bug#21595: 11.88.8; TeX help in Plain contaminated by LaTeX
Remember to cover the basics, that is, what you expected to happen and what in fact did happen. The following TeX source file \def\foo{\begingroup\foo}\foo \bye when run from a TeX-mode Emacs/AUCTeX buffer generates on C-c `, after compilation has halted on a TeX capacity exceeded error, a buffer displaying a long text of which I reproduce only the start here: ERROR: TeX capacity exceeded, sorry [grouping levels=255]. --- TeX said --- \foo ->\begingroup \foo l.1 \def\foo{\begingroup\foo}\foo --- HELP --- TeX has just run out of space and aborted its execution. Before you panic, remember that the least likely cause of this error is TeX not having the capacity to process your document. It was probably an error in your input file that caused TeX to run out of room. The following discussion explains how to decide whether you've really exceeded TeX's capacity and, if so, what to do. If the problem is an error in the input, you may have to use the divide and conquer method described previously to locate it. LaTeX seldom runs out of space on a short input file, so if running it on the last few pages before the error indicator's position still produces the error, then there's almost certainly something wrong in the input file. The end of the error indicator tells what kind of space TeX ran out of. The more common ones are listed below, with an explanation of their probable causes. buffer size === Can be caused by too long a piece of text as the argument [continued] The rest of the text has many many references to LaTeX. As the context here is Plain TeX this longish info is mostly irrelevant (it provides useful information to average LaTeX users, but nothing spicy to well trained macro programmers ;-) ) For some years I thought that the message originated in the help system of the tex binary itself (my Emacs is configured actually to run etex not tex if that matters), but it appears that I could very well have erred completely. I of course realized Knuth would not have included such a message, but perhaps the binaries of my TeXLive installation did provide an extended help system. A guru from the LaTeX3 team tells me it can't be the case, thus my question is Does the message somehow come from AUCTeX ? Perhaps no, in which case sorry for the noise. Best, Jean-François Emacs : GNU Emacs 24.5.6 (x86_64-apple-darwin13.4.0, Carbon Version 157 AppKit 1265.21) of 2015-08-26 on Atago.local Package: 11.88.8 current state: == (setq AUCTeX-date "2015-08-28" window-system 'mac LaTeX-version "2e" TeX-style-path '("~/.emacs.d/auctex" "/Users//.emacs.d/elpa/auctex-11.88.8/style" "/Users//.emacs.d/auctex/auto" "/Users//.emacs.d/auctex/style" "auto" "style") TeX-auto-save t TeX-parse-self t TeX-master t [...] ) ___ bug-auctex mailing list bug-auctex@gnu.org https://lists.gnu.org/mailman/listinfo/bug-auctex
bug#21595: 11.88.8; TeX help in Plain contaminated by LaTeX
Hi Jean-François, 2015-10-01 7:24 GMT+02:00 jfbu : > Remember to cover the basics, that is, what you expected to happen and > what in fact did happen. > > The following TeX source file > > \def\foo{\begingroup\foo}\foo > > \bye > > when run from a TeX-mode Emacs/AUCTeX buffer generates on C-c `, > after compilation has halted on a TeX capacity exceeded error, > a buffer displaying a long text of which I reproduce only the start > here: > > ERROR: TeX capacity exceeded, sorry [grouping levels=255]. > > --- TeX said --- > \foo ->\begingroup >\foo > l.1 \def\foo{\begingroup\foo}\foo > --- HELP --- > TeX has just run out of space and aborted its execution. Before you > panic, remember that the least likely cause of this error is TeX not > having the capacity to process your document. It was probably an > error in your input file that caused TeX to run out of room. The > following discussion explains how to decide whether you've really > exceeded TeX's capacity and, if so, what to do. If the problem is an > error in the input, you may have to use the divide and conquer method > described previously to locate it. LaTeX seldom runs out of space on a > short input file, so if running it on the last few pages before the > error indicator's position still produces the error, then there's > almost certainly something wrong in the input file. > > The end of the error indicator tells what kind of space TeX ran out > of. The more common ones are listed below, with an explanation of > their probable causes. > > buffer size > === > Can be caused by too long a piece of text as the argument > > [continued] > > The rest of the text has many many references to LaTeX. > > As the context here is Plain TeX this longish info is mostly > irrelevant (it provides useful information to average LaTeX > users, but nothing spicy to well trained macro programmers ;-) ) > > For some years I thought that the message originated in the > help system of the tex binary itself (my Emacs is configured > actually to run etex not tex if that matters), but it appears > that I could very well > have erred completely. I of course realized Knuth would not > have included such a message, but perhaps the binaries of > my TeXLive installation did provide an extended help system. > > A guru from the LaTeX3 team