# locators

**URL:** <https://discourse.citationstyles.org/t/locators/430>\
**Category:** CSL Development\
**Created:** [February 17, 2008, 1:30pm UTC](https://discourse.citationstyles.org/t/locators/430 "2008-02-17T13:30:07Z")\
**Posts on this page:** 20\
**Page:** 1

<div class="post-metadata">

**Author:** ![Elena\_Razlogova](https://avatars.discourse-cdn.com/v4/letter/e/c89c15/32.png) [@Elena\_Razlogova](https://discourse.citationstyles.org/u/Elena_Razlogova)\
**Post date:** [February 17, 2008, 1:30pm UTC](https://discourse.citationstyles.org/t/locators/430/1 "2008-02-17T13:30:07Z")

</div>

Multiple locators would also be useful for citing multivolume  
collections–i.e. in a citation below locators are volume 3, page 25.

Luigi Albertini, The Origins of the War of 1914, trans. Isabella  
Massey (New York, NY: Enigma Books, 2005), 3:25.

Is there an existing style that does this correctly?

Thanks,  
Elena

---

<div class="post-metadata">

**Author:** ![Bruce\_D\_Arcus1](https://avatars.discourse-cdn.com/v4/letter/b/4bbf92/32.png) [@Bruce\_D\_Arcus1](https://discourse.citationstyles.org/u/Bruce_D_Arcus1)\
**Post date:** [February 17, 2008, 2:50pm UTC](https://discourse.citationstyles.org/t/locators/430/2 "2008-02-17T14:50:01Z")

</div>

Elena Razlogova wrote:

> Multiple locators would also be useful for citing multivolume  
> collections–i.e. in a citation below locators are volume 3, page 25.
> 
> Luigi Albertini, The Origins of the War of 1914, trans. Isabella  
> Massey (New York, NY: Enigma Books, 2005), 3:25.
> 
> Is there an existing style that does this correctly?

AFAIK, no.

The typical way this is handled currently assumes one point citation  
locator. This isn’t really very good.

I think with some hindsight and the subsequent adding of the macro  
functionality, the correct may to handle this is to probably to use a  
macro and explicitly layout the locators, such that whee constructing a  
citation style, you just do:

```
<text macro="point-locators"/>

```

I’m happy to discuss this change, but we need to give it some thought,  
and make sure to fix a problem that’s already present in Zotero, and  
which got brought over to CSL: the ambiguity of locators (pages, etc.)  
within a cited item (the cited pages, etc.) vs. those that locate the  
item within some larger collection.

BTW, I’ve been working on a web application for storing and creating CSL  
styles. I don’t yet have anything to show, but the basic models includes  
classes like:

```
Style
Context (citation or bibliography)
Macro

```

So, for example, a macro object/row will contain the XML fragment for  
that chunk, in turn validated against the CSL schema.

The idea is to incrementally build up the pieces that make up styles  
around macros, and to make it easy to select them.

The above use case would be a good example of how this would work.

Bruce

---

<div class="post-metadata">

**Author:** ![nicturcan](https://avatars.discourse-cdn.com/v4/letter/n/ba9def/32.png) [@nicturcan](https://discourse.citationstyles.org/u/nicturcan)\
**Post date:** [February 17, 2008, 7:59pm UTC](https://discourse.citationstyles.org/t/locators/430/3 "2008-02-17T19:59:30Z")

</div>

Elena Razlogova wrote:

> Multiple locators would also be useful for citing multivolume  
> collections–i.e. in a citation below locators are volume 3, page 25.
> 
> Luigi Albertini, The Origins of the War of 1914, trans. Isabella  
> Massey (New York, NY: Enigma Books, 2005), 3:25.
> 
> Is there an existing style that does this correctly?
> 
> Thanks,  
> Elena

I don’t know if there is an existing style…

It would be very usefull a possibility for test locators type, for  
putting them at the end (pages), or after the title (sources of ancient  
philosophy).  
Thanks,  
Nicolae

