This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
mktables: change variable name for clarity
[perl5.git] / lib / unicore / mktables
CommitLineData
d73e5302 1#!/usr/bin/perl -w
99870f4d
KW
2
3# !!!!!!!!!!!!!! IF YOU MODIFY THIS FILE !!!!!!!!!!!!!!!!!!!!!!!!!
4# Any files created or read by this program should be listed in 'mktables.lst'
5# Use -makelist to regenerate it.
6
23e33b60
KW
7# Needs 'no overloading' to run faster on miniperl. Code commented out at the
8# subroutine objaddr can be used instead to work as far back (untested) as
f998e60c
KW
9# 5.8: needs pack "U". But almost all occurrences of objaddr have been
10# removed in favor of using 'no overloading'. You also would have to go
11# through and replace occurrences like:
ffe43484 12# my $addr = do { no overloading; pack 'J', $self; }
f998e60c
KW
13# with
14# my $addr = main::objaddr $self;
6c68572b 15# (or reverse commit 9b01bafde4b022706c3d6f947a0963f821b2e50b
051df77b
NC
16# that instituted the change to main::objaddr, and subsequent commits that
17# changed 0+$self to pack 'J', $self.)
6c68572b 18
cdcef19a
KW
19my $start_time;
20BEGIN { # Get the time the script started running; do it at compiliation to
21 # get it as close as possible
22 $start_time= time;
23}
24
25
23e33b60 26require 5.010_001;
d73e5302 27use strict;
99870f4d 28use warnings;
cf25bb62 29use Carp;
99870f4d
KW
30use File::Find;
31use File::Path;
d07a55ed 32use File::Spec;
99870f4d
KW
33use Text::Tabs;
34
35sub DEBUG () { 0 } # Set to 0 for production; 1 for development
36
37##########################################################################
38#
39# mktables -- create the runtime Perl Unicode files (lib/unicore/.../*.pl),
40# from the Unicode database files (lib/unicore/.../*.txt), It also generates
41# a pod file and a .t file
42#
43# The structure of this file is:
44# First these introductory comments; then
45# code needed for everywhere, such as debugging stuff; then
46# code to handle input parameters; then
47# data structures likely to be of external interest (some of which depend on
48# the input parameters, so follows them; then
49# more data structures and subroutine and package (class) definitions; then
50# the small actual loop to process the input files and finish up; then
51# a __DATA__ section, for the .t tests
52#
53# This program works on all releases of Unicode through at least 5.2. The
54# outputs have been scrutinized most intently for release 5.1. The others
55# have been checked for somewhat more than just sanity. It can handle all
56# existing Unicode character properties in those releases.
57#
99870f4d
KW
58# This program is mostly about Unicode character (or code point) properties.
59# A property describes some attribute or quality of a code point, like if it
60# is lowercase or not, its name, what version of Unicode it was first defined
61# in, or what its uppercase equivalent is. Unicode deals with these disparate
62# possibilities by making all properties into mappings from each code point
63# into some corresponding value. In the case of it being lowercase or not,
64# the mapping is either to 'Y' or 'N' (or various synonyms thereof). Each
65# property maps each Unicode code point to a single value, called a "property
66# value". (Hence each Unicode property is a true mathematical function with
67# exactly one value per code point.)
68#
69# When using a property in a regular expression, what is desired isn't the
70# mapping of the code point to its property's value, but the reverse (or the
71# mathematical "inverse relation"): starting with the property value, "Does a
72# code point map to it?" These are written in a "compound" form:
73# \p{property=value}, e.g., \p{category=punctuation}. This program generates
74# files containing the lists of code points that map to each such regular
75# expression property value, one file per list
76#
77# There is also a single form shortcut that Perl adds for many of the commonly
78# used properties. This happens for all binary properties, plus script,
79# general_category, and block properties.
80#
81# Thus the outputs of this program are files. There are map files, mostly in
82# the 'To' directory; and there are list files for use in regular expression
83# matching, all in subdirectories of the 'lib' directory, with each
84# subdirectory being named for the property that the lists in it are for.
85# Bookkeeping, test, and documentation files are also generated.
86
87my $matches_directory = 'lib'; # Where match (\p{}) files go.
88my $map_directory = 'To'; # Where map files go.
89
90# DATA STRUCTURES
91#
92# The major data structures of this program are Property, of course, but also
93# Table. There are two kinds of tables, very similar to each other.
94# "Match_Table" is the data structure giving the list of code points that have
95# a particular property value, mentioned above. There is also a "Map_Table"
96# data structure which gives the property's mapping from code point to value.
97# There are two structures because the match tables need to be combined in
98# various ways, such as constructing unions, intersections, complements, etc.,
99# and the map ones don't. And there would be problems, perhaps subtle, if
100# a map table were inadvertently operated on in some of those ways.
101# The use of separate classes with operations defined on one but not the other
102# prevents accidentally confusing the two.
103#
104# At the heart of each table's data structure is a "Range_List", which is just
105# an ordered list of "Ranges", plus ancillary information, and methods to
106# operate on them. A Range is a compact way to store property information.
107# Each range has a starting code point, an ending code point, and a value that
108# is meant to apply to all the code points between the two end points,
109# inclusive. For a map table, this value is the property value for those
110# code points. Two such ranges could be written like this:
111# 0x41 .. 0x5A, 'Upper',
112# 0x61 .. 0x7A, 'Lower'
113#
114# Each range also has a type used as a convenience to classify the values.
115# Most ranges in this program will be Type 0, or normal, but there are some
116# ranges that have a non-zero type. These are used only in map tables, and
117# are for mappings that don't fit into the normal scheme of things. Mappings
118# that require a hash entry to communicate with utf8.c are one example;
119# another example is mappings for charnames.pm to use which indicate a name
120# that is algorithmically determinable from its code point (and vice-versa).
121# These are used to significantly compact these tables, instead of listing
122# each one of the tens of thousands individually.
123#
124# In a match table, the value of a range is irrelevant (and hence the type as
125# well, which will always be 0), and arbitrarily set to the null string.
126# Using the example above, there would be two match tables for those two
127# entries, one named Upper would contain the 0x41..0x5A range, and the other
128# named Lower would contain 0x61..0x7A.
129#
130# Actually, there are two types of range lists, "Range_Map" is the one
131# associated with map tables, and "Range_List" with match tables.
132# Again, this is so that methods can be defined on one and not the other so as
133# to prevent operating on them in incorrect ways.
134#
135# Eventually, most tables are written out to files to be read by utf8_heavy.pl
136# in the perl core. All tables could in theory be written, but some are
137# suppressed because there is no current practical use for them. It is easy
138# to change which get written by changing various lists that are near the top
139# of the actual code in this file. The table data structures contain enough
140# ancillary information to allow them to be treated as separate entities for
141# writing, such as the path to each one's file. There is a heading in each
142# map table that gives the format of its entries, and what the map is for all
143# the code points missing from it. (This allows tables to be more compact.)
678f13d5 144#
99870f4d
KW
145# The Property data structure contains one or more tables. All properties
146# contain a map table (except the $perl property which is a
147# pseudo-property containing only match tables), and any properties that
148# are usable in regular expression matches also contain various matching
149# tables, one for each value the property can have. A binary property can
150# have two values, True and False (or Y and N, which are preferred by Unicode
151# terminology). Thus each of these properties will have a map table that
152# takes every code point and maps it to Y or N (but having ranges cuts the
153# number of entries in that table way down), and two match tables, one
154# which has a list of all the code points that map to Y, and one for all the
155# code points that map to N. (For each of these, a third table is also
156# generated for the pseudo Perl property. It contains the identical code
157# points as the Y table, but can be written, not in the compound form, but in
158# a "single" form like \p{IsUppercase}.) Many properties are binary, but some
159# properties have several possible values, some have many, and properties like
160# Name have a different value for every named code point. Those will not,
161# unless the controlling lists are changed, have their match tables written
162# out. But all the ones which can be used in regular expression \p{} and \P{}
163# constructs will. Generally a property will have either its map table or its
164# match tables written but not both. Again, what gets written is controlled
165# by lists which can easily be changed.
678f13d5 166#
99870f4d
KW
167# For information about the Unicode properties, see Unicode's UAX44 document:
168
169my $unicode_reference_url = 'http://www.unicode.org/reports/tr44/';
170
171# As stated earlier, this program will work on any release of Unicode so far.
172# Most obvious problems in earlier data have NOT been corrected except when
173# necessary to make Perl or this program work reasonably. For example, no
174# folding information was given in early releases, so this program uses the
175# substitute of lower case, just so that a regular expression with the /i
176# option will do something that actually gives the right results in many
177# cases. There are also a couple other corrections for version 1.1.5,
178# commented at the point they are made. As an example of corrections that
179# weren't made (but could be) is this statement from DerivedAge.txt: "The
180# supplementary private use code points and the non-character code points were
181# assigned in version 2.0, but not specifically listed in the UCD until
182# versions 3.0 and 3.1 respectively." (To be precise it was 3.0.1 not 3.0.0)
183# More information on Unicode version glitches is further down in these
184# introductory comments.
185#
186# This program works on all properties as of 5.2, though the files for some
678f13d5
KW
187# are suppressed from apparent lack of demand for them. You can change which
188# are output by changing lists in this program.
189#
99870f4d
KW
190# The old version of mktables emphasized the term "Fuzzy" to mean Unocde's
191# loose matchings rules (from Unicode TR18):
192#
193# The recommended names for UCD properties and property values are in
194# PropertyAliases.txt [Prop] and PropertyValueAliases.txt
195# [PropValue]. There are both abbreviated names and longer, more
196# descriptive names. It is strongly recommended that both names be
197# recognized, and that loose matching of property names be used,
198# whereby the case distinctions, whitespace, hyphens, and underbar
199# are ignored.
200# The program still allows Fuzzy to override its determination of if loose
201# matching should be used, but it isn't currently used, as it is no longer
202# needed; the calculations it makes are good enough.
678f13d5 203#
99870f4d
KW
204# SUMMARY OF HOW IT WORKS:
205#
206# Process arguments
207#
208# A list is constructed containing each input file that is to be processed
209#
210# Each file on the list is processed in a loop, using the associated handler
211# code for each:
212# The PropertyAliases.txt and PropValueAliases.txt files are processed
213# first. These files name the properties and property values.
214# Objects are created of all the property and property value names
215# that the rest of the input should expect, including all synonyms.
216# The other input files give mappings from properties to property
217# values. That is, they list code points and say what the mapping
218# is under the given property. Some files give the mappings for
219# just one property; and some for many. This program goes through
220# each file and populates the properties from them. Some properties
221# are listed in more than one file, and Unicode has set up a
222# precedence as to which has priority if there is a conflict. Thus
223# the order of processing matters, and this program handles the
224# conflict possibility by processing the overriding input files
225# last, so that if necessary they replace earlier values.
226# After this is all done, the program creates the property mappings not
227# furnished by Unicode, but derivable from what it does give.
228# The tables of code points that match each property value in each
229# property that is accessible by regular expressions are created.
230# The Perl-defined properties are created and populated. Many of these
231# require data determined from the earlier steps
232# Any Perl-defined synonyms are created, and name clashes between Perl
678f13d5 233# and Unicode are reconciled and warned about.
99870f4d
KW
234# All the properties are written to files
235# Any other files are written, and final warnings issued.
678f13d5 236#
99870f4d
KW
237# For clarity, a number of operators have been overloaded to work on tables:
238# ~ means invert (take all characters not in the set). The more
239# conventional '!' is not used because of the possibility of confusing
240# it with the actual boolean operation.
241# + means union
242# - means subtraction
243# & means intersection
244# The precedence of these is the order listed. Parentheses should be
245# copiously used. These are not a general scheme. The operations aren't
246# defined for a number of things, deliberately, to avoid getting into trouble.
247# Operations are done on references and affect the underlying structures, so
248# that the copy constructors for them have been overloaded to not return a new
249# clone, but the input object itself.
678f13d5 250#
99870f4d
KW
251# The bool operator is deliberately not overloaded to avoid confusion with
252# "should it mean if the object merely exists, or also is non-empty?".
99870f4d
KW
253#
254# WHY CERTAIN DESIGN DECISIONS WERE MADE
678f13d5
KW
255#
256# This program needs to be able to run under miniperl. Therefore, it uses a
257# minimum of other modules, and hence implements some things itself that could
258# be gotten from CPAN
259#
260# This program uses inputs published by the Unicode Consortium. These can
261# change incompatibly between releases without the Perl maintainers realizing
262# it. Therefore this program is now designed to try to flag these. It looks
263# at the directories where the inputs are, and flags any unrecognized files.
264# It keeps track of all the properties in the files it handles, and flags any
265# that it doesn't know how to handle. It also flags any input lines that
266# don't match the expected syntax, among other checks.
267#
268# It is also designed so if a new input file matches one of the known
269# templates, one hopefully just needs to add it to a list to have it
270# processed.
271#
272# As mentioned earlier, some properties are given in more than one file. In
273# particular, the files in the extracted directory are supposedly just
274# reformattings of the others. But they contain information not easily
275# derivable from the other files, including results for Unihan, which this
276# program doesn't ordinarily look at, and for unassigned code points. They
277# also have historically had errors or been incomplete. In an attempt to
278# create the best possible data, this program thus processes them first to
279# glean information missing from the other files; then processes those other
280# files to override any errors in the extracted ones. Much of the design was
281# driven by this need to store things and then possibly override them.
282#
283# It tries to keep fatal errors to a minimum, to generate something usable for
284# testing purposes. It always looks for files that could be inputs, and will
285# warn about any that it doesn't know how to handle (the -q option suppresses
286# the warning).
99870f4d
KW
287#
288# Why have files written out for binary 'N' matches?
289# For binary properties, if you know the mapping for either Y or N; the
678f13d5
KW
290# other is trivial to construct, so could be done at Perl run-time by just
291# complementing the result, instead of having a file for it. That is, if
292# someone types in \p{foo: N}, Perl could translate that to \P{foo: Y} and
293# not need a file. The problem is communicating to Perl that a given
294# property is binary. Perl can't figure it out from looking at the N (or
295# No), as some non-binary properties have these as property values. So
296# rather than inventing a way to communicate this info back to the core,
297# which would have required changes there as well, it was simpler just to
298# add the extra tables.
299#
300# Why is there more than one type of range?
301# This simplified things. There are some very specialized code points that
302# have to be handled specially for output, such as Hangul syllable names.
303# By creating a range type (done late in the development process), it
304# allowed this to be stored with the range, and overridden by other input.
305# Originally these were stored in another data structure, and it became a
306# mess trying to decide if a second file that was for the same property was
307# overriding the earlier one or not.
308#
309# Why are there two kinds of tables, match and map?
310# (And there is a base class shared by the two as well.) As stated above,
311# they actually are for different things. Development proceeded much more
312# smoothly when I (khw) realized the distinction. Map tables are used to
313# give the property value for every code point (actually every code point
314# that doesn't map to a default value). Match tables are used for regular
315# expression matches, and are essentially the inverse mapping. Separating
316# the two allows more specialized methods, and error checks so that one
317# can't just take the intersection of two map tables, for example, as that
318# is nonsensical.
99870f4d
KW
319#
320# There are no match tables generated for matches of the null string. These
c1739a4a 321# would look like qr/\p{JSN=}/ currently without modifying the regex code.
678f13d5
KW
322# Perhaps something like them could be added if necessary. The JSN does have
323# a real code point U+110B that maps to the null string, but it is a
324# contributory property, and therefore not output by default. And it's easily
325# handled so far by making the null string the default where it is a
326# possibility.
99870f4d 327#
23e33b60
KW
328# DEBUGGING
329#
678f13d5
KW
330# This program is written so it will run under miniperl. Occasionally changes
331# will cause an error where the backtrace doesn't work well under miniperl.
332# To diagnose the problem, you can instead run it under regular perl, if you
333# have one compiled.
334#
335# There is a good trace facility. To enable it, first sub DEBUG must be set
336# to return true. Then a line like
337#
338# local $to_trace = 1 if main::DEBUG;
339#
340# can be added to enable tracing in its lexical scope or until you insert
341# another line:
342#
343# local $to_trace = 0 if main::DEBUG;
344#
345# then use a line like "trace $a, @b, %c, ...;
346#
347# Some of the more complex subroutines already have trace statements in them.
348# Permanent trace statements should be like:
349#
350# trace ... if main::DEBUG && $to_trace;
351#
352# If there is just one or a few files that you're debugging, you can easily
353# cause most everything else to be skipped. Change the line
354#
355# my $debug_skip = 0;
356#
357# to 1, and every file whose object is in @input_file_objects and doesn't have
358# a, 'non_skip => 1,' in its constructor will be skipped.
359#
99870f4d
KW
360# FUTURE ISSUES
361#
362# The program would break if Unicode were to change its names so that
363# interior white space, underscores, or dashes differences were significant
364# within property and property value names.
365#
366# It might be easier to use the xml versions of the UCD if this program ever
367# would need heavy revision, and the ability to handle old versions was not
368# required.
369#
370# There is the potential for name collisions, in that Perl has chosen names
371# that Unicode could decide it also likes. There have been such collisions in
372# the past, with mostly Perl deciding to adopt the Unicode definition of the
373# name. However in the 5.2 Unicode beta testing, there were a number of such
374# collisions, which were withdrawn before the final release, because of Perl's
375# and other's protests. These all involved new properties which began with
376# 'Is'. Based on the protests, Unicode is unlikely to try that again. Also,
377# many of the Perl-defined synonyms, like Any, Word, etc, are listed in a
378# Unicode document, so they are unlikely to be used by Unicode for another
379# purpose. However, they might try something beginning with 'In', or use any
380# of the other Perl-defined properties. This program will warn you of name
381# collisions, and refuse to generate tables with them, but manual intervention
382# will be required in this event. One scheme that could be implemented, if
383# necessary, would be to have this program generate another file, or add a
384# field to mktables.lst that gives the date of first definition of a property.
385# Each new release of Unicode would use that file as a basis for the next
386# iteration. And the Perl synonym addition code could sort based on the age
387# of the property, so older properties get priority, and newer ones that clash
388# would be refused; hence existing code would not be impacted, and some other
389# synonym would have to be used for the new property. This is ugly, and
390# manual intervention would certainly be easier to do in the short run; lets
391# hope it never comes to this.
678f13d5 392#
99870f4d
KW
393# A NOTE ON UNIHAN
394#
395# This program can generate tables from the Unihan database. But it doesn't
396# by default, letting the CPAN module Unicode::Unihan handle them. Prior to
397# version 5.2, this database was in a single file, Unihan.txt. In 5.2 the
398# database was split into 8 different files, all beginning with the letters
399# 'Unihan'. This program will read those file(s) if present, but it needs to
400# know which of the many properties in the file(s) should have tables created
401# for them. It will create tables for any properties listed in
402# PropertyAliases.txt and PropValueAliases.txt, plus any listed in the
403# @cjk_properties array and the @cjk_property_values array. Thus, if a
404# property you want is not in those files of the release you are building
405# against, you must add it to those two arrays. Starting in 4.0, the
406# Unicode_Radical_Stroke was listed in those files, so if the Unihan database
407# is present in the directory, a table will be generated for that property.
408# In 5.2, several more properties were added. For your convenience, the two
409# arrays are initialized with all the 5.2 listed properties that are also in
410# earlier releases. But these are commented out. You can just uncomment the
411# ones you want, or use them as a template for adding entries for other
412# properties.
413#
414# You may need to adjust the entries to suit your purposes. setup_unihan(),
415# and filter_unihan_line() are the functions where this is done. This program
416# already does some adjusting to make the lines look more like the rest of the
417# Unicode DB; You can see what that is in filter_unihan_line()
418#
419# There is a bug in the 3.2 data file in which some values for the
420# kPrimaryNumeric property have commas and an unexpected comment. A filter
421# could be added for these; or for a particular installation, the Unihan.txt
422# file could be edited to fix them.
