texlive[64831] Build/source/texk/web2c/hitexdir: HiTeX: changing
commits+mruckert at tug.org
commits+mruckert at tug.org
Thu Oct 27 11:18:41 CEST 2022
Revision: 64831
http://tug.org/svn/texlive?view=revision&revision=64831
Author: mruckert
Date: 2022-10-27 11:18:40 +0200 (Thu, 27 Oct 2022)
Log Message:
-----------
HiTeX: changing prefered to preferred, adding a TeX Live compliant \input
Modified Paths:
--------------
trunk/Build/source/texk/web2c/hitexdir/ChangeLog
trunk/Build/source/texk/web2c/hitexdir/hitex.w
trunk/Build/source/texk/web2c/hitexdir/tex/hiltxpage.tex
trunk/Build/source/texk/web2c/hitexdir/tex/hiplainpage.tex
Modified: trunk/Build/source/texk/web2c/hitexdir/ChangeLog
===================================================================
--- trunk/Build/source/texk/web2c/hitexdir/ChangeLog 2022-10-26 23:49:22 UTC (rev 64830)
+++ trunk/Build/source/texk/web2c/hitexdir/ChangeLog 2022-10-27 09:18:40 UTC (rev 64831)
@@ -1,3 +1,12 @@
+2022-10-27 Martin Ruckert <ruckert at cs.hm.edu>
+
+ * documentation and typograhic improvements
+ * changing the keyword prefered to preferred
+ * rewriting the scan_file_name function to comply with TeX Live
+ Eliminating the special file name buffer and reverting to the
+ traditional use of the string pool.
+ * initializing hlog early to allow for early HINT format errors
+
2022-06-04 Martin Ruckert <ruckert at cs.hm.edu>
* documentation and typograhic improvements
Modified: trunk/Build/source/texk/web2c/hitexdir/hitex.w
===================================================================
--- trunk/Build/source/texk/web2c/hitexdir/hitex.w 2022-10-26 23:49:22 UTC (rev 64830)
+++ trunk/Build/source/texk/web2c/hitexdir/hitex.w 2022-10-27 09:18:40 UTC (rev 64831)
@@ -208,10 +208,37 @@
\def\RQ{{\tt\char'23}} % right quote in a string
\def\UL{{\tt\char`\_}} % underline character in a C identifier
\def\dotdot{\mathrel{.\,.}} % double dot, used only in math mode
+\setbox\MGbox=\hbox{\kern1pt$\rightarrow$\kern1pt} % C pointer to field
+\def\MG{\copy\MGbox}
@s dotdot TeX
@s alpha_file int
@s byte_file int
@s word_file int
+ at s uint8_t int
+ at s int16_t int
+ at s uint16_t int
+ at s int32_t int
+ at s uint32_t int
+ at s halfword int
+ at s nonnegative_integer int
+ at s small_number int
+ at s glue_ratio double
+ at s Xdimen int
+ at s Kern int
+ at s Font int
+ at s Glue int
+ at s Disc int
+ at s Lig int
+ at s Rule int
+ at s List int
+ at s Kind int
+ at s Info int
+ at s kpse_file_format_type int
+ at s Stretch int
+ at s in TeX
+ at s line normal
+ at s to do
+
@* Introduction.
This is Hi\TeX, a program derived from \TeX, extending its capabilities
using \eTeX and \Prote, and adding functions common to other engines from
@@ -381,7 +408,7 @@
therefore chosen to substitute their own home-grown rounding procedure.
@ The following is an outline of the program, whose
-components will be filled in later, using the conventions of \.{WEB}.
+components will be filled in later, using the conventions of \.{cweb}.
@.WEB@>
For example, the portion of the program called `\X\glob:Global
variables\X' below will be replaced by a sequence of variable declarations
@@ -393,6 +420,10 @@
all global variables, if desired. Similar remarks apply to the other
portions of the program.
+The program starts with inserting header files and occassionaly a function
+must be placed before declaring \TeX's macros, because the function
+uses identifiers that \TeX will declare as macros.
+
@p @<Header files and function declarations@>@;
@h
enum {@+@<Constants in the outer block@>@+};
@@ -418,16 +449,8 @@
intended to combine the advantages of the ``bottom up'' and ``top down''
approaches to the problem of understanding a somewhat complicated system.
-@ There is no need to declare labels in \CEE/, but occasionally
-it is necessary to insert header files and declare functions
-very early in the program. For example, the function |s_no|
-which uses |make_string| to convert a \CEE/ string into
-a string number is used in |initialize| and needs a forward
-declaration.
+@ There is no need to declare labels in \CEE/.
-@<Header files and function declarations@>=
-static int s_no(const char *str);
-
@ Some of the code below is intended to be used only when diagnosing the
strange behavior that sometimes occurs when \TeX\ is being installed or
when system wizards are fooling around with \TeX\ without quite knowing
@@ -457,24 +480,13 @@
@<Initialize whatever...@>=
@<Set initial values of key variables@>@/
#ifdef @!INIT
-if (iniversion) {@+@<Initialize table entries (done by \.{INITEX} only)@>@;@+}
+if (iniversion) /* \TeX\ Live*/
+ {@+@<Initialize table entries (done by \.{INITEX} only)@>@;@+}
#endif
@ The declaration of all basic type definitions needed by Hi\TeX\ are
contained in a system dependent header file.
- at s uint8_t int
- at s int16_t int
- at s uint16_t int
- at s int32_t int
- at s uint32_t int
- at s halfword int
- at s nonnegative_integer int
- at s small_number int
- at s glue_ratio double
- at s in TeX
- at s line normal
- at s to do
@<Header files and function declarations@>=
#include "hibasetypes.h"
@@ -900,9 +912,10 @@
relevant in |name_of_file| (the rest are blank)*/
@ To open files, \TeX\ used \PASCAL's |reset| function.
-Now we use the {\tt kpathsearch} library to implement new functions.
-Most of the code to find and open files is contained in two functions,
-|open_in| and |open_out|, defined later.
+We use the {\tt kpathsearch} library to implement new functions
+in the section on \TeX\ Live Integration.
+Here we give only the function prototypes.
+
\TeX's file-opening functions do not issue their own
error messages if something goes wrong. If a file identified by
|name_of_file| cannot be found,
@@ -913,44 +926,15 @@
@^system dependencies@>
@p
-static FILE*open_in(char*filename,kpse_file_format_type t,const char*rwb);
-static bool a_open_in(alpha_file *f) /*open a text file for input*/
-{@+f->f= open_in((char *)name_of_file+1,kpse_tex_format,"r");
- if (f->f!=NULL) get(*f);
- return f->f!=NULL && ferror(f->f)==0;
-}
-
-static bool b_open_in(byte_file *f) /*open a binary file for input*/
-{@+f->f= open_in((char *)name_of_file+1,kpse_tfm_format,"rb");
- if (f->f!=NULL) get(*f);
- return f->f!=NULL && ferror(f->f)==0;
-}
-
-static bool w_open_in(word_file *f) /*open a word file for input*/
-{@+
- f->f=NULL;
- if (name_of_file[1]!=0)
- f->f= open_in((char*)name_of_file+1,kpse_fmt_format,"rb");
- if (f->f!=NULL) get(*f);
- return f->f!=NULL && ferror(f->f)==0;
-}
-
-static FILE *open_out(const char *file_name, const char *file_mode);
-static bool a_open_out(alpha_file *f) /*open a text file for output*/
-{@+f->f=open_out((char*)name_of_file+1,"w");
- return f->f!=NULL && ferror(f->f)==0;@+
-}
-
-static bool b_open_out(byte_file *f) /*open a binary file for output*/
-{@+f->f=open_out((char *)name_of_file+1,"wb");
-return f->f!=NULL && ferror(f->f)==0;@+
-}
-
+static FILE*open_in(char*filename,kpse_file_format_type t,const char*rwb); /* \TeX\ Live*/
+static bool a_open_in(alpha_file *f); /*open a text file for input*/
+static bool b_open_in(byte_file *f); /*open a binary file for input*/
+static bool w_open_in(word_file *f); /*open a word file for input*/
+static FILE *open_out(const char *file_name, const char *file_mode); /* \TeX\ Live*/
+static bool a_open_out(alpha_file *f); /*open a text file for output*/
+static bool b_open_out(byte_file *f); /*open a binary file for output*/
#ifdef @!INIT
-static bool w_open_out(word_file *f) /*open a word file for output*/
-{@+f->f=open_out((char *)name_of_file+1,"wb");
- return f->f!=NULL && ferror(f->f)==0;@+
-}
+static bool w_open_out(word_file *f); /*open a word file for output*/
#endif
@ Files can be closed with the \ph\ routine `|pascal_close(f)|', which
@@ -1321,7 +1305,7 @@
@ Here is a similar routine, but it compares two strings in the string pool,
and it does not assume that they have the same length.
- at p static bool strn_eq_str(str_number @!s, str_number @!t)
+ at p static bool str_eq_str(str_number @!s, str_number @!t)
/*test equality of strings*/
{@+ /*loop exit*/
pool_pointer j, @!k; /*running indices*/
@@ -1338,16 +1322,6 @@
}
@t\4@>@<Declare \Prote\ procedures for strings@>@;
-static bool str_eq_str(str_number @!s, char *@!t)
- /*test equality of strings*/
-{@+int j, @!k; /*running indices*/
-if (length(s)!=(int)strlen(t)) return false;
-j=str_start[s];k=0;
-while (j < str_start[s+1])
- if (str_pool[j++]!=t[k++]) return false;
-return true;
-}
-
@ The initial values of |str_pool|, |str_start|, |pool_ptr|,
and |str_ptr| are computed by the \.{INITEX} program, based in part
on the information that \.{WEB} has output while processing \TeX.
@@ -1355,12 +1329,13 @@
@^string pool@>
@p
-static void get_strings_started(void) /*initializes the string pool*/
+static bool get_strings_started(void) /*initializes the string pool*/
{@+
int k, @!l; /*small indices or counters*/
pool_ptr=0;str_ptr=0;str_start[0]=0;
@<Make the first 256 strings@>;
@<Add the empty string to the string pool@>;
+return true;
}
@ @d app_lc_hex(A) l=A;
@@ -1411,7 +1386,7 @@
@ The |pool_file| variable is no longer needed and has been removed.
-@ Instead of reading the other strings from the \.{TEX.POOL} file,
+Instead of reading the other strings from the \.{TEX.POOL} file,
it is sufficient here to add the empty string.
@<Add the empty string to the string pool@>=
make_string();
@@ -1430,6 +1405,12 @@
return make_string();
}
+@ The function |s_no| is used in |initialize| and needs a forward
+declaration.
+
+@<Forward declarations@>=
+static int s_no(const char *str);
+
@* On-line and off-line printing.
Messages that are sent to a user's terminal and to the transcript-log file
are produced by several `|print|' procedures. These procedures will
@@ -1586,7 +1567,8 @@
@<Basic print...@>=
static void print(char *s) /* the simple version */
-{ @+while (*s!=0) print_char(*s++);@+
+{ @+if (s == NULL) s="???"; /*this can't happen*/
+ while (*s!=0) print_char(*s++);@+
}
static void printn(int @!s) /*prints string |s|*/
@@ -1821,7 +1803,7 @@
@^system dependencies@>
@<Error handling...@>=
-void print_err(char *s)
+static void print_err(char *s)
{@+if (interaction==error_stop_mode) wake_up_terminal;
if (filelineerrorstylep) print_file_line(); /* \TeX\ Live */
else print_nl("! ");
@@ -1841,7 +1823,7 @@
@ @<Set init...@>=
if (interaction_option<0) interaction=error_stop_mode;
-else interaction=interaction_option;
+else interaction=interaction_option; /* \TeX\ Live */
@ \TeX\ is careful not to call |error| when the print |selector| setting
might be unusual. The only possible values of |selector| at the time of
@@ -5488,7 +5470,7 @@
@^creation date@>
@^reference time@>
@^system dependencies@>
-k\TeX\ calls |tl_now| to obtain the current time as a |tm| structure.
+\TeX\ Live calls |tl_now| to obtain the current time as a |tm| structure.
@p static void fix_date_and_time(void)
{@+ struct tm *t=tl_now();
time=sys_time= t->tm_hour*60+t->tm_min;/*minutes since midnight*/
@@ -7516,7 +7498,7 @@
if (first==buf_size) overflow("buffer size", buf_size);
@:TeX capacity exceeded buffer size}{\quad buffer size@>
incr(in_open);push_input;index=in_open;@/
-source_filename_stack[index]=0; /* k\TeX\ */
+source_filename_stack[index]=0; /* \TeX\ Live */
eof_seen[index]=false;
grp_stack[index]=cur_boundary;if_stack[index]=cond_ptr;
line_stack[index]=line;start=first;state=mid_line;
@@ -7585,12 +7567,18 @@
@ The value of |par_loc| is the |eqtb| address of `\.{\\par}'. This quantity
is needed because a blank line of input is supposed to be exactly equivalent
-to the appearance of \.{\\par}; we must set |cur_cs=par_loc|
+to the appearance of \.{\\par}; we must set |cur_cs:=par_loc|
when detecting a blank line.
+The same is true for the input, for the warning message, since input
+is expected by default before every scanning and hence setting of
+\.{cur_cs}.
+
@<Glob...@>=
static pointer @!par_loc; /*location of `\.{\\par}' in |eqtb|*/
static halfword @!par_token; /*token representing `\.{\\par}'*/
+static pointer @!input_loc; /*location of `\.{\\input}' in |eqtb|*/
+static halfword @!input_token; /*token representing `\.{\\input}'*/
@ @<Put each...@>=
primitive("par", par_end, 256); /*cf.\ |scan_file_name|*/
@@ -8280,6 +8268,7 @@
@<Put each...@>=
primitive("input", input, 0);@/
@!@:input\_}{\.{\\input} primitive@>
+input_loc=cur_val;input_token=cs_token_flag+input_loc;
primitive("endinput", input, 1);@/
@!@:end\_input\_}{\.{\\endinput} primitive@>
@@ -10593,8 +10582,8 @@
@<Glob...@>=
static str_number @!cur_name; /*name of file just scanned*/
-static char *@!cur_area; /*file area just scanned, or \.{""}*/
-static char *@!cur_ext; /*file extension just scanned, or \.{""}*/
+static str_number @!cur_area; /*file area just scanned, or \.{""}*/
+static str_number @!cur_ext; /*file extension just scanned, or \.{""}*/
@ The file names we shall deal with for illustrative purposes have the
following structure: If the name contains `\.>' or `\.:', the file area
@@ -10610,7 +10599,6 @@
@<Glob...@>=
static pool_pointer @!area_delimiter; /*the most recent `\.>' or `\.:', if any*/
static pool_pointer @!ext_delimiter; /*the relevant `\..', if any*/
-static int @!cur_file_name_length;
@ Input files that can't be found in the user's area may appear in a standard
system area called |TEX_area|. Font metric files whose areas are not given
@@ -10627,14 +10615,8 @@
@ Here now is the first of the system-dependent routines for file name scanning.
@^system dependencies@>
-The filename is scanned into |cur_file_name|; |cur_ext| and |cur_aire|
-will point into this buffer.
-
- at d MAX_CUR_FILE_NAME 1024
-
- at p static char cur_file_name[MAX_CUR_FILE_NAME+1];
-static void begin_name(void)
-{@+area_delimiter=ext_delimiter=cur_file_name_length=0;
+ at p static void begin_name(void)
+{@+area_delimiter=0;ext_delimiter=0;
}
@ And here's the second. The string pool might change as the file name is
@@ -10643,21 +10625,17 @@
string, instead of assigning an absolute address like |pool_ptr| to them.