---

<div class="post-metadata">

**Author:** ![Elena\_Razlogova](https://avatars.discourse-cdn.com/v4/letter/e/c89c15/32.png) [@Elena\_Razlogova](https://discourse.citationstyles.org/u/Elena_Razlogova)\
**Post date:** [February 19, 2008, 4:33am UTC](https://discourse.citationstyles.org/t/locators/430/4 "2008-02-19T04:33:44Z")

</div>

> Elena Razlogova wrote:
> 
> > Multiple locators would also be useful for citing multivolume  
> > collections–i.e. in a citation below locators are volume 3, page 25.
> > 
> > Luigi Albertini, The Origins of the War of 1914, trans. Isabella  
> > Massey (New York, NY: Enigma Books, 2005), 3:25.
> > 
> > Is there an existing style that does this correctly?
> 
> I think with some hindsight and the subsequent adding of the macro  
> functionality, the correct may to handle this is to probably to use a  
> macro and explicitly layout the locators, such that whee  
> constructing a  
> citation style, you just do:

Can we make this possible (this seems simpler than nicolae’s  
suggestion of conditionals based on locator type):

> I’m happy to discuss this change, but we need to give it some thought,  
> and make sure to fix a problem that’s already present in Zotero, and  
> which got brought over to CSL: the ambiguity of locators (pages, etc.)  
> within a cited item (the cited pages, etc.) vs. those that locate the  
> item within some larger collection.

Not sure what you mean here.

Thanks,  
Elena

---

<div class="post-metadata">

**Author:** ![Simon\_Kornblith](https://avatars.discourse-cdn.com/v4/letter/s/ce7236/32.png) [@Simon\_Kornblith](https://discourse.citationstyles.org/u/Simon_Kornblith)\
**Post date:** [February 19, 2008, 9:42am UTC](https://discourse.citationstyles.org/t/locators/430/5 "2008-02-19T09:42:46Z")

</div>

I’m not opposed to this, but what happens if the user tries to cite a  
locator type that the style doesn’t specify how to format? Or should  
Zotero only show locator types for which the selected style has a  
 element?

Simon

---

<div class="post-metadata">

**Author:** ![Elena\_Razlogova](https://avatars.discourse-cdn.com/v4/letter/e/c89c15/32.png) [@Elena\_Razlogova](https://discourse.citationstyles.org/u/Elena_Razlogova)\
**Post date:** [February 19, 2008, 1:48pm UTC](https://discourse.citationstyles.org/t/locators/430/6 "2008-02-19T13:48:35Z")

</div>

> > Can we make this possible (this seems simpler than nicolae’s  
> > suggestion of conditionals based on locator type):
> 
> I’m not opposed to this, but what happens if the user tries to cite a  
> locator type that the style doesn’t specify how to format? Or should  
> Zotero only show locator types for which the selected style has a  
> element?

I think most locators (except maybe pages) need a term or a particular  
place in citation defined to be cited correctly, so if we don’t cite  
locators unless specified we won’t lose much.

Elena

---

<div class="post-metadata">

**Author:** ![Bruce\_D\_Arcus1](https://avatars.discourse-cdn.com/v4/letter/b/4bbf92/32.png) [@Bruce\_D\_Arcus1](https://discourse.citationstyles.org/u/Bruce_D_Arcus1)\
**Post date:** [February 19, 2008, 2:17pm UTC](https://discourse.citationstyles.org/t/locators/430/7 "2008-02-19T14:17:18Z")

</div>

forgot to reply-all …

…

> Can we make this possible (this seems simpler than nicolae’s suggestion of  
> conditionals based on locator type):

Are you asking if we can get rid of the cs:label element in favor of a  
an attribute on the cs:text element?

What I was trying to emphasize in this example is that a volume and an  
issue are not point locators, but which I mean specific locations  
within a cited item.

I’m also highlighting that there’s a bug in Zotero’s design that has  
made its way into CSL, and that is that it’s unclear what “page”  
means. Does it mean the numbers of pages in a book (as it does in  
Zotero)? Does it mean the page range of a chapter within a book (as it  
also does in Zotero)? Or does it mean the specific page from which I’m  
extracting a quote (what I mean by a point locator)?

