Revision history for Biber 2.18 (2022-06) * Added support for biblatex "pluralothers" option which allows "et al" to be forced to replace more than one name. * Universal x86_64/ARM64 binary * Support for citation-only rules for name templates in sorting (see \visibility macro in sorting name key templates in biblatex) 2.17 (2022-01) * Tool mode now supports tab indentation in output as well as spaces (see --output-indent) * More compact match/replace syntax for serial items in data maps * Native ARM64 binary for OSX now provided 2.16 (2020-12) * Better support for Unicode on Windows (10 1803+) with the --winunicode option * Option "output_macro_fields" no longer does anything as macro fields are auto-detected. * New options "output_no_macrodefs" and "output_all_macrodefs" control BibTeX macros (@STRING) output for bibtex format output. 2.15 (2020-08) * Added --output-legacy-dates and --output-field-replace options to aid in outputting legacy bibtex format in tool and non-tool mode * UTF-8 filenames supported on Windows * INCOMPATIBLE CHANGE - datasource globbing now has to be turned on explicitly with the --glob-datasources option. Globbing causes some issues with UTF8 encoding detection on some OSes and since it is a niche requirement, it is turned off by default. * DEPRECATION WARNING - The BBLXML output format will be removed in version 4. It is rarely used and complicates test suite maintenance considerably. 2.14 (2019-12) * XDATA now supports granular references so that specific fields/items in a list field may be used and inserted into fields/list positions. This allow, for example, re-use of names from a general XDATA entry containing data about authors. See the biblatex documentation. * Remote Zotero data source requests now send correct header * Annotations are now correctly output for related entries * Better sorting of some name edge cases * New map_entrykey_* mapping verbs * New options --noremove-tmp-dir and --show-tmp-dir to disable of removal on exit and showing location of temporary directory used for intermediate data. Useful for debugging. * Tool mode now accepts multiple data files on the command-line so that data can be spread over several files (e.g. main bib and crossrefs can be kept in separate files) * Biber no longer uses Text::BibTeX/btparse to generation initials from names in normal name parsing mode because it can't handle some Unicode situations. 2.13 (2019-08-13) * biber now exits with error on some very malformed names instead of skipping them as this can lead to data loss in bibliographies. * Support for named annotations 2.12 (2018-10-30) * Added support for biblatex literal and named annotation features * Added support for biblatex \ifnocite test * Case insensitive matching operators in sourcemaps (matchi/notmatchi) * Support for biblatex min/maxsortnames options * Added support for biblatex 'noroman' option to supporess roman numeral interpretation in ranges. * Misc bug fixes 2.11 (2018-02-14) * New entrynocite option for maps which create/clone entries * Misc bug fixes 2.10 (2017-12-19) * Misc bug fixes 2.9 (2017-11-04) * Misc bug fixes 2.8 (2017-11-04) * Updated copyright dates * Support for many new biblatex 3.8 features (see bibatex change doc and PDF manual). * --output-resolve option is now a meta-alias for the more granular --output-resolve-* options - see the biber -? output * The --nouri-encode option is removed. All urls are URL encoded and a new field urlraw is always output with an unencoded version. 2.7 (2016-12-08) * New option (--nouri-encode) to prevent URL encoding of URLs. Avoids breaking display of URLS which contain non-ASCII. * Misc bug fixes 2.6 (2016-09-10) * Improved tool mode bibtex output. This has some normalisation consequences for YEAR and custom fields. See PDF doc. * Support for new \ifuniquebaretitle test * Sourcemaps can now be restricted to particular refsections * Perl 5.24 is now required due to use of postfix deref notation * Removed RIS input driver due to lack of use * Added extended name format - see documentation * Added notfield function to sourcemaps - see documentation 2.5 (2016-05-10) * Added data annotations feature - see biblatex documentation. * Removed the "latinkeys.txt" Unicode::Collate table. This is now pointless because the compiled 'allkeys.txt' is faster and complete. * Incompatible biblatex/biber version is now a fatal error. * Significant performance improvements, particularly in sorting subsystem. * Addition of new bblXML output format and associated RelaxNG XML schema based on active datamodel. * Integration