Institutional names

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Institutional names

fbennett
Chris Maloney has a thread on the citeproc-js tracker concerning names handling:

    https://bitbucket.org/fbennett/citeproc-js/issue/173/fix-institutional-authors-jiggery-pokery

As the title says, the issue is a request to turn off (some of) the
institutional handling in citeproc-js by default. I suspect this will
be a welcome thing for everyone generally, but it would change default
output, so I'd like to run it past the list before I release anything.

The behaviour is currently triggered by names set as a "literal"
element in a CSL JSON name object. It has two effects.

First, the processor recognizes such names as institutions, which
enables a couple of special operations on them in the CSL-m schema.
Specifically, if the string contains multiple elements separated by
pipe characters, the elements to be rendered can be controlled by
attributes in the CSL-m; and both individual elements and the entire
name string as a whole can be abbreviated through the Abbreviation
Filter. These features are unintrusive WRT official CSL, and important
for legal materials (at least). I'd like to retain them, and they do
not seem to be an issue.

Second, the processor treats personal names that precede an
institution name as affiliated with the institution. This yields (for
example) "Smith & Jones, IBM" rather than "Smith, Jones & IBM."
Authors at the end of the list are treated as unaffiliated authors,
and are moved to the front in the rendered citation.

It is this second aspect of institutional names handling that Chris
strongly suggests should be turned off. To be honest, just writing the
description in the paragraph immediately above kind of made be cringe,
and I do suspect there will be a chorus of agreement. But as I say, I
want to be sure.

For reference, the original description of the name-rearrangement
logic, in input examples etc., is here:

    http://citationstylist.org/docs/citeproc-js-csl.html#cs-institution-extension

Bedtime here now, and an early morning at work tomorrow; but I'll look
forward to picking up responses later in the day if anyone has views
on this.

Frank

------------------------------------------------------------------------------
Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
from Actuate! Instantly Supercharge Your Business Reports and Dashboards
with Interactivity, Sharing, Native Excel Exports, App Integration & more
Get technology previously reserved for billion-dollar corporations, FREE
http://pubads.g.doubleclick.net/gampad/clk?id=190641631&iu=/4140/ostg.clktrk
_______________________________________________
xbiblio-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/xbiblio-devel
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Institutional names

rmzelle
Administrator
I'd like to see better institutional name support in future CSL
releases, but I agree it makes sense to turn off the feature by
default for vanilla citeproc-js processing.

Rintze

On Thu, Feb 19, 2015 at 10:52 AM, Frank Bennett <[hidden email]> wrote:

> Chris Maloney has a thread on the citeproc-js tracker concerning names handling:
>
>     https://bitbucket.org/fbennett/citeproc-js/issue/173/fix-institutional-authors-jiggery-pokery
>
> As the title says, the issue is a request to turn off (some of) the
> institutional handling in citeproc-js by default. I suspect this will
> be a welcome thing for everyone generally, but it would change default
> output, so I'd like to run it past the list before I release anything.
>
> The behaviour is currently triggered by names set as a "literal"
> element in a CSL JSON name object. It has two effects.
>
> First, the processor recognizes such names as institutions, which
> enables a couple of special operations on them in the CSL-m schema.
> Specifically, if the string contains multiple elements separated by
> pipe characters, the elements to be rendered can be controlled by
> attributes in the CSL-m; and both individual elements and the entire
> name string as a whole can be abbreviated through the Abbreviation
> Filter. These features are unintrusive WRT official CSL, and important
> for legal materials (at least). I'd like to retain them, and they do
> not seem to be an issue.
>
> Second, the processor treats personal names that precede an
> institution name as affiliated with the institution. This yields (for
> example) "Smith & Jones, IBM" rather than "Smith, Jones & IBM."
> Authors at the end of the list are treated as unaffiliated authors,
> and are moved to the front in the rendered citation.
>
> It is this second aspect of institutional names handling that Chris
> strongly suggests should be turned off. To be honest, just writing the
> description in the paragraph immediately above kind of made be cringe,
> and I do suspect there will be a chorus of agreement. But as I say, I
> want to be sure.
>
> For reference, the original description of the name-rearrangement
> logic, in input examples etc., is here:
>
>     http://citationstylist.org/docs/citeproc-js-csl.html#cs-institution-extension
>
> Bedtime here now, and an early morning at work tomorrow; but I'll look
> forward to picking up responses later in the day if anyone has views
> on this.
>
> Frank
>
> ------------------------------------------------------------------------------
> Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
> from Actuate! Instantly Supercharge Your Business Reports and Dashboards
> with Interactivity, Sharing, Native Excel Exports, App Integration & more
> Get technology previously reserved for billion-dollar corporations, FREE
> http://pubads.g.doubleclick.net/gampad/clk?id=190641631&iu=/4140/ostg.clktrk
> _______________________________________________
> xbiblio-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/xbiblio-devel

------------------------------------------------------------------------------
Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
from Actuate! Instantly Supercharge Your Business Reports and Dashboards
with Interactivity, Sharing, Native Excel Exports, App Integration & more
Get technology previously reserved for billion-dollar corporations, FREE
http://pubads.g.doubleclick.net/gampad/clk?id=190641631&iu=/4140/ostg.clktrk
_______________________________________________
xbiblio-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/xbiblio-devel
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Institutional names

fbennett
In reply to this post by fbennett
As per the earlier note, I have followed the suggestion of Chris
Maloney (and probably the quiet hope of many others), and disabled
name reordering by default in citeproc-js. The change takes effect
from tag 1.0.559.

    https://bitbucket.org/fbennett/citeproc-js/issue/173/fix-institutional-authors-jiggery-pokery

There is another small change in names handling in this release: I
noticed that decorations (italics etc) were being applied across name
lists, including separating punctuation. I have adjusted things so
that decorations are applied instead to individual names in the list,
which is surely more correct.