> I’m happy to discuss this change, but we need to give it some thought,  
> and make sure to fix a problem that’s already present in Zotero, and  
> which got brought over to CSL: the ambiguity of locators (pages, etc.)  
> within a cited item (the cited pages, etc.) vs. those that locate the  
> item within some larger collection.
> 
> Not sure what you mean here.

See above.

Bruce

---

<div class="post-metadata">

**Author:** ![Elena\_Razlogova](https://avatars.discourse-cdn.com/v4/letter/e/c89c15/32.png) [@Elena\_Razlogova](https://discourse.citationstyles.org/u/Elena_Razlogova)\
**Post date:** [February 19, 2008, 3:33pm UTC](https://discourse.citationstyles.org/t/locators/430/8 "2008-02-19T15:33:57Z")

</div>

> Are you asking if we can get rid of the cs:label element in favor of a  
> an attribute on the cs:text element?

Can we have both?

> What I was trying to emphasize in this example is that a volume and an  
> issue are not point locators, but which I mean specific locations  
> within a cited item.

A volume number entered by the user when citing a multivolume work  
seems like a point-locator to me:

1. See the article “Feathers,” in Johnson’s Universal Cyclopaedia,  
rev. ed. (New York: A. J. Johnson, 1886), vol. 3.

> **[The Chicago Manual of Style, 17th Edition](https://www.chicagomanualofstyle.org/ch17/ch17_sec129.html)**
>
> Find it. Write it. Cite it. The Chicago Manual of Style Online is the venerable, time-tested guide to style, usage, and grammar in an accessible online format. ¶ It is the indispensable reference for writers, editors, proofreaders, indexers,...

I think it would be easiest to assume that any locator entered by the  
user is a point-locator, whereas any locator passed from Zotero is a  
“regular” locator.

> I’m also highlighting that there’s a bug in Zotero’s design that has  
> made its way into CSL, and that is that it’s unclear what “page”  
> means. Does it mean the numbers of pages in a book (as it does in  
> Zotero)? Does it mean the page range of a chapter within a book (as it  
> also does in Zotero)?

There is a plan to resolve this:  
[https://www.zotero.org/trac/ticket/786](https://www.zotero.org/trac/ticket/786)

> Or does it mean the specific page from which I’m  
> extracting a quote (what I mean by a point locator)?

Pages don’t seem to be any different in this respect than any other  
locator defined in CSL–issue, section, volume, etc.–each of these  
can be passed from Zotero or entered by the user. We could enter  
variable=“page” if it’s a Zotero value, and locator=“page” if it’s the  
user’s value, to distinguish between the two.

Elena

---

<div class="post-metadata">

**Author:** ![Bruce\_D\_Arcus1](https://avatars.discourse-cdn.com/v4/letter/b/4bbf92/32.png) [@Bruce\_D\_Arcus1](https://discourse.citationstyles.org/u/Bruce_D_Arcus1)\
**Post date:** [February 19, 2008, 4:00pm UTC](https://discourse.citationstyles.org/t/locators/430/9 "2008-02-19T16:00:14Z")

</div>

> > Are you asking if we can get rid of the cs:label element in favor of a  
> > an attribute on the cs:text element?
> 
> Can we have both?

Why?

> > What I was trying to emphasize in this example is that a volume and an  
> > issue are not point locators, but which I mean specific locations  
> > within a cited item.
> 
> A volume number entered by the user when citing a multivolume work  
> seems like a point-locator to me:

Hmm … I guess one might argue that, except that neither Zotero, nor  
CSL, nor any other implementation I’m aware of really has any notion  
of sub-document/item content. We don’t have a type for “Page” or  
“Fragment”, for example. So we cite the whole object, and add a  
parameters that specifies the specific context within it.

…

> Pages don’t seem to be any different in this respect than any other  
> locator defined in CSL–issue, section, volume, etc.–each of these  
> can be passed from Zotero or entered by the user. We could enter  
> variable=“page” if it’s a Zotero value, and locator=“page” if it’s the  
> user’s value, to distinguish between the two.

That might work. Simon? Julian? Any opinions on this?

