This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
mktables: White-space only
[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 19my $start_time;
98dc9551 20BEGIN { # Get the time the script started running; do it at compilation to
cdcef19a
KW
21 # get it as close as possible
22 $start_time= time;
23}
24
23e33b60 25require 5.010_001;
d73e5302 26use strict;
99870f4d 27use warnings;
cf25bb62 28use Carp;
bd9ebcfd 29use Config;
99870f4d
KW
30use File::Find;
31use File::Path;
d07a55ed 32use File::Spec;
99870f4d 33use Text::Tabs;
6b64c11c 34use re "/aa";
99870f4d
KW
35
36sub DEBUG () { 0 } # Set to 0 for production; 1 for development
bd9ebcfd 37my $debugging_build = $Config{"ccflags"} =~ /-DDEBUGGING/;
99870f4d 38
74cd47d0
KW
39sub NON_ASCII_PLATFORM { ord("A") != 65 }
40
99870f4d
KW
41##########################################################################
42#
43# mktables -- create the runtime Perl Unicode files (lib/unicore/.../*.pl),
44# from the Unicode database files (lib/unicore/.../*.txt), It also generates
232ed87f 45# a pod file and .t files, depending on option parameters.
99870f4d
KW
46#
47# The structure of this file is:
48# First these introductory comments; then
49# code needed for everywhere, such as debugging stuff; then
50# code to handle input parameters; then
51# data structures likely to be of external interest (some of which depend on
52# the input parameters, so follows them; then
53# more data structures and subroutine and package (class) definitions; then
54# the small actual loop to process the input files and finish up; then
55# a __DATA__ section, for the .t tests
56#
232ed87f
KW
57# This program works on all releases of Unicode so far. The outputs have been
58# scrutinized most intently for release 5.1. The others have been checked for
59# somewhat more than just sanity. It can handle all non-provisional Unicode
60# character properties in those releases.
99870f4d 61#
99870f4d
KW
62# This program is mostly about Unicode character (or code point) properties.
63# A property describes some attribute or quality of a code point, like if it
64# is lowercase or not, its name, what version of Unicode it was first defined
65# in, or what its uppercase equivalent is. Unicode deals with these disparate
66# possibilities by making all properties into mappings from each code point
67# into some corresponding value. In the case of it being lowercase or not,
68# the mapping is either to 'Y' or 'N' (or various synonyms thereof). Each
69# property maps each Unicode code point to a single value, called a "property
232ed87f
KW
70# value". (Some more recently defined properties, map a code point to a set
71# of values.)
99870f4d
KW
72#
73# When using a property in a regular expression, what is desired isn't the
74# mapping of the code point to its property's value, but the reverse (or the
75# mathematical "inverse relation"): starting with the property value, "Does a
76# code point map to it?" These are written in a "compound" form:
77# \p{property=value}, e.g., \p{category=punctuation}. This program generates
78# files containing the lists of code points that map to each such regular
79# expression property value, one file per list
80#
81# There is also a single form shortcut that Perl adds for many of the commonly
82# used properties. This happens for all binary properties, plus script,
83# general_category, and block properties.
84#
85# Thus the outputs of this program are files. There are map files, mostly in
86# the 'To' directory; and there are list files for use in regular expression
87# matching, all in subdirectories of the 'lib' directory, with each
88# subdirectory being named for the property that the lists in it are for.
89# Bookkeeping, test, and documentation files are also generated.
90
91my $matches_directory = 'lib'; # Where match (\p{}) files go.
92my $map_directory = 'To'; # Where map files go.
93
94# DATA STRUCTURES
95#
96# The major data structures of this program are Property, of course, but also
97# Table. There are two kinds of tables, very similar to each other.
98# "Match_Table" is the data structure giving the list of code points that have
99# a particular property value, mentioned above. There is also a "Map_Table"
100# data structure which gives the property's mapping from code point to value.
101# There are two structures because the match tables need to be combined in
102# various ways, such as constructing unions, intersections, complements, etc.,
103# and the map ones don't. And there would be problems, perhaps subtle, if
104# a map table were inadvertently operated on in some of those ways.
105# The use of separate classes with operations defined on one but not the other
106# prevents accidentally confusing the two.
107#
108# At the heart of each table's data structure is a "Range_List", which is just
109# an ordered list of "Ranges", plus ancillary information, and methods to
110# operate on them. A Range is a compact way to store property information.
111# Each range has a starting code point, an ending code point, and a value that
112# is meant to apply to all the code points between the two end points,
113# inclusive. For a map table, this value is the property value for those
114# code points. Two such ranges could be written like this:
115# 0x41 .. 0x5A, 'Upper',
116# 0x61 .. 0x7A, 'Lower'
117#
118# Each range also has a type used as a convenience to classify the values.
119# Most ranges in this program will be Type 0, or normal, but there are some
120# ranges that have a non-zero type. These are used only in map tables, and
121# are for mappings that don't fit into the normal scheme of things. Mappings
122# that require a hash entry to communicate with utf8.c are one example;
123# another example is mappings for charnames.pm to use which indicate a name
232ed87f 124# that is algorithmically determinable from its code point (and the reverse).
99870f4d
KW
125# These are used to significantly compact these tables, instead of listing
126# each one of the tens of thousands individually.
127#
128# In a match table, the value of a range is irrelevant (and hence the type as
129# well, which will always be 0), and arbitrarily set to the null string.
130# Using the example above, there would be two match tables for those two
131# entries, one named Upper would contain the 0x41..0x5A range, and the other
132# named Lower would contain 0x61..0x7A.
133#
134# Actually, there are two types of range lists, "Range_Map" is the one
135# associated with map tables, and "Range_List" with match tables.
232ed87f
KW
136# Again, this is so that methods can be defined on one and not the others so
137# as to prevent operating on them in incorrect ways.
99870f4d
KW
138#
139# Eventually, most tables are written out to files to be read by utf8_heavy.pl
140# in the perl core. All tables could in theory be written, but some are
141# suppressed because there is no current practical use for them. It is easy
142# to change which get written by changing various lists that are near the top
143# of the actual code in this file. The table data structures contain enough
144# ancillary information to allow them to be treated as separate entities for
145# writing, such as the path to each one's file. There is a heading in each
146# map table that gives the format of its entries, and what the map is for all
147# the code points missing from it. (This allows tables to be more compact.)
678f13d5 148#
99870f4d
KW
149# The Property data structure contains one or more tables. All properties
150# contain a map table (except the $perl property which is a
151# pseudo-property containing only match tables), and any properties that
152# are usable in regular expression matches also contain various matching
153# tables, one for each value the property can have. A binary property can
154# have two values, True and False (or Y and N, which are preferred by Unicode
155# terminology). Thus each of these properties will have a map table that
156# takes every code point and maps it to Y or N (but having ranges cuts the
157# number of entries in that table way down), and two match tables, one
158# which has a list of all the code points that map to Y, and one for all the
232ed87f 159# code points that map to N. (For each binary property, a third table is also
99870f4d 160# generated for the pseudo Perl property. It contains the identical code
232ed87f
KW
161# points as the Y table, but can be written in regular expressions, not in the
162# compound form, but in a "single" form like \p{IsUppercase}.) Many
163# properties are binary, but some properties have several possible values,
164# some have many, and properties like Name have a different value for every
165# named code point. Those will not, unless the controlling lists are changed,
166# have their match tables written out. But all the ones which can be used in
167# regular expression \p{} and \P{} constructs will. Prior to 5.14, generally
168# a property would have either its map table or its match tables written but
169# not both. Again, what gets written is controlled by lists which can easily
170# be changed. Starting in 5.14, advantage was taken of this, and all the map
171# tables needed to reconstruct the Unicode db are now written out, while
172# suppressing the Unicode .txt files that contain the data. Our tables are
173# much more compact than the .txt files, so a significant space savings was
174# achieved. Also, tables are not written out that are trivially derivable
175# from tables that do get written. So, there typically is no file containing
176# the code points not matched by a binary property (the table for \P{} versus
177# lowercase \p{}), since you just need to invert the True table to get the
178# False table.
179
180# Properties have a 'Type', like 'binary', or 'string', or 'enum' depending on
181# how many match tables there are and the content of the maps. This 'Type' is
c12f2655
KW
182# different than a range 'Type', so don't get confused by the two concepts
183# having the same name.
678f13d5 184#
99870f4d
KW
185# For information about the Unicode properties, see Unicode's UAX44 document:
186
187my $unicode_reference_url = 'http://www.unicode.org/reports/tr44/';
188
189# As stated earlier, this program will work on any release of Unicode so far.
190# Most obvious problems in earlier data have NOT been corrected except when
be864b6c 191# necessary to make Perl or this program work reasonably, and to keep out
232ed87f
KW
192# potential security issues. For example, no folding information was given in
193# early releases, so this program substitutes lower case instead, just so that
194# a regular expression with the /i option will do something that actually
195# gives the right results in many cases. There are also a couple other
196# corrections for version 1.1.5, commented at the point they are made. As an
197# example of corrections that weren't made (but could be) is this statement
198# from DerivedAge.txt: "The supplementary private use code points and the
199# non-character code points were assigned in version 2.0, but not specifically
200# listed in the UCD until versions 3.0 and 3.1 respectively." (To be precise
201# it was 3.0.1 not 3.0.0) More information on Unicode version glitches is
202# further down in these introductory comments.
99870f4d 203#
232ed87f
KW
204# This program works on all non-provisional properties as of the current
205# Unicode release, though the files for some are suppressed for various
206# reasons. You can change which are output by changing lists in this program.
678f13d5 207#
dc85bd38 208# The old version of mktables emphasized the term "Fuzzy" to mean Unicode's
99870f4d
KW
209# loose matchings rules (from Unicode TR18):
210#
211# The recommended names for UCD properties and property values are in
212# PropertyAliases.txt [Prop] and PropertyValueAliases.txt
213# [PropValue]. There are both abbreviated names and longer, more
214# descriptive names. It is strongly recommended that both names be
215# recognized, and that loose matching of property names be used,
216# whereby the case distinctions, whitespace, hyphens, and underbar
217# are ignored.
232ed87f 218#
99870f4d
KW
219# The program still allows Fuzzy to override its determination of if loose
220# matching should be used, but it isn't currently used, as it is no longer
221# needed; the calculations it makes are good enough.
678f13d5 222#
99870f4d
KW
223# SUMMARY OF HOW IT WORKS:
224#
225# Process arguments
226#
227# A list is constructed containing each input file that is to be processed
228#
229# Each file on the list is processed in a loop, using the associated handler
230# code for each:
231# The PropertyAliases.txt and PropValueAliases.txt files are processed
232# first. These files name the properties and property values.
233# Objects are created of all the property and property value names
234# that the rest of the input should expect, including all synonyms.
235# The other input files give mappings from properties to property
236# values. That is, they list code points and say what the mapping
237# is under the given property. Some files give the mappings for
238# just one property; and some for many. This program goes through
232ed87f
KW
239# each file and populates the properties and their map tables from
240# them. Some properties are listed in more than one file, and
241# Unicode has set up a precedence as to which has priority if there
242# is a conflict. Thus the order of processing matters, and this
243# program handles the conflict possibility by processing the
244# overriding input files last, so that if necessary they replace
245# earlier values.
99870f4d
KW
246# After this is all done, the program creates the property mappings not
247# furnished by Unicode, but derivable from what it does give.
248# The tables of code points that match each property value in each
249# property that is accessible by regular expressions are created.
250# The Perl-defined properties are created and populated. Many of these
251# require data determined from the earlier steps
252# Any Perl-defined synonyms are created, and name clashes between Perl
678f13d5 253# and Unicode are reconciled and warned about.
99870f4d
KW
254# All the properties are written to files
255# Any other files are written, and final warnings issued.
678f13d5 256#
99870f4d
KW
257# For clarity, a number of operators have been overloaded to work on tables:
258# ~ means invert (take all characters not in the set). The more
259# conventional '!' is not used because of the possibility of confusing
260# it with the actual boolean operation.
261# + means union
262# - means subtraction
263# & means intersection
264# The precedence of these is the order listed. Parentheses should be
265# copiously used. These are not a general scheme. The operations aren't
266# defined for a number of things, deliberately, to avoid getting into trouble.
267# Operations are done on references and affect the underlying structures, so
268# that the copy constructors for them have been overloaded to not return a new
269# clone, but the input object itself.
678f13d5 270#
99870f4d
KW
271# The bool operator is deliberately not overloaded to avoid confusion with
272# "should it mean if the object merely exists, or also is non-empty?".
99870f4d
KW
273#
274# WHY CERTAIN DESIGN DECISIONS WERE MADE
678f13d5
KW
275#
276# This program needs to be able to run under miniperl. Therefore, it uses a
277# minimum of other modules, and hence implements some things itself that could
278# be gotten from CPAN
279#
280# This program uses inputs published by the Unicode Consortium. These can
281# change incompatibly between releases without the Perl maintainers realizing
282# it. Therefore this program is now designed to try to flag these. It looks
283# at the directories where the inputs are, and flags any unrecognized files.
284# It keeps track of all the properties in the files it handles, and flags any
285# that it doesn't know how to handle. It also flags any input lines that
286# don't match the expected syntax, among other checks.
287#
288# It is also designed so if a new input file matches one of the known
289# templates, one hopefully just needs to add it to a list to have it
290# processed.
291#
292# As mentioned earlier, some properties are given in more than one file. In
293# particular, the files in the extracted directory are supposedly just
294# reformattings of the others. But they contain information not easily
295# derivable from the other files, including results for Unihan, which this
296# program doesn't ordinarily look at, and for unassigned code points. They
297# also have historically had errors or been incomplete. In an attempt to
298# create the best possible data, this program thus processes them first to
299# glean information missing from the other files; then processes those other
300# files to override any errors in the extracted ones. Much of the design was
301# driven by this need to store things and then possibly override them.
302#
303# It tries to keep fatal errors to a minimum, to generate something usable for
304# testing purposes. It always looks for files that could be inputs, and will
305# warn about any that it doesn't know how to handle (the -q option suppresses
306# the warning).
99870f4d 307#
678f13d5
KW
308# Why is there more than one type of range?
309# This simplified things. There are some very specialized code points that
310# have to be handled specially for output, such as Hangul syllable names.
311# By creating a range type (done late in the development process), it
312# allowed this to be stored with the range, and overridden by other input.
313# Originally these were stored in another data structure, and it became a
314# mess trying to decide if a second file that was for the same property was
315# overriding the earlier one or not.
316#
317# Why are there two kinds of tables, match and map?
318# (And there is a base class shared by the two as well.) As stated above,
319# they actually are for different things. Development proceeded much more
320# smoothly when I (khw) realized the distinction. Map tables are used to
321# give the property value for every code point (actually every code point
322# that doesn't map to a default value). Match tables are used for regular
323# expression matches, and are essentially the inverse mapping. Separating
324# the two allows more specialized methods, and error checks so that one
325# can't just take the intersection of two map tables, for example, as that
326# is nonsensical.
99870f4d 327#
232ed87f
KW
328# What about 'fate' and 'status'. The concept of a table's fate was created
329# late when it became clear that something more was needed. The difference
330# between this and 'status' is unclean, and could be improved if someone
331# wanted to spend the effort.
