View source for Text manipulation with GAMAP

== String basics ==

=== Creating text and numeric strings ===

We may form a string of text characters in IDL in the following ways:

# by placing text between single and double quotes
# by parsing a number with IDL's STRING function
# by concatenating one string with another

For example:

 ; Create a text string 
 IDL> str1 = 'hello world'
 IDL> help, str1
 STR1            STRING    = 'hello world'
  
 ; Create a numeric string
 IDL> num2 = 3.14159
 IDL> str2 = string( num2 )   
 IDL> help, str2  
 STR2            STRING    = '      3.14159'
 
 ; Strip leading and trailing white space 
 IDL> str2 = strtrim( str2, 2 )
 IDL> help, str2
 STR2            STRING    = '3.14159'
 
 ; Create a new string by concatenating 2 strings
 IDL> a = 'I am string 1'
 IDL> b = 'I am string 2'
 IDL> c = a + ':' + b
 IDL> print, c
 I am string 1:I am string 2

You can use IDL's STRTRIM function to strip the leading and trailing whitespace.

=== Equivalence of strings and byte arrays ===

In IDL, a string of text characters is equivalent to an array of byte values.  A byte is a collection of 8 bits and may express values from 0-255.  The [http://www.asciitable.com/ ASCII collating sequence] has 255 values.  (Actually, the original ASCII table had 128 values, but this was later extended to 255 values to include special characters.)  One byte represents a single ASCII text character.

This means that it is easy to convert between strings and bytes in IDL.  If you have an array of bytes, you can use any of the IDL string routines on them, for example:

 IDL> byte_array = [ 72B, 69B, 76B, 76B, 79B ]
 IDL> help, byte_array    
 BYTE_ARRAY      BYTE      = Array[5]
 IDL> print, strtrim( byte_array, 2 ) 
 HELLO

GAMAP comes with a very useful routine called STR2BYTE.  This allows you to take a text string and to convert it into the equivalent array of bytes.  

 IDL> str = 'IDL is neat!'
 IDL> byte_array = str2byte( str, strlen( str ) )
 IDL> help, byte_array
 BYTE_ARRAY      BYTE      = Array[12]
 IDL> print, byte_array   
   73  68  76  32 105 115  32 110 101  97 116  33

Note that we used IDL's STRLEN function to return the length of the string.

=== Representing special characters ===

We must specify some special non-printing ASCII characters with their byte value.  For exaaple, the horizontal tab character is the 9th character in the ASCII table, so we may specify that as:

 IDL> tab = 9B
 IDL> help, tab
 TAB             BYTE      =    9
 IDL> str = 'hello' + string(tab) + 'world' 
 IDL> print, str
 hello   world

For more information about IDL's string functions, please see [http://idlastro.gsfc.nasa.gov/idl_html_help/Strings.html http://idlastro.gsfc.nasa.gov/idl_html_help/Strings.html.]

== Locating text within a string ==

The following routines can be used to locate text within a string variable:

:;STRCMP: IDL routine to compare text in one string to another
:;STRPOS: IDL routine to test for the existence of a substring within a string
:;STRMATCH: IDL routine that will test for strings that match certain patterns
:;STRMID: IDL routine that will return a substring from a string
:;STRWHERE: GAMAP routine that returns the locations of a single character within a string
:;RSEARCH: GAMAP routine to look for text in a string starting from the end of the string
:;STRRIGHT: GAMAP routine that returns the last N characters from a string

STRCMP can be used to test if two strings are equivalent.  You can do the test for all characters in a string, or just for the first N characters.  For example:

 ; Test to see if 2 strings are equivalent
 ; Specifying /FOLD_CASE keyword will do a case-insensitive search
 IDL> str1 = 'My baloney has a first name, it is O.S.C.A.R.'
 IDL> str2 = 'My baloney has a second name, it is M.A.Y.E.R.'
 IDL> print, strcmp( str1, str2, /fold_case )
    0
 
 ; This time only test the 1st 16 characters in both strings
 IDL> print, strcmp( str1, str2, 16, /fold_case )
    1

STRPOS is an easy way to test if a given substring is located within larger string:

 IDL> print, strpos( 'She sells seashells by the seashore', 'sea' )
           10

Note that even though the substring "sea" occurs twice in the above string, STRPOS will only return the location of the first occurrence.

STRMATCH can be used to test if certain strings match a given pattern.  You can use the following wild cards in your search string. 