Bruce

---

<div class="post-metadata">

**Author:** ![Elena\_Razlogova](https://avatars.discourse-cdn.com/v4/letter/e/c89c15/32.png) [@Elena\_Razlogova](https://discourse.citationstyles.org/u/Elena_Razlogova)\
**Post date:** [February 21, 2008, 10:00pm UTC](https://discourse.citationstyles.org/t/locators/430/10 "2008-02-21T22:00:05Z")

</div>

> > Pages don’t seem to be any different in this respect than any other  
> > locator defined in CSL–issue, section, volume, etc.–each of these  
> > can be passed from Zotero or entered by the user. We could enter  
> > variable=“page” if it’s a Zotero value, and locator=“page” if it’s  
> > the  
> > user’s value, to distinguish between the two.
> 
> That might work. Simon? Julian? Any opinions on this?
> 
> Bruce

In the interest of moving things along, Simon had posted on this  
earlier:

---

<div class="post-metadata">

**Author:** ![Bruce\_D\_Arcus1](https://avatars.discourse-cdn.com/v4/letter/b/4bbf92/32.png) [@Bruce\_D\_Arcus1](https://discourse.citationstyles.org/u/Bruce_D_Arcus1)\
**Post date:** [February 21, 2008, 10:32pm UTC](https://discourse.citationstyles.org/t/locators/430/11 "2008-02-21T22:32:09Z")

</div>

argh … forgot again …---------- Forwarded message ----------  
From: Bruce D’Arcus \<@Bruce\_D\_Arcus1\>  
Date: Thu, Feb 21, 2008 at 5:31 PM  
Subject: Re: [xbiblio-devel] locators  
To: [zotero-dev@googlegroups.com](mailto:zotero-dev@googlegroups.com)

On Thu, Feb 21, 2008 at 5:00 PM, Elena Razlogova \<@Elena\_Razlogova\> wrote:

> In the interest of moving things along, Simon had posted on this  
> earlier:
> 
> > On Feb 19, 2008, at 12:42 PM, Simon Kornblith wrote:
> > 
> > >
> 
> > > I’m not opposed to this, but what happens if the user tries to cite a  
> > > locator type that the style doesn’t specify how to format? Or should  
> > > Zotero only show locator types for which the selected style has a  
> > > element?
> > 
> > I think most locators (except maybe pages) need a term or a  
> > particular place in citation defined to be cited correctly, so if we  
> > don’t cite locators unless specified we won’t lose much.

I’ve got about 10 things I’m juggling ATM, and it’s felt like a long  
day, so forgive me  
if I’ve lost this discussion.

IIRC, this started with a desire for more complex markup of point  
citation details; being  
able to specify, for example “page 3, paragraph 2” and such.

This then led to recognition of some ambiguity between user-supplied,  
context-specific  
locators on the one hand (page 23 of this article), and generic  
locators that belong to the  
data record per se (pages 20-45 as the page range for an article  
within an issue).

So IIRC, Elena, you were proposing adding a “locator” attribute, which  
I was thinking would  
make the cs:locator element redundant. Your response (“can we have  
both”) suggested  
perhaps not, but I ended up confused.

It might help for those us that are being a bit dim if you could  
summarize your suggestion  
with a quick “we need a, b, c because x, y, z.”

Aside: I have a feeling that using “locator” to refer to the first  
case will be confusing. It is for  
me at least.

Bruce

---

<div class="post-metadata">

**Author:** ![Elena\_Razlogova](https://avatars.discourse-cdn.com/v4/letter/e/c89c15/32.png) [@Elena\_Razlogova](https://discourse.citationstyles.org/u/Elena_Razlogova)\
**Post date:** [February 22, 2008, 5:42am UTC](https://discourse.citationstyles.org/t/locators/430/12 "2008-02-22T05:42:21Z")

</div>

> IIRC, this started with a desire for more complex markup of point  
> citation details; being  
> able to specify, for example “page 3, paragraph 2” and such.  
> yes  
> This then led to recognition of some ambiguity between user-supplied,  
> context-specific  
> locators on the one hand (page 23 of this article), and generic  
> locators that belong to the  
> data record per se (pages 20-45 as the page range for an article  
> within an issue).  
> yes
> 
> So IIRC, Elena, you were proposing adding a “locator” attribute, which  
> I was thinking would  
> make the cs:locator element redundant. Your response (“can we have  
> both”) suggested  
> perhaps not, but I ended up confused.

actually, you had asked:

> > > Are you asking if we can get rid of the cs:label element in favor  
> > > of a  
> > > an attribute on the cs:text element?

