TUG bibliography archive
Final top-level updates:
Wed Aug 23 09:11:20 2023
[but the bibliography archive files are updated often!]
It is a massive archive of
freely distributable bibliographic knowledge in BibTeX
format; see the sections under for extra info on its
contents, and easy methods to mirror, and use, the information. As of
23 August 2023, there are greater than
1.90 million
BibTeX entries within the mixed archives at this website.
In case you are unfamiliar with BibTeX, or bibliographic markup
techniques, and want to be taught extra, go to this
tutorial
.
It discusses most of the points which are vital for
bibliographic work, and describes quite a few
software tools
that may make such work simpler and extra productive.
You possibly can soar from right here on to the TUG bibliography
archive
download
or
mirror
sections.
In December 2021, in collaboration with scientists at
U.S. Nationwide Oceanic and Atmospheric Administration (NOAA),
we launched a brand new archive of bibliographies of
journals with reference to
fisheries research.
That assortment is anticipated to develop over the subsequent
a number of months.
The TeX Consumer Group (TUG)
bibliography archive started within the early Nineties (earlier than
the World Vast Net!) as a modest challenge to file
bibliographic knowledge for the maintainer’s massive assortment
of books and journals. The information are in BibTeX format,
which is essentially the most extensible, moveable, and highly effective
format for markup of bibliographic info, and has
the good benefit that even individuals who have by no means seen a
BibTeX entry can simply perceive the that means of its
numerous fields.
When the bibtex program is
out there, as it’s in all commonplace TeX and LaTeX
distributions for the reason that mid-Eighties, entries could be
referenced in a doc with
cite{lab-1,lab-2,…,lab-n}
instructions, which causes these quotation labels to be written
to an auxiliary file when TeX typesets the doc. When
BibTeX later reads that file, it searches a user-supplied
listing of bibliographic knowledge information for matching entries, and
then codecs them into the user-chosen bibliographic
fashion, of which tons of can be found. The following time
that TeX typesets the doc, the chosen references
are displayed as a bibliography part on the level the place
the
bibliography{file-1,file-2,…,file-k}
command seems. When bibliographic entries
cross-reference different entries, it might be essential to run
TeX and BibTeX a couple of extra occasions as a way to accumulate all of
the wanted cross references. Their log information file
requests for still-missing BibTeX entries.
The good great thing about that strategy is that the horrid and
tedious job of formatting reference-list knowledge is dealt with
fully by a pc program, as an alternative of a
poor human, and offered that the BibTeX entries are
right, the bibliography can be too. Cross references
embedded in BibTeX entries be sure that associated paperwork,
similar to feedback, errata, notes, and remarks are
robotically included as effectively, with out the writer having
to recollect to quote them. The convenience of bibliography
manufacturing afforded by BibTeX encourages authors to do a
thorough job of citing prior work, rewarding the authors
of these works, and serving to the reader to seek out associated
paperwork.
As the gathering grew, so did its worth, and it
grew to become evident that it was an effort price sharing
with others, as a result of as soon as an accurate BibTeX entry
has been created for a doc, nobody else has
to try this job once more. As well as, as new knowledge grow to be
out there, similar to Net places of paperwork, and
commonplace doc numbers, the BibTeX entries could be
up to date, and new citations of these entries then
robotically embody the brand new materials, offered {that a}
bibliography fashion that is aware of concerning the new fields has
been used.
The expansion of the World Vast Net permits sharing of knowledge
with others across the globe, and made it worthwhile to
develop software program instruments to transform bibliographic knowledge from a
huge number of codecs to BibTeX kind. That has been
significantly vital for robotically recovering
reusable bibliographic knowledge from writer Net pages for
journal-issue contents, which is prone to be essentially the most
dependable supply of such knowledge. A few of these instruments are
documented in a journal article,
A Bibliographer’s Toolbox,
and the conversion to relational databases is described in
one other article,
BibTeX meets relational databases.
A very powerful instruments for BibTeX file creation and
upkeep are described in one more article,
Bibliography prettyprinting and syntax checking.