@^system dependencies@>
- at p static bool more_name(ASCII_code @!c)
-{@+if (c==' ') return false;
- if (cur_file_name_length>= MAX_CUR_FILE_NAME )
- { overflow("file name length", MAX_CUR_FILE_NAME);
- return false;
- }
- else
- { /*contribute |c| to the current string*/
- cur_file_name[cur_file_name_length]=c;
- if (c=='/')
- {@+area_delimiter=cur_file_name_length;ext_delimiter=0;
+
+The |ASCII_code| to use as a delimiter is specified also as a parameter.
+
+ at p static bool more_name(ASCII_code @!c, ASCII_code d)
+{@+if (c==d) return false;
+else{@+str_room(1);append_char(c); /*contribute |c| to the current string*/
+ if ((c=='>')||(c==':'))
+ {@+area_delimiter=cur_length;ext_delimiter=0;
}
- else if (c=='.') ext_delimiter=cur_file_name_length;
- cur_file_name_length++;
- return true;
+ else if (c=='.') ext_delimiter=cur_length;
+ return true;
}
}
@@ -10665,34 +10643,20 @@
@^system dependencies@>
@p static void end_name(void)
-{@+int i=0;
- if (str_ptr+1 > max_strings)
+{@+if (str_ptr+3 > max_strings)
overflow("number of strings", max_strings-init_str_ptr);
@:TeX capacity exceeded number of strings}{\quad number of strings@>
-cur_file_name[cur_file_name_length]=0;
-if (area_delimiter==0) cur_area=cur_file_name+cur_file_name_length;
-else @+cur_area=cur_file_name,i=area_delimiter+1;
-if (ext_delimiter==0) ext_delimiter=cur_file_name_length;
-cur_ext=cur_file_name+ext_delimiter;
-for (;i<ext_delimiter;i++) append_char(cur_file_name[i]);
-if (area_delimiter!=0) cur_file_name[area_delimiter+1]=0;
-cur_name=make_string();
-}
-
-static void set_cur_area_ext(str_number a, str_number e)
-{ int i;
- if (length(a)+length(e)+2>=MAX_CUR_FILE_NAME)
- { overflow("file name length", MAX_CUR_FILE_NAME);
- return;
+if (area_delimiter==0) cur_area=empty_string;
+else{@+cur_area=str_ptr;
+ str_start[str_ptr+1]=str_start[str_ptr]+area_delimiter;incr(str_ptr);
}
- cur_file_name_length=0;
- area_delimiter=0;
- for(i=str_start[a]; i<str_start[a+1];i++) cur_file_name[cur_file_name_length++]=str_pool[i];
- cur_file_name[cur_file_name_length++]=0;
- cur_area=cur_file_name;
- cur_ext=cur_file_name+cur_file_name_length;
- for(i=str_start[e]; i<str_start[e+1];i++) cur_file_name[cur_file_name_length++]=str_pool[i];
- cur_file_name[cur_file_name_length++]=0;
+if (ext_delimiter==0)
+ {@+cur_ext=empty_string;cur_name=make_string();
+ }
+else{@+cur_name=str_ptr;
+ str_start[str_ptr+1]=str_start[str_ptr]+ext_delimiter-area_delimiter-1;
+ incr(str_ptr);cur_ext=make_string();
+ }
}
@ Conversely, here is a routine that takes three strings and prints a file
@@ -10701,11 +10665,7 @@
@^system dependencies@>
@<Basic printing...@>=
-static void print_file_name(int @!n, char *@!a, char *@!e)
-{@+print(a);slow_print(n);print(e);
-}
-
-static void printn_file_name(int @!n, int @!a, int @!e)
+static void print_file_name(int @!n, int @!a, int @!e)
{@+slow_print(a);slow_print(n);slow_print(e);
}
@@ -10719,28 +10679,32 @@
if (k <= file_name_size) name_of_file[k]=xchr[c];
}
- at p static void pack_file_name(str_number @!n, char *@!a, char *@!e)
+ at p static void pack_file_name(str_number @!n, str_number @!a, str_number @!e, char *@!f)
{@+int k; /*number of positions filled in |name_of_file|*/
ASCII_code @!c; /*character being packed*/
int @!j; /*index into |str_pool|*/
k=0;
-while(*a!=0) append_to_name(so(*a++));
-for (j=str_start[n]; j<=str_start[n+1]-1; j++) append_to_name(so(str_pool[j]));
-while(*e!=0) append_to_name(so(*e++));
+for (j=str_start[a]; j<=str_start[a+1]-1; j++) append_to_name(so(str_pool[j]))@;
+for (j=str_start[n]; j<=str_start[n+1]-1; j++) append_to_name(so(str_pool[j]))@;
+if (f==NULL)
+ for (j=str_start[e]; j<=str_start[e+1]-1; j++) append_to_name(so(str_pool[j]))@;
+else
+ while(*f!=0) append_to_name(so(*f++))@;
if (k <= file_name_size) name_length=k;@+else name_length=file_name_size;
name_of_file[name_length+1]=0;
}
-@ The global variable |TEX_format_default| is no longer needed
+
+@ \TeX\ Live does not use the global variable |TEX_format_default|. It is no longer needed
to supply the text for default system areas
and extensions related to format files.
-@ Consequently there is no initialization of |TEX_format_default| either.
+@ Consequently \TeX\ Live does not need the initialization of |TEX_format_default| either.
-@ And there is no need to check the length of |TEX_format_default|.
+@ And \TeX\ Live does not check the length of |TEX_format_default|.
-@ The |format_extension|, however, is needed later on
+@ The |format_extension|, however, is needed by \TeX\ Live
to create the format name from the job name.
@d format_extension ".fmt"
@@ -10752,9 +10716,9 @@
contains the first line of input in |buffer[loc dotdot(last-1)]|, where
|loc < last| and |buffer[loc]!=' '|.
-k\TeX\ uses the {\tt kpathsearch} library to implement access to files.
-To do so we declare |open_fmt_file| here and postpone the
-actual implementation.
+\TeX\ Live uses the {\tt kpathsearch} library to implement access to files.
+|open_fmt_file| is declared here and the
+actual implementation is in the section on \TeX\ Live Integration.
@<Declare the function called |open_fmt_file|@>=
static bool open_fmt_file(void);
@@ -10793,16 +10757,37 @@
@ Now let's consider the ``driver''
routines by which \TeX\ deals with file names
in a system-independent manner. First comes a procedure that looks for a
-file name in the input by calling |get_x_token| for the information.
+file name.
+There are two ways to specify the file name: as a general text argument
+or as a token (after expansion).
+The traditional token delimiter is the space. For a file name, however,
+a double quote is used as the token delimiter if the token starts
+with a double quote.
+Once the |area_delimiter| and the |ext_delimiter| are defined,
+the final processing is shared for all variants.
+
+When starting, \.{\\relax} is skipped as well as blanks and non-calls.
+Then a test for the |left_brace| will branch to the code for scanning
+a general text.
+% consider the case where a double quote is made an active character
+% equivalent to a left brace!
+
@p static void scan_file_name(void)
{@+
+ASCII_code d; /*the delimiter for |more_name|*/
+pool_pointer @!j, k; /*index into |str_pool|*/
+int @!old_setting; /*holds |selector| setting*/
name_in_progress=true;begin_name();
-@<Get the next non-blank non-call...@>;
+d=' '; /*traditional token delimiter*/
+@<Get the next non-blank non-relax...@>;
+if (cur_cmd==left_brace)
+ @<Define a general text file name and |goto done|@>@;
+if (cur_chr=='"') {@+d='"';get_x_token();}
loop at +{@+if ((cur_cmd > other_char)||(cur_chr > 255)) /*not a character*/
{@+back_input();goto done;
}
- if (!more_name(cur_chr)) goto done;
+ if (!more_name(cur_chr, d)) goto done;
get_x_token();
}
done: end_name();name_in_progress=false;
@@ -10836,18 +10821,23 @@
|job_name!=0|. It ignores and changes the current settings of |cur_area|
and |cur_ext|.
- at d pack_cur_name pack_file_name(cur_name, cur_area, cur_ext)
+ at d pack_cur_name(A) if (cur_ext==empty_string) pack_file_name(cur_name, cur_area, cur_ext,A);
+ else pack_file_name(cur_name, cur_area, cur_ext,NULL)
@p static void pack_job_name(char *@!s) /*|s==".log"|, |".dvi"|, or
|format_extension|*/
-{@+cur_area="";cur_ext=s;
-cur_name=job_name;pack_cur_name;
+{@+cur_area=empty_string;cur_ext=empty_string;
+cur_name=job_name;pack_cur_name(s);
}
@ If some trouble arises when \TeX\ tries to open a file, the following
routine calls upon the user to supply another file name. Parameter~|s|
is used in the error message to identify the type of file; parameter~|e|
-is the default extension if none is given. Upon exit from the routine,
+is the default extension if none is given.
+We handle the specification of a file name with possibly spaces in
+double quotes (the last one is optional if this is the end of line
+i.e. the end of the buffer).
+Upon exit from the routine,
variables |cur_name|, |cur_area|, |cur_ext|, and |name_of_file| are
ready for another attempt at file opening.
@@ -10867,15 +10857,15 @@
fatal_error("*** (job aborted, file error in nonstop mode)");
@.job aborted, file error...@>
clear_terminal;prompt_input(": ");@<Scan file name in the buffer@>;
-if (cur_ext[0]==0) cur_ext=e;
-pack_cur_name;
+pack_cur_name(e);
}
@ @<Scan file name in the buffer@>=
-{@+begin_name();k=first;
+{@+ASCII_code d=' '; /*traditional delimiter*/
+begin_name();k=first;
while ((buffer[k]==' ')&&(k < last)) incr(k);
loop at +{@+if (k==last) goto done;
- if (!more_name(buffer[k])) goto done;
+ if (!more_name(buffer[k], d)) goto done;
incr(k);
}
done: end_name();
@@ -10908,7 +10898,7 @@
int @!l; /*end of first input line*/
char @!months[]=" JANFEBMARAPRMAYJUNJULAUGSEPOCTNOVDEC"; /*abbreviations of month names*/
old_setting=selector;
-if (job_name==0) job_name=s_no(c_job_name?c_job_name:"texput"); /* k\TeX\ */
+if (job_name==0) job_name=s_no(c_job_name?c_job_name:"texput"); /* \TeX\ Live */
pack_job_name(".fls");
recorder_change_filename((char *)name_of_file+1);
@.texput@>
@@ -10970,8 +10960,7 @@
@p static void start_input(void) /*\TeX\ will \.{\\input} something*/
{@+
scan_file_name(); /*set |cur_name| to desired file name*/
-if (cur_ext[0]==0) cur_ext=".tex";
-pack_cur_name;
+pack_cur_name(".tex");
loop at +{@+begin_file_reading(); /*set up |cur_file| and new level of input*/
if (a_open_in(&cur_file)) goto done;
end_file_reading(); /*remove the level that didn't work*/
@@ -10978,10 +10967,10 @@
prompt_file_name("input file name",".tex");
}
done: name=a_make_name_string(&cur_file);@/
-source_filename_stack[in_open]=name; /* k\TeX\ */
+source_filename_stack[in_open]=name; /* \TeX\ Live*/
if (job_name==0)
{@+if (c_job_name==NULL) job_name=cur_name;
- else job_name=s_no(c_job_name); open_log_file(); /* k\TeX\ */
+ else job_name=s_no(c_job_name); open_log_file(); /* \TeX\ Live*/
} /*|open_log_file| doesn't |show_context|, so |limit|
and |loc| needn't be set to meaningful values yet*/
if (term_offset+length(name) > max_print_line-2) print_ln();
@@ -11522,7 +11511,7 @@
@d abort goto bad_tfm /*do this when the \.{TFM} data is wrong*/
- at p static internal_font_number read_font_info(pointer @!u, str_number @!nom, char *@!aire,
+ at p static internal_font_number read_font_info(pointer @!u, str_number @!nom, str_number @!aire,
scaled @!s) /*input a \.{TFM} file*/
{@+
int k; /*index into |font_info|*/
@@ -11555,7 +11544,7 @@
@.TFtoPL@> @.PLtoTF@>
@d start_font_error_message print_err("Font ");sprint_cs(u);
- print_char('=');print_file_name(nom, aire,"");
+ print_char('=');print_file_name(nom, aire, empty_string);
if (s >= 0)
{@+print(" at ");print_scaled(s);print("pt");
}
@@ -11589,7 +11578,7 @@
@ @<Open |tfm_file| for input@>=
file_opened=false;
-pack_file_name(nom, "",".tfm"); /* k\TeX\ */
+pack_file_name(nom, empty_string,empty_string,".tfm"); /* \TeX\ Live */
if (!b_open_in(&tfm_file)) abort;
file_opened=true
@@ -11830,7 +11819,7 @@
if (char_exists(qw)) font_false_bchar[f]=non_char;
}
font_name[f]=nom;
-font_area[f]=s_no(aire);
+font_area[f]=aire;
font_bc[f]=bc;font_ec[f]=ec;font_glue[f]=null;
adjust(char_base);adjust(width_base);adjust(lig_kern_base);
adjust(kern_base);adjust(exten_base);
@@ -16341,7 +16330,7 @@
static void fin_align(void)
{@+pointer @!p, @!q, @!r, @!s, @!u, @!v; /*registers for the list operations*/
scaled @!t, @!w; /*width of column*/
-bool x=false; /* indicates an extende alignment */
+bool x=false; /* indicates an extended alignment */
scaled @!o; /*shift offset for unset boxes*/
halfword @!n; /*matching span amount*/
scaled @!rule_save; /*temporary storage for |overfull_rule|*/
@@ -20120,7 +20109,7 @@
control point.
@p
-void update_last_values(pointer p)
+static void update_last_values(pointer p)
{@<Update the values of |last_glue|, |last_penalty|, and |last_kern|@>;
}
@@ -24255,7 +24244,7 @@
@<If this font has already been loaded...@>=
flushable_string=str_ptr-1;
for (f=font_base+1; f<=font_ptr; f++)
- if (strn_eq_str(font_name[f], cur_name)&&str_eq_str(font_area[f], cur_area))
+ if (str_eq_str(font_name[f], cur_name)&&str_eq_str(font_area[f], cur_area))
{@+if (cur_name==flushable_string)
{@+flush_string;cur_name=font_name[f];
}
@@ -24363,9 +24352,8 @@
}
if (c!=0)
{@+scan_optional_equals();scan_file_name();
- if (cur_ext[0]==0) cur_ext=".tex";
- pack_cur_name;
- if (a_open_in(&read_file[n])) read_open[n]=just_open;
+ pack_cur_name(".tex");
+ if (a_open_in(&read_file[n])) read_open[n]=just_open;
}
}
@@ -24607,14 +24595,16 @@
before \TeX's tables are loaded.
@<Glob...@>=
-static str_number @!format_ident;
+static str_number @!format_ident, frozen_format_ident;
@ @<Set init...@>=
-format_ident=0;
+format_ident=frozen_format_ident=0;
-@ @<Initialize table entries...@>=
-format_ident=s_no(" (INITEX)");
+@ We keep a copy of the initial value, be able to test for it later.