332#
23e33b60
KW
333# DEBUGGING
334#
678f13d5
KW
335# This program is written so it will run under miniperl. Occasionally changes
336# will cause an error where the backtrace doesn't work well under miniperl.
337# To diagnose the problem, you can instead run it under regular perl, if you
338# have one compiled.
339#
340# There is a good trace facility. To enable it, first sub DEBUG must be set
341# to return true. Then a line like
342#
343# local $to_trace = 1 if main::DEBUG;
344#
232ed87f
KW
345# can be added to enable tracing in its lexical scope (plus dynamic) or until
346# you insert another line:
678f13d5
KW
347#
348# local $to_trace = 0 if main::DEBUG;
349#
232ed87f 350# To actually trace, use a line like "trace $a, @b, %c, ...;
678f13d5
KW
351#
352# Some of the more complex subroutines already have trace statements in them.
353# Permanent trace statements should be like:
354#
355# trace ... if main::DEBUG && $to_trace;
356#
357# If there is just one or a few files that you're debugging, you can easily
358# cause most everything else to be skipped. Change the line
359#
360# my $debug_skip = 0;
361#
362# to 1, and every file whose object is in @input_file_objects and doesn't have
232ed87f
KW
363# a, 'non_skip => 1,' in its constructor will be skipped. However, skipping
364# Jamo.txt or UnicodeData.txt will likely cause fatal errors.
678f13d5 365#
b4a0206c 366# To compare the output tables, it may be useful to specify the -annotate
97a8a595
KW
367# flag. (As of this writing, this can't be done on a clean workspace, due to
368# requirements in Text::Tabs used in this option; so first run mktables
369# without this option.) This option adds comment lines to each table, one for
370# each non-algorithmically named character giving, currently its code point,
371# name, and graphic representation if printable (and you have a font that
372# knows about it). This makes it easier to see what the particular code
373# points are in each output table. Non-named code points are annotated with a
374# description of their status, and contiguous ones with the same description
375# will be output as a range rather than individually. Algorithmically named
376# characters are also output as ranges, except when there are just a few
377# contiguous ones.
c4019d52 378#
99870f4d
KW
379# FUTURE ISSUES
380#
381# The program would break if Unicode were to change its names so that
382# interior white space, underscores, or dashes differences were significant
383# within property and property value names.
384#
385# It might be easier to use the xml versions of the UCD if this program ever
386# would need heavy revision, and the ability to handle old versions was not
387# required.
388#
389# There is the potential for name collisions, in that Perl has chosen names
390# that Unicode could decide it also likes. There have been such collisions in
391# the past, with mostly Perl deciding to adopt the Unicode definition of the
392# name. However in the 5.2 Unicode beta testing, there were a number of such
393# collisions, which were withdrawn before the final release, because of Perl's
394# and other's protests. These all involved new properties which began with
395# 'Is'. Based on the protests, Unicode is unlikely to try that again. Also,
396# many of the Perl-defined synonyms, like Any, Word, etc, are listed in a
397# Unicode document, so they are unlikely to be used by Unicode for another
398# purpose. However, they might try something beginning with 'In', or use any
399# of the other Perl-defined properties. This program will warn you of name
400# collisions, and refuse to generate tables with them, but manual intervention
401# will be required in this event. One scheme that could be implemented, if
402# necessary, would be to have this program generate another file, or add a
403# field to mktables.lst that gives the date of first definition of a property.
404# Each new release of Unicode would use that file as a basis for the next
405# iteration. And the Perl synonym addition code could sort based on the age
406# of the property, so older properties get priority, and newer ones that clash
407# would be refused; hence existing code would not be impacted, and some other
408# synonym would have to be used for the new property. This is ugly, and
409# manual intervention would certainly be easier to do in the short run; lets
410# hope it never comes to this.
678f13d5 411#
99870f4d
KW
412# A NOTE ON UNIHAN
413#
414# This program can generate tables from the Unihan database. But it doesn't
415# by default, letting the CPAN module Unicode::Unihan handle them. Prior to
416# version 5.2, this database was in a single file, Unihan.txt. In 5.2 the
417# database was split into 8 different files, all beginning with the letters
418# 'Unihan'. This program will read those file(s) if present, but it needs to
419# know which of the many properties in the file(s) should have tables created
420# for them. It will create tables for any properties listed in
421# PropertyAliases.txt and PropValueAliases.txt, plus any listed in the
422# @cjk_properties array and the @cjk_property_values array. Thus, if a
423# property you want is not in those files of the release you are building
424# against, you must add it to those two arrays. Starting in 4.0, the
425# Unicode_Radical_Stroke was listed in those files, so if the Unihan database
426# is present in the directory, a table will be generated for that property.
427# In 5.2, several more properties were added. For your convenience, the two
5f7264c7 428# arrays are initialized with all the 6.0 listed properties that are also in
99870f4d
KW
429# earlier releases. But these are commented out. You can just uncomment the
430# ones you want, or use them as a template for adding entries for other
431# properties.
432#
433# You may need to adjust the entries to suit your purposes. setup_unihan(),
434# and filter_unihan_line() are the functions where this is done. This program
435# already does some adjusting to make the lines look more like the rest of the
436# Unicode DB; You can see what that is in filter_unihan_line()
437#
438# There is a bug in the 3.2 data file in which some values for the
439# kPrimaryNumeric property have commas and an unexpected comment. A filter
440# could be added for these; or for a particular installation, the Unihan.txt
441# file could be edited to fix them.
99870f4d 442#
678f13d5
KW
443# HOW TO ADD A FILE TO BE PROCESSED
444#
445# A new file from Unicode needs to have an object constructed for it in
446# @input_file_objects, probably at the end or at the end of the extracted
447# ones. The program should warn you if its name will clash with others on
448# restrictive file systems, like DOS. If so, figure out a better name, and
449# add lines to the README.perl file giving that. If the file is a character
232ed87f 450# property, it should be in the format that Unicode has implicitly
678f13d5
KW
451# standardized for such files for the more recently introduced ones.
452# If so, the Input_file constructor for @input_file_objects can just be the
453# file name and release it first appeared in. If not, then it should be
454# possible to construct an each_line_handler() to massage the line into the
455# standardized form.
456#
457# For non-character properties, more code will be needed. You can look at
458# the existing entries for clues.
459#
460# UNICODE VERSIONS NOTES
461#
462# The Unicode UCD has had a number of errors in it over the versions. And
463# these remain, by policy, in the standard for that version. Therefore it is
464# risky to correct them, because code may be expecting the error. So this
465# program doesn't generally make changes, unless the error breaks the Perl
466# core. As an example, some versions of 2.1.x Jamo.txt have the wrong value
467# for U+1105, which causes real problems for the algorithms for Jamo
468# calculations, so it is changed here.
469#
470# But it isn't so clear cut as to what to do about concepts that are
471# introduced in a later release; should they extend back to earlier releases
472# where the concept just didn't exist? It was easier to do this than to not,
473# so that's what was done. For example, the default value for code points not
474# in the files for various properties was probably undefined until changed by
475# some version. No_Block for blocks is such an example. This program will
476# assign No_Block even in Unicode versions that didn't have it. This has the
477# benefit that code being written doesn't have to special case earlier
478# versions; and the detriment that it doesn't match the Standard precisely for
479# the affected versions.
480#
481# Here are some observations about some of the issues in early versions:
482#
232ed87f
KW
483# Prior to version 3.0, there were 3 character decompositions. These are not
484# handled by Unicode::Normalize, nor will it compile when presented a version
485# that has them. However, you can trivially get it to compile by simply
486# ignoring those decompositions, by changing the croak to a carp. At the time
487# of this writing, the line (in cpan/Unicode-Normalize/mkheader) reads
488#
489# croak("Weird Canonical Decomposition of U+$h");
490#
28807e1d
KW
491# Simply comment it out. It will compile, but will not know about any three
492# character decompositions. If using the .pm version, there is a similar
493# line.
232ed87f
KW
494
495# The number of code points in \p{alpha=True} halved in 2.1.9. It turns out
496# that the reason is that the CJK block starting at 4E00 was removed from
497# PropList, and was not put back in until 3.1.0. The Perl extension (the
498# single property name \p{alpha}) has the correct values. But the compound
499# form is simply not generated until 3.1, as it can be argued that prior to
500# this release, this was not an official property. The comments for
501# filter_old_style_proplist() give more details.
678f13d5
KW
502#
503# Unicode introduced the synonym Space for White_Space in 4.1. Perl has
504# always had a \p{Space}. In release 3.2 only, they are not synonymous. The
505# reason is that 3.2 introduced U+205F=medium math space, which was not
506# classed as white space, but Perl figured out that it should have been. 4.0
507# reclassified it correctly.
508#
509# Another change between 3.2 and 4.0 is the CCC property value ATBL. In 3.2
232ed87f
KW
510# this was erroneously a synonym for 202 (it should be 200). In 4.0, ATB
511# became 202, and ATBL was left with no code points, as all the ones that
512# mapped to 202 stayed mapped to 202. Thus if your program used the numeric
513# name for the class, it would not have been affected, but if it used the
514# mnemonic, it would have been.
678f13d5
KW
515#
516# \p{Script=Hrkt} (Katakana_Or_Hiragana) came in 4.0.1. Before that code
517# points which eventually came to have this script property value, instead
518# mapped to "Unknown". But in the next release all these code points were
519# moved to \p{sc=common} instead.
99870f4d
KW
520#
521# The default for missing code points for BidiClass is complicated. Starting
522# in 3.1.1, the derived file DBidiClass.txt handles this, but this program
523# tries to do the best it can for earlier releases. It is done in
524# process_PropertyAliases()
525#
232ed87f
KW
526# In version 2.1.2, the entry in UnicodeData.txt:
527# 0275;LATIN SMALL LETTER BARRED O;Ll;0;L;;;;;N;;;;019F;
528# should instead be
529# 0275;LATIN SMALL LETTER BARRED O;Ll;0;L;;;;;N;;;019F;;019F
530# Without this change, there are casing problems for this character.
531#
7803ad2d
KW
532# Search for $string_compare_versions to see how to compare changes to
533# properties between Unicode versions
534#
99870f4d
KW
535##############################################################################
536
537my $UNDEF = ':UNDEF:'; # String to print out for undefined values in tracing
538 # and errors
539my $MAX_LINE_WIDTH = 78;
540
541# Debugging aid to skip most files so as to not be distracted by them when
542# concentrating on the ones being debugged. Add
543# non_skip => 1,
544# to the constructor for those files you want processed when you set this.
545# Files with a first version number of 0 are special: they are always
c12f2655
KW
546# processed regardless of the state of this flag. Generally, Jamo.txt and
547# UnicodeData.txt must not be skipped if you want this program to not die
548# before normal completion.
99870f4d
KW
549my $debug_skip = 0;
550
e9c4b4f8
KW
551
552# Normally these are suppressed.
553my $write_Unicode_deprecated_tables = 0;
554
99870f4d
KW
555# Set to 1 to enable tracing.
556our $to_trace = 0;
557
558{ # Closure for trace: debugging aid
559 my $print_caller = 1; # ? Include calling subroutine name
560 my $main_with_colon = 'main::';
561 my $main_colon_length = length($main_with_colon);
562
563 sub trace {
564 return unless $to_trace; # Do nothing if global flag not set
565
566 my @input = @_;
567
568 local $DB::trace = 0;
569 $DB::trace = 0; # Quiet 'used only once' message
570
571 my $line_number;
572
573 # Loop looking up the stack to get the first non-trace caller
574 my $caller_line;
575 my $caller_name;
576 my $i = 0;
577 do {
578 $line_number = $caller_line;
579 (my $pkg, my $file, $caller_line, my $caller) = caller $i++;
580 $caller = $main_with_colon unless defined $caller;
581
582 $caller_name = $caller;
583
584 # get rid of pkg
585 $caller_name =~ s/.*:://;
586 if (substr($caller_name, 0, $main_colon_length)
587 eq $main_with_colon)
588 {
589 $caller_name = substr($caller_name, $main_colon_length);
590 }
591
592 } until ($caller_name ne 'trace');
593
594 # If the stack was empty, we were called from the top level
595 $caller_name = 'main' if ($caller_name eq ""
596 || $caller_name eq 'trace');
597
598 my $output = "";
599 foreach my $string (@input) {
600 #print STDERR __LINE__, ": ", join ", ", @input, "\n";
601 if (ref $string eq 'ARRAY' || ref $string eq 'HASH') {
602 $output .= simple_dumper($string);
603 }
604 else {
605 $string = "$string" if ref $string;
606 $string = $UNDEF unless defined $string;
607 chomp $string;
608 $string = '""' if $string eq "";
609 $output .= " " if $output ne ""
610 && $string ne ""
611 && substr($output, -1, 1) ne " "
612 && substr($string, 0, 1) ne " ";
613 $output .= $string;
614 }
615 }
616
99f78760
KW
617 print STDERR sprintf "%4d: ", $line_number if defined $line_number;
618 print STDERR "$caller_name: " if $print_caller;
99870f4d
KW
619 print STDERR $output, "\n";
620 return;
621 }
622}
623
624# This is for a rarely used development feature that allows you to compare two
625# versions of the Unicode standard without having to deal with changes caused
c12f2655
KW
626# by the code points introduced in the later version. Change the 0 to a
627# string containing a SINGLE dotted Unicode release number (e.g. "2.1"). Only
628# code points introduced in that release and earlier will be used; later ones
629# are thrown away. You use the version number of the earliest one you want to
630# compare; then run this program on directory structures containing each
631# release, and compare the outputs. These outputs will therefore include only
632# the code points common to both releases, and you can see the changes caused
633# just by the underlying release semantic changes. For versions earlier than
634# 3.2, you must copy a version of DAge.txt into the directory.
635my $string_compare_versions = DEBUG && 0; # e.g., "2.1";
99870f4d
KW
636my $compare_versions = DEBUG
637 && $string_compare_versions
638 && pack "C*", split /\./, $string_compare_versions;
639
640sub uniques {
641 # Returns non-duplicated input values. From "Perl Best Practices:
642 # Encapsulated Cleverness". p. 455 in first edition.
643
644 my %seen;
0e407844
NC
645 # Arguably this breaks encapsulation, if the goal is to permit multiple
646 # distinct objects to stringify to the same value, and be interchangeable.
647 # However, for this program, no two objects stringify identically, and all
648 # lists passed to this function are either objects or strings. So this
649 # doesn't affect correctness, but it does give a couple of percent speedup.
650 no overloading;
99870f4d
KW
651 return grep { ! $seen{$_}++ } @_;
652}
653
654$0 = File::Spec->canonpath($0);
655
656my $make_test_script = 0; # ? Should we output a test script
6b5ab373 657my $make_norm_test_script = 0; # ? Should we output a normalization test script
99870f4d
KW
658my $write_unchanged_files = 0; # ? Should we update the output files even if
659 # we don't think they have changed
660my $use_directory = ""; # ? Should we chdir somewhere.
661my $pod_directory; # input directory to store the pod file.
662my $pod_file = 'perluniprops';
663my $t_path; # Path to the .t test file
664my $file_list = 'mktables.lst'; # File to store input and output file names.
