BIB2GLS(1)

🌈 bib2gls — BIB2GLS(1) 🌈

✨ SYNOPSIS ✨

bib2gls [option]+ auxfile

💅 DESCRIPTION 💅

okay so bib2gls is literally designed to work with the glossaries-extra LaTeX package. (The record package option is required — no cap.) This app does two functions in one: (1) selects entries from .bib files according to information found in the auxfile (similar to bibtex); (2) hierarchically sorts entries and collates location lists (similar to makeindex and xindy). The .aux extension may be omitted from auxfile.

The auxfile (and corresponding .log file) should either be in the current directory or in the directory specified by --dir. Bib files can either be relative to the directory the auxfile is in or in a location that can be found by kpsewhich.

---

🎀 OPTIONS 🎀

• --aux-input-action setting — Determines what to do when bib2gls encounters \@input in the aux file. The setting may be one of:
  • follow — Follow the reference (that is, input the file as usual).
  • skip-after-bibglsaux (or skip after bibglsaux) — Follow the reference for each \@input encountered until \@bibgls@input is encountered, after which skip all remaining instances of \@input (default).
  • skip — Skip all instances of \@input (but not \@bibgls@input, which is always followed).

The \@bibgls@input command will be in the aux file if the package option bibglsaux has been used.

• --break-space — Interpret ~ (tilde) character and \nobreakspace as a normal space.
• --cite-as-record — Treat \citation as an ignored record.
• --collapse-same-location-range — Collapse an explicit range that has a duplicate start and end location into a normal record (default).
• --custom-packages list — Instruct the TeX parser library to attempt to parse the packages listed in list. This is intended for simple custom packages that don't contain complex code.
• --datatool-sort-markers — The interpreter will use the datatool-base.sty marker commands to match their localised definitions within \DTLsortwordlist (datatool v3.0+). You will need a sort method that doesn't ignore the control characters used by those commands.
• --date-in-header (or -D) — The header comment at the start of the .glstex file will include the file modification date.
• --debug [n] — Switch on the debug mode at the given setting n, which should be a non-negative integer. If n is 0 then the debug mode is switched off. If omitted, n is set to 1, which corresponds to --debug-mode io. Any positive number will enable all of bib2gls's debugging messages. The amount of TeX Parser Library debugging messages depends on bitwise operation of n.
• --debug-mode setting — May be used instead of --debug where the level number is difficult to remember. The setting may be a comma-separated list of any of the keywords:
  • all — Highest level (likely to result in a large transcript file). Equivalent to switching on all settings.
  • catcode — TeX Parser category code changes.
  • cs — TeX Parser command definitions.
  • decl — TeX Parser declarations.
  • expansion — TeX Parser expansions (may result in a large transcript file).
  • expansion-list — TeX Parser stack expansions (may result in a large transcript file).
  • expansion-once — TeX Parser one-level expansions.
  • expansion-once-list — TeX Parser one-level list expansions.
  • io — TeX Parser I/O operations (open, close and fetch token).
  • popped — TeX Parser stack popping.
  • process — TeX Parser macro process (may result in a large transcript file).
  • process-generic-cs — TeX Parser generic command process.
  • process-stack — TeX Parser stack process (may result in a large transcript file).
  • process-stack-list — TeX Parser detailed stack process (may result in a large transcript file).
  • read — TeX Parser codepoint read from file (likely to result in a large transcript file).
  • sty-data — TeX Parser data (typically relates to internal objects that don't have an analogous TeX macro). This includes data gathered from the aux file.

For example: bib2gls --debug-mode io,catcode,sty-data mydoc

• --default-encoding name — Set the default character encoding for input and output files. If omitted, the JVM default encoding is assumed.
• --dir directory (or -d <directory>) — Files are relative to directory. (This doesn't change the current working directory.)
• --expand-fields — Don't write \glsnoexpandfields to the .glstex file.
• --force-cross-resource-refs or -x — Force cross-resource referencing mode on.
• --group or -g — Add group field to entries.
• --help or -h — Display help message and exit.
• --ignore-packages list (or -k list) — Don't parse the log file for the packages listed in list. Note that --packages overrides this option, so if the same package is listed in both --ignore-packages and --packages then the interpreter will check if it's supported. This option has a cumulative action. Only known packages may be included in list.
• --interpret — Use the TeX Parser Library to interpret known (La)TeX commands (default).
• --list-known-packages — Lists all the packages that are known to the TeX parser library and then exits (with exit code 0). Any of the listed packages may be used in --packages or --ignore-packages.
• --locale lang (or -l lang) — Use language resource file for lang (which should be an IETF locale tag).
• --log-encoding name — Set the character encoding for the log files (which may not be the same as the encoding for the document LaTeX files).
• --log-file file (or -t <file>) — Set the transcript file.
• --merge-nameref-on rule — Rule for merging locations created with the record=nameref package option (requires glossaries-extra.sty version 1.37+). The rule may be one of: hcounter, href, title or location.
• --merge-wrglossary-records — Merge an entry's wrglossary records for the same page locations. (For use with the indexcounter package option.)
• --mfirstuc-math-protection — Switch on the auto-insertion of an empty group for math-shift ($).

