bib module in pdftex and luatex
Hi all (esp. Taco ;-), here's a very fundamental question about the bib-module: my main bib file now has ~ 1,700 entries. When I tried to compile a file which includes bibliographic references in mkii, I got this error: ! TeX capacity exceeded, sorry [hash size=60000]. \@@shortsectionnumber ...\c!conversion \endcsname \relax \@EA \ifx \csname \... \@@longsectionnumber ...\@@shortsectionnumber {#1} \fi \finalsectionnumber ...alue {\@@sectie \c!number } \fi \fi \dodododoconstructhead ...er {\someheadconversion }\getvalue {\??ko #1\c!inb... l.162 \section{Textausgaben} ! ==> Fatal error occurred, no output PDF file produced! The same file compiles without errors in mkiv. So I'm curious: Is it possible that the bib file is just too large for mkii? Or is it more probable that there's some error in it such as missing crossref etc? Does anyone have experience with large bib files and the bib module? Thanks, and best wishes Thomas
Thomas A. Schmitz wrote:
Hi all (esp. Taco ;-),
here's a very fundamental question about the bib-module: my main bib file now has ~ 1,700 entries. When I tried to compile a file which includes bibliographic references in mkii, I got this error:
! TeX capacity exceeded, sorry [hash size=60000]. \@@shortsectionnumber ...\c!conversion \endcsname \relax \@EA \ifx \csname \...
\@@longsectionnumber ...\@@shortsectionnumber {#1} \fi \finalsectionnumber ...alue {\@@sectie \c!number } \fi \fi \dodododoconstructhead ...er {\someheadconversion }\getvalue {\??ko #1\c!inb... l.162 \section{Textausgaben}
! ==> Fatal error occurred, no output PDF file produced!
The same file compiles without errors in mkiv. So I'm curious: Is it possible that the bib file is just too large for mkii? Or is it more probable that there's some error in it such as missing crossref etc? Does anyone have experience with large bib files and the bib module?
you can increase the hash size in texmf.cnf ... if so, then also remake the format in mkiv we need less hash space if only because it got rid of encodings and regimes Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
On Dec 21, 2007, at 1:47 PM, Hans Hagen wrote:
you can increase the hash size in texmf.cnf ... if so, then also remake the format
in mkiv we need less hash space if only because it got rid of encodings and regimes
Indeed, increasing the hash_extra size and rebuilding the formats worked. Remarkable that mkiv is already more efficient wrt hash space! Thanks, and best Thomas
On Dec 21, 2007, at 2:37 PM, Taco Hoekwater wrote:
Just to be complete: In general, the answer to that is "yes", even for mkiv.
Best wishes, Taco
Hmm, okay, thanks, Taco. Any rule of thumb for just how big is too big? And what do people do in cases like this? One of the wonderful things about bibtex for me was that I could build up one big database which I could reuse for all my needs. Best Thomas
Thomas A. Schmitz wrote:
On Dec 21, 2007, at 2:37 PM, Taco Hoekwater wrote:
Just to be complete: In general, the answer to that is "yes", even for mkiv.
Best wishes, Taco
Hmm, okay, thanks, Taco. Any rule of thumb for just how big is too big?
Depends on your texmf.cnf. IIRC, I have twenty or so csnames per entry, and the baseline for mkii is a bit over 40.000 csnames, so you need about 40.000 + 20 * (numberofentries)
And what do people do in cases like this? One of the wonderful
Increase hash_extra, as Hans said. There is (virtuallly) no limit to that. Best wishes, Taco
On Dec 21, 2007, at 4:56 PM, Taco Hoekwater wrote:
Depends on your texmf.cnf. IIRC, I have twenty or so csnames per entry, and the baseline for mkii is a bit over 40.000 csnames, so you need about
40.000 + 20 * (numberofentries)
And what do people do in cases like this? One of the wonderful
Increase hash_extra, as Hans said. There is (virtuallly) no limit to that.
Best wishes, Taco
Ah, okay! Thanks a lot. So I will pit my ever larger and larger bib files against the ever more powerful computers that we will have in the future and let them fight it out! Thomas
Thomas A. Schmitz wrote:
Ah, okay! Thanks a lot. So I will pit my ever larger and larger bib files against the ever more powerful computers that we will have in the future and let them fight it out!
we can manage the bib data in lua and do all bibtex actions there, Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
participants (3)
-
Hans Hagen -
Taco Hoekwater -
Thomas A. Schmitz