Official Zotero added syncing of full text index tables this week. Today, MLZ follows suit: we are now level with Zotero version 4.0.14. That’s the main news, but there have been a few other bugfixes to MLZ in the past two weeks, which are summarised below. Most are related to multilingual functionality, which is starting to attract the attention of style developers.
- Terms in multilingual sort keys
- Sort keys were being generated with the default locale of the style always. To get a correct sort in multi-layout styles, terms from the layout’s master locale (instead of the global default locale) must be used. Thanks to Ming-Li Wang for diagnosing the problem.
- Automatic CSL-m mode
- Force CSL-m mode on valid CSL-m styles (i.e. those with version 1.1mlz1).
- Fix title-short outside of MLZ
- In MLZ, the content of the Short Title field is ignored when <text variable=”title” form=”short”/> is used on the legal_case type. This is what we want in MLZ, but it violates the official CSL schema. The behaviour has been rectified for official CSL styles run outside of MLZ.
- Test of genre field content
- To resolve an MLZ-specific issue relating to localisation of the email, interview, podcast, radio-broadcast, and television-broadcast types, a test for content automatically inserted into the genre field by these types has been added to the CSL-m schema, and implemented in the processor.
- Labels on nested name elements
- The processor was quashing the labels on names elements nested within a macro. This has been fixed.
- Redundant calls to substituteStart() and substituteEnd()
- The processor was adding redunant calls to substituteStart() and substituteEnd() on macro blocks. There were no known side-effects, but the unnecessary calls have been removed.
- Multilingual supplement affixes
- A bug that erroneously trimmed the suffix on the style element when a multilingual element with affixes is added has been fixed.
- Add default-locale-sort attribute (CSL-m only)
- CSL-m styles with multiple locale-specific layouts require a global setting for the sort locale. The new attribute supplies this value and the processor makes use of it.
- Sort fixes
- There were some infelicities in the processor’s sort function. These have been fixed.
- Recognize full stops inside quotes for term capitalisation purposes, and leading full stops on a prefix for delimiter purposes.
- Greek entries
- Extend the pattern match used for converting names to initials to cover Greek names. The family of recognised Greek characters has also been extended.