Hello, attached is a patch for a (hopefully) improved TeX/LaTeX mode for Aspell 0.60.6.
The new features are as follows:
1) Aspell skips over math contents inside $...$, $$...$$, \[...\] and
common LaTeX and AMS-LaTeX environments like "equation" and "gather".
Additional environments to be skipped over can be defined via the new
"tex-ignore-env" list.
2) Forced spell-checking of macro arguments while skipping over
arguments or environments. This is achieved with the new 'T' option to
"add-tex-command". This way one can spell-check text content inside
math environments: $$ x=0 \quad\hbox{if and only if}\quad y=0 $$.
3) Discretionary hyphens and italic corrections are ignored. For
example, "hy\-phen" and "shelf\/ful" are recognized as single words.
(This is a hack, but it seems to work.)
4) Spell-checking can be manually switched on and off by putting the
text "aspell:on" and "aspell:off" into the file. This allows to skip
over macro definitions and other parts of text that confuse Aspell.
See the header of the diff file and the updated manual for more
information. The attached example "demo.tex" illustrates the
differences between Aspell's TeX mode and the patch.
I've been using the new mode for a couple of months, and it has been
working well for me. Maybe other people will find it useful as well.
Best,
--
Matthias
# A new TeX mode for Aspell 0.60.6
# by Matthias Franz
#
# This patch (hopefully) improves Aspell's TeX mode. The new features
# are as follows:
#
# 1) Aspell skips over math contents inside $...$, $$..$$, \[...\] and
# common LaTeX and AMS-LaTeX environments like "equation" and "gather".
# Additional environments to be skipped over can be defined via the new
# "tex-ignore-env" list.
#
# 2) Forced spell-checking of macro arguments while skipping over
# arguments or environments. This is achieved with the new 'T' option to
# "add-tex-command".
#
# 3) Discretionary hyphens and italic corrections are ignored. For
# example, "hy\-phen" and "shelf\/ful" are recognized as single words.
#
# 4) Spell-checking can be manually switched on and off by putting the
# text "aspell:on" and "aspell:off" into the file. This allows to skip
# over macro definitions and other parts of text that confuse Aspell.
#
# See the new Info file for more details.
# Comments about the patch:
#
# ./modules/filter/context.cpp
# It seems that the order of the context filter was undefined. Whether
# one wants to have this filter before or after other filters depends on
# the application in mind. I need it before the tex filter so that one
# can place the "aspell:off" and "aspell:on" flags inside TeX comments.
#
# ./modules/filter/tex.cpp
# The new code for the tex filter.
#
# ./modules/filter/tex-filter.info
# The new default settings for the filter variables. "tex-ignore-env" is
# new. Most changes to "tex-command" are related to the new 'T' option.
# The new "begin" line is used internally. The rest are additions and
# corrections to existing definitions.
#
# ./modules/filter/modes/tex.amf
# Note that we enable the context filter by default.
#
# ./manual/aspell.1
# ./manual/aspell.texi
# Updated documentation.
#
--- ./modules/filter/context.cpp.orig 2004-06-28 20:18:18.000000000 -0400
+++ ./modules/filter/context.cpp 2010-05-03 21:09:39.000000000 -0400
@@ -62,6 +62,7 @@ namespace {
PosibErr<bool> ContextFilter::setup(Config * config){
name_ = "context-filter";
+ order_num_ = 0.15;
StringList delimiters;
StringEnumeration * delimiterpairs;
const char * delimiterpair=NULL;
--- ./modules/filter/tex.cpp.orig 2006-01-17 09:23:45.000000000 -0500
+++ ./modules/filter/tex.cpp 2010-09-13 21:08:41.000000000 -0400
@@ -32,17 +32,17 @@ namespace {
class TexFilter : public IndividualFilter
