with contributions from Frank G. Bennett, Jr. and Bruce D’Arcus.
Table of Contents
- CSL 1.0 Release Notes
- Project Home(s)
- Updating CSL 0.8 Styles
- Version Number
- Options as Attributes
- Style Metadata
- Citation Collapsing
- Ordinal Numbers
- Original Publisher
- Rich Text Markup within Fields
- Stripping Periods
- Year Suffix
- Layout Control with the Display Attribute
- Back-References in Note Styles
Citation Style Language 1.0 represents a significant update of the Citation Style Language (CSL), an open XML language to describe citation styles. This document describes the changes compared to CSL 0.8 (released in March 2009).
Nota Bene These notes include many XML examples to demonstrate new CSL features. Please note that these examples are reduced to a minimum, and often don’t constitute valid CSL styles. In addition, when references are made to the XML elements of CSL, the CSL namespace (“cs”) is always attached as a prefix (e.g. cs:citation for the <citation/> element). However, for CSL styles it is customary to declare the default CSL namespace on the root cs:style element (<style xmlns="http://purl.org/net/xbiblio/csl"/>), which eliminates the need to include this namespace for each element.
The first change is that CitationStyles.org is the new home of the CSL project (the old home was located at http://xbiblio.sourceforge.net/csl/). The new website will be the central location to find CSL schemas, documentation and styles.
For CSL development the SourceForge xbiblio mailing list will be kept in use. Code hosting has moved around a bit. A switch from SourceForge.net to Bitbucket in 2010 has been followed by a move to GitHub in 2011.
Prior to the release of CSL 1.0, CSL documentation was rather scarce. Most of what was available was hosted on the Zotero wiki. With CSL 1.0, we improved on this situation. In addition to these upgrade notes, a full specification has been made available. As before, the schema itself is also a source of information, although reading the schema requires some understanding of RELAX NG Compact, the XML schema language in which the CSL schema is written.
Whereas the CSL 0.8 schema consists of a single file, the CSL 1.0 schema has been split into the following files:
- csl.rnc - The core of the CSL schema. This file contains most of the schema logic and calls the other schema files
The main advantage of splitting up the schema is that the schema will be easier to maintain. If you wish to validate styles against the CSL schema, make sure all the files are located in the same directory, and validate against csl.rnc.
The CSL 1.0 schema has been extended with two Schematron rules to make sure styles don’t use cs:text and cs:key elements that call non-existing cs:macro elements. Note that not all validators support embedded Schematron code (e.g. Jing just ignores the rules). In addition, the CSL 1.0 schema can now be used to validate the locales files (e.g. “locales-en-US.xml”), which contain localizations of terms, date formats and style options.
CSL 1.0 is backward incompatible with CSL 0.8, which means that CSL 0.8 styles don’t work with CSL 1.0 processors. Fortunately, it is possible to (automatically) update CSL 0.8 styles to the CSL 1.0 format using the upgrade.xsl XSLT stylesheet. This conversion has been performed for all the styles in the Zotero Style Repository. However, if you use (custom) CSL styles that aren’t included in this style repository, you might need to do this yourself.
First, check whether the styles that you wish to update validate against the CSL 0.8.1 schema.
Then use an XSLT processor to update the styles. Available options are the command-line tools Saxon and xsltproc. Alternatively, one of the (more user-friendly) online converters, such as the one offered by www.shell-tools.net can be used. For the latter tool, the instructions are:
- Paste the contents of upgrade.xsl into the “xslt” text box at http://www.shell-tools.net/index.php?op=xslt
- Paste the contents of the CSL 0.8 style into the “xml” text box
- Click the “Submit Query”-button
- Copy the text from the “output” text box to a suitable text editor (e.g. Notepad on Windows) and save the file with a .csl-extension.
Finally, after the conversion, it is recommended to validate the converted style, this time against the CSL 1.0 schema.
Starting with CSL 1.0, styles (and locales files) must indicate the CSL version with which they are compatible. All CSL 1.0 styles should include the version attribute with the value “1.0” on the cs:style element, e.g.:
<style version="1.0" class="in-text"/>
For “locale-xx-XX.xml” files this attribute should be set on the root cs:locale element.