665 # This is used to speed up the build, by not
666 # executing the main body of the program if
667 # nothing on the list has changed since the
668 # previous build
669my $make_list = 1; # ? Should we write $file_list. Set to always
670 # make a list so that when the pumpking is
671 # preparing a release, s/he won't have to do
672 # special things
673my $glob_list = 0; # ? Should we try to include unknown .txt files
674 # in the input.
bd9ebcfd
KW
675my $output_range_counts = $debugging_build; # ? Should we include the number
676 # of code points in ranges in
677 # the output
558712cf 678my $annotate = 0; # ? Should character names be in the output
9ef2b94f 679
99870f4d
KW
680# Verbosity levels; 0 is quiet
681my $NORMAL_VERBOSITY = 1;
682my $PROGRESS = 2;
683my $VERBOSE = 3;
684
685my $verbosity = $NORMAL_VERBOSITY;
686
0458fbc1
KW
687# Stored in mktables.lst so that if this program is called with different
688# options, will regenerate even if the files otherwise look like they're
689# up-to-date.
690my $command_line_arguments = join " ", @ARGV;
691
99870f4d
KW
692# Process arguments
693while (@ARGV) {
cf25bb62
JH
694 my $arg = shift @ARGV;
695 if ($arg eq '-v') {
99870f4d
KW
696 $verbosity = $VERBOSE;
697 }
698 elsif ($arg eq '-p') {
699 $verbosity = $PROGRESS;
700 $| = 1; # Flush buffers as we go.
701 }
702 elsif ($arg eq '-q') {
703 $verbosity = 0;
704 }
705 elsif ($arg eq '-w') {
706 $write_unchanged_files = 1; # update the files even if havent changed
707 }
708 elsif ($arg eq '-check') {
6ae7e459
YO
709 my $this = shift @ARGV;
710 my $ok = shift @ARGV;
711 if ($this ne $ok) {
712 print "Skipping as check params are not the same.\n";
713 exit(0);
714 }
00a8df5c 715 }
99870f4d
KW
716 elsif ($arg eq '-P' && defined ($pod_directory = shift)) {
717 -d $pod_directory or croak "Directory '$pod_directory' doesn't exist";
718 }
3df51b85
KW
719 elsif ($arg eq '-maketest' || ($arg eq '-T' && defined ($t_path = shift)))
720 {
99870f4d 721 $make_test_script = 1;
99870f4d 722 }
6b5ab373
KW
723 elsif ($arg eq '-makenormtest')
724 {
725 $make_norm_test_script = 1;
726 }
99870f4d
KW
727 elsif ($arg eq '-makelist') {
728 $make_list = 1;
729 }
730 elsif ($arg eq '-C' && defined ($use_directory = shift)) {
731 -d $use_directory or croak "Unknown directory '$use_directory'";
732 }
733 elsif ($arg eq '-L') {
734
735 # Existence not tested until have chdir'd
736 $file_list = shift;
737 }
738 elsif ($arg eq '-globlist') {
739 $glob_list = 1;
740 }
741 elsif ($arg eq '-c') {
742 $output_range_counts = ! $output_range_counts
743 }
b4a0206c 744 elsif ($arg eq '-annotate') {
558712cf 745 $annotate = 1;
bd9ebcfd
KW
746 $debugging_build = 1;
747 $output_range_counts = 1;
9ef2b94f 748 }
99870f4d
KW
749 else {
750 my $with_c = 'with';
751 $with_c .= 'out' if $output_range_counts; # Complements the state
752 croak <<END;
753usage: $0 [-c|-p|-q|-v|-w] [-C dir] [-L filelist] [ -P pod_dir ]
754 [ -T test_file_path ] [-globlist] [-makelist] [-maketest]
755 [-check A B ]
756 -c : Output comments $with_c number of code points in ranges
757 -q : Quiet Mode: Only output serious warnings.
758 -p : Set verbosity level to normal plus show progress.
759 -v : Set Verbosity level high: Show progress and non-serious
760 warnings
761 -w : Write files regardless
762 -C dir : Change to this directory before proceeding. All relative paths
763 except those specified by the -P and -T options will be done
764 with respect to this directory.
765 -P dir : Output $pod_file file to directory 'dir'.
3df51b85 766 -T path : Create a test script as 'path'; overrides -maketest
99870f4d
KW
767 -L filelist : Use alternate 'filelist' instead of standard one
768 -globlist : Take as input all non-Test *.txt files in current and sub
769 directories
3df51b85
KW
770 -maketest : Make test script 'TestProp.pl' in current (or -C directory),
771 overrides -T
99870f4d 772 -makelist : Rewrite the file list $file_list based on current setup
b4a0206c 773 -annotate : Output an annotation for each character in the table files;
97a8a595
KW
774 useful for debugging mktables, looking at diffs; but is slow
775 and memory intensive
99870f4d
KW
776 -check A B : Executes $0 only if A and B are the same
777END
778 }
779}
780
781# Stores the most-recently changed file. If none have changed, can skip the
782# build
aeab6150 783my $most_recent = (stat $0)[9]; # Do this before the chdir!
99870f4d
KW
784
785# Change directories now, because need to read 'version' early.
786if ($use_directory) {
3df51b85 787 if ($pod_directory && ! File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d
KW
788 $pod_directory = File::Spec->rel2abs($pod_directory);
789 }
3df51b85 790 if ($t_path && ! File::Spec->file_name_is_absolute($t_path)) {
99870f4d 791 $t_path = File::Spec->rel2abs($t_path);
00a8df5c 792 }
99870f4d 793 chdir $use_directory or croak "Failed to chdir to '$use_directory':$!";
3df51b85 794 if ($pod_directory && File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d 795 $pod_directory = File::Spec->abs2rel($pod_directory);
02b1aeec 796 }
3df51b85 797 if ($t_path && File::Spec->file_name_is_absolute($t_path)) {
99870f4d 798 $t_path = File::Spec->abs2rel($t_path);
02b1aeec 799 }
00a8df5c
YO
800}
801
99870f4d
KW
802# Get Unicode version into regular and v-string. This is done now because
803# various tables below get populated based on it. These tables are populated
804# here to be near the top of the file, and so easily seeable by those needing
805# to modify things.
806open my $VERSION, "<", "version"
807 or croak "$0: can't open required file 'version': $!\n";
808my $string_version = <$VERSION>;
809close $VERSION;
810chomp $string_version;
811my $v_version = pack "C*", split /\./, $string_version; # v string
812
813# The following are the complete names of properties with property values that
814# are known to not match any code points in some versions of Unicode, but that
815# may change in the future so they should be matchable, hence an empty file is
816# generated for them.
a9c9e371
KW
817my @tables_that_may_be_empty;
818push @tables_that_may_be_empty, 'Joining_Type=Left_Joining'
819 if $v_version lt v6.3.0;
99870f4d
KW
820push @tables_that_may_be_empty, 'Script=Common' if $v_version le v4.0.1;
821push @tables_that_may_be_empty, 'Title' if $v_version lt v2.0.0;
822push @tables_that_may_be_empty, 'Script=Katakana_Or_Hiragana'
823 if $v_version ge v4.1.0;
82aed44a
KW
824push @tables_that_may_be_empty, 'Script_Extensions=Katakana_Or_Hiragana'
825 if $v_version ge v6.0.0;
f583b44c
KW
826push @tables_that_may_be_empty, 'Grapheme_Cluster_Break=Prepend'
827 if $v_version ge v6.1.0;
1e958ea9
KW
828push @tables_that_may_be_empty, 'Canonical_Combining_Class=CCC133'
829 if $v_version ge v6.2.0;
99870f4d
KW
830
831# The lists below are hashes, so the key is the item in the list, and the
832# value is the reason why it is in the list. This makes generation of
833# documentation easier.
834
835my %why_suppressed; # No file generated for these.
836
837# Files aren't generated for empty extraneous properties. This is arguable.
838# Extraneous properties generally come about because a property is no longer
839# used in a newer version of Unicode. If we generated a file without code
840# points, programs that used to work on that property will still execute
841# without errors. It just won't ever match (or will always match, with \P{}).
842# This means that the logic is now likely wrong. I (khw) think its better to
843# find this out by getting an error message. Just move them to the table
844# above to change this behavior
845my %why_suppress_if_empty_warn_if_not = (
846
847 # It is the only property that has ever officially been removed from the
848 # Standard. The database never contained any code points for it.
849 'Special_Case_Condition' => 'Obsolete',
850
851 # Apparently never official, but there were code points in some versions of
852 # old-style PropList.txt
853 'Non_Break' => 'Obsolete',
854);
855
856# These would normally go in the warn table just above, but they were changed
857# a long time before this program was written, so warnings about them are
858# moot.
859if ($v_version gt v3.2.0) {
860 push @tables_that_may_be_empty,
861 'Canonical_Combining_Class=Attached_Below_Left'
862}
863
5f7264c7 864# These are listed in the Property aliases file in 6.0, but Unihan is ignored
99870f4d
KW
865# unless explicitly added.
866if ($v_version ge v5.2.0) {
867 my $unihan = 'Unihan; remove from list if using Unihan';
ea25a9b2 868 foreach my $table (qw (
99870f4d
KW
869 kAccountingNumeric
870 kOtherNumeric
871 kPrimaryNumeric
872 kCompatibilityVariant
873 kIICore
874 kIRG_GSource
875 kIRG_HSource
876 kIRG_JSource
877 kIRG_KPSource
878 kIRG_MSource
879 kIRG_KSource
880 kIRG_TSource
881 kIRG_USource
882 kIRG_VSource
883 kRSUnicode
ea25a9b2 884 ))
99870f4d
KW
885 {
886 $why_suppress_if_empty_warn_if_not{$table} = $unihan;
887 }
ca12659b
NC
888}
889
272501f6
KW
890# Enum values for to_output_map() method in the Map_Table package.
891my $EXTERNAL_MAP = 1;
892my $INTERNAL_MAP = 2;
ce712c88 893my $OUTPUT_ADJUSTED = 3;
272501f6 894
fcf1973c
KW
895# To override computed values for writing the map tables for these properties.
896# The default for enum map tables is to write them out, so that the Unicode
897# .txt files can be removed, but all the data to compute any property value
898# for any code point is available in a more compact form.
899my %global_to_output_map = (
900 # Needed by UCD.pm, but don't want to publicize that it exists, so won't
c12f2655
KW
901 # get stuck supporting it if things change. Since it is a STRING
902 # property, it normally would be listed in the pod, but INTERNAL_MAP
903 # suppresses that.
fcf1973c
KW
904 Unicode_1_Name => $INTERNAL_MAP,
905
906 Present_In => 0, # Suppress, as easily computed from Age
74cd47d0
KW
907 Block => (NON_ASCII_PLATFORM) ? 1 : 0, # Suppress, as Blocks.txt is
908 # retained, but needed for
909 # non-ASCII
53d34b6c
KW
910
911 # Suppress, as mapping can be found instead from the
912 # Perl_Decomposition_Mapping file
913 Decomposition_Type => 0,
fcf1973c
KW
914);
915
99870f4d 916# Properties that this program ignores.
230e0c16
KW
917my @unimplemented_properties;
918
919# With this release, it is automatically handled if the Unihan db is
920# downloaded
431c1f00 921push @unimplemented_properties, 'Unicode_Radical_Stroke' if $v_version lt v5.2.0;
d73e5302 922
99870f4d
KW
923# There are several types of obsolete properties defined by Unicode. These
924# must be hand-edited for every new Unicode release.
925my %why_deprecated; # Generates a deprecated warning message if used.
926my %why_stabilized; # Documentation only
927my %why_obsolete; # Documentation only
928
929{ # Closure
8364f9eb 930 my $simple = 'Perl uses the more complete version';
99870f4d
KW
931 my $unihan = 'Unihan properties are by default not enabled in the Perl core. Instead use CPAN: Unicode::Unihan';
932
933 my $other_properties = 'other properties';
934 my $contributory = "Used by Unicode internally for generating $other_properties and not intended to be used stand-alone";
5d294d41 935 my $why_no_expand = "Deprecated by Unicode. These are characters that expand to more than one character in the specified normalization form, but whether they actually take up more bytes or not depends on the encoding being used. For example, a UTF-8 encoded character may expand to a different number of bytes than a UTF-32 encoded character.";
99870f4d
KW
936
937 %why_deprecated = (
5f7264c7 938 'Grapheme_Link' => 'Deprecated by Unicode: Duplicates ccc=vr (Canonical_Combining_Class=Virama)',
99870f4d
KW
939 'Jamo_Short_Name' => $contributory,
940 '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',
941 'Other_Alphabetic' => $contributory,
942 'Other_Default_Ignorable_Code_Point' => $contributory,
943 'Other_Grapheme_Extend' => $contributory,
944 'Other_ID_Continue' => $contributory,
945 'Other_ID_Start' => $contributory,
946 'Other_Lowercase' => $contributory,
947 'Other_Math' => $contributory,
948 'Other_Uppercase' => $contributory,
e22aaf5c
KW
949 'Expands_On_NFC' => $why_no_expand,
950 'Expands_On_NFD' => $why_no_expand,
951 'Expands_On_NFKC' => $why_no_expand,
952 'Expands_On_NFKD' => $why_no_expand,
99870f4d
KW
953 );
954
955 %why_suppressed = (
5f7264c7 956 # There is a lib/unicore/Decomposition.pl (used by Normalize.pm) which
99870f4d
KW
957 # contains the same information, but without the algorithmically
958 # determinable Hangul syllables'. This file is not published, so it's
959 # existence is not noted in the comment.
12fee290 960 'Decomposition_Mapping' => 'Accessible via Unicode::Normalize or prop_invmap() or charprop() in Unicode::UCD::',
99870f4d 961
ac71d2a0 962 'Indic_Matra_Category' => "Withdrawn by Unicode while still provisional",
3111abc0 963
5f8d1a89
KW
964 # Don't suppress ISO_Comment, as otherwise special handling is needed
965 # to differentiate between it and gc=c, which can be written as 'isc',
966 # which is the same characters as ISO_Comment's short name.
99870f4d 967
12fee290 968 'Name' => "Accessible via \\N{...} or 'use charnames;' or charprop() or prop_invmap() in Unicode::UCD::",
e0b29447 969
12fee290
KW
970 'Simple_Case_Folding' => "$simple. Can access this through casefold(), charprop(), or prop_invmap() in Unicode::UCD",
971 'Simple_Lowercase_Mapping' => "$simple. Can access this through charinfo(), charprop(), or prop_invmap() in Unicode::UCD",
972 'Simple_Titlecase_Mapping' => "$simple. Can access this through charinfo(), charprop(), or prop_invmap() in Unicode::UCD",
973 'Simple_Uppercase_Mapping' => "$simple. Can access this through charinfo(), charprop(), or prop_invmap() in Unicode::UCD",
99870f4d 974
dac6f618 975 FC_NFKC_Closure => 'Deprecated by Unicode, and supplanted in usage by NFKC_Casefold; otherwise not useful',
99870f4d
KW
976 );
977
1704a0ea
KW
978 foreach my $property (
979
980 # The following are suppressed because they were made contributory
981 # or deprecated by Unicode before Perl ever thought about
982 # supporting them.