+@<Initialize table entries...@>=
+format_ident=frozen_format_ident=s_no(" (INITEX)");
+
@ @<Declare act...@>=
#ifdef @!INIT
static void store_fmt_file(void)
@@ -24837,6 +24827,7 @@
@<Dump regions 1 to 4 of |eqtb|@>;
@<Dump regions 5 and 6 of |eqtb|@>;
dump_int(par_loc);dump_int(write_loc);@/
+dump_int(input_loc);@/
@<Dump the hash table@>@;
@ @<Undump the table of equivalents@>=
@@ -24844,6 +24835,8 @@
undump(hash_base, frozen_control_sequence, par_loc);
par_token=cs_token_flag+par_loc;@/
undump(hash_base, frozen_control_sequence, write_loc);@/
+undump(hash_base, frozen_control_sequence, input_loc);
+input_token=cs_token_flag+input_loc;@/
@<Undump the hash table@>@;
@ The table of equivalents usually contains repeated information, so we dump it
@@ -24967,7 +24960,7 @@
dump_int(font_bchar[k]);
dump_int(font_false_bchar[k]);@/
print_nl("\\font");printn_esc(font_id_text(k));print_char('=');
-printn_file_name(font_name[k], font_area[k],empty_string);
+print_file_name(font_name[k], font_area[k],empty_string);
if (font_size[k]!=font_dsize[k])
{@+print(" at ");print_scaled(font_size[k]);print("pt");
}
@@ -25077,7 +25070,7 @@
@ @<Undump a couple more things and the closing check word@>=
undump(batch_mode, error_stop_mode, interaction);
-if (interaction_option>=0) interaction=interaction_option;
+if (interaction_option>=0) interaction=interaction_option; /* \TeX\ Live */
undump(0, str_ptr, format_ident);
undump_int(x);
if ((x!=69069)||eof(fmt_file)) goto bad_fmt
@@ -25165,7 +25158,8 @@
@p
int main(int argc, char *argv[]) {@! /*|start_here|*/
- main_init(argc, argv); /* k\TeX\ */
+hlog=stderr;
+ main_init(argc, argv); /* \TeX\ Live */
history=fatal_error_stop; /*in case we quit during initialization*/
t_open_out; /*open the terminal for output*/
if (ready_already==314159) goto start_of_TEX;
@@ -25179,7 +25173,7 @@
get_strings_started();
initialize(); /*set global variables to their starting values*/
#ifdef @!INIT
-if (iniversion)
+if (iniversion) /* \TeX\ Live */
{ init_prim(); /*call |primitive| for each primitive*/
init_str_ptr=str_ptr;init_pool_ptr=pool_ptr;fix_date_and_time(); }
#endif
@@ -25575,7 +25569,7 @@
@d setstream_number(A) type(A+1)
@d setstream_insertion(A) subtype(A+1)
@d setstream_mag(A) link(A+1) /* magnification factor */
- at d setstream_prefered(A) type(A+2)
+ at d setstream_preferred(A) type(A+2)
@d setstream_next(A) subtype(A+2)
@d setstream_ratio(A) link(A+2) /* split ratio */
@d setstream_max(A) info(A+3) /* extended dimension number */
@@ -25718,11 +25712,11 @@
case setstream_node: print_esc("HINTsetstream");@+break;
case stream_node: print_esc("HINTstream");@+break;
case param_node: print("[HINT internal: parameter list]");@+break;
- case par_node: print("[HINT internal: paragraf]");@+break;
+ case par_node: print("[HINT internal: paragraph]");@+break;
case disp_node: print("[HINT internal: display]");@+break;
case baseline_node: print("[HINT internal: baselineskip]");@+break;
case hpack_node: print("[HINT internal: hpack]");@+break;
- case vpack_node: print("[HINT internal: vpacky");@+break;
+ case vpack_node: print("[HINT internal: vpack");@+break;
case hset_node: print("[HINT internal: hset]");@+break;
case vset_node: print("[HINT internal: vset]");@+break;
case align_node: print("[HINT internal: align]");@+break;
@@ -25783,7 +25777,7 @@
}
case start_link_node:
if (abs(mode) == vmode)
- fatal_error("HiTeX startlink cannot be used in vertical mode");
+ fatal_error("HINTstartlink cannot be used in vertical mode");
else
{ new_whatsit(start_link_node,link_node_size);
scan_label(tail);
@@ -25791,7 +25785,7 @@
break;
case end_link_node:
if (abs(mode) == vmode)
- fatal_error("HiTeX endlink cannot be used in vertical mode");
+ fatal_error("HINTendlink cannot be used in vertical mode");
else
new_whatsit(end_link_node,link_node_size);
break;
@@ -25858,10 +25852,10 @@
s=new_setstream_node(n);
link(s)=setpage_streams(t); setpage_streams(t)=s;
loop {
- if (scan_keyword("prefered"))
+ if (scan_keyword("preferred"))
{@+scan_eight_bit_int();
if (cur_val!=255)
- setstream_prefered(s)=hget_stream_no(cur_val); }
+ setstream_preferred(s)=hget_stream_no(cur_val); }
else if (scan_keyword("next"))
{@+scan_eight_bit_int();
if (cur_val!=255)
@@ -25896,11 +25890,11 @@
}
@ @<Declare procedures needed in |do_extension|@>=
-void scan_spaces(void)
+static void scan_spaces(void)
{ @<Get the next non-blank non-call token@>;
back_input();
}
-void scan_destination(pointer p)
+static void scan_destination(pointer p)
{@+if (scan_keyword("name"))
{ label_has_name(p)=1;
scan_toks(false,true); label_ptr(p)=def_ref;
@@ -25916,7 +25910,7 @@
}
scan_spaces();
}
-void scan_label(pointer p)
+static void scan_label(pointer p)
{ if (!scan_keyword("goto"))
print_err("keyword `goto' inserted");
scan_destination(p);
@@ -25950,7 +25944,7 @@
@ @<Implement \.{\\openout}@>=
{@+new_write_whatsit(open_node_size);
scan_optional_equals();scan_file_name();@/
-open_name(tail)=cur_name;open_area(tail)=s_no(cur_area);open_ext(tail)=s_no(cur_ext);
+open_name(tail)=cur_name;open_area(tail)=cur_area;open_ext(tail)=cur_ext;
}
@ When `\.{\\write 12\{...\}}' appears, we scan the token list `\.{\{...\}}'
@@ -25998,7 +25992,7 @@
@ @<Display the whatsit...@>=
switch (subtype(p)) {
case open_node: {@+print_write_whatsit("openout", p);
- print_char('=');printn_file_name(open_name(p), open_area(p), open_ext(p));
+ print_char('=');print_file_name(open_name(p), open_area(p), open_ext(p));
} @+break;
case write_node: {@+print_write_whatsit("write", p);
print_mark(write_tokens(p));
@@ -26017,7 +26011,7 @@
print_int(param_type(p));print_char(',');print_int(param_no(p));
print_char(':');print_int(param_value(p).i);
break;
-case par_node: print_esc("paragraf(");
+case par_node: print_esc("paragraph(");
print_xdimen(par_extent(p));
print(", ");
print_int(par_penalty(p));
@@ -26090,7 +26084,7 @@
case setstream_node:
print_esc("HINTsetstream");print_int(setstream_insertion(p));
print_char('(');print_int(setstream_number(p));print_char(')');
- if (setstream_prefered(p)!=255) { print(" prefered ");print_int(setstream_prefered(p)); }
+ if (setstream_preferred(p)!=255) { print(" preferred ");print_int(setstream_preferred(p)); }
if (setstream_ratio(p)>0) { print(" ratio ");print_int(setstream_ratio(p)); }
if (setstream_next(p)!=255) { print(" next ");print_int(setstream_next(p)); }
append_char('.');
@@ -26487,10 +26481,9 @@
else{@+if (write_open[j]) a_close(&write_file[j]);
if (subtype(p)==close_node) write_open[j]=false;
else if (j < 16)
- {@+cur_name=open_name(p);
- set_cur_area_ext(open_area(p),open_ext(p));
- if (cur_ext[0]==0) cur_ext=".tex";
- pack_cur_name;
+ {@+cur_name=open_name(p);cur_area=open_area(p);
+ cur_ext=open_ext(p);
+ pack_cur_name(".tex");
while (!a_open_out(&write_file[j]))
prompt_file_name("output file name",".tex");
write_open[j]=true;
@@ -26567,10 +26560,10 @@
@<Enable \eTeX\ and furthermore Prote, if requested@>=
#ifdef @!INIT
-if (iniversion && (buffer[loc]=='*'||etexp))
+if (iniversion && (buffer[loc]=='*'||etexp)) /* \TeX\ Live */
{@+no_new_control_sequence=false;
@<Generate all \eTeX\ primitives@>@;
- if (buffer[loc]=='*') incr(loc);
+ if (buffer[loc]=='*') incr(loc); /* \TeX\ Live */
eTeX_mode=1; /*enter extended mode*/
@<Initialize variables for \eTeX\ extended mode@>@;
if (buffer[loc]=='*'||ltxp) {@+
@@ -29063,7 +29056,6 @@
and converts it to a string number that is returned in |info(garbage)|.
Neither the token list nor the string (obviously) are flushed.
-
@<Forward declarations@>=
static void toks_to_str(void);
@@ -29671,7 +29663,6 @@
{@+dig[0]=md5_digest[k]%16;@+dig[1]=md5_digest[k]/16;print_the_digs(2);
} @+break;
-@ If the string obtained is supposed to be a file name,
@ This is something that we will be doing several times. We have scanned
a general text. The result is a token list that we will interpret as a
file name. We must then put this name in |name_of_file| and try to open
@@ -30350,9 +30341,9 @@
@* Hi\TeX.
In the following we present macros, variables, and routines that
implement the various features that have been used above to replace
-\TeX's native behaviour.
+\TeX's native behavior.
-@ Following the implemenentation of other engines,
+@ Following the implementation of other engines,
the new engine returns a version number as an integer
extending the cases for |last_item|. Since the additional
primitives that we define are specific to the \HINT\ format,
@@ -30362,7 +30353,7 @@
@d HINT_version_code (eTeX_last_last_item_cmd_mod+7) /* \.{\\HINTversion} */
@d HINT_subversion_code (eTeX_last_last_item_cmd_mod+8) /* \.{\\HINTsubversion} */
-@ Now this new primitive needs its implementation
+@ Now this new primitive needs its implementation.
@<Cases of |last_item| for |print_cmd_chr|@>=
case HINT_version_code: print_esc("HINTversion");@+break;
@@ -30374,32 +30365,32 @@
@ The implementation reuses code that has been written as part of
-the \HINT\ file format specification; therfore we start with three
+the \HINT\ file format specification; therefore we start with three
include files containing the necessary declarations.
-We end with a list of forward declarations for all the functions
-and variables that are used above but are defined below.
- at p
+@<Header files and function declarations@>=
#include "hierror.h"
#include "hiformat.h"
#include "hiput.h"
+@ @p
@<Hi\TeX\ macros@>@;
@<Hi\TeX\ variables@>@;
@<Hi\TeX\ function declarations@>@;
-@<Hi\TeX\ auxiliar routines@>@;
+@<Hi\TeX\ auxiliary routines@>@;
@<Hi\TeX\ routines@>@;
-@ @<Forward declarations@>=
+@ This is a list of forward declarations for all the functions
+and variables that are used above but are defined below.
+
+
+ @<Forward declarations@>=
static void hout_allocate(void);
-static void hout_init(void);
static void hint_open(void);
static void hint_close(void);
-static void hint_debug_help(void);
-
static void hyphenate_word(void);
static void hline_break(int final_widow_penalty);
static void execute_output(pointer p);
@@ -30420,39 +30411,13 @@
static pointer new_setstream_node(eight_bits n);
static pointer new_setpage_node(eight_bits k, str_number n);
static pointer new_disp_node(void);
-static pointer new_image_node(str_number n, char *a, char *e);
+static pointer new_image_node(str_number n, str_number a, str_number e);
static void new_param_node(eight_bits t, eight_bits n, int v);
-@*1 Debugging options.
-Hi\TeX\ has a few additional command line options that we define next.
-
-@<Hi\TeX\ routines@>=
-#ifdef DEBUG
-static void hint_debug_help(void)
-{
-fprintf(stderr,@/
- "To generate HINT format debug output use the option\n"
- " -hint-debug=XX "@/
- at t\qquad@>"\t XX is a hexadecimal value. OR together these values:\n");@/
-fprintf(stderr,"\t\t\t XX=%04X \t basic debugging\n", DBGBASIC);@/
-fprintf(stderr,"\t\t\t XX=%04X \t tag debugging\n", DBGTAGS);@/
-fprintf(stderr,"\t\t\t XX=%04X \t node debugging\n",DBGNODE);@/
-fprintf(stderr,"\t\t\t XX=%04X \t definition debugging\n", DBGDEF);@/
-fprintf(stderr,"\t\t\t XX=%04X \t directory debugging\n", DBGDIR);@/
-fprintf(stderr,"\t\t\t XX=%04X \t range debugging\n",DBGRANGE);@/
-fprintf(stderr,"\t\t\t XX=%04X \t float debugging\n", DBGFLOAT);@/
-fprintf(stderr,"\t\t\t XX=%04X \t compression debugging\n", DBGCOMPRESS);@/
-fprintf(stderr,"\t\t\t XX=%04X \t buffer debugging\n", DBGBUFFER);@/
-fprintf(stderr,"\t\t\t XX=%04X \t TeX debugging\n", DBGTEX);@/
-fprintf(stderr,"\t\t\t XX=%04X \t page debugging\n", DBGPAGE);@/
-fprintf(stderr,"\t\t\t XX=%04X \t font debugging\n", DBGFONT);@/
-exit(0);
-}
-#endif
@*1 Creating new whatsit nodes.
The following functions create nodes for paragraphs, displayed equations, baseline skips,
-hpack nodes, vpack nodes, hset nodes, vset nodes, and image nodes.
+hpack nodes, vpack nodes, hset nodes, and vset nodes.
@<Hi\TeX\ routines@>=
static pointer new_par_node(void)
@@ -30460,9 +30425,7 @@
p=get_node(par_node_size);
type(p)=whatsit_node;
subtype(p)=par_node;
- par_params(p)=null;
- par_list(p)=null;
- par_extent(p)=null;
+ par_params(p)= par_list(p)= par_extent(p)=null;
depth(p)=0;
return p;
}
@@ -30473,9 +30436,7 @@
p=get_node(disp_node_size);
type(p)=whatsit_node;
subtype(p)=disp_node;
- display_params(p)=null;
- display_formula(p)=null;
- display_eqno(p)=null;
+ display_params(p)= display_formula(p)= display_eqno(p)=null;
return p;
}
@@ -30495,8 +30456,7 @@
subtype(p)=hpack_node;
width(p)=depth(p)=height(p)=shift_amount(p)=0;
pack_limit(p)=max_dimen;
- pack_extent(p)=null;
- list_ptr(p)=null;
+ pack_extent(p)= list_ptr(p)=null;
return p;
}
@@ -30506,15 +30466,17 @@
type(p)=whatsit_node;
subtype(p)=hset_node;
width(p)=depth(p)=height(p)=shift_amount(p)=set_stretch(p)=set_shrink(p)=0;
- set_extent(p)=null;
- list_ptr(p)=null;
+ set_extent(p)= list_ptr(p)=null;
return p;
}
+@ When creating a new image node, we could use the |kpse_find_tex|
+function to get image files from the same directory, where we also
+get the \TeX\ input files. Here we use the simpler method from plain \TeX.