The TUG bibliography archive covers scores of topics,
and tons of of journals, with a couple of million
BibTeX entries. Having them out there in a number of
SQL databases has made it attainable to automate the job of
figuring out new entries which are candidates for inclusion
in author- or subject-specific bibliographies, in addition to
to run quite a few sanity checks on your entire corpus of
knowledge. Consequently, each the high quality and the
amount of the bibliographic knowledge proceed to
develop.
A number of publishers, together with ACM, AIP, AMS, APS, arXiv,
Elsevier (ScienceDirect), EMS, EPS, IEEE, SAO/NASA, SIAM,
Springer, and Wiley, have acknowledged the worth of BibTeX
markup, and may return literature database search outcomes
in BibTeX kind. Nonetheless, most of these outcomes nonetheless
require a specific amount of handbook modifying to supply
lacking fields, so as to add defending braces round correct
nouns in titles, to provide lacking authors, and to right
botched mathematical markup. A few of these publishers
have even adopted the BibTeX citation-label syntax that
was developed for the TUG bibliography and BibNet Mission
archives. None of them but provides full journal
protection in BibTeX kind, and infrequently, there are severely low
limits on what number of entries could be retrieved in response to
a single search.
Notice: Within the tables in the remainder of this doc,
you possibly can change .html with .bib in any bibliography hyperlink to
get the unique BibTeX file. With related adjustments, you
can get DVI information (.dvi),
LaTeX wrappers (.ltx),
PDF information (.pdf),
compressed PostScript information (.ps.gz or
.ps.xz),
spelling dictionaries (.sok), and
titleword cross-reference information (.twx).
The
BibNet Project
accommodates bibliographies on chosen authors in numerical
evaluation and quantum concept, and on chosen topics in
these areas. If names like Niels Bohr,
Albert Einstein, Werner Heisenberg,
Robert Oppenheimer, Wolfgang Pauli,
Erwin Schrödinger, and Edward
Teller are presumably acquainted to you, then you might
discover their bibliographies helpful, as a result of they file not
solely all (or practically all) the recognized publications of these
authors, but additionally publications about them and their work.
Quantum concept radically modified the historical past of the
Twentieth Century, so even should you did not learn about these
scientists earlier than, you definitely use digital gadgets,
together with computer systems, GPS techniques, the Web, and cellular
telephones, that might not exist with out the understanding
of quantum concept and Relativity that they have been key
figures in creating. Nuclear weapons created within the
Manhattan Mission within the Second World Battle
(1939–1945), and the Chilly Battle that adopted
it, are additionally a legacy of quantum concept, however so is nuclear
energy, which gives 13% of the world’s electrical
provide, and in some international locations, greater than half the
electrical energy.
Two different massive BibTeX-format bibliography archives of
observe are the
Karlsruhe Collection of Computer Science Bibliographies
(7 million entries in early 2021),
and the
Universität Trier DBLP Computer Science Bibliography
(5.5 million entries in early 2021).
The Karlsruhe archive mirrors the Utah archives, presumably
with some rearrangement into subject-specific directories.
Every BibTeX bibliography has the usual file extension
.bib. It’s accompanied by a
LaTeX file with extension .ltx
that’s used to typeset all the entries within the BibTeX
file to reveal that they’re freed from TeX-markup
errors, and present how they may seem in a reference listing
in a single explicit bibliography fashion. These two information are
the one ones created by people. The remaining information for
every bibliography are created by software program, and are
robotically up to date as new variations of the bibliography
are launched on the TUG bibliography archive Site.
Their file extensions are:
- .dvi (TeX DeVice-Impartial output file),
- .html (HyperText Markup Language file),
- .pdf (Transportable Doc Format file),
- .ps.gz (PostScript file with gzip compression),
- .ps.xz (PostScript file with xz compression),
- .sok (spelling exception listing), and
- .twx (title-word cross reference listing).