* <tt>*</tt> matches any string
* <tt>?</tt> matches any single character
* <tt>[a..c]</tt> matches any of the enclosed characters (in this case a thru c)

For example:

 ; Find ALL 4-LETTER WORDS in a string array 
 ; that begin with “f” or “F” and end with “t” or “T”:
 IDL> str = ['foot', 'Feet', 'fate', 'FAST', 'ferret', 'fort']
 IDL> ind = where( strmatch( str, 'f??t', /fold_case ) eq 1 )
 IDL> print, ind
           0           1           3           5
 IDL> print, str[ind]
 foot Feet FAST fort
 
 ; Find all words OF ANY LENGTH in a string array 
 ; that begin with “f” or “F” and end with “t” or “T”:
 IDL> ind = where( strmatch( str, 'f*t', /fold_case ) eq 1 )
 IDL> print, ind
            0           1           3           4           5
 IDL> print, str[ind]
 foot Feet FAST ferret fort
 

STRMID can be used to extract any sized substring from a larger string.  You need to specify the starting location and number of characters to be extracted.  For example:

 IDL> str = 'The quick brown fox jumped over the lazy dog'
 IDL> print, strmid( str, 4, 5 )
 quick
 IDL> print, strmid( str, 0, 3 )
 The
 IDL> print, strmid( str, 10, 5 )
 brown

'''''TIP: Always remember that the first character has index 0!'''''

STRWHERE returns the location of a single character in a larger string.

 IDL> print, strwhere( 'anthony aardvark asked about auditory access', 'a' )
     0    8    9   13   17   23   29   38

If you want to search for text starting from the end of the string, you can use the RSEARCH function.  This can be useful in removing or replacing from a file name.  For example:

 ; Replace the *.pro extension with *.txt
 IDL> str = '/home/bmy/IDL/gamap2/gamap_util/gamap.pro'
 IDL> ind = rsearch ( str, '.' )
 IDL> str2 = strmid( str, 0, ind ) + '.txt
 IDL> print, str2
 /home/bmy/IDL/gamap2/gamap_util/gamap.txt

Finally, STRRIGHT returns the last N characters from a string.

 IDL> print, strright( 'anthony aardvark asked about auditory access', 6 )              
 access

== Replacing characters in a string ==

The following routines can be used to replace text within a string variable:

:;STRPUT: IDL routine to insert text into a string
:;REPLACE_TOKEN: GAMAP routine that replaces occurrences of tokens with text. Can also be used to expand wildcards with a name list.
:;STRREPL: GAMAP routine that replaces all occurences of one character in a string with another character.

IDL's STRPUT function is one way to insert characters into a string of text:

 IDL> str1 = 'Now is the winter of our discontent'
 IDL> strput, str1, 'summer', 11
 IDL> print, str1
 Now is the summer of our discontent

However, this requires that you provide the location in the string where the text replacement will take place.  In the above example, we insert the text at character 11 (the 1st character in a string is always character 0).

The above task is much more easily accomplished with GAMAP's REPLACE_TOKEN function:

 IDL> str1 = 'Now is the winter of our discontent'
 IDL> str2 = replace_token( str1, 'winter', 'summer', delim="" )
 IDL> print, str2
 Now is the summer of our discontent

With REPLACE_TOKEN you do not need to know the position in the string where the replacement text will be inserted.  Also, if the text to be replaced occurs more than once in the string, REPLACE_TOKEN will do the text replacement for the whole string in one fell swoop.  For example:

 IDL> print, replace_token( 'She sells seashells by the seashore', 'sea', 'ocean', delim="")
 She sells oceanshells by the oceanshore

Note that REPLACE_TOKEN can replace strings that are of different lengths (e.g. in the above example "sea" is replaced by "ocean").

STRREPL that allows you to replace multiple instances of a single character in a string.  For example:

 IDL> print, strrepl( 'Mississippi', 'i', 'a' )
 Massassappa

But if you need to replace an entire word rather than just single characters it's better to use REPLACE_TOKEN.

== Splitting strings into substrings ==

You can split a string into individual substrings with GAMAP's STRBREAK function.

 ; Use STRBREAK to split the line by spaces
 IDL> result = strbreak( 'The sunshine of our li..ii..ii..ii..ife', ' ' )
 IDL> for j = 0, n_elements( result )-1 do print, fix(j), ':', result[j]
        0:The
        1:sunshine
        2:of
        3:our
        4:li..ii..ii..ii..ife
 
 ; Use STRBREAK to split the line by commas
 IDL> result = strbreak( 'Parsley,Sage,Rosemary,and Thyme', ',' )
 IDL> for j = 0, n_elements( result )-1 do print, fix(j), ':', result[j]
        0:Parsley
        1:Sage
        2:Rosemary
        3:and Thyme

STRBREAK will return an array of values.  The first element is the first word, the second element is the second word, etc.

We recommend that you use GAMAP's STRBREAK rather than IDL's STRSPLIT or STR_SEP routines.  STR_SEP was the standard routine to separate strings until IDL 5.2.  In IDL 5.3 and higher, STR_SEP was obsoleted and replaced with the new STRSPLIT routine.

* If you are using IDL 5.2 or lower, then STRBREAK will call STR_SEP to break the string.  
* If you are using IDL 5.3 or higher, then STRBREAK will call STRSPLIT to break the string.