983 'Jamo_Short_Name',
984 'Grapheme_Link',
985 'Expands_On_NFC',
986 'Expands_On_NFD',
987 'Expands_On_NFKC',
988 'Expands_On_NFKD',
989
990 # The following are suppressed because they have been marked
991 # as deprecated for a sufficient amount of time
992 'Other_Alphabetic',
993 'Other_Default_Ignorable_Code_Point',
994 'Other_Grapheme_Extend',
995 'Other_ID_Continue',
996 'Other_ID_Start',
997 'Other_Lowercase',
998 'Other_Math',
999 'Other_Uppercase',
e22aaf5c 1000 ) {
99870f4d
KW
1001 $why_suppressed{$property} = $why_deprecated{$property};
1002 }
cf25bb62 1003
99870f4d
KW
1004 # Customize the message for all the 'Other_' properties
1005 foreach my $property (keys %why_deprecated) {
1006 next if (my $main_property = $property) !~ s/^Other_//;
1007 $why_deprecated{$property} =~ s/$other_properties/the $main_property property (which should be used instead)/;
1008 }
1009}
1010
e9c4b4f8
KW
1011if ($write_Unicode_deprecated_tables) {
1012 foreach my $property (keys %why_suppressed) {
1013 delete $why_suppressed{$property} if $property =~
1014 / ^ Other | Grapheme /x;
1015 }
1016}
1017
99870f4d
KW
1018if ($v_version ge 4.0.0) {
1019 $why_stabilized{'Hyphen'} = 'Use the Line_Break property instead; see www.unicode.org/reports/tr14';
5f7264c7
KW
1020 if ($v_version ge 6.0.0) {
1021 $why_deprecated{'Hyphen'} = 'Supplanted by Line_Break property values; see www.unicode.org/reports/tr14';
1022 }
99870f4d 1023}
5f7264c7 1024if ($v_version ge 5.2.0 && $v_version lt 6.0.0) {
99870f4d 1025 $why_obsolete{'ISO_Comment'} = 'Code points for it have been removed';
5f7264c7 1026 if ($v_version ge 6.0.0) {
63f74647 1027 $why_deprecated{'ISO_Comment'} = 'No longer needed for Unicode\'s internal chart generation; otherwise not useful, and code points for it have been removed';
5f7264c7 1028 }
99870f4d
KW
1029}
1030
1031# Probably obsolete forever
1032if ($v_version ge v4.1.0) {
82aed44a
KW
1033 $why_suppressed{'Script=Katakana_Or_Hiragana'} = 'Obsolete. All code points previously matched by this have been moved to "Script=Common".';
1034}
1035if ($v_version ge v6.0.0) {
caa75395 1036 $why_suppressed{'Script=Katakana_Or_Hiragana'} .= ' Consider instead using "Script_Extensions=Katakana" or "Script_Extensions=Hiragana" (or both)';
2b352efd 1037 $why_suppressed{'Script_Extensions=Katakana_Or_Hiragana'} = 'All code points that would be matched by this are matched by either "Script_Extensions=Katakana" or "Script_Extensions=Hiragana"';
99870f4d
KW
1038}
1039
1040# This program can create files for enumerated-like properties, such as
1041# 'Numeric_Type'. This file would be the same format as for a string
1042# property, with a mapping from code point to its value, so you could look up,
1043# for example, the script a code point is in. But no one so far wants this
1044# mapping, or they have found another way to get it since this is a new
1045# feature. So no file is generated except if it is in this list.
1046my @output_mapped_properties = split "\n", <<END;
1047END
1048
c12f2655
KW
1049# If you are using the Unihan database in a Unicode version before 5.2, you
1050# need to add the properties that you want to extract from it to this table.
1051# For your convenience, the properties in the 6.0 PropertyAliases.txt file are
1052# listed, commented out
99870f4d
KW
1053my @cjk_properties = split "\n", <<'END';
1054#cjkAccountingNumeric; kAccountingNumeric
1055#cjkOtherNumeric; kOtherNumeric
1056#cjkPrimaryNumeric; kPrimaryNumeric
1057#cjkCompatibilityVariant; kCompatibilityVariant
1058#cjkIICore ; kIICore
1059#cjkIRG_GSource; kIRG_GSource
1060#cjkIRG_HSource; kIRG_HSource
1061#cjkIRG_JSource; kIRG_JSource
1062#cjkIRG_KPSource; kIRG_KPSource
1063#cjkIRG_KSource; kIRG_KSource
1064#cjkIRG_TSource; kIRG_TSource
1065#cjkIRG_USource; kIRG_USource
1066#cjkIRG_VSource; kIRG_VSource
1067#cjkRSUnicode; kRSUnicode ; Unicode_Radical_Stroke; URS
1068END
1069
1070# Similarly for the property values. For your convenience, the lines in the
5f7264c7 1071# 6.0 PropertyAliases.txt file are listed. Just remove the first BUT NOT both
c12f2655 1072# '#' marks (for Unicode versions before 5.2)
99870f4d
KW
1073my @cjk_property_values = split "\n", <<'END';
1074## @missing: 0000..10FFFF; cjkAccountingNumeric; NaN
1075## @missing: 0000..10FFFF; cjkCompatibilityVariant; <code point>
1076## @missing: 0000..10FFFF; cjkIICore; <none>
1077## @missing: 0000..10FFFF; cjkIRG_GSource; <none>
1078## @missing: 0000..10FFFF; cjkIRG_HSource; <none>
1079## @missing: 0000..10FFFF; cjkIRG_JSource; <none>
1080## @missing: 0000..10FFFF; cjkIRG_KPSource; <none>
1081## @missing: 0000..10FFFF; cjkIRG_KSource; <none>
1082## @missing: 0000..10FFFF; cjkIRG_TSource; <none>
1083## @missing: 0000..10FFFF; cjkIRG_USource; <none>
1084## @missing: 0000..10FFFF; cjkIRG_VSource; <none>
1085## @missing: 0000..10FFFF; cjkOtherNumeric; NaN
1086## @missing: 0000..10FFFF; cjkPrimaryNumeric; NaN
1087## @missing: 0000..10FFFF; cjkRSUnicode; <none>
1088END
1089
1090# The input files don't list every code point. Those not listed are to be
1091# defaulted to some value. Below are hard-coded what those values are for
1092# non-binary properties as of 5.1. Starting in 5.0, there are
caa75395 1093# machine-parsable comment lines in the files that give the defaults; so this
99870f4d
KW
1094# list shouldn't have to be extended. The claim is that all missing entries
1095# for binary properties will default to 'N'. Unicode tried to change that in
1096# 5.2, but the beta period produced enough protest that they backed off.
1097#
1098# The defaults for the fields that appear in UnicodeData.txt in this hash must
1099# be in the form that it expects. The others may be synonyms.
1100my $CODE_POINT = '<code point>';
1101my %default_mapping = (
1102 Age => "Unassigned",
1103 # Bidi_Class => Complicated; set in code
1104 Bidi_Mirroring_Glyph => "",
1105 Block => 'No_Block',
1106 Canonical_Combining_Class => 0,
1107 Case_Folding => $CODE_POINT,
1108 Decomposition_Mapping => $CODE_POINT,
1109 Decomposition_Type => 'None',
1110 East_Asian_Width => "Neutral",
1111 FC_NFKC_Closure => $CODE_POINT,
1112 General_Category => 'Cn',
1113 Grapheme_Cluster_Break => 'Other',
1114 Hangul_Syllable_Type => 'NA',
1115 ISO_Comment => "",
1116 Jamo_Short_Name => "",
1117 Joining_Group => "No_Joining_Group",
1118 # Joining_Type => Complicated; set in code
1119 kIICore => 'N', # Is converted to binary
1120 #Line_Break => Complicated; set in code
1121 Lowercase_Mapping => $CODE_POINT,
1122 Name => "",
1123 Name_Alias => "",
1124 NFC_QC => 'Yes',
1125 NFD_QC => 'Yes',
1126 NFKC_QC => 'Yes',
1127 NFKD_QC => 'Yes',
1128 Numeric_Type => 'None',
1129 Numeric_Value => 'NaN',
1130 Script => ($v_version le 4.1.0) ? 'Common' : 'Unknown',
1131 Sentence_Break => 'Other',
1132 Simple_Case_Folding => $CODE_POINT,
1133 Simple_Lowercase_Mapping => $CODE_POINT,
1134 Simple_Titlecase_Mapping => $CODE_POINT,
1135 Simple_Uppercase_Mapping => $CODE_POINT,
1136 Titlecase_Mapping => $CODE_POINT,
1137 Unicode_1_Name => "",
1138 Unicode_Radical_Stroke => "",
1139 Uppercase_Mapping => $CODE_POINT,
1140 Word_Break => 'Other',
1141);
1142
232ed87f
KW
1143# Below are files that Unicode furnishes, but this program ignores, and why.
1144# NormalizationCorrections.txt requires some more explanation. It documents
1145# the cumulative fixes to erroneous normalizations in earlier Unicode
1146# versions. Its main purpose is so that someone running on an earlier version
1147# can use this file to override what got published in that earlier release.
1148# It would be easy for mktables to read and handle this file. But all the
1149# corrections in it should already be in the other files for the release it
1150# is. To get it to actually mean something useful, someone would have to be
1151# using an earlier Unicode release, and copy it to the files for that release
1152# and recomplile. So far there has been no demand to do that, so this hasn't
1153# been implemented.
99870f4d 1154my %ignored_files = (
73ba1144
KW
1155 'CJKRadicals.txt' => 'Maps the kRSUnicode property values to corresponding code points',
1156 'Index.txt' => 'Alphabetical index of Unicode characters',
1157 'NamedSqProv.txt' => 'Named sequences proposed for inclusion in a later version of the Unicode Standard; if you need them now, you can append this file to F<NamedSequences.txt> and recompile perl',
1158 'NamesList.txt' => 'Annotated list of characters',
524a8e5e 1159 'NamesList.html' => 'Describes the format and contents of F<NamesList.txt>',
73ba1144
KW
1160 'NormalizationCorrections.txt' => 'Documentation of corrections already incorporated into the Unicode data base',
1161 'Props.txt' => 'Only in very early releases; is a subset of F<PropList.txt> (which is used instead)',
1162 'ReadMe.txt' => 'Documentation',
1163 'StandardizedVariants.txt' => 'Certain glyph variations for character display are standardized. This lists the non-Unihan ones; the Unihan ones are also not used by Perl, and are in a separate Unicode data base L<http://www.unicode.org/ivd>',
524a8e5e 1164 'StandardizedVariants.html' => 'Provides a visual display of the standard variant sequences derived from F<StandardizedVariants.txt>.',
73ba1144 1165 'EmojiSources.txt' => 'Maps certain Unicode code points to their legacy Japanese cell-phone values',
8ee2793f 1166 'USourceData.txt' => 'Documentation of status and cross reference of proposals for encoding by Unicode of Unihan characters',
524a8e5e 1167 'USourceGlyphs.pdf' => 'Pictures of the characters in F<USourceData.txt>',
73ba1144
KW
1168 'auxiliary/WordBreakTest.html' => 'Documentation of validation tests',
1169 'auxiliary/SentenceBreakTest.html' => 'Documentation of validation tests',
1170 'auxiliary/GraphemeBreakTest.html' => 'Documentation of validation tests',
1171 'auxiliary/LineBreakTest.html' => 'Documentation of validation tests',
99870f4d
KW
1172);
1173
1fec9f60
KW
1174my %skipped_files; # List of files that we skip
1175
678f13d5 1176### End of externally interesting definitions, except for @input_file_objects
99870f4d
KW
1177
1178my $HEADER=<<"EOF";
1179# !!!!!!! DO NOT EDIT THIS FILE !!!!!!!
3df51b85
KW
1180# This file is machine-generated by $0 from the Unicode
1181# database, Version $string_version. Any changes made here will be lost!
cf25bb62
JH
1182EOF
1183
126c3d4e 1184my $INTERNAL_ONLY_HEADER = <<"EOF";
99870f4d
KW
1185
1186# !!!!!!! INTERNAL PERL USE ONLY !!!!!!!
fac53429
KW
1187# This file is for internal use by core Perl only. The format and even the
1188# name or existence of this file are subject to change without notice. Don't
d9ae3878
KW
1189# use it directly. Use Unicode::UCD to access the Unicode character data
1190# base.
99870f4d
KW
1191EOF
1192
1193my $DEVELOPMENT_ONLY=<<"EOF";
1194# !!!!!!! DEVELOPMENT USE ONLY !!!!!!!
1195# This file contains information artificially constrained to code points
1196# present in Unicode release $string_compare_versions.
1197# IT CANNOT BE RELIED ON. It is for use during development only and should
23e33b60 1198# not be used for production.
b6922eda
KW
1199
1200EOF
1201
346a20cf
KW
1202my $MAX_UNICODE_CODEPOINT_STRING = ($v_version ge v2.0.0)
1203 ? "10FFFF"
1204 : "FFFF";
6189eadc
KW
1205my $MAX_UNICODE_CODEPOINT = hex $MAX_UNICODE_CODEPOINT_STRING;
1206my $MAX_UNICODE_CODEPOINTS = $MAX_UNICODE_CODEPOINT + 1;
99870f4d 1207
2d88a86a
KW
1208# We work with above-Unicode code points, up to UV_MAX. But when you get
1209# that high, above IV_MAX, some operations don't work, and you can easily get
1210# overflow. Therefore for internal use, we use a much smaller number,
1211# translating it to UV_MAX only for output. The exact number is immaterial
1212# (all Unicode code points are treated exactly the same), but the algorithm
1213# requires it to be at least 2 * $MAX_UNICODE_CODEPOINTS + 1;
1214my $MAX_WORKING_CODEPOINTS= $MAX_UNICODE_CODEPOINT * 8;
1215my $MAX_WORKING_CODEPOINT = $MAX_WORKING_CODEPOINTS - 1;
1216my $MAX_WORKING_CODEPOINT_STRING = sprintf("%X", $MAX_WORKING_CODEPOINT);
1217
1218my $MAX_PLATFORM_CODEPOINT = ~0;
1219
99870f4d
KW
1220# Matches legal code point. 4-6 hex numbers, If there are 6, the first
1221# two must be 10; if there are 5, the first must not be a 0. Written this way
92199589
KW
1222# to decrease backtracking. The first regex allows the code point to be at
1223# the end of a word, but to work properly, the word shouldn't end with a valid
1224# hex character. The second one won't match a code point at the end of a
1225# word, and doesn't have the run-on issue
8c32d378
KW
1226my $run_on_code_point_re =
1227 qr/ (?: 10[0-9A-F]{4} | [1-9A-F][0-9A-F]{4} | [0-9A-F]{4} ) \b/x;
1228my $code_point_re = qr/\b$run_on_code_point_re/;
99870f4d
KW
1229
1230# This matches the beginning of the line in the Unicode db files that give the
1231# defaults for code points not listed (i.e., missing) in the file. The code
1232# depends on this ending with a semi-colon, so it can assume it is a valid
1233# field when the line is split() by semi-colons
346a20cf 1234my $missing_defaults_prefix = qr/^#\s+\@missing:\s+0000\.\.10FFFF\s*;/;
99870f4d
KW
1235
1236# Property types. Unicode has more types, but these are sufficient for our
1237# purposes.