-
-static pointer new_image_node( str_number n, char *a, char *e)
+@<Hi\TeX\ routines@>=
+static pointer new_image_node( str_number n, str_number a, str_number e)
{ pointer p;
int i;
char *fn;
@@ -30521,16 +30483,10 @@
int l;
p=get_node(image_node_size);type(p)=whatsit_node;subtype(p)=image_node;
image_name(p)=n;
- image_area(p)=s_no(a);
- image_ext(p)=s_no(e);
+ image_area(p)=a;
+ image_ext(p)=e;
fn=hfile_name(n,a,e);
-#if 0
- fn=kpse_find_tex(fn);
-#endif
i=hnew_file_section(fn);
-#if 0
- free(fn);
-#endif
image_no(p)=i;
image_xwidth(p)=image_xheight(p)=image_alt(p)=null;
return p;
@@ -30537,14 +30493,26 @@
}
@*1 Creating parameter nodes.
-Parameter nodes are added to the current list using the |new_param_node| function.
+The |new_param_node| function
+adds parameter nodes to the current list.
It should be possible to check the parameter values against those
stored in the definition section and remove the ones that
are unchanged. It would make the parameter lists shorter, saving
some time when setting and restoring them later.
There is probably not much savings in memory space, because
-most of the times a reference number is found for the parameter list.
-@<Create the parameter node@>=
+most of the time a reference number is found for the parameter list.
+
+
+@<Hi\TeX\ routines@>=
+static void new_param_node(uint8_t t, uint8_t n, int v)
+{ @+pointer p;
+ @<Create the parameter node@>@;
+ @<Initialize the parameter node@>@;
+ link(p)=link(temp_head);
+ link(temp_head)=p;
+}
+
+@ @<Create the parameter node@>=
p=get_node(param_node_size);
type(p)=whatsit_node;
subtype(p)=param_node;
@@ -30561,15 +30529,6 @@
QUIT("Undefined parameter type %d",t);
}
-@ @<Hi\TeX\ routines@>=
-static void new_param_node(uint8_t t, uint8_t n, int v)
-{ @+pointer p;
- @<Create the parameter node@>@;
- @<Initialize the parameter node@>@;
- link(p)=link(temp_head);
- link(temp_head)=p;
-}
-
@*1 Hyphenation.
While the breaking of a paragraph into lines must be postponed because
{\tt hsize} is not known, hyphenation should be done as part of Hi\TeX\
@@ -30578,19 +30537,21 @@
There is a fine point to observe here: \TeX\ will consider a word as
a candidate for automatic hyphenation only if the world ``follows'' after a
-glue. (For the exact rules, see Appendix H.)
+glue. (For the exact rules, see Appendix H of the \TeX-book.)
As a consequence, \TeX\ usually does not submit the first word of a
paragraph to its hyphenation routine.
Viewing paragraphs that start with a lengthy word on a narrow display
therefore often look more unsightly than necessary: the long word sticks out
into the right margin as much as it can. To remedy this situation,
-Hi\TeX\ has a ``{\tt -f}'' option. If set Hi\TeX\ will deviate from
-\TeX's rules and submit the first word of a paragraph to the hyphenation algorithm.
+Hi\TeX\ has a ``{\tt [-no]-hyphenate-first-word}'' option.
+If set, which is the default, Hi\TeX\ will deviate from
+\TeX's rules and submit the first word of a paragraph
+to the hyphenation algorithm.
The next problem arises from \TeX's multipass approach to line breaking
and the attempt to have Hi\TeX\ choose exactly the same line breaks as
\TeX\ does:
-\TeX\ distingishes between discretionary breaks inserted by the author of a
+\TeX\ distinguishes between discretionary breaks inserted by the author of a
text, and discretionary breaks discovered by the hyphenation routine.
The latter, called here ``automatic'', are used only in pass two and three
of the line breaking routine.
@@ -30622,7 +30583,7 @@
par_extent(pp)=new_xdimen(dimen_par(hsize_code),
dimen_par_hfactor(hsize_code),dimen_par_vfactor(hsize_code));
else
- @<fix the use of parshape = 1 indent length@>@;
+ @<fix simple use of parshape@>@;
link(temp_head)=link(head);
if (is_char_node(tail))
{ tail_append(new_penalty(inf_penalty))@;
@@ -30759,13 +30720,13 @@
}
@ Currently Hi\TeX\ dos not implement the parshape feature of \TeX.
-The implementation of {\tt \BS list} in \LaTeX\ does however depends
-on a simple us of parshape where all lines have the same length
+The implementation of {\tt \BS list} in \LaTeX\ does, however, depend
+on a simple use of parshape where all lines have the same length
and indentation. We cover this special case be using a hanging
indentation and adjusting the paragraph width by the difference
of the normal {\tt \BS hsize} and the given length.
-@<fix the use of parshape = 1 indent length@>=
+@<fix simple use of parshape@>=
{ last_special_line= info(par_shape_ptr)-1;
if (last_special_line!=0)
DBG(DBGTEX,"Warning parshape with n=%d not yet implemented",info(par_shape_ptr));
@@ -30784,8 +30745,8 @@
When generating a short format \HINT\ file, links are part of
the content section, where as labels and outlines are found in
the definition section. Because labels are defined while
-writing the content section, the writing of labels and outlines, which
-reference the labels, must be postponed. For that reason,
+writing the content section, the writing of labels and outlines
+ must be postponed. For that reason,
we store information about labels and outlines in dynamic arrays,
and map labels, which are identified by a name or a number,
to their index using a dynamic hash table.
@@ -30792,7 +30753,7 @@
We start with two functions that allocate new entries in the
dynamic arrays increasing their size if necessary.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static int next_label(void)
{ static int label_no=-1;
static int labels_allocated =0;
@@ -30823,13 +30784,13 @@
return outline_no;
}
@ While processing the content nodes, access to the labels is provided either
-by name or by number through a hash table. We store tabel entries in linked
+by name or by number through a hash table. We store table entries in linked
lists starting with a reasonably sized table of pointers. This keeps
the fixed costs low and guards against overflow and rapidly increasing
inefficiency. We start with a function to insert a new entry into
the hash table.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
typedef struct hash_entry
{int num; char *nom; uint16_t n; struct hash_entry *next;} HashEntry;
#define LABEL_HASH 1009 /* MIX a prime number */
@@ -30851,9 +30812,9 @@
}
@ There are two cases: finding a label by name or by number.
We start with the simpler case where the number is given.
-The process is straigth forward:
+The process is straight forward:
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static int find_label_by_number(int p)
{ unsigned int h=(unsigned int)p%LABEL_HASH;
HashEntry *e= label_hash[h];
@@ -30863,18 +30824,17 @@
return insert_hash(h,p,NULL);
}
@ To look up a label by its name as given by a token list,
-we prepare ourselfs by implementing two functions:
+we prepare ourselves by implementing two functions:
one to extract the character codes from the token list
forming the ``name''
and one to compute the hash value for a name.
The routine to find the label by name is then equivalent to the
routine we have just seen. Given a pointer |p| to
-either a label, a link, or an outline node, the function |find_label_by_name|
+either a label, a link, or an outline node, the function |find_label|
returns the correct label reference.
-Currently, we limit label names to at most 255 byte not counting the
-zero byte.
+Currently, we limit label names to at most 255 significant byte.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static char *tokens_to_name(pointer p)
{ static char s[256];
int i=0;
@@ -30910,7 +30870,7 @@
}
@ We combine both ways of finding a label reference in the following function:
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static int find_label(pointer p)
{@+ if (label_has_name(p)) return find_label_by_name(label_ptr(p));
else return find_label_by_number(label_ptr(p));
@@ -30925,7 +30885,7 @@
to allow the efficient adjustment of label positions when
moving lists.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static void new_label(pointer p)
{ int n=find_label(p);
if (n!=zero_label_no && labels[n].where!=LABEL_UNDEF)
@@ -30945,7 +30905,7 @@
@ When a link node is written to the output, we can check
that start links and end links properly match.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static int last_link=-1;
static int new_start_link(pointer p)
{ int n=find_label(p);
@@ -30972,7 +30932,7 @@
horizontal list, but at least it should be able to extract
the UTF8 character codes and display those.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static void new_outline(pointer p)
{ int r=find_label(p);
int m=next_outline();
@@ -31005,12 +30965,13 @@
pointer q=null; /* for output nodes */
pointer *t=NULL; /*the tail of the output nodes*/
bool eject=(type(p)==penalty_node && penalty(p)<=eject_penalty);
+ int page_penalty=0;
+ if (eject) page_penalty=penalty(p);
@<Record the bottom mark@>@;
@<Suppress empty pages if requested@>@;
link(contrib_head)= link(p);link(p)= null;
if (link(contrib_head)==null)
- { if(nest_ptr==0) tail= contrib_head;
- else contrib_tail= contrib_head;
+ { @<Make the contribution list empty by setting its tail to |contrib_head|@>;
}
update_last_values(p);
@<Freeze the page specs if called for@>@;
@@ -31020,8 +30981,9 @@
{ hpos0=hpos; hout_node(p); }
recycle_p:
flush_node_list(p);
- if (q!=null||eject)
+ if (q!=null||(eject&&page_contents>=box_there))
{
+ geq_word_define(int_base+output_penalty_code, page_penalty);
empty_output:
@<Fire up the output routine for |q|@>@;
}
@@ -31029,10 +30991,11 @@
DBG(DBGBUFFER,"after build page dyn_used= %d\n", dyn_used);
}
@ When the |page_contents| changes from |empty| to not |empty|,
-the function |hint_open| will open the output file. This place
-is choosen to match as close as possible the behaviour of the
-original \TeX. The output file is needed only much later in the
-function |hput_hint|.
+the function |hint_open| will open the output file.
+While the output file is needed only much later in the
+function |hput_hint|, this place was chosen to match,
+as close as possible, the behavior of the
+original \TeX.
@<Freeze the page specs if called for@>=
if (page_contents<box_there)
@@ -31046,7 +31009,7 @@
break; /* else fall through */
case hlist_node: case vlist_node: case rule_node:
if (page_contents==empty)
- { hint_open(); hout_init();
+ { hint_open();
freeze_page_specs(box_there);
hfix_defaults();
}
@@ -31054,7 +31017,7 @@
break;
case ins_node:
if (page_contents==empty)
- { hint_open(); hout_init();
+ { hint_open();
freeze_page_specs(inserts_only);
hfix_defaults();
}
@@ -31071,7 +31034,7 @@
This makes sense for a printed book but not for a screen reader where
there are no page numbers nor right or left hand pages.
Using a screen reader, empty pages are just annoying.
-The common way to achive an empty page is the use of {\tt \BS eject}
+The common way to achieve an empty page is the use of {\tt \BS eject}
followed by a an empty box, a fill glue, and another {\tt \BS eject}.
The following code tries to detect such a sequence of nodes and will eliminate
@@ -31099,7 +31062,8 @@
{ q=p;
link(prev_r)=null;
link(contrib_head)=r;
- DBG(DBGPAGE,"Eliminating empty page preceeding penalty %d\n",penalty(r));
+ DBG(DBGPAGE,"Eliminating empty page preceding penalty %d\n",penalty(r));
+ geq_word_define(int_base+output_penalty_code, penalty(r));
goto empty_output;
}
prev_r=r;
@@ -31114,7 +31078,7 @@
side effects. Possibly it would be better to regard them as invisible,
but still pass them on to the rest of the output routine.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static bool is_visible(pointer p)
{ switch (type(p))
{ case penalty_node:
@@ -31200,7 +31164,7 @@
hsize=0; vsize=0;
dimen_par_hfactor(hsize_code)= unity;
dimen_par_vfactor(vsize_code)= unity;
-@ There is one point where we can not simpy forego the
+@ There is one point where we can not simply forgo the
output routine: \.{\\write} commands. Unless the \.{\\write} is
decorated with an \.{\\immediate}, the whatsit node generated from it
will lay dormant in the contribution list (and later the page) until
@@ -31207,7 +31171,7 @@
the output routine passes it as part of the finished page to the |ship_out|
routine. There it will come to life and write its token list out.
The whatsit nodes from \.{\\openout} and \.{\\closeout} commands
-behave similary.
+behave similarly.
It is not possible to ignore the output routine
because the output routine may change the environment in which the
@@ -31220,7 +31184,7 @@
removes the output nodes and appends them to |q|, with |q|
always pointing to the tail pointer.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static pointer *collect_output(pointer *p, pointer *q)
{ while (*p!=null)
{ @<Collect output nodes from |*p|@>@;
@@ -31228,7 +31192,7 @@
}
return q;
}
-@ \TeX\ does not permit output nodes in leaders, so we dont check them;
+@ \TeX\ does not permit output nodes in leaders, so we don't check them;
further we do not check the pre- and post-break lists of
discretionary breaks.
@@ -31271,11 +31235,13 @@
default: break;
}
}
+
+
@ @<Fire up the output routine for |q|@>=
{ pointer r=new_null_box();type(r)=vlist_node;
subtype(r)=0;shift_amount(r)=0;height(r)=hvsize;
if (t==NULL) list_ptr(r)=null;
- else { list_ptr(r)=q; *t=new_glue(ss_glue); }
+ else { list_ptr(r)=q; *t=new_glue(fill_glue); }
flush_node_list(box(255)); /* just in case \dots */
box(255)=r;
if (output_routine!=null)
@@ -31286,6 +31252,7 @@
if (first_mark!=null) delete_token_ref(first_mark);
first_mark=bot_mark;add_token_ref(first_mark);
}
+ DBG(DBGPAGE,"Starting the output routine (output penalty=%d)\n",output_penalty);
push_nest();mode=-vmode;prev_depth=ignore_depth;mode_line=-line;
begin_token_list(output_routine, output_text);
new_save_level(output_group);normal_paragraph();
@@ -31297,6 +31264,7 @@
ship_out(box(255)); box(255)=null;
}
}
+
@ The |ship_out| routine just calls |execute_output|.
Because the output routine might have added plenty
of decorations around the list of output nodes,
@@ -31727,15 +31695,15 @@
The predefined stream for the main content has stream number 0.
@<Hi\TeX\ variables@>=
-int insert2stream[0x100]={0};
+static int insert2stream[0x100]={0};
@ The following function returns the stream number for a given insert number $i$
with $255>|i|\ge 0$. A new stream number is allocated if necessary.
Note that no overflow test is necessary since \TeX\ allocates less
than 233 inserts.
The initial value of |max_ref[stream_kind]| is 0 and therefore
-stream number 0, reserved for the main content
+stream number 0, reserved for the main content,
is never allocated. Stream definitions might also be loaded
-as part of a format file. Then the maximum stream number ist stored in |max_stream|.
+as part of a format file. Then the maximum stream number is stored in |max_stream|.
So if we do not find a stream number
in the |insert2stream| array, we scan the stream definitions
once and cache the associations found there.
@@ -31767,32 +31735,33 @@
\def\item{\par\noindent\hbox to 0pt{$\bullet$\hss}\quad\ignorespaces }
\item |setstream_number(p)| the \HINT\ stream number $n$.
\item |setstream_insertion(p)| the corresponding \TeX\ insertion number $i$.
-\item |setstream_max(p)| the maximum height $x$: This extended dimension is the maximum size
+\item |setstream_max(p)| the maximum height $x$:
+ This extended dimension is the maximum size
per page for this insertion.
\item |setstream_mag(p)| the magnification factor $f$:
Inserting a box of height $h$ will contribute $h*f/1000$
to the main page.