Therefore, STRBREAK will work properly regardless of which version of IDL you are using.

== GAMAP's string inquiry functions ==

GAMAP ships with the following string inquiry functions:

:;ISALGEBRAIC: Locates the position of algebraic characters in a string (e.g. locations that are EITHER digits '.' OR +/- signs).
:;ISALNUM: Locates the position of alphanumeric characters ( A...Z, a...z, 0..9 ) in a string.
:;ISALPHA: Locates the positions of alphabetic characters ( A...Z, a...z ) in a string.
:;ISDIGIT: Locates the positions of numeric characters ( '0' ... '9') in a string. 
:;ISGRAPH: Locates the positions of graphics characters (i.e. printable characters excluding SPACE) in a string.
:;ISLOWER: Locates the positions of lowercase alphabetic characters in a string.
:;ISPRINT: Locates the positions of all printable characters (including SPACE) in a string.
:;ISSPACE: Locates the positions of all white space characters in a string.
:;ISUPPER: Locates the positions of all uppercase alphabetic characters in a string.

Each of the above routines return a vector of 0's and 1's corresponding to each character in the string that satisfies the given criteria.

Some examples:

 IDL> str = '#99# Bottles of *Beer* on the Wall!'  
 
 IDL> print, isalgebraic( str ), format='(35i1)'
 01100000000000000000000000000000000
 
 IDL> print, isalnum( str ), format='(35i1)'
 01100111111101100111100110111011110
 
 IDL> print, isalpha( str ), format='(35i1)'
 00000111111101100111100110111011110
 
 IDL> print, isdigit( str ), format='(35i1)'
 01100000000000000000000000000000000
 
 IDL> print, isgraph( str ), format='(35i1)'
 11110111111101101111110110111011111
 
 IDL> print, islower( str ), format='(35i1)'
 00000011111101100011100110111001110
 
 IDL> print, isprint( str ), format='(35i1)'
 11111111111111111111111111111111111
 
 IDL> print, isspace( str ), format='(35i1)'
 00001000000010010000001001000100000
 
 IDL> print, isupper( str ), format='(35i1)'
 00000100000000000100000000000010000

== GAMAP's string formatting functions ==

GAMAP ships with the following string formatting functions:

:;STRSCI: Converts a number to a string in scientific notation format ( e.g. A x 10^B )
:;STRCHEM: Superscripts or subscripts numbers and special characters ('x', 'y') found in strings containing names of chemical species.

STRSCI can be used to put a string into scientific notation.  The string will contain the appropriate Hershey characters so that it can be passed to PLOT or XYOUTS. 

 IDL> str = STRSCI( 2000000, format='(i1)' )
 IDL> print, str
 2 x 10!u6!n

STRCHEM can be used to create strings with superscripts and subscripts (e.g. H<sub>2</sub>O, <sup>222</sup>Rn) for plotting purposes:

 IDL> print, strchem( 'NOx', /sub )            
 NO!lx!n
 
 IDL> print, strchem( '222Rn', /sup )
 !u2!n!u2!n!u2!nRn

== Functions for working with file and path names ==

The following string-handling functions are specially geared towards working with file and path names:

:;EXPAND_PATH: IDL routine to expand wild cards in a file name or path name
:;EXTRACT_FILENAME: GAMAP routine to extract a file name from a fully qualified file path
:;EXTRACT_PATH: GAMAP routine to extract the directory from a fully qualified file path
:;ADD_SEPARATOR: GAMAP routine to make sure a file name ends with a directory path separator

Sometimes it is necessary to fully expand the wild cards in a filename.  (For example, the IDL routine HDF_BROWSER will choke if it finds the Unix wild card character ~ in the file path, so you have to expand to the full file path.)  This can be easily done with EXPAND_PATH:

 IDL> print, expand_path( '~bmy/IDL/gamap2/gamap_util/gamap.pro' )
 /home/bmy/IDL/gamap2/gamap_util/gamap.pro

With EXTRACT_FILENAME, you can extract just the filename part from a full file path:

 IDL> print, extract_filename( '/home/bmy/IDL/gamap2/gamap_util/gamap.pro' )
 gamap.pro

and with EXTRACT_PATH, you can extract just the directory part from a full file path:

 IDL> print, extract_path( '/home/bmy/IDL/gamap2/gamap_util/gamap.pro' )
 /home/bmy/IDL/gamap2/gamap_util/

Finally, with ADD_SEPARATOR, you can ensure that your directory name always ends in a separator.  For example:

 ; Make sure the directory has a separator character
 ; before we append the file name
 IDL> pwd, mydir                    
 IDL> print, mydir
 /home/bmy/IDL
 IDL> myfile = add_separator( mydir ) + 'myfile.pro'
 IDL> print, myfile
 /home/bmy/IDL/perl/myfile.pro

--[[User:Bmy|Bmy]] 16:12, 22 April 2008 (EDT)