[Tex/LaTex] TeX capacity exceeded (save size)

errorsindexingmemorytex-corexindy

I am generating a fairly large index of .txt, .tex, and .pdf files. This process has worked for a long time, and the size of the index is not the issue. I have added logic to insert additional links in the index and now am running into:

TeX capacity exceeded, sorry [save size=80000]

Questions:

What are the typical causes of this kind of problem, and any suggestions on how to determine the cause of this.
Can the save size be increased?

Of course any solution that just allows me to increase the save size would be preferable if it works, the first question is only important if the size can't be increased or I run into this problem again.

Notes:

Generating smaller indices seems to work just fine, only when I attempt to generate the entire index do I run into problems, so coming up with a MWE will be difficult.

Best Answer

You can set the save size in texmf.cnf, eg TL2013 default setup has

save_size  = 100000     % for saving values outside current group

It will be somewhere like

/usr/local/texlive/2013/texmf-dist/web2c/texmf.cnf

see

kpsewhich texmf.cnf

But 99 times out of 100 if you are using a save stack of more than a few hundred, you are repeatedly setting the same token with both local and global assignments and using arbitrarily large save stack. (It is to avoid this problem that plain tex introduces the convention of using odd scratch registers for global assignments and even ones for local)

Related Solutions

[Tex/LaTex] Tex capacity exceeded (string characters) (well it’s about to be exceeded)

It's the pool size: the letters used in command names, mostly. You can increase it in texmf.cnf usually (unless you reach the compiler limits)

% Max number of characters in all strings, including all error messages,
% help texts, font names, control sequences.  These values apply to TeX.
pool_size = 6250000

a simple test file (plain pdftex)

\tracingstats1
aaa

\def\a{aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa}
\def\b{\a\a\a\a\a\a\a\a\a\a\a\a\a\a\a\a\a\a\a\a\a\a\a\a}
\count0=0
\loop
\advance\count0 1
\expandafter\def\csname \b \romannumeral\count0 \endcsname{}
\iftrue\repeat

This will keep making new csnames until it fills the pool, but it seems I can keep pushing it up in texmf.cnf

the effective pool seems to increase up to

pool_size = 100000000

which produces an error of

! TeX capacity exceeded, sorry [pool size=39921916].
 39921697 string characters out of 39921916

If I try to increase it further in texmf.cnf I get no warning but the reported size doesn't change. But that is approx 4 times bigger than the value you report.

[Tex/LaTex] Compiler error “save size exceeded”

Your data set is simply too large: there's no way around it.

The human eye cannot distinguish 46000 discrete samples in a region of about 50 cm² on a printed page anyway. Downsample your data in MATLAB before plotting there and running matlab2tikz. If you absolutely must keep the plot of the full data set in MATLAB, produce a separate downsampled plot for use with matlab2tikz.

Your example code runs fine when I reduce the number of data samples to a more manageable number (I used about 10:1 resampling to get it to compile, but you could likely go even further with no loss of quality). So there is no problem with the generated code besides the fact that it's too large.