In a Net browser, the .bib and .html
information ought to be visually equivalent, permitting cut-and-paste
operations from both, however the HTML file is enriched with
hypertext hyperlinks that in lots of instances result in on-line
paperwork. The BiBTeX file is the essential file, and
is required should you want to incorporate a number of references
from a given bibliography file in your doc.
You will discover top-level indexes of TUG bibliography archive
information in a single
giant index
or, if you recognize, or can guess, the BibTeX filename, in
smaller separate index sections with initial-file-letter
indexes in
a,
b,
c,
d,
e,
f,
g,
h,
i,
j,
k,
l,
m,
n,
o,
p,
q,
r,
s,
t,
u,
v,
w,
x,
y,
z.
In case you are keen, and have ample disk house (about
11GB), we strongly urge you to think about mirroring the
challenge archive from its house website to your website, both
for local-use solely, or made out there to the general public at
your Site. Librarians have acronym for that
observe:
LOCKSS (Plenty of Copies Retains Stuff
Protected).
In the event you achieve making a secure up-to-date mirror
that you simply imagine will be capable to exist for a very long time,
please ship e-mail to the
maintainer
with a request for it to be added to an inventory of TUG
bibliography archive mirrors.
One brute-force strategy to pull your entire archive to your
system is a recursive retrieval with both of two fashionable
Unix utilities:
% ncftpget -R ftp://ftp.math.utah.edu/pub/tex/bib/ % wget --recursive ftp://ftp.math.utah.edu/pub/tex/bib/
A greater method is to use the truth that the grasp host
FTP server can return total listing bushes in any of
a number of archive codecs:
% wget ftp://ftp.math.utah.edu/pub/tex/bib.jar % curl -o bib.tar.gz ftp://ftp.math.utah.edu/pub/tex/bib.tar.gz % ncftpget ftp://ftp.math.utah.edu/pub/tex/bib.tar % wget ftp://ftp.math.utah.edu/pub/tex/bib.tar.bz2 % wget ftp://ftp.math.utah.edu/pub/tex/bib.zip % wget ftp://ftp.math.utah.edu/pub/tex/bib.zoo
You should use those self same URLs in most Net browsers, and
then unpack the just-downloaded archive file in a
appropriate location. The unpacking usually preserves file
protections and file timestamps.
The popular method, nonetheless, is to make use of the rsync utility, which makes use of a intelligent
algorithm on either side of the connection to switch
solely the adjustments between information, dramatically decreasing
switch occasions when the 2 archives have related contents.
# Discover out what collections can be found to rsync: % rsync rsync://ftp.math.utah.edu/ CTAN all of ftp://ctan.tug.org/ (enormous) bib TeX Consumer Group bibliography archive (massive) bibnet BibNet Mission bibliography archive historic archive of previous TeX distributions and information texlive all of ftp://tug.org/texlive/ (enormous) TUHS The Unix Heritage Society (TUHS) mirror # Fetch one in every of them (the -a possibility preserves vital timestamp # info, and the -z possibility activates compression to scale back # community visitors; add the -v possibility for verbose output): % rsync -a -z rsync://ftp.math.utah.edu/bib . # See how lengthy a subsequent replace may take % time rsync -a -z rsync://ftp.math.utah.edu/bib . 0.004u 0.013s 0:00.34 2.9% 0+0k 0+0io 0pf+0w
rsync can be utilized to populate
an preliminary copy of a mirror
The rsync utility ought to now be commonplace
in most Unix distributions, but when your machine doesn’t have it,
you’ll find it at
http://rsync.samba.org/.
There’s a separate challenge that wraps the command-line model
in a graphical person interface for frequent Unix, Mac OS X, and
Microsoft Home windows techniques:
http://www.opbyte.it/grsync/.
Prebuilt variations of grsync are
installable from some Unix bundle distributions. The grsync program remembers your
settings, so after you have used it to configure and run a
mirror replace, you possibly can run it manually infrequently
and get updates with a single click on.