1238my $UNKNOWN = -1; # initialized to illegal value
1239my $NON_STRING = 1; # Either binary or enum
1240my $BINARY = 2;
06f26c45
KW
1241my $FORCED_BINARY = 3; # Not a binary property, but, besides its normal
1242 # tables, additional true and false tables are
1243 # generated so that false is anything matching the
1244 # default value, and true is everything else.
1245my $ENUM = 4; # Include catalog
1246my $STRING = 5; # Anything else: string or misc
99870f4d
KW
1247
1248# Some input files have lines that give default values for code points not
1249# contained in the file. Sometimes these should be ignored.
1250my $NO_DEFAULTS = 0; # Must evaluate to false
1251my $NOT_IGNORED = 1;
1252my $IGNORED = 2;
1253
1254# Range types. Each range has a type. Most ranges are type 0, for normal,
1255# and will appear in the main body of the tables in the output files, but
1256# there are other types of ranges as well, listed below, that are specially
1257# handled. There are pseudo-types as well that will never be stored as a
1258# type, but will affect the calculation of the type.
1259
1260# 0 is for normal, non-specials
1261my $MULTI_CP = 1; # Sequence of more than code point
1262my $HANGUL_SYLLABLE = 2;
1263my $CP_IN_NAME = 3; # The NAME contains the code point appended to it.
1264my $NULL = 4; # The map is to the null string; utf8.c can't
1265 # handle these, nor is there an accepted syntax
1266 # for them in \p{} constructs
f86864ac 1267my $COMPUTE_NO_MULTI_CP = 5; # Pseudo-type; means that ranges that would
99870f4d
KW
1268 # otherwise be $MULTI_CP type are instead type 0
1269
1270# process_generic_property_file() can accept certain overrides in its input.
1271# Each of these must begin AND end with $CMD_DELIM.
1272my $CMD_DELIM = "\a";
1273my $REPLACE_CMD = 'replace'; # Override the Replace
1274my $MAP_TYPE_CMD = 'map_type'; # Override the Type
1275
1276my $NO = 0;
1277my $YES = 1;
1278
1279# Values for the Replace argument to add_range.
1280# $NO # Don't replace; add only the code points not
1281 # already present.
1282my $IF_NOT_EQUIVALENT = 1; # Replace only under certain conditions; details in
1283 # the comments at the subroutine definition.
1284my $UNCONDITIONALLY = 2; # Replace without conditions.
9470941f 1285my $MULTIPLE_BEFORE = 4; # Don't replace, but add a duplicate record if
99870f4d 1286 # already there
7f4b1e25
KW
1287my $MULTIPLE_AFTER = 5; # Don't replace, but add a duplicate record if
1288 # already there
1289my $CROAK = 6; # Die with an error if is already there
99870f4d
KW
1290
1291# Flags to give property statuses. The phrases are to remind maintainers that
1292# if the flag is changed, the indefinite article referring to it in the
1293# documentation may need to be as well.
1294my $NORMAL = "";
99870f4d
KW
1295my $DEPRECATED = 'D';
1296my $a_bold_deprecated = "a 'B<$DEPRECATED>'";
1297my $A_bold_deprecated = "A 'B<$DEPRECATED>'";
1298my $DISCOURAGED = 'X';
1299my $a_bold_discouraged = "an 'B<$DISCOURAGED>'";
1300my $A_bold_discouraged = "An 'B<$DISCOURAGED>'";
1301my $STRICTER = 'T';
1302my $a_bold_stricter = "a 'B<$STRICTER>'";
1303my $A_bold_stricter = "A 'B<$STRICTER>'";
1304my $STABILIZED = 'S';
1305my $a_bold_stabilized = "an 'B<$STABILIZED>'";
1306my $A_bold_stabilized = "An 'B<$STABILIZED>'";
1307my $OBSOLETE = 'O';
1308my $a_bold_obsolete = "an 'B<$OBSOLETE>'";
1309my $A_bold_obsolete = "An 'B<$OBSOLETE>'";
1310
1311my %status_past_participles = (
1312 $DISCOURAGED => 'discouraged',
99870f4d
KW
1313 $STABILIZED => 'stabilized',
1314 $OBSOLETE => 'obsolete',
37e2e78e 1315 $DEPRECATED => 'deprecated',
99870f4d
KW
1316);
1317
395dfc19
KW
1318# Table fates. These are somewhat ordered, so that fates < $MAP_PROXIED should be
1319# externally documented.
301ba948 1320my $ORDINARY = 0; # The normal fate.
395dfc19
KW
1321my $MAP_PROXIED = 1; # The map table for the property isn't written out,
1322 # but there is a file written that can be used to
1323 # reconstruct this table
3cdaf629 1324my $INTERNAL_ONLY = 2; # The file for this table is written out, but it is
301ba948 1325 # for Perl's internal use only
277b7b16
KW
1326my $LEGACY_ONLY = 3; # Like $INTERNAL_ONLY, but not actually used by Perl.
1327 # Is for backwards compatibility for applications that
1328 # read the file directly, so it's format is
1329 # unchangeable.
1330my $SUPPRESSED = 4; # The file for this table is not written out, and as a
3cdaf629
KW
1331 # result, we don't bother to do many computations on
1332 # it.
277b7b16 1333my $PLACEHOLDER = 5; # Like $SUPPRESSED, but we go through all the
3cdaf629
KW
1334 # computations anyway, as the values are needed for
1335 # things to work. This happens when we have Perl
1336 # extensions that depend on Unicode tables that
1337 # wouldn't normally be in a given Unicode version.
301ba948 1338
f5817e0a
KW
1339# The format of the values of the tables:
1340my $EMPTY_FORMAT = "";
99870f4d
KW
1341my $BINARY_FORMAT = 'b';
1342my $DECIMAL_FORMAT = 'd';
1343my $FLOAT_FORMAT = 'f';
1344my $INTEGER_FORMAT = 'i';
1345my $HEX_FORMAT = 'x';
1346my $RATIONAL_FORMAT = 'r';
1347my $STRING_FORMAT = 's';
d11155ec 1348my $ADJUST_FORMAT = 'a';
24303724 1349my $HEX_ADJUST_FORMAT = 'ax';
a14f3cb1 1350my $DECOMP_STRING_FORMAT = 'c';
c3ff2976 1351my $STRING_WHITE_SPACE_LIST = 'sw';
99870f4d
KW
1352
1353my %map_table_formats = (
1354 $BINARY_FORMAT => 'binary',
1355 $DECIMAL_FORMAT => 'single decimal digit',
1356 $FLOAT_FORMAT => 'floating point number',
1357 $INTEGER_FORMAT => 'integer',
add63c13 1358 $HEX_FORMAT => 'non-negative hex whole number; a code point',
99870f4d 1359 $RATIONAL_FORMAT => 'rational: an integer or a fraction',
1a9d544b 1360 $STRING_FORMAT => 'string',
d11155ec 1361 $ADJUST_FORMAT => 'some entries need adjustment',
24303724 1362 $HEX_ADJUST_FORMAT => 'mapped value in hex; some entries need adjustment',
92f9d56c 1363 $DECOMP_STRING_FORMAT => 'Perl\'s internal (Normalize.pm) decomposition mapping',
c3ff2976 1364 $STRING_WHITE_SPACE_LIST => 'string, but some elements are interpreted as a list; white space occurs only as list item separators'
99870f4d
KW
1365);
1366
1367# Unicode didn't put such derived files in a separate directory at first.
1368my $EXTRACTED_DIR = (-d 'extracted') ? 'extracted' : "";
1369my $EXTRACTED = ($EXTRACTED_DIR) ? "$EXTRACTED_DIR/" : "";
1370my $AUXILIARY = 'auxiliary';
1371
3854b4b8
KW
1372# Hashes and arrays that will eventually go into Heavy.pl for the use of
1373# utf8_heavy.pl and into UCD.pl for the use of UCD.pm
99870f4d
KW
1374my %loose_to_file_of; # loosely maps table names to their respective
1375 # files
1376my %stricter_to_file_of; # same; but for stricter mapping.
315bfd4e 1377my %loose_property_to_file_of; # Maps a loose property name to its map file
3854b4b8
KW
1378my @inline_definitions = "V0"; # Each element gives a definition of a unique
1379 # inversion list. When a definition is inlined,
1380 # its value in the hash it's in (one of the two
1381 # defined just above) will include an index into
1382 # this array. The 0th element is initialized to
1383 # the definition for a zero length invwersion list
89cf10cc
KW
1384my %file_to_swash_name; # Maps the file name to its corresponding key name
1385 # in the hash %utf8::SwashInfo
99870f4d
KW
1386my %nv_floating_to_rational; # maps numeric values floating point numbers to
1387 # their rational equivalent
c12f2655
KW
1388my %loose_property_name_of; # Loosely maps (non_string) property names to
1389 # standard form
86a52d1e 1390my %string_property_loose_to_name; # Same, for string properties.
c15fda25
KW
1391my %loose_defaults; # keys are of form "prop=value", where 'prop' is
1392 # the property name in standard loose form, and
1393 # 'value' is the default value for that property,
1394 # also in standard loose form.
9e4a1e86
KW
1395my %loose_to_standard_value; # loosely maps table names to the canonical
1396 # alias for them
2df7880f
KW
1397my %ambiguous_names; # keys are alias names (in standard form) that
1398 # have more than one possible meaning.
5d1df013
KW
1399my %prop_aliases; # Keys are standard property name; values are each
1400 # one's aliases
1e863613
KW
1401my %prop_value_aliases; # Keys of top level are standard property name;
1402 # values are keys to another hash, Each one is
1403 # one of the property's values, in standard form.
1404 # The values are that prop-val's aliases.
2df7880f 1405my %ucd_pod; # Holds entries that will go into the UCD section of the pod
99870f4d 1406
d867ccfb
KW
1407# Most properties are immune to caseless matching, otherwise you would get
1408# nonsensical results, as properties are a function of a code point, not
1409# everything that is caselessly equivalent to that code point. For example,
1410# Changes_When_Case_Folded('s') should be false, whereas caselessly it would
1411# be true because 's' and 'S' are equivalent caselessly. However,
1412# traditionally, [:upper:] and [:lower:] are equivalent caselessly, so we
1413# extend that concept to those very few properties that are like this. Each
1414# such property will match the full range caselessly. They are hard-coded in
1415# the program; it's not worth trying to make it general as it's extremely
1416# unlikely that they will ever change.
1417my %caseless_equivalent_to;
1418
99870f4d
KW
1419# These constants names and values were taken from the Unicode standard,
1420# version 5.1, section 3.12. They are used in conjunction with Hangul
6e5a209b
KW
1421# syllables. The '_string' versions are so generated tables can retain the
1422# hex format, which is the more familiar value
1423my $SBase_string = "0xAC00";
1424my $SBase = CORE::hex $SBase_string;
1425my $LBase_string = "0x1100";
1426my $LBase = CORE::hex $LBase_string;
1427my $VBase_string = "0x1161";
1428my $VBase = CORE::hex $VBase_string;
1429my $TBase_string = "0x11A7";
1430my $TBase = CORE::hex $TBase_string;
99870f4d
KW
1431my $SCount = 11172;
1432my $LCount = 19;
1433my $VCount = 21;
1434my $TCount = 28;
1435my $NCount = $VCount * $TCount;
1436
1437# For Hangul syllables; These store the numbers from Jamo.txt in conjunction
1438# with the above published constants.
1439my %Jamo;
1440my %Jamo_L; # Leading consonants
1441my %Jamo_V; # Vowels
1442my %Jamo_T; # Trailing consonants
1443
bb1dd3da
KW
1444# For code points whose name contains its ordinal as a '-ABCD' suffix.
1445# The key is the base name of the code point, and the value is an
1446# array giving all the ranges that use this base name. Each range
1447# is actually a hash giving the 'low' and 'high' values of it.
1448my %names_ending_in_code_point;
1449my %loose_names_ending_in_code_point; # Same as above, but has blanks, dashes
1450 # removed from the names
1451# Inverse mapping. The list of ranges that have these kinds of
1452# names. Each element contains the low, high, and base names in an
1453# anonymous hash.
1454my @code_points_ending_in_code_point;
1455
6b5ab373
KW
1456# To hold Unicode's normalization test suite
1457my @normalization_tests;
1458
bb1dd3da
KW
1459# Boolean: does this Unicode version have the hangul syllables, and are we
1460# writing out a table for them?
1461my $has_hangul_syllables = 0;
1462
1463# Does this Unicode version have code points whose names end in their
1464# respective code points, and are we writing out a table for them? 0 for no;
1465# otherwise points to first property that a table is needed for them, so that
1466# if multiple tables are needed, we don't create duplicates
1467my $needing_code_points_ending_in_code_point = 0;
1468
37e2e78e 1469my @backslash_X_tests; # List of tests read in for testing \X
06ae2722 1470my @SB_tests; # List of tests read in for testing \b{sb}
ae3bb8ea 1471my @WB_tests; # List of tests read in for testing \b{wb}
99870f4d
KW
1472my @unhandled_properties; # Will contain a list of properties found in
1473 # the input that we didn't process.
f86864ac 1474my @match_properties; # Properties that have match tables, to be
99870f4d
KW
1475 # listed in the pod
1476my @map_properties; # Properties that get map files written
1477my @named_sequences; # NamedSequences.txt contents.
1478my %potential_files; # Generated list of all .txt files in the directory
1479 # structure so we can warn if something is being
1480 # ignored.
1481my @files_actually_output; # List of files we generated.
1482my @more_Names; # Some code point names are compound; this is used
1483 # to store the extra components of them.
1484my $MIN_FRACTION_LENGTH = 3; # How many digits of a floating point number at
1485 # the minimum before we consider it equivalent to a
1486 # candidate rational
1487my $MAX_FLOATING_SLOP = 10 ** - $MIN_FRACTION_LENGTH; # And in floating terms
1488
1489# These store references to certain commonly used property objects
3c88a801 1490my $ccc;
99870f4d
KW
1491my $gc;
1492my $perl;
1493my $block;
3e20195b
KW
1494my $perl_charname;
1495my $print;
ac7dbdb6 1496my $All;
359523e2 1497my $script;
99870f4d
KW
1498
1499# Are there conflicting names because of beginning with 'In_', or 'Is_'
1500my $has_In_conflicts = 0;
1501my $has_Is_conflicts = 0;
1502
1503sub internal_file_to_platform ($) {
1504 # Convert our file paths which have '/' separators to those of the
1505 # platform.
1506
1507 my $file = shift;
1508 return undef unless defined $file;
1509
1510 return File::Spec->join(split '/', $file);
d07a55ed 1511}
5beb625e 1512
99870f4d
KW
1513sub file_exists ($) { # platform independent '-e'. This program internally
1514 # uses slash as a path separator.
1515 my $file = shift;
1516 return 0 if ! defined $file;
1517 return -e internal_file_to_platform($file);
1518}
5beb625e 1519
99870f4d 1520sub objaddr($) {
23e33b60
KW
1521 # Returns the address of the blessed input object.
1522 # It doesn't check for blessedness because that would do a string eval
1523 # every call, and the program is structured so that this is never called
1524 # for a non-blessed object.
99870f4d 1525
23e33b60 1526 no overloading; # If overloaded, numifying below won't work.
99870f4d
KW
1527
1528 # Numifying a ref gives its address.