-\item |setstream_prefered(p)| the prefered stream $p$:
+\item |setstream_preferred(p)| the preferred stream $p$:
If $p\ge0$ we move the insert to stream $p$ if possible.
\item |setstream_next(p)| the next stream $n$:
If $n\ge0$ we move the insert to stream $n$ if it can not be
- accomodated otherwise.
+ accommodated otherwise.
\item |setstream_ratio(p)| the split ratio $r$:
- If $r>0$ split the final contribution of this streams between
- stream $p$ and $n$ in the ratio $r/1000$ for $p$ and $1-r/1000$ for $n$
- before contributing streams $p$ and $r$ to the page.
+ If $r>0$ split the final contribution of this stream between
+ stream $p$ and $n$ in the ratio $r/1000$ for $p$ and $1-r/1000$ for $n$
+ before contributing streams $p$ and $n$ to the page.
\item |setstream_before(p)| the ``before'' list $b$:
For a nonempty stream the material that is added before the stream content.
\item |setstream_after(p)| the ``after'' list $a$:
- For a nonempty stream, material that is added after the stream conten.
+ For a nonempty stream the material that is added after the stream content.
\item |setstream_topskip(p)| the top skip glue $t$: This glue is inserted between
- the $b$ list and the stream content and ajusted for the height for the first box
- of the stream content.
+ the $b$ list and the stream content and adjusted for the height for
+ the first box of the stream content.
\item |setstream_width(p)| the width $w$:
This extended dimension is the width used for example
to break paragraphs in the stream content into lines.
\item |setstream_height(p)| a glue specification $h$ reflecting the total height,
- strechability and shrinkability of the material in lists $a$ and $b$.
+ stretchability and shrinkability of the material in lists $a$ and $b$.
Currently Hi\TeX\ handles only normal streams. First or last streams will come later.
@@ -31805,7 +31774,7 @@
setstream_insertion(p)=n;
setstream_number(p)=hget_stream_no(n);
setstream_mag(p)=1000;
- setstream_prefered(p)=255;
+ setstream_preferred(p)=255;
setstream_next(p)=255;
setstream_ratio(p)=0;
setstream_max(p)=new_xdimen(0,0,ONE);
@@ -31817,7 +31786,7 @@
return p;
}
-@ The prefered stream, the next stream, and the split ratio are scanned as part of the
+@ The preferred stream, the next stream, and the split ratio are scanned as part of the
{\tt \BS setstream} primitive.
When \TeX\ finds the right brace that terminates the stream definition,
it calls |handle_right_brace|. Then it is time to obtain the remaining parts of the
@@ -31852,7 +31821,7 @@
pop_nest();
}
@ The before list $b$ and the after list $a$ are defined using the
-{\tt \BS before} and {\tt \BS after} primitives. When the corresponding list
+{\tt \BS HINTbefore} and {\tt \BS HINTafter} primitives. When the corresponding list
has ended with a right brace, \TeX\ calls |handle_right_brace| and we can store
the lists.
@@ -31890,7 +31859,7 @@
to reference this page template.
\item |setpage_priority(p)|: The priority helps in selecting a page template.
\item |setpage_topskip(p)|: The topskip glue is added at the top of a page and
- adjusted by the height of the first bos on the page.
+ adjusted by the height of the first box on the page.
\item |setpage_height(p)|: The height of the full page including the margins.
\item |setpage_width(p)|: The width of the full page including the margins.
\item |setpage_depth(p)|: The maximum depth of the page content. If the last box is deeper
@@ -31961,7 +31930,7 @@
unsave();
pop_nest();
}
-@ @<Hi\TeX\ auxiliar routines@>=
+@ @<Hi\TeX\ auxiliary routines@>=
static pointer hget_current_page(void)
{ pointer p=link(setpage_head);
@@ -31983,7 +31952,7 @@
@* \HINT\ Output.
Here are the routines to initialize and terminate the output.
The initialization is done in three steps:
-First we allocate the data structurs to write nodes into buffers;
+First we allocate the data structures to write nodes into buffers;
this requires a directory and buffers for sections 0, 1, and 2.
@ @<Hi\TeX\ routines@>=
@@ -32004,15 +31973,9 @@
part of the termination routines. It is placed, however, much earlier
because asking for the output file name---according to \TeX's
conventions---should come before the first item is put on the first
-page by the page builder. For this reason, |hint_open| is called
-when calling |hout_init|.
+page by the page builder.
@<Hi\TeX\ routines@>=
-static void hout_init(void)
-{
-
-}
-
static void hint_open(void)
{ if (job_name==0) open_log_file();
pack_job_name(".hnt");
@@ -32019,7 +31982,6 @@
while (!(hout=open_out((char *)name_of_file+1,"wb")))
prompt_file_name("file name for output",".hnt");
output_file_name=make_name_string();
- hlog=stderr;
DBG(DBGBASIC,"Output file %s opened\n",(char *)name_of_file+1);
}
@@ -32051,7 +32013,7 @@
record the names of the files that are added as optional sections.
This feature is not part of the |hput_optional_sections| function
which is called from |hput_hint|. The following simple
-loop will achive this.
+loop will achieve this.
@<record the names of files in optional sections@>=
{ int i;
@@ -32062,7 +32024,7 @@
@* The \HINT\ Directory.
There is not much to do here: some code to find a new or existing directory entry,
a variable to hold the number of directory entries allocated,
-a function to allocate a new file section, and an auxiliar function to
+a function to allocate a new file section, and an auxiliary function to
convert \TeX's file names to ordinary \CEE/ strings.
@<Find an existing directory entry@>=
@@ -32093,7 +32055,7 @@
@ @<Hi\TeX\ variables@>=
static int dir_entries=4;
-@ @<Hi\TeX\ auxiliar routines@>=
+@ @<Hi\TeX\ auxiliary routines@>=
static uint16_t hnew_file_section(char *file_name)
{ uint16_t i;
@<Find an existing directory entry@>@;
@@ -32103,7 +32065,7 @@
}
@ The following function uses \TeX's function |pack_file_name|
-to create a new filename from a name |n|, a direcory or ``area'' |a|,
+to create a new filename from a name |n|, a directory or ``area'' |a|,
and an extension |e|. \TeX\ will truncate the new filename
to |file_name_size| characters without warning. The new function
will take a |name_length| equal to | file_name_size| as an
@@ -32111,9 +32073,9 @@
program. The return value converts a {\mc Pascal} array, starting with index 1,
into a \CEE/ array starting with index 0.
-@<Hi\TeX\ auxiliar routines@>=
-static char *hfile_name(str_number n, char *a, char *e)
-{ pack_file_name(n,a,e);
+@<Hi\TeX\ auxiliary routines@>=
+static char *hfile_name(str_number n, str_number a, str_number e)
+{ pack_file_name(n,a,e,NULL);
if (name_length>=file_name_size)
QUIT("File name too long %d >= %d",name_length,file_name_size);
return (char *)name_of_file+1;
@@ -32139,9 +32101,8 @@
The procedure that generates these definitions is called |hfix_defaults|:
-@<Hi\TeX\ auxiliar routines@>=
-
-void hfix_defaults(void)
+@<Hi\TeX\ auxiliary routines@>=
+static void hfix_defaults(void)
{ @+int i;
DBG(DBGDEF,"Freezing HINT file defaults\n");
@<Compute the page size@>@;
@@ -32173,7 +32134,7 @@
Here we initialize the necessary data structures for definitions.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static void hdef_init(void)
{@+ int i;
@<Switch |hsize| and |vsize| to extended dimensions@>@;
@@ -32188,10 +32149,10 @@
@ After all definitions are ready, we write them using the function
|hput_definitions|. When we output the definitions,
we have to make sure to define references before we use them.
-This is achived by using a specific ordering of the
+This is achieved by using a specific ordering of the
definitions in the function |hput_definitions| and by preventing
the allocation of new definitions as soon as the output of the definition
-section has started. The latter has the aditional benefit that the
+section has started. The latter has the additional benefit that the
maximum values do no longer change.
@<Hi\TeX\ routines@>=
@@ -32235,7 +32196,7 @@
parameter numbers to \HINT\ definition numbers. While it seems more convenient here
to have the reverse mapping, we need the mapping only once to record parameter definitions,
but we will need it repeatedly in the function |hdef_param_node| and the overhead here does
-not warant having the mapping in both directions.
+not warrant having the mapping in both directions.
@<Hi\TeX\ variables@>=
static const int hmap_int[] ={@/
@@ -32295,7 +32256,7 @@
@ The function |hget_int_no| tries to allocate a predefined integer number;
if not successful, it returns $-1$.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static int hget_int_no(int32_t n)
{ int i;
int m =max_ref[int_kind];
@@ -32307,7 +32268,7 @@
return -1;
}
@ Before we give the code to output an integer definition, we declare a macro that
-is usefull for all the definitions. |HPUTDEF| takes a function |F| and a reference number |R|.
+is useful for all the definitions. |HPUTDEF| takes a function |F| and a reference number |R|.
It is assumed that |F| writes a definition into the output and returns a tag. The macro
will then add the reference number and both tags to the output.
@<Hi\TeX\ macros@>=
@@ -32368,7 +32329,7 @@
dimen_defined[quad_no]=quad(cur_font);
dimen_defined[math_quad_no]=math_quad(text_size);
max_ref[dimen_kind]=MAX_DIMEN_DEFAULT;
-@ @<Hi\TeX\ auxiliar routines@>=
+@ @<Hi\TeX\ auxiliary routines@>=
static int hget_dimen_no(scaled s)
/* tries to allocate a predefined dimension number in the range 0 to 0xFF
if not successful return -1 */
@@ -32405,8 +32366,8 @@
array and if no match was found, we allocate a new entry,
reallocating the array if needed.
-@<Hi\TeX\ auxiliar routines@>=
-int hget_xdimen_no(pointer p)
+@<Hi\TeX\ auxiliary routines@>=
+static int hget_xdimen_no(pointer p)
{ int i;
for (i=0;i<=max_ref[xdimen_kind];i++)
{ if (xdimen_defined[i].w== xdimen_width(p) &&
@@ -32487,20 +32448,20 @@
if ( hmap_glue[i]>=0)
{ glue_defined[hmap_glue[i]]=glue_par(i); incr(glue_ref_count(glue_par(i)));}
max_ref[glue_kind]=MAX_GLUE_DEFAULT;
-@ Next we define some auxiliar routines to compare glues for equality and to convert glues beween the different representations.
+@ Next we define some auxiliary routines to compare glues for equality and to convert glues between the different representations.
-@<Hi\TeX\ auxiliar routines@>=
-int glue_spec_equal(pointer p, pointer q)
+@<Hi\TeX\ auxiliary routines@>=
+static int glue_spec_equal(pointer p, pointer q)
{ return (width(q)==width(p) && stretch(q)==stretch(p) && shrink(q)==shrink(p) &&
(stretch_order(q)==stretch_order(p) || stretch(q)==0) &&
(shrink_order(q)==shrink_order(p)|| shrink(q)==0));
}
-int glue_equal(pointer p, pointer q)
+static int glue_equal(pointer p, pointer q)
{ return p==q || glue_spec_equal(p,q);
}
-int Glue_equal(Glue *p, Glue *q)
+static int Glue_equal(Glue *p, Glue *q)
{ return(p->w.w==q->w.w && p->w.h==q->w.h && p->w.v==q->w.v &&
p->p.f== q->p.f && p->m.f==q->m.f &&
(p->p.o==q->p.o || p->p.f==0.0) &&
@@ -32512,7 +32473,7 @@
we also compare values. An alternative approach to speed up searching
is used for parameter lists as described below.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static int hget_glue_no(pointer p)
{ static int rover=0;
int i;
@@ -32564,8 +32525,8 @@
% Care should be taken for the statically allocated glue specs
% deallocation of glue specs is relatively simple
% extended glue and kern values can be restricted to non math mode
-@<Hi\TeX\ auxiliar routines@>=
-void to_Glue(pointer p, Glue *g)
+@<Hi\TeX\ auxiliary routines@>=
+static void to_Glue(pointer p, Glue *g)
{ g->w.w=width(p);
g->w.h=g->w.v=0.0;
g->p.f=stretch(p)/(double)ONE; g->p.o= stretch_order(p);
@@ -32573,8 +32534,8 @@
}
@*1 Baseline Skips.
TeX's baseline nodes just store a baseline skip reference number.
-We have seen this sitation before when dealing with extended dimensions
-and the solution here is the same: a dynamicaly allocated array.
+We have seen this situation before when dealing with extended dimensions
+and the solution here is the same: a dynamically allocated array.
@<Hi\TeX\ variables@>=
typedef struct {
pointer ls, bs; /* line skip and baselineskip gluespecs */
@@ -32595,8 +32556,8 @@
bl_defined[zero_baseline_no].lsl=0;
bl_used= MAX_BASELINE_DEFAULT+1;
max_ref[baseline_kind]= MAX_BASELINE_DEFAULT;
-@ @<Hi\TeX\ auxiliar routines@>=
-int hget_baseline_no(pointer bs, pointer ls, scaled lsl)
+@ @<Hi\TeX\ auxiliary routines@>=
+static int hget_baseline_no(pointer bs, pointer ls, scaled lsl)
{
static int rover=0;
int i;
@@ -32626,10 +32587,10 @@
}
@ The following routine does not allocate a new glue definition, because the
-baselinedefinitions are output after the glue definitions. This is not perfect.
-@<Hi\TeX\ auxiliar routines@>=
+baseline definitions are output after the glue definitions. This is not perfect.
+@<Hi\TeX\ auxiliary routines@>=
static uint8_t hout_glue_spec(pointer p);
-uint8_t hout_baselinespec(int n)
+static uint8_t hout_baselinespec(int n)
{ Info i=b000;
pointer p;
scaled s;
@@ -32688,7 +32649,7 @@
@<Hi\TeX\ variables@>=
static pointer dc_defined[0x100];
@ There are no predefined discretionary breaks and so we start with
-three auxiliar functions and the
+three auxiliary functions and the
function to get a ``disc'' number.
The first two routines are used to compare discretionary breaks
@@ -32700,7 +32661,7 @@
The following routine will not make a complete comparison but will give
up if the box content is ``too complex''.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static bool list_equal(pointer p, pointer q);
static bool node_equal(pointer p, pointer q)
{ if (is_char_node(p) && is_char_node(q) &&
@@ -32794,13 +32755,13 @@
#define PLH_SIZE 313 /* a prime number $\approx2^8\times 1.2$. */
-struct {int l; /* link */
+static struct {int l; /* link */
uint32_t h; /* hash */
uint32_t n; /* number */
uint32_t s; /* size */
uint8_t *p; /* pointer */} pl_defined[PLH_SIZE]={{0}};
-int pl_head=-1, *pl_tail=&pl_head;
-@ Next we define three short auxiliar routines and the |hget_param_list_no| function.
+static int pl_head=-1, *pl_tail=&pl_head;
+@ Next we define three short auxiliary routines and the |hget_param_list_no| function.
@<Hi\TeX\ routines@>=
@@ -32914,9 +32875,9 @@
number to a \HINT\ font number using the function |hget_font_no|, and while
doing so discover that the corresponding \HINT\ font number does not yet exist.
Because the |Font| structure must be initialized after allocating it,
-we start with some auxiliar routines for that purpose.
+we start with some auxiliary routines for that purpose.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static pointer find_space_glue(internal_font_number f)
{@+font_index @!main_k;
pointer main_p=font_glue[f];
@@ -32969,7 +32930,7 @@
%|kpse_find_file(name,kpse_fontmap_format,false);|
%with name "ps2pk.map" or "psfonts.map" or "ttfonts.map" or cmfonts.map
%to get from the tfm name the postscript name.