In CSL 0.8, citation- and bibliography-specific style options were set with cs:option elements. In CSL 1.0 this element is no longer used. Instead, options are set using attributes. CSL 1.0 includes global options, which affect the output of both the citation and bibliography sections, and which are set as attributes on the cs:style element. Options that are citation-specific are now set on the cs:citation element, while bibliography-specific options are set on the cs:bibliography element.
The cs:category element has two purposes: to indicate for which field(s) of study a style is relevant (e.g. “biology”) and to indicate the format of the style (e.g. “author-date”). In CSL 0.8 the term attribute was used for both cases. With CSL 1.0, term has been replaced with two attributes: citation-format, to indicate the citation format, and field, to indicate the field of study. An example:
<style> <info> <category citation-format="author-date"/> <category field="biology"/> </info> </style>
In CSL 0.8 the possible citation formats were: “author-date”, “label”, “note”, “numeric” and “in-text”. In CSL 1.0 “in-text” has been replaced with “author” (a format that only shows author names in in-text citations, like the MLA style).
ISSN-identifiers unambiguously identify journals. While CSL 0.8 allowed only a single ISSN identifier to be included in the style metadata section, CSL 1.0 now supports multiple ISSNs (e.g. the ISSNs of the print and online editions of a journal), as well as the relatively new ISSN-L identifier. For example:
<style> <info> <issn>0099-2240</issn> <eissn>1098-5336</eissn> <issnl>0099-2240</issnl> </info> </style>
The “locales-xx-XX.xml” files (with “xx-XX” indicating the locale, e.g. “en-US” for “English - United States”) that are part of CSL previously had only the role of supplying localized terms. With CSL 1.0, these locale files also contain localized style options and localized date formats. Because of this, some changes have been made to the XML format of these files: cs:locale has replaced terms as the root element, the xml:lang and xmlns attributes are now applied to the cs:locales element, and several new elements have been introduced for the localization of dates and options (cs:date, cs:date-part and cs:style-options).
As before, it is possible to use the cs:locale element in styles to override any content of the “locales-xx-XX.xml” files. The cs:locale element can be used with or without the xml:lang attribute. If xml:lang is not set, the contents of the cs:locale element will be used for all locales. If xml:lang is set to a locale code, the content of the cs:locale element will override the content of the specified locale. N.B. a cs:locale element with the xml:lang attribute takes priority over a cs:locale element without the attribute. For example,
<style> <locale> <terms> <term name="et-al">et alii</term> </terms> </locale> <locale xml:lang="en"> <style-options punctuation-in-quote="true" /> <terms> <term name="et-al">and others</term> </terms> <date form="text"> <date-part name="month" suffix=" " form="short"/> <date-part name="day" suffix=", "/> <date-part name="year"/> </date> </locale> </style>
with regard to the “et-al” term, this will result in the use of “and others” for the English locales, and of “et alii” for all other locales.
CSL 1.0 introduces support for date localization. This feature is optional: styles can still define dates in the usual non-localized format. To use a localized date, all you need to do is use cs:date with the form attribute set to either text (for dates like ‘April 21, 2008’) or numeric (e.g. ‘4/21/08’). As demonstrated in the example below, it is not necessary to specify any cs:date-part elements for localized dates:
<style> <bibliography> <layout> <!-- old-fashioned, unlocalized date --> <date variable="accessed"> <date-part name="year"/> <date-part name="month" form="numeric" prefix="-"/> <date-part name="day" prefix="-"/> </date> <!-- default localized date --> <date variable="accessed" form="numeric"/> </layout> </bibliography> </style>
The format of localized dates (i.e. punctuation and the order of the date-parts) is specified in the “locales-xx-XX.xml”. As with terms, localized date formats can be overridden within styles. In addition, localized dates can be customized within a style via two options. First, the date-parts attribute can be added to cs:date to control which date-parts are shown. With the default value of year-month-day the whole date is shown. With year-month and year only the year/month and year date-parts are shown, respectively. The second option is the ability to redefine how one or more cs:date-part element are formatted. Note that the order of cs:date-part elements for a localized date within cs:layout doesn’t affect the rendering order of the date-parts (this in contrast with non-localized dates or dates specified within cs:locale, where the order of the cs:date-part elements does control the rendering order). Neither does the presence or absence of cs:date-part elements affect which date-parts are shown (this is controlled via the date-parts attribute described above). Instead, cs:date-part elements allow you to override specific properties of the localized date-parts (e.g. the form attribute of the month-date-part can be set to “short”). Note that changes made in this way affect all locales. An example illustrating the different options:
<style> <!-- a modified date format for the English locale --> <locale xml:lang="en"> <date form="text"> <date-part name="month" suffix=" " form="short"/> <date-part name="day" suffix=", "/> <date-part name="year"/> </date> </locale> <bibliography> <layout> <!-- localized date that only shows the year and month --> <date form="text" date-parts="year-month"/> <!-- localized date in numeric format with leading zeros --> <date form="numeric"> <date-part name="month" form="numeric-leading-zeros"/> <date-part name="day" form="numeric-leading-zeros"/> </date> </layout> </bibliography> </style>
In developing CSL 1.0 it was recognized that robust date localization requires a clear distinction between style-specific and locale-dependent formatting of dates. As a result, some limitations have been placed on the use of cs:date when used for localized dates. First, affixes (prefixes and suffixes) on cs:date are considered style-specific formatting (e.g. parentheses around the date: “(2000)”). It is therefore not allowed to apply affixes to cs:date when this element is used within cs:locale (in both styles and “locales-xx-XX.xml” files). Instead, all locale-specific affixes should be applied to the cs:date-part elements. Conversely, it is not allowed to apply affixes to cs:date-part elements when the parent cs:date calls a localized date. Secondly, cs:date may not carry the delimiter attribute when used in a style to call a localized date. In CSL 1.0 this attribute can be used to specify a delimiter for the date-parts, which is considered locale-specific formatting.
N.B. When creating a localized date format, consider graceful scaling of dates when applying affixes to the cs:date-part elements. As an example, consider the date format “May 1, 2008”. By using the following arrangement of affixes, correct dates are obtained for any value of the date-parts attribute:
<date form="text"> <date-part name="month" suffix=" "/> <date-part name="day" suffix=", "/> <date-part name="year"/> </date>
|“year-month-day”||“May 1, 2008”|
In addition to localized dates and terms, CSL 1.0 now also supports localized options (although for now, there is only one such an option, punctuation-in-quote). The default value of localized options is set for each locale in the “locales-xx-XX.xml” files, but these values can be overridden using the cs:style-options element within cs:locale in a CSL style. An example:
<style> <locale xml:lang="en"> <style-options punctuation-in-quote="true"/> </locale> </style>
To prevent localization of styles (which might be desirable for journal-specific styles) the default-locale attribute can be included on the cs:style element (this attribute already existed in CSL 0.8, but was not supported by Zotero). Its value should be a locale code (e.g. “fr-FR” for French). An example:
N.B. With CSL 0.8 there was some confusion about the use of default-locale, and some style authors included the xml:lang attribute instead. In CSL 1.0 xml:lang is no longer allowed as an attribute on cs:style.
CSL 1.0 offers finer control of citation collapsing. First, two new options have been introduced, both of which are set as attributes on cs:citation: year-suffix-delimiter, which defines the delimiter for subsequent year suffixes (e.g. the comma in “Doe 2000a,b, Smith 1999”), and after-collapse-delimiter, which defines the delimiter between a group of collapsed citations and the subsequent citation (e.g. the semicolon in “Doe 2000a, b; Smith 1999, Williams 2002”). Both attributes default to the delimiter set on the cs:layout element within cs:citation. Secondly, “year-suffix-ranged” has been added as a possible value of the collapse attribute of cs:citation. If collapse is set this value, citations are collapsed as with “year-suffix”, but ranges of year-suffixes are collapsed as well (e.g. “Doe 2000a,b,c,e” would become “Doe 2000a-c,e”). An example of how these attributes are set:
<style> <citation collapse="year-suffix-ranged" year-suffix-delimiter="," after-collapse-delimiter=";"> <layout delimiter=", " /> </citation> </style>
CSL 1.0 includes two new terms, ad and bc (Anno Domini and Before Christ). These terms are automatically appended to years: bc is added to negative years (e.g. 2500BC), while ad is added to positive years of less than four digits (79AD).
CSL 1.0 adds support for date ranges. By default, date ranges are delimited by an en-dash (e.g. May–July 2008). Custom delimiters can be set on the cs:date-part elements with the new range-delimiter attribute. The attribute value set on the largest date-part (“day”, “month” or “year”) that differs between the two dates of the date range will then be used instead of the en-dash. For example,
<style> <citation> <layout> <date variable="issued"> <date-part name="month" suffix=" "/> <date-part name="year" range-delimiter="/"/> </date> </layout> </citation> </style>
would result in “May–July 2008” and “May 2008/June 2009”.