We’d still need labels to add terms (“p”, “pp”, etc) to locators added  
by the user. We won’t need the variable “locator”–is that what you’re  
asking?

> It might help for those us that are being a bit dim if you could  
> summarize your suggestion  
> with a quick “we need a, b, c because x, y, z.”

you summarize everything nicely above

> Aside: I have a feeling that using “locator” to refer to the first  
> case will be confusing.

point-locator then

> I’ve got about 10 things I’m juggling ATM

me too

Elena\> It is for

---

<div class="post-metadata">

**Author:** ![Bruce\_D\_Arcus1](https://avatars.discourse-cdn.com/v4/letter/b/4bbf92/32.png) [@Bruce\_D\_Arcus1](https://discourse.citationstyles.org/u/Bruce_D_Arcus1)\
**Post date:** [February 26, 2008, 8:43pm UTC](https://discourse.citationstyles.org/t/locators/430/13 "2008-02-26T20:43:00Z")

</div>

Sorry I keep losing this thread …

Elena Razlogova wrote:

…

> > So IIRC, Elena, you were proposing adding a “locator” attribute, which  
> > I was thinking would  
> > make the cs:locator element redundant. Your response (“can we have  
> > both”) suggested  
> > perhaps not, but I ended up confused.
> 
> actually, you had asked:
> 
> > > > Are you asking if we can get rid of the cs:label element in favor  
> > > > of a  
> > > > an attribute on the cs:text element?
> 
> We’d still need labels to add terms (“p”, “pp”, etc) to locators added  
> by the user. We won’t need the variable “locator”–is that what you’re  
> asking?

Ah, my bad.

So now we have something like:

This is nice and concise (no need to explicitly account for each kind of  
locator), but breaks down if you need to do multiple. The question is  
how to get the best of both worlds? I’m actually a little stuck on this.

In other words, what exactly would you be adding to the Chicago style,  
say, to get the effect you’d like to achieve?

Bruce

---

<div class="post-metadata">

**Author:** ![Elena\_Razlogova](https://avatars.discourse-cdn.com/v4/letter/e/c89c15/32.png) [@Elena\_Razlogova](https://discourse.citationstyles.org/u/Elena_Razlogova)\
**Post date:** [February 27, 2008, 1:05am UTC](https://discourse.citationstyles.org/t/locators/430/14 "2008-02-27T01:05:11Z")

</div>

> So now we have something like:

This doesn’t work for Chicago with the current “one locator” setup  
anyway, because every locator requires a label except page. Because  
page is the most common locator I have to exclude the label and all  
other locators are cited incorrectly.

> This is nice and concise (no need to explicitly account for each  
> kind of  
> locator), but breaks down if you need to do multiple. The question is  
> how to get the best of both worlds? I’m actually a little stuck on  
> this.
> 
> In other words, what exactly would you be adding to the Chicago style,  
> say, to get the effect you’d like to achieve?

Using previous examples, with this macro:

Will let you cite both of these (with locators volume 3, page 25):

Luigi Albertini, The Origins of the War of 1914, trans. Isabella  
Massey (New York, NY: Enigma Books, 2005), vol. 3.

Luigi Albertini, The Origins of the War of 1914, trans. Isabella  
Massey (New York, NY: Enigma Books, 2005), 3:25.

or, in Nicolae’s example, the following macros (with locators book IX,  
chapter III, paragraph 2, page 50-52):

Will let you cite both of these:

Herodot, History, IX, III, 2, edition by…, Place, 2007.

Turcan, /Title about Herodot/, Bucharest, Humanitas Publisher, 2008,  
pp. 50-52.