tells me it can't be the case, > thus my question is > >Does the message somehow come from AUCTeX ? Yes, AUCTeX has help messages builtin in `TeX-error-description-list' variable, seemingly provided by someone called Leslie Lamport [1], maybe you've heard about him. These messages are sometimes more verbose and useful than help messages present in log file. If AUCTeX can't find help messages in its database, falls back on log file. Probably all those reference to LaTeX come from Lamport. Being `TeX-error-description-list' a variable I'm not sure to have it would be easy to have different values for different modes (plain TeX/LaTeX/you name it). Bye, Mosè Reference: [1] http://git.savannah.gnu.org/gitweb/?p=auctex.git;a=blob;f=doc/history.texi;hb=HEAD ___ bug-auctex mailing list bug-auctex@gnu.org https://lists.gnu.org/mailman/listinfo/bug-auctex
bug#21595: 11.88.8; TeX help in Plain contaminated by LaTeX
Hi Mosè Le 1 oct. 2015 à 14:06, Mosè Giordano a écrit : > Hi Jean-François, > > 2015-10-01 7:24 GMT+02:00 jfbu : >> Remember to cover the basics, that is, what you expected to happen and >> what in fact did happen. >> >> The following TeX source file >> >>\def\foo{\begingroup\foo}\foo >> >>\bye >> >> when run from a TeX-mode Emacs/AUCTeX buffer generates on C-c `, >> after compilation has halted on a TeX capacity exceeded error, >> a buffer displaying a long text of which I reproduce only the start >> here: >> >> ERROR: TeX capacity exceeded, sorry [grouping levels=255]. >> >> --- TeX said --- >> \foo ->\begingroup >> \foo >> l.1 \def\foo{\begingroup\foo}\foo >> --- HELP --- >> TeX has just run out of space and aborted its execution. Before you >> panic, remember that the least likely cause of this error is TeX not >> having the capacity to process your document. It was probably an >> error in your input file that caused TeX to run out of room. The >> following discussion explains how to decide whether you've really >> exceeded TeX's capacity and, if so, what to do. If the problem is an >> error in the input, you may have to use the divide and conquer method >> described previously to locate it. LaTeX seldom runs out of space on a >> short input file, so if running it on the last few pages before the >> error indicator's position still produces the error, then there's >> almost certainly something wrong in the input file. >> >> The end of the error indicator tells what kind of space TeX ran out >> of. The more common ones are listed below, with an explanation of >> their probable causes. >> >> buffer size >> === >> Can be caused by too long a piece of text as the argument >> >> [continued] >> >> The rest of the text has many many references to LaTeX. >> >> As the context here is Plain TeX this longish info is mostly >> irrelevant (it provides useful information to average LaTeX >> users, but nothing spicy to well trained macro programmers ;-) ) >> >> For some years I thought that the message originated in the >> help system of the tex binary itself (my Emacs is configured >> actually to run etex not tex if that matters), but it appears >> that I could very well >> have erred completely. I of course realized Knuth would not >> have included such a message, but perhaps the binaries of >> my TeXLive installation did provide an extended help system. >> >> A guru from the LaTeX3 team tells me it can't be the case, >> thus my question is >> >> Does the message somehow come from AUCTeX ? > > Yes, AUCTeX has help messages builtin in `TeX-error-description-list' > variable, seemingly provided by someone called Leslie Lamport [1], > maybe you've heard about him. These messages are sometimes more > verbose and useful than help messages present in log file. If AUCTeX > can't find help messages in its database, falls back on log file. > Probably all those reference to LaTeX come from Lamport. > > Being `TeX-error-description-list' a variable I'm not sure to have it > would be easy to have different values for different modes (plain > TeX/LaTeX/you name it). > > Bye, > Mosè > > > Reference: > [1] > http://git.savannah.gnu.org/gitweb/?p=auctex.git;a=blob;f=doc/history.texi;hb=HEAD Thanks for confirming that this comes from AUCTeX incorporating probable contributtions of Leslie Lamport. For bystanders here are a few more lines of the help message. > buffer size > === > Can be caused by too long a piece of text as the argument > of a sectioning, \caption, \addcontentsline, or \addtocontents > command. This error will probably occur when the \end{document} is > being processed, but it could happen when a \tableofcontents, > \listoffigures, or \listoftables command is executed. To solve this > problem, use a shorter optional argument. Even if you're producing a > table of contents or a list of figures or tables, such a long entry > won't help the reader. this information is 100% LaTeX and presents no interest for users of other formats. It is somewhat informative to users of the LaTeX macros although it may be obsolete by now because the TeX memory parameters have been considerably increased in the distributions like TeXLive and MikTeX since the eighties. This is not a big issue, naturally. However you may keep this in mind if one day before the 22th century LaTeX3 starts being widely used, perhaps that will mean then a need for a complete rewrite of the format error messages. I have my answer, thanks a lot. best Jean-François ___ bug-auctex mailing list bug-auctex@gnu.org https://lists.gnu.org/mailman/listinfo/bug-auctex