99870f4d 423#
678f13d5
KW
424# HOW TO ADD A FILE TO BE PROCESSED
425#
426# A new file from Unicode needs to have an object constructed for it in
427# @input_file_objects, probably at the end or at the end of the extracted
428# ones. The program should warn you if its name will clash with others on
429# restrictive file systems, like DOS. If so, figure out a better name, and
430# add lines to the README.perl file giving that. If the file is a character
431# property, it should be in the format that Unicode has by default
432# standardized for such files for the more recently introduced ones.
433# If so, the Input_file constructor for @input_file_objects can just be the
434# file name and release it first appeared in. If not, then it should be
435# possible to construct an each_line_handler() to massage the line into the
436# standardized form.
437#
438# For non-character properties, more code will be needed. You can look at
439# the existing entries for clues.
440#
441# UNICODE VERSIONS NOTES
442#
443# The Unicode UCD has had a number of errors in it over the versions. And
444# these remain, by policy, in the standard for that version. Therefore it is
445# risky to correct them, because code may be expecting the error. So this
446# program doesn't generally make changes, unless the error breaks the Perl
447# core. As an example, some versions of 2.1.x Jamo.txt have the wrong value
448# for U+1105, which causes real problems for the algorithms for Jamo
449# calculations, so it is changed here.
450#
451# But it isn't so clear cut as to what to do about concepts that are
452# introduced in a later release; should they extend back to earlier releases
453# where the concept just didn't exist? It was easier to do this than to not,
454# so that's what was done. For example, the default value for code points not
455# in the files for various properties was probably undefined until changed by
456# some version. No_Block for blocks is such an example. This program will
457# assign No_Block even in Unicode versions that didn't have it. This has the
458# benefit that code being written doesn't have to special case earlier
459# versions; and the detriment that it doesn't match the Standard precisely for
460# the affected versions.
461#
462# Here are some observations about some of the issues in early versions:
463#
464# The number of code points in \p{alpha} halve in 2.1.9. It turns out that
465# the reason is that the CJK block starting at 4E00 was removed from PropList,
466# and was not put back in until 3.1.0
467#
468# Unicode introduced the synonym Space for White_Space in 4.1. Perl has
469# always had a \p{Space}. In release 3.2 only, they are not synonymous. The
470# reason is that 3.2 introduced U+205F=medium math space, which was not
471# classed as white space, but Perl figured out that it should have been. 4.0
472# reclassified it correctly.
473#
474# Another change between 3.2 and 4.0 is the CCC property value ATBL. In 3.2
475# this was erroneously a synonym for 202. In 4.0, ATB became 202, and ATBL
476# was left with no code points, as all the ones that mapped to 202 stayed
477# mapped to 202. Thus if your program used the numeric name for the class,
478# it would not have been affected, but if it used the mnemonic, it would have
479# been.
480#
481# \p{Script=Hrkt} (Katakana_Or_Hiragana) came in 4.0.1. Before that code
482# points which eventually came to have this script property value, instead
483# mapped to "Unknown". But in the next release all these code points were
484# moved to \p{sc=common} instead.
99870f4d
KW
485#
486# The default for missing code points for BidiClass is complicated. Starting
487# in 3.1.1, the derived file DBidiClass.txt handles this, but this program
488# tries to do the best it can for earlier releases. It is done in
489# process_PropertyAliases()
490#
491##############################################################################
492
493my $UNDEF = ':UNDEF:'; # String to print out for undefined values in tracing
494 # and errors
495my $MAX_LINE_WIDTH = 78;
496
497# Debugging aid to skip most files so as to not be distracted by them when
498# concentrating on the ones being debugged. Add
499# non_skip => 1,
500# to the constructor for those files you want processed when you set this.
501# Files with a first version number of 0 are special: they are always
502# processed regardless of the state of this flag.
503my $debug_skip = 0;
504
505# Set to 1 to enable tracing.
506our $to_trace = 0;
507
508{ # Closure for trace: debugging aid
509 my $print_caller = 1; # ? Include calling subroutine name
510 my $main_with_colon = 'main::';
511 my $main_colon_length = length($main_with_colon);
512
513 sub trace {
514 return unless $to_trace; # Do nothing if global flag not set
515
516 my @input = @_;
517
518 local $DB::trace = 0;
519 $DB::trace = 0; # Quiet 'used only once' message
520
521 my $line_number;
522
523 # Loop looking up the stack to get the first non-trace caller
524 my $caller_line;
525 my $caller_name;
526 my $i = 0;
527 do {
528 $line_number = $caller_line;
529 (my $pkg, my $file, $caller_line, my $caller) = caller $i++;
530 $caller = $main_with_colon unless defined $caller;
531
532 $caller_name = $caller;
533
534 # get rid of pkg
535 $caller_name =~ s/.*:://;
536 if (substr($caller_name, 0, $main_colon_length)
537 eq $main_with_colon)
538 {
539 $caller_name = substr($caller_name, $main_colon_length);
540 }
541
542 } until ($caller_name ne 'trace');
543
544 # If the stack was empty, we were called from the top level
545 $caller_name = 'main' if ($caller_name eq ""
546 || $caller_name eq 'trace');
547
548 my $output = "";
549 foreach my $string (@input) {
550 #print STDERR __LINE__, ": ", join ", ", @input, "\n";
551 if (ref $string eq 'ARRAY' || ref $string eq 'HASH') {
552 $output .= simple_dumper($string);
553 }
554 else {
555 $string = "$string" if ref $string;
556 $string = $UNDEF unless defined $string;
557 chomp $string;
558 $string = '""' if $string eq "";
559 $output .= " " if $output ne ""
560 && $string ne ""
561 && substr($output, -1, 1) ne " "
562 && substr($string, 0, 1) ne " ";
563 $output .= $string;
564 }
565 }
566
99f78760
KW
567 print STDERR sprintf "%4d: ", $line_number if defined $line_number;
568 print STDERR "$caller_name: " if $print_caller;
99870f4d
KW
569 print STDERR $output, "\n";
570 return;
571 }
572}
573
574# This is for a rarely used development feature that allows you to compare two
575# versions of the Unicode standard without having to deal with changes caused
576# by the code points introduced in the later verson. Change the 0 to a SINGLE
577# dotted Unicode release number (e.g. 2.1). Only code points introduced in
578# that release and earlier will be used; later ones are thrown away. You use
579# the version number of the earliest one you want to compare; then run this
580# program on directory structures containing each release, and compare the
581# outputs. These outputs will therefore include only the code points common
582# to both releases, and you can see the changes caused just by the underlying
583# release semantic changes. For versions earlier than 3.2, you must copy a
584# version of DAge.txt into the directory.
585my $string_compare_versions = DEBUG && 0; # e.g., v2.1;
586my $compare_versions = DEBUG
587 && $string_compare_versions
588 && pack "C*", split /\./, $string_compare_versions;
589
590sub uniques {
591 # Returns non-duplicated input values. From "Perl Best Practices:
592 # Encapsulated Cleverness". p. 455 in first edition.
593
594 my %seen;
0e407844
NC
595 # Arguably this breaks encapsulation, if the goal is to permit multiple
596 # distinct objects to stringify to the same value, and be interchangeable.
597 # However, for this program, no two objects stringify identically, and all
598 # lists passed to this function are either objects or strings. So this
599 # doesn't affect correctness, but it does give a couple of percent speedup.
600 no overloading;
99870f4d
KW
601 return grep { ! $seen{$_}++ } @_;
602}
603
604$0 = File::Spec->canonpath($0);
605
606my $make_test_script = 0; # ? Should we output a test script
607my $write_unchanged_files = 0; # ? Should we update the output files even if
608 # we don't think they have changed
609my $use_directory = ""; # ? Should we chdir somewhere.
610my $pod_directory; # input directory to store the pod file.
611my $pod_file = 'perluniprops';
612my $t_path; # Path to the .t test file
613my $file_list = 'mktables.lst'; # File to store input and output file names.
614 # This is used to speed up the build, by not
615 # executing the main body of the program if
616 # nothing on the list has changed since the
617 # previous build
618my $make_list = 1; # ? Should we write $file_list. Set to always
619 # make a list so that when the pumpking is
620 # preparing a release, s/he won't have to do
621 # special things
622my $glob_list = 0; # ? Should we try to include unknown .txt files
623 # in the input.
624my $output_range_counts = 1; # ? Should we include the number of code points
625 # in ranges in the output
9ef2b94f
KW
626my $output_names = 0; # ? Should character names be in the output
627my @viacode; # Contains the 1 million character names, if
628 # $output_names is true
629
99870f4d
KW
630# Verbosity levels; 0 is quiet
631my $NORMAL_VERBOSITY = 1;
632my $PROGRESS = 2;
633my $VERBOSE = 3;
634
635my $verbosity = $NORMAL_VERBOSITY;
636
637# Process arguments
638while (@ARGV) {
cf25bb62
JH
639 my $arg = shift @ARGV;
640 if ($arg eq '-v') {
99870f4d
KW
641 $verbosity = $VERBOSE;
642 }
643 elsif ($arg eq '-p') {
644 $verbosity = $PROGRESS;
645 $| = 1; # Flush buffers as we go.
646 }
647 elsif ($arg eq '-q') {
648 $verbosity = 0;
649 }
650 elsif ($arg eq '-w') {
651 $write_unchanged_files = 1; # update the files even if havent changed
652 }
653 elsif ($arg eq '-check') {
6ae7e459
YO
654 my $this = shift @ARGV;
655 my $ok = shift @ARGV;
656 if ($this ne $ok) {
657 print "Skipping as check params are not the same.\n";
658 exit(0);
659 }
00a8df5c 660 }
99870f4d
KW
661 elsif ($arg eq '-P' && defined ($pod_directory = shift)) {
662 -d $pod_directory or croak "Directory '$pod_directory' doesn't exist";
663 }
3df51b85
KW
664 elsif ($arg eq '-maketest' || ($arg eq '-T' && defined ($t_path = shift)))
665 {
99870f4d 666 $make_test_script = 1;
99870f4d
KW
667 }
668 elsif ($arg eq '-makelist') {
669 $make_list = 1;
670 }
671 elsif ($arg eq '-C' && defined ($use_directory = shift)) {
672 -d $use_directory or croak "Unknown directory '$use_directory'";
673 }
674 elsif ($arg eq '-L') {
675
676 # Existence not tested until have chdir'd
677 $file_list = shift;
678 }
679 elsif ($arg eq '-globlist') {
680 $glob_list = 1;
681 }
682 elsif ($arg eq '-c') {
683 $output_range_counts = ! $output_range_counts
684 }
9ef2b94f
KW
685 elsif ($arg eq '-output_names') {
686 $output_names = 1;
687 }
99870f4d
KW
688 else {
689 my $with_c = 'with';
690 $with_c .= 'out' if $output_range_counts; # Complements the state
691 croak <<END;
692usage: $0 [-c|-p|-q|-v|-w] [-C dir] [-L filelist] [ -P pod_dir ]
693 [ -T test_file_path ] [-globlist] [-makelist] [-maketest]
694 [-check A B ]
695 -c : Output comments $with_c number of code points in ranges
696 -q : Quiet Mode: Only output serious warnings.
697 -p : Set verbosity level to normal plus show progress.
698 -v : Set Verbosity level high: Show progress and non-serious
699 warnings
700 -w : Write files regardless
701 -C dir : Change to this directory before proceeding. All relative paths
702 except those specified by the -P and -T options will be done
703 with respect to this directory.
704 -P dir : Output $pod_file file to directory 'dir'.
3df51b85 705 -T path : Create a test script as 'path'; overrides -maketest
99870f4d
KW
706 -L filelist : Use alternate 'filelist' instead of standard one
707 -globlist : Take as input all non-Test *.txt files in current and sub
708 directories
3df51b85
KW
709 -maketest : Make test script 'TestProp.pl' in current (or -C directory),
710 overrides -T
99870f4d 711 -makelist : Rewrite the file list $file_list based on current setup
9ef2b94f
KW
712 -output_names : Output each character's name in the table files; useful for
713 doing what-ifs, looking at diffs; is slow, memory intensive,
714 resulting tables are usable but very large.
99870f4d
KW
715 -check A B : Executes $0 only if A and B are the same
716END
717 }
718}
719
720# Stores the most-recently changed file. If none have changed, can skip the
721# build
aeab6150 722my $most_recent = (stat $0)[9]; # Do this before the chdir!
99870f4d
KW
723
724# Change directories now, because need to read 'version' early.
725if ($use_directory) {
3df51b85 726 if ($pod_directory && ! File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d
KW
727 $pod_directory = File::Spec->rel2abs($pod_directory);
728 }
3df51b85 729 if ($t_path && ! File::Spec->file_name_is_absolute($t_path)) {
99870f4d 730 $t_path = File::Spec->rel2abs($t_path);
00a8df5c 731 }
99870f4d 732 chdir $use_directory or croak "Failed to chdir to '$use_directory':$!";
3df51b85 733 if ($pod_directory && File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d 734 $pod_directory = File::Spec->abs2rel($pod_directory);
02b1aeec 735 }
3df51b85 736 if ($t_path && File::Spec->file_name_is_absolute($t_path)) {
99870f4d 737 $t_path = File::Spec->abs2rel($t_path);
02b1aeec 738 }
00a8df5c
YO
739}
740
99870f4d
KW
741# Get Unicode version into regular and v-string. This is done now because
742# various tables below get populated based on it. These tables are populated
743# here to be near the top of the file, and so easily seeable by those needing
744# to modify things.
745open my $VERSION, "<", "version"
746 or croak "$0: can't open required file 'version': $!\n";
747my $string_version = <$VERSION>;
748close $VERSION;
749chomp $string_version;
750my $v_version = pack "C*", split /\./, $string_version; # v string
751
752# The following are the complete names of properties with property values that
753# are known to not match any code points in some versions of Unicode, but that
754# may change in the future so they should be matchable, hence an empty file is
755# generated for them.
756my @tables_that_may_be_empty = (
757 'Joining_Type=Left_Joining',
758 );
759push @tables_that_may_be_empty, 'Script=Common' if $v_version le v4.0.1;
760push @tables_that_may_be_empty, 'Title' if $v_version lt v2.0.0;
761push @tables_that_may_be_empty, 'Script=Katakana_Or_Hiragana'
762 if $v_version ge v4.1.0;
763
764# The lists below are hashes, so the key is the item in the list, and the
765# value is the reason why it is in the list. This makes generation of
766# documentation easier.
767
768my %why_suppressed; # No file generated for these.
769
770# Files aren't generated for empty extraneous properties. This is arguable.
771# Extraneous properties generally come about because a property is no longer
772# used in a newer version of Unicode. If we generated a file without code
773# points, programs that used to work on that property will still execute
774# without errors. It just won't ever match (or will always match, with \P{}).
775# This means that the logic is now likely wrong. I (khw) think its better to
776# find this out by getting an error message. Just move them to the table
777# above to change this behavior
778my %why_suppress_if_empty_warn_if_not = (
779
780 # It is the only property that has ever officially been removed from the
781 # Standard. The database never contained any code points for it.
782 'Special_Case_Condition' => 'Obsolete',
783
784 # Apparently never official, but there were code points in some versions of
785 # old-style PropList.txt
786 'Non_Break' => 'Obsolete',
787);
788
789# These would normally go in the warn table just above, but they were changed
790# a long time before this program was written, so warnings about them are
791# moot.
792if ($v_version gt v3.2.0) {
793 push @tables_that_may_be_empty,
794 'Canonical_Combining_Class=Attached_Below_Left'
795}
796
797# These are listed in the Property aliases file in 5.2, but Unihan is ignored
798# unless explicitly added.
799if ($v_version ge v5.2.0) {
800 my $unihan = 'Unihan; remove from list if using Unihan';
ea25a9b2 801 foreach my $table (qw (
99870f4d
KW
802 kAccountingNumeric
803 kOtherNumeric
804 kPrimaryNumeric
805 kCompatibilityVariant
806 kIICore
807 kIRG_GSource
808 kIRG_HSource
809 kIRG_JSource
810 kIRG_KPSource
811 kIRG_MSource
812 kIRG_KSource
813 kIRG_TSource
814 kIRG_USource
815 kIRG_VSource
816 kRSUnicode
ea25a9b2 817 ))
99870f4d
KW
818 {
819 $why_suppress_if_empty_warn_if_not{$table} = $unihan;
820 }
ca12659b
NC
821}
822
99870f4d
KW
823# Properties that this program ignores.
824my @unimplemented_properties = (
825'Unicode_Radical_Stroke' # Remove if changing to handle this one.
826);
d73e5302 827
99870f4d
KW
828# There are several types of obsolete properties defined by Unicode. These
829# must be hand-edited for every new Unicode release.
830my %why_deprecated; # Generates a deprecated warning message if used.
831my %why_stabilized; # Documentation only
832my %why_obsolete; # Documentation only
833
834{ # Closure
835 my $simple = 'Perl uses the more complete version of this property';
836 my $unihan = 'Unihan properties are by default not enabled in the Perl core. Instead use CPAN: Unicode::Unihan';
837
838 my $other_properties = 'other properties';
839 my $contributory = "Used by Unicode internally for generating $other_properties and not intended to be used stand-alone";
840 my $why_no_expand = "Easily computed, and yet doesn't cover the common encoding forms (UTF-16/8)",
841
842 %why_deprecated = (
843 'Grapheme_Link' => 'Deprecated by Unicode. Use ccc=vr (Canonical_Combining_Class=Virama) instead',
844 'Jamo_Short_Name' => $contributory,
845 'Line_Break=Surrogate' => 'Deprecated by Unicode because surrogates should never appear in well-formed text, and therefore shouldn\'t be the basis for line breaking',
846 'Other_Alphabetic' => $contributory,
847 'Other_Default_Ignorable_Code_Point' => $contributory,
848 'Other_Grapheme_Extend' => $contributory,
849 'Other_ID_Continue' => $contributory,
850 'Other_ID_Start' => $contributory,
851 'Other_Lowercase' => $contributory,
852 'Other_Math' => $contributory,
853 'Other_Uppercase' => $contributory,
854 );
855
856 %why_suppressed = (
857 # There is a lib/unicore/Decomposition.pl (used by normalize.pm) which
858 # contains the same information, but without the algorithmically
859 # determinable Hangul syllables'. This file is not published, so it's
860 # existence is not noted in the comment.
861 'Decomposition_Mapping' => 'Accessible via Unicode::Normalize',
862
863 'ISO_Comment' => 'Apparently no demand for it, but can access it through Unicode::UCD::charinfo. Obsoleted, and code points for it removed in Unicode 5.2',
864 'Unicode_1_Name' => "$simple, and no apparent demand for it, but can access it through Unicode::UCD::charinfo. If there is no later name for a code point, then this one is used instead in charnames",
865
866 'Simple_Case_Folding' => "$simple. Can access this through Unicode::UCD::casefold",
867 'Simple_Lowercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo",
868 'Simple_Titlecase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo",
869 'Simple_Uppercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo",
870
871 'Name' => "Accessible via 'use charnames;'",
872 'Name_Alias' => "Accessible via 'use charnames;'",
873
874 # These are sort of jumping the gun; deprecation is proposed for
875 # Unicode version 6.0, but they have never been exposed by Perl, and
876 # likely are soon to be deprecated, so best not to expose them.
877 FC_NFKC_Closure => 'Use NFKC_Casefold instead',
878 Expands_On_NFC => $why_no_expand,
879 Expands_On_NFD => $why_no_expand,
880 Expands_On_NFKC => $why_no_expand,
881 Expands_On_NFKD => $why_no_expand,
882 );
883
884 # The following are suppressed because they were made contributory or
885 # deprecated by Unicode before Perl ever thought about supporting them.