-%then I get the |psfont_name| ususaly its the same font name
+%then I get the |psfont_name| usually its the same font name
%with .pfb appended
%so I can skip it
%#ifdef HAVE_KPSE_ENC_FORMATS
@@ -32979,7 +32940,7 @@
%#endif
%reading the encoding is found in dvipng/dvipnd-src/enc.c
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static char *hfind_glyphs(char *filename)
{ char *fname=NULL;
kpse_glyph_file_type file_ret;
@@ -33013,13 +32974,13 @@
hmap_font[f]=g;
hfonts[g]->g=hget_font_space(f);
hfonts[g]->h=hget_font_hyphen(f);
- pack_file_name(font_name[f], "",".tfm");
+ pack_file_name(font_name[f], empty_string,empty_string,".tfm");
n = kpse_find_tfm((char*)name_of_file+1);
if (n==NULL)
QUIT("Unable to find .tfm file for font %s",(char*)name_of_file+1);
hfonts[g]->m= hnew_file_section(n);
free(n);
- pack_file_name(font_name[f], "","");
+ pack_file_name(font_name[f], empty_string,empty_string,"");
n= hfind_glyphs((char*)name_of_file+1);
if (n==NULL)
QUIT("Unable to find glyph file for font %s",(char*)name_of_file+1);
@@ -33032,9 +32993,9 @@
inside the pre- or post-break list of a predefined discretionary break.
To make sure that the fonts necessary for these characters are included
in the final \HINT\ file, we check these lists to make sure all \TeX\ font
-numbers have a corresponting \HINT\ font number.
+numbers have a corresponding \HINT\ font number.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static void ensure_font_no(pointer p)
{ while (p!=null)
{ if (is_char_node(p))
@@ -33069,9 +33030,8 @@
}
}
@ We used the following function to write a \TeX\ string to the \HINT\ file:
-@<Hi\TeX\ auxiliar routines@>=
-
-void hout_string(int s)
+@<Hi\TeX\ auxiliary routines@>=
+static void hout_string(int s)
{ pool_pointer j;
uint8_t c;
j= str_start[s];
@@ -33128,7 +33088,7 @@
Once we start producing content nodes, we update the maximum numbers
of page templates and streams from |max_page| and |max_stream|.
-These values might have changed because the templates were stored in a
+These values might have changed because templates were loaded from a
format file.
@<Fix definitions of page templates@>=
@@ -33176,7 +33136,7 @@
HPUT8(n);
hout_xdimen_node(setstream_max(p)); /* maximum height */
HPUT16(setstream_mag(p)); /* factor */
- HPUT8(setstream_prefered(p)); /* prefered */
+ HPUT8(setstream_preferred(p)); /* preferred */
HPUT8(setstream_next(p)); /* next */
HPUT16(setstream_ratio(p)); /* ratio */
q=setstream_before(p);setstream_before(p)=null;
@@ -33266,15 +33226,15 @@
}
break;
@*1 Extended Dimensions.
-Extended dimensions do not consitute content on their own, but nodes
+Extended dimensions do not constitute content on their own, but nodes
containing an extended dimension are part of other nodes. Here we
-define an auxiliar function that checks for a predefined extended
+define an auxiliary function that checks for a predefined extended
dimension and if found outputs the reference number and returns false;
otherwise it outputs the extended dimension and returns true.
-@<Hi\TeX\ auxiliar routines@>=
-void hout_xdimen_node(pointer p)
+@<Hi\TeX\ auxiliary routines@>=
+static void hout_xdimen_node(pointer p)
{ Xdimen x;
x.w=xdimen_width(p);
x.h=xdimen_hfactor(p)/(double)ONE;
@@ -33281,7 +33241,8 @@
x.v=xdimen_vfactor(p)/(double)ONE;
hput_xdimen_node(&x);
}
-bool hout_xdimen(pointer p)
+
+static bool hout_xdimen(pointer p)
{int n = hget_xdimen_no(p);
if (n>=0) { HPUT8(n); return false; @+}
else
@@ -33289,8 +33250,7 @@
}
@*1 Languages.
-We map the language numbers of \TeX\ to \HINT\ language numbers
-using the |hlanguage| array.
+The |hlanguage| array maps the language numbers of \TeX\ to \HINT\ language numbers.
@<Hi\TeX\ variables@>=
static struct {
uint8_t n;
@@ -33298,12 +33258,12 @@
} hlanguage[0x100];
@ For any language number of \TeX, the following function returns
the corresponding \HINT\ language number.
-Since \TeX\ knowns about a maximum of 255 languages, there is
+Since \TeX\ knows about a maximum of 255 languages, there is
no need for overflow checking. The next function writes a language
node to the output stream.
-@<Hi\TeX\ auxiliar routines@>=
-uint8_t hget_language_no(uint8_t n)
+@<Hi\TeX\ auxiliary routines@>=
+static uint8_t hget_language_no(uint8_t n)
{ int i;
for (i=0;i<=max_ref[language_kind]; i++)
if (hlanguage[i].n==n) return i;
@@ -33313,7 +33273,7 @@
return i;
}
-uint8_t hout_language(uint8_t n)
+static uint8_t hout_language(uint8_t n)
{ n=hget_language_no(n);
if (n<7) return TAG(language_kind,n+1);
else
@@ -33326,7 +33286,8 @@
case language_node:
tag=hout_language(what_lang(p));
break;
-@ Normaly \TeX\ does not produce an initial language node and
+
+@ Normally \TeX\ does not produce an initial language node and
then the language in the \HINT\ file would not be known until
it changes for the first time.
@@ -33388,11 +33349,11 @@
break;
@*1 Glue and Leaders.
Because glue specifications and glue nodes are sometimes part of other
-nodes, we start with three auxiliar functions: The first simply
+nodes, we start with three auxiliary functions: The first simply
converts a Hi\TeX\ glue node into a \HINT\ |Glue|, outputs it and
returns the tag; the second checks for predefined glues, and the third
outputs a complete glue node including tags.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static uint8_t hout_glue_spec(pointer p)
{ @+Glue g;
@@ -33442,9 +33403,8 @@
Therefore we define a function that converts \TeX's |disc_node| pointers
to \HINT's |Disc|, outputs the discretionary break, and returns the tag.
-@<Hi\TeX\ auxiliar routines@>=
-
-uint8_t hout_disc(pointer p)
+@<Hi\TeX\ auxiliary routines@>=
+static uint8_t hout_disc(pointer p)
{ Disc h;
h.x=!is_auto_disc(p);
h.r=replace_count(p);
@@ -33481,7 +33441,7 @@
@*1 Ligatures.
The subtype giving information on left and right boundary characters
is ignored since the \HINT\ viewer will not do ligature or kerning
-programms and neither attempt hyphenation.
+programs and neither attempt hyphenation.
@<cases to output content nodes@>=
case ligature_node:
@@ -33542,8 +33502,8 @@
used. This is probably also for \TeX\ the most common situation.
It is, however, possible to supply more than one page template with different defaults
and while not very common, \TeX\ might change the parameters at any time.
-Because we dont know which is the current page template,
-it is not posible to compare the current parameter values against the defaults,
+Because we don't know which is the current page template,
+it is not possible to compare the current parameter values against the defaults,
and we have to supply all the parameters always.
In a future version, we might have a \TeX\ primitive that allows us to
signal ``use the defaults''.
@@ -33591,7 +33551,7 @@
@ For \TeX's whatsit nodes that handle output files, no code is generated;
hence, we call |out_what| and simply remove the tag byte that is already
in the output.
-When the \.{\\write} node arrives here, it is normaly handled
+When the \.{\\write} node arrives here, it is normally handled
in |hlist_out| or |vlist_out| in an environment determined by
the output routine. For example \LaTeX\ redefines \.{\\protect}
as \.{\\noexpand} and these redefinitions need to be made
@@ -33610,7 +33570,7 @@
The parameter list is given by a reference number but the extended dimension
needs an |xdimen| node. In this case the reference number for the parameter
list comes first, while otherwise the extended dimension would come first.
-To determine whether ther is a reference number for the parameter list,
+To determine whether there is a reference number for the parameter list,
the function |hout_param_list| is writing the parameter list to the output.
\noindent
@<cases to output whatsit content nodes@>=
@@ -33679,8 +33639,8 @@
When we output an extended box, we have to consider a special case: the page templates.
Page templates are boxes that contain insertion points. These insertion points look
like regular insertions but with an empty content list. As a result the |hpack| and
-|vpackage| routines might belive that they can compute the dimensions of the box
-conntent when in fact they can not.
+|vpackage| routines might believe that they can compute the dimensions of the box
+content when in fact they can not.
@<cases to output whatsit content nodes@>=
@@ -33728,8 +33688,8 @@
}
break;
@ In the preamble, we remove the unset nodes and retain only the list of tabskip glues.
-@<Hi\TeX\ auxiliar routines@>=
-void hout_preamble(pointer p)
+@<Hi\TeX\ auxiliary routines@>=
+static void hout_preamble(pointer p)
{ pointer q,r;
DBG(DBGBASIC,"Writing Preamble\n");
q=p;
@@ -33751,7 +33711,7 @@
When the viewer reads an item node, it will package the extended boxes to their natural size.
This is the size that is needed to compute the maximum width of a column.
-@<Hi\TeX\ auxiliar routines@>=
+@<Hi\TeX\ auxiliary routines@>=
static void hout_item(pointer p, uint8_t t, uint8_t s)
{ Info i=b000;
@@ -33794,7 +33754,7 @@
DBG(DBGBASIC,"End Item List\n");
}
-void hout_align_list(pointer p, bool v)
+static void hout_align_list(pointer p, bool v)
{ List l;
uint32_t pos;
DBG(DBGBASIC,"Writing Align List\n");
@@ -33879,7 +33839,7 @@
The parameter |p| is a pointer to a param node list.
The function either finds a reference number to a predefined parameter list
and returns the reference number,
- or it outputs the node list at position pos (thats where the tag goes),
+ or it outputs the node list at position pos (that's where the tag goes),
sets |l->k|, |l->p| and |l->s|, and returns $-1$.
@<Hi\TeX\ routines@>=
@@ -34087,6 +34047,9 @@
\item Tables where the width of a column depends on \.{\\hsize} or
\.{\\vsize} are not tested and probably not yet supported.
+\item \.{\\vtop} and \.{\\vcenter} will not work if any dimension of the
+ vertical list depends on \.{\\hsize} or \.{\\vsize}.
+
\item The encoding of horizontal lists as texts is not yet supported,
but it would make the \HINT\ file shorter and much better to read
when stretched into long \HINT\ format.
@@ -34122,16 +34085,50 @@
that the code is presented in literate programming style.
The code that follows is organized in three parts.
-Some code for k\TeX\ must come before
+Some code for \TeX\ Live must come before
the definition of \TeX's macros because
it uses include files containing identifiers that are in conflict
-with \TeX's macros. The remaining two parts are first auxiliar functions and then
+with \TeX's macros. The remaining two parts are first auxiliary functions and then
those functions that are called from the ``classic'' \TeX\ code.
- at p @<k\TeX\ auxiliar functions@>@;
- @<k\TeX\ functions@>@;
+ at p @<\TeX\ Live auxiliary functions@>@;
+ @<\TeX\ Live functions@>@;
-@ Let's begin with the beginning: the command line.
+@
+Most of the code that we present next comes together in the
+function |main_init|
+which is the first function called in the main program of a \TeX\ engine
+belonging to \TeX\ Live. Before doing so, we make copies of
+argument count and argument vector putting them in global variables.
+@<Global...@>=
+static char **argv;
+static int argc;
+
+@ @<\TeX\ Live functions@>=
+static void main_init(int ac, char *av[])
+{@+
+ char* main_input_file;
+ argc = ac;
+ argv = av;
+ interaction = error_stop_mode;
+ kpse_record_input = recorder_record_input;
+ kpse_record_output = recorder_record_output;
+
+ @<parse options@>@;
+ @<set the program and engine name@>@;
+ @<activate configuration lines@>@;
+ @<set the input file name@>@;
+ @<set defaults from the {\tt texmf.cfg} file@>@;
+ @<set the format name@>@;
+ @<enable the generation of input files@>@;
+}
+
+@ @<Forward declarations@>=
+static void main_init(int ac, char *av[]);
+
+
+@*1 Command Line.
+ Let's begin with the beginning: the command line.
To see how a command line is structured, we first look at the
help text that is displayed if the user asks for it (or if \TeX\
decides that the user needs it). The help text is produced by the
@@ -34138,7 +34135,7 @@
function |usage_help|.
\def\SP{ }% disable visible spaces in strings.
-@<k\TeX\ auxiliar functions@>=
+@<\TeX\ Live auxiliary functions@>=
static void usage_help(void)
{@+@<explain the command line@>@;
@<explain the options@>@;
@@ -34146,7 +34143,7 @@
exit(0);
}
-@ The command line commes in three slightly different versions:
+@ The command line comes in three slightly different versions:
@<explain the command line@>=
fprintf(stdout,@/
@@ -34174,8 +34171,10 @@
" If no arguments or options are specified, prompt for input.\n"@/
"\n");
-@ Here is the list of possible options and their explanation:
+@*1 Options.
+Here is the list of possible options and their explanation:
+
@<explain the options@>=
fprintf(stdout,
"Options:\n"@/
@@ -34212,36 +34211,36 @@
@t\qquad@>"\t disable/enable parsing of the first line of\n"@/
@t\qquad@>"\t\t\t the input file\n"@/
" [-no]-file-line-error"@/
- @t\qquad@>"\t Disable/Enable file:line:error style\n"@/
+ @t\qquad@>"\t disable/enable file:line:error style\n"@/
" -cnf-line=STRING"@/
- @t\qquad@>"\t Process STRING like a line in texmf.cnf\n"@/
+ @t\qquad@>"\t process STRING like a line in texmf.cnf\n"@/
" -compress "@/
@t\qquad@>"\t enable compression of section 1 and 2\n"@/
" [-no]-empty-page "@/
- @t\qquad@>"\t Disable/Enable empty pages\n"@/
- " -hyphenate-first-word "@/
- @t\qquad@>"\t hyphenate the first word of a paragraph\n"@/
- " -no-hyphenate-first-word "@/
- @t\qquad@>"\t don't hyphenate the first word of a paragraph\n"@/
+ @t\qquad@>"\t disable/enable empty pages\n"@/
+ " [-no]-hyphenate-first-word "@/
+ @t\qquad@>"\t disable/enable hyphenation of\n"@/
+ @t\qquad@>"\t\t\t the first word of a paragraph\n"@/
" -resolution=NUMBER "@/
@t\qquad@>"\t set the resolution to NUMBER dpi\n"@/
" -mfmode=MODE "@/
@t\qquad@>"\t set the METAFONT mode to MODE\n"@/
#ifdef DEBUG
- " -hint-debug=FLAGS "@/
- @t\qquad@>"\t set flags to controll hint debug output\n"@/
+ " -hint-debug=FLAGS "@/
+ @t\qquad@>"\t set flags to control hint debug output\n"@/
" -hint-debug-help "@/
@t\qquad@>"\t give help on hint debugging\n"@/
#endif
);
-
@ The processing of command line options is controlled by the |long_options|
\def\SP{{\tt\char`\ }}% restore the visible space in a string
array. Each entry in this array contains first the name of the option,
-then a flag that tells whether the option takes an argument or not,
-then the (optional) address of a flag variable that gets set
-by this option, and finally the value to store in the flag variable.
