The following settings are not exposed in the UI, but can be found under Preferences/Advanced/Config editor.
All are prefixed with extensions.zotero.translators.better-bibtex. in the table you will find there
default: <not set>
If you have unicode turned on you can still selectively replace some characters to plain-text commands; any characters entered here will always be replaced by their LaTeX-command counterparts.
default: 1
If you have auto-exports set up, BBT will wait this many seconds before actually kicking off the exports to buffer multiple changes in quick succession setting off an unreasonable number of auto-exports. Minimum is 1 second. Changes to this preference take effect after restarting Zotero.
default: 10
Number of seconds to wait after your system goes idle before kicking off auto-exports.
default: 4
If cache priming is kicked off (see autoExportPrimeExportCacheThreshold), the cache will be primed in batches of autoExportPrimeExportCacheBatch. The lower this number is, the longer the priming will take (there’s a fixed per-prime overhead), but (since the priming uses exports under the hood) the larger it is, the longer the priming actions will lock the UI.
default: 100
Cache priming (see autoExportPrimeExportCacheThreshold), happens in a tight loop, which will still make Zotero sluggish while the priming runs. By default, BBT will wait a tenths of a second between prime batches;- longer waits mean a more responsive Zotero during the priming, but the priming will take longer to complete.
default: 0
If an auto-export is triggered and there are more than autoExportPrimeExportCacheThreshold un-cached entries, prime the cache before starting the auto-export. This makes the export take longer in total, but since exports in Zotero lock up the UI, priming helps making the actual export run as fast as possible. Really only useful for large exports, but what counts as large is system-dependent, so you can play with this value. The default of 0 disables priming. This is a temporary measure until https://groups.google.com/d/msg/zotero-dev/lHYEtdgPHCE/nDUyxpFmAgAJ is implemented in Zotero.
default: no
When on, BBT will automatically pin the first citekey it generates for an item.
default: yes
Support for EDTF dates in biblatex
default: 5
How often the Better BibTeX database should be saved to disk. Defaults to once every 5 seconds. Note that your database is always saved when your computer goes idle, or when you exit Zotero.
default: no
Replaces the “note” field with the bibtex key during citation rendering in Word/Libreoffice. Main use-case is to help migrating word documents to pandoc. This setting only takes effect during startup, so if you change it, you will have to restart Zotero to have this take effect (or to disable it. Please disable it when done). You will want to use a custom CSL style (such as this) to make this work.
default: <not set>
if you set csquotes to a string of character pairs, each pair will be assumed to be the open and close parts of a pair and will be replaced with a \\enquote{...} construct.
default: config
Can be ‘off’, ‘config’ or ‘always’
default: 100
I’ve had reports where Zotero notifies extensions that references have changed, but if BBT then actually retrieves those same references, Zotero complains they “haven’t been saved yet”. Super. This preference sets the number of microseconds BBT should wait after being notified before acting on the changed references.
default: no
When on, BBT will load kuroshiro for romajization in citekeys. This uses a lot of memory, easily 100MB. If you don’t have Japanese titles/names, keep this off.
default: no
BBT locks the UI during startup because I have been told in no uncertain terms I am not to touch the Zotero database before I get an all-clear from Zotero. This all-clear takes a fair amount of time. As BBT needs database access for generating keys, and everything in BBT depends on the keys being present, it is absolutely safest to make sure BBT initialization has completed before freeing the UI. I want to stress that during most of the lock-time, BBT is simply waiting for Zotero to complete its own initialization; try to do an export of any kind (not just BBT) or to import new references directly after Zotero has started and you’ll notice that it may take a while before Zotero reacts. The lockout just puts a face on this hidden init, and prevents nasty race conditions between the BBT and Zotero initialization leading to unpredictable breakage occasionally.
By default you only get a flaoting popup to tell you how far along this process Zotero is; if you really want to see how long BBT has to wait before it can do anything, you can enable the lockout during startup, but it seems fairly safe now if you’re patient enough to wait until the floating window goes away. There is absolutely nothing I can do about the long time it takes before BBT moves from “Waiting for Zotero”. After that, initialization is very quick.
default: <not set>
Any characters entered here will prefer a math-mode LaTeX-command counterpart over a math-mode, if a math-mode command is available. Only useful when mapUnicode is conservative.
default: <not set>
Any characters entered here will prefer a text-mode LaTeX-command counterpart over a math-mode, if a text-mode command is available. Only useful when mapUnicode is conservative. Characters specified in mapMath take presedence over characters specified in mapText.
default: conservative
When a unicode character can be exported as either a math-mode or text-mode command, map to:
conservative: if both a math-mode and a text-mode mapping is available, stay in the mode of the previously mapped character if possible. This minimizes the number of generated $s in the output.text: if both a math-mode and a text-mode mapping is available, prefer text.math: if both a math-mode and a text-mode mapping is available, prefer math.default: yes
Name particle handling. Only turn on when requested and we’re talking about it on github.
default: #LaTeX
When an item has this tag, all its fields will be assumed to hold raw LaTeX and will undergo no further transformation. If you set this to *, all items will be assumed to have raw LaTeX.
default: no
When exporting a Bib(La)TeX file, if the attachments are stored anywhere under the directory the bibliography is exported to, use relative paths to those attachments. Caching is disabled when this option is on, so it affects performance.
default: no
Finds potential problems in the database and fixes those. This runs extremely slow at startup, don’t enable this unless explicitly asked to.
default: a,ab,aboard,about,above,across,after,against,al,along,amid,among,an,and,anti,around,as,at,before,behind,below,beneath,beside,besides,between,beyond,but,by,d,da,das,de,del,dell,dello,dei,degli,della,dell,delle,dem,den,der,des,despite,die,do,down,du,during,ein,eine,einem,einen,einer,eines,el,en,et,except,for,from,gli,i,il,in,inside,into,is,l,la,las,le,les,like,lo,los,near,nor,of,off,on,onto,or,over,past,per,plus,round,save,since,so,some,sur,than,the,through,to,toward,towards,un,una,unas,under,underneath,une,unlike,uno,unos,until,up,upon,versus,via,von,while,with,within,without,yet,zu,zum
list of words to skip in title when generating citation keys
default: no
If you’re dead-set on ignoring both BibTeX/BibLaTeX best practice and the Zotero recommendations on title/sentence casing, set this preference to “true” to suppress automatic brace-protection for words with uppercase letters.
default: no
On import, BBT will add case-protection (<span class=“nocase”>…<span>) to titles that have words in {Braces}. There’s plenty of bib(la)tex files out there that do this a little overzealously, and you may not like the resulting HTML code in your items, even though this is what the braces mean in bib(la)tex, and Zotero supports it. If you turn this off, the markup is omitted during import.
default: no
Bib(La)TeX references ought to be stored in Title Case; Zotero references are expected to be entered as sentence-case. With this option on, BBT will try to sentence-case during import. This sentence-casing uses heuristics, no natural language processing is performed, and the results are not perfect. You can suppress this, but you may then also want to enable suppressTitleCase (which has its own problems, see the help entry for that option on this page).
default: no
If you’re dead-set on ignoring both BibTeX/BibLaTeX best practice and the Zotero recommendations on title/sentence casing, set this preference to “true” to suppress title casing for English references.
default: url,doi,file,eprint,verba,verbb,verbc
list of fields to treat as verbatim during import. If you’re importing e.g. Mendeley-generated BibTeX, try removing file from this list before import.