886 foreach my $property ('Jamo_Short_Name', 'Grapheme_Link') {
887 $why_suppressed{$property} = $why_deprecated{$property};
888 }
cf25bb62 889
99870f4d
KW
890 # Customize the message for all the 'Other_' properties
891 foreach my $property (keys %why_deprecated) {
892 next if (my $main_property = $property) !~ s/^Other_//;
893 $why_deprecated{$property} =~ s/$other_properties/the $main_property property (which should be used instead)/;
894 }
895}
896
897if ($v_version ge 4.0.0) {
898 $why_stabilized{'Hyphen'} = 'Use the Line_Break property instead; see www.unicode.org/reports/tr14';
899}
900if ($v_version ge 5.2.0) {
901 $why_obsolete{'ISO_Comment'} = 'Code points for it have been removed';
902}
903
904# Probably obsolete forever
905if ($v_version ge v4.1.0) {
906 $why_suppressed{'Script=Katakana_Or_Hiragana'} = 'Obsolete. All code points previously matched by this have been moved to "Script=Common"';
907}
908
909# This program can create files for enumerated-like properties, such as
910# 'Numeric_Type'. This file would be the same format as for a string
911# property, with a mapping from code point to its value, so you could look up,
912# for example, the script a code point is in. But no one so far wants this
913# mapping, or they have found another way to get it since this is a new
914# feature. So no file is generated except if it is in this list.
915my @output_mapped_properties = split "\n", <<END;
916END
917
918# If you are using the Unihan database, you need to add the properties that
919# you want to extract from it to this table. For your convenience, the
920# properties in the 5.2 PropertyAliases.txt file are listed, commented out
921my @cjk_properties = split "\n", <<'END';
922#cjkAccountingNumeric; kAccountingNumeric
923#cjkOtherNumeric; kOtherNumeric
924#cjkPrimaryNumeric; kPrimaryNumeric
925#cjkCompatibilityVariant; kCompatibilityVariant
926#cjkIICore ; kIICore
927#cjkIRG_GSource; kIRG_GSource
928#cjkIRG_HSource; kIRG_HSource
929#cjkIRG_JSource; kIRG_JSource
930#cjkIRG_KPSource; kIRG_KPSource
931#cjkIRG_KSource; kIRG_KSource
932#cjkIRG_TSource; kIRG_TSource
933#cjkIRG_USource; kIRG_USource
934#cjkIRG_VSource; kIRG_VSource
935#cjkRSUnicode; kRSUnicode ; Unicode_Radical_Stroke; URS
936END
937
938# Similarly for the property values. For your convenience, the lines in the
939# 5.2 PropertyAliases.txt file are listed. Just remove the first BUT NOT both
940# '#' marks
941my @cjk_property_values = split "\n", <<'END';
942## @missing: 0000..10FFFF; cjkAccountingNumeric; NaN
943## @missing: 0000..10FFFF; cjkCompatibilityVariant; <code point>
944## @missing: 0000..10FFFF; cjkIICore; <none>
945## @missing: 0000..10FFFF; cjkIRG_GSource; <none>
946## @missing: 0000..10FFFF; cjkIRG_HSource; <none>
947## @missing: 0000..10FFFF; cjkIRG_JSource; <none>
948## @missing: 0000..10FFFF; cjkIRG_KPSource; <none>
949## @missing: 0000..10FFFF; cjkIRG_KSource; <none>
950## @missing: 0000..10FFFF; cjkIRG_TSource; <none>
951## @missing: 0000..10FFFF; cjkIRG_USource; <none>
952## @missing: 0000..10FFFF; cjkIRG_VSource; <none>
953## @missing: 0000..10FFFF; cjkOtherNumeric; NaN
954## @missing: 0000..10FFFF; cjkPrimaryNumeric; NaN
955## @missing: 0000..10FFFF; cjkRSUnicode; <none>
956END
957
958# The input files don't list every code point. Those not listed are to be
959# defaulted to some value. Below are hard-coded what those values are for
960# non-binary properties as of 5.1. Starting in 5.0, there are
961# machine-parsable comment lines in the files the give the defaults; so this
962# list shouldn't have to be extended. The claim is that all missing entries
963# for binary properties will default to 'N'. Unicode tried to change that in
964# 5.2, but the beta period produced enough protest that they backed off.
965#
966# The defaults for the fields that appear in UnicodeData.txt in this hash must
967# be in the form that it expects. The others may be synonyms.
968my $CODE_POINT = '<code point>';
969my %default_mapping = (
970 Age => "Unassigned",
971 # Bidi_Class => Complicated; set in code
972 Bidi_Mirroring_Glyph => "",
973 Block => 'No_Block',
974 Canonical_Combining_Class => 0,
975 Case_Folding => $CODE_POINT,
976 Decomposition_Mapping => $CODE_POINT,
977 Decomposition_Type => 'None',
978 East_Asian_Width => "Neutral",
979 FC_NFKC_Closure => $CODE_POINT,
980 General_Category => 'Cn',
981 Grapheme_Cluster_Break => 'Other',
982 Hangul_Syllable_Type => 'NA',
983 ISO_Comment => "",
984 Jamo_Short_Name => "",
985 Joining_Group => "No_Joining_Group",
986 # Joining_Type => Complicated; set in code
987 kIICore => 'N', # Is converted to binary
988 #Line_Break => Complicated; set in code
989 Lowercase_Mapping => $CODE_POINT,
990 Name => "",
991 Name_Alias => "",
992 NFC_QC => 'Yes',
993 NFD_QC => 'Yes',
994 NFKC_QC => 'Yes',
995 NFKD_QC => 'Yes',
996 Numeric_Type => 'None',
997 Numeric_Value => 'NaN',
998 Script => ($v_version le 4.1.0) ? 'Common' : 'Unknown',
999 Sentence_Break => 'Other',
1000 Simple_Case_Folding => $CODE_POINT,
1001 Simple_Lowercase_Mapping => $CODE_POINT,
1002 Simple_Titlecase_Mapping => $CODE_POINT,
1003 Simple_Uppercase_Mapping => $CODE_POINT,
1004 Titlecase_Mapping => $CODE_POINT,
1005 Unicode_1_Name => "",
1006 Unicode_Radical_Stroke => "",
1007 Uppercase_Mapping => $CODE_POINT,
1008 Word_Break => 'Other',
1009);
1010
1011# Below are files that Unicode furnishes, but this program ignores, and why
1012my %ignored_files = (
1013 'CJKRadicals.txt' => 'Unihan data',
1014 'Index.txt' => 'An index, not actual data',
1015 'NamedSqProv.txt' => 'Not officially part of the Unicode standard; Append it to NamedSequences.txt if you want to process the contents.',
1016 'NamesList.txt' => 'Just adds commentary',
1017 'NormalizationCorrections.txt' => 'Data is already in other files.',
1018 'Props.txt' => 'Adds nothing to PropList.txt; only in very early releases',
1019 'ReadMe.txt' => 'Just comments',
1020 'README.TXT' => 'Just comments',
1021 'StandardizedVariants.txt' => 'Only for glyph changes, not a Unicode character property. Does not fit into current scheme where one code point is mapped',
1022);
1023
678f13d5 1024### End of externally interesting definitions, except for @input_file_objects
99870f4d
KW
1025
1026my $HEADER=<<"EOF";
1027# !!!!!!! DO NOT EDIT THIS FILE !!!!!!!
3df51b85
KW
1028# This file is machine-generated by $0 from the Unicode
1029# database, Version $string_version. Any changes made here will be lost!
cf25bb62
JH
1030EOF
1031
b6922eda 1032my $INTERNAL_ONLY=<<"EOF";
99870f4d
KW
1033
1034# !!!!!!! INTERNAL PERL USE ONLY !!!!!!!
b6922eda 1035# This file is for internal use by the Perl program only. The format and even
99870f4d
KW
1036# the name or existence of this file are subject to change without notice.
1037# Don't use it directly.
1038EOF
1039
1040my $DEVELOPMENT_ONLY=<<"EOF";
1041# !!!!!!! DEVELOPMENT USE ONLY !!!!!!!
1042# This file contains information artificially constrained to code points
1043# present in Unicode release $string_compare_versions.
1044# IT CANNOT BE RELIED ON. It is for use during development only and should
23e33b60 1045# not be used for production.
b6922eda
KW
1046
1047EOF
1048
99870f4d
KW
1049my $LAST_UNICODE_CODEPOINT_STRING = "10FFFF";
1050my $LAST_UNICODE_CODEPOINT = hex $LAST_UNICODE_CODEPOINT_STRING;
1051my $MAX_UNICODE_CODEPOINTS = $LAST_UNICODE_CODEPOINT + 1;
1052
1053# Matches legal code point. 4-6 hex numbers, If there are 6, the first
1054# two must be 10; if there are 5, the first must not be a 0. Written this way
1055# to decrease backtracking
1056my $code_point_re =
1057 qr/ \b (?: 10[0-9A-F]{4} | [1-9A-F][0-9A-F]{4} | [0-9A-F]{4} ) \b/x;
1058
1059# This matches the beginning of the line in the Unicode db files that give the
1060# defaults for code points not listed (i.e., missing) in the file. The code
1061# depends on this ending with a semi-colon, so it can assume it is a valid
1062# field when the line is split() by semi-colons
1063my $missing_defaults_prefix =
1064 qr/^#\s+\@missing:\s+0000\.\.$LAST_UNICODE_CODEPOINT_STRING\s*;/;
1065
1066# Property types. Unicode has more types, but these are sufficient for our
1067# purposes.
1068my $UNKNOWN = -1; # initialized to illegal value
1069my $NON_STRING = 1; # Either binary or enum
1070my $BINARY = 2;
1071my $ENUM = 3; # Include catalog
1072my $STRING = 4; # Anything else: string or misc
1073
1074# Some input files have lines that give default values for code points not
1075# contained in the file. Sometimes these should be ignored.
1076my $NO_DEFAULTS = 0; # Must evaluate to false
1077my $NOT_IGNORED = 1;
1078my $IGNORED = 2;
1079
1080# Range types. Each range has a type. Most ranges are type 0, for normal,
1081# and will appear in the main body of the tables in the output files, but
1082# there are other types of ranges as well, listed below, that are specially
1083# handled. There are pseudo-types as well that will never be stored as a
1084# type, but will affect the calculation of the type.
1085
1086# 0 is for normal, non-specials
1087my $MULTI_CP = 1; # Sequence of more than code point
1088my $HANGUL_SYLLABLE = 2;
1089my $CP_IN_NAME = 3; # The NAME contains the code point appended to it.
1090my $NULL = 4; # The map is to the null string; utf8.c can't
1091 # handle these, nor is there an accepted syntax
1092 # for them in \p{} constructs
f86864ac 1093my $COMPUTE_NO_MULTI_CP = 5; # Pseudo-type; means that ranges that would
99870f4d
KW
1094 # otherwise be $MULTI_CP type are instead type 0
1095
1096# process_generic_property_file() can accept certain overrides in its input.
1097# Each of these must begin AND end with $CMD_DELIM.
1098my $CMD_DELIM = "\a";
1099my $REPLACE_CMD = 'replace'; # Override the Replace
1100my $MAP_TYPE_CMD = 'map_type'; # Override the Type
1101
1102my $NO = 0;
1103my $YES = 1;
1104
1105# Values for the Replace argument to add_range.
1106# $NO # Don't replace; add only the code points not
1107 # already present.
1108my $IF_NOT_EQUIVALENT = 1; # Replace only under certain conditions; details in
1109 # the comments at the subroutine definition.
1110my $UNCONDITIONALLY = 2; # Replace without conditions.
1111my $MULTIPLE = 4; # Don't replace, but add a duplicate record if
1112 # already there
1113
1114# Flags to give property statuses. The phrases are to remind maintainers that
1115# if the flag is changed, the indefinite article referring to it in the
1116# documentation may need to be as well.
1117my $NORMAL = "";
1118my $SUPPRESSED = 'z'; # The character should never actually be seen, since
1119 # it is suppressed
37e2e78e 1120my $PLACEHOLDER = 'P'; # Implies no pod entry generated
99870f4d
KW
1121my $DEPRECATED = 'D';
1122my $a_bold_deprecated = "a 'B<$DEPRECATED>'";
1123my $A_bold_deprecated = "A 'B<$DEPRECATED>'";
1124my $DISCOURAGED = 'X';
1125my $a_bold_discouraged = "an 'B<$DISCOURAGED>'";
1126my $A_bold_discouraged = "An 'B<$DISCOURAGED>'";
1127my $STRICTER = 'T';
1128my $a_bold_stricter = "a 'B<$STRICTER>'";
1129my $A_bold_stricter = "A 'B<$STRICTER>'";
1130my $STABILIZED = 'S';
1131my $a_bold_stabilized = "an 'B<$STABILIZED>'";
1132my $A_bold_stabilized = "An 'B<$STABILIZED>'";
1133my $OBSOLETE = 'O';
1134my $a_bold_obsolete = "an 'B<$OBSOLETE>'";
1135my $A_bold_obsolete = "An 'B<$OBSOLETE>'";
1136
1137my %status_past_participles = (
1138 $DISCOURAGED => 'discouraged',
1139 $SUPPRESSED => 'should never be generated',
1140 $STABILIZED => 'stabilized',
1141 $OBSOLETE => 'obsolete',
37e2e78e 1142 $DEPRECATED => 'deprecated',
99870f4d
KW
1143);
1144
1145# The format of the values of the map tables:
1146my $BINARY_FORMAT = 'b';
1147my $DECIMAL_FORMAT = 'd';
1148my $FLOAT_FORMAT = 'f';
1149my $INTEGER_FORMAT = 'i';
1150my $HEX_FORMAT = 'x';
1151my $RATIONAL_FORMAT = 'r';
1152my $STRING_FORMAT = 's';
1153
1154my %map_table_formats = (
1155 $BINARY_FORMAT => 'binary',
1156 $DECIMAL_FORMAT => 'single decimal digit',
1157 $FLOAT_FORMAT => 'floating point number',
1158 $INTEGER_FORMAT => 'integer',
1159 $HEX_FORMAT => 'positive hex whole number; a code point',
1160 $RATIONAL_FORMAT => 'rational: an integer or a fraction',
1a9d544b 1161 $STRING_FORMAT => 'string',
99870f4d
KW
1162);
1163
1164# Unicode didn't put such derived files in a separate directory at first.
1165my $EXTRACTED_DIR = (-d 'extracted') ? 'extracted' : "";
1166my $EXTRACTED = ($EXTRACTED_DIR) ? "$EXTRACTED_DIR/" : "";
1167my $AUXILIARY = 'auxiliary';
1168
1169# Hashes that will eventually go into Heavy.pl for the use of utf8_heavy.pl
1170my %loose_to_file_of; # loosely maps table names to their respective
1171 # files
1172my %stricter_to_file_of; # same; but for stricter mapping.
1173my %nv_floating_to_rational; # maps numeric values floating point numbers to
1174 # their rational equivalent
1175my %loose_property_name_of; # Loosely maps property names to standard form
1176
1177# These constants names and values were taken from the Unicode standard,
1178# version 5.1, section 3.12. They are used in conjunction with Hangul
6e5a209b
KW
1179# syllables. The '_string' versions are so generated tables can retain the
1180# hex format, which is the more familiar value
1181my $SBase_string = "0xAC00";
1182my $SBase = CORE::hex $SBase_string;
1183my $LBase_string = "0x1100";
1184my $LBase = CORE::hex $LBase_string;
1185my $VBase_string = "0x1161";
1186my $VBase = CORE::hex $VBase_string;
1187my $TBase_string = "0x11A7";
1188my $TBase = CORE::hex $TBase_string;
99870f4d
KW
1189my $SCount = 11172;
1190my $LCount = 19;
1191my $VCount = 21;
1192my $TCount = 28;
1193my $NCount = $VCount * $TCount;
1194
1195# For Hangul syllables; These store the numbers from Jamo.txt in conjunction
1196# with the above published constants.
1197my %Jamo;
1198my %Jamo_L; # Leading consonants
1199my %Jamo_V; # Vowels
1200my %Jamo_T; # Trailing consonants
1201
37e2e78e 1202my @backslash_X_tests; # List of tests read in for testing \X
99870f4d
KW
1203my @unhandled_properties; # Will contain a list of properties found in
1204 # the input that we didn't process.
f86864ac 1205my @match_properties; # Properties that have match tables, to be
99870f4d
KW
1206 # listed in the pod
1207my @map_properties; # Properties that get map files written
1208my @named_sequences; # NamedSequences.txt contents.
1209my %potential_files; # Generated list of all .txt files in the directory
1210 # structure so we can warn if something is being
1211 # ignored.
1212my @files_actually_output; # List of files we generated.
1213my @more_Names; # Some code point names are compound; this is used
1214 # to store the extra components of them.
1215my $MIN_FRACTION_LENGTH = 3; # How many digits of a floating point number at
1216 # the minimum before we consider it equivalent to a
1217 # candidate rational
1218my $MAX_FLOATING_SLOP = 10 ** - $MIN_FRACTION_LENGTH; # And in floating terms
1219
1220# These store references to certain commonly used property objects
1221my $gc;
1222my $perl;
1223my $block;
3e20195b
KW
1224my $perl_charname;
1225my $print;
99870f4d
KW
1226
1227# Are there conflicting names because of beginning with 'In_', or 'Is_'
1228my $has_In_conflicts = 0;
1229my $has_Is_conflicts = 0;
1230
1231sub internal_file_to_platform ($) {
1232 # Convert our file paths which have '/' separators to those of the
1233 # platform.
1234
1235 my $file = shift;
1236 return undef unless defined $file;
1237
1238 return File::Spec->join(split '/', $file);
d07a55ed 1239}
5beb625e 1240
99870f4d
KW
1241sub file_exists ($) { # platform independent '-e'. This program internally
1242 # uses slash as a path separator.
1243 my $file = shift;
1244 return 0 if ! defined $file;
1245 return -e internal_file_to_platform($file);
1246}
5beb625e 1247
99870f4d 1248sub objaddr($) {
23e33b60
KW
1249 # Returns the address of the blessed input object.
1250 # It doesn't check for blessedness because that would do a string eval
1251 # every call, and the program is structured so that this is never called
1252 # for a non-blessed object.
99870f4d 1253
23e33b60 1254 no overloading; # If overloaded, numifying below won't work.
99870f4d
KW
1255
1256 # Numifying a ref gives its address.
051df77b 1257 return pack 'J', $_[0];
99870f4d
KW
1258}
1259
23e33b60
KW
1260# Commented code below should work on Perl 5.8.
1261## This 'require' doesn't necessarily work in miniperl, and even if it does,
1262## the native perl version of it (which is what would operate under miniperl)
1263## is extremely slow, as it does a string eval every call.
1264#my $has_fast_scalar_util = $\18 !~ /miniperl/
1265# && defined eval "require Scalar::Util";
1266#
1267#sub objaddr($) {
1268# # Returns the address of the blessed input object. Uses the XS version if
1269# # available. It doesn't check for blessedness because that would do a
1270# # string eval every call, and the program is structured so that this is
1271# # never called for a non-blessed object.
1272#
1273# return Scalar::Util::refaddr($_[0]) if $has_fast_scalar_util;
1274#
1275# # Check at least that is a ref.
1276# my $pkg = ref($_[0]) or return undef;
1277#
1278# # Change to a fake package to defeat any overloaded stringify
1279# bless $_[0], 'main::Fake';
1280#
1281# # Numifying a ref gives its address.
051df77b 1282# my $addr = pack 'J', $_[0];
23e33b60
KW
1283#
1284# # Return to original class
1285# bless $_[0], $pkg;
1286# return $addr;
1287#}
1288
99870f4d
KW
1289sub max ($$) {
1290 my $a = shift;
1291 my $b = shift;
1292 return $a if $a >= $b;
1293 return $b;
1294}
1295
1296sub min ($$) {
1297 my $a = shift;
1298 my $b = shift;
1299 return $a if $a <= $b;
1300 return $b;
1301}
1302
1303sub clarify_number ($) {
1304 # This returns the input number with underscores inserted every 3 digits
1305 # in large (5 digits or more) numbers. Input must be entirely digits, not
1306 # checked.
1307
1308 my $number = shift;
1309 my $pos = length($number) - 3;
1310 return $number if $pos <= 1;
1311 while ($pos > 0) {
1312 substr($number, $pos, 0) = '_';
1313 $pos -= 3;
5beb625e 1314 }
99870f4d 1315 return $number;
99598c8c
JH
1316}
1317
12ac2576 1318
99870f4d 1319package Carp;
7ebf06b3 1320
99870f4d
KW
1321# These routines give a uniform treatment of messages in this program. They
1322# are placed in the Carp package to cause the stack trace to not include them,
1323# although an alternative would be to use another package and set @CARP_NOT
1324# for it.