+then a flag that tells whether the option takes an argument or not.
+If next the (optional) address of a flag variable is given,
+it is followed by the value to store in the flag variable.
+In this case, setting the flag variable is handled by
+the |getopt_long_only| function.
Besides the flag variables that occur in the table,
a few string variables may be set using the options.
@@ -34250,18 +34249,13 @@
string variables are initialized with |NULL|.
@<Global...@>=
-static int iniversion=0;
-static int etexp=0;
-static int ltxp=0;
-static int recorder_enabled=0;
-static int parsefirstlinep=-1;
-static int filelineerrorstylep=-1;
-static int interaction_option=-1;
-static const char *user_progname=NULL, *output_directory=NULL, *c_job_name=NULL;
-static char *dump_name=NULL;@#
-int option_no_empty_page=true, option_hyphen_first=true;
-int option_dpi=600;
-const char *option_mfmode="ljfour", *option_dpi_str="600";
+static int iniversion=false, etexp=false, ltxp=false, recorder_enabled=false;
+static int parsefirstlinep=-1, filelineerrorstylep=-1, interaction_option=-1;
+static const char *user_progname=NULL, *output_directory=NULL, *c_job_name=NULL,
+ *dump_name=NULL;@#
+static int option_no_empty_page=true, option_hyphen_first=true;
+static int option_dpi=600;
+static const char *option_mfmode="ljfour", *option_dpi_str="600";
extern int option_compress;
extern unsigned int debugflags;
@@ -34304,13 +34298,13 @@
function from the \CEE/ library. This function returns 0 and sets the
|option_index| parameter to the option found, or it returns $-1$ if
the end of all options is reached.
-@<k\TeX\ functions@>=
+@<\TeX\ Live functions@>=
static void parse_options (int argc, char *argv[])
{@+ while (true) {
int option_index;
int g = getopt_long_only (argc, argv, "+", long_options, &option_index);
if (g==0)
- { @<handle the option at |option_index|@>@;@+ }
+ { @<handle the options@>@;@+ }
else if (g == '?')
{ fprintf(stderr,"Try '%s --help' for more information\n",argv[0]);
exit(1);
@@ -34319,12 +34313,32 @@
}
}
+@ @<Forward declarations@>=
+static void parse_options (int argc, char *argv[]);
+
+@ Before we can call the |parse_options| function,
+we might need some special preparations for Windows.
+
+@<parse options@>=
+#if defined(WIN32)
+{@+ char* enc;
+ kpse_set_program_name (argv[0], NULL);
+ enc = kpse_var_value("command_line_encoding");
+ get_command_line_args_utf8(enc, &argc, &argv);@/
+ parse_options (argc, argv);
+ @<record {\tt texmf.cnf}@>@;
+}
+#else
+ parse_options (ac, av);
+#endif
+
+
@ To handle the options, we compare the name at the given |option_index| with
the different option names. This is not a very efficient method, but the
impact is low and it's simple to write.
Comparing the name of the argument with the |name| field in the |option|
-structure is done in the auxiliar function |argument_is|.
+structure is done in the auxiliary function |argument_is|.
Unfortunately the |name| field is in conflict with the |name| macro
defined by \TeX. To avoid the conflict, the |argument_is| function
goes just after the \.{kpathsea.h} header file that defines
@@ -34339,12 +34353,12 @@
@ Now we can handle the first two options:
-@<handle the option at |option_index|@>=
+@<handle the options@>=
if (ARGUMENT_IS("help")) usage_help();
else if (ARGUMENT_IS("version")){@+
- printf(banner "\n"
- "eTeX version " eTeX_version_string "\n"
- "Prote version " Prote_version_string "\n");
+ printf(banner@, "\n"@/
+ "eTeX version "@, eTeX_version_string@, "\n"@/
+ "Prote version "@, Prote_version_string@, "\n");
exit(0);@+
}
@@ -34354,7 +34368,7 @@
If defined, the |interaction_option| will be used to set \TeX's
|interaction| variable in the |initialize| and the |undump| functions.
-@<handle the option at |option_index|@>=
+@<handle the options@>=
else @+if (ARGUMENT_IS ("interaction"))@t\2@> {
if (STREQ (optarg, "batchmode")) interaction_option = batch_mode;
else if (STREQ (optarg, "nonstopmode")) interaction_option = nonstop_mode;
@@ -34366,7 +34380,7 @@
@ The next two options pass the string argument to the
\.{kpathsearch} library.
-@<handle the option at |option_index|@>=
+@<handle the options@>=
else @+if (ARGUMENT_IS ("mktex")) @t\2@> kpse_maketex_option (optarg, true);
else if (ARGUMENT_IS ("no-mktex")) kpse_maketex_option (optarg, false);
@@ -34375,17 +34389,16 @@
the following option can be used.
The argument value 3 is a good choice to start with.
-@<handle the option at |option_index|@>=
+@<handle the options@>=
else @+if (ARGUMENT_IS ("kpathsea-debug")) @t\2@>
kpathsea_debug |= atoi (optarg);
-@ The remaining options either set a flag and are handled by
-the |getopt_long_only| function or they take a string argument
+@ The next set of options take a string argument
and assign it to the corresponding string variable.
-@<handle the option at |option_index|@>=
+@<handle the options@>=
else @+if (ARGUMENT_IS ("progname"))@t\2@>
user_progname = normalize_quotes(optarg,"program name");
else if (ARGUMENT_IS ("fmt"))
@@ -34394,23 +34407,85 @@
output_directory = normalize_quotes(optarg,"output directory");
else if (ARGUMENT_IS ("jobname"))
c_job_name = normalize_quotes (optarg, "job name");
-else @+if (ARGUMENT_IS ("cnf-line"))
- add_cnf_line(optarg);@t\2@>
-else if (ARGUMENT_IS("resolution")) @t\2@>
-{ option_dpi_str=optarg;
- option_dpi=strtol(option_dpi_str,NULL,10);
+
+@ When string arguments specify files or directories,
+special care is needed if arguments are quoted and/or contain spaces.
+The function |normalize_quotes| makes sure that arguments containing
+spaces get quotes around them and it checks for unbalanced quotes.
+
+
+@<\TeX\ Live auxiliary functions@>=
+static char *normalize_quotes (const char *nom, const char *mesg)
+{@+ int quoted = false;
+ int must_quote = (strchr(nom, ' ') != NULL);
+ char *ret = xmalloc(strlen(nom)+3);/*room for two quotes and NUL*/
+ char *p=ret;
+ const char *q;
+
+ if (must_quote)
+ *p++ = '"';
+ for (q = nom; *q; q++)
+ if (*q == '"')
+ quoted = !quoted;@+
+ else
+ *p++ = *q;
+
+ if (must_quote)
+ *p++ = '"';
+ *p = '\0';
+ if (quoted) {
+ fprintf(stderr, "! Unbalanced quotes in %s %s\n", mesg, nom);
+ exit(1);
+ }
+ return ret;
}
-else if (ARGUMENT_IS("mfmode"))
- option_mfmode=optarg;
-#ifdef DEBUG at t\1@>
-else @+if (ARGUMENT_IS("hint-debug"))
- debugflags=strtol(optarg,NULL,16);
-else @+if (ARGUMENT_IS("hint-debug-help"))
- hint_debug_help();
-#endif
+@*1 Passing a file name as a general text argument.
-@ The recorder option can be used to enable the file name recorder.
+|scan_file_name| uses the following code to parse a file name given
+as a general text argument. Such an argument can be any token list
+starting with a left brace and ending with a right brace.
+This token list is then expanded (without the leading and trailing braces)
+and printed into the string pool without making it yet an official string.
+After removing all double quotes, because this is current practice
+for \TeX\ engines that are part of \TeX\ Live,
+and setting the area and extension delimiters,
+all temporary garbage used so far is freed.
+
+Due to the expansion of the token list, this code and hence the
+|scan_file_name| procedure is recursive.
+One can provide the name of a file as the content of an other file.
+
+
+% Thierry Laronde
+@<Define a general text file name and |goto done|@>=
+{@+back_input();name_in_progress=false; /*this version is recursive...*/
+ cur_cs=input_loc; /*|scan_toks| will set |warning_index| from it*/
+ scan_general_x_text();
+ old_setting=selector;selector=new_string;
+ token_show(link(garbage));selector=old_setting;
+ @<Suppress double quotes in braced input file name@>@;
+ j=pool_ptr-1;while ((j >= str_start[str_ptr])&&(area_delimiter==0))
+ {@+
+ if ((str_pool[j]=='/')) area_delimiter=j-str_start[str_ptr];
+ if ((ext_delimiter==0)&&(str_pool[j]=='.'))
+ ext_delimiter=j-str_start[str_ptr];
+ decr(j);
+ }
+ flush_list(link(garbage));
+ goto done;
+}
+
+@ A simple loop removes the double quotes and adjusts the |pool_ptr|.
+
+@<Suppress double quotes in braced input file name@>=
+for (k=j=str_start[str_ptr]; k < pool_ptr; k++) {@+
+ if (str_pool[k]!='"') {@+str_pool[j]=str_pool[k];incr(j);}
+}
+pool_ptr=j;
+
+@*1 The {\tt -recorder} Option.
+The recorder option can be used to enable the file name recorder.
It is crucial for getting a reliable list of files used in a given run.
Many post-processors use it, and it is used in \TeX\ Live for
checking the format building infrastructure.
@@ -34423,7 +34498,7 @@
work. Windows, however, seems to have no |pid_t|, so instead of storing the
value returned by |getpid|, we immediately consume it.
-@<k\TeX\ auxiliar functions@>=
+@<\TeX\ Live auxiliary functions@>=
static char *recorder_name=NULL;
static FILE *recorder_file=NULL;
static void
@@ -34454,7 +34529,7 @@
called with does not; it is just the log file name with {\tt .log}
replaced by {\tt .fls}.
-@ @<k\TeX\ auxiliar functions@>=
+@ @<\TeX\ Live auxiliary functions@>=
static void
recorder_change_filename (const char *new_name)
{ char *temp = NULL;
@@ -34461,7 +34536,7 @@
if (!recorder_file)
return;
#if defined(_WIN32)
- fclose (recorder_file); /* An opened file cannot be renamed. */
+ fclose (recorder_file); /* An open file cannot be renamed. */
#endif /* |_WIN32| */
if (output_directory) {
temp = concat3(output_directory, DIR_SEP_STRING, new_name);
@@ -34468,15 +34543,14 @@
new_name = temp;
}
- /* On windows, renaming fails if a file with |new_name| exists. */
#if defined(_WIN32)
- remove (new_name); /* Renaming fails if a file with the |new_name| exists. */
+ remove (new_name); /* A file with the |new_name| must not exist. */
#endif /* |_WIN32| */
rename(recorder_name, new_name);
free(recorder_name);
recorder_name = xstrdup(new_name);
#if defined(_WIN32)
- recorder_file = xfopen (recorder_name, FOPEN_A_MODE);
+ recorder_file = xfopen (recorder_name, FOPEN_A_MODE); /* A closed file must be opened. */
#endif /* |_WIN32| */
if (temp)
free (temp);
@@ -34483,14 +34557,14 @@
}
@ Now we are ready to record file names. The prefix INPUT is added
-to an input file and the prefiex OUTPUT to an output file.
+to an input file and the prefix OUTPUT to an output file.
But both functions for recording a file name use the same
function otherwise, which on first use will start the recorder.
-@<k\TeX\ auxiliar functions@>=
+@<\TeX\ Live auxiliary functions@>=
static void
recorder_record_name (const char *pfx, const char *fname)
-{ if (recorder_enabled) {
+{ @+if (recorder_enabled) {
if (!recorder_file)
recorder_start();
fprintf(recorder_file, "%s %s\n", pfx, fname);
@@ -34535,39 +34609,10 @@
}
-@ When string arguments specify files or directories,
-special care is needed if arguments are quoted and/or contain spaces.
-The function |normalize_quotes| makes shure that arguments containing
-spaces get quotes around them and it checks for unballanced quotes.
-@<k\TeX\ auxiliar functions@>=
-static char *normalize_quotes (const char *nom, const char *mesg)
-{@+ int quoted = false;
- int must_quote = (strchr(nom, ' ') != NULL);
- char *ret = xmalloc(strlen(nom)+3);/*room for two quotes and NUL*/
- char *p=ret;
- const char *q;
-
- if (must_quote)
- *p++ = '"';
- for (q = nom; *q; q++)
- if (*q == '"')
- quoted = !quoted;@+
- else
- *p++ = *q;
-
- if (must_quote)
- *p++ = '"';
- *p = '\0';
- if (quoted) {
- fprintf(stderr, "! Unbalanced quotes in %s %s\n", mesg, nom);
- exit(1);
- }
- return ret;
-}
-
-@ With the \.{-cnf-line} option it is possible to specify
+@*1 The {\tt -cnf-line} Option.
+With the \.{-cnf-line} option it is possible to specify
a line of text as if this line were part of \TeX's configuration
file---even taking precedence over conflicting lines in the configuration
file. For example it is possible to change \TeX's
@@ -34578,13 +34623,17 @@
initialized sufficiently.
@<Global...@>=
-char **cnf_lines=NULL;
-int cnf_count=0;
+static char **cnf_lines=NULL;
+static int cnf_count=0;
+@ @<handle the options@>=
+else @+if (ARGUMENT_IS ("cnf-line"))
+ add_cnf_line(optarg);
+
@ The function |add_cnf_line| stores the
given command line argument in the variable |cnf_lines|.
-@<k\TeX\ auxiliar functions@>=
+@<\TeX\ Live auxiliary functions@>=
static void add_cnf_line(char *arg)
{@+ cnf_count++;
cnf_lines=xrealloc(cnf_lines,sizeof(char*)*cnf_count);
@@ -34595,7 +34644,7 @@
\.{kpathsearch} library.
@<activate configuration lines@>=
-#if 0 /* this function does not exists allways */
+#if 1 /* this function does not exists always */
{@+ int i;
for (i=0;i<cnf_count;i++)
kpathsea_cnf_line_env_progname (kpse_def, cnf_lines[i]);
@@ -34603,11 +34652,50 @@
}
#endif
-@ After we are done with the options,
+
+@*1 Hi\TeX\ specific command line options.
+Hi\TeX\ provides options to set the \MF\ mode and the resolution if
+{\tt .pk} fonts must be rendered and/or included in the {\tt .hnt}
+output file. Further, a lot of debug output can be generated if
+Hi\TeX\ was compiled with debugging enabled. The {\tt -hint-debug-help}
+option gives a short summary of what to expect.