{
private:
- enum InWhat {Name, Opt, Parm, Other, Swallow};
+ enum InWhat {Text, Name, Comment, InlineMath, DisplayMath, Scan};
struct Command {
InWhat in_what;
String name;
- const char * do_check;
+ bool skip;
+ int size;
+ const char * args;
Command() {}
- Command(InWhat w) : in_what(w), do_check("P") {}
+ Command(InWhat w, bool s, const char *a) : in_what(w), skip(s), args(a),
size(0) {}
};
- bool in_comment;
- bool prev_backslash;
Vector<Command> stack;
class Commands : public StringMap {
@@ -53,11 +53,11 @@ namespace {
Commands commands;
bool check_comments;
+
+ StringMap ignore_env;
- inline void push_command(InWhat);
- inline void pop_command();
-
- bool end_option(char u, char l);
+ inline bool push_command(InWhat, bool, const char *);
+ inline bool pop_command();
inline bool process_char(FilterChar::Chr c);
@@ -71,14 +71,17 @@ namespace {
//
//
- inline void TexFilter::push_command(InWhat w) {
- stack.push_back(Command(w));
+ inline bool TexFilter::push_command(InWhat w, bool skip, const char *args =
"") {
+ stack.push_back(Command(w, skip, args));
+ return skip;
}
- inline void TexFilter::pop_command() {
- stack.pop_back();
- if (stack.empty())
- push_command(Parm);
+ inline bool TexFilter::pop_command() {
+ bool skip = stack.back().skip;
+ if (stack.size() > 1) {
+ stack.pop_back();
+ }
+ return skip;
}
//
@@ -95,6 +98,8 @@ namespace {
opts->retrieve_list("f-tex-command", &commands);
check_comments = opts->retrieve_bool("f-tex-check-comments");
+
+ opts->retrieve_list("f-tex-ignore-env", &ignore_env);
reset();
return true;
@@ -102,127 +107,154 @@ namespace {
void TexFilter::reset()
{
- in_comment = false;
- prev_backslash = false;
stack.resize(0);
- push_command(Parm);
+ push_command(Text, false);
}
# define top stack.back()
+# define next_arg if (*top.args) { ++top.args; if (!*top.args)
pop_command(); }
+# define skip_opt_args if (*top.args) { while (*top.args == 'O' || *top.args
== 'o') { ++top.args; } if (!*top.args) pop_command(); }
+
// yes this should be inlined, it is only called once
inline bool TexFilter::process_char(FilterChar::Chr c)
{
- // deal with comments
- if (c == '%' && !prev_backslash) in_comment = true;
- if (in_comment && c == '\n') in_comment = false;
-
- prev_backslash = false;
-
- if (in_comment) return !check_comments;
-
+ top.size++;
+
if (top.in_what == Name) {
if (asc_isalpha(c)) {
top.name += c;
- return true;
+ return top.skip;
} else {
-
- if (top.name.empty() && (c == '@')) {
- top.name += c;
- return true;
- }
-
- top.in_what = Other;
+ bool in_name;
if (top.name.empty()) {
- top.name.clear();
top.name += c;
- top.do_check = commands.lookup(top.name.c_str());
- if (top.do_check == 0) top.do_check = "";
- return !asc_isspace(c);
+ in_name = true;
+ } else {
+ top.size--; // necessary?
+ in_name = false;
}
- top.do_check = commands.lookup(top.name.c_str());
- if (top.do_check == 0) top.do_check = "";
+ String name = top.name;
- if (asc_isspace(c)) { // swallow extra spaces
- top.in_what = Swallow;
- return true;
- } else if (c == '*') { // ignore * at end of commands
- return true;
+ pop_command();
+
+ const char *args = commands.lookup(name.c_str());
+
+ // later?
+ if (name == "begin")
+ push_command(top.in_what, top.skip);
+ // args = "s";
+ else if (name == "end")
+ pop_command();
+
+ // we might still be waiting for arguments
+ skip_opt_args;
+ if (*top.args) {
+ next_arg;
+ } else if (name == "[") {
+ // \[
+ push_command(DisplayMath, true);
+ } else if (name == "]") {
+ // \]
+ pop_command(); // pop DisplayMath
+ } else if (args && *args) {
+ push_command(top.in_what, top.skip, args);
}
- // continue o...
+ if (in_name || c == '*') // better way to deal with "*"?