12ac2576 1325
99870f4d 1326our $Verbose = 1 if main::DEBUG; # Useful info when debugging
12ac2576 1327
99f78760
KW
1328# This is a work-around suggested by Nicholas Clark to fix a problem with Carp
1329# and overload trying to load Scalar:Util under miniperl. See
1330# http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2009-11/msg01057.html
1331undef $overload::VERSION;
1332
99870f4d
KW
1333sub my_carp {
1334 my $message = shift || "";
1335 my $nofold = shift || 0;
7ebf06b3 1336
99870f4d
KW
1337 if ($message) {
1338 $message = main::join_lines($message);
1339 $message =~ s/^$0: *//; # Remove initial program name
1340 $message =~ s/[.;,]+$//; # Remove certain ending punctuation
1341 $message = "\n$0: $message;";
12ac2576 1342
99870f4d
KW
1343 # Fold the message with program name, semi-colon end punctuation
1344 # (which looks good with the message that carp appends to it), and a
1345 # hanging indent for continuation lines.
1346 $message = main::simple_fold($message, "", 4) unless $nofold;
1347 $message =~ s/\n$//; # Remove the trailing nl so what carp
1348 # appends is to the same line
1349 }
12ac2576 1350
99870f4d 1351 return $message if defined wantarray; # If a caller just wants the msg
12ac2576 1352
99870f4d
KW
1353 carp $message;
1354 return;
1355}
7ebf06b3 1356
99870f4d
KW
1357sub my_carp_bug {
1358 # This is called when it is clear that the problem is caused by a bug in
1359 # this program.
7ebf06b3 1360
99870f4d
KW
1361 my $message = shift;
1362 $message =~ s/^$0: *//;
1363 $message = my_carp("Bug in $0. Please report it by running perlbug or if that is unavailable, by sending email to perbug\@perl.org:\n$message");
1364 carp $message;
1365 return;
1366}
7ebf06b3 1367
99870f4d
KW
1368sub carp_too_few_args {
1369 if (@_ != 2) {
1370 my_carp_bug("Wrong number of arguments: to 'carp_too_few_arguments'. No action taken.");
1371 return;
12ac2576 1372 }
7ebf06b3 1373
99870f4d
KW
1374 my $args_ref = shift;
1375 my $count = shift;
7ebf06b3 1376
99870f4d
KW
1377 my_carp_bug("Need at least $count arguments to "
1378 . (caller 1)[3]
1379 . ". Instead got: '"
1380 . join ', ', @$args_ref
1381 . "'. No action taken.");
1382 return;
12ac2576
JP
1383}
1384
99870f4d
KW
1385sub carp_extra_args {
1386 my $args_ref = shift;
1387 my_carp_bug("Too many arguments to 'carp_extra_args': (" . join(', ', @_) . "); Extras ignored.") if @_;
12ac2576 1388
99870f4d
KW
1389 unless (ref $args_ref) {
1390 my_carp_bug("Argument to 'carp_extra_args' ($args_ref) must be a ref. Not checking arguments.");
1391 return;
1392 }
1393 my ($package, $file, $line) = caller;
1394 my $subroutine = (caller 1)[3];
cf25bb62 1395
99870f4d
KW
1396 my $list;
1397 if (ref $args_ref eq 'HASH') {
1398 foreach my $key (keys %$args_ref) {
1399 $args_ref->{$key} = $UNDEF unless defined $args_ref->{$key};
cf25bb62 1400 }
99870f4d 1401 $list = join ', ', each %{$args_ref};
cf25bb62 1402 }
99870f4d
KW
1403 elsif (ref $args_ref eq 'ARRAY') {
1404 foreach my $arg (@$args_ref) {
1405 $arg = $UNDEF unless defined $arg;
1406 }
1407 $list = join ', ', @$args_ref;
1408 }
1409 else {
1410 my_carp_bug("Can't cope with ref "
1411 . ref($args_ref)
1412 . " . argument to 'carp_extra_args'. Not checking arguments.");
1413 return;
1414 }
1415
1416 my_carp_bug("Unrecognized parameters in options: '$list' to $subroutine. Skipped.");
1417 return;
d73e5302
JH
1418}
1419
99870f4d
KW
1420package main;
1421
1422{ # Closure
1423
1424 # This program uses the inside-out method for objects, as recommended in
1425 # "Perl Best Practices". This closure aids in generating those. There
1426 # are two routines. setup_package() is called once per package to set
1427 # things up, and then set_access() is called for each hash representing a
1428 # field in the object. These routines arrange for the object to be
1429 # properly destroyed when no longer used, and for standard accessor
1430 # functions to be generated. If you need more complex accessors, just
1431 # write your own and leave those accesses out of the call to set_access().
1432 # More details below.
1433
1434 my %constructor_fields; # fields that are to be used in constructors; see
1435 # below
1436
1437 # The values of this hash will be the package names as keys to other
1438 # hashes containing the name of each field in the package as keys, and
1439 # references to their respective hashes as values.
1440 my %package_fields;
1441
1442 sub setup_package {
1443 # Sets up the package, creating standard DESTROY and dump methods
1444 # (unless already defined). The dump method is used in debugging by
1445 # simple_dumper().
1446 # The optional parameters are:
1447 # a) a reference to a hash, that gets populated by later
1448 # set_access() calls with one of the accesses being
1449 # 'constructor'. The caller can then refer to this, but it is
1450 # not otherwise used by these two routines.
1451 # b) a reference to a callback routine to call during destruction
1452 # of the object, before any fields are actually destroyed
1453
1454 my %args = @_;
1455 my $constructor_ref = delete $args{'Constructor_Fields'};
1456 my $destroy_callback = delete $args{'Destroy_Callback'};
1457 Carp::carp_extra_args(\@_) if main::DEBUG && %args;
1458
1459 my %fields;
1460 my $package = (caller)[0];
1461
1462 $package_fields{$package} = \%fields;
1463 $constructor_fields{$package} = $constructor_ref;
1464
1465 unless ($package->can('DESTROY')) {
1466 my $destroy_name = "${package}::DESTROY";
1467 no strict "refs";
1468
1469 # Use typeglob to give the anonymous subroutine the name we want
1470 *$destroy_name = sub {
1471 my $self = shift;
ffe43484 1472 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1473
1474 $self->$destroy_callback if $destroy_callback;
1475 foreach my $field (keys %{$package_fields{$package}}) {
1476 #print STDERR __LINE__, ": Destroying ", ref $self, " ", sprintf("%04X", $addr), ": ", $field, "\n";
1477 delete $package_fields{$package}{$field}{$addr};
1478 }
1479 return;
1480 }
1481 }
1482
1483 unless ($package->can('dump')) {
1484 my $dump_name = "${package}::dump";
1485 no strict "refs";
1486 *$dump_name = sub {
1487 my $self = shift;
1488 return dump_inside_out($self, $package_fields{$package}, @_);
1489 }
1490 }
1491 return;
1492 }
1493
1494 sub set_access {
1495 # Arrange for the input field to be garbage collected when no longer
1496 # needed. Also, creates standard accessor functions for the field
1497 # based on the optional parameters-- none if none of these parameters:
1498 # 'addable' creates an 'add_NAME()' accessor function.
1499 # 'readable' or 'readable_array' creates a 'NAME()' accessor
1500 # function.
1501 # 'settable' creates a 'set_NAME()' accessor function.
1502 # 'constructor' doesn't create an accessor function, but adds the
1503 # field to the hash that was previously passed to
1504 # setup_package();
1505 # Any of the accesses can be abbreviated down, so that 'a', 'ad',
1506 # 'add' etc. all mean 'addable'.
1507 # The read accessor function will work on both array and scalar
1508 # values. If another accessor in the parameter list is 'a', the read
1509 # access assumes an array. You can also force it to be array access
1510 # by specifying 'readable_array' instead of 'readable'
1511 #
1512 # A sort-of 'protected' access can be set-up by preceding the addable,
1513 # readable or settable with some initial portion of 'protected_' (but,
1514 # the underscore is required), like 'p_a', 'pro_set', etc. The
1515 # "protection" is only by convention. All that happens is that the
1516 # accessor functions' names begin with an underscore. So instead of
1517 # calling set_foo, the call is _set_foo. (Real protection could be
c1739a4a 1518 # accomplished by having a new subroutine, end_package, called at the
99870f4d
KW
1519 # end of each package, and then storing the __LINE__ ranges and
1520 # checking them on every accessor. But that is way overkill.)
1521
1522 # We create anonymous subroutines as the accessors and then use
1523 # typeglobs to assign them to the proper package and name
1524
1525 my $name = shift; # Name of the field
1526 my $field = shift; # Reference to the inside-out hash containing the
1527 # field
1528
1529 my $package = (caller)[0];
1530
1531 if (! exists $package_fields{$package}) {
1532 croak "$0: Must call 'setup_package' before 'set_access'";
1533 }
d73e5302 1534
99870f4d
KW
1535 # Stash the field so DESTROY can get it.
1536 $package_fields{$package}{$name} = $field;
cf25bb62 1537
99870f4d
KW
1538 # Remaining arguments are the accessors. For each...
1539 foreach my $access (@_) {
1540 my $access = lc $access;
cf25bb62 1541
99870f4d 1542 my $protected = "";
cf25bb62 1543
99870f4d
KW
1544 # Match the input as far as it goes.
1545 if ($access =~ /^(p[^_]*)_/) {
1546 $protected = $1;
1547 if (substr('protected_', 0, length $protected)
1548 eq $protected)
1549 {
1550
1551 # Add 1 for the underscore not included in $protected
1552 $access = substr($access, length($protected) + 1);
1553 $protected = '_';
1554 }
1555 else {
1556 $protected = "";
1557 }
1558 }
1559
1560 if (substr('addable', 0, length $access) eq $access) {
1561 my $subname = "${package}::${protected}add_$name";
1562 no strict "refs";
1563
1564 # add_ accessor. Don't add if already there, which we
1565 # determine using 'eq' for scalars and '==' otherwise.
1566 *$subname = sub {
1567 use strict "refs";
1568 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
1569 my $self = shift;
1570 my $value = shift;
ffe43484 1571 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1572 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1573 if (ref $value) {
f998e60c 1574 return if grep { $value == $_ } @{$field->{$addr}};
99870f4d
KW
1575 }
1576 else {
f998e60c 1577 return if grep { $value eq $_ } @{$field->{$addr}};
99870f4d 1578 }
f998e60c 1579 push @{$field->{$addr}}, $value;
99870f4d
KW
1580 return;
1581 }
1582 }
1583 elsif (substr('constructor', 0, length $access) eq $access) {
1584 if ($protected) {
1585 Carp::my_carp_bug("Can't set-up 'protected' constructors")
1586 }
1587 else {
1588 $constructor_fields{$package}{$name} = $field;
1589 }
1590 }
1591 elsif (substr('readable_array', 0, length $access) eq $access) {
1592
1593 # Here has read access. If one of the other parameters for
1594 # access is array, or this one specifies array (by being more
1595 # than just 'readable_'), then create a subroutine that
1596 # assumes the data is an array. Otherwise just a scalar
1597 my $subname = "${package}::${protected}$name";
1598 if (grep { /^a/i } @_
1599 or length($access) > length('readable_'))
1600 {
1601 no strict "refs";
1602 *$subname = sub {
1603 use strict "refs";
23e33b60 1604 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
ffe43484 1605 my $addr = do { no overloading; pack 'J', $_[0]; };
99870f4d
KW
1606 if (ref $field->{$addr} ne 'ARRAY') {
1607 my $type = ref $field->{$addr};
1608 $type = 'scalar' unless $type;
1609 Carp::my_carp_bug("Trying to read $name as an array when it is a $type. Big problems.");
1610 return;
1611 }
1612 return scalar @{$field->{$addr}} unless wantarray;
1613
1614 # Make a copy; had problems with caller modifying the
1615 # original otherwise
1616 my @return = @{$field->{$addr}};
1617 return @return;
1618 }
1619 }
1620 else {
1621
1622 # Here not an array value, a simpler function.
1623 no strict "refs";
1624 *$subname = sub {
1625 use strict "refs";
23e33b60 1626 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
f998e60c 1627 no overloading;
051df77b 1628 return $field->{pack 'J', $_[0]};
99870f4d
KW
1629 }
1630 }
1631 }
1632 elsif (substr('settable', 0, length $access) eq $access) {
1633 my $subname = "${package}::${protected}set_$name";
1634 no strict "refs";
1635 *$subname = sub {
1636 use strict "refs";
23e33b60
KW
1637 if (main::DEBUG) {
1638 return Carp::carp_too_few_args(\@_, 2) if @_ < 2;
1639 Carp::carp_extra_args(\@_) if @_ > 2;
1640 }
1641 # $self is $_[0]; $value is $_[1]
f998e60c 1642 no overloading;
051df77b 1643 $field->{pack 'J', $_[0]} = $_[1];
99870f4d
KW
1644 return;
1645 }
1646 }
1647 else {
1648 Carp::my_carp_bug("Unknown accessor type $access. No accessor set.");
1649 }
cf25bb62 1650 }
99870f4d 1651 return;
cf25bb62 1652 }
99870f4d
KW
1653}
1654
1655package Input_file;
1656
1657# All input files use this object, which stores various attributes about them,
1658# and provides for convenient, uniform handling. The run method wraps the
1659# processing. It handles all the bookkeeping of opening, reading, and closing
1660# the file, returning only significant input lines.
1661#
1662# Each object gets a handler which processes the body of the file, and is
1663# called by run(). Most should use the generic, default handler, which has
1664# code scrubbed to handle things you might not expect. A handler should
1665# basically be a while(next_line()) {...} loop.
1666#
1667# You can also set up handlers to
1668# 1) call before the first line is read for pre processing
1669# 2) call to adjust each line of the input before the main handler gets them
1670# 3) call upon EOF before the main handler exits its loop
1671# 4) call at the end for post processing
1672#
1673# $_ is used to store the input line, and is to be filtered by the
1674# each_line_handler()s. So, if the format of the line is not in the desired
1675# format for the main handler, these are used to do that adjusting. They can
1676# be stacked (by enclosing them in an [ anonymous array ] in the constructor,
1677# so the $_ output of one is used as the input to the next. None of the other
1678# handlers are stackable, but could easily be changed to be so.
1679#
1680# Most of the handlers can call insert_lines() or insert_adjusted_lines()
1681# which insert the parameters as lines to be processed before the next input
1682# file line is read. This allows the EOF handler to flush buffers, for
1683# example. The difference between the two routines is that the lines inserted
1684# by insert_lines() are subjected to the each_line_handler()s. (So if you
1685# called it from such a handler, you would get infinite recursion.) Lines
1686# inserted by insert_adjusted_lines() go directly to the main handler without
1687# any adjustments. If the post-processing handler calls any of these, there
1688# will be no effect. Some error checking for these conditions could be added,
1689# but it hasn't been done.
1690#
1691# carp_bad_line() should be called to warn of bad input lines, which clears $_
1692# to prevent further processing of the line. This routine will output the
1693# message as a warning once, and then keep a count of the lines that have the
1694# same message, and output that count at the end of the file's processing.
1695# This keeps the number of messages down to a manageable amount.
1696#
1697# get_missings() should be called to retrieve any @missing input lines.
1698# Messages will be raised if this isn't done if the options aren't to ignore
1699# missings.
1700
1701sub trace { return main::trace(@_); }
1702
99870f4d
KW
1703{ # Closure
1704 # Keep track of fields that are to be put into the constructor.
1705 my %constructor_fields;
1706
1707 main::setup_package(Constructor_Fields => \%constructor_fields);
1708
1709 my %file; # Input file name, required
1710 main::set_access('file', \%file, qw{ c r });
1711
1712 my %first_released; # Unicode version file was first released in, required
1713 main::set_access('first_released', \%first_released, qw{ c r });
1714
1715 my %handler; # Subroutine to process the input file, defaults to
1716 # 'process_generic_property_file'
1717 main::set_access('handler', \%handler, qw{ c });
1718
1719 my %property;
1720 # name of property this file is for. defaults to none, meaning not
1721 # applicable, or is otherwise determinable, for example, from each line.
1722 main::set_access('property', \%property, qw{ c });
1723
1724 my %optional;
1725 # If this is true, the file is optional. If not present, no warning is
1726 # output. If it is present, the string given by this parameter is
1727 # evaluated, and if false the file is not processed.
1728 main::set_access('optional', \%optional, 'c', 'r');
1729
1730 my %non_skip;
1731 # This is used for debugging, to skip processing of all but a few input
1732 # files. Add 'non_skip => 1' to the constructor for those files you want
1733 # processed when you set the $debug_skip global.
1734 main::set_access('non_skip', \%non_skip, 'c');
1735
37e2e78e
KW
1736 my %skip;
1737 # This is used to skip processing of this input file semi-permanently.
1738 # It is used for files that we aren't planning to process anytime soon,
1739 # but want to allow to be in the directory and not raise a message that we
1740 # are not handling. Mostly for test files. This is in contrast to the
1741 # non_skip element, which is supposed to be used very temporarily for
1742 # debugging. Sets 'optional' to 1
1743 main::set_access('skip', \%skip, 'c');
1744
99870f4d
KW
1745 my %each_line_handler;
1746 # list of subroutines to look at and filter each non-comment line in the
1747 # file. defaults to none. The subroutines are called in order, each is
1748 # to adjust $_ for the next one, and the final one adjusts it for
1749 # 'handler'
1750 main::set_access('each_line_handler', \%each_line_handler, 'c');
1751
1752 my %has_missings_defaults;
1753 # ? Are there lines in the file giving default values for code points
1754 # missing from it?. Defaults to NO_DEFAULTS. Otherwise NOT_IGNORED is
1755 # the norm, but IGNORED means it has such lines, but the handler doesn't
1756 # use them. Having these three states allows us to catch changes to the
1757 # UCD that this program should track
1758 main::set_access('has_missings_defaults',
1759 \%has_missings_defaults, qw{ c r });
1760
1761 my %pre_handler;
1762 # Subroutine to call before doing anything else in the file. If undef, no
1763 # such handler is called.
1764 main::set_access('pre_handler', \%pre_handler, qw{ c });
1765
1766 my %eof_handler;
1767 # Subroutine to call upon getting an EOF on the input file, but before
1768 # that is returned to the main handler. This is to allow buffers to be
1769 # flushed. The handler is expected to call insert_lines() or
1770 # insert_adjusted() with the buffered material
1771 main::set_access('eof_handler', \%eof_handler, qw{ c r });
1772
1773 my %post_handler;
1774 # Subroutine to call after all the lines of the file are read in and
1775 # processed. If undef, no such handler is called.
1776 main::set_access('post_handler', \%post_handler, qw{ c });
1777
1778 my %progress_message;
1779 # Message to print to display progress in lieu of the standard one
1780 main::set_access('progress_message', \%progress_message, qw{ c });
1781
1782 my %handle;
1783 # cache open file handle, internal. Is undef if file hasn't been
1784 # processed at all, empty if has;
1785 main::set_access('handle', \%handle);
1786
1787 my %added_lines;
1788 # cache of lines added virtually to the file, internal
1789 main::set_access('added_lines', \%added_lines);
1790
1791 my %errors;
1792 # cache of errors found, internal
1793 main::set_access('errors', \%errors);
1794
1795 my %missings;
1796 # storage of '@missing' defaults lines
1797 main::set_access('missings', \%missings);
1798
1799 sub new {
1800 my $class = shift;
1801
1802 my $self = bless \do{ my $anonymous_scalar }, $class;
ffe43484 1803 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1804
1805 # Set defaults
1806 $handler{$addr} = \&main::process_generic_property_file;
1807 $non_skip{$addr} = 0;
37e2e78e 1808 $skip{$addr} = 0;
99870f4d
KW
1809 $has_missings_defaults{$addr} = $NO_DEFAULTS;
1810 $handle{$addr} = undef;
1811 $added_lines{$addr} = [ ];
1812 $each_line_handler{$addr} = [ ];
1813 $errors{$addr} = { };
1814 $missings{$addr} = [ ];
1815
1816 # Two positional parameters.