CSL 1.0 includes four new season terms, season-01 to season-04 (respectively Spring, Summer, Autumn and Winter). If a date includes a season instead of a month, the season term will substituted the month date-part. E.g.,
<style> <citation> <layout> <date variable="issued"> <date-part name="month" suffix=" "/> <date-part name="year"/> </date> </layout> </citation> </style>
would result in “May 2008” and “Winter 2009”.
Two new features of CSL 1.0 allow for special formatting of uncertain dates. First, CSL 1.0 introduces the is-uncertain-date conditional. This conditional tests “true” when a date is flagged as uncertain. The second addition is the new “circa” term. For example,
<style> <citation> <layout delimiter="; "> <choose> <if is-uncertain-date="issued"> <text term="circa" form="short" suffix=" "/> </if> </choose> <date variable="issued"> <date-part name="year"/> </date> </layout> </citation> </style>
would result in “2005” (certain date) and “ca. 2003” (uncertain date).
The delimiter-precedes-last attribute on cs:names controls the use of the name delimiter between the last and next-to-last name in name lists. In CSL 0.8, this attribute could be set to either always or never. To include the delimiter for lists of three or more names (“Doe, Smith, and Williams”) and to exclude it for lists of only two names (“Doe and Smith”), you would have to leave out the attribute. Now, in CSL 1.0, it is also possible to explicitly set the last behavior by using the value “contextual”.
CSL 1.0 includes a new name variable, editorial-director. This addition is mostly specific to French, where the “Directeur de la publication” role is common.
If a cs:names element has its name attribute set to “editor translator” (or “translator editor”), CSL 1.0 collapses both name lists when their contents is identical. If a label is specified, and collapsing occurs, the newly added editortranslator term is used. For example,
<names variable="editor translator"> <name /> <label form="short" prefix=" (" suffix=")" /> </names>
could result in “John Doe (ed. & trans.)”.
A new element, cs:et-al, can now be included within cs:names. This adds two important features. First, formatting can now be set independently for the et-al term. For example,
<names variable="author"> <name/> <et-al font-style="italic" prefix=" "/> </names>
results in “Doe et al.”. Secondly, it is now possible to use two different et-al terms within a single style (e.g. one for in-text citations and one for the bibliography). The desired term, “et-al” (the default) or “and others”, is set with the name attribute on the cs:et-al element. For example,
<names variable="author"> <name/> <et-al term="and others" prefix=" "/> </names>
would yield “Doe and others” (note that both terms are localized).
CSL 1.0 introduces the ability to separately format given and family names. Formatting is specified via the new cs:name-part element, a child of cs:name. The name attribute of this element should be set to either “family” or “given”. Note that the order of name-part elements does not affect the order in which the name parts are shown. An example, resulting in names like “John SMITH”:
<names variable="author"> <name form="long"> <name-part name="family" text-case="uppercase"/> </name> </names>
In CSL 0.8, any attribute used for name formatting had to be included for each occurrence of the cs:names element, even if names were identically formatted for all these elements. To reduce the need for duplication, CSL 1.0 introduces inheritable options: the attributes and, delimiter-precedes-last, initialize-with, name-as-sort-order and sort-separator can now also be set on cs:style, cs:citation and cs:bibliography. The attributes form and delimiter have been made available as name-form and name-delimiter, respectively, as the original attribute names have different uses when set on cs:style, cs:citation and cs:bibliography. Similarly, the names-delimiter attribute has been introduced as a companion of the delimiter attribute on cs:names.
When a name attribute is set on cs:style, cs:citation or cs:bibliography, its value is used for all cs:names elements within the element carrying the attribute. However, when an element lower in the hierarchy carries the same attribute with a different value, this value will override the value(s) specified higher in the hierarchy.
In addition to these changes, CSL 1.0 also includes more fine-grained control for et-al settings. The attributes et-al-min, et-al-use-first, et-al-subsequent-min, et-al-subsequent-use-first now behave like any other cs:name attribute, and thus can be set on cs:style, cs:citation, cs:bibliography and cs:name.