051df77b 1529 return pack 'J', $_[0];
99870f4d
KW
1530}
1531
558712cf 1532# These are used only if $annotate is true.
c4019d52
KW
1533# The entire range of Unicode characters is examined to populate these
1534# after all the input has been processed. But most can be skipped, as they
1535# have the same descriptive phrases, such as being unassigned
1536my @viacode; # Contains the 1 million character names
1537my @printable; # boolean: And are those characters printable?
1538my @annotate_char_type; # Contains a type of those characters, specifically
1539 # for the purposes of annotation.
1540my $annotate_ranges; # A map of ranges of code points that have the same
98dc9551 1541 # name for the purposes of annotation. They map to the
c4019d52
KW
1542 # upper edge of the range, so that the end point can
1543 # be immediately found. This is used to skip ahead to
1544 # the end of a range, and avoid processing each
1545 # individual code point in it.
1546my $unassigned_sans_noncharacters; # A Range_List of the unassigned
1547 # characters, but excluding those which are
1548 # also noncharacter code points
1549
1550# The annotation types are an extension of the regular range types, though
1551# some of the latter are folded into one. Make the new types negative to
1552# avoid conflicting with the regular types
1553my $SURROGATE_TYPE = -1;
1554my $UNASSIGNED_TYPE = -2;
1555my $PRIVATE_USE_TYPE = -3;
1556my $NONCHARACTER_TYPE = -4;
1557my $CONTROL_TYPE = -5;
2d88a86a
KW
1558my $ABOVE_UNICODE_TYPE = -6;
1559my $UNKNOWN_TYPE = -7; # Used only if there is a bug in this program
c4019d52
KW
1560
1561sub populate_char_info ($) {
558712cf 1562 # Used only with the $annotate option. Populates the arrays with the
c4019d52
KW
1563 # input code point's info that are needed for outputting more detailed
1564 # comments. If calling context wants a return, it is the end point of
1565 # any contiguous range of characters that share essentially the same info
1566
1567 my $i = shift;
1568 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1569
1570 $viacode[$i] = $perl_charname->value_of($i) || "";
1571
1572 # A character is generally printable if Unicode says it is,
1573 # but below we make sure that most Unicode general category 'C' types
1574 # aren't.
1575 $printable[$i] = $print->contains($i);
1576
1577 $annotate_char_type[$i] = $perl_charname->type_of($i) || 0;
1578
1579 # Only these two regular types are treated specially for annotations
1580 # purposes
1581 $annotate_char_type[$i] = 0 if $annotate_char_type[$i] != $CP_IN_NAME
1582 && $annotate_char_type[$i] != $HANGUL_SYLLABLE;
1583
1584 # Give a generic name to all code points that don't have a real name.
1585 # We output ranges, if applicable, for these. Also calculate the end
1586 # point of the range.
1587 my $end;
1588 if (! $viacode[$i]) {
1d025d66 1589 my $nonchar;
2d88a86a
KW
1590 if ($i > $MAX_UNICODE_CODEPOINT) {
1591 $viacode[$i] = 'Above-Unicode';
1592 $annotate_char_type[$i] = $ABOVE_UNICODE_TYPE;
1593 $printable[$i] = 0;
1594 $end = $MAX_WORKING_CODEPOINT;
1595 }
1596 elsif ($gc-> table('Private_use')->contains($i)) {
c4019d52
KW
1597 $viacode[$i] = 'Private Use';
1598 $annotate_char_type[$i] = $PRIVATE_USE_TYPE;
1599 $printable[$i] = 0;
1600 $end = $gc->table('Private_Use')->containing_range($i)->end;
1601 }
1d025d66
KW
1602 elsif ((defined ($nonchar =
1603 Property::property_ref('Noncharacter_Code_Point'))
1604 && $nonchar->table('Y')->contains($i)))
c4019d52
KW
1605 {
1606 $viacode[$i] = 'Noncharacter';
1607 $annotate_char_type[$i] = $NONCHARACTER_TYPE;
1608 $printable[$i] = 0;
1609 $end = property_ref('Noncharacter_Code_Point')->table('Y')->
1610 containing_range($i)->end;
1611 }
1612 elsif ($gc-> table('Control')->contains($i)) {
c71dea7f 1613 $viacode[$i] = property_ref('Name_Alias')->value_of($i) || 'Control';
c4019d52
KW
1614 $annotate_char_type[$i] = $CONTROL_TYPE;
1615 $printable[$i] = 0;
c4019d52
KW
1616 }
1617 elsif ($gc-> table('Unassigned')->contains($i)) {
c4019d52
KW
1618 $annotate_char_type[$i] = $UNASSIGNED_TYPE;
1619 $printable[$i] = 0;
1d025d66
KW
1620 if ($v_version lt v2.0.0) { # No blocks in earliest releases
1621 $viacode[$i] = 'Unassigned';
1622 $end = $gc-> table('Unassigned')->containing_range($i)->end;
1623 }
1624 else {
1625 $viacode[$i] = 'Unassigned, block=' . $block-> value_of($i);
c4019d52 1626
bf06c733
KW
1627 # Because we name the unassigned by the blocks they are in, it
1628 # can't go past the end of that block, and it also can't go
1629 # past the unassigned range it is in. The special table makes
1630 # sure that the non-characters, which are unassigned, are
1631 # separated out.
1632 $end = min($block->containing_range($i)->end,
1633 $unassigned_sans_noncharacters->
1634 containing_range($i)->end);
1d025d66
KW
1635 }
1636 }
1637 elsif ($v_version lt v2.0.0) { # No surrogates in earliest releases
1638 $viacode[$i] = $gc->value_of($i);
1639 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1640 $printable[$i] = 0;
1641 }
1642 elsif ($gc-> table('Surrogate')->contains($i)) {
1643 $viacode[$i] = 'Surrogate';
1644 $annotate_char_type[$i] = $SURROGATE_TYPE;
1645 $printable[$i] = 0;
1646 $end = $gc->table('Surrogate')->containing_range($i)->end;
13ca76ff
KW
1647 }
1648 else {
1649 Carp::my_carp_bug("Can't figure out how to annotate "
1650 . sprintf("U+%04X", $i)
1651 . ". Proceeding anyway.");
c4019d52
KW
1652 $viacode[$i] = 'UNKNOWN';
1653 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1654 $printable[$i] = 0;
1655 }
1656 }
1657
1658 # Here, has a name, but if it's one in which the code point number is
1659 # appended to the name, do that.
1660 elsif ($annotate_char_type[$i] == $CP_IN_NAME) {
1661 $viacode[$i] .= sprintf("-%04X", $i);
1662 $end = $perl_charname->containing_range($i)->end;
1663 }
1664
1665 # And here, has a name, but if it's a hangul syllable one, replace it with
1666 # the correct name from the Unicode algorithm
1667 elsif ($annotate_char_type[$i] == $HANGUL_SYLLABLE) {
1668 use integer;
1669 my $SIndex = $i - $SBase;
1670 my $L = $LBase + $SIndex / $NCount;
1671 my $V = $VBase + ($SIndex % $NCount) / $TCount;
1672 my $T = $TBase + $SIndex % $TCount;
1673 $viacode[$i] = "HANGUL SYLLABLE $Jamo{$L}$Jamo{$V}";
1674 $viacode[$i] .= $Jamo{$T} if $T != $TBase;
1675 $end = $perl_charname->containing_range($i)->end;
1676 }
1677
1678 return if ! defined wantarray;
1679 return $i if ! defined $end; # If not a range, return the input
1680
1681 # Save this whole range so can find the end point quickly
1682 $annotate_ranges->add_map($i, $end, $end);
1683
1684 return $end;
1685}
1686
23e33b60
KW
1687# Commented code below should work on Perl 5.8.
1688## This 'require' doesn't necessarily work in miniperl, and even if it does,
1689## the native perl version of it (which is what would operate under miniperl)
1690## is extremely slow, as it does a string eval every call.
7e017d6d 1691#my $has_fast_scalar_util = $^X !~ /miniperl/
23e33b60
KW
1692# && defined eval "require Scalar::Util";
1693#
1694#sub objaddr($) {
1695# # Returns the address of the blessed input object. Uses the XS version if
1696# # available. It doesn't check for blessedness because that would do a
1697# # string eval every call, and the program is structured so that this is
1698# # never called for a non-blessed object.
1699#
1700# return Scalar::Util::refaddr($_[0]) if $has_fast_scalar_util;
1701#
1702# # Check at least that is a ref.
1703# my $pkg = ref($_[0]) or return undef;
1704#
1705# # Change to a fake package to defeat any overloaded stringify
1706# bless $_[0], 'main::Fake';
1707#
1708# # Numifying a ref gives its address.
051df77b 1709# my $addr = pack 'J', $_[0];
23e33b60
KW
1710#
1711# # Return to original class
1712# bless $_[0], $pkg;
1713# return $addr;
1714#}
1715
99870f4d
KW
1716sub max ($$) {
1717 my $a = shift;
1718 my $b = shift;
1719 return $a if $a >= $b;
1720 return $b;
1721}
1722
1723sub min ($$) {
1724 my $a = shift;
1725 my $b = shift;
1726 return $a if $a <= $b;
1727 return $b;
1728}
1729
1730sub clarify_number ($) {
1731 # This returns the input number with underscores inserted every 3 digits
1732 # in large (5 digits or more) numbers. Input must be entirely digits, not
1733 # checked.
1734
1735 my $number = shift;
1736 my $pos = length($number) - 3;
1737 return $number if $pos <= 1;
1738 while ($pos > 0) {
1739 substr($number, $pos, 0) = '_';
1740 $pos -= 3;
5beb625e 1741 }
99870f4d 1742 return $number;
99598c8c
JH
1743}
1744
731cb813
KW
1745sub clarify_code_point_count ($) {
1746 # This is like clarify_number(), but the input is assumed to be a count of
1747 # code points, rather than a generic number.
1748
2d88a86a
KW
1749 my $append = "";
1750
1751 my $number = shift;
1752 if ($number > $MAX_UNICODE_CODEPOINTS) {
1753 $number -= ($MAX_WORKING_CODEPOINTS - $MAX_UNICODE_CODEPOINTS);
1754 return "All above-Unicode code points" if $number == 0;
1755 $append = " + all above-Unicode code points";
1756 }
1757 return clarify_number($number) . $append;
731cb813 1758}
12ac2576 1759
99870f4d 1760package Carp;
7ebf06b3 1761
99870f4d
KW
1762# These routines give a uniform treatment of messages in this program. They
1763# are placed in the Carp package to cause the stack trace to not include them,
1764# although an alternative would be to use another package and set @CARP_NOT
1765# for it.
12ac2576 1766
99870f4d 1767our $Verbose = 1 if main::DEBUG; # Useful info when debugging
12ac2576 1768
99f78760
KW
1769# This is a work-around suggested by Nicholas Clark to fix a problem with Carp
1770# and overload trying to load Scalar:Util under miniperl. See
1771# http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2009-11/msg01057.html
1772undef $overload::VERSION;
1773
99870f4d
KW
1774sub my_carp {
1775 my $message = shift || "";
1776 my $nofold = shift || 0;
7ebf06b3 1777
99870f4d
KW
1778 if ($message) {
1779 $message = main::join_lines($message);
1780 $message =~ s/^$0: *//; # Remove initial program name
1781 $message =~ s/[.;,]+$//; # Remove certain ending punctuation
1782 $message = "\n$0: $message;";
12ac2576 1783
99870f4d
KW
1784 # Fold the message with program name, semi-colon end punctuation
1785 # (which looks good with the message that carp appends to it), and a
1786 # hanging indent for continuation lines.
1787 $message = main::simple_fold($message, "", 4) unless $nofold;
1788 $message =~ s/\n$//; # Remove the trailing nl so what carp
1789 # appends is to the same line
1790 }
12ac2576 1791
99870f4d 1792 return $message if defined wantarray; # If a caller just wants the msg
12ac2576 1793
99870f4d
KW
1794 carp $message;
1795 return;
1796}
7ebf06b3 1797
99870f4d
KW
1798sub my_carp_bug {
1799 # This is called when it is clear that the problem is caused by a bug in
1800 # this program.
7ebf06b3 1801
99870f4d
KW
1802 my $message = shift;
1803 $message =~ s/^$0: *//;
1804 $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");
1805 carp $message;
1806 return;
1807}
7ebf06b3 1808
99870f4d
KW
1809sub carp_too_few_args {
1810 if (@_ != 2) {
1811 my_carp_bug("Wrong number of arguments: to 'carp_too_few_arguments'. No action taken.");
1812 return;
12ac2576 1813 }
7ebf06b3 1814
99870f4d
KW
1815 my $args_ref = shift;
1816 my $count = shift;
7ebf06b3 1817
99870f4d
KW
1818 my_carp_bug("Need at least $count arguments to "
1819 . (caller 1)[3]
1820 . ". Instead got: '"
1821 . join ', ', @$args_ref
1822 . "'. No action taken.");
1823 return;
12ac2576
JP
1824}
1825
99870f4d
KW
1826sub carp_extra_args {
1827 my $args_ref = shift;
1828 my_carp_bug("Too many arguments to 'carp_extra_args': (" . join(', ', @_) . "); Extras ignored.") if @_;
12ac2576 1829
99870f4d
KW
1830 unless (ref $args_ref) {
1831 my_carp_bug("Argument to 'carp_extra_args' ($args_ref) must be a ref. Not checking arguments.");
1832 return;
1833 }
1834 my ($package, $file, $line) = caller;
1835 my $subroutine = (caller 1)[3];
cf25bb62 1836
99870f4d
KW
1837 my $list;
1838 if (ref $args_ref eq 'HASH') {
1839 foreach my $key (keys %$args_ref) {
1840 $args_ref->{$key} = $UNDEF unless defined $args_ref->{$key};
cf25bb62 1841 }
99870f4d 1842 $list = join ', ', each %{$args_ref};
cf25bb62 1843 }
99870f4d
KW
1844 elsif (ref $args_ref eq 'ARRAY') {
1845 foreach my $arg (@$args_ref) {
1846 $arg = $UNDEF unless defined $arg;
1847 }
1848 $list = join ', ', @$args_ref;
1849 }
1850 else {
1851 my_carp_bug("Can't cope with ref "
1852 . ref($args_ref)
1853 . " . argument to 'carp_extra_args'. Not checking arguments.");
1854 return;
1855 }
1856
1857 my_carp_bug("Unrecognized parameters in options: '$list' to $subroutine. Skipped.");
1858 return;
d73e5302
JH
1859}
1860
99870f4d
KW
1861package main;
1862
1863{ # Closure
1864
1865 # This program uses the inside-out method for objects, as recommended in
1b6b3fa9
KW
1866 # "Perl Best Practices". (This is the best solution still, since this has
1867 # to run under miniperl.) This closure aids in generating those. There
99870f4d
KW
1868 # are two routines. setup_package() is called once per package to set
1869 # things up, and then set_access() is called for each hash representing a
1870 # field in the object. These routines arrange for the object to be
1871 # properly destroyed when no longer used, and for standard accessor
1872 # functions to be generated. If you need more complex accessors, just
1873 # write your own and leave those accesses out of the call to set_access().
1874 # More details below.
