Re: text to bibliography?

On Aug 16, 8:23 am, p0 <yves.dho...@xxxxxxxxx> wrote:
grammatim schreef:



On Aug 15, 7:28 pm, p0 <yves.dho...@xxxxxxxxx> wrote:
grammatim schreef:

On Aug 14, 6:34 pm, p0 <yves.dho...@xxxxxxxxx> wrote:
The software does not support what you want. And I think it is
doubtful it ever will.

An "author" of a book is not necessarely represented by a b:Author/
b:Author element in your XML.

You seem to be overlooking the bit where I said this could be done by
inserting a code, which would still be vastly preferable to retyping
the entire contents of a full subject bibliography.

I was talking about automatically processing lists without codes. The
second part of my reply stated that as soon as you start using codes,
you can just as well use xml tags and there would be no point I doing
anything automatically.

How is an "xml tag" not a code?

It is a code, the point is, once you add the xml tags, there is no
more need for processing. You have done the entire processing by hand.
All that is left to do is copy/pasting the sources into the database
and you are done.

How is typing (or pasting) a few-letter code not better than typing
entire names, titles, etc.?

Xml was intended to be human readable. The disadvantage is that tags
tend to be long, the major advantage is that you don't have to learn a
few dozen non-descriptive codes by heart (would @c be city, comments,
chapternumber, conferencename, country, court, or casenumber?). Also
the use of closing tags is a good way in getting around punctuation
issues.

The following simple example:

John Doe. "A book by myself". London, 2008.

would result in the following string with xml 'codes':

<b:Source><b:SourceType>Book</
b:SourceType><b:Author><b:Author><b:NameList><b:Person><b:First>John</
b:First><b:Last>Doe</b:Last></b:Person></b:NameList></b:Author></
b:Author><b:Garbage>. "</b:Garbage><b:Title>A book by myself</
b:Title><b:Garbage>". </b:Garbage><b:City>London</b:City><b:Garbage>,
</b:Garbage><b:Year>2008</b:Year><b:Garbage>.</b:Garbage><b:Source>

What a stupid system.

I would expect something like

<au>Doe, John</au><ti>A Book by Myself</ti><pl>London</pl><pu>Smith &
Wesson</pu><yr>2008</yr>

And </yr> isn't needed because it's always 4 digits, and the place and
publisher would use codes rather than spelling out: <pl>L<pu>SW. <au>
is known to always select an item from the Name list -- as are also
<ed>, <tr>, etc.

where b:Garbage elements are meaningless elements and should be
removed from the source at a later stage. Do you really think typing/
copying/pasting out all those 'codes' is easier then copy/pasting 5
elements into a form?

A boo with a 20-word title can have six editors.

If the book would be an edited book,
than the author would be a b:Author/b:Editor element. If the book
would be a translated work, the author would be a b:Author/
b:Translator element.

If it were a proper relational database, then there would be a list of
"names," and in any particular instance, a name could be an "author,"
an "editor," a "translator," or even some combination of the above.

I am not a specialist when it comes to relational database, but I
agree that the current layout is not in full normal form. However the
way the names are stored seems to be similar as with other programs
(EndNote -http://www.endnote.com/support/helpdocs/endnote.zip).

Personally, I also see no gain in going for a relational database in
full normal form where names (b:Person elements) are put in a separate
list. For starters, if you would share the same names across multiple
sources, you would have to create a two-way link: source to one or
more names, and name to one or more sources. The reason for the first
link is obvious: indicating which names participated in the source.
The second link is necessary in case you would remove a source. You
would have to know if a name became obsolete or not. I would suspect
the overhead being greater than the benefits for this case.

Not sure why I'd remove a source ...

Space constraints, access to better sources, overlap, ...

This is scholarship, not a public lending library that only has room
for x number of books on its shelves.

Alternatively, if you would not share the names across sources, and
just use them within one source, it is highly unlikely that one name
will be used multiple times. There are exceptions such as the author
of a book section also being the editor of the book, but in most cases
the 'database' for a single source would actually be larger and
require more processing to obtain the same result.

So there are several options. And what if the
book would be replaced with a film. Then you wouldn't have an author
at all, you would have performers, directors, writers, and producers.

Each one of which is a "name," and which, again, could be associated
with several of the above categories.

You adding a code in front of every element of every entry is pretty
much the same as putting each element in between xml tags. So I guess
you could create the database yourself. You can always create a book
and a journal article through Word 2007 and study the resulting
sources.xml (located at %appdata%\Microsoft\Bibliography) and then
copy paste all other data into that xml file. That should be a bit
faster than copy/pasting everything into a source form for every
entry.

Yves

grammatim schreef:

I have, of course, many articles with bibliographies, and it would be
a lot quicker to insert tab delimiters -- or even a code -- before
each element in each entry in the list, so that the bibliography
database could be created/added to (as if a computer were involved)
without retyping every entry in toto

I should hope the software is smart enough to realize that the
"Author" of a journal article is the same sort of beast as the
"Author" of a book!

On Aug 14, 10:02 am, p0 <yves.dho...@xxxxxxxxx> wrote:
On 14 aug, 13:53, grammatim <gramma...@xxxxxxxxxxx> wrote:

The latest estimate is that I'll have my new computer (with Office 07)
this afternoon ...

Is there a way to convert an existingbibliography, i.e. formatted
list of references, into a Word 2007 table of sources (or whatever
it's called) -- are they, like, like Excel tables, or (heaven forfend)
Access tables, so that something along the lines of tab delimiters
might work?

I'm not really sure what you are trying to accomplish.

Some background: in Word 2007, bibliographic entries are actually
stored inside a custom XML file in the docx. The file, often called
item1.xml, has the following format:

<b:Sources SelectedStyle="\something.xsl" StyleName="A style called
something" xmlns:b="http://schemas.openxmlformats.org/officeDocument/
2006/bibliography" xmlns="http://schemas.openxmlformats.org/
officeDocument/2006/bibliography">
<b:Source>...<b:Source>
<b:Source>...<b:Source>
<b:Source>...<b:Source>
</b:Sources>

where every b:Source element represents a bibliographic source. For a
description of the content of a b:Source element, you can check out
section 7.6 of http://www.ecma-international.org/publications/files/ECMA-ST/Office%2...

Then, when the sources need to be displayed, one of the stylesheets
with the different bibliographic styles (APA, MLA, ...) gets a piece
of XML containing one or more b:Source elements and outputs a piece of
HTML. That HTML is then displayed in Word 2007 as your in-text
citation or bibliography.

Doing the reverse operation (at least that's what I think you try to
do) is not possible. You could try to create your own parser for that
but it seems overly complex to me. How would you expect a parser to be
able to identify the type of a bibliographic entry: Book, BookSection,
JournalArticle, ArticleInAPeriodical, ... And how would you make the
difference between the different contributors of a work: Author,
Artist, Editor, Translator, Writer, Producer, Performer, ...

If references to a work are available online somewhere, it might be
possible to more easily get them that way according to the following
articlehttp://savas.parastatidis.name/2007/01/25/595c0ffb-6595-41bb-9f81-952....
However, I have never seen any code actually doing this and as far as
I know, Microsoft pulled the plug on their academic search.

BR,

Yves
--http://www.codeplex.com/bibliography

.