with new biblatex 3.4 labelprefix name change and refcontext option. * Integration with new biblatex 3.4 \DeclareDatafieldSet functionality. * Improved sourcemap looping functionality 2.4 (2016-03-01) * Misc bug fixes * There is now a 64-bit windows build built on windows 10 * Biblatexml datasources now support sourcemapping and have a schema automatically generated from the datamodel. See new biber options 'validate-bltxml' and 'no-bltxml-schema' in --help output. * New functionality in sourcemaps for creating new entries and looping over specified fields. * Sorting key used to sort names is now customisable. See \DeclareSortingNamekeyScheme in biblatex manual. * Support for Zotero RDF/XML and Endnote removed. These were experimental and messy. 2.3 (2015-12-28) * Misc bug fixes 2.2 (2015-10-21) * Tool mode now preserves comments at the end of the file * New ISBN formatting options for biber (--isbn10, --isnb13, --isbn-normalise) * Implemented way of ignoring certain regexps when counting fixed-width string in labels (see biblatex \DeclareNolabelwidthcount macro) * Updated binary builds to perl 5.22 * Cygwin binary build is now handled by Texlive builders * Misc bug fixes 2.1 (2015-05-06) * Misc bug fixes 2.0 (2015-04-20) * Misc bug fixes * New negated "per type" sourcemap verb for negating the application of maps to specific entry types * New "entry clone" sourcemap verb for cloning entries during sourcemapping. See PDF documentation for biber. * Added range length calculation support. Biber now calculates the length of ranges for biblatex and can handle roman numerals, including just about any UTF-8 composed, non-composed or compat versions * Complete re-write of the biblatex use* name functionality now automatically supports all name fields defined in the data model. See biblatex manual * Moved labelname and labeltitle resolution from biber to biblatex for more flexibility and future extensions * Moved to perl 5.20 for all internals and builds 1.9 (2014-05-25) * Misc bug fixes * Option names accept dashes as well as underscores for general TeX option compatibility. * New "--input-directory" option * Tool mode now also supports data validation * Perl 5.18 is now used to build all biber binaries * Locale-sensitive sorting now more granular (see biblatex "sortlocale" documentation) * Source mapping feature now interpolates from previous matches * Now supports generalised separated values lists using new "--xsvsep" option. * Macro <-> UTF-8 conversion now more robust * New Text::BibTeX library now supports generating initials from general UTF-8 strings (for, e.g. Russian names) 1.8 (2013-10-21) * Perl 5.18 compatibility (hash order randomisation) * Removed Data::Diver module due to linux licensing restrictions * Misc Bug Fixes 1.7 (2013-07-07) * Added clrmacros option * Added map_nomatch * Added --tool_macro_fields option * Exposed sorting option to user-level config file (so sorting in tool mode is possible) 1.6 (2013-04-30) * Bugfixes * Now uses perl 5.16. Non-binary (unpacked) use requires perl 5.16+ * Improved tool mode with some formatting options * Made Unicode NFD/NFC boundaries more robust * labelyear is now labeldate to match biblatex 2.6 * Support for new CLONESOURCEKEY field (see biblatex 2.6 docs) * New padding character option in labelalphatemplate (see biblatex 2.6 docs) 1.5 (2013-01-10) * Added new source mapping option "map_entry_null" * Renamed some options (old ones remain as aliases): bibencoding -> input_encoding bblencoding -> output_encoding bblsafechars -> output_safechars bblsafecharsset -> output_safecharsset outfile -> output_file outformat -> output_format * New "tool" mode with "--tool" option which takes a data source as argument and just performs reencoding and source mappings etc. as per options and config file. * Support for customisable BibTeX data source list and name separators * Better debug output for source mapping * Better defined source mapping hierarchy with bugfixes 1.4 (2012-11-28) * Support for biblatex 2.4 style sourcemaps (See \DeclareStyleSourcemap in biblatex docs) * entrykey is now a possible read-only pseudo field in sourcemaps * Reencode data interface completely re-written. New user option --recodedata allows users to specify a new data file for defining reencoding sets. * Internal changes to allow multiple script/lang variants of fields in biblatex in the future. * Citation key aliases now supported more comprehensively, in crossrefs etc. * Support for new RELATEDOPTIONS field (see biblatex docs) * Misc bug fixes 1.3 (01-11-2012) * New "append" mode