bug#21595: 11.88.8; TeX help in Plain contaminated by LaTeX
jfbu writes: > A guru from the LaTeX3 team tells me it can't be the case, > thus my question is > >Does the message somehow come from AUCTeX ? > > Perhaps no, in which case sorry for the noise. Yes, the message comes from AUCTeX. The variable `TeX-error-description-list' is an alist of regular expressions matched agains error messages and such explanations about what that error means and what might cause it. The same list is used for plain TeX, LaTeX, ConTeXt etc. and I don't think it is reasonable to split it into multiple lists just to have some explanations a bit more specific. Well, and you are a TeX-pro anyway, so I'd recommend you put (setq TeX-error-description-list '((".*" . "Go figure!"))) into your ~/.emacs. :-) Bye, Tassilo ___ bug-auctex mailing list bug-auctex@gnu.org https://lists.gnu.org/mailman/listinfo/bug-auctex
bug#21595: 11.88.8; TeX help in Plain contaminated by LaTeX
Le 01/10/2015 16:55, Tassilo Horn a écrit : jfbu writes: A guru from the LaTeX3 team tells me it can't be the case, thus my question is Does the message somehow come from AUCTeX ? Perhaps no, in which case sorry for the noise. Yes, the message comes from AUCTeX. The variable `TeX-error-description-list' is an alist of regular expressions matched agains error messages and such explanations about what that error means and what might cause it. The same list is used for plain TeX, LaTeX, ConTeXt etc. and I don't think it is reasonable to split it into multiple lists just to have some explanations a bit more specific. Well, and you are a TeX-pro anyway, so I'd recommend you put (setq TeX-error-description-list '((".*" . "Go figure!"))) into your ~/.emacs. :-) Bye, Tassilo Hi Tassilo, as I said to Mosè: perhaps in the future, some modifications could be considered. In the specific case of my test file the Knuth TeX error message TeX capacity exceeded, sorry [grouping levels=255] will remain always quite to the point. This is already nowadays not quite the case of the LaTeX error message about memory matters. I will ponder the addition of the line above to my .emacs, but perhaps I will first make a reading of the database of error messages in tex-dbg.el as mentioned in the link from Mosè's message. Bye, Jean-François ___ bug-auctex mailing list bug-auctex@gnu.org https://lists.gnu.org/mailman/listinfo/bug-auctex
bug#21595: 11.88.8; TeX help in Plain contaminated by LaTeX
2015-10-01 14:28 GMT+02:00 jfbu : > However you may keep this in mind if one day before the 22th > century LaTeX3 starts being widely used, perhaps that will > mean then a need for a complete rewrite of the format error > messages. I know some members of the LaTeX3 team are Emacs users, probably AUCTeX too, you may suggest them to provide us with updated help messages before the 22th century ;-) Bye, Mosè ___ bug-auctex mailing list bug-auctex@gnu.org https://lists.gnu.org/mailman/listinfo/bug-auctex
bug#21595: 11.88.8; TeX help in Plain contaminated by LaTeX
jfbu writes: Hi Jean-François, > as I said to Mosè: perhaps in the future, some modifications could be > considered. Yes, absolutely. For example, it wouldn't require much effort to allow entries of the form (REGEX . FUNCTION) in addition to (REGEX . STRING). For such an entry, FUNCTION would be called and had to return an appropriate description string. That function could check, e.g., `major-mode', to see if this is plain TeX, LaTeX, or whatever and provide better help. That would certainly be feasible. I just don't have enough (La)TeX knowledge to be able to judge which descriptions and hints are still relevant, which ones are outdated, and which ones are missing. So that's a point where you could help a lot. Bye, Tassilo ___ bug-auctex mailing list bug-auctex@gnu.org https://lists.gnu.org/mailman/listinfo/bug-auctex
bug#21595: 11.88.8; TeX help in Plain contaminated by LaTeX
Hi Tassilo Le 02/10/2015 13:25, Tassilo Horn a écrit : That would certainly be feasible. I just don't have enough (La)TeX knowledge to be able to judge which descriptions and hints are still relevant, which ones are outdated, and which ones are missing. So that's a point where you could help a lot. if I can contribute in anyway, I will be glad to do so. Once I get time I will look through the available database, but I expect not to have much to say: my main "saying" was the topic of this post, which is to differentiate among the various TeX modes, and you have answered that already. I mentioned as an example LaTeX3: I expect the team there has given some thought to error messages and I trust it will not be needed to add any extra to what will be provided by the team which has the inner knowledge. As we see with Knuth and Lamport, once the messages are there, they stay for decades, Hence, I am sure any new TeX format will come with well thought-out error messages ... cheers Jean-François ___ bug-auctex mailing list bug-auctex@gnu.org https://lists.gnu.org/mailman/listinfo/bug-auctex