If mfirstuc v2.08+, glossaries v4.50+ and glossaries-extra v1.49+ are detected, the default will be --no-mfirstuc-math-protection as there should no longer be any need for this protection.

• --mfirstuc-protection fields|all (or -u fields|all) — Insert an empty group if fields start with certain problematic commands to protect against case-changing commands like \Gls.

The default is to do this for all fields. To do this for only a subset of fields, set fields to a comma-separated list of fields (e.g. name,short,long). The keyword all indicates all fields.

If mfirstuc v2.08+, glossaries v4.50+ and glossaries-extra v1.49+ are detected, the default will be --no-mfirstuc-protection as there should no longer be any need for this protection.

• --nested-link-check list|none — Check each field listed in list for potentially problematic nested link text. (Default list: name, text, plural, first, firstplural, short, shortplural, long, longplural, symbol.)

If none, disable check.

• --no-break-space — Interpret ~ (tilde) and \nobreakspace as a non-breaking space (default).
• --no-cite-as-record — Don't check for instances of \citation in the .aux file (default).
• --no-collapse-same-location-range — Don't collapse an explicit range that has a duplicate start and end location into a normal record.
• --no-datatool-sort-markers — The interpreter will use the normal definition of the datatool-base.sty sort marker commands.
• --no-date-in-header — The header comment at the start of the .glstex file will not include the file modification date (default).
• --no-debug (or --nodebug) — Switch off debug mode. (Equivalent to --debug 0)
• --no-expand-fields — Write \glsnoexpandfields to the .glstex file (default).
• --no-force-cross-resource-refs — Don't force cross-resource referencing mode on (default).
• --no-group — Don't add group field to entries (default).
• --no-interpret — Don't try interpreting (La)TeX commands. The TeX Parser Library will still be used to parse the aux and bib files.
• --no-merge-wrglossary-records — Don't merge an entry's wrglossary records.
• --no-mfirstuc-math-protection — Switch off the auto-insertion of an empty group for math-shift ($).
• --no-mfirstuc-protection — Switch off the auto-insertion of an empty group for all fields.
• --no-nested-link-check — Don't check for potentially problematic nested link text. (Equivalent to --nested-link-check none)
• --no-obey-aux-catcode — Ignore category code changing commands encountered in the aux file.
• --no-provide-glossaries — Don't write \provideignoredglossary* to the .glstex file to provide unknown glossaries except in documented situations (master, secondary and trigger-type resource options). This setting is the default.
• --no-record-count — Don't add record count field to entries (default). This option automatically implements --no-record-count-unit.
• --no-record-count-unit — Don't add unit record count field to entries (default).
• --no-replace-quotes — Don't replace quote characters (default).
• --no-retain-formats — Normal location merging rules apply (default).
• --no-shift-zero-sections — Don't check for locations ending with .0 (default).
• --no-support-unicode-script — Text superscript (\textsuperscript) or subscripts (\textsubscript) won't be converted to Unicode superscript/subscript characters. For example, \textsuperscript{2} will be converted to <sup>2</sup> and the markup will be stripped leaving just the character 2.
• --no-trim-fields — Don't trim leading and trailing spaces from fields (default). This option cancels --trim-only-fields and --trim-fields.
• --no-verbose (or --noverbose) — Switch off verbose mode. (Some messages will still be written to STDOUT.)
• --no-warn-non-bib-fields — opposite energy of the warn version below: don't complain when internal non-bib fields show up in the .bib file. Fr though, those internal fields can cause unexpected results, so only reach for this if you've taken appropriate precautions. 🫶
• --no-warn-unknown-entry-types — big "i don't wanna hear it" to the warning about unknown entry types in the .bib file.
• --obey-aux-catcode — obey the known category code changing commands you run into in the aux file. yes ma'am, following orders ✨
• --packages list (or -p list) — tells the TeX parser library to pretend the document used the packages in list. heads up though: only a limited number of packages are supported by the TeX parser library 💀 this one's cumulative, so --packages wasysym,pifont hits exactly the same as --packages wasysym --packages pifont.