99f78760 1817 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
99870f4d
KW
1818 $file{$addr} = main::internal_file_to_platform(shift);
1819 $first_released{$addr} = shift;
1820
1821 # The rest of the arguments are key => value pairs
1822 # %constructor_fields has been set up earlier to list all possible
1823 # ones. Either set or push, depending on how the default has been set
1824 # up just above.
1825 my %args = @_;
1826 foreach my $key (keys %args) {
1827 my $argument = $args{$key};
1828
1829 # Note that the fields are the lower case of the constructor keys
1830 my $hash = $constructor_fields{lc $key};
1831 if (! defined $hash) {
1832 Carp::my_carp_bug("Unrecognized parameters '$key => $argument' to new() for $self. Skipped");
1833 next;
1834 }
1835 if (ref $hash->{$addr} eq 'ARRAY') {
1836 if (ref $argument eq 'ARRAY') {
1837 foreach my $argument (@{$argument}) {
1838 next if ! defined $argument;
1839 push @{$hash->{$addr}}, $argument;
1840 }
1841 }
1842 else {
1843 push @{$hash->{$addr}}, $argument if defined $argument;
1844 }
1845 }
1846 else {
1847 $hash->{$addr} = $argument;
1848 }
1849 delete $args{$key};
1850 };
1851
1852 # If the file has a property for it, it means that the property is not
1853 # listed in the file's entries. So add a handler to the list of line
1854 # handlers to insert the property name into the lines, to provide a
1855 # uniform interface to the final processing subroutine.
1856 # the final code doesn't have to worry about that.
1857 if ($property{$addr}) {
1858 push @{$each_line_handler{$addr}}, \&_insert_property_into_line;
1859 }
1860
1861 if ($non_skip{$addr} && ! $debug_skip && $verbosity) {
1862 print "Warning: " . __PACKAGE__ . " constructor for $file{$addr} has useless 'non_skip' in it\n";
a3a8c5f0 1863 }
99870f4d 1864
37e2e78e
KW
1865 $optional{$addr} = 1 if $skip{$addr};
1866
99870f4d 1867 return $self;
d73e5302
JH
1868 }
1869
cf25bb62 1870
99870f4d
KW
1871 use overload
1872 fallback => 0,
1873 qw("") => "_operator_stringify",
1874 "." => \&main::_operator_dot,
1875 ;
cf25bb62 1876
99870f4d
KW
1877 sub _operator_stringify {
1878 my $self = shift;
cf25bb62 1879
99870f4d 1880 return __PACKAGE__ . " object for " . $self->file;
d73e5302 1881 }
d73e5302 1882
99870f4d
KW
1883 # flag to make sure extracted files are processed early
1884 my $seen_non_extracted_non_age = 0;
d73e5302 1885
99870f4d
KW
1886 sub run {
1887 # Process the input object $self. This opens and closes the file and
1888 # calls all the handlers for it. Currently, this can only be called
1889 # once per file, as it destroy's the EOF handler
d73e5302 1890
99870f4d
KW
1891 my $self = shift;
1892 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
b6922eda 1893
ffe43484 1894 my $addr = do { no overloading; pack 'J', $self; };
b6922eda 1895
99870f4d 1896 my $file = $file{$addr};
d73e5302 1897
99870f4d
KW
1898 # Don't process if not expecting this file (because released later
1899 # than this Unicode version), and isn't there. This means if someone
1900 # copies it into an earlier version's directory, we will go ahead and
1901 # process it.
1902 return if $first_released{$addr} gt $v_version && ! -e $file;
1903
1904 # If in debugging mode and this file doesn't have the non-skip
1905 # flag set, and isn't one of the critical files, skip it.
1906 if ($debug_skip
1907 && $first_released{$addr} ne v0
1908 && ! $non_skip{$addr})
1909 {
1910 print "Skipping $file in debugging\n" if $verbosity;
1911 return;
1912 }
1913
1914 # File could be optional
37e2e78e 1915 if ($optional{$addr}) {
99870f4d
KW
1916 return unless -e $file;
1917 my $result = eval $optional{$addr};
1918 if (! defined $result) {
1919 Carp::my_carp_bug("Got '$@' when tried to eval $optional{$addr}. $file Skipped.");
1920 return;
1921 }
1922 if (! $result) {
1923 if ($verbosity) {
1924 print STDERR "Skipping processing input file '$file' because '$optional{$addr}' is not true\n";
1925 }
1926 return;
1927 }
1928 }
1929
1930 if (! defined $file || ! -e $file) {
1931
1932 # If the file doesn't exist, see if have internal data for it
1933 # (based on first_released being 0).
1934 if ($first_released{$addr} eq v0) {
1935 $handle{$addr} = 'pretend_is_open';
1936 }
1937 else {
1938 if (! $optional{$addr} # File could be optional
1939 && $v_version ge $first_released{$addr})
1940 {
1941 print STDERR "Skipping processing input file '$file' because not found\n" if $v_version ge $first_released{$addr};
1942 }
1943 return;
1944 }
1945 }
1946 else {
1947
37e2e78e
KW
1948 # Here, the file exists. Some platforms may change the case of
1949 # its name
99870f4d 1950 if ($seen_non_extracted_non_age) {
517956bf 1951 if ($file =~ /$EXTRACTED/i) {
99870f4d 1952 Carp::my_carp_bug(join_lines(<<END
99f78760 1953$file should be processed just after the 'Prop...Alias' files, and before
99870f4d
KW
1954anything not in the $EXTRACTED_DIR directory. Proceeding, but the results may
1955have subtle problems
1956END
1957 ));
1958 }
1959 }
1960 elsif ($EXTRACTED_DIR
1961 && $first_released{$addr} ne v0
517956bf
CB
1962 && $file !~ /$EXTRACTED/i
1963 && lc($file) ne 'dage.txt')
99870f4d
KW
1964 {
1965 # We don't set this (by the 'if' above) if we have no
1966 # extracted directory, so if running on an early version,
1967 # this test won't work. Not worth worrying about.
1968 $seen_non_extracted_non_age = 1;
1969 }
1970
1971 # And mark the file as having being processed, and warn if it
1972 # isn't a file we are expecting. As we process the files,
1973 # they are deleted from the hash, so any that remain at the
1974 # end of the program are files that we didn't process.
517956bf
CB
1975 my $fkey = File::Spec->rel2abs($file);
1976 my $expecting = delete $potential_files{$fkey};
1977 $expecting = delete $potential_files{lc($fkey)} unless defined $expecting;
678f13d5
KW
1978 Carp::my_carp("Was not expecting '$file'.") if
1979 ! $expecting
99870f4d
KW
1980 && ! defined $handle{$addr};
1981
37e2e78e
KW
1982 # Having deleted from expected files, we can quit if not to do
1983 # anything. Don't print progress unless really want verbosity
1984 if ($skip{$addr}) {
1985 print "Skipping $file.\n" if $verbosity >= $VERBOSE;
1986 return;
1987 }
1988
99870f4d
KW
1989 # Open the file, converting the slashes used in this program
1990 # into the proper form for the OS
1991 my $file_handle;
1992 if (not open $file_handle, "<", $file) {
1993 Carp::my_carp("Can't open $file. Skipping: $!");
1994 return 0;
1995 }
1996 $handle{$addr} = $file_handle; # Cache the open file handle
1997 }
1998
1999 if ($verbosity >= $PROGRESS) {
2000 if ($progress_message{$addr}) {
2001 print "$progress_message{$addr}\n";
2002 }
2003 else {
2004 # If using a virtual file, say so.
2005 print "Processing ", (-e $file)
2006 ? $file
2007 : "substitute $file",
2008 "\n";
2009 }
2010 }
2011
2012
2013 # Call any special handler for before the file.
2014 &{$pre_handler{$addr}}($self) if $pre_handler{$addr};
2015
2016 # Then the main handler
2017 &{$handler{$addr}}($self);
2018
2019 # Then any special post-file handler.
2020 &{$post_handler{$addr}}($self) if $post_handler{$addr};
2021
2022 # If any errors have been accumulated, output the counts (as the first
2023 # error message in each class was output when it was encountered).
2024 if ($errors{$addr}) {
2025 my $total = 0;
2026 my $types = 0;
2027 foreach my $error (keys %{$errors{$addr}}) {
2028 $total += $errors{$addr}->{$error};
2029 delete $errors{$addr}->{$error};
2030 $types++;
2031 }
2032 if ($total > 1) {
2033 my $message
2034 = "A total of $total lines had errors in $file. ";
2035
2036 $message .= ($types == 1)
2037 ? '(Only the first one was displayed.)'
2038 : '(Only the first of each type was displayed.)';
2039 Carp::my_carp($message);
2040 }
2041 }
2042
2043 if (@{$missings{$addr}}) {
2044 Carp::my_carp_bug("Handler for $file didn't look at all the \@missing lines. Generated tables likely are wrong");
2045 }
2046
2047 # If a real file handle, close it.
2048 close $handle{$addr} or Carp::my_carp("Can't close $file: $!") if
2049 ref $handle{$addr};
2050 $handle{$addr} = ""; # Uses empty to indicate that has already seen
2051 # the file, as opposed to undef
2052 return;
2053 }
2054
2055 sub next_line {
2056 # Sets $_ to be the next logical input line, if any. Returns non-zero
2057 # if such a line exists. 'logical' means that any lines that have
2058 # been added via insert_lines() will be returned in $_ before the file
2059 # is read again.
2060
2061 my $self = shift;
2062 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2063
ffe43484 2064 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2065
2066 # Here the file is open (or if the handle is not a ref, is an open
2067 # 'virtual' file). Get the next line; any inserted lines get priority
2068 # over the file itself.
2069 my $adjusted;
2070
2071 LINE:
2072 while (1) { # Loop until find non-comment, non-empty line
2073 #local $to_trace = 1 if main::DEBUG;
2074 my $inserted_ref = shift @{$added_lines{$addr}};
2075 if (defined $inserted_ref) {
2076 ($adjusted, $_) = @{$inserted_ref};
2077 trace $adjusted, $_ if main::DEBUG && $to_trace;
2078 return 1 if $adjusted;
2079 }
2080 else {
2081 last if ! ref $handle{$addr}; # Don't read unless is real file
2082 last if ! defined ($_ = readline $handle{$addr});
2083 }
2084 chomp;
2085 trace $_ if main::DEBUG && $to_trace;
2086
2087 # See if this line is the comment line that defines what property
2088 # value that code points that are not listed in the file should
2089 # have. The format or existence of these lines is not guaranteed
2090 # by Unicode since they are comments, but the documentation says
2091 # that this was added for machine-readability, so probably won't
2092 # change. This works starting in Unicode Version 5.0. They look
2093 # like:
2094 #
2095 # @missing: 0000..10FFFF; Not_Reordered
2096 # @missing: 0000..10FFFF; Decomposition_Mapping; <code point>
2097 # @missing: 0000..10FFFF; ; NaN
2098 #
2099 # Save the line for a later get_missings() call.
2100 if (/$missing_defaults_prefix/) {
2101 if ($has_missings_defaults{$addr} == $NO_DEFAULTS) {
2102 $self->carp_bad_line("Unexpected \@missing line. Assuming no missing entries");
2103 }
2104 elsif ($has_missings_defaults{$addr} == $NOT_IGNORED) {
2105 my @defaults = split /\s* ; \s*/x, $_;
2106
2107 # The first field is the @missing, which ends in a
2108 # semi-colon, so can safely shift.
2109 shift @defaults;
2110
2111 # Some of these lines may have empty field placeholders
2112 # which get in the way. An example is:
2113 # @missing: 0000..10FFFF; ; NaN
2114 # Remove them. Process starting from the top so the
2115 # splice doesn't affect things still to be looked at.
2116 for (my $i = @defaults - 1; $i >= 0; $i--) {
2117 next if $defaults[$i] ne "";
2118 splice @defaults, $i, 1;
2119 }
2120
2121 # What's left should be just the property (maybe) and the
2122 # default. Having only one element means it doesn't have
2123 # the property.
2124 my $default;
2125 my $property;
2126 if (@defaults >= 1) {
2127 if (@defaults == 1) {
2128 $default = $defaults[0];
2129 }
2130 else {
2131 $property = $defaults[0];
2132 $default = $defaults[1];
2133 }
2134 }
2135
2136 if (@defaults < 1
2137 || @defaults > 2
2138 || ($default =~ /^</
2139 && $default !~ /^<code *point>$/i
2140 && $default !~ /^<none>$/i))
2141 {
2142 $self->carp_bad_line("Unrecognized \@missing line: $_. Assuming no missing entries");
2143 }
2144 else {
2145
2146 # If the property is missing from the line, it should
2147 # be the one for the whole file
2148 $property = $property{$addr} if ! defined $property;
2149
2150 # Change <none> to the null string, which is what it
2151 # really means. If the default is the code point
2152 # itself, set it to <code point>, which is what
2153 # Unicode uses (but sometimes they've forgotten the
2154 # space)
2155 if ($default =~ /^<none>$/i) {
2156 $default = "";
2157 }
2158 elsif ($default =~ /^<code *point>$/i) {
2159 $default = $CODE_POINT;
2160 }
2161
2162 # Store them as a sub-arrays with both components.
2163 push @{$missings{$addr}}, [ $default, $property ];
2164 }
2165 }
2166
2167 # There is nothing for the caller to process on this comment
2168 # line.
2169 next;
2170 }
2171
2172 # Remove comments and trailing space, and skip this line if the
2173 # result is empty
2174 s/#.*//;
2175 s/\s+$//;
2176 next if /^$/;
2177
2178 # Call any handlers for this line, and skip further processing of
2179 # the line if the handler sets the line to null.
2180 foreach my $sub_ref (@{$each_line_handler{$addr}}) {
2181 &{$sub_ref}($self);
2182 next LINE if /^$/;
2183 }
2184
2185 # Here the line is ok. return success.
2186 return 1;
2187 } # End of looping through lines.
2188
2189 # If there is an EOF handler, call it (only once) and if it generates
2190 # more lines to process go back in the loop to handle them.
2191 if ($eof_handler{$addr}) {
2192 &{$eof_handler{$addr}}($self);
2193 $eof_handler{$addr} = ""; # Currently only get one shot at it.
2194 goto LINE if $added_lines{$addr};
2195 }
2196
2197 # Return failure -- no more lines.
2198 return 0;
2199
2200 }
2201
2202# Not currently used, not fully tested.
2203# sub peek {
2204# # Non-destructive look-ahead one non-adjusted, non-comment, non-blank
2205# # record. Not callable from an each_line_handler(), nor does it call
2206# # an each_line_handler() on the line.
2207#
2208# my $self = shift;
ffe43484 2209# my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2210#
2211# foreach my $inserted_ref (@{$added_lines{$addr}}) {
2212# my ($adjusted, $line) = @{$inserted_ref};
2213# next if $adjusted;
2214#
2215# # Remove comments and trailing space, and return a non-empty
2216# # resulting line
2217# $line =~ s/#.*//;
2218# $line =~ s/\s+$//;
2219# return $line if $line ne "";
2220# }
2221#
2222# return if ! ref $handle{$addr}; # Don't read unless is real file
2223# while (1) { # Loop until find non-comment, non-empty line
2224# local $to_trace = 1 if main::DEBUG;
2225# trace $_ if main::DEBUG && $to_trace;
2226# return if ! defined (my $line = readline $handle{$addr});
2227# chomp $line;
2228# push @{$added_lines{$addr}}, [ 0, $line ];
2229#
2230# $line =~ s/#.*//;
2231# $line =~ s/\s+$//;
2232# return $line if $line ne "";
2233# }
2234#
2235# return;
2236# }
2237
2238
2239 sub insert_lines {
2240 # Lines can be inserted so that it looks like they were in the input
2241 # file at the place it was when this routine is called. See also
2242 # insert_adjusted_lines(). Lines inserted via this routine go through
2243 # any each_line_handler()
2244
2245 my $self = shift;
2246
2247 # Each inserted line is an array, with the first element being 0 to
2248 # indicate that this line hasn't been adjusted, and needs to be
2249 # processed.
f998e60c 2250 no overloading;
051df77b 2251 push @{$added_lines{pack 'J', $self}}, map { [ 0, $_ ] } @_;
99870f4d
KW
2252 return;
2253 }
2254
2255 sub insert_adjusted_lines {
2256 # Lines can be inserted so that it looks like they were in the input
2257 # file at the place it was when this routine is called. See also
2258 # insert_lines(). Lines inserted via this routine are already fully
2259 # adjusted, ready to be processed; each_line_handler()s handlers will
2260 # not be called. This means this is not a completely general
2261 # facility, as only the last each_line_handler on the stack should
2262 # call this. It could be made more general, by passing to each of the
2263 # line_handlers their position on the stack, which they would pass on
2264 # to this routine, and that would replace the boolean first element in
2265 # the anonymous array pushed here, so that the next_line routine could
2266 # use that to call only those handlers whose index is after it on the
2267 # stack. But this is overkill for what is needed now.
2268
2269 my $self = shift;
2270 trace $_[0] if main::DEBUG && $to_trace;
2271
2272 # Each inserted line is an array, with the first element being 1 to
2273 # indicate that this line has been adjusted
f998e60c 2274 no overloading;
051df77b 2275 push @{$added_lines{pack 'J', $self}}, map { [ 1, $_ ] } @_;
99870f4d
KW
2276 return;
2277 }
2278
2279 sub get_missings {
2280 # Returns the stored up @missings lines' values, and clears the list.
2281 # The values are in an array, consisting of the default in the first
2282 # element, and the property in the 2nd. However, since these lines
2283 # can be stacked up, the return is an array of all these arrays.
2284
2285 my $self = shift;
2286 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2287
ffe43484 2288 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2289
2290 # If not accepting a list return, just return the first one.
2291 return shift @{$missings{$addr}} unless wantarray;
2292
2293 my @return = @{$missings{$addr}};
2294 undef @{$missings{$addr}};
2295 return @return;
2296 }
2297
2298 sub _insert_property_into_line {
2299 # Add a property field to $_, if this file requires it.
2300
f998e60c 2301 my $self = shift;
ffe43484 2302 my $addr = do { no overloading; pack 'J', $self; };
f998e60c 2303 my $property = $property{$addr};
99870f4d
KW
2304 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2305
2306 $_ =~ s/(;|$)/; $property$1/;
2307 return;
2308 }
2309
2310 sub carp_bad_line {
2311 # Output consistent error messages, using either a generic one, or the
2312 # one given by the optional parameter. To avoid gazillions of the
2313 # same message in case the syntax of a file is way off, this routine
2314 # only outputs the first instance of each message, incrementing a
2315 # count so the totals can be output at the end of the file.
2316
2317 my $self = shift;
2318 my $message = shift;
2319 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2320
ffe43484 2321 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2322
2323 $message = 'Unexpected line' unless $message;
2324
2325 # No trailing punctuation so as to fit with our addenda.
2326 $message =~ s/[.:;,]$//;
2327
2328 # If haven't seen this exact message before, output it now. Otherwise
2329 # increment the count of how many times it has occurred
2330 unless ($errors{$addr}->{$message}) {
2331 Carp::my_carp("$message in '$_' in "
f998e60c 2332 . $file{$addr}
99870f4d
KW
2333 . " at line $.. Skipping this line;");
2334 $errors{$addr}->{$message} = 1;
2335 }
2336 else {
2337 $errors{$addr}->{$message}++;
2338 }
2339
2340 # Clear the line to prevent any further (meaningful) processing of it.
2341 $_ = "";
2342
2343 return;
2344 }
2345} # End closure
2346
2347package Multi_Default;
2348
2349# Certain properties in early versions of Unicode had more than one possible
2350# default for code points missing from the files. In these cases, one
2351# default applies to everything left over after all the others are applied,
2352# and for each of the others, there is a description of which class of code
2353# points applies to it. This object helps implement this by storing the
2354# defaults, and for all but that final default, an eval string that generates
2355# the class that it applies to.
2356
2357
2358{ # Closure
2359
2360 main::setup_package();
2361
2362 my %class_defaults;
2363 # The defaults structure for the classes
2364 main::set_access('class_defaults', \%class_defaults);
2365
2366 my %other_default;
2367 # The default that applies to everything left over.