+ return true;
+ else
+ return process_char(c); // start over
+ }
+
+ }
+
+ if (top.in_what == Comment) {
+ if (c == '\n') {
+ pop_command();
+ return false; // preserve newlines
+ } else {
+ return top.skip;
}
-
- } else if (top.in_what == Swallow) {
-
- if (asc_isspace(c))
- return true;
- else
- top.in_what = Other;
}
- if (c == '{')
- while (*top.do_check == 'O' || *top.do_check == 'o')
- ++top.do_check;
-
- if (*top.do_check == '\0')
- pop_command();
-
- if (c == '{') {
-
- if (top.in_what == Parm || top.in_what == Opt || top.do_check == '\0')
- push_command(Parm);
-
- top.in_what = Parm;
+ if (c == '%') {
+ push_command(Comment, !check_comments);
return true;
}
-
- if (top.in_what == Other) {
-
- if (c == '[') {
-
- top.in_what = Opt;
- return true;
-
- } else if (asc_isspace(c)) {
-
- return true;
-
+
+ if (c == '$') {
+ if (top.in_what != InlineMath) {
+ // $ begin
+ return push_command(InlineMath, true);
+ } else if (top.size > 1) {
+ // $ end
+ return pop_command();
} else {
-
- pop_command();
-
+ // $ -> $$
+ pop_command(); // pop InlineMath
+ if (top.in_what == DisplayMath)
+ // $$ end
+ return pop_command();
+ else
+ // $$ start
+ return push_command(DisplayMath, true);
}
-
- }
+ }
if (c == '\\') {
- prev_backslash = true;
- push_command(Name);
- return true;
+ return push_command(Name, true);
}
- if (top.in_what == Parm) {
-
- if (c == '}')
- return end_option('P','p');
- else
- return *top.do_check == 'p';
-
- } else if (top.in_what == Opt) {
+ if (c == '}' || c == ']') {
+ if (top.in_what == Scan) {
+ String env = top.name;
+ if (env.back() == '*')
+ env.pop_back();
+ bool skip = pop_command();
+ next_arg;
+ if (ignore_env.have(env)) {
+ stack[stack.size()-2].skip = true;
+ }
+ return skip;
+ } else {
+ bool skip = pop_command();
+ next_arg;
+ return skip;
+ }
+ }
- if (c == ']')
- return end_option('O', 'o');
+ if (c == '{') {
+ skip_opt_args;
+ if (*top.args == 'T')
+ return push_command(Text, false);
+ else if (*top.args == 's') // also do it below?
+ return push_command(Scan, true);
else
- return *top.do_check == 'o';
-
+ return push_command(top.in_what, top.skip || *top.args == 'p');
+ }
+
+ if (c == '[') {
+ if (*top.args == 'O' || *top.args == 'o' || !*top.args) {
+ return push_command(top.in_what, top.skip || *top.args == 'o');
+ }
+ // else: fall-through to treat it as argument
}
+
+ if (top.in_what == Scan)
+ top.name += c;
- return false;
+ // we might still be waiting for arguments
+ if (!asc_isspace(c)) {
+ skip_opt_args;
+ next_arg;
+ }
+
+ return top.skip;
}
void TexFilter::process(FilterChar * & str, FilterChar * & stop)
@@ -230,19 +262,24 @@ namespace {
FilterChar * cur = str;
while (cur != stop) {
- if (process_char(*cur))
+ bool hyphen = top.in_what == Name && top.size == 0
+ && (*cur == '-' || *cur == '/') && cur-str >= 2;
+ if (process_char(*cur)) {
*cur = ' ';
+ }
+ if (hyphen) {
+ FilterChar *i = cur-2, *j = cur+1;
+ *i = FilterChar(*i, FilterChar::sum(i, j));
+ i++;
+ while (j != stop)
+ *(i++) = *(j++);
+ *(stop-2) = *(stop-1) = FilterChar(0, 0);
+ cur--;
+ }
++cur;
}
}
- bool TexFilter::end_option(char u, char l) {
- top.in_what = Other;
- if (*top.do_check == u || *top.do_check == l)
- ++top.do_check;
- return true;
- }
-
//
// TexFilter::Commands
//