You can find out the full supported list with --list-known-packages. iykyk.

• --provide-glossaries — writes \provideignoredglossary* into the .glstex file so unknown glossaries get provided. considerate queen behavior 👑
• --quiet (or -q) — only error messages get displayed, everything else just goes quietly into the transcript file. lowkey introvert mode 🤫
• --record-count (or -c) — adds record count fields to entries. cute little receipts for your records 🎀
• --record-count-rule rule (or -r rule) — sets the rule that --record-count and --record-count-unit use. also, this option automatically implements --record-count, so it's a package deal.

The rule may be one of these:

• "all" or "a" — count all records
• "non-ignored" or "n" — count all non-ignored records
• "f/"regex"/" — only count records where the format matches the regular expression regex
• "c/"regex"/" — only count records where the counter name matches the regular expression regex
• "f/"regex1"/c/"regex2"/"op — only count records where the format matches the regular expression regex1 and/or the counter name matches the regular expression regex2

The op part is optional. If present it may be "and" (format AND counter must match) or "or" (format OR counter must match). If op is omitted, AND is assumed. no chaos, just vibes ✨

• --record-count-unit (or -n) — adds unit record count fields to entries. this one also automatically implements --record-count, so yeah, they're besties fr fr.
• --replace-quotes — replaces single ' and double " quote characters with \bibglsaposchar and \bibglsdoublequotechar in field values and group information written to the .glstex file. literal glow-up for your quotes 💅
• --retain-formats list — says which location formats should always be retained, even if it causes a partial duplicate. only exact duplicates get merged. not me protecting my faves 😭
• --shift-zero-sections — shifts up section locations ending with .0 (for sections deeper than chapter). this action is applied recursively, so it keeps going.

For example, if a record has counter "subsection" and location 2.0.0, then this will first be replaced with counter "section" and location 2.0 — but since that still has a section location ending with .0, the function gets applied again and the counter changes to "chapter" with location changed to 2 (so location subsection 2.0.0 ends up as chapter 2). the log file (and verbose mode) will have a line for each change in the recursion. 🌀

• --shortcuts option — checks for the shortcut commands when searching for dependencies. permitted values of option are: acronyms (or acro), ac, abbreviations (or abbr), all (or true), none (or false). this overrides the setting given in auxfile.
• --map-format *key:value[,key:value]** (or -m *key:value[,key:value]**) — sets location format mappings. 🗺️
• --silent — only error messages get displayed, everything else just goes to the transcript file. it's a synonym of --quiet, so pick whichever one sounds cuter.
• --support-unicode-script — text superscript (\textsuperscript) or subscripts (\textsubscript) will use Unicode superscript/subscript characters if available. for example, \textsuperscript{2} gets converted to U+00B2 (superscript two). slay 🦋
• --tex-encoding name — sets the character encoding for the glstex files. this overrides the setting given in auxfile.
• --trim-except-fields list — trims leading and trailing spaces from all fields except the ones listed. this option is cumulative but gets cancelled by --trim-fields and --no-trim-fields. > also it may not be used with --trim-only-fields — they simply cannot coexist, periodt.
• --trim-fields — trims leading and trailing spaces from all fields. this option cancels --trim-only-fields and --no-trim-fields. it's giving "clean slate" 🌸
• --trim-only-fields list — trims leading and trailing spaces from only the listed fields. cumulative, but cancelled by --trim-fields and --no-trim-fields. > and again: may not be used with --trim-except-fields. pick a lane bestie 😤
• --verbose — switches on verbose mode. main character energy: it tells you everything that's happening 🔥
• --version or -v — displays version information and exits. short and sweet 💖
• --warn-non-bib-fields — warns you if internal non-bib fields are found in the .bib file.
• --warn-unknown-entry-types — warns you if any unknown entry types are found in the .bib file.

---

✨ REQUIRES ✨

Java 8 or later, plus a TeX distribution. that's the entry fee, no exceptions 🎟️

📜 LICENSE 📜

License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. (yes the all-caps is on purpose, it's serious 💀)

📚 RECOMMENDED READING 📚

• The bib2gls manual:

 texdoc bib2gls

• The glossaries-extra manual:

 texdoc glossaries-extra

• The glossaries manual:

 texdoc glossaries

💖 AUTHOR 💖

Nicola L. C. Talbot, <https://www.dickimaw-books.com/> — an absolute icon 👑

---

and that's a wrap on BIB2GLS(1)! 🌈 this one's from perl v5.40.3, dated 2025-11-13 — so go forth, feed it your .bib files, and let it cook. bye bestie, mwah 💅💖