2368 main::set_access('other_default', \%other_default, 'r');
2369
2370
2371 sub new {
2372 # The constructor is called with default => eval pairs, terminated by
2373 # the left-over default. e.g.
2374 # Multi_Default->new(
2375 # 'T' => '$gc->table("Mn") + $gc->table("Cf") - 0x200C
2376 # - 0x200D',
2377 # 'R' => 'some other expression that evaluates to code points',
2378 # .
2379 # .
2380 # .
2381 # 'U'));
2382
2383 my $class = shift;
2384
2385 my $self = bless \do{my $anonymous_scalar}, $class;
ffe43484 2386 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2387
2388 while (@_ > 1) {
2389 my $default = shift;
2390 my $eval = shift;
2391 $class_defaults{$addr}->{$default} = $eval;
2392 }
2393
2394 $other_default{$addr} = shift;
2395
2396 return $self;
2397 }
2398
2399 sub get_next_defaults {
2400 # Iterates and returns the next class of defaults.
2401 my $self = shift;
2402 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2403
ffe43484 2404 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2405
2406 return each %{$class_defaults{$addr}};
2407 }
2408}
2409
2410package Alias;
2411
2412# An alias is one of the names that a table goes by. This class defines them
2413# including some attributes. Everything is currently setup in the
2414# constructor.
2415
2416
2417{ # Closure
2418
2419 main::setup_package();
2420
2421 my %name;
2422 main::set_access('name', \%name, 'r');
2423
2424 my %loose_match;
2425 # Determined by the constructor code if this name should match loosely or
2426 # not. The constructor parameters can override this, but it isn't fully
2427 # implemented, as should have ability to override Unicode one's via
2428 # something like a set_loose_match()
2429 main::set_access('loose_match', \%loose_match, 'r');
2430
2431 my %make_pod_entry;
2432 # Some aliases should not get their own entries because they are covered
2433 # by a wild-card, and some we want to discourage use of. Binary
2434 main::set_access('make_pod_entry', \%make_pod_entry, 'r');
2435
2436 my %status;
2437 # Aliases have a status, like deprecated, or even suppressed (which means
2438 # they don't appear in documentation). Enum
2439 main::set_access('status', \%status, 'r');
2440
2441 my %externally_ok;
2442 # Similarly, some aliases should not be considered as usable ones for
2443 # external use, such as file names, or we don't want documentation to
2444 # recommend them. Boolean
2445 main::set_access('externally_ok', \%externally_ok, 'r');
2446
2447 sub new {
2448 my $class = shift;
2449
2450 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2451 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2452
2453 $name{$addr} = shift;
2454 $loose_match{$addr} = shift;
2455 $make_pod_entry{$addr} = shift;
2456 $externally_ok{$addr} = shift;
2457 $status{$addr} = shift;
2458
2459 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2460
2461 # Null names are never ok externally
2462 $externally_ok{$addr} = 0 if $name{$addr} eq "";
2463
2464 return $self;
2465 }
2466}
2467
2468package Range;
2469
2470# A range is the basic unit for storing code points, and is described in the
2471# comments at the beginning of the program. Each range has a starting code
2472# point; an ending code point (not less than the starting one); a value
2473# that applies to every code point in between the two end-points, inclusive;
2474# and an enum type that applies to the value. The type is for the user's
2475# convenience, and has no meaning here, except that a non-zero type is
2476# considered to not obey the normal Unicode rules for having standard forms.
2477#
2478# The same structure is used for both map and match tables, even though in the
2479# latter, the value (and hence type) is irrelevant and could be used as a
2480# comment. In map tables, the value is what all the code points in the range
2481# map to. Type 0 values have the standardized version of the value stored as
2482# well, so as to not have to recalculate it a lot.
2483
2484sub trace { return main::trace(@_); }
2485
2486{ # Closure
2487
2488 main::setup_package();
2489
2490 my %start;
2491 main::set_access('start', \%start, 'r', 's');
2492
2493 my %end;
2494 main::set_access('end', \%end, 'r', 's');
2495
2496 my %value;
2497 main::set_access('value', \%value, 'r');
2498
2499 my %type;
2500 main::set_access('type', \%type, 'r');
2501
2502 my %standard_form;
2503 # The value in internal standard form. Defined only if the type is 0.
2504 main::set_access('standard_form', \%standard_form);
2505
2506 # Note that if these fields change, the dump() method should as well
2507
2508 sub new {
2509 return Carp::carp_too_few_args(\@_, 3) if main::DEBUG && @_ < 3;
2510 my $class = shift;
2511
2512 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2513 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2514
2515 $start{$addr} = shift;
2516 $end{$addr} = shift;
2517
2518 my %args = @_;
2519
2520 my $value = delete $args{'Value'}; # Can be 0
2521 $value = "" unless defined $value;
2522 $value{$addr} = $value;
2523
2524 $type{$addr} = delete $args{'Type'} || 0;
2525
2526 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2527
2528 if (! $type{$addr}) {
2529 $standard_form{$addr} = main::standardize($value);
2530 }
2531
2532 return $self;
2533 }
2534
2535 use overload
2536 fallback => 0,
2537 qw("") => "_operator_stringify",
2538 "." => \&main::_operator_dot,
2539 ;
2540
2541 sub _operator_stringify {
2542 my $self = shift;
ffe43484 2543 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2544
2545 # Output it like '0041..0065 (value)'
2546 my $return = sprintf("%04X", $start{$addr})
2547 . '..'
2548 . sprintf("%04X", $end{$addr});
2549 my $value = $value{$addr};
2550 my $type = $type{$addr};
2551 $return .= ' (';
2552 $return .= "$value";
2553 $return .= ", Type=$type" if $type != 0;
2554 $return .= ')';
2555
2556 return $return;
2557 }
2558
2559 sub standard_form {
2560 # The standard form is the value itself if the standard form is
2561 # undefined (that is if the value is special)
2562
2563 my $self = shift;
2564 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2565
ffe43484 2566 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2567
2568 return $standard_form{$addr} if defined $standard_form{$addr};
2569 return $value{$addr};
2570 }
2571
2572 sub dump {
2573 # Human, not machine readable. For machine readable, comment out this
2574 # entire routine and let the standard one take effect.
2575 my $self = shift;
2576 my $indent = shift;
2577 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2578
ffe43484 2579 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2580
2581 my $return = $indent
2582 . sprintf("%04X", $start{$addr})
2583 . '..'
2584 . sprintf("%04X", $end{$addr})
2585 . " '$value{$addr}';";
2586 if (! defined $standard_form{$addr}) {
2587 $return .= "(type=$type{$addr})";
2588 }
2589 elsif ($standard_form{$addr} ne $value{$addr}) {
2590 $return .= "(standard '$standard_form{$addr}')";
2591 }
2592 return $return;
2593 }
2594} # End closure
2595
2596package _Range_List_Base;
2597
2598# Base class for range lists. A range list is simply an ordered list of
2599# ranges, so that the ranges with the lowest starting numbers are first in it.
2600#
2601# When a new range is added that is adjacent to an existing range that has the
2602# same value and type, it merges with it to form a larger range.
2603#
2604# Ranges generally do not overlap, except that there can be multiple entries
2605# of single code point ranges. This is because of NameAliases.txt.
2606#
2607# In this program, there is a standard value such that if two different
2608# values, have the same standard value, they are considered equivalent. This
2609# value was chosen so that it gives correct results on Unicode data
2610
2611# There are a number of methods to manipulate range lists, and some operators
2612# are overloaded to handle them.
2613
99870f4d
KW
2614sub trace { return main::trace(@_); }
2615
2616{ # Closure
2617
2618 our $addr;
2619
2620 main::setup_package();
2621
2622 my %ranges;
2623 # The list of ranges
2624 main::set_access('ranges', \%ranges, 'readable_array');
2625
2626 my %max;
2627 # The highest code point in the list. This was originally a method, but
2628 # actual measurements said it was used a lot.
2629 main::set_access('max', \%max, 'r');
2630
2631 my %each_range_iterator;
2632 # Iterator position for each_range()
2633 main::set_access('each_range_iterator', \%each_range_iterator);
2634
2635 my %owner_name_of;
2636 # Name of parent this is attached to, if any. Solely for better error
2637 # messages.
2638 main::set_access('owner_name_of', \%owner_name_of, 'p_r');
2639
2640 my %_search_ranges_cache;
2641 # A cache of the previous result from _search_ranges(), for better
2642 # performance
2643 main::set_access('_search_ranges_cache', \%_search_ranges_cache);
2644
2645 sub new {
2646 my $class = shift;
2647 my %args = @_;
2648
2649 # Optional initialization data for the range list.
2650 my $initialize = delete $args{'Initialize'};
2651
2652 my $self;
2653
2654 # Use _union() to initialize. _union() returns an object of this
2655 # class, which means that it will call this constructor recursively.
2656 # But it won't have this $initialize parameter so that it won't
2657 # infinitely loop on this.
2658 return _union($class, $initialize, %args) if defined $initialize;
2659
2660 $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2661 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2662
2663 # Optional parent object, only for debug info.
2664 $owner_name_of{$addr} = delete $args{'Owner'};
2665 $owner_name_of{$addr} = "" if ! defined $owner_name_of{$addr};
2666
2667 # Stringify, in case it is an object.
2668 $owner_name_of{$addr} = "$owner_name_of{$addr}";
2669
2670 # This is used only for error messages, and so a colon is added
2671 $owner_name_of{$addr} .= ": " if $owner_name_of{$addr} ne "";
2672
2673 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2674
2675 # Max is initialized to a negative value that isn't adjacent to 0,
2676 # for simpler tests
2677 $max{$addr} = -2;
2678
2679 $_search_ranges_cache{$addr} = 0;
2680 $ranges{$addr} = [];
2681
2682 return $self;
2683 }
2684
2685 use overload
2686 fallback => 0,
2687 qw("") => "_operator_stringify",
2688 "." => \&main::_operator_dot,
2689 ;
2690
2691 sub _operator_stringify {
2692 my $self = shift;
ffe43484 2693 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2694
2695 return "Range_List attached to '$owner_name_of{$addr}'"
2696 if $owner_name_of{$addr};
2697 return "anonymous Range_List " . \$self;
2698 }
2699
2700 sub _union {
2701 # Returns the union of the input code points. It can be called as
2702 # either a constructor or a method. If called as a method, the result
2703 # will be a new() instance of the calling object, containing the union
2704 # of that object with the other parameter's code points; if called as
2705 # a constructor, the first parameter gives the class the new object
2706 # should be, and the second parameter gives the code points to go into
2707 # it.
2708 # In either case, there are two parameters looked at by this routine;
2709 # any additional parameters are passed to the new() constructor.
2710 #
2711 # The code points can come in the form of some object that contains
2712 # ranges, and has a conventionally named method to access them; or
2713 # they can be an array of individual code points (as integers); or
2714 # just a single code point.
2715 #
2716 # If they are ranges, this routine doesn't make any effort to preserve
2717 # the range values of one input over the other. Therefore this base
2718 # class should not allow _union to be called from other than
2719 # initialization code, so as to prevent two tables from being added
2720 # together where the range values matter. The general form of this
2721 # routine therefore belongs in a derived class, but it was moved here
2722 # to avoid duplication of code. The failure to overload this in this
2723 # class keeps it safe.
2724 #
2725
2726 my $self;
2727 my @args; # Arguments to pass to the constructor
2728
2729 my $class = shift;
2730
2731 # If a method call, will start the union with the object itself, and
2732 # the class of the new object will be the same as self.
2733 if (ref $class) {
2734 $self = $class;
2735 $class = ref $self;
2736 push @args, $self;
2737 }
2738
2739 # Add the other required parameter.
2740 push @args, shift;
2741 # Rest of parameters are passed on to the constructor
2742
2743 # Accumulate all records from both lists.
2744 my @records;
2745 for my $arg (@args) {
2746 #local $to_trace = 0 if main::DEBUG;
2747 trace "argument = $arg" if main::DEBUG && $to_trace;
2748 if (! defined $arg) {
2749 my $message = "";
2750 if (defined $self) {
f998e60c 2751 no overloading;
051df77b 2752 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
2753 }
2754 Carp::my_carp_bug($message .= "Undefined argument to _union. No union done.");
2755 return;
2756 }
2757 $arg = [ $arg ] if ! ref $arg;
2758 my $type = ref $arg;
2759 if ($type eq 'ARRAY') {
2760 foreach my $element (@$arg) {
2761 push @records, Range->new($element, $element);
2762 }
2763 }
2764 elsif ($arg->isa('Range')) {
2765 push @records, $arg;
2766 }
2767 elsif ($arg->can('ranges')) {
2768 push @records, $arg->ranges;
2769 }
2770 else {
2771 my $message = "";
2772 if (defined $self) {
f998e60c 2773 no overloading;
051df77b 2774 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
2775 }
2776 Carp::my_carp_bug($message . "Cannot take the union of a $type. No union done.");
2777 return;
2778 }
2779 }
2780
2781 # Sort with the range containing the lowest ordinal first, but if
2782 # two ranges start at the same code point, sort with the bigger range
2783 # of the two first, because it takes fewer cycles.
2784 @records = sort { ($a->start <=> $b->start)
2785 or
2786 # if b is shorter than a, b->end will be
2787 # less than a->end, and we want to select
2788 # a, so want to return -1
2789 ($b->end <=> $a->end)
2790 } @records;
2791
2792 my $new = $class->new(@_);
2793
2794 # Fold in records so long as they add new information.
2795 for my $set (@records) {
2796 my $start = $set->start;
2797 my $end = $set->end;
2798 my $value = $set->value;
2799 if ($start > $new->max) {
2800 $new->_add_delete('+', $start, $end, $value);
2801 }
2802 elsif ($end > $new->max) {
2803 $new->_add_delete('+', $new->max +1, $end, $value);
2804 }
2805 }
2806
2807 return $new;
2808 }
2809
2810 sub range_count { # Return the number of ranges in the range list
2811 my $self = shift;
2812 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2813
f998e60c 2814 no overloading;
051df77b 2815 return scalar @{$ranges{pack 'J', $self}};
99870f4d
KW
2816 }
2817
2818 sub min {
2819 # Returns the minimum code point currently in the range list, or if
2820 # the range list is empty, 2 beyond the max possible. This is a
2821 # method because used so rarely, that not worth saving between calls,
2822 # and having to worry about changing it as ranges are added and
2823 # deleted.
2824
2825 my $self = shift;
2826 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2827
ffe43484 2828 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2829
2830 # If the range list is empty, return a large value that isn't adjacent
2831 # to any that could be in the range list, for simpler tests
2832 return $LAST_UNICODE_CODEPOINT + 2 unless scalar @{$ranges{$addr}};
2833 return $ranges{$addr}->[0]->start;
2834 }
2835
2836 sub contains {
2837 # Boolean: Is argument in the range list? If so returns $i such that:
2838 # range[$i]->end < $codepoint <= range[$i+1]->end
2839 # which is one beyond what you want; this is so that the 0th range
2840 # doesn't return false
2841 my $self = shift;
2842 my $codepoint = shift;
2843 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2844
99870f4d
KW
2845 my $i = $self->_search_ranges($codepoint);
2846 return 0 unless defined $i;
2847
2848 # The search returns $i, such that
2849 # range[$i-1]->end < $codepoint <= range[$i]->end
2850 # So is in the table if and only iff it is at least the start position
2851 # of range $i.
f998e60c 2852 no overloading;
051df77b 2853 return 0 if $ranges{pack 'J', $self}->[$i]->start > $codepoint;
99870f4d
KW
2854 return $i + 1;
2855 }
2856
2f7a8815
KW
2857 sub containing_range {
2858 # Returns the range object that contains the code point, undef if none
2859
2860 my $self = shift;
2861 my $codepoint = shift;
2862 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2863
2864 my $i = $self->contains($codepoint);
2865 return unless $i;
2866
2867 # contains() returns 1 beyond where we should look
2868 no overloading;
2869 return $ranges{pack 'J', $self}->[$i-1];
2870 }
2871
99870f4d
KW
2872 sub value_of {
2873 # Returns the value associated with the code point, undef if none
2874
2875 my $self = shift;
2876 my $codepoint = shift;
2877 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2878
d69c231b
KW
2879 my $range = $self->containing_range($codepoint);
2880 return unless defined $range;
99870f4d 2881
d69c231b 2882 return $range->value;
99870f4d
KW
2883 }
2884
0a9dbafc
KW
2885 sub type_of {
2886 # Returns the type of the range containing the code point, undef if
2887 # the code point is not in the table
2888
2889 my $self = shift;
2890 my $codepoint = shift;
2891 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2892
2893 my $range = $self->containing_range($codepoint);
2894 return unless defined $range;
2895
2896 return $range->type;
2897 }
2898
99870f4d
KW
2899 sub _search_ranges {
2900 # Find the range in the list which contains a code point, or where it
2901 # should go if were to add it. That is, it returns $i, such that:
2902 # range[$i-1]->end < $codepoint <= range[$i]->end
2903 # Returns undef if no such $i is possible (e.g. at end of table), or
2904 # if there is an error.
2905
2906 my $self = shift;
2907 my $code_point = shift;
2908 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2909
ffe43484 2910 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2911
2912 return if $code_point > $max{$addr};
2913 my $r = $ranges{$addr}; # The current list of ranges
2914 my $range_list_size = scalar @$r;
2915 my $i;
2916
2917 use integer; # want integer division
2918
2919 # Use the cached result as the starting guess for this one, because,
2920 # an experiment on 5.1 showed that 90% of the time the cache was the
2921 # same as the result on the next call (and 7% it was one less).
2922 $i = $_search_ranges_cache{$addr};
2923 $i = 0 if $i >= $range_list_size; # Reset if no longer valid (prob.
2924 # from an intervening deletion
2925 #local $to_trace = 1 if main::DEBUG;
2926 trace "previous \$i is still valid: $i" if main::DEBUG && $to_trace && $code_point <= $r->[$i]->end && ($i == 0 || $r->[$i-1]->end < $code_point);
2927 return $i if $code_point <= $r->[$i]->end
2928 && ($i == 0 || $r->[$i-1]->end < $code_point);
2929
2930 # Here the cache doesn't yield the correct $i. Try adding 1.
2931 if ($i < $range_list_size - 1
2932 && $r->[$i]->end < $code_point &&
2933 $code_point <= $r->[$i+1]->end)
2934 {
2935 $i++;
2936 trace "next \$i is correct: $i" if main::DEBUG && $to_trace;
2937 $_search_ranges_cache{$addr} = $i;
2938 return $i;
2939 }
2940
2941 # Here, adding 1 also didn't work. We do a binary search to
2942 # find the correct position, starting with current $i
2943 my $lower = 0;
2944 my $upper = $range_list_size - 1;
2945 while (1) {
2946 trace "top of loop i=$i:", sprintf("%04X", $r->[$lower]->start), "[$lower] .. ", sprintf("%04X", $r->[$i]->start), "[$i] .. ", sprintf("%04X", $r->[$upper]->start), "[$upper]" if main::DEBUG && $to_trace;
2947
2948 if ($code_point <= $r->[$i]->end) {
2949
2950 # Here we have met the upper constraint. We can quit if we
2951 # also meet the lower one.
2952 last if $i == 0 || $r->[$i-1]->end < $code_point;
2953
2954 $upper = $i; # Still too high.
2955
2956 }
2957 else {
2958
2959 # Here, $r[$i]->end < $code_point, so look higher up.
2960 $lower = $i;
2961 }
2962
2963 # Split search domain in half to try again.
2964 my $temp = ($upper + $lower) / 2;
2965
2966 # No point in continuing unless $i changes for next time
2967 # in the loop.
2968 if ($temp == $i) {
2969
2970 # We can't reach the highest element because of the averaging.
2971 # So if one below the upper edge, force it there and try one
2972 # more time.
2973 if ($i == $range_list_size - 2) {
2974
2975 trace "Forcing to upper edge" if main::DEBUG && $to_trace;
2976 $i = $range_list_size - 1;
2977
2978 # Change $lower as well so if fails next time through,
2979 # taking the average will yield the same $i, and we will
2980 # quit with the error message just below.