If anyone needs the old name-reordering behavior, it can be enabled by
setting the following option on an instantiated citation processor:

    citeproc.sys.opt.development_extensions.spoof_institutional_affiliations
= true;

For posterity, name reordering is documented here:

    http://citationstylist.org/docs/citeproc-js-csl.html#cs-institution-extension

Frank


On Fri, Feb 20, 2015 at 2:47 PM, Frank Bennett <[hidden email]> wrote:

> Andrew,
>
> Yep.l I need to write out the full story sometime soon, but standard
> CSL has the use cases you describe pretty well covered. If a name is
> to be always written in a fixed form, it can be fed to the processor
> in the "family" field of the CSL JSON schema, with nothing in the
> "given" field. That will render as a personal name, but never change
> form.
>
> (MLZ and the CSL-m schema provide more flexibility, with name
> formatting parameters optionally specified for specific languages.)
>
> Frank
>
>
> On Fri, Feb 20, 2015 at 10:27 AM, Andrew Dunning <[hidden email]> wrote:
>> Thanks for your continued work on citeproc-js!
>>
>> One thing I wanted to verify (forgive me if I am pointing out the painfully obvious): it sounded from the thread as if it is only non-Western names that are considered as not fitting into the usual model. Even in Western history, however, pre-modern figures are not typically treated as having a family/given name; for example, you would have ‘William of Malmesbury’ and ‘Martianus Capella’ in an index (not ‘Malmesbury, William of’ and ‘Capella, Martianus’) so character detection won’t be helpful in such cases. But I am speaking as a Zotero/Pandoc user, and not someone familiar with the citeproc-js internals.
>>
>> All best,
>>
>> Andrew Dunning
>>
>>> On 19 Feb 2015, at 10:52 AM, Frank Bennett <[hidden email]> wrote:
>>>
>>> Chris Maloney has a thread on the citeproc-js tracker concerning names handling:
>>>
>>>    https://bitbucket.org/fbennett/citeproc-js/issue/173/fix-institutional-authors-jiggery-pokery
>>>
>>> As the title says, the issue is a request to turn off (some of) the
>>> institutional handling in citeproc-js by default. I suspect this will
>>> be a welcome thing for everyone generally, but it would change default
>>> output, so I'd like to run it past the list before I release anything.
>>>
>>> The behaviour is currently triggered by names set as a "literal"
>>> element in a CSL JSON name object. It has two effects.
>>>
>>> First, the processor recognizes such names as institutions, which
>>> enables a couple of special operations on them in the CSL-m schema.
>>> Specifically, if the string contains multiple elements separated by
>>> pipe characters, the elements to be rendered can be controlled by
>>> attributes in the CSL-m; and both individual elements and the entire
>>> name string as a whole can be abbreviated through the Abbreviation
>>> Filter. These features are unintrusive WRT official CSL, and important
>>> for legal materials (at least). I'd like to retain them, and they do
>>> not seem to be an issue.
>>>
>>> Second, the processor treats personal names that precede an
>>> institution name as affiliated with the institution. This yields (for
>>> example) "Smith & Jones, IBM" rather than "Smith, Jones & IBM."
>>> Authors at the end of the list are treated as unaffiliated authors,
>>> and are moved to the front in the rendered citation.
>>>
>>> It is this second aspect of institutional names handling that Chris
>>> strongly suggests should be turned off. To be honest, just writing the
>>> description in the paragraph immediately above kind of made be cringe,
>>> and I do suspect there will be a chorus of agreement. But as I say, I
>>> want to be sure.
>>>
>>> For reference, the original description of the name-rearrangement
>>> logic, in input examples etc., is here:
>>>
>>>    http://citationstylist.org/docs/citeproc-js-csl.html#cs-institution-extension
>>>
>>> Bedtime here now, and an early morning at work tomorrow; but I'll look
>>> forward to picking up responses later in the day if anyone has views
>>> on this.
>>>
>>> Frank
>>>
>>> ------------------------------------------------------------------------------
>>> Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
>>> from Actuate! Instantly Supercharge Your Business Reports and Dashboards
>>> with Interactivity, Sharing, Native Excel Exports, App Integration & more
>>> Get technology previously reserved for billion-dollar corporations, FREE
>>> http://pubads.g.doubleclick.net/gampad/clk?id=190641631&iu=/4140/ostg.clktrk
>>> _______________________________________________
>>> xbiblio-devel mailing list
>>> [hidden email]
>>> https://lists.sourceforge.net/lists/listinfo/xbiblio-devel
>>

------------------------------------------------------------------------------
Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
from Actuate! Instantly Supercharge Your Business Reports and Dashboards
with Interactivity, Sharing, Native Excel Exports, App Integration & more
Get technology previously reserved for billion-dollar corporations, FREE
http://pubads.g.doubleclick.net/gampad/clk?id=190641631&iu=/4140/ostg.clktrk
_______________________________________________
xbiblio-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/xbiblio-devel
Loading...