@@ -252,14 +289,14 @@ namespace {
while (!asc_isspace(value[p1])) {
if (value[p1] == '\0')
return make_err(bad_value, value,"",
- _("a string of 'o','O','p',or 'P'"));
+ _("a string of 'o', 'O', 'p', 'P', 's' or 'T'"));
++p1;
}
int p2 = p1 + 1;
while (asc_isspace(value[p2])) {
if (value[p2] == '\0')
return make_err(bad_value, value,"",
- _("a string of 'o','O','p',or 'P'"));
+ _("a string of 'o', 'O', 'p', 'P', 's' or 'T'"));
++p2;
}
String t1; t1.assign(value,p1);
--- ./modules/filter/tex-filter.info.orig 2004-06-27 00:54:00.000000000
-0400
+++ ./modules/filter/tex-filter.info 2011-02-05 16:44:49.000000000 -0500
@@ -16,15 +16,35 @@ DESCRIPTION check TeX comments
DEFAULT false
ENDOPTION
+OPTION ignore-env
+TYPE list
+DESCRIPTION LaTeX environments to be ignored
+# LaTeX
+#DEFAULT thebibliography
+DEFAULT equation
+DEFAULT eqnarray
+# AMS-LaTeX
+DEFAULT gather
+DEFAULT multline
+DEFAULT align
+DEFAULT flalign
+DEFAULT alignat
+# Babel
+DEFAULT otherlanguage
+ENDOPTION
+
OPTION command
TYPE list
DESCRIPTION TeX commands
+# plain TeX / LaTeX
DEFAULT addtocounter pp
DEFAULT addtolength pp
-DEFAULT alpha p
+DEFAULT alph p
+DEFAULT Alph p
DEFAULT arabic p
DEFAULT fnsymbol p
DEFAULT roman p
+DEFAULT Roman p
DEFAULT stepcounter p
DEFAULT setcounter pp
DEFAULT usecounter p
@@ -42,7 +62,8 @@ DEFAULT newtheorem poPo
DEFAULT newfont pp
DEFAULT documentclass op
DEFAULT usepackage op
-DEFAULT begin po
+# Do NOT change the next line!
+DEFAULT begin so
DEFAULT end p
DEFAULT setlength pp
DEFAULT addtolength pp
@@ -54,15 +75,17 @@ DEFAULT hyphenation p
DEFAULT pagenumbering p
DEFAULT pagestyle p
DEFAULT addvspace p
-DEFAULT framebox ooP
+DEFAULT framebox ooT
DEFAULT hspace p
DEFAULT vspace p
-DEFAULT makebox ooP
-DEFAULT parbox ooopP
-DEFAULT raisebox pooP
+DEFAULT hbox T
+DEFAULT vbox T
+DEFAULT makebox ooT
+DEFAULT parbox ooopT
+DEFAULT raisebox pooT
DEFAULT rule opp
DEFAULT sbox pO
-DEFAULT savebox pooP
+DEFAULT savebox pooT
DEFAULT usebox p
DEFAULT include p
DEFAULT includeonly p
@@ -76,13 +99,30 @@ DEFAULT fontshape p
DEFAULT fontsize pp
DEFAULT usefont pppp
DEFAULT documentstyle op
-DEFAULT cite p
+DEFAULT cite Op
DEFAULT nocite p
DEFAULT psfig p
DEFAULT selectlanguage p
DEFAULT includegraphics op
DEFAULT bibitem op
+DEFAULT bibliography p
+DEFAULT bibliographystyle p
DEFAULT geometry p
+# AMS-LaTeX
+DEFAULT address p
+DEFAULT email p
+DEFAULT mathbb p
+DEFAULT mathfrak p
+DEFAULT eqref p
+DEFAULT text T
+DEFAULT intertext T
+DEFAULT DeclareMathOperator pp
+DEFAULT DeclareMathAlphabet ppppp
+# hyperref
+DEFAULT href pP
+DEFAULT autoref p
+DEFAULT url p
+DEFAULT texorpdfstring Pp
ENDOPTION
#OPTION multi-byte
--- ./modules/filter/modes/tex.amf.orig 2004-10-30 12:00:07.000000000 -0400
+++ ./modules/filter/modes/tex.amf 2010-05-30 20:49:28.000000000 -0400
@@ -6,5 +6,8 @@ MAGIC /0:256:^[ \t]*\\documentclass\[[^\
DESCRIPTION mode for checking TeX/LaTeX documents
-FILTER url
+FILTER context
+OPTION clear-context-delimiters
+OPTION add-context-delimiters aspell:off aspell:on
+OPTION enable-context-visible-first
FILTER tex
--- ./manual/aspell.1.orig 2006-12-19 05:55:08.000000000 -0500
+++ ./manual/aspell.1 2011-02-05 14:30:00.000000000 -0500
@@ -181,7 +181,7 @@ encoding the document is expected to be
current locale.