2981 $lower = $i;
2982 next;
2983 }
2984 Carp::my_carp_bug("$owner_name_of{$addr}Can't find where the range ought to go. No action taken.");
2985 return;
2986 }
2987 $i = $temp;
2988 } # End of while loop
2989
2990 if (main::DEBUG && $to_trace) {
2991 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i;
2992 trace "i= [ $i ]", $r->[$i];
2993 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < $range_list_size - 1;
2994 }
2995
2996 # Here we have found the offset. Cache it as a starting point for the
2997 # next call.
2998 $_search_ranges_cache{$addr} = $i;
2999 return $i;
3000 }
3001
3002 sub _add_delete {
3003 # Add, replace or delete ranges to or from a list. The $type
3004 # parameter gives which:
3005 # '+' => insert or replace a range, returning a list of any changed
3006 # ranges.
3007 # '-' => delete a range, returning a list of any deleted ranges.
3008 #
3009 # The next three parameters give respectively the start, end, and
3010 # value associated with the range. 'value' should be null unless the
3011 # operation is '+';
3012 #
3013 # The range list is kept sorted so that the range with the lowest
3014 # starting position is first in the list, and generally, adjacent
c1739a4a 3015 # ranges with the same values are merged into a single larger one (see
99870f4d
KW
3016 # exceptions below).
3017 #
c1739a4a 3018 # There are more parameters; all are key => value pairs:
99870f4d
KW
3019 # Type gives the type of the value. It is only valid for '+'.
3020 # All ranges have types; if this parameter is omitted, 0 is
3021 # assumed. Ranges with type 0 are assumed to obey the
3022 # Unicode rules for casing, etc; ranges with other types are
3023 # not. Otherwise, the type is arbitrary, for the caller's
3024 # convenience, and looked at only by this routine to keep
3025 # adjacent ranges of different types from being merged into
3026 # a single larger range, and when Replace =>
3027 # $IF_NOT_EQUIVALENT is specified (see just below).
3028 # Replace determines what to do if the range list already contains
3029 # ranges which coincide with all or portions of the input
3030 # range. It is only valid for '+':
3031 # => $NO means that the new value is not to replace
3032 # any existing ones, but any empty gaps of the
3033 # range list coinciding with the input range
3034 # will be filled in with the new value.
3035 # => $UNCONDITIONALLY means to replace the existing values with
3036 # this one unconditionally. However, if the
3037 # new and old values are identical, the
3038 # replacement is skipped to save cycles
3039 # => $IF_NOT_EQUIVALENT means to replace the existing values
3040 # with this one if they are not equivalent.
3041 # Ranges are equivalent if their types are the
c1739a4a 3042 # same, and they are the same string; or if
99870f4d
KW
3043 # both are type 0 ranges, if their Unicode
3044 # standard forms are identical. In this last
3045 # case, the routine chooses the more "modern"
3046 # one to use. This is because some of the
3047 # older files are formatted with values that
3048 # are, for example, ALL CAPs, whereas the
3049 # derived files have a more modern style,
3050 # which looks better. By looking for this
3051 # style when the pre-existing and replacement
3052 # standard forms are the same, we can move to
3053 # the modern style
3054 # => $MULTIPLE means that if this range duplicates an
3055 # existing one, but has a different value,
3056 # don't replace the existing one, but insert
3057 # this, one so that the same range can occur
3058 # multiple times.
3059 # => anything else is the same as => $IF_NOT_EQUIVALENT
3060 #
c1739a4a
KW
3061 # "same value" means identical for non-type-0 ranges, and it means
3062 # having the same standard forms for type-0 ranges.
99870f4d
KW
3063
3064 return Carp::carp_too_few_args(\@_, 5) if main::DEBUG && @_ < 5;
3065
3066 my $self = shift;
3067 my $operation = shift; # '+' for add/replace; '-' for delete;
3068 my $start = shift;
3069 my $end = shift;
3070 my $value = shift;
3071
3072 my %args = @_;
3073
3074 $value = "" if not defined $value; # warning: $value can be "0"
3075
3076 my $replace = delete $args{'Replace'};
3077 $replace = $IF_NOT_EQUIVALENT unless defined $replace;
3078
3079 my $type = delete $args{'Type'};
3080 $type = 0 unless defined $type;
3081
3082 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3083
ffe43484 3084 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3085
3086 if ($operation ne '+' && $operation ne '-') {
3087 Carp::my_carp_bug("$owner_name_of{$addr}First parameter to _add_delete must be '+' or '-'. No action taken.");
3088 return;
3089 }
3090 unless (defined $start && defined $end) {
3091 Carp::my_carp_bug("$owner_name_of{$addr}Undefined start and/or end to _add_delete. No action taken.");
3092 return;
3093 }
3094 unless ($end >= $start) {
3095 Carp::my_carp_bug("$owner_name_of{$addr}End of range (" . sprintf("%04X", $end) . ") must not be before start (" . sprintf("%04X", $start) . "). No action taken.");
3096 return;
3097 }
3098 #local $to_trace = 1 if main::DEBUG;
3099
3100 if ($operation eq '-') {
3101 if ($replace != $IF_NOT_EQUIVALENT) {
3102 Carp::my_carp_bug("$owner_name_of{$addr}Replace => \$IF_NOT_EQUIVALENT is required when deleting a range from a range list. Assuming Replace => \$IF_NOT_EQUIVALENT.");
3103 $replace = $IF_NOT_EQUIVALENT;
3104 }
3105 if ($type) {
3106 Carp::my_carp_bug("$owner_name_of{$addr}Type => 0 is required when deleting a range from a range list. Assuming Type => 0.");
3107 $type = 0;
3108 }
3109 if ($value ne "") {
3110 Carp::my_carp_bug("$owner_name_of{$addr}Value => \"\" is required when deleting a range from a range list. Assuming Value => \"\".");
3111 $value = "";
3112 }
3113 }
3114
3115 my $r = $ranges{$addr}; # The current list of ranges
3116 my $range_list_size = scalar @$r; # And its size
3117 my $max = $max{$addr}; # The current high code point in
3118 # the list of ranges
3119
3120 # Do a special case requiring fewer machine cycles when the new range
3121 # starts after the current highest point. The Unicode input data is
3122 # structured so this is common.
3123 if ($start > $max) {
3124
3125 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) type=$type" if main::DEBUG && $to_trace;
3126 return if $operation eq '-'; # Deleting a non-existing range is a
3127 # no-op
3128
3129 # If the new range doesn't logically extend the current final one
3130 # in the range list, create a new range at the end of the range
3131 # list. (max cleverly is initialized to a negative number not
3132 # adjacent to 0 if the range list is empty, so even adding a range
3133 # to an empty range list starting at 0 will have this 'if'
3134 # succeed.)
3135 if ($start > $max + 1 # non-adjacent means can't extend.
3136 || @{$r}[-1]->value ne $value # values differ, can't extend.
3137 || @{$r}[-1]->type != $type # types differ, can't extend.
3138 ) {
3139 push @$r, Range->new($start, $end,
3140 Value => $value,
3141 Type => $type);
3142 }
3143 else {
3144
3145 # Here, the new range starts just after the current highest in
3146 # the range list, and they have the same type and value.
3147 # Extend the current range to incorporate the new one.
3148 @{$r}[-1]->set_end($end);
3149 }
3150
3151 # This becomes the new maximum.
3152 $max{$addr} = $end;
3153
3154 return;
3155 }
3156 #local $to_trace = 0 if main::DEBUG;
3157
3158 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) replace=$replace" if main::DEBUG && $to_trace;
3159
3160 # Here, the input range isn't after the whole rest of the range list.
3161 # Most likely 'splice' will be needed. The rest of the routine finds
3162 # the needed splice parameters, and if necessary, does the splice.
3163 # First, find the offset parameter needed by the splice function for
3164 # the input range. Note that the input range may span multiple
3165 # existing ones, but we'll worry about that later. For now, just find
3166 # the beginning. If the input range is to be inserted starting in a
3167 # position not currently in the range list, it must (obviously) come
3168 # just after the range below it, and just before the range above it.
3169 # Slightly less obviously, it will occupy the position currently
3170 # occupied by the range that is to come after it. More formally, we
3171 # are looking for the position, $i, in the array of ranges, such that:
3172 #
3173 # r[$i-1]->start <= r[$i-1]->end < $start < r[$i]->start <= r[$i]->end
3174 #
3175 # (The ordered relationships within existing ranges are also shown in
3176 # the equation above). However, if the start of the input range is
3177 # within an existing range, the splice offset should point to that
3178 # existing range's position in the list; that is $i satisfies a
3179 # somewhat different equation, namely:
3180 #
3181 #r[$i-1]->start <= r[$i-1]->end < r[$i]->start <= $start <= r[$i]->end
3182 #
3183 # More briefly, $start can come before or after r[$i]->start, and at
3184 # this point, we don't know which it will be. However, these
3185 # two equations share these constraints:
3186 #
3187 # r[$i-1]->end < $start <= r[$i]->end
3188 #
3189 # And that is good enough to find $i.
3190
3191 my $i = $self->_search_ranges($start);
3192 if (! defined $i) {
3193 Carp::my_carp_bug("Searching $self for range beginning with $start unexpectedly returned undefined. Operation '$operation' not performed");
3194 return;
3195 }
3196
3197 # The search function returns $i such that:
3198 #
3199 # r[$i-1]->end < $start <= r[$i]->end
3200 #
3201 # That means that $i points to the first range in the range list
3202 # that could possibly be affected by this operation. We still don't
3203 # know if the start of the input range is within r[$i], or if it
3204 # points to empty space between r[$i-1] and r[$i].
3205 trace "[$i] is the beginning splice point. Existing range there is ", $r->[$i] if main::DEBUG && $to_trace;
3206
3207 # Special case the insertion of data that is not to replace any
3208 # existing data.
3209 if ($replace == $NO) { # If $NO, has to be operation '+'
3210 #local $to_trace = 1 if main::DEBUG;
3211 trace "Doesn't replace" if main::DEBUG && $to_trace;
3212
3213 # Here, the new range is to take effect only on those code points
3214 # that aren't already in an existing range. This can be done by
3215 # looking through the existing range list and finding the gaps in
3216 # the ranges that this new range affects, and then calling this
3217 # function recursively on each of those gaps, leaving untouched
3218 # anything already in the list. Gather up a list of the changed
3219 # gaps first so that changes to the internal state as new ranges
3220 # are added won't be a problem.
3221 my @gap_list;
3222
3223 # First, if the starting point of the input range is outside an
3224 # existing one, there is a gap from there to the beginning of the
3225 # existing range -- add a span to fill the part that this new
3226 # range occupies
3227 if ($start < $r->[$i]->start) {
3228 push @gap_list, Range->new($start,
3229 main::min($end,
3230 $r->[$i]->start - 1),
3231 Type => $type);
3232 trace "gap before $r->[$i] [$i], will add", $gap_list[-1] if main::DEBUG && $to_trace;
3233 }
3234
3235 # Then look through the range list for other gaps until we reach
3236 # the highest range affected by the input one.
3237 my $j;
3238 for ($j = $i+1; $j < $range_list_size; $j++) {
3239 trace "j=[$j]", $r->[$j] if main::DEBUG && $to_trace;
3240 last if $end < $r->[$j]->start;
3241
3242 # If there is a gap between when this range starts and the
3243 # previous one ends, add a span to fill it. Note that just
3244 # because there are two ranges doesn't mean there is a
3245 # non-zero gap between them. It could be that they have
3246 # different values or types
3247 if ($r->[$j-1]->end + 1 != $r->[$j]->start) {
3248 push @gap_list,
3249 Range->new($r->[$j-1]->end + 1,
3250 $r->[$j]->start - 1,
3251 Type => $type);
3252 trace "gap between $r->[$j-1] and $r->[$j] [$j], will add: $gap_list[-1]" if main::DEBUG && $to_trace;
3253 }
3254 }
3255
3256 # Here, we have either found an existing range in the range list,
3257 # beyond the area affected by the input one, or we fell off the
3258 # end of the loop because the input range affects the whole rest
3259 # of the range list. In either case, $j is 1 higher than the
3260 # highest affected range. If $j == $i, it means that there are no
3261 # affected ranges, that the entire insertion is in the gap between
3262 # r[$i-1], and r[$i], which we already have taken care of before
3263 # the loop.
3264 # On the other hand, if there are affected ranges, it might be
3265 # that there is a gap that needs filling after the final such
3266 # range to the end of the input range
3267 if ($r->[$j-1]->end < $end) {
3268 push @gap_list, Range->new(main::max($start,
3269 $r->[$j-1]->end + 1),
3270 $end,
3271 Type => $type);
3272 trace "gap after $r->[$j-1], will add $gap_list[-1]" if main::DEBUG && $to_trace;
3273 }
3274
3275 # Call recursively to fill in all the gaps.
3276 foreach my $gap (@gap_list) {
3277 $self->_add_delete($operation,
3278 $gap->start,
3279 $gap->end,
3280 $value,
3281 Type => $type);
3282 }
3283
3284 return;
3285 }
3286
3287 # Here, we have taken care of the case where $replace is $NO, which
3288 # means that whatever action we now take is done unconditionally. It
3289 # still could be that this call will result in a no-op, if duplicates
3290 # aren't allowed, and we are inserting a range that merely duplicates
3291 # data already in the range list; or also if deleting a non-existent
3292 # range.
3293 # $i still points to the first potential affected range. Now find the
3294 # highest range affected, which will determine the length parameter to
3295 # splice. (The input range can span multiple existing ones.) While
3296 # we are looking through the range list, see also if this is an
3297 # insertion that will change the values of at least one of the
3298 # affected ranges. We don't need to do this check unless this is an
3299 # insertion of non-multiples, and also since this is a boolean, we
3300 # don't need to do it if have already determined that it will make a
3301 # change; just unconditionally change them. $cdm is created to be 1
3302 # if either of these is true. (The 'c' in the name comes from below)
3303 my $cdm = ($operation eq '-' || $replace == $MULTIPLE);
3304 my $j; # This will point to the highest affected range
3305
3306 # For non-zero types, the standard form is the value itself;
3307 my $standard_form = ($type) ? $value : main::standardize($value);
3308
3309 for ($j = $i; $j < $range_list_size; $j++) {
3310 trace "Looking for highest affected range; the one at $j is ", $r->[$j] if main::DEBUG && $to_trace;
3311
3312 # If find a range that it doesn't overlap into, we can stop
3313 # searching
3314 last if $end < $r->[$j]->start;
3315
3316 # Here, overlaps the range at $j. If the value's don't match,
3317 # and this is supposedly an insertion, it becomes a change
3318 # instead. This is what the 'c' stands for in $cdm.
3319 if (! $cdm) {
3320 if ($r->[$j]->standard_form ne $standard_form) {
3321 $cdm = 1;
3322 }
3323 else {
3324
3325 # Here, the two values are essentially the same. If the
3326 # two are actually identical, replacing wouldn't change
3327 # anything so skip it.
3328 my $pre_existing = $r->[$j]->value;
3329 if ($pre_existing ne $value) {
3330
3331 # Here the new and old standardized values are the
3332 # same, but the non-standardized values aren't. If
3333 # replacing unconditionally, then replace
3334 if( $replace == $UNCONDITIONALLY) {
3335 $cdm = 1;
3336 }
3337 else {
3338
3339 # Here, are replacing conditionally. Decide to
3340 # replace or not based on which appears to look
3341 # the "nicest". If one is mixed case and the
3342 # other isn't, choose the mixed case one.
3343 my $new_mixed = $value =~ /[A-Z]/
3344 && $value =~ /[a-z]/;
3345 my $old_mixed = $pre_existing =~ /[A-Z]/
3346 && $pre_existing =~ /[a-z]/;
3347
3348 if ($old_mixed != $new_mixed) {
3349 $cdm = 1 if $new_mixed;
3350 if (main::DEBUG && $to_trace) {
3351 if ($cdm) {
3352 trace "Replacing $pre_existing with $value";
3353 }
3354 else {
3355 trace "Retaining $pre_existing over $value";
3356 }
3357 }
3358 }
3359 else {
3360
3361 # Here casing wasn't different between the two.
3362 # If one has hyphens or underscores and the
3363 # other doesn't, choose the one with the
3364 # punctuation.
3365 my $new_punct = $value =~ /[-_]/;
3366 my $old_punct = $pre_existing =~ /[-_]/;
3367
3368 if ($old_punct != $new_punct) {
3369 $cdm = 1 if $new_punct;
3370 if (main::DEBUG && $to_trace) {
3371 if ($cdm) {
3372 trace "Replacing $pre_existing with $value";
3373 }
3374 else {
3375 trace "Retaining $pre_existing over $value";
3376 }
3377 }
3378 } # else existing one is just as "good";
3379 # retain it to save cycles.
3380 }
3381 }
3382 }
3383 }
3384 }
3385 } # End of loop looking for highest affected range.
3386
3387 # Here, $j points to one beyond the highest range that this insertion
3388 # affects (hence to beyond the range list if that range is the final
3389 # one in the range list).
3390
3391 # The splice length is all the affected ranges. Get it before
3392 # subtracting, for efficiency, so we don't have to later add 1.
3393 my $length = $j - $i;
3394
3395 $j--; # $j now points to the highest affected range.
3396 trace "Final affected range is $j: $r->[$j]" if main::DEBUG && $to_trace;
3397
3398 # If inserting a multiple record, this is where it goes, after all the
3399 # existing ones for this range. This implies an insertion, and no
3400 # change to any existing ranges. Note that $j can be -1 if this new
3401 # range doesn't actually duplicate any existing, and comes at the
3402 # beginning of the list, in which case we can handle it like any other
3403 # insertion, and is easier to do so.
3404 if ($replace == $MULTIPLE && $j >= 0) {
3405
3406 # This restriction could be remedied with a little extra work, but
3407 # it won't hopefully ever be necessary
3408 if ($r->[$j]->start != $r->[$j]->end) {
3409 Carp::my_carp_bug("$owner_name_of{$addr}Can't cope with adding a multiple when the other range ($r->[$j]) contains more than one code point. No action taken.");
3410 return;
3411 }
3412
3413 # Don't add an exact duplicate, as it isn't really a multiple
3414 return if $value eq $r->[$j]->value && $type eq $r->[$j]->type;
3415
3416 trace "Adding multiple record at $j+1 with $start..$end, $value" if main::DEBUG && $to_trace;
3417 my @return = splice @$r,
3418 $j+1,
3419 0,
3420 Range->new($start,
3421 $end,
3422 Value => $value,
3423 Type => $type);
3424 if (main::DEBUG && $to_trace) {
3425 trace "After splice:";
3426 trace 'j-2=[', $j-2, ']', $r->[$j-2] if $j >= 2;
3427 trace 'j-1=[', $j-1, ']', $r->[$j-1] if $j >= 1;
3428 trace "j =[", $j, "]", $r->[$j] if $j >= 0;
3429 trace 'j+1=[', $j+1, ']', $r->[$j+1] if $j < @$r - 1;
3430 trace 'j+2=[', $j+2, ']', $r->[$j+2] if $j < @$r - 2;
3431 trace 'j+3=[', $j+3, ']', $r->[$j+3] if $j < @$r - 3;
3432 }
3433 return @return;
3434 }
3435
3436 # Here, have taken care of $NO and $MULTIPLE replaces.
3437 # $j points to the highest affected range. But it can be < $i or even
3438 # -1. These happen only if the insertion is entirely in the gap
3439 # between r[$i-1] and r[$i]. Here's why: j < i means that the j loop
3440 # above exited first time through with $end < $r->[$i]->start. (And
3441 # then we subtracted one from j) This implies also that $start <
3442 # $r->[$i]->start, but we know from above that $r->[$i-1]->end <
3443 # $start, so the entire input range is in the gap.
3444 if ($j < $i) {
3445
3446 # Here the entire input range is in the gap before $i.