1875
1876 my %constructor_fields; # fields that are to be used in constructors; see
1877 # below
1878
1879 # The values of this hash will be the package names as keys to other
1880 # hashes containing the name of each field in the package as keys, and
1881 # references to their respective hashes as values.
1882 my %package_fields;
1883
1884 sub setup_package {
1885 # Sets up the package, creating standard DESTROY and dump methods
1886 # (unless already defined). The dump method is used in debugging by
1887 # simple_dumper().
1888 # The optional parameters are:
1889 # a) a reference to a hash, that gets populated by later
1890 # set_access() calls with one of the accesses being
1891 # 'constructor'. The caller can then refer to this, but it is
1892 # not otherwise used by these two routines.
1893 # b) a reference to a callback routine to call during destruction
1894 # of the object, before any fields are actually destroyed
1895
1896 my %args = @_;
1897 my $constructor_ref = delete $args{'Constructor_Fields'};
1898 my $destroy_callback = delete $args{'Destroy_Callback'};
1899 Carp::carp_extra_args(\@_) if main::DEBUG && %args;
1900
1901 my %fields;
1902 my $package = (caller)[0];
1903
1904 $package_fields{$package} = \%fields;
1905 $constructor_fields{$package} = $constructor_ref;
1906
1907 unless ($package->can('DESTROY')) {
1908 my $destroy_name = "${package}::DESTROY";
1909 no strict "refs";
1910
1911 # Use typeglob to give the anonymous subroutine the name we want
1912 *$destroy_name = sub {
1913 my $self = shift;
ffe43484 1914 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1915
1916 $self->$destroy_callback if $destroy_callback;
1917 foreach my $field (keys %{$package_fields{$package}}) {
1918 #print STDERR __LINE__, ": Destroying ", ref $self, " ", sprintf("%04X", $addr), ": ", $field, "\n";
1919 delete $package_fields{$package}{$field}{$addr};
1920 }
1921 return;
1922 }
1923 }
1924
1925 unless ($package->can('dump')) {
1926 my $dump_name = "${package}::dump";
1927 no strict "refs";
1928 *$dump_name = sub {
1929 my $self = shift;
1930 return dump_inside_out($self, $package_fields{$package}, @_);
1931 }
1932 }
1933 return;
1934 }
1935
1936 sub set_access {
1937 # Arrange for the input field to be garbage collected when no longer
1938 # needed. Also, creates standard accessor functions for the field
1939 # based on the optional parameters-- none if none of these parameters:
1940 # 'addable' creates an 'add_NAME()' accessor function.
1941 # 'readable' or 'readable_array' creates a 'NAME()' accessor
1942 # function.
1943 # 'settable' creates a 'set_NAME()' accessor function.
1944 # 'constructor' doesn't create an accessor function, but adds the
1945 # field to the hash that was previously passed to
1946 # setup_package();
1947 # Any of the accesses can be abbreviated down, so that 'a', 'ad',
1948 # 'add' etc. all mean 'addable'.
1949 # The read accessor function will work on both array and scalar
1950 # values. If another accessor in the parameter list is 'a', the read
1951 # access assumes an array. You can also force it to be array access
1952 # by specifying 'readable_array' instead of 'readable'
1953 #
1954 # A sort-of 'protected' access can be set-up by preceding the addable,
1955 # readable or settable with some initial portion of 'protected_' (but,
1956 # the underscore is required), like 'p_a', 'pro_set', etc. The
1957 # "protection" is only by convention. All that happens is that the
1958 # accessor functions' names begin with an underscore. So instead of
1959 # calling set_foo, the call is _set_foo. (Real protection could be
c1739a4a 1960 # accomplished by having a new subroutine, end_package, called at the
99870f4d
KW
1961 # end of each package, and then storing the __LINE__ ranges and
1962 # checking them on every accessor. But that is way overkill.)
1963
1964 # We create anonymous subroutines as the accessors and then use
1965 # typeglobs to assign them to the proper package and name
1966
1967 my $name = shift; # Name of the field
1968 my $field = shift; # Reference to the inside-out hash containing the
1969 # field
1970
1971 my $package = (caller)[0];
1972
1973 if (! exists $package_fields{$package}) {
1974 croak "$0: Must call 'setup_package' before 'set_access'";
1975 }
d73e5302 1976
99870f4d
KW
1977 # Stash the field so DESTROY can get it.
1978 $package_fields{$package}{$name} = $field;
cf25bb62 1979
99870f4d
KW
1980 # Remaining arguments are the accessors. For each...
1981 foreach my $access (@_) {
1982 my $access = lc $access;
cf25bb62 1983
99870f4d 1984 my $protected = "";
cf25bb62 1985
99870f4d
KW
1986 # Match the input as far as it goes.
1987 if ($access =~ /^(p[^_]*)_/) {
1988 $protected = $1;
1989 if (substr('protected_', 0, length $protected)
1990 eq $protected)
1991 {
1992
1993 # Add 1 for the underscore not included in $protected
1994 $access = substr($access, length($protected) + 1);
1995 $protected = '_';
1996 }
1997 else {
1998 $protected = "";
1999 }
2000 }
2001
2002 if (substr('addable', 0, length $access) eq $access) {
2003 my $subname = "${package}::${protected}add_$name";
2004 no strict "refs";
2005
2006 # add_ accessor. Don't add if already there, which we
2007 # determine using 'eq' for scalars and '==' otherwise.
2008 *$subname = sub {
2009 use strict "refs";
2010 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
2011 my $self = shift;
2012 my $value = shift;
ffe43484 2013 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2014 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2015 if (ref $value) {
f998e60c 2016 return if grep { $value == $_ } @{$field->{$addr}};
99870f4d
KW
2017 }
2018 else {
f998e60c 2019 return if grep { $value eq $_ } @{$field->{$addr}};
99870f4d 2020 }
f998e60c 2021 push @{$field->{$addr}}, $value;
99870f4d
KW
2022 return;
2023 }
2024 }
2025 elsif (substr('constructor', 0, length $access) eq $access) {
2026 if ($protected) {
2027 Carp::my_carp_bug("Can't set-up 'protected' constructors")
2028 }
2029 else {
2030 $constructor_fields{$package}{$name} = $field;
2031 }
2032 }
2033 elsif (substr('readable_array', 0, length $access) eq $access) {
2034
2035 # Here has read access. If one of the other parameters for
2036 # access is array, or this one specifies array (by being more
2037 # than just 'readable_'), then create a subroutine that
2038 # assumes the data is an array. Otherwise just a scalar
2039 my $subname = "${package}::${protected}$name";
2040 if (grep { /^a/i } @_
2041 or length($access) > length('readable_'))
2042 {
2043 no strict "refs";
2044 *$subname = sub {
2045 use strict "refs";
23e33b60 2046 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
ffe43484 2047 my $addr = do { no overloading; pack 'J', $_[0]; };
99870f4d
KW
2048 if (ref $field->{$addr} ne 'ARRAY') {
2049 my $type = ref $field->{$addr};
2050 $type = 'scalar' unless $type;
2051 Carp::my_carp_bug("Trying to read $name as an array when it is a $type. Big problems.");
2052 return;
2053 }
2054 return scalar @{$field->{$addr}} unless wantarray;
2055
2056 # Make a copy; had problems with caller modifying the
2057 # original otherwise
2058 my @return = @{$field->{$addr}};
2059 return @return;
2060 }
2061 }
2062 else {
2063
2064 # Here not an array value, a simpler function.
2065 no strict "refs";
2066 *$subname = sub {
2067 use strict "refs";
23e33b60 2068 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
f998e60c 2069 no overloading;
051df77b 2070 return $field->{pack 'J', $_[0]};
99870f4d
KW
2071 }
2072 }
2073 }
2074 elsif (substr('settable', 0, length $access) eq $access) {
2075 my $subname = "${package}::${protected}set_$name";
2076 no strict "refs";
2077 *$subname = sub {
2078 use strict "refs";
23e33b60
KW
2079 if (main::DEBUG) {
2080 return Carp::carp_too_few_args(\@_, 2) if @_ < 2;
2081 Carp::carp_extra_args(\@_) if @_ > 2;
2082 }
2083 # $self is $_[0]; $value is $_[1]
f998e60c 2084 no overloading;
051df77b 2085 $field->{pack 'J', $_[0]} = $_[1];
99870f4d
KW
2086 return;
2087 }
2088 }
2089 else {
2090 Carp::my_carp_bug("Unknown accessor type $access. No accessor set.");
2091 }
cf25bb62 2092 }
99870f4d 2093 return;
cf25bb62 2094 }
99870f4d
KW
2095}
2096
2097package Input_file;
2098
2099# All input files use this object, which stores various attributes about them,
2100# and provides for convenient, uniform handling. The run method wraps the
2101# processing. It handles all the bookkeeping of opening, reading, and closing
2102# the file, returning only significant input lines.
2103#
2104# Each object gets a handler which processes the body of the file, and is
74cd47d0
KW
2105# called by run(). All character property files must use the generic,
2106# default handler, which has code scrubbed to handle things you might not
2107# expect, including automatic EBCDIC handling. For files that don't deal with
2108# mapping code points to a property value, such as test files,
2109# PropertyAliases, PropValueAliases, and named sequences, you can override the
2110# handler to be a custom one. Such a handler should basically be a
2111# while(next_line()) {...} loop.
99870f4d
KW
2112#
2113# You can also set up handlers to
537124e4 2114# 1) call before the first line is read, for pre processing
83b68635
KW
2115# 2) call to adjust each line of the input before the main handler gets
2116# them. This can be automatically generated, if appropriately simple
2117# enough, by specifiying a Properties parameter in the constructor.
99870f4d 2118# 3) call upon EOF before the main handler exits its loop
537124e4 2119# 4) call at the end, for post processing
99870f4d
KW
2120#
2121# $_ is used to store the input line, and is to be filtered by the
2122# each_line_handler()s. So, if the format of the line is not in the desired
2123# format for the main handler, these are used to do that adjusting. They can
2124# be stacked (by enclosing them in an [ anonymous array ] in the constructor,
2125# so the $_ output of one is used as the input to the next. None of the other
2126# handlers are stackable, but could easily be changed to be so.
2127#
2128# Most of the handlers can call insert_lines() or insert_adjusted_lines()
2129# which insert the parameters as lines to be processed before the next input
2130# file line is read. This allows the EOF handler to flush buffers, for
2131# example. The difference between the two routines is that the lines inserted
2132# by insert_lines() are subjected to the each_line_handler()s. (So if you
2133# called it from such a handler, you would get infinite recursion.) Lines
2134# inserted by insert_adjusted_lines() go directly to the main handler without
2135# any adjustments. If the post-processing handler calls any of these, there
2136# will be no effect. Some error checking for these conditions could be added,
2137# but it hasn't been done.
2138#
2139# carp_bad_line() should be called to warn of bad input lines, which clears $_
2140# to prevent further processing of the line. This routine will output the
2141# message as a warning once, and then keep a count of the lines that have the
2142# same message, and output that count at the end of the file's processing.
2143# This keeps the number of messages down to a manageable amount.
2144#
2145# get_missings() should be called to retrieve any @missing input lines.
2146# Messages will be raised if this isn't done if the options aren't to ignore
2147# missings.
2148
2149sub trace { return main::trace(@_); }
2150
99870f4d
KW
2151{ # Closure
2152 # Keep track of fields that are to be put into the constructor.
2153 my %constructor_fields;
2154
2155 main::setup_package(Constructor_Fields => \%constructor_fields);
2156
2157 my %file; # Input file name, required
2158 main::set_access('file', \%file, qw{ c r });
2159
2160 my %first_released; # Unicode version file was first released in, required
2161 main::set_access('first_released', \%first_released, qw{ c r });
2162
2163 my %handler; # Subroutine to process the input file, defaults to
2164 # 'process_generic_property_file'
2165 main::set_access('handler', \%handler, qw{ c });
2166
2167 my %property;
2168 # name of property this file is for. defaults to none, meaning not
2169 # applicable, or is otherwise determinable, for example, from each line.
696609bf 2170 main::set_access('property', \%property, qw{ c r });
99870f4d
KW
2171
2172 my %optional;
2173 # If this is true, the file is optional. If not present, no warning is
2174 # output. If it is present, the string given by this parameter is
2175 # evaluated, and if false the file is not processed.
2176 main::set_access('optional', \%optional, 'c', 'r');
2177
2178 my %non_skip;
2179 # This is used for debugging, to skip processing of all but a few input
2180 # files. Add 'non_skip => 1' to the constructor for those files you want
2181 # processed when you set the $debug_skip global.
2182 main::set_access('non_skip', \%non_skip, 'c');
2183
37e2e78e 2184 my %skip;
09ca89ce
KW
2185 # This is used to skip processing of this input file semi-permanently,
2186 # when it evaluates to true. The value should be the reason the file is
2187 # being skipped. It is used for files that we aren't planning to process
2188 # anytime soon, but want to allow to be in the directory and not raise a
2189 # message that we are not handling. Mostly for test files. This is in
2190 # contrast to the non_skip element, which is supposed to be used very
2191 # temporarily for debugging. Sets 'optional' to 1. Also, files that we
2192 # pretty much will never look at can be placed in the global
1fec9f60 2193 # %ignored_files instead. Ones used here will be added to %skipped files
37e2e78e
KW
2194 main::set_access('skip', \%skip, 'c');
2195
99870f4d
KW
2196 my %each_line_handler;
2197 # list of subroutines to look at and filter each non-comment line in the
2198 # file. defaults to none. The subroutines are called in order, each is
2199 # to adjust $_ for the next one, and the final one adjusts it for
2200 # 'handler'
2201 main::set_access('each_line_handler', \%each_line_handler, 'c');
2202
83b68635
KW
2203 my %properties; # Optional ordered list of the properties that occur in each
2204 # meaningful line of the input file. If present, an appropriate
2205 # each_line_handler() is automatically generated and pushed onto the stack
2206 # of such handlers. This is useful when a file contains multiple
2207 # proerties per line, but no other special considerations are necessary.
2208 # The special value "<ignored>" means to discard the corresponding input
2209 # field.
2210 # Any @missing lines in the file should also match this syntax; no such
2211 # files exist as of 6.3. But if it happens in a future release, the code
2212 # could be expanded to properly parse them.
2213 main::set_access('properties', \%properties, qw{ c r });
2214
99870f4d
KW
2215 my %has_missings_defaults;
2216 # ? Are there lines in the file giving default values for code points
2217 # missing from it?. Defaults to NO_DEFAULTS. Otherwise NOT_IGNORED is
2218 # the norm, but IGNORED means it has such lines, but the handler doesn't
2219 # use them. Having these three states allows us to catch changes to the
83b68635
KW
2220 # UCD that this program should track. XXX This could be expanded to
2221 # specify the syntax for such lines, like %properties above.
99870f4d
KW
2222 main::set_access('has_missings_defaults',
2223 \%has_missings_defaults, qw{ c r });
2224
2225 my %pre_handler;
2226 # Subroutine to call before doing anything else in the file. If undef, no
2227 # such handler is called.
