ISO/IEC JTC1/SC22/WG15 RIN N092

Title: reorder_after and reorder_end keywords in LC_COLLATE

Source: DS

Date: 1992-10-14



The following section is inserted in the description of LC_COLLATE
keywords in POSIX.2 D11.3 section 2.5.5.2.

line 2037: change:
"no other keyword shall be present." to
"only the keywords 'reorder_after' and 'reorder_end' may be present."

Line 2047 add:

'reorder_after' 	  Specify a position in the previously
			  defined collation order, to be used as
			  a starting point for reordering collating
			  elements.
			  This statement is followed by one or more
			  collation reorder statements, reassigning
			  character collation weights to collating
			  elements.
			  This keyword is optional.

'reorder_end'		  Specify the end of the collation
			  reordering statements.
			  This keyword is required when a
			  'reorder_after' statement has been used.

line 2376 add:

2.5.2.2.6 'reorder_after' keyword

The 'reorder_after' keyword specifies a starting point for
reordering collating elements. It is followed by one or more
collation reorder statements, reassigning character
collation weights to collating elements. The syntax is:

	 "reorder_after %s\n",<collating-symbol>


2.5.2.2.6 Collation Reordering

Each 'reorder_after' statement shall be followed by one or
more collation element reordering entries. The definition of
collation element reordering entries are equivalent to the
collating element entries in 2.5.2.2.4, specifying collation elements
and associated weights. The collating element
reordring entries are terminated by a 'reorder_after' keyword
or a 'reorder_end' keyword.

Each collation element specified via a collation element reordering entry
is removed from the current collating sequence, if present,
and inserted in the collating sequence
after the previous reordering collation elements. The collating
element specified on the previous 'reorder_after' statement specifies
the first reordering collation element. The last reordering
collation element is followed by the follower to the collation
element specified on the 'replace-after' statement.

Example:

 order_start
 <collating-el1>
 <collating-el2>
 <collating-el3>
 <collating-el4>
 <collating-el5>
 order_end

 reorder_after <collating-el4>
 <collating-el1>
 <collating-el2>
 reorder_after <collating-el2>
 <collating-el6>
 reorder_end

The resulting order is then:
 <collating-el3>
 <collating-el4>
 <collating-el1>
 <collating-el2>
 <collating-el6>
 <collating-el5>


2.5.2.2.8 'reorder_end' keyword

The collating reorder entries shall be terminated with a
'reorder_end' keyword.


2.5.2.2.9.1 'reorder_after' rationale

Much work is done on locales, making them applicable to large
character sets and also to numerous character sets via a character
set independent specification.
WG15 RIN has on its programme of work to harmonize locales
as far as it is feasible.  The POSIX.2 Draft 11 introduced
a copy command for all sections of the locale. This is useful for
many purposes and it ensures that two locales are equivalent
for this category. The 'reorder_after' specification
faciliates a further refined way of building on previous work.

The collating sequences of characters vary a bit from country
to country, but generally much of the collating sequence is the same.
For example the Danish collating sequence is quite equal to the German,
English or French, but for about a dozen letters it differs.
The same can be said for Swedish or Spanish: generally the Latin
collating sequence is the same, but a few charcters are different.

With the advent of the quite general coded character set
independent locales like the example Danish in POSIX.2 draft 11
annex F, it would be convenient if the few differences could
be specified just as changes to an existing one. This also
improves the overview of what the changes really are.

2.5.2.2.9.2 'awk' script for 'reorder_after'

A script has been written in the 'awk' language defined in POSIX.2
to implement the 'reorder_after' construct.

   BEGIN	{
	comment = "%";
	back[0]= follow[0] = 0
	}

   /LC_COLLATE/ { coll=1 }

   /END LC_COLLATE/ {
	coll=0;
	for (lnr= 1; lnr; lnr= follow[lnr])
		print cont[lnr]
   }

   { if (coll == 0) print $0  ;
      else {
	if ($1 == "copy")   {
		file = $2
		while (getline < file )
		if ( $1 == "LC_COLLATE" ) copy_lc = 1
		else if ( $1 == "END"
		&&&& $2 == "LC_COLLATE" ) copy_lc =0
		else if (copy_lc) {
			lnr++
			follow[lnr-1] = lnr
			back [ lnr ] = lnr-1
			cont[lnr] = $0
			symb[ $1 ] = lnr
		}
		close (file )
	}
	else if ($1 == "reorder_after")
	{ ra=1 ; after = symb [ $2 ] }
	else if ($1 == "reorder_end") ra = 0
	else {
		lnr++
		if (ra) follow [ lnr ] = follow [ after ]
		if (ra) back [ follow [ after ] ] = lnr
		follow[after] = lnr
		back [ lnr ] = after
		cont[lnr] = $0
		if ( ra &&&& $1 != comment &&&& $1 != "" )  {
			old = symb [ $1 ]
			follow [ back [ old ] ] = follow [ old ]
			back [ follow [ old ] ] = back [ old ]
			symb[ $1 ] = lnr
		}
		after = lnr
	}
      }
   }