A new attribute, initialize-with-hyphen, can be set on cs:style to control hyphenation of compound given names (e.g. “Jean-Luc Picard”). When set to “true” (the default), a hyphen is added when the given name is initialized (“J.-L. Picard”). With “false” the hyphen is left out (“J.L. Picard”).
Many Western names consist not only of a given and a family name, but also of one or more name particles. To control how these particles are handled, CSL 1.0 introduces a new option, demote-non-dropping-particle, which can be set as an attribute on cs:style. To understand how this option works, it is important to recognize that not all particles are equal: name particles can be either kept or dropped when only the surname is shown (from now on we will refer to these two types as non-dropping-particle and dropping-particle, respectively). A single name can contain particles of both types (in this case the non-dropping-particle always comes after the dropping-particle). For example, the French name “Gérard de la Martinière” can be deconstructed into:
When only the surname is shown, only the non-dropping-particle is kept: “La Martinière”. However, when names are inverted (with the family name preceding the given name), styles often differ in name particle handling. First, the non-dropping-particle can be either prepended to the family name (e.g. “de Koning, W.”) or appended (after initials or given names, e.g. “Koning, W. de”). Note that the dropping-particle is always appended in inverted names. Secondly, if the non-dropping-particle is prepended to the family name, names can be sorted in two ways: the non-dropping-particle can remain part of the family name (as part of the primary sort key; example A), or it may be separated from the family name and become (part of) a secondary sort key, joining the dropping-particle, if available (example B). The different sort orders are illustrated below:
Sort order A: non-dropping-particle not demoted
- primary sort key: “la Martinière”
- secondary sort key: “de”
- tertiary sort key: “Gérard”
Sort order B: non-dropping-particle demoted
- primary sort key: “Martinière”
- secondary sort key: “de la”
- tertiary sort key: “Gérard”
The demote-non-dropping-particle attribute can be set to the following values:
- “never”: the non-dropping-particle is treated as part of the family name, whereas the dropping-particle is appended (e.g. “de Koning, W.”, “la Martinière, Gérard de”). The non-dropping-particle is part of the primary sort key (example A, e.g. “de Koning, W.” appears under “D”).
- “sort-only”: as “never”, with the exception that non-dropping-particle is demoted to a secondary sort key (see example B, e.g. “de Koning, W.” appears under “K”).
- “display-and-sort” (default): the dropping and non-dropping-particle are appended to the rest of the name (e.g. “Koning, W. de” and “Martinière, Gérard de la”). When names are sorted, both particles are part of the secondary sort key (see example B, e.g. “Koning, W. de” appears under “K”).
To allow for localization of ordinal numbers, CSL 1.0 includes the new terms ordinal-01 to ordinal-04. For the en-US locale, these terms have the values “st”, “nd”, “rd” and “th” (resulting in ordinal numbers of “1st”, “2nd”, “3rd”, “4th”, etc.). In addition, support for long ordinals has been introduced with the terms long-ordinal-01 to long-ordinal-10 (“first”, “second”, ..., “tenth”). Long ordinals can be selected by using cs:number and setting the form attribute to “long-ordinal”.
Sometimes (older) books are republished by a different publisher. To indicate the original publisher, and the location of the original publisher, CSL 1.0 adds two new variables, original-publisher and original-publisher-place. Note that CSL 0.8 already included the name variable original-publisher, which could only be used with cs:names. The variables original-publisher and original-publisher-place in CSL 1.0 are ‘normal’ variables, and can be used with cs:text.
CSL 1.0 introduces two new page variables: “page-first” and “number-of-pages”. The existing variable “pages” is still used for page ranges (e.g. of journal articles and book chapters). The variable “page-first” holds the first page of the page range. The variable “number-of-pages” is used to indicate the total number of pages of an item (e.g. a book or thesis).