+@<handle the options@>=
+else @+if (ARGUMENT_IS("resolution")) @t\2@>
+{ option_dpi_str=optarg;
+ option_dpi=strtol(option_dpi_str,NULL,10);
+}
+else if (ARGUMENT_IS("mfmode"))
+ option_mfmode=optarg;
+#ifdef DEBUG at t\1@>
+else @+if (ARGUMENT_IS("hint-debug"))@t\2@>
+ debugflags=strtol(optarg,NULL,16);
+else @+if (ARGUMENT_IS("hint-debug-help"))
+{ fprintf(stderr,@/
+ "To generate HINT format debug output use the option\n"
+ " -hint-debug=XX"@/
+ at t\qquad@>"\t\t XX is a hexadecimal value. OR together these values:\n");@/
+fprintf(stderr,"\t XX=%04X \t basic debugging\n", DBGBASIC);@/
+fprintf(stderr,"\t XX=%04X \t tag debugging\n", DBGTAGS);@/
+fprintf(stderr,"\t XX=%04X \t node debugging\n",DBGNODE);@/
+fprintf(stderr,"\t XX=%04X \t definition debugging\n", DBGDEF);@/
+fprintf(stderr,"\t XX=%04X \t directory debugging\n", DBGDIR);@/
+fprintf(stderr,"\t XX=%04X \t range debugging\n",DBGRANGE);@/
+fprintf(stderr,"\t XX=%04X \t float debugging\n", DBGFLOAT);@/
+fprintf(stderr,"\t XX=%04X \t compression debugging\n", DBGCOMPRESS);@/
+fprintf(stderr,"\t XX=%04X \t buffer debugging\n", DBGBUFFER);@/
+fprintf(stderr,"\t XX=%04X \t TeX debugging\n", DBGTEX);@/
+fprintf(stderr,"\t XX=%04X \t page debugging\n", DBGPAGE);@/
+fprintf(stderr,"\t XX=%04X \t font debugging\n", DBGFONT);@/
+exit(0);
+}
+#endif
+
+@*1 The Input File.
+After we are done with the options,
we inform the \.{kpathsearch} library about the program name.
This is an important piece of information for the library
because the library serves quite different programs
-and its behaviour can be customized for each program
+and its behavior can be customized for each program
using configuration files.
After the program and engine name is set,
the library is ready to use.
@@ -34625,7 +34713,7 @@
@ After the options, the command line usually continues with the name of the input
file. Getting a hold of the input file name can be quite complicated,
-and we will have the first opportunity to use the \.{kpathsearch} library.
+but the \.{kpathsearch} library will help us to do the job.
We start by looking at the first argument after the options:
If it does not start with an
@@ -34634,9 +34722,9 @@
followed by a colon and a ``\.{\BS}'' which is used to separate
directory names. Finally, if the filename is a quoted string, we need
to remove the quotes before we use the \.{kpathsearch} library to find it
-and reattach the quotes afterwards.
+and reattach the quotes afterward.
-@<k\TeX\ auxiliar functions@>=
+@<\TeX\ Live auxiliary functions@>=
#ifdef WIN32
static void clean_windows_filename(char *filename)
{@+if (strlen (filename) > 2 && isalpha (filename[0]) &&
@@ -34713,7 +34801,7 @@
parsefirstlinep = texmf_yesno ("parse_first_line");@/@t~@>
@ We needed:
-@<k\TeX\ auxiliar functions@>=
+@<\TeX\ Live auxiliary functions@>=
static int texmf_yesno(const char *var)
{@+
char *value = kpse_var_value (var);
@@ -34722,15 +34810,15 @@
@ We need a stack, matching the |line_stack| that
contains the source file names;
-we postpone k\TeX\'s |full_source_filename_stack| to a later time.
+we postpone \TeX\ Live's |full_source_filename_stack| to a later time.
@<Global...@>=
-int @!source_filename_stack0[max_in_open], *const @!source_filename_stack = @!source_filename_stack0-1;
+static int @!source_filename_stack0[max_in_open], *const @!source_filename_stack = @!source_filename_stack0-1;
@ The function |print_file_line|
-prints ``file:line:error'' style messages. It looks for a
-filename in |source_filename_stack|, and if it fails to find
-one falls back on the ``non-file:line:error'' style.
+prints ``file:line:error'' style messages using
+the |source_filename_stack|. If it fails to find the file name, it
+falls back to the ``non-file:line:error'' style.
@<Basic printing...@>=
static void print_file_line(void)
@@ -34745,7 +34833,8 @@
}
}
-@ Most of the time \TeX\ is not running as \.{initex} or \.{virtex},
+@*1 The Format File.
+Most of the time \TeX\ is not running as \.{initex} or \.{virtex},
but it runs with a format file preloaded.
To set the format name, we first check if the format name was given on the
command line with an ``\.{\AM}'' prefix,
@@ -34784,8 +34873,8 @@
the file for a \TeX\ comment of the form ``\%\AM format''\footnote{${}^1$}{The idea of using this format came from Wlodzimierz Bzyl.}.
If found, we will use the format given there.
-@<k\TeX\ functions@>=
-static void parse_first_line (char*filename)
+@<\TeX\ Live auxiliary functions@>=
+static void parse_first_line(char*filename)
{@+ FILE *f=NULL;
if (filename==NULL) return;
f = open_in(filename,kpse_tex_format,"r");
@@ -34814,31 +34903,14 @@
}
}
-
-@ The \TeX\ Live infrastructure is able to generate format files,
-font metric files,
-and even some tex files, if required.
-
-@<enable the generation of input files@>=
- kpse_set_program_enabled (kpse_tfm_format, MAKE_TEX_TFM_BY_DEFAULT,
- kpse_src_compile);@/
- kpse_set_program_enabled (kpse_tex_format, MAKE_TEX_TEX_BY_DEFAULT,
- kpse_src_compile);@/
- kpse_set_program_enabled (kpse_fmt_format, MAKE_TEX_FMT_BY_DEFAULT,
- kpse_src_compile);@/
- kpse_set_program_enabled (kpse_pk_format, MAKE_TEX_PK_BY_DEFAULT,
- kpse_src_compile);@/
- xputenv("MAKETEX_BASE_DPI", option_dpi_str);
- xputenv("MAKETEX_MODE", option_mfmode);
-
-
-@ In the old days, \TeX\ was a \PASCAL\ program, and standard \PASCAL\
+@*1 Commands.
+In the old days, \TeX\ was a \PASCAL\ program, and standard \PASCAL\
did say nothing about a command line. So \TeX\ would open the terminal
file for input and read all the information from the terminal.
If you don't give \TeX\ command line arguments, this is still true today.
-In our present time, people got so much used to control the behaviour
+In our present time, people got so much used to control the behavior
of a program using command line arguments---especially when writing
-scripts---that k\TeX\ allows the specification of commands on
+scripts---that \TeX\ Live allows the specification of commands on
the command line which \TeX\ would normally expect on
the first line of its terminal input.
@@ -34850,7 +34922,7 @@
Further it skips initial spaces and replaces trailing spaces
and line endings by a single space.
-@<k\TeX\ auxiliar functions@>=
+@<\TeX\ Live auxiliary functions@>=
static void input_add_char(unsigned int c)
{@+ if (last >= max_buf_stack)
{@+max_buf_stack=last+1;
@@ -34883,9 +34955,12 @@
return (loc < last);
@+ }
+@ @<Forward declarations@>=
+static int input_command_line(void);
-@ When we open an output file, there is usually no searching necessary.
+@*1 Opening Files.
+When we open an output file, there is usually no searching necessary.
In the best case, we have an absolute path and can open it.
If the path is relative, we try in this order:
the |file_name| prefixed by the |output_directory|,
@@ -34895,7 +34970,7 @@
If we were successful with one of the modified names, we update
|name_of_file|.
-@<k\TeX\ functions@>=
+@<\TeX\ Live functions@>=
static FILE *open_out(const char *file_name, const char *file_mode)
{@+ FILE *f=NULL;
char *new_name=NULL;
@@ -34927,6 +35002,24 @@
return f;
}
+static bool a_open_out(alpha_file *f) /*open a text file for output*/
+{@+f->f=open_out((char*)name_of_file+1,"w");
+ return f->f!=NULL && ferror(f->f)==0;@+
+}
+
+static bool b_open_out(byte_file *f) /*open a binary file for output*/
+{@+f->f=open_out((char *)name_of_file+1,"wb");
+return f->f!=NULL && ferror(f->f)==0;@+
+}
+
+#ifdef @!INIT
+static bool w_open_out(word_file *f) /*open a word file for output*/
+{@+f->f=open_out((char *)name_of_file+1,"wb");
+ return f->f!=NULL && ferror(f->f)==0;@+
+}
+#endif
+
+
@ Format file names must be scanned
before \TeX's string mechanism has been initialized.
The function |update_name_of_file| will set |name_of_file| from
@@ -34938,8 +35031,8 @@
isn't found.
@^system dependencies@>
-@<k\TeX\ auxiliar functions@>=
-static void update_name_of_file(char *@!s, int @!k)
+@<\TeX\ Live auxiliary functions@>=
+static void update_name_of_file(const char *@!s, int @!k)
{@+ int j;
if (k <= file_name_size) name_length=k;@+
else name_length=file_name_size;
@@ -34948,12 +35041,11 @@
}
@ In standard \TeX, the |reset| macro is used to open input files.
-The \.{kpathsearch} library uses different search pathes for
+The \.{kpathsearch} library uses different search paths for
different types of files and therefore different functions are needed to
open these files. The common code is in the function |open_in|.
-
-@<k\TeX\ auxiliar functions@>=
+@<\TeX\ Live auxiliary functions@>=
static FILE*open_in(char*filename,kpse_file_format_type t,const char*rwb)
{@+ char *fname=NULL;
FILE*f= NULL;
@@ -34967,6 +35059,27 @@
return f;
}
+static bool a_open_in(alpha_file *f) /*open a text file for input*/
+{@+f->f= open_in((char *)name_of_file+1,kpse_tex_format,"r");
+ if (f->f!=NULL) get(*f);
+ return f->f!=NULL && ferror(f->f)==0;
+}
+
+static bool b_open_in(byte_file *f) /*open a binary file for input*/
+{@+f->f= open_in((char *)name_of_file+1,kpse_tfm_format,"rb");
+ if (f->f!=NULL) get(*f);
+ return f->f!=NULL && ferror(f->f)==0;
+}
+
+static bool w_open_in(word_file *f) /*open a word file for input*/
+{@+
+ f->f=NULL;
+ if (name_of_file[1]!=0)
+ f->f= open_in((char*)name_of_file+1,kpse_fmt_format,"rb");
+ if (f->f!=NULL) get(*f);
+ return f->f!=NULL && ferror(f->f)==0;
+}
+
@ \TeX's |open_fmt_file| function will call the following function
either with the name of a format file as given with an ``\.{\AM}''
prefix in the input or with |NULL| if no such name was specified.
@@ -34974,7 +35087,7 @@
-@<k\TeX\ functions@>=
+@<\TeX\ Live functions@>=
static bool open_fmt_file(void)
{@+int j=loc;
if (buffer[loc]=='&')
@@ -34993,60 +35106,26 @@
}
+@ The \TeX\ Live infrastructure is able to generate format files,
+font metric files,
+and even some tex files, if required.
-@ All the code that we have seen so far commes together in the function |main_init|
-which is the first function called in the main program of a \TeX\ engine
-belonging to \TeX\ Live. Before doing so, we make copies of
-argument count and argument vector putting them in global variables.
-@<Global...@>=
-static char **argv;
-static int argc;
+@<enable the generation of input files@>=
+ kpse_set_program_enabled (kpse_tfm_format, MAKE_TEX_TFM_BY_DEFAULT,
+ kpse_src_compile);@/
+ kpse_set_program_enabled (kpse_tex_format, MAKE_TEX_TEX_BY_DEFAULT,
+ kpse_src_compile);@/
+ kpse_set_program_enabled (kpse_fmt_format, MAKE_TEX_FMT_BY_DEFAULT,
+ kpse_src_compile);@/
+ kpse_set_program_enabled (kpse_pk_format, MAKE_TEX_PK_BY_DEFAULT,
+ kpse_src_compile);@/
+ xputenv("MAKETEX_BASE_DPI", option_dpi_str);
+ xputenv("MAKETEX_MODE", option_mfmode);
-@ @<k\TeX\ functions@>=
-static void main_init(int ac, char *av[])
-{@+
- char* main_input_file;
- argc = ac;
- argv = av;
- interaction = error_stop_mode;
- kpse_record_input = recorder_record_input;
- kpse_record_output = recorder_record_output;
- @<parse options@>@;
- @<set the program and engine name@>@;
- @<activate configuration lines@>@;
- @<set the input file name@>@;
- @<set defaults from the {\tt texmf.cfg} file@>@;
- @<set the format name@>@;
- @<enable the generation of input files@>@;
-}
-
-@ Before we can call the |parse_options| function,
-we might need some special preparations for Windows.
-
-@<parse options@>=
-#if defined(WIN32)
-{@+ char* enc;
- kpse_set_program_name (argv[0], NULL);
- enc = kpse_var_value("command_line_encoding");
- get_command_line_args_utf8(enc, &argc, &argv);@/
- parse_options (argc, argv);
- @<record {\tt texmf.cnf}@>@;
-}
-#else
- parse_options (ac, av);
-#endif
-
-@ Adding prototypes for the functions just defined allows \TeX\
-to call them.
-
-@<Header files and function declarations@>=
-
-static int input_command_line(void);
-static void main_init(int ac, char *av[]);
-
-@ We conclude this chapter using \.{time.h} to provide a function
+@*1 Date and Time.
+We conclude this chapter using \.{time.h} to provide a function
that is used to initialize
\TeX's date and time information. Because |time| is one of \TeX's
macros, we add the function |tl_now| before including \TeX's macros
@@ -35101,7 +35180,8 @@
return tp;
}
-@ To support \LaTeX, a few more time related functions are needed.
+@*1 Retrieving File Properties.
+To support \LaTeX, a few more time related functions are needed.
@<Header files and function declarations@>=
#define TIME_STR_SIZE 30
static char time_str[TIME_STR_SIZE];
@@ -35120,7 +35200,7 @@
the \TeX\ Live distribution and slightly modified.
-@<k\TeX\ auxiliar functions@>=
+@<\TeX\ Live auxiliary functions@>=
static void
make_time_str(time_t t, bool utc)
@@ -35241,7 +35321,7 @@
memset(md5_digest,0,DIGEST_SIZE);
if (file)
{ char *fname;
- pack_file_name(s,"","");
+ pack_file_name(s,empty_string,empty_string,NULL);
fname= find_input_file();
if(fname!=NULL)
{ FILE *f;
Modified: trunk/Build/source/texk/web2c/hitexdir/tex/hiltxpage.tex
===================================================================
--- trunk/Build/source/texk/web2c/hitexdir/tex/hiltxpage.tex 2022-10-26 23:49:22 UTC (rev 64830)
+++ trunk/Build/source/texk/web2c/hitexdir/tex/hiltxpage.tex 2022-10-27 09:18:40 UTC (rev 64831)
@@ -55,12 +55,12 @@
\HINTstream0 %main text
\HINTstream\footins
\HINTsetstream\topins = %topinsert
- prefered 0
+ preferred 0
{%
\HINTafter = {}
}
\HINTsetstream\footins =%footnotes
- prefered 255
+ preferred 255
ratio 0
{%
\hsize=300pt
Modified: trunk/Build/source/texk/web2c/hitexdir/tex/hiplainpage.tex
===================================================================
--- trunk/Build/source/texk/web2c/hitexdir/tex/hiplainpage.tex 2022-10-26 23:49:22 UTC (rev 64830)
+++ trunk/Build/source/texk/web2c/hitexdir/tex/hiplainpage.tex 2022-10-27 09:18:40 UTC (rev 64831)
@@ -45,12 +45,12 @@
\HINTstream0 %main text
\HINTstream\footins
\HINTsetstream\topins = %topinsert
- prefered 0
+ preferred 0
{%
\HINTafter = {}
}
\HINTsetstream\footins =%footnotes
- prefered 255
+ preferred 255
ratio 0
{%
\hsize=300pt
More information about the tex-live-commits
mailing list.