tedit: Edit a table.

Package: nttools

Usage

tedit table

Description

This task is a screen editor for STSDAS tables. You edit a table by moving the cursor around the screen with the cursor keys and typing in the new value of the table element. The screen scrolls both sideways and up and down as you move the cursor, so all elements of the table can be reached. Other editing commands are entered on the command line. To switch from table editing mode to command line mode, you press the EXIT key (usually Control-Z, however, you can change this). After performing your command, the editor returns to table editing mode, unless the command exits the editor. The most important commands in command mode are `help', `exit', and `quit'. The `help' command displays all the editing key bindings and the command line commands. The `exit' command will get you out of the editor and automatically save the edited table. The `quit' command will get you out of the editor after asking you whether you want to save the table. By default, the editor modifies a copy instead of the original table, so if you quit without saving the table, the original table is still there without any modifications.

If you try to edit a table that does not exist, the editor will ask if you want to create the table. If you answer "no", the editor will exit. If you answer "yes", the editor will ask you for each column name, type, unit, and print format. When you have finished entering all the columns, press the return key instead of entering another column name. The editor will create the table and put you in table editing mode.

To add a new, blank line to the end of a table, press the return key while you are on the last line of the table. You can add blank lines anywhere in the table with the `add row' command, which will be described later.

Some editing commands are entered from the command line in command mode. To get to command line mode, press the exit key. This key is bound to Control-Z by default. If you enter a blank line, the editor will return to table editing mode. Some commands take arguments. They can be included when the command is entered, or if they are omitted, the editor will prompt you for their values. If the argument has embedded blanks, the argument should be enclosed in quotes if passed on the command line. No quotes should be used if the argument is entered interactively. When the editor interactively prompts you for a command argument it will also display a default value for the argument. Pressing the return key gets the default value. Some command names are two words long, for example, "add row". Usually the second word is optional and modifies the meaning of the first, for example "copy append". If the second word is not optional and you omit it, the editor will prompt you for it. All command names can be abbreviated to one or more letters. If the command name is two words long, both words can be abbreviated to one or more letters.

The following is a list of the available commands:

add column <name> <type> <format> <units>: Add a new column to the table with the specified name and data type.

add row <row> <number>: Add new, blank rows after row number <row>. The legal range of <row> is 0 to the number of rows in the table. The number of blank rows to add is <number>.

copy <first> <last>: Copy the rows between <first> and <last> into the paste buffer. The current contents of the paste buffer are destroyed before the copy. The table is not modified by this command. The contents of the paste buffer can be put back into the table by the 'insert' command.

copy append <first> <last>: Copy the rows between <first> and <last> into the paste buffer. The current contents of the paste buffer are preserved and the new rows are inserted after them.

delete <first> <last>: Delete the rows between <first> and <last>. The deleted rows are placed into the paste buffer and the current contents of the paste buffer are destroyed.

delete append <first> <last>: Delete the rows between <first> and <last>. The deleted rows are appended to the paste buffer.

exit: Exit the table editor, saving any changes made to the table.

find <expression>: Find the next row in the table which makes <expression> true and move the cursor to that row. The expression has the same syntax as an expression in a Fortran if statement. The variables in the expression are column names. For more information on the syntax of the expression, read the help for 'tselect'. The direction of the search depends upon previous 'find' commands. By default the search direction is forward; however, if a "find backwards" command has been executed previously, searches will be done in a backwards direction until a "find forward" command is executed.

find forward <expression>: Find the next row in the table which makes <expression> true and move the cursor to that row. The search is done in the forwards direction.

find backwards <expression>: Find the next row in the table which makes <expression> true and move the cursor to that row. The search is done in the backwards direction.

goto <row> <column>: Move the cursor to <row> and <column>.

help: Display online help information for the table editor. The help includes a brief description of each command line command and the key bindings for table editing commands.

insert <row>: Insert the contents of the paste buffer after row number <row>. The contents of the paste buffer are not changed.

lower <column>: Convert <column> to lower case. Only string columns can be converted.

next: Repeat the previous find command, using the same expression and search direction that was used with it.

next forward: Repeat the previous find command, changing the search direction to forwards.

next backwards: Repeat the previous find command, changing the search direction to backwards.

quit: Exit the table editor. If the table has been changed, the table editor will ask you whether to save it before exiting.

set <column> <expression>: Set a column equal to an expression. If the column is a string column, the expression must be a constant. If the column is numeric, the expression can either be a constant or a Fortran-like expression. For the exact syntax of the expression, see the help file for tcalc.

