to work right.

Here are two sample entries (pared down). The one with the middle initial comes out fine, but the one without does not. Add an initial to one, or take it away from the other, and I can reliably predict that the one without the initial will be a mess, and the other one ok. Can you help?

Thanks!

@article{Abeille:2002,
Author = {Abeille, Anne},
Date-Added = {2006-07-31 09:13:16 -0400},
Date-Modified = {2006-07-31 09:14:44 -0400},
Journal = {Language},
Local-Url = {file://localhost/Users/kyle/Documents/Bookends/Attachments/*Language/78.3abeille.pdf},
Number = {3},
Pages = {404--452},
Title = {The Syntactic Structure of {F}rench Auxiliaries},
Volume = {78},
Year = {2002},
Abstract = {While a consensus has been reached}}

@article{Beaver:2004,
Author = {Beaver, David I.},
Date-Added = {2006-07-28 09:44:45 -0400},
Date-Modified = {2006-07-28 09:45:55 -0400},
Journal = {Linguistics and Philosophy},
Local-Url = {file://localhost/Users/kyle/Documents/Bookends/Attachments/*LandP/LP27(1)_Beaver.pdf},
Number = {1},
Pages = {3--56},
Title = {The Optimization of Discourse Anaphora},
Volume = {27},
Year = {2004},
Abstract = {In this paper the Centering model of anaphoraresolution and discourse coherence}}

[Jon]

The BibTex import filter is one of the few that uses a "valid tag" to determin end-of-field. That means you have to account for every tag (ignore the ones you don't want imported). For example,

Date-Added

is not a tag in the default filter we provide. Add it (to Ignore, if you like) and any other tags that you expect to encounter that Bookends doesn't know about yet.

Jon
Sonny Software

[Massachusetts user]

OK, I've done that (set them to "ignore"), but nothing changes. What makes the difference is whether the name ends with a middle initial.

I'm running the test not on the full database that I wish to import, but on a simple file with the two entries I included in my previous message, so you can see what I am trying to do. Any ideas about what is going wrong.

[Massachusetts user]

Actually it seems to be working completely, but here is what it took.

The last and first names in the bibtex data base are, as you can see, separated by a comma.

In the panel for editing the BibTex filter, following your instructions, I set the pulldown menu to "Surname First". Now right below this is another pulldown menu with the heading "First and last names separated by" -- with two options: "Space" and "Smart".

Since they are separated by a comma, I picked "Smart".

What makes it work for me now is in fact picking "Space" -- which makes no sense to me at all. But it does seem to work now.

Thanks you for your help. Is there, however, either a bug in the program or a mislabeled menu item that caused this odd behavior? Or am I misunderstanding something?

[Jon]

The Smart refers to the ability of Bookends to determine if a compound surname (like Van Allen) is really a surname or a first (Van) and last name (Allen). It only works reliably when the source provides initials instead of full first names. If they don't, then "space" is what to use (and there is a space after the comma, which is what Bookends notices).

Jon
Sonny Software

[Massachusetts user]

I understand now, but -- with respect -- perhaps these labels will confuse others as well.

I now see the discussion on p.70 of the manual that clarifies what "smart" means in this context, but given that the other choice was "space" and I *thought* I knew what that meant -- it never occured to me to check what "smart" might mean here.

Might I suggest different text in the panel and pull-down menu? For example:

Special handling of compound names? [pulldown: Yes/No]

-- with a warning if you pick "yes" that gives you the warning that I now see as the second sentence on p.71 of the manual.

It would have saved me tons of time!

But thanks once again, regardless.