In theory, we could keep locator variables and have something like  
this to make sure all locators get cited:

Any of the above will require complicated conditionals but I don’t see  
how that can be avoided.

Best,  
Elena

---

<div class="post-metadata">

**Author:** ![Simon\_Kornblith](https://avatars.discourse-cdn.com/v4/letter/s/ce7236/32.png) [@Simon\_Kornblith](https://discourse.citationstyles.org/u/Simon_Kornblith)\
**Post date:** [February 27, 2008, 5:08am UTC](https://discourse.citationstyles.org/t/locators/430/15 "2008-02-27T05:08:20Z")

</div>

Barring a Zotero bug, the following should work:

Simon

---

<div class="post-metadata">

**Author:** ![Elena\_Razlogova](https://avatars.discourse-cdn.com/v4/letter/e/c89c15/32.png) [@Elena\_Razlogova](https://discourse.citationstyles.org/u/Elena_Razlogova)\
**Post date:** [February 27, 2008, 8:36am UTC](https://discourse.citationstyles.org/t/locators/430/16 "2008-02-27T08:36:59Z")

</div>

> > > So now we have something like:
> > 
> > This doesn’t work for Chicago with the current “one locator” setup  
> > anyway, because every locator requires a label except page. Because  
> > page is the most common locator I have to exclude the label and all  
> > other locators are cited incorrectly.
> 
> Barring a Zotero bug, the following should work:

Thanks! I didn’t know this was supposed to work. There is indeed a bug  
however:

I tried this:

```
     <group prefix=", ">
        <choose>
       <if locator="page" match="none">
	  <label variable="locator" form="short" suffix=". "/>
        </if>
   </choose>
        <text variable="locator"/>
 </group>

```

And got this error (and no label):

Error: citation is not defined  
Source File: chrome://zotero/content/xpcom/cite.js  
Line: 1313

I created a ticket.

Best,  
Elena

---

<div class="post-metadata">

**Author:** ![Bruce\_D\_Arcus1](https://avatars.discourse-cdn.com/v4/letter/b/4bbf92/32.png) [@Bruce\_D\_Arcus1](https://discourse.citationstyles.org/u/Bruce_D_Arcus1)\
**Post date:** [February 27, 2008, 3:57pm UTC](https://discourse.citationstyles.org/t/locators/430/17 "2008-02-27T15:57:48Z")

</div>

OK …

Elena Razlogova wrote:

> Using previous examples, with this macro:

So looking at the schema, we currently have within the conditional  
pattern this:

```
 ## A conditional on the locator for this specific entry
 attribute locator { cs-terms.locator }?,

```

So it can’t be referenced outside that context.

We thus need to create a “locator” pattern so that it can be reused in  
cs:text.

We also need to make sure to document it correctly in the different  
contexts.

So would this work?

1. generic pattern:

2. change conditional to use this pattern:

3. add pattern to text:

This is more complex, but I’ll work it out. I just need some  
documentation that will unambiguously explain how this would work, and  
which works across style classes (number, author-date, etc.).

Finally, a question that goes this issue of definition:

What becomes the difference between this:

\<cs:text variable=“volume”/\>

… and:

\<cs:text locator=“volume”/\>

..? Do we remove the locators values as option for “variable”?

Bruce

---

<div class="post-metadata">

**Author:** ![Simon\_Kornblith](https://avatars.discourse-cdn.com/v4/letter/s/ce7236/32.png) [@Simon\_Kornblith](https://discourse.citationstyles.org/u/Simon_Kornblith)\
**Post date:** [February 28, 2008, 7:51am UTC](https://discourse.citationstyles.org/t/locators/430/18 "2008-02-28T07:51:37Z")

</div>

> So looking at the schema, we currently have within the conditional  
> pattern this:
> 
> ```
> ## A conditional on the locator for this specific entry
> attribute locator { cs-terms.locator }?,
> 
> ```
> 
> So it can’t be referenced outside that context.
> 
> We thus need to create a “locator” pattern so that it can be reused in  
> cs:text.
> 
> We also need to make sure to document it correctly in the different  
> contexts.
> 
> So would this work?
> 
> 1. generic pattern:
> 
> 2. change conditional to use this pattern:
> 
> 3. add pattern to text:
> 
> This is more complex, but I’ll work it out. I just need some  
> documentation that will unambiguously explain how this would work, and  
> which works across style classes (number, author-date, etc.).