substitute <column> <target> <replacement>

Search for and replace text patterns in a column. The syntax for the target and replacement pattern strings largely follows that used in the substitute command by the Unix text editors `ed' and `ex'. The pattern consists of a sequence of ordinary characters, which match themselves, and meta-characters, which match a set of characters. A meta-character can be matched as if it were an ordinary character by preceding it with the escape character, `\'. For example, the escape character itself is indicated in a pattern by `\\'. The meta-characters which can be used in the target pattern are:

beginning of string     ^       end of string           $
white space             #       escape character        \
ignore case             {       end ignore case         }
begin character class   [       end character class     ]
not, in char class      ^       range, in char class    -
one character           ?       zero or more occurrences *
begin tagged string     \(      end tagged string       \)

A set of characters is indicated in the target string by the character class construct. For example, punctuation could be indicated by `[,;.!]'. A range of characters contiguous in the underlying character set can be abbreviated by the range construct. For example, `[a-z]' matches any lower case character. The complement of a character set is indicated by making `^' the first character in a class. For example, `[^0-9]' matches any non-digit. Repetition of a character or character class is indicated by the following it with the `*' meta-character. Thus, zero or more occurrences of a lower case character is indicated by `[a-z]*'. The tagged string meta-characters have no effect on the match, they only serve to identify portions of the matched string for the replacement pattern. The meta-characters which are used in the replacement pattern are the following:

entire string           &       tagged string           \n
capitalize              \u      upper case              \U
lower case              \L      end case conversion     \e \E

The ditto meta-character, `&`, indicates that the entire portion of the string that was matched by the target pattern. The tag meta-character indicates that the n-th tagged string. For example, `\1' indicates the first tagged string and `\2' the second. The remaining meta-characters affect the case of the output string. The capitalization meta-character only affects the immediately following meta-character, but the upper and lower case meta-characters must be turned off explicitly with `\e' or `\E'.

upper <column>: Convert <column> to upper case. Only string columns can be converted.

The bindings to the table editing keys are read from the edcap file. This is the same file which is used to define the key bindings for the parameter editor and history editor. The edcap file defines key bindings which resemble those of commonly used text editors. Three edcap files are distributed with IRAF. They define key bindings which resemble EDT, Emacs, and vi. These edcap files are located in the 'dev$' directory and have the extension '.ed'. The appropriate file is chosen according to the value of the environment variable 'EDITOR'. If you want to customize the key bindings of the table editor, copy the appropriate edcap file from the 'dev$' directory to your 'home$' directory and edit the second column of the file. The table editor searches your home directory first for the edcap file and if it does not find it, then it searches the 'dev$' directory.

The table editor also uses the termcap file to determine the screen size and the escape sequences used to modify the screen. There are entries in the termcap file for almost all terminal types. The proper entry is selected according to the environment variable 'TERMINAL'. To change your terminal type or the screen size, use the IRAF 'stty' command.

The 'tread' task can also be used to view a file in readonly mode.

Parameters

table [string]: The name of the table to be edited. The editor checks for the existence of the table and its access mode before editing. If the table does not exist, the editor will ask whether you want to create a new table. If you do not have write access to a table you can only edit it by setting 'rdonly=yes'.

(columns = "") [string]

The names of the columns to be edited. A null or blank string means edit all columns. A column template consists of a list of either column names or column patterns containing the usual pattern matching meta-characters. The names or patterns are separated by commas or white space. The list can be placed in a file and the name of the file preceded by an "@" given in its place. If the first character in the column template is a bang (!), all columns NOT named will be displayed. The 'tlcol' task (with the 'nlist' parameter set to 1) may be used to generate a list of column names so there is no question about spelling. This list may be edited to rearrange or delete the names, and then the list file is given preceded by an '@' sign, for example:

tt> tedit junk columns=@colnames.lis

(silent = no) [boolean]: Turn off the bell indicating warning messages?

(rdonly = no) [boolean]: View a table without modifying it? This parameter prevents you from executing any command that would modify the file.

(inplace = no) [boolean]: Replace existing table? If 'rdonly' is set to "yes" the table is always edited in place.

Examples

1. Make a copy of the table 'm12b.tab' (if it exists) and edit the copy. If the table does not exist then a temporary table is created, and you will be prompted for the name of the first column to be created. In either case, if you exit (rather than quitting) the temporary table will be renamed to 'm12b.tab'.

tt> tedit m12b

2. Display the columns 'SHARP' and 'ROUND' in an existing table. Rows may be added or deleted, and columns may be added.

tt> tedit m12b columns="SHARP,ROUND"

Bugs

References

This task was written by Bernie Simon.