In addition, a new global (non-localized) option, page-range-format, has been added to control the collapsing of page ranges. This attribute, set on cs:style, can have the values “expanded” (e.g. “321-328”), “minimal” (“321-8”), and “chicago” (“321-28”, which follows the collapsing rules of the Chicago Manual of Style). When the attribute isn’t present, the content of the “page”-variable is shown as is. An example:
<style page-range-format="chicago"> <bibliography> <layout> <text variable="page"/> </layout> </bibliography> </style>
In CSL 1.0 small changes have been made to the use of plural. As with CSL 0.8, this attribute can be set on cs:text and cs:label. When used on cs:text, plural can still be set to “false” to use the singular form of a term (the default), or to “true” to use the plural form. But when used on cs:label, different values are now available. With “contextual” (the default value), the plurality of the variable determines whether the singular or plural form of the term is used (e.g. “page 43” and “pages 3-5”). With “never” and “always” respectively the singular or plural form of the term is used, regardless of the plurality of the variable. An example:
<number variable="edition" form="ordinal"/> <text term="edition" plural="false"/> <group> <label variable="page" plural="always"/> <text variable="page"/> </group>
N.B. The plural attribute was one of the few cases where the implementation in Zotero did not follow the CSL 0.8 schema.
CSL 1.0 introduces new terms for inner (“open-inner-quote” and “close-inner-quote”, e.g. and ) and outer quotes (“open-quote” and “close-quote”, e.g. and ). Together with the new punctuation-in-quote-option (see Punctuation around Quotes), quotes applied with the quotes attribute are now fully localized.
If a field (e.g. a variable) contains a matching set of quotation marks (”, ‘, or the quotation marks defined by the open-inner-quote and close-inner-quote terms), then these quotation marks are replaced by those defined by the open-quote and close-quote terms. For example:
<text value="Voyage of 'The Beagle'"/>
will render as: Voyage of The Beagle
Flipflopping occurs when a field (e.g. a variable) contains a matching set of quotation marks (”, ‘, or the quotation marks defined by the open-quote, close-quote, open-inner-quote and close-inner-quote terms), or when it contains markup for italics or boldfacing (see Rich Text Markup within Fields). For example:
<text prefix="Speak, " value="'friend'" suffix=", and enter" quotes="true"/>
will render as Speak friend, and enter. Quotes flipflop between inner (open-inner-quote and close-inner-quote) and outer (open-quote and close-quote) quotes. Italics flipflop between italics and the normal font-style, and boldface between bold and the normal font-weight.
The localized option punctuation-in-quote is used to specify whether punctuation (commas and periods) should appear within (“true”, e.g. for American English) or outside quotation marks that have been applied by the style (“false” (default value), e.g. for British English). As such, it can toggle the style output between
Douglas Adams, “The Hitchhiker’s Guide to the Galaxy,” 1979.
Douglas Adams, “The Hitchhiker’s Guide to the Galaxy”, 1979.
Although not part of the CSL 1.0 specification, the new citeproc-js CSL processor used by Zotero supports an exciting new feature: the ability to use rich text markup within item fields. This markup is applied with a small set of HTML(-like) tags:
- <b> - bold
- <i> - italics
- <sc> - small-caps
- <sub> - subscript
- <sup> - superscript
E.g. if a Zotero item has the title “Ca<sup>2+</sup> levels in <i>Homo sapiens</i>”, this will render as “Ca2+levels in Homo sapiens”. Rich text markup can also be used with the value attribute of cs:text, but here special XML characters (“<”, “>”) have to be escaped, e.g.:
<text value="<b>some bold text</b>"/>
In contrast, markup used with the prefix and suffix attributes is not recognized. Finally, note that bold and italics markup are subject to flipflopping.
The second-field-align attribute can be used to align any subsequent lines of a bibliography entry with the beginning of the second field. In CSL 0.8 the value of this attribute could be set to “true” or “margin” to place the first field respectively in the margin, or flush against it. In CSL 1.0 “true” has been renamed to “flush”.
A new attribute, strip-periods, can now be set on cs:date-part (only for name=”month”), cs:label and cs:text. The attribute is inactive when set to “false” (the default value), but if set to “true”, any periods are stripped from the variable contents. strip-periods replaces the include-period attribute that was part of CSL 0.8.
strip-periods is especially useful for journal abbreviations. There are plans to improve support for journal abbreviations in future versions of CSL (e.g. by using lookup lists to find the correct journal abbreviation given a certain journal title), but for now it is recommended that users include periods for journal abbreviations in their (Zotero) libraries. With the help of strip-periods, styles can then either use the journal abbrevation as is, or use a version without periods. An example:
<text variable="container-title" form="short" strip-periods="true"/>
would output “Appl Environ Microbiol” if the journal abbreviation for the Zotero item is “Appl. Environ. Microbiol.”.
