1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
|
Detex - Version 2.7
Detex is a program to remove TeX constructs from a text file. It recognizes
the \input command.
This program assumes it is dealing with LaTeX input if it sees the string
"\begin{document}" in the text. It recognizes the \include and \includeonly
commands.
This directory contains the following files:
README - you're looking at it.
Makefile - makefile for generating detex on a 4.2BSD Unix system.
detex.1l - troff source for the detex manual page.
Assuming you have the -man macros, use "make man-page" to
generate it.
detex.h - Various global definitions. These should be modified to suit
the local installation.
detex.l - Lex and C source for the detex program.
lexout.c - C code generated from detex.l using lex on a Sun (SunOS 4.1.1)
This can be useful for DOS, OS/2 or systems that don't have
lex.
states.sed - sed(1) script to munge the state names in detex.l so that
reasonable names can be used in the source without causing
lex(1) to overflow.
os2 - subdirectory containing support for compilation on
OS/2 and DOS systems
Feel free to redistribute this program, but distribute the complete contents
of this directory. The latest version is available at
http://www.cs.purdue.edu/homes/trinkle/detex/ Send comments and fixes
to me via email.
Daniel Trinkle <trinkle@cs.purdue.edu>
Department of Computer Sciences
Purdue University
West Lafayette, IN 47907-1398
April 26, 1986
Modified -- June 4, 1986
Changed so that it automatically recognizes LaTeX source and ignores several
environment modes such as array.
Modified (Version 2.0) -- June 30, 1984
Now handles white space in sequences like "\begin { document }". Detex is not
as easily confused by such things as display mode ends and begins that don't
match up.
Modified -- September 19, 1986
Added the "-e <environment-list>" option to ignore specified LaTeX
environments.
Modified -- June 30, 1987
Added the "-n" no-follow option, to allow detex to ignore \input and \include
commands. Also changed the algorithm for locating the input files. It now
interprets the "." more reasonably (i.e. it is not always the beginning of an
extension).
Modified -- December 15, 1988
Added handling of verbatim environment in LaTeX mode and added it to the list
of modes ignored by default. Because of limitations with lex, it was
necessary to shorten the names of some of the existing start states before
adding a new one (ugh).
Modified -- January 3, 1988
Added ignore of \$ inside $$ (math) pair.
Modified (Version 2.2) -- June 25, 1990
Control sequences are no longer replaced by space, but just removed. This
means accents no longer cause output words to be broken. The "-c" option was
added to show the arguments of \cite, \ref, and \pageref macros. This is
useful when using something like style on the output.
Modified (Version 2.3) -- September 7, 1990
Fixed the handling of Ctl mode a little better and added an exception
for \index on suggestions from kcb@hss.caltech.edu (KC Border). Also
changed the value for DEFAULTINPUTS to coincide with a local change.
Modified -- February 10, 1991
Added -t option to force TeX mode even when seeing the "\begin{document}"
string in the input.
Modified -- February 23, 1991
Based on suggestions from pinard@iro.umontreal.ca (Francois Pinard), I
added support for the SysV string routines (-DUSG), added defines for
the flex lexical scanner (-DFLEX_SCANNER), changed NULL to '\0' when
using it as a character (his sys defined NULL as (void *)0), changed
the Makefile to use ${CC} instead of cc, and added comments about the
new compile time options.
Modified (Version 2.4) -- September 2, 1992
Corrected the way CITEBEGIN worked. Due to serious braindeath I had
the condition wrong in the if test. It should be (fLatex && !fCite).
This solves the problem a couple people reported with amstex style
\ref entries.
Added a preprocessing sed(1) command to replace all the long, easy to
read state names with short two letter state names (SA-S?) so that lex
won't overflow and I don't have to keep shortening the state names
every time I add one. If a state is added, it must also be added to
states.sed (order is important) along with its unique S? replacement.
Added \pagestyle, \setcounter, and \verb handling from
K.Sattar@cs.exeter.ac.uk (Khalid Sattar). Also allows invocation as
"delatex" to force LaTeX mode.
Applied patches from queinnec@cenatls.cena.dgac.fr (Philippe Queinnec)
to handle nested {}s in state <LaMacro2> (\bibitem, \cite, \index).
Added special ligature handling (\aa, \ae, \oe, \l, \o, \i, \j, \ss)
at the suggestion of gwp@dido.caltech.edu (G. W. Pigman II).
Cleaned up the comments on detex.h, added mathmatica to DEFAULTENV.
Modified (Version 2.5) -- January 28, 1993
Leading spaces in macros are no longer stripped. This means
"foo\footnote{ bar}" comes out as "foo bar" instead of "foobar".
Fixed special ligature handling so it works in cases line {\ss} instead of
just when followed by a space.
Modified (Version 2.6) -- July 30, 1993
Added OS/2 port from hankedr@mail.auburn.edu (Darrel R Hankerson).
Added handling of leading and trailing ':' in TEXINPUTS per the latest
TeX as suggested by jnp@tfl.dk (J|rgen N|rgaard).
Changed the way the input path is constructed in SetInputPaths() so we
never try to modify a constant string.
Changed the way the the ignore environment list is contructed in
SetEnvIgnore() so we never try to modify a constant string.
Changed the USG define to HAVE_STRING_H.
Fixed the states.sed script so it only replaces "Input" in the correct
places. I would like to use the \< \> word separator patterns but
they are not supported by all versions of sed. This as least works.
Changed the detex.c target in the Makefile to use a temporary file
because I experienced problems (segmentation fault) with lex on
Solaris 2.1 when input was from stdin.
Modified (Version 2.7) -- September 10, 1997
Removed line breaks in detex.l between a few patterns and actions. It
appears that flex is no longer able to handle this. Thanks to Anthony
Harris <harris@hebb.neurology.pitt.edu> and Marty Leisner
<leisner@sdsp.mc.xerox.com> for reporting this.
Porting notes -- March 30, 1992
When using gcc, or compiling on a NeXT, you should compile with
-fwritable-strings. With the change to SetInputPaths() in 2.6 this
should no longer be necessary.
On a NeXT, it has been reported that lex replaces the '\0' with NULL,
and then the compiler complains about it. I think this is an old bug
that is no longer applicable.
July 30, 1993
The flex scanner generator does not work because it does not handle
input buffering the same way as lex. I don't know of any reasonable
way to rewrite detex to get around this problem.
May 25, 1995
According to alain@ia1.u-strasbg.fr (Alain Ketterlin), using flex
allows 8-bit characters to be handled correctly.
|