for sourcemap to allow combining of fields * map_final now possible for map_set_field steps which causes map to stop completely if map_overwrite is false and the field to set already exists * Allow extrayear generation when there is no labelyear 1.2 (20-08-2012) * Misc bug fixes 1.1 (01-08-2012) * Misc bug fixes * Fixed later CentOS libz issue by including latest libz for linux builds * Fixed DOT output (SF bug 3544894) 1.0 (01-07-2012) * Internal data model processing changes * New labeltitleyear and extratitleyear disambiguation tracker * New labeltitle and extratitle disambiguation tracker * Sorting per list enabled for biblatex 2.0 * New noinit option for ignoring parts of names when generating initials * Some optimisations for large data sources * Fixed build issues which broke PAR::Packer binary under latest Debian (Wheezy) 0.9.9 (30-03-2012) * Added sortfirstinits option * HTTPS support for remote data sources * Biber now ignores names when it can't determine minimal information like a lastname. Before, it would throw lots of errors and die. * Strips UTF-8 BOMs before passing to Text::BibTeX otherwise it complains * Merged --graph and --bltxml options into --outformat option * Updated some internal Unicode 6.0 things * Minor bug fixes 0.9.8 (20-12-2012) * Changed config file sourcemap syntax and docs * Support for citekey aliases * New DOT output format 0.9.7 (27-11-2011) * per-datasource mapping restrictions * Completely new config file format to allow for extensions * Fixed --help issue with PAR::Packer * Fixed .bcf XSL inheritance attribute value issue * Moved XDATA output skip into structure where it should have been originally 0.9.6 (13-11-2011) * Fixed binary issue with Readonly::XS not being detected * Fixed bug tracker ticket #3436870 * Match/replace functionality for data source field filtering * New AUCTeX patches for better biblatex/biber support * Now requires perl 5.14 for full Unicode 6.0 support * New XDATA entrytype and field to allow more complex macro-style data inheritance * crossref cascading is in guaranteed resolve order now * Warnings/error summary is now a more consistent format with all other biber messages. * Matching of .bcf keys to datasource keys is now case-sensitive * Fixed biblatex bug tracker ticket #3399744 * Fixed bug tracker ticket #3402114 * max/minnames -> max/mincitenames * max/min* now all per-type and per-entry * uniquelist/uniquename now per-entry * Re-factored recode module with XML data and XSL->HTML transform doc 0.9.5 (18-08-2011) * Fixed auto-bblsafechars bug * refactored explicit "and others" in bibtex data sources * Fixed issue with sorting=none and \nocite{*} 0.9.4 (30-07-2011) * Added support for "SORTSHORTHAND" field * New "map" option to allow user field/entry type mapping (aliasing) in addition to .dcf settings * Made the relevant places depend on max/min{alpha,bib}names instead of just max/minnames * Bug fixes to bblsafechars * Implemented user-customisable labelalpha (not yet in biblatex) * Implemented per-name hashes 0.9.3 (05-06-2011) * New mininit and minfull uniquename options * New minyear uniquelist option * Major revamp of uniqueness code * Moved to new external process calling module because IPC::Run doesn't work on Windows in packed binaries (kpsewhich was returning nothing). 0.9.2 (09-05-2011) * Some internal changes for perl 5.14 * Bug fixes for uniquelist processing 0.9.1 (29-04-2011) * Made temp files all use a global temp dir. * Moved temp .utf8 conversions to temp dir so biber works using .bib files which are in non-writable locations. * Added back an info log message about finding data sources (Lyx compat) 0.9 (29-03-2011) * No longer auto-escapes any TeX markup in bibtex data sources. This is fraught with danger and is a mess. Best not to do it at all. * Supports biblatex \bibnamedelimi macro * Implementation of uniquename and uniquelist options for biblatex 1.4+ 0.8.4 (19-03-2011) * Bug fixes (wrong escaping of "[", "_" and "]" in .bbl) 0.8.3 14-03-2011 * Biber now in TexLive * Added several new binary builds * All name part and element formatting is now abstracted into macros so that it can be controlled via biblatex. * Much more documentation on binary builds for packagers * Bug fixes 0.8.2 (14-02-2011) * Bug fixes 0.8.1 (12-02-2011) * Small bug with default bibtex month macro support * Made remote file support a little more robust (URI return codes) 0.8 (11-02-2011) * Forces biblatex format for month field * Added support for remote data sources * Added experimental support for RIS data sources * Took out -a and -d