![[BACK]](/cvsweb/icons/back.gif){kind=icon}
Return to e6 CVS log ![[TXT]](/cvsweb/icons/text.gif){kind=icon}
![[DIR]](/cvsweb/icons/dir.gif){kind=icon}
Up to [CSRG BSD Unix] / 43BSDReno / share / doc / usd / 12.edtut|
Return to e6 CVS log | Up to [CSRG BSD Unix] / 43BSDReno / share / doc / usd / 12.edtut |
1.1 root 1: .\" @(#)e6 6.1 (Berkeley) 5/22/86
2: .\"
3: .SH
4: Special Characters
5: .PP
6: You may have noticed that things just don't work right when you used
7: some characters like
8: \*.,
9: .UL * ,
10: .UL $ ,
11: and others in
12: context searches and the substitute command.
13: The reason is rather complex, although the cure is simple.
14: Basically,
15: .ul
16: ed
17: treats these characters as special, with special meanings.
18: For instance,
19: .ul
20: in a context search or the first string of the substitute command only,
21: \*.
22: means ``any character,'' not a period, so
23: .P1
24: /x\*.y/
25: .P2
26: means ``a line with an
27: .UL x ,
28: .ul
29: any character,
30: and a
31: .UL y ,''
32: .ul
33: not
34: just ``a line with an
35: .UL x ,
36: a period, and a
37: .UL y .''
38: A complete list of the special characters
39: that can cause trouble is the following:
40: .P1
41: ^ \*. $ [ * \e
42: .P2
43: .ul
44: Warning:
45: The backslash character
46: .UL \e
47: is special to
48: .ul
49: ed.
50: For safety's sake,
51: avoid it where possible.
52: If you have to use one of the special characters
53: in a substitute command,
54: you can turn off its magic meaning temporarily
55: by preceding it with the backslash.
56: Thus
57: .P1
58: s/\e\e\e\*.\e*/backslash dot star/
59: .P2
60: will change
61: .UL \e.*
62: into ``backslash dot star''.
63: .PP
64: Here is a hurried synopsis of the other special characters.
65: First, the circumflex
66: .UL ^
67: signifies
68: the beginning of a line.
69: Thus
70: .P1
71: /^string/
72: .P2
73: finds
74: .UL string
75: only if it is at the beginning of a line:
76: it will find
77: .P1
78: string
79: .P2
80: but not
81: .P1
82: the string...
83: .P2
84: The dollar-sign
85: .UL $
86: is just the opposite of the circumflex;
87: it means the end of a line:
88: .P1
89: /string$/
90: .P2
91: will only find an occurrence of
92: .UL string
93: that is at the end of some line.
94: This implies, of course,
95: that
96: .P1
97: /^string$/
98: .P2
99: will find only a line that contains just
100: .UL string ,
101: and
102: .P1
103: /^\*.$/
104: .P2
105: finds a line containing exactly one character.
106: .PP
107: The character
108: .UL . ,
109: as we mentioned above,
110: matches anything;
111: .P1
112: /x\*.y/
113: .P2
114: matches any of
115: .P1
116: x+y
117: x-y
118: x y
119: x\*.y
120: .P2
121: This is useful in conjunction with
122: .UL * ,
123: which is a repetition character;
124: .UL a*
125: is a shorthand for ``any number of
126: .UL a 's,''
127: so
128: .UL .*
129: matches any number of anythings.
130: This is used like this:
131: .P1
132: s/\*.*/stuff/
133: .P2
134: which changes an entire line,
135: or
136: .P1
137: s/\*.*,//
138: .P2
139: which deletes all characters in the line up to and
140: including the last comma.
141: (Since
142: .UL .*
143: finds the longest possible match,
144: this goes up to the last comma.)
145: .PP
146: .UL [
147: is used with
148: .UL ]
149: to form ``character classes'';
150: for example,
151: .P1
152: /[0123456789]/
153: .P2
154: matches any single digit \-
155: any one of the characters inside the braces
156: will cause a match.
157: This can be abbreviated to
158: .UL [0\-9] .
159: .PP
160: Finally, the
161: .UL &
162: is another shorthand character \-
163: it is used only on the right-hand part of a substitute command
164: where it means ``whatever was matched on the left-hand side''.
165: It is used to save typing.
166: Suppose the current line contained
167: .P1
168: Now is the time
169: .P2
170: and you wanted to put parentheses around it.
171: You could just retype the line, but
172: this is tedious.
173: Or you could say
174: .P1
175: s/^/(/
176: s/$/)/
177: .P2
178: using your knowledge of
179: .UL ^
180: and
181: .UL $ .
182: But the easiest way uses the
183: .UL & :
184: .P1
185: s/\*.*/(&)/
186: .P2
187: This says ``match the whole line, and replace it
188: by itself surrounded by parentheses.''
189: The
190: .UL &
191: can be used several times in a line;
192: consider
193: using
194: .P1
195: s/\*.*/&? &!!/
196: .P2
197: to produce
198: .P1
199: Now is the time? Now is the time!!
200: .P2
201: .PP
202: You don't have to match the whole line, of course:
203: if the buffer contains
204: .P1
205: the end of the world
206: .P2
207: you could type
208: .P1
209: /world/s//& is at hand/
210: .P2
211: to produce
212: .P1
213: the end of the world is at hand
214: .P2
215: Observe this expression carefully,
216: for it illustrates how to take advantage of
217: .ul
218: ed
219: to save typing.
220: The string
221: .UL /world/
222: found the desired line;
223: the shorthand
224: .UL //
225: found the same
226: word in the line;
227: and the
228: .UL &
229: saves you from typing it again.
230: .PP
231: The
232: .UL &
233: is a special character only within
234: the replacement text of a substitute command,
235: and has no special meaning elsewhere.
236: You can turn off the special meaning of
237: .UL &
238: by preceding it with a
239: .UL \e :
240: .P1
241: s/ampersand/\e&/
242: .P2
243: will convert the word ``ampersand'' into the literal symbol
244: .UL &
245: in the current line.
This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.