.TP
\fB\-\-add-email\-quote=\fR\fI<list>\fR,
\fB\-\-rem-email\-quote=\fR\fI<list>\fR
-Add or Remove a list of email quote characters.
+Add or remove a list of email quote characters.
.TP
\fB\-\-email\-margin=\fR\fI<integer>\fR
Number of chars that can appear before the quote char.
@@ -191,14 +191,14 @@ Add or remove a list of HTML attributes
look inside alt= tags.
.TP
\fB\-\-add\-html\-skip=\fR\fI<list>\fR, \fB\-\-rem\-html\-skip=\fR\fI<list>\fR
-Add or remove a list of HTML attributes to always skip while spell
+Add or remove a list of HTML tags to always skip while spell
checking.
.TP
\fB\-\-add\-sgml\-check=\fR\fI<list>\fR,
\fB\-\-rem\-sgml\-check=\fR\fI<list>\fR
Add or remove a list of SGML attributes to always check for spelling.
.TP
\fB\-\-add\-sgml\-skip=\fR\fI<list>\fR, \fB\-\-rem\-sgml\-skip=\fR\fI<list>\fR
-Add or remove a list of SGML attributes to always skip while spell
+Add or remove a list of SGML tags to always skip while spell
checking.
.TP
\fB\-\-sgml\-extension=\fR\fI<list>\fR
@@ -208,7 +208,22 @@ SGML file extensions.
Check TeX comments.
.TP
\fB\-\-add\-tex\-command=\fR\fI<list>\fR,
\fB\-\-rem\-tex\-command=\fR\fI<list>\fR
-Add or Remove a list of TeX commands.
+Add or remove a list of TeX commands.
+.TP
+\fB\-\-add\-tex\-ignore\-env=\fR\fI<list>\fR,
\fB\-\-rem\-tex\-ignore\-env=\fR\fI<list>\fR
+Add or remove a list of LaTeX environments to skip while spell checking.
+.TP
+\fB\-\-add\-texinfo\-ignore=\fR\fI<list>\fR,
\fB\-\-rem\-texinfo\-ignore=\fR\fI<list>\fR
+Add or remove a list of Texinfo commands.
+.TP
+\fB\-\-add\-texinfo\-ignore\-env=\fR\fI<list>\fR,
\fB\-\-rem\-texinfo\-ignore\-env=\fR\fI<list>\fR
+Add or remove a list of Texinfo environments to ignore.
+.TP
+\fB\-\-context\-visible\-first, \fB\-\-dont\-context\-visible\-first
+Switch the context which should be visible to Aspell.
+.TP
+\fB\-\-add\-context\-delimiters=\fR\fI<list>\fR,
\fB\-\-rem\-context\-delimiters=\fR\fI<list>\fR
+Add or remove pairs of delimiters.
.SH RUN\-TOGETHER WORD OPTIONS
These may be used to control the behavior of run\-together words.