2228 main::set_access('pre_handler', \%pre_handler, qw{ c });
2229
2230 my %eof_handler;
2231 # Subroutine to call upon getting an EOF on the input file, but before
2232 # that is returned to the main handler. This is to allow buffers to be
2233 # flushed. The handler is expected to call insert_lines() or
2234 # insert_adjusted() with the buffered material
2235 main::set_access('eof_handler', \%eof_handler, qw{ c r });
2236
2237 my %post_handler;
2238 # Subroutine to call after all the lines of the file are read in and
2239 # processed. If undef, no such handler is called.
2240 main::set_access('post_handler', \%post_handler, qw{ c });
2241
2242 my %progress_message;
2243 # Message to print to display progress in lieu of the standard one
2244 main::set_access('progress_message', \%progress_message, qw{ c });
2245
2246 my %handle;
2247 # cache open file handle, internal. Is undef if file hasn't been
2248 # processed at all, empty if has;
2249 main::set_access('handle', \%handle);
2250
2251 my %added_lines;
2252 # cache of lines added virtually to the file, internal
2253 main::set_access('added_lines', \%added_lines);
2254
74cd47d0
KW
2255 my %remapped_lines;
2256 # cache of lines added virtually to the file, internal
2257 main::set_access('remapped_lines', \%remapped_lines);
2258
99870f4d
KW
2259 my %errors;
2260 # cache of errors found, internal
2261 main::set_access('errors', \%errors);
2262
2263 my %missings;
2264 # storage of '@missing' defaults lines
2265 main::set_access('missings', \%missings);
2266
74cd47d0
KW
2267 sub _next_line;
2268 sub _next_line_with_remapped_range;
2269
99870f4d
KW
2270 sub new {
2271 my $class = shift;
2272
2273 my $self = bless \do{ my $anonymous_scalar }, $class;
ffe43484 2274 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2275
2276 # Set defaults
2277 $handler{$addr} = \&main::process_generic_property_file;
2278 $non_skip{$addr} = 0;
37e2e78e 2279 $skip{$addr} = 0;
99870f4d
KW
2280 $has_missings_defaults{$addr} = $NO_DEFAULTS;
2281 $handle{$addr} = undef;
2282 $added_lines{$addr} = [ ];
74cd47d0 2283 $remapped_lines{$addr} = [ ];
99870f4d
KW
2284 $each_line_handler{$addr} = [ ];
2285 $errors{$addr} = { };
2286 $missings{$addr} = [ ];
2287
2288 # Two positional parameters.
99f78760 2289 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
99870f4d
KW
2290 $file{$addr} = main::internal_file_to_platform(shift);
2291 $first_released{$addr} = shift;
2292
71bd4c0b
KW
2293 undef $file{$addr} if $first_released{$addr} gt $v_version;
2294
99870f4d
KW
2295 # The rest of the arguments are key => value pairs
2296 # %constructor_fields has been set up earlier to list all possible
2297 # ones. Either set or push, depending on how the default has been set
2298 # up just above.
2299 my %args = @_;
2300 foreach my $key (keys %args) {
2301 my $argument = $args{$key};
2302
2303 # Note that the fields are the lower case of the constructor keys
2304 my $hash = $constructor_fields{lc $key};
2305 if (! defined $hash) {
2306 Carp::my_carp_bug("Unrecognized parameters '$key => $argument' to new() for $self. Skipped");
2307 next;
2308 }
2309 if (ref $hash->{$addr} eq 'ARRAY') {
2310 if (ref $argument eq 'ARRAY') {
2311 foreach my $argument (@{$argument}) {
2312 next if ! defined $argument;
2313 push @{$hash->{$addr}}, $argument;
2314 }
2315 }
2316 else {
2317 push @{$hash->{$addr}}, $argument if defined $argument;
2318 }
2319 }
2320 else {
2321 $hash->{$addr} = $argument;
2322 }
2323 delete $args{$key};
2324 };
2325
2326 # If the file has a property for it, it means that the property is not
2327 # listed in the file's entries. So add a handler to the list of line
2328 # handlers to insert the property name into the lines, to provide a
2329 # uniform interface to the final processing subroutine.
2330 # the final code doesn't have to worry about that.
2331 if ($property{$addr}) {
2332 push @{$each_line_handler{$addr}}, \&_insert_property_into_line;
2333 }
2334
2335 if ($non_skip{$addr} && ! $debug_skip && $verbosity) {
2336 print "Warning: " . __PACKAGE__ . " constructor for $file{$addr} has useless 'non_skip' in it\n";
a3a8c5f0 2337 }
99870f4d 2338
09ca89ce
KW
2339 # If skipping, set to optional, and add to list of ignored files,
2340 # including its reason
2341 if ($skip{$addr}) {
2342 $optional{$addr} = 1;
71bd4c0b 2343 $skipped_files{$file{$addr}} = $skip{$addr} if $file{$addr};
09ca89ce 2344 }
83b68635
KW
2345 elsif ($properties{$addr}) {
2346
2347 # Add a handler for each line in the input so that it creates a
2348 # separate input line for each property in those input lines, thus
2349 # making them suitable for process_generic_property_file().
2350
2351 push @{$each_line_handler{$addr}},
2352 sub {
2353 my $file = shift;
2354 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2355
2356 my @fields = split /\s*;\s*/, $_, -1;
2357
2358 if (@fields - 1 > @{$properties{$addr}}) {
2359 $file->carp_bad_line('Extra fields');
2360 $_ = "";
2361 return;
2362 }
2363 my $range = shift @fields; # 0th element is always the
2364 # range
2365
2366 # The next fields in the input line correspond
2367 # respectively to the stored properties.
2368 for my $i (0 .. @{$properties{$addr}} - 1) {
2369 my $property_name = $properties{$addr}[$i];
2370 next if $property_name eq '<ignored>';
2371 $file->insert_adjusted_lines(
2372 "$range; $property_name; $fields[$i]");
2373 }
2374 $_ = "";
2375
2376 return;
2377 };
2378 }
37e2e78e 2379
74cd47d0
KW
2380 { # On non-ascii platforms, we use a special handler
2381 no strict;
2382 no warnings 'once';
2383 *next_line = (main::NON_ASCII_PLATFORM)
2384 ? *_next_line_with_remapped_range
2385 : *_next_line;
2386 }
2387
99870f4d 2388 return $self;
d73e5302
JH
2389 }
2390
cf25bb62 2391
99870f4d
KW
2392 use overload
2393 fallback => 0,
2394 qw("") => "_operator_stringify",
2395 "." => \&main::_operator_dot,
1285127e 2396 ".=" => \&main::_operator_dot_equal,
99870f4d 2397 ;
cf25bb62 2398
99870f4d
KW
2399 sub _operator_stringify {
2400 my $self = shift;
cf25bb62 2401
99870f4d 2402 return __PACKAGE__ . " object for " . $self->file;
d73e5302 2403 }
d73e5302 2404
99870f4d
KW
2405 # flag to make sure extracted files are processed early
2406 my $seen_non_extracted_non_age = 0;
d73e5302 2407
99870f4d
KW
2408 sub run {
2409 # Process the input object $self. This opens and closes the file and
2410 # calls all the handlers for it. Currently, this can only be called
2411 # once per file, as it destroy's the EOF handler
d73e5302 2412
99870f4d
KW
2413 my $self = shift;
2414 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
b6922eda 2415
ffe43484 2416 my $addr = do { no overloading; pack 'J', $self; };
b6922eda 2417
99870f4d 2418 my $file = $file{$addr};
d73e5302 2419
99870f4d
KW
2420 # Don't process if not expecting this file (because released later
2421 # than this Unicode version), and isn't there. This means if someone
2422 # copies it into an earlier version's directory, we will go ahead and
2423 # process it.
71bd4c0b
KW
2424 return if $first_released{$addr} gt $v_version
2425 && (! defined $file || ! -e $file);
99870f4d
KW
2426
2427 # If in debugging mode and this file doesn't have the non-skip
2428 # flag set, and isn't one of the critical files, skip it.
2429 if ($debug_skip
2430 && $first_released{$addr} ne v0
2431 && ! $non_skip{$addr})
2432 {
2433 print "Skipping $file in debugging\n" if $verbosity;
2434 return;
2435 }
2436
2437 # File could be optional
37e2e78e 2438 if ($optional{$addr}) {
99870f4d
KW
2439 return unless -e $file;
2440 my $result = eval $optional{$addr};
2441 if (! defined $result) {
2442 Carp::my_carp_bug("Got '$@' when tried to eval $optional{$addr}. $file Skipped.");
2443 return;
2444 }
2445 if (! $result) {
2446 if ($verbosity) {
2447 print STDERR "Skipping processing input file '$file' because '$optional{$addr}' is not true\n";
2448 }
2449 return;
2450 }
2451 }
2452
2453 if (! defined $file || ! -e $file) {
2454
2455 # If the file doesn't exist, see if have internal data for it
2456 # (based on first_released being 0).
2457 if ($first_released{$addr} eq v0) {
2458 $handle{$addr} = 'pretend_is_open';
2459 }
2460 else {
2461 if (! $optional{$addr} # File could be optional
2462 && $v_version ge $first_released{$addr})
2463 {
2464 print STDERR "Skipping processing input file '$file' because not found\n" if $v_version ge $first_released{$addr};
2465 }
2466 return;
2467 }
2468 }
2469 else {
2470
37e2e78e
KW
2471 # Here, the file exists. Some platforms may change the case of
2472 # its name
99870f4d 2473 if ($seen_non_extracted_non_age) {
517956bf 2474 if ($file =~ /$EXTRACTED/i) {
1675ea0d 2475 Carp::my_carp_bug(main::join_lines(<<END
99f78760 2476$file should be processed just after the 'Prop...Alias' files, and before
99870f4d
KW
2477anything not in the $EXTRACTED_DIR directory. Proceeding, but the results may
2478have subtle problems
2479END
2480 ));
2481 }
2482 }
2483 elsif ($EXTRACTED_DIR
2484 && $first_released{$addr} ne v0
517956bf
CB
2485 && $file !~ /$EXTRACTED/i
2486 && lc($file) ne 'dage.txt')
99870f4d
KW
2487 {
2488 # We don't set this (by the 'if' above) if we have no
2489 # extracted directory, so if running on an early version,
2490 # this test won't work. Not worth worrying about.
2491 $seen_non_extracted_non_age = 1;
2492 }
2493
2494 # And mark the file as having being processed, and warn if it
2495 # isn't a file we are expecting. As we process the files,
2496 # they are deleted from the hash, so any that remain at the
2497 # end of the program are files that we didn't process.
517956bf 2498 my $fkey = File::Spec->rel2abs($file);
faf3cf6b
KW
2499 my $expecting = delete $potential_files{lc($fkey)};
2500
678f13d5
KW
2501 Carp::my_carp("Was not expecting '$file'.") if
2502 ! $expecting
99870f4d
KW
2503 && ! defined $handle{$addr};
2504
37e2e78e
KW
2505 # Having deleted from expected files, we can quit if not to do
2506 # anything. Don't print progress unless really want verbosity
2507 if ($skip{$addr}) {
2508 print "Skipping $file.\n" if $verbosity >= $VERBOSE;
2509 return;
2510 }
2511
99870f4d
KW
2512 # Open the file, converting the slashes used in this program
2513 # into the proper form for the OS
2514 my $file_handle;
2515 if (not open $file_handle, "<", $file) {
2516 Carp::my_carp("Can't open $file. Skipping: $!");
2517 return 0;
2518 }
2519 $handle{$addr} = $file_handle; # Cache the open file handle
9e65c3f4 2520
96f226dc
KW
2521 if ($v_version ge v3.2.0 && lc($file) ne 'unicodedata.txt') {
2522 if ($file !~ /^Unihan/i) {
cafe9cf0
KW
2523 $_ = <$file_handle>;
2524 if ($_ !~ / - $string_version \. /x) {
2525 chomp;
2526 $_ =~ s/^#\s*//;
2527 die Carp::my_carp("File '$file' is version '$_'. It should be version $string_version");
2528 }
96f226dc
KW
2529 }
2530 else {
2531 while (<$file_handle>) {
2532 if ($_ !~ /^#/) {
2533 Carp::my_carp_bug("Could not find the expected version info in file '$file'");
2534 last;
2535 }
2536 chomp;
2537 $_ =~ s/^#\s*//;
2538 next if $_ !~ / version: /x;
2539 last if $_ =~ /$string_version/;
2540 die Carp::my_carp("File '$file' is '$_'. It should be version $string_version");
2541 }
2542 }
9e65c3f4 2543 }
99870f4d
KW
2544 }
2545
2546 if ($verbosity >= $PROGRESS) {
2547 if ($progress_message{$addr}) {
2548 print "$progress_message{$addr}\n";
2549 }
2550 else {
2551 # If using a virtual file, say so.
2552 print "Processing ", (-e $file)
2553 ? $file
2554 : "substitute $file",
2555 "\n";
2556 }
2557 }
2558
2559
2560 # Call any special handler for before the file.
2561 &{$pre_handler{$addr}}($self) if $pre_handler{$addr};
2562
2563 # Then the main handler
2564 &{$handler{$addr}}($self);
2565
2566 # Then any special post-file handler.
2567 &{$post_handler{$addr}}($self) if $post_handler{$addr};
2568
2569 # If any errors have been accumulated, output the counts (as the first
2570 # error message in each class was output when it was encountered).
2571 if ($errors{$addr}) {
2572 my $total = 0;
2573 my $types = 0;
2574 foreach my $error (keys %{$errors{$addr}}) {
2575 $total += $errors{$addr}->{$error};
2576 delete $errors{$addr}->{$error};
2577 $types++;
2578 }
2579 if ($total > 1) {
2580 my $message
2581 = "A total of $total lines had errors in $file. ";
2582
2583 $message .= ($types == 1)
2584 ? '(Only the first one was displayed.)'
2585 : '(Only the first of each type was displayed.)';
2586 Carp::my_carp($message);
2587 }
2588 }
2589
2590 if (@{$missings{$addr}}) {
2591 Carp::my_carp_bug("Handler for $file didn't look at all the \@missing lines. Generated tables likely are wrong");
2592 }
2593
2594 # If a real file handle, close it.
2595 close $handle{$addr} or Carp::my_carp("Can't close $file: $!") if
2596 ref $handle{$addr};
2597 $handle{$addr} = ""; # Uses empty to indicate that has already seen
2598 # the file, as opposed to undef
2599 return;
2600 }
2601
74cd47d0 2602 sub _next_line {
99870f4d
KW
2603 # Sets $_ to be the next logical input line, if any. Returns non-zero
2604 # if such a line exists. 'logical' means that any lines that have
2605 # been added via insert_lines() will be returned in $_ before the file
2606 # is read again.
2607
2608 my $self = shift;
2609 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2610
ffe43484 2611 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2612
2613 # Here the file is open (or if the handle is not a ref, is an open
2614 # 'virtual' file). Get the next line; any inserted lines get priority
2615 # over the file itself.
2616 my $adjusted;
2617
2618 LINE:
2619 while (1) { # Loop until find non-comment, non-empty line
2620 #local $to_trace = 1 if main::DEBUG;
2621 my $inserted_ref = shift @{$added_lines{$addr}};
2622 if (defined $inserted_ref) {
2623 ($adjusted, $_) = @{$inserted_ref};
2624 trace $adjusted, $_ if main::DEBUG && $to_trace;
2625 return 1 if $adjusted;
2626 }
2627 else {
2628