3447
3448 if (main::DEBUG && $to_trace) {
3449 if ($i) {
3450 trace "Entire range is between $r->[$i-1] and $r->[$i]";
3451 }
3452 else {
3453 trace "Entire range is before $r->[$i]";
3454 }
3455 }
3456 return if $operation ne '+'; # Deletion of a non-existent range is
3457 # a no-op
3458 }
3459 else {
3460
3461 # Here the entire input range is not in the gap before $i. There
3462 # is an affected one, and $j points to the highest such one.
3463
3464 # At this point, here is the situation:
3465 # This is not an insertion of a multiple, nor of tentative ($NO)
3466 # data.
3467 # $i points to the first element in the current range list that
3468 # may be affected by this operation. In fact, we know
3469 # that the range at $i is affected because we are in
3470 # the else branch of this 'if'
3471 # $j points to the highest affected range.
3472 # In other words,
3473 # r[$i-1]->end < $start <= r[$i]->end
3474 # And:
3475 # r[$i-1]->end < $start <= $end <= r[$j]->end
3476 #
3477 # Also:
3478 # $cdm is a boolean which is set true if and only if this is a
3479 # change or deletion (multiple was handled above). In
3480 # other words, it could be renamed to be just $cd.
3481
3482 # We now have enough information to decide if this call is a no-op
3483 # or not. It is a no-op if it is a deletion of a non-existent
3484 # range, or an insertion of already existing data.
3485
3486 if (main::DEBUG && $to_trace && ! $cdm
3487 && $i == $j
3488 && $start >= $r->[$i]->start)
3489 {
3490 trace "no-op";
3491 }
3492 return if ! $cdm # change or delete => not no-op
3493 && $i == $j # more than one affected range => not no-op
3494
3495 # Here, r[$i-1]->end < $start <= $end <= r[$i]->end
3496 # Further, $start and/or $end is >= r[$i]->start
3497 # The test below hence guarantees that
3498 # r[$i]->start < $start <= $end <= r[$i]->end
3499 # This means the input range is contained entirely in
3500 # the one at $i, so is a no-op
3501 && $start >= $r->[$i]->start;
3502 }
3503
3504 # Here, we know that some action will have to be taken. We have
3505 # calculated the offset and length (though adjustments may be needed)
3506 # for the splice. Now start constructing the replacement list.
3507 my @replacement;
3508 my $splice_start = $i;
3509
3510 my $extends_below;
3511 my $extends_above;
3512
3513 # See if should extend any adjacent ranges.
3514 if ($operation eq '-') { # Don't extend deletions
3515 $extends_below = $extends_above = 0;
3516 }
3517 else { # Here, should extend any adjacent ranges. See if there are
3518 # any.
3519 $extends_below = ($i > 0
3520 # can't extend unless adjacent
3521 && $r->[$i-1]->end == $start -1
3522 # can't extend unless are same standard value
3523 && $r->[$i-1]->standard_form eq $standard_form
3524 # can't extend unless share type
3525 && $r->[$i-1]->type == $type);
3526 $extends_above = ($j+1 < $range_list_size
3527 && $r->[$j+1]->start == $end +1
3528 && $r->[$j+1]->standard_form eq $standard_form
3529 && $r->[$j-1]->type == $type);
3530 }
3531 if ($extends_below && $extends_above) { # Adds to both
3532 $splice_start--; # start replace at element below
3533 $length += 2; # will replace on both sides
3534 trace "Extends both below and above ranges" if main::DEBUG && $to_trace;
3535
3536 # The result will fill in any gap, replacing both sides, and
3537 # create one large range.
3538 @replacement = Range->new($r->[$i-1]->start,
3539 $r->[$j+1]->end,
3540 Value => $value,
3541 Type => $type);
3542 }
3543 else {
3544
3545 # Here we know that the result won't just be the conglomeration of
3546 # a new range with both its adjacent neighbors. But it could
3547 # extend one of them.
3548
3549 if ($extends_below) {
3550
3551 # Here the new element adds to the one below, but not to the
3552 # one above. If inserting, and only to that one range, can
3553 # just change its ending to include the new one.
3554 if ($length == 0 && ! $cdm) {
3555 $r->[$i-1]->set_end($end);
3556 trace "inserted range extends range to below so it is now $r->[$i-1]" if main::DEBUG && $to_trace;
3557 return;
3558 }
3559 else {
3560 trace "Changing inserted range to start at ", sprintf("%04X", $r->[$i-1]->start), " instead of ", sprintf("%04X", $start) if main::DEBUG && $to_trace;
3561 $splice_start--; # start replace at element below
3562 $length++; # will replace the element below
3563 $start = $r->[$i-1]->start;
3564 }
3565 }
3566 elsif ($extends_above) {
3567
3568 # Here the new element adds to the one above, but not below.
3569 # Mirror the code above
3570 if ($length == 0 && ! $cdm) {
3571 $r->[$j+1]->set_start($start);
3572 trace "inserted range extends range to above so it is now $r->[$j+1]" if main::DEBUG && $to_trace;
3573 return;
3574 }
3575 else {
3576 trace "Changing inserted range to end at ", sprintf("%04X", $r->[$j+1]->end), " instead of ", sprintf("%04X", $end) if main::DEBUG && $to_trace;
3577 $length++; # will replace the element above
3578 $end = $r->[$j+1]->end;
3579 }
3580 }
3581
3582 trace "Range at $i is $r->[$i]" if main::DEBUG && $to_trace;
3583
3584 # Finally, here we know there will have to be a splice.
3585 # If the change or delete affects only the highest portion of the
3586 # first affected range, the range will have to be split. The
3587 # splice will remove the whole range, but will replace it by a new
3588 # range containing just the unaffected part. So, in this case,
3589 # add to the replacement list just this unaffected portion.
3590 if (! $extends_below
3591 && $start > $r->[$i]->start && $start <= $r->[$i]->end)
3592 {
3593 push @replacement,
3594 Range->new($r->[$i]->start,
3595 $start - 1,
3596 Value => $r->[$i]->value,
3597 Type => $r->[$i]->type);
3598 }
3599
3600 # In the case of an insert or change, but not a delete, we have to
3601 # put in the new stuff; this comes next.
3602 if ($operation eq '+') {
3603 push @replacement, Range->new($start,
3604 $end,
3605 Value => $value,
3606 Type => $type);
3607 }
3608
3609 trace "Range at $j is $r->[$j]" if main::DEBUG && $to_trace && $j != $i;
3610 #trace "$end >=", $r->[$j]->start, " && $end <", $r->[$j]->end if main::DEBUG && $to_trace;
3611
3612 # And finally, if we're changing or deleting only a portion of the
3613 # highest affected range, it must be split, as the lowest one was.
3614 if (! $extends_above
3615 && $j >= 0 # Remember that j can be -1 if before first
3616 # current element
3617 && $end >= $r->[$j]->start
3618 && $end < $r->[$j]->end)
3619 {
3620 push @replacement,
3621 Range->new($end + 1,
3622 $r->[$j]->end,
3623 Value => $r->[$j]->value,
3624 Type => $r->[$j]->type);
3625 }
3626 }
3627
3628 # And do the splice, as calculated above
3629 if (main::DEBUG && $to_trace) {
3630 trace "replacing $length element(s) at $i with ";
3631 foreach my $replacement (@replacement) {
3632 trace " $replacement";
3633 }
3634 trace "Before splice:";
3635 trace 'i-2=[', $i-2, ']', $r->[$i-2] if $i >= 2;
3636 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i >= 1;
3637 trace "i =[", $i, "]", $r->[$i];
3638 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < @$r - 1;
3639 trace 'i+2=[', $i+2, ']', $r->[$i+2] if $i < @$r - 2;
3640 }
3641
3642 my @return = splice @$r, $splice_start, $length, @replacement;
3643
3644 if (main::DEBUG && $to_trace) {
3645 trace "After splice:";
3646 trace 'i-2=[', $i-2, ']', $r->[$i-2] if $i >= 2;
3647 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i >= 1;
3648 trace "i =[", $i, "]", $r->[$i];
3649 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < @$r - 1;
3650 trace 'i+2=[', $i+2, ']', $r->[$i+2] if $i < @$r - 2;
3651 trace "removed @return";
3652 }
3653
3654 # An actual deletion could have changed the maximum in the list.
3655 # There was no deletion if the splice didn't return something, but
3656 # otherwise recalculate it. This is done too rarely to worry about
3657 # performance.
3658 if ($operation eq '-' && @return) {
3659 $max{$addr} = $r->[-1]->end;
3660 }
3661 return @return;
3662 }
3663
3664 sub reset_each_range { # reset the iterator for each_range();
3665 my $self = shift;
3666 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3667
f998e60c 3668 no overloading;
051df77b 3669 undef $each_range_iterator{pack 'J', $self};
99870f4d
KW
3670 return;
3671 }
3672
3673 sub each_range {
3674 # Iterate over each range in a range list. Results are undefined if
3675 # the range list is changed during the iteration.
3676
3677 my $self = shift;
3678 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3679
ffe43484 3680 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3681
3682 return if $self->is_empty;
3683
3684 $each_range_iterator{$addr} = -1
3685 if ! defined $each_range_iterator{$addr};
3686 $each_range_iterator{$addr}++;
3687 return $ranges{$addr}->[$each_range_iterator{$addr}]
3688 if $each_range_iterator{$addr} < @{$ranges{$addr}};
3689 undef $each_range_iterator{$addr};
3690 return;
3691 }
3692
3693 sub count { # Returns count of code points in range list
3694 my $self = shift;
3695 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3696
ffe43484 3697 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3698
3699 my $count = 0;
3700 foreach my $range (@{$ranges{$addr}}) {
3701 $count += $range->end - $range->start + 1;
3702 }
3703 return $count;
3704 }
3705
3706 sub delete_range { # Delete a range
3707 my $self = shift;
3708 my $start = shift;
3709 my $end = shift;
3710
3711 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3712
3713 return $self->_add_delete('-', $start, $end, "");
3714 }
3715
3716 sub is_empty { # Returns boolean as to if a range list is empty
3717 my $self = shift;
3718 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3719
f998e60c 3720 no overloading;
051df77b 3721 return scalar @{$ranges{pack 'J', $self}} == 0;
99870f4d
KW
3722 }
3723
3724 sub hash {
3725 # Quickly returns a scalar suitable for separating tables into
3726 # buckets, i.e. it is a hash function of the contents of a table, so
3727 # there are relatively few conflicts.
3728
3729 my $self = shift;
3730 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3731
ffe43484 3732 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3733
3734 # These are quickly computable. Return looks like 'min..max;count'
3735 return $self->min . "..$max{$addr};" . scalar @{$ranges{$addr}};
3736 }
3737} # End closure for _Range_List_Base
3738
3739package Range_List;
3740use base '_Range_List_Base';
3741
3742# A Range_List is a range list for match tables; i.e. the range values are
3743# not significant. Thus a number of operations can be safely added to it,
3744# such as inversion, intersection. Note that union is also an unsafe
3745# operation when range values are cared about, and that method is in the base
3746# class, not here. But things are set up so that that method is callable only
3747# during initialization. Only in this derived class, is there an operation
3748# that combines two tables. A Range_Map can thus be used to initialize a
3749# Range_List, and its mappings will be in the list, but are not significant to
3750# this class.
3751
3752sub trace { return main::trace(@_); }
3753
3754{ # Closure
3755
3756 use overload
3757 fallback => 0,
3758 '+' => sub { my $self = shift;
3759 my $other = shift;
3760
3761 return $self->_union($other)
3762 },
3763 '&' => sub { my $self = shift;
3764 my $other = shift;
3765
3766 return $self->_intersect($other, 0);
3767 },
3768 '~' => "_invert",
3769 '-' => "_subtract",
3770 ;
3771
3772 sub _invert {
3773 # Returns a new Range_List that gives all code points not in $self.
3774
3775 my $self = shift;
3776
3777 my $new = Range_List->new;
3778
3779 # Go through each range in the table, finding the gaps between them
3780 my $max = -1; # Set so no gap before range beginning at 0
3781 for my $range ($self->ranges) {
3782 my $start = $range->start;
3783 my $end = $range->end;
3784
3785 # If there is a gap before this range, the inverse will contain
3786 # that gap.
3787 if ($start > $max + 1) {
3788 $new->add_range($max + 1, $start - 1);
3789 }
3790 $max = $end;
3791 }
3792
3793 # And finally, add the gap from the end of the table to the max
3794 # possible code point
3795 if ($max < $LAST_UNICODE_CODEPOINT) {
3796 $new->add_range($max + 1, $LAST_UNICODE_CODEPOINT);
3797 }
3798 return $new;
3799 }
3800
3801 sub _subtract {
3802 # Returns a new Range_List with the argument deleted from it. The
3803 # argument can be a single code point, a range, or something that has
3804 # a range, with the _range_list() method on it returning them
3805
3806 my $self = shift;
3807 my $other = shift;
3808 my $reversed = shift;
3809 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3810
3811 if ($reversed) {
3812 Carp::my_carp_bug("Can't cope with a "
3813 . __PACKAGE__
3814 . " being the second parameter in a '-'. Subtraction ignored.");
3815 return $self;
3816 }
3817
3818 my $new = Range_List->new(Initialize => $self);
3819
3820 if (! ref $other) { # Single code point
3821 $new->delete_range($other, $other);
3822 }
3823 elsif ($other->isa('Range')) {
3824 $new->delete_range($other->start, $other->end);
3825 }
3826 elsif ($other->can('_range_list')) {
3827 foreach my $range ($other->_range_list->ranges) {
3828 $new->delete_range($range->start, $range->end);
3829 }
3830 }
3831 else {
3832 Carp::my_carp_bug("Can't cope with a "
3833 . ref($other)
3834 . " argument to '-'. Subtraction ignored."
3835 );
3836 return $self;
3837 }
3838
3839 return $new;
3840 }
3841
3842 sub _intersect {
3843 # Returns either a boolean giving whether the two inputs' range lists
3844 # intersect (overlap), or a new Range_List containing the intersection
3845 # of the two lists. The optional final parameter being true indicates
3846 # to do the check instead of the intersection.
3847
3848 my $a_object = shift;
3849 my $b_object = shift;
3850 my $check_if_overlapping = shift;
3851 $check_if_overlapping = 0 unless defined $check_if_overlapping;
3852 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3853
3854 if (! defined $b_object) {
3855 my $message = "";
3856 $message .= $a_object->_owner_name_of if defined $a_object;
3857 Carp::my_carp_bug($message .= "Called with undefined value. Intersection not done.");
3858 return;
3859 }
3860
3861 # a & b = !(!a | !b), or in our terminology = ~ ( ~a + -b )
3862 # Thus the intersection could be much more simply be written:
3863 # return ~(~$a_object + ~$b_object);
3864 # But, this is slower, and when taking the inverse of a large
3865 # range_size_1 table, back when such tables were always stored that
3866 # way, it became prohibitively slow, hence the code was changed to the
3867 # below
3868
3869 if ($b_object->isa('Range')) {
3870 $b_object = Range_List->new(Initialize => $b_object,
3871 Owner => $a_object->_owner_name_of);
3872 }
3873 $b_object = $b_object->_range_list if $b_object->can('_range_list');
3874
3875 my @a_ranges = $a_object->ranges;
3876 my @b_ranges = $b_object->ranges;
3877
3878 #local $to_trace = 1 if main::DEBUG;
3879 trace "intersecting $a_object with ", scalar @a_ranges, "ranges and $b_object with", scalar @b_ranges, " ranges" if main::DEBUG && $to_trace;
3880
3881 # Start with the first range in each list
3882 my $a_i = 0;
3883 my $range_a = $a_ranges[$a_i];
3884 my $b_i = 0;
3885 my $range_b = $b_ranges[$b_i];
3886
3887 my $new = __PACKAGE__->new(Owner => $a_object->_owner_name_of)
3888 if ! $check_if_overlapping;
3889
3890 # If either list is empty, there is no intersection and no overlap
3891 if (! defined $range_a || ! defined $range_b) {
3892 return $check_if_overlapping ? 0 : $new;
3893 }
3894 trace "range_a[$a_i]=$range_a; range_b[$b_i]=$range_b" if main::DEBUG && $to_trace;
3895
3896 # Otherwise, must calculate the intersection/overlap. Start with the
3897 # very first code point in each list
3898 my $a = $range_a->start;
3899 my $b = $range_b->start;
3900
3901 # Loop through all the ranges of each list; in each iteration, $a and
3902 # $b are the current code points in their respective lists
3903 while (1) {
3904
3905 # If $a and $b are the same code point, ...
3906 if ($a == $b) {
3907
3908 # it means the lists overlap. If just checking for overlap
3909 # know the answer now,
3910 return 1 if $check_if_overlapping;
3911
3912 # The intersection includes this code point plus anything else
3913 # common to both current ranges.
3914 my $start = $a;
3915 my $end = main::min($range_a->end, $range_b->end);
3916 if (! $check_if_overlapping) {
3917 trace "adding intersection range ", sprintf("%04X", $start) . ".." . sprintf("%04X", $end) if main::DEBUG && $to_trace;
3918 $new->add_range($start, $end);
3919 }
3920
3921 # Skip ahead to the end of the current intersect
3922 $a = $b = $end;
3923
3924 # If the current intersect ends at the end of either range (as
3925 # it must for at least one of them), the next possible one
3926 # will be the beginning code point in it's list's next range.
3927 if ($a == $range_a->end) {
3928 $range_a = $a_ranges[++$a_i];
3929 last unless defined $range_a;
3930 $a = $range_a->start;
3931 }
3932 if ($b == $range_b->end) {
3933 $range_b = $b_ranges[++$b_i];
3934 last unless defined $range_b;
3935 $b = $range_b->start;
3936 }
3937
3938 trace "range_a[$a_i]=$range_a; range_b[$b_i]=$range_b" if main::DEBUG && $to_trace;
3939 }
3940 elsif ($a < $b) {
3941
3942 # Not equal, but if the range containing $a encompasses $b,
3943 # change $a to be the middle of the range where it does equal
3944 # $b, so the next iteration will get the intersection
3945 if ($range_a->end >= $b) {
3946 $a = $b;
3947 }
3948 else {
3949
3950 # Here, the current range containing $a is entirely below
3951 # $b. Go try to find a range that could contain $b.
3952 $a_i = $a_object->_search_ranges($b);
3953
3954 # If no range found, quit.
3955 last unless defined $a_i;
3956
3957 # The search returns $a_i, such that
3958 # range_a[$a_i-1]->end < $b <= range_a[$a_i]->end
3959 # Set $a to the beginning of this new range, and repeat.
3960 $range_a = $a_ranges[$a_i];
3961 $a = $range_a->start;
3962 }
3963 }
3964 else { # Here, $b < $a.
3965
3966 # Mirror image code to the leg just above
3967 if ($range_b->end >= $a) {
3968 $b = $a;
3969 }
3970 else {
3971 $b_i = $b_object->_search_ranges($a);
3972 last unless defined $b_i;
3973 $range_b = $b_ranges[$b_i];
3974 $b = $range_b->start;
3975 }
3976 }
3977 } # End of looping through ranges.
3978
3979 # Intersection fully computed, or now know that there is no overlap
3980 return $check_if_overlapping ? 0 : $new;
3981 }
3982
3983 sub overlaps {
3984 # Returns boolean giving whether the two arguments overlap somewhere
3985
3986 my $self = shift;
3987 my $other = shift;
3988 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3989
3990 return $self->_intersect($other, 1);
3991 }
3992
3993 sub add_range {
3994 # Add a range to the list.
3995
3996 my $self = shift;
3997 my $start = shift;
3998 my $end = shift;
3999 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
4000
4001 return $self->_add_delete('+', $start, $end, "");
4002 }
4003
99870f4d
KW
4004 sub is_code_point_usable {
4005 # This used only for making the test script. See if the input
4006 # proposed trial code point is one that Perl will handle. If second
4007 # parameter is 0, it won't select some code points for various
4008 # reasons, noted below.
4009
4010 my $code = shift;
4011 my $try_hard = shift;
4012 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
4013
4014 return 0 if $code < 0; # Never use a negative
4015
99870f4d
KW
4016 # shun null. I'm (khw) not sure why this was done, but NULL would be
4017 # the character very frequently used.
4018 return $try_hard if $code == 0x0000;
4019
4020 retu