ex.html

unix 下的C开发手册,还用详细的例程。

<b>$HOME/.exrc</b>
sets the editor option
<b>exrc</b>,
<i>ex</i>
finally will attempt to read
commands from a file
<b>.exrc</b>
in the current directory.
In the event that
<i>EXINIT</i>
is not set and the current directory is the
user's home directory, any
<b>.exrc</b>
file will only be processed once.
No
<b>.exrc</b>
file will be read unless it is owned by the
same user ID as the effective user ID of the process.
After any
<b>.exrc</b>
files are processed, any commands specified by the
<b>-c</b>
option will be processed.
<p>
By default,
<i>ex</i>
starts in the command mode, which is indicated by the <b>:</b> prompt.
The input mode can be entered by
<b>append</b>,
<b>insert</b>
or
<b>change</b>
commands;
it can be exited (and command mode reentered) by typing a period
(.)
alone at the beginning of a line.
There is one other mode, visual mode, in which
full-screen editing is available.
This is described more fully under the
<b>visual</b>
command and in the
<i><a href="vi.html">vi</a></i>
utility description.
The command line can consist of multiple
<i>ex</i>
commands separated
by vertical-line characters
(|).
The use of commands that enter input or
visual modes in this manner, unless they are the final command
on the line, produces undefined results.
<p>
Command lines beginning with the double-quote character
( )
are ignored.
This can be used for comments in an editor script.
<h5><a name = "tag_001_014_598_001">&nbsp;</a>Addressing in ex</h5>
<xref type="5" name="exaddr"></xref>
Addressing in
<i>ex</i>
relates to the
<i>current line</i>.
In general, the current line is
the last line affected by a command;
the exact effect on the current line
is discussed under the description of
each command.
When the buffer contains no lines, the current line is set to zero.
<p>
Addresses are constructed by one of the following methods:
<ol>
<p>
<li>
The address
.
(period)
refers to the current line.
<p>
<li>
The address "$" refers to the last line of the buffer.
<p>
<li>
The address
<i>n</i>,
where
<i>n</i>
is a decimal number, refers to the
<i>n</i>th
line of the buffer.
<p>
<li>
The address
<i>'x</i> refers to the line marked with the
mark-name character
<i>x</i>,
which must be a lower-case letter of the POSIX locale.
Lines can be marked with the
<b>ma</b>
or
k
commands described below.
<p>
<li>
A regular expression
(RE)
enclosed by slashes
(/)
is an address,
and refers to the first line found by searching
forward from the line following the current line
toward the end of the
buffer and stopping at the first line containing a
string matching the RE.
The second slash can be omitted at the end of a command line.
If the
<b>wrapscan</b>
option is set, the search will wrap
around to the beginning of the buffer
and continue up to and including the current line, so that the entire
buffer is searched.
<p>
<li>
An RE enclosed in question marks
(?)
addresses
the first line found by searching
backward
from the line
preceding
the current line
toward the beginning of
the buffer and stopping at the first line containing
a string matching the RE.
The second question mark can be omitted at the end of a command line.
If the
<b>wrapscan</b>
option is set, the search will wrap around
from the beginning of the buffer to the
end of the buffer and continue up to and including the
current line, so that the entire buffer is searched.
<p>
<li>
An address followed by a plus sign
(+)
or a minus sign
(-)
followed by a
decimal number
is an offset address, and refers to the first address plus
(respectively minus) the indicated number of lines.
If the address is omitted, the addition or subtraction
is taken with respect to the current line.
<p>
<li>
An address of "+" or "-" followed by a number
is taken with respect to the current line number; for example,
-5
is understood to mean
.-5.
<p>
<li>
An address ending with "+" or "-" has 1 added to or subtracted
from the address, respectively.
As a consequence of this rule and of rule 8 above, the address "-"
refers to the line preceding the current line.
Moreover, trailing "+" and "-" characters have a cumulative effect;
for example,
--
refers to the current line less 2.
<p>
<li>
A percent sign
(%)
stands for the
address pair
1,$.
<p>
</ol>
<p>
Commands require zero, one or two addresses.
See the descriptions of
<i>line</i>
and
<i>range</i>
in
<xref href=excmd><a href="#tag_001_014_598_002">
Command Descriptions in ex
</a></xref>.
Commands that require zero addresses regard the presence
of an address as an error.
<p>
Adjacent addresses in a
<i>range</i>
must be separated from each other by a comma
(,)
or a semicolon
(;).
In the latter case, the current line
(.)
is set to
the first address,
and only then is the second address calculated.
This feature can be used to determine the starting
line for forward and backward searches (see rules 5 and 6 above).
The second address of any two-address sequence
corresponds to a line that follows,
in the buffer, the line corresponding to the first address.
The first address must be less than or equal to the second address.
The first address must be greater than or equal to the first line
of the editing buffer and the last address must be less than or equal
to the last line of the editing buffer.
Any other case is an error.
<p>
All of the following examples are valid
<i>addresses</i>:
<dl compact>