Year-suffixes are included automatically when the disambiguate-add-year-suffix attribute on cs:citation is set to “true”. However, some styles desire special markup of year-suffixes, such as italics (e.g. “2000a, b”). For this CSL 1.0 introduces the year-suffix variable, which can be used to explicitly specify the location and formatting of year-suffixes. An example:
<style> <citation> <layout delimiter=", "> <date variable="issued"> <date-part name="year"/> </date> <text variable="year-suffix" font-style="italic"/> </layout> </citation> </style>
CSL 0.8 included a display attribute, with possible values of “block” or “inline-block”, intended to provide some control over the layout of bibliography entries. However, it remained unimplemented by any known processor, and was not used in any known styles. In CSL 1.0, the display attribute has been refined and extended: it is now restricted to rendering-element children of cs:layout under cs:bibliography, and has possible values of “block”, “left-margin”, “right-inline”, and “indent”.
By leveraging the styling features of the target rendering platform (HTML, a word processor, a document processing system), the enhancements to display permit the implementation of sophisticated formatting effects, such as publication listings headed by the name of each author. See the CSL 1.0 Specification for further details on the use of this attribute.
The disambiguation algorithm specified in CSL 0.8 followed the Chicago Manual of Style. CSL 1.0 supports additional disambiguation methods through the addition of a new givenname-disambiguation-rule attribute, which can be used in combination with the existing disambiguate-add-names and disambiguate-add-givenname attributes.
The givenname-disambiguate-rule option accepts values of “all-names”, “all-names-with-initials”, “primary-name”, “primary-name-with-initials”, and “by-cite” [*]. The first value specifies the Chicago Manual of Style method, which assures that all names included in citations uniquely identify the relevant author. The second does the same, but will not expand initialized names. The third and fourth values specify analogous methods, but here the transformation of names is limited to the first-listed name. The last option transforms names only as necessary to uniquely identify references listed in the bibliography. A more detailed discussion of the disambiguation options can be found in the CSL 1.0 Specification.
|[*]||A hat tip to user komrade of the Zotero forums, whose review of the major style guides led to this set of disambiguation options.|
CSL 1.0 includes several new features to allow for more complex reference sorting. The first change is that the form attribute on cs:name can now be set to “count”. With this value, the enclosing cs:names returns the number of names in the name variable instead of the names themselves. When used for a sort key, this makes it possible to sort according to the number of authors (or any other kind of contributor). An example:
<style> <macro name="author"> <names variable="author"> <name form="count"/> </names> </macro> <bibliography> <sort> <key macro="author"/> </sort> <layout/> </bibliography> </style>
The second change consists of two new attributes for the cs:key element, names-min and names-use-first. These attributes, when set, override the values of et-al-min and et-al-min-first, respectively. The following example shows how with these attributes a bibliography can be sorted alphabetically, while only taking the first author into account:
<style> <macro name="author"> <names variable="author"> <name/> </names> </macro> <bibliography> <sort> <key macro="author" names-min="1" names-use-first="1"/> </sort> <layout/> </bibliography> </style>
CSL 1.0 adds two features related to back-referencing in note styles. First, the position conditional supports a new value, “near-note”. It tests true when an item has been previously cited, and the distance between the current and most recent use (measured in number of footnotes or endnotes) does not exceed the value of the new near-note-distance attribute. This attribute, which has a default value of 5, may be set on cs:citation. An example:
<style class="note"> <citation near-note-distance="3"> <layout> <choose> <if position="near-note"> ... </if> <else> ... </else> </choose> </layout> </citation> </style>
N.B. The value of “near-note” is always false for references that are not in a footnote/endnote.
The second feature is the new first-reference-note-number variable. When an item has been previously cited, this variable holds the number of the first note to cite the item.
CSL 1.0 introduces a number of new variables:
CSL 1.0 also introduces a number of new terms:
- “ad” and “bc”. See AD and BC.
- “author”. This term can be applied as a label to author names. For most styles, this term will consist of an empty string (“”).
- “editortranslator”. See Editor/Translator Name Collapsing.
- “ordinal-01`` to “ordinal-04” and “ordinal-01” to “ordinal-10”. See Ordinal Numbers.
- “original-publisher” and “original-publisher-place”. See Original Publisher.
- “reference”. This term replace the “references” term from CSL 0.8.
- “season-01” to “season-04”. These terms map to respectively spring, summer, fall and winter. See Seasons.