.TP
--- ./manual/aspell.texi.orig 2008-04-16 01:14:36.000000000 -0400
+++ ./manual/aspell.texi 2011-02-05 16:49:09.000000000 -0500
@@ -1065,6 +1065,10 @@ being readable text in LaTeX output from
@i{(list)}
@TeX{} commands
+@item tex-ignore-env
+@i{(list)}
+@TeX{} environments to always skip over
+
@item tex-check-comments
@i{(boolean)}
check @TeX{} comments
@@ -1363,13 +1367,15 @@ all their parameters and/or options chec
is
@example
-<command> <a list of p,P,o and Os>
+<command> <a list of 'p', 'P', 'o', 'O' and 'T's>
@end example
The first item is simply the command name. The second item controls
which parameters to skip over. A 'p' skips over a parameter while a
'P' doesn't. Similarly an 'o' will skip over an optional parameter
-while an 'O' doesn't. The first letter on the list will apply to the
+while an 'O' doesn't. A 'T' will force spell-checking of a parameter
+even if the command occurs within a parameter or an environment Aspell
+is told to skipped over. The first letter on the list will apply to the
first parameter, the second letter will apply to the second parameter
etc. If there are more parameters than letters Aspell will simply
check them as normal. For example the option
@@ -1392,6 +1398,25 @@ over the next optional parameter, if it
the second parameter --- even if the optional parameter is not present
--- and will check any additional parameters.
+@example
+add-tex-command foo T
+@end example
+
+@noindent
+will @emph{check} the first parameter of the @code{foo} command even
+if Aspell is currently skipping over an argument or environment. For
+example, if Aspell has been told to skip over the @code{bar}
+environment (@pxref{Ignoring LaTeX Environments}), then in the text
+
+@example
+\begin@{bar@} don't check \foo@{check@} don't check \end@{bar@}
+@end example
+
+@noindent
+it will nevertheless @emph{check} the argument to @code{foo}. This is
+useful to force checking of arguments to text-related commands like
+@code{hbox}, @code{text} or @code{intertext} inside math environments.
+
A @samp{*} at the end of the command is simply ignored. For example
the option
@@ -1414,6 +1439,50 @@ rem-tex-command foo
will remove the command foo, if present, from the list of @TeX{}
commands.
+@anchor{Ignoring LaTeX Environments}
+The option
+@option{add|rem-tex-ignore-env}
+controls which @TeX{} environments are skipped over. By default,
+Aspell will skip over math formulas inside @code{$...$}, @code{$$...$$}
+and @code{\[...\]} and over several common LaTeX and AMS-LaTeX math
+environments like @code{equation} and @code{gather}. For example,
+
+@example
+add-tex-ignore-env thebibliography
+@end example
+
+@noindent
+will tell Aspell to skip over the bibliography as well (which may or
+may not be a good idea). As with commands, skipping applies to the
+starred form of the environment as well.
+
+@example
+rem-tex-ignore-env equation
+@end example
+
+@noindent
+will make Aspell spell-check the contents of @code{equation} and
+@code{equation*} environments. Skipping the contents of @code{$...$},
+@code{$$...$$} and @code{\[...\]} cannot be turned off.
+
+Note that one can force spell-checking of arguments to TeX commands
+inside ignored environments with the 'T' parameter to the
+@option{add-tex-command} option.
+
+As a last resort, spell checking can be switched off by putting the
+text @code{aspell:off} into the file. Similarly, with @code{aspell:on}
+one can turn it on again. This can be useful for macro definitions,
+for example
+
+@example
+% aspell:off
+\def\doi#1@{\href@{http://dx.doi.org/#1@}@{doi:#1@}@}
+% aspell:on
+@end example
+
+@noindent
+This feature is implemented via the @ref{Context Filter}.
+
The TeX filter mode is also available via @option{latex} alias name.
@c The TeXfilter mode also contains a decoding and a encoding filter for
@@ -1514,6 +1583,7 @@ escapes
(@code{\(}) and extended (@code{\[comp1 comp2 @dots{}]}) form.
@end itemize
+@anchor{Context Filter}
@subsubsection Context Filter
The @emph{context} filter allows Aspell to distinguish between visible
demo.tex
Description: TeX document
_______________________________________________ Aspell-devel mailing list Aspell-devel@gnu.org http://lists.gnu.org/mailman/listinfo/aspell-devel