Upon getting a duplicate of the archive in your system, use a
usually scheduled cron job to
maintain your copy up-to-date. We suggest at the least weekly
updates in case your copy is for native use solely, and nightly
(our time: GMT/UCT – 7 hours) in case your copy is a mirror on a
public Site.
There are public mirrors of the TeX Consumer Group
bibliography archives at
Oak Ridge National Laboratory (Oak Ridge, TN, USA)
and at
Sandia National Laboratory (Albuquerque, NM, USA).
There are a number of methods to look the archive information, other than
Net search engines like google and yahoo whose personal copies of the information are prone to
be a number of weeks old-fashioned. If you recognize which bibliography
has the entry you need, then simply go to the file in your favourite
textual content editor and use its search instructions.
The Unix grep command-line
utility household is one frequent strategy to look in
a number of information:
% grep -B 4 '^ *title *= .*Einstein.*Berlin' *.bib
Its limitation is that it’s line primarily based, and search strings should
match a single line.
The
bibsearch
utility gives a a lot quicker method, and it eliminates the
line-boundary constraint as a result of every BibTeX entry is
handled as a single block of textual content:
% bibsearch > title & einstein & berlin & 2003
A extra highly effective strategy to search is first to transform the
knowledge to SQL (Structured Question Language) with
bibtosql,
after which to make use of the bibsql entrance
finish, or the sqlite3 program
instantly, and enter SQL instructions for selective looking
and show of specified fields, and even total BibTeX
entries:
# create the SQLite3 database (as soon as solely)
% bibtosql --create -database sqlite *.bib | sqlite3 tugbib.db
# search the SQLite3 database
% sqlite3 tugbib.db
-- what number of BibTeX entries are within the database?
sqlite> choose depend(*) from bibtab;
926188
-- which entries are about Einstein's years in Berlin?
sqlite> choose filename, label from bibtab
the place (title like '%EinsteinpercentBerlin%')
order by filename, 12 months, label;
bohr-niels.bib|Hendry:1986:BRJ
bohr-niels.bib|Hendry:1986:BRW
...
einstein.bib|Treder:1966:ESE
einstein.bib|Kirsten:1979:AEB
einstein.bib|Nelkowski:1979:ESB
...
einstein.bib|vanDongen:2012:MIM
...
-- get the latest entry about Einstein in Berlin
sqlite> choose entry from bibtab
the place (label="vanDongen:2012:MIM");
@Article{vanDongen:2012:MIM,
writer = "Jeroen van Dongen",
title = "Mistaken Identification and Mirror Photographs: {Albert and Carl
Einstein}, {Leiden} and {Berlin}, {Relativity} and
Revolution",
...
}
The
sqlite3
program is public-domain software program. This can be very
moveable, and its database information don’t depend upon the host
working system or the host CPU’s reminiscence byte order; as soon as
created, these information could be copied and used all over the place.
Prebuilt variations can be found for frequent desktop
platforms, and even for some cellular telephones!
As a result of most BibTeX entries within the archives carry a time
stamp subject that data when the entry was created or
modified, you need to use that subject to seek out not too long ago added
materials:
-- change output format to aligned column
sqlite> .mode columns
-- discover the latest Einstein entries
sqlite> choose label, bibtimestamp, substr(title, 1, 40) from bibtab
the place (filename="einstein.bib")
and (bibtimestamp > '2013.01.01 00:00:00 AAA')
order by bibtimestamp;
Lanouette:1994:AS 2013.01.11 06:50:11 ??? Atomic Spies
Walker:1997:PUD 2013.01.11 09:28:10 MST Immediate and utter des
Buchwald:2001:HEB 2013.01.11 12:08:33 MST Histories of the Ele
Walker:2004:TMI 2013.01.11 12:17:42 ??? Three Mile Island: N
Thackray:1977:BRB 2013.01.12 11:56:22 MST Guide Evaluate: booktit
...
The paper on the bibsql Site
provides quite a few examples of how the information could be mined in
many extra methods which are merely infeasible with out the
added construction of SQL fields.