Sounds fine to me.

> Finally, a question that goes this issue of definition:
> 
> What becomes the difference between this:
> 
> \<cs:text variable=“volume”/\>
> 
> … and:
> 
> \<cs:text locator=“volume”/\>
> 
> …? Do we remove the locators values as option for “variable”?

It seems like the former should show up in the bibliography as  
information about the source, whereas the latter would be a “point  
locator” as we’ve been calling them and show up in the citation.

Simon

---

<div class="post-metadata">

**Author:** ![Elena\_Razlogova](https://avatars.discourse-cdn.com/v4/letter/e/c89c15/32.png) [@Elena\_Razlogova](https://discourse.citationstyles.org/u/Elena_Razlogova)\
**Post date:** [February 28, 2008, 11:03am UTC](https://discourse.citationstyles.org/t/locators/430/19 "2008-02-28T11:03:42Z")

</div>

> > So looking at the schema, we currently have within the conditional  
> > pattern this:
> > 
> > ## A conditional on the locator for this specific entry
> > 
> > attribute locator { cs-terms.locator }?,
> > 
> > So it can’t be referenced outside that context.
> > 
> > We thus need to create a “locator” pattern so that it can be reused  
> > in  
> > cs:text.
> > 
> > We also need to make sure to document it correctly in the different  
> > contexts.
> > 
> > So would this work?
> > 
> > 1. generic pattern:
> > 
> > 2. change conditional to use this pattern:
> > 
> > 3. add pattern to text:
> > 
> > This is more complex, but I’ll work it out. I just need some  
> > documentation that will unambiguously explain how this would work,  
> > and  
> > which works across style classes (number, author-date, etc.).
> 
> Sounds fine to me.
> 
> > Finally, a question that goes this issue of definition:
> > 
> > What becomes the difference between this:
> > 
> > \<cs:text variable=“volume”/\>
> > 
> > … and:
> > 
> > \<cs:text locator=“volume”/\>
> > 
> > …? Do we remove the locators values as option for “variable”?
> 
> It seems like the former should show up in the bibliography as  
> information about the source, whereas the latter would be a “point  
> locator” as we’ve been calling them and show up in the citation.

Yes, we need both. Variables would also show up in a citation, in a  
different place than locators:

variable:  
John Donne, The Variorum Edition of the Poetry of John Donne, ed.  
Gary A. Stringer, vol. 6 (Bloomington: Indiana University Press,  
1995), 5

locator:  
John Donne, The Variorum Edition of the Poetry of John Donne, ed. Gary  
A. Stringer (Bloomington: Indiana University Press, 1995), vol. 6.

Elena

---

<div class="post-metadata">

**Author:** ![Bruce\_D\_Arcus1](https://avatars.discourse-cdn.com/v4/letter/b/4bbf92/32.png) [@Bruce\_D\_Arcus1](https://discourse.citationstyles.org/u/Bruce_D_Arcus1)\
**Post date:** [February 28, 2008, 12:35pm UTC](https://discourse.citationstyles.org/t/locators/430/20 "2008-02-28T12:35:21Z")

</div>

Elena Razlogova wrote:

…

> > > ..? Do we remove the locators values as option for “variable”?  
> > > It seems like the former should show up in the bibliography as  
> > > information about the source, whereas the latter would be a “point  
> > > locator” as we’ve been calling them and show up in the citation.
> 
> Yes, we need both. Variables would also show up in a citation, in a  
> different place than locators:

We need to make this distinction more clear. Your first case is still a  
locator.

Maybe:

## A descriptor that locates sub-content within a cited resource. Used

## in some styles to indicate specific page numbers for excerpted

## content, for example.

point-locator.att = attribute point-locator { cs-terms.locator }

We then leave cs:text@variable alone I guess.

Bruce

[Next page](https://discourse.citationstyles.org/t/locators/430.md?page=2)