<dt>+++<dd>three lines after the current line

<dt>/<i>re</i>/-<dd>one line before the next occurrence of
<i>re</i>

<dt>-2<dd>two lines before the current line.

</dl>
<h5><a name = "tag_001_014_598_002">&nbsp;</a>Command Descriptions in ex</h5>
<xref type="5" name="excmd"></xref>
The following symbols are used in this section to represent
optional modifiers.
Any or all can be omitted; the defaults are shown.
<dl compact>

<dt><i>line</i><dd>A single line address, given in any of the
forms described in
<xref href=exaddr><a href="#tag_001_014_598_001">
Addressing in ex
</a></xref>;
the default for
<i>line</i>
is the current line.

<dt><i>range</i><dd>A
<i>line</i>,
or a pair of line addresses, separated by a comma or semicolon
(see
<xref href=exaddr><a href="#tag_001_014_598_001">
Addressing in ex
</a></xref>
for the difference between the two);
the default for
<i>range</i>
is the current line only
(.,.).
A percent sign
(%)
stands for the range
(1,$).
If the range specified is such that the starting address exceeds
the ending address, the range is invalid and the command will
not be performed.
If more than the expected number of addresses are given in a range,
the greatest valid number of the last ones given will be used.
For example,
1,3,5p
prints lines 3 to 5, inclusive (because two is the
greatest valid number in the range accepted by
<b>print</b>).

<dt><i>count</i><dd>A positive integer, specifying the number of lines to be
affected by the command;
the default for
<i>count</i>
is 1.

<dt><i>flags</i><dd>One of the characters
"#",
p
or
l
(ell),
or both
"#"
and
l
to add numbers to list-format output.
When a command with such a flag
completes, the
addressed lines
will be written out as if by the corresponding
<b>#</b>,
<b>p</b>
or
<b>l</b>
command.
The use of
<i>flags</i>
applies to all lines written by the
<b>list</b>,
<b>number</b>,
<b>open</b>,
<b>print</b>,
<b>substitute</b>,
<b>visual</b>,
<b>&amp;</b>
and
z
commands; for other commands, it applies to the
current line at the completion of the command.
In addition, any number of "+" or "-" characters can also be
given after the flags,
in which case the line written is not the one affected by the
command, but rather the line addressed by the offset
address as described above.
The default for
<i>flags</i>
is null.

<dt><i>buffer</i><dd>One of a number of named areas for saving text.
The named buffers
are specified by the lower-case letters of the POSIX locale.
Specifying
<i>buffer</i>
causes the area of text affected by
the command to be stored into the buffer as it was before
the command took effect.
This argument is also used on the
<b>put</b>
command and the visual mode put commands
(<b>p</b>
and
<b>P</b>)
to specify the buffer that will provide the text to insert.