flags. These are no longer very useful as the increasingly sophisticated biblatex options cannot be set on the command line anyway and so these flags are less and less relevant * New inheritance model in sync with biblatex defaults * Bug fixes 0.7.4 (05-02-2011) * Fix for mincrossrefs leaking across sections (biblatex SF bug #3173032) 0.7.3 (03-02-2011) * Performance tuning for sorting * Text::BibTeX 0.51 for increased macro stack space * Support for old BibTeX month macros by default * Re-write of datasource processing to allow for modular plugins * Generalised nosort option. * Merged LaTeX::Decode into Biber::LaTeX::Recode and added encoding to deal with inputenc's limitations for PDFLaTeX users * new options for forcing/tuning UTF-8 -> LaTeX macro conversion * Bug fixes 0.7.2 (08-01-2011) * Improved latexmk compatibility more 0.7.1 (07-01-2011) * Improved latexmk compatibility * Added "x-ascii" encoding alias for "ascii" to deal with inputenx ascii. * Bug fixes 0.7 (05-01-2011) * Completely re-written sorting routine using Schwartzian transform for real multi-field support and per-field sorting settings. * Bug fixes 0.6.9 (dev beta) * Bug fixes 0.6.8 (dev beta) * Generalised sorting behaviour in several respects * Bug fixes 0.6.7 (dev beta) * Bug fixes 0.6.6 (dev beta) * More generalised sorting options * Bug fixes 0.6.5 (dev beta) * Bug fixes 0.6.4 (dev beta) * Bug fixes 0.6.3 (dev beta) * Outputs labelname name list to the .bbl * Support for dynamic sets and simpler static set interface * Misc bug fixes 0.6.2 (dev beta) * Internal cleanup of structure code * Misc bug fixes 0.6.1 (dev beta) * More comprehensive structural checks on allowed entrytypes/fields and field combinations/formats * Added option validate_structure to turn on structural checks * Renamed option validate -> validate_control * Renamed option output-directory -> output_directory * Added labelname and labelyear as possible sort specifications * Misc bug fixes 0.6 (20-11-2010) * First non-beta release * More abstracted and extensible crossref inheritance model * U::C dependency to 0.67 * sortcase and sortupper options now take same argument forms as in biblatex * Misc bug fixes 0.5.7 (09-10-2010) * Misc bug fixes 0.5.6 (04-10-2010) * Latex character macro decoding is now better defined in the whole (re)encoding workflow. * "inputenc" option is now "bblencoding" * "locale" option is now "sortlocale" * New option "sortupper" * Unicode::Collate sorting is now default as version 0.61 of this has XS module and CLDR tailoring. * Added binary dist build scripts * Removed bundled Module::Build as there are some conflicts with later versions. This is less of a problem anyway because the binary build makes it less necessary to make builds easy on perl<5.10 * Misc bug fixes 0.5.5 (29-08-2010) * Biber::BibTeX moved to Biber::Input::BibTeX. If upgrading, please remove your installed /Biber/BibTeX.pm, just to be tidy * New aliases for EPRINT classes and types implemented (see BibLaTeX docs). * Biber now resolves XREFs as well as CROSSREFs so one less LaTeX pass is needed * Allows UTF-8 citekeys. Requires Text::BibTeX >= 0.46 and a fully UTF-8 aware TeX macro package (e.g. XeTeX or LuaLaTeX). Normal LaTeX with \inputenc will not work because the UTF-8 it uses is incomplete, macro based and not native. * RecDescent parser is no longer supported or included. If upgrading, please remove your installed /Biber/BibTeX directory, just to be tidy * Now requires LaTeX::Decode >= 0.03 * Stricter version dependency for Text::BibTeX. Older than 0.42 is rather useless for complex names. * Re-factored and enhanced auto-re-encoding of .bib/.bbl. With biblatex 0.9c, Biber handles all reencoding. * Misc bug fixes 0.5.4 (04-08-2010) * Case-sensitive sorting option * Multi-section support. Only one biber pass needed to process all bib sections using extended biblatex 0.9b .bbl format. * .aux file is no longer used by biber - all information is passed in .bcf file. * Misc bug fixes * LaTeX::Decode is now a separate CPAN module and is no longer included 0.5.3 (19-03-2010) * Changed semantics of --outfile and --output-directory. --outfile is now relative to --output-directory. If --outfile is absolute, it is stripped to the filename and the path discarded. Absolute .aux paths are stripped to filename for appending to --output-directory. * Updated PRD to use new encoding option names. Now works with -e. * sortlos was not settable to use bib sort order * With Text::BibTeX >= 0.42, will do all name/initials generation with btparse library. * Abstracted BBL output into