If the buffer name is specified in upper-case,
and the buffer is to be modified (as with a
deletion or yanking command),
the buffer will be appended to
rather than being overwritten.
If the buffer is not to be modified (as in a visual mode
<b>put</b>
command) the buffer name can be specified in lower-case or
upper-case with the same results.
There is also one unnamed buffer, which is the repository
for all text deleted (with the
<b>delete</b>
or visual mode
<b>d</b>
command) or yanked (with the
<b>yank</b>
or visual mode
<b>y</b>
command) when no buffer is specified.

There are also
numbered buffers, 1 to 9, inclusive, which are accessible only from
visual mode.
These buffers are special in that,
in visual mode, when deleted text is placed in the unnamed buffer,
it also is placed in buffer 1, the previous contents of
buffer 1 are placed in buffer 2, and so on.
Any text in buffer 9 will be lost.
Text that is yanked (or otherwise copied) into the unnamed
buffer does not modify the numbered buffers.
Text cannot be placed directly into the numbered
buffers although it can be retrieved from them by using a
visual mode
<b>put</b>
command with the buffer name given as a number.
When the
<i>buffer</i>
modifier is not used in the commands below, the unnamed buffer
is the default.

<dt><i>file</i><dd>A pattern used to derive a pathname; the default is the
current file, as defined above.
If no current file has yet been established, a warning will be written
and the command will be aborted, except where specifically noted
in the individual command descriptions that follow.
The pattern will be subjected to the process of shell
word expansions (see
<xref href=wordexp><a href="chap2.html#tag_001_006">
Word Expansions
</a></xref>);
if more than a single pathname results
and the command is expecting one file, the effects are unspecified.

<dt><i>word</i><dd>In the POSIX locale, a
<i>word</i>
consists of a maximal sequence of letters, digits and underscores,
delimited at both ends by characters other than
letters, digits or underscores, or by the beginning or end of
a line or the file.

<dt><b>!</b><dd>A character that can be
appended to the command to modify its operation, as detailed
in the individual command descriptions.

</dl>
<p>
If both a
<i>count</i>
and a
<i>range</i>
are specified for a
command that uses them, the number of lines affected will be
taken from the
<i>count</i>
value rather than the
<i>range</i>.
The starting line for the command is taken to be the
first line addressed by the
<i>range</i>.
<p>
When only a
<i>line</i>
or
<i>range</i>
is specified with no
command, the implied command is either a
<b>print</b>,
<b>list</b>
or
<b>number</b>
(<b>p</b>,
<b>l</b>
or
<b>#</b>).
The command selected will be the last
of these three commands to be used, including use as a
<i>flag</i>.
When no range or count is specified and the command line
is a blank line, the current line will be
written, and the current line will be set to
.+1.
<p>
Zero or more blank characters can precede or follow the addresses,
<i>count</i>,
<i>flags</i>
or command name.
Any object following a command name
(such as
<i>buffer</i>,
<i>file</i>
and so on)
that begins with an alphabetic character
will be separated from the command name
with at least one
blank character.
<p>
For each of the
commands listed below, the command can be entered as the
abbreviation (those characters in the Synopsis command word preceding the
<b>[),</b>
the full command (all characters shown for the command word, omitting the
<b>[</b>
and
<b>]),</b>
or any subset of the characters of the full command down to the abbreviation.
For example, the
<b>args</b>
command (shown as
<b>ar[gs]</b>
in the Synopsis)
can be entered as
<b>ar</b>,
<b>arg</b>
or
<b>args</b>.
<h5><a name = "tag_001_014_598_003">&nbsp;</a>Abbreviate</h5>
<i>Synopsis</i>:   
ab<b>[</b>brev<b>]</b><i> word rhs</i>
<p>
Add the named abbreviation to the current abbreviation list.
In visual mode, if
<i>word</i>
is typed so that it is preceded
and followed by characters that cannot be part of a
<i>word</i>
(as defined previously),
it will be replaced by the string
<i>rhs</i>.