Biber::Output class for better re-use and potential for output formats other than .bbl. * Synced uniquename .bcf values with change in biblatex 0.9a 0.5.2 (04-03-2010) * Locale fixes * Uniquename fixed to match BibLaTeX option values * Improved error message for malformed names * Fixed config file problems under windows and also added option to ignore config file * Bug with conflicting case between citations and bib entries * Moved pure perl parsename to PRD.pm * Re-write of name parsing to use Text::BibTeX::Name by default * Fields which can be null are explicitly flagged * Add THANKS file * Die early when biber is called without argument * Fixed Mac config path error * .bib parse continues after certain errors * regularised collate options processing * field nameaddon was missing from Biber::Constants * Improved warning message for missing backend=biber * SF bug 2952347 fix (empty date range end) * uniquename counter was being provided incorrectly 0.5.1 (15-02-2010) * Remove support for legacy control files (\jobname-blx.bib): Biber now only reads .bcf files written by biblatex with the option "backend=biber". * Improved handling of "uniquename" and "namehash". * Don't apply latexescape in bbl output, as it breaks math formatting. 0.5 (13-02-2010) * Supports all changes introduced by biblatex 0.9 * Moved all parts relating to the experimental BibLaTeXML format to the "biblatexml" branch on the git repo * Communication between biblatex and Biber can be done through a XML-based config file (extension .bcf): this is enabled with the experimental biblatex option "backend=biber" * Biber can now be customized with a configuration file * Considerable refactoring and cleanup of the code * Further enhancements to the test suite * Improved logging mechanism * More robust parsing of personal names * Ignore certain prefixes and diacritics in sort strings, for instance ʿAlī is sorted under A, and al-Hasan under H. This can be customized in the configuration file (biber.conf). * New command-line option --output-directory (like TeX) * Many, many bugfixes 0.4.3 (23-07-2009) * With biblatex option "backend=biber" and biber option --v2 the new experimental control file \jobname.bcf is used, which brings in many new features, such as per-entrytype settings. * More sophisticated/customizable sorting mechanism * More extensive test suite * Improved BibLaTeXML schema * improved handling of BibLaTeXML (enable nested children, formatting elements, nosort elements) * Do not apply latexescape() by default on xml strings * Skip title handling in BibLaTeXML when not existing * Ignore maxline, add option wraplines * Built-in sort is now default (option --fastsort) * Sorting with Unicode::Collate is possible with option --collate * Make tests succeed w/o Parse::RD and Regexp::Common * Don't output date and urldate in the bbl * Add missing fields "date" "urldate" "organization" "annotator" "foreword" and "eventtitle" * Use Data::Dump instead of Data::Dumper for debugging * Fields "library" and "remarks" are now skipped * Bugfix for aliases (e.g. address, school) * Bugfix: process urldate to get urlyear,urlmonth,urlday * Remove illegal value shorttranslator from labelname option * Changed biberdebug option to debug * Changed option inputencoding to bibencoding 0.4.2 (25-04-2009) * New option --inputencoding to handle 8-bit encoded bib files * $BIBINPUT environment is now searched * Improved parsing of personal names * @comments are now silently ignored * Fixed handling of @preamble; multiple preambles are allowed * Warnings and errors behave more like BibTeX * Alias fields are automatically converted (e.g. address -> location) * Take care of "phdthesis" and "mathesis" types * Cite keys of bib files are now case-insensitive * Default Unicode collation now uses the table "latinkeys.txt" (this is much faster than "allkeys.txt" which contains rules for all Unicode glyphs) * Unicode::Collate parameters can be passed as a command-line option * Faster sorting (with option -f) is also possible for specific locales given with option --locale (this is OS-dependent however) * PRD grammar is now precompiled * bbl output is now wrapped acc to maxline option * Empty fields are no longer printed in the bbl * Sorting is much improved. * Use sortstring separators 0 1 2 like biblatex.bst * Handle truncation of name lists acc to minnames/maxnames * bib2biblatexml has been improved * Windows version of Text::BibTeX provided as ppm. * Many more bugfixes 0.4.1 (31-03-2009) * Text::BibTeX now optional dependency 0.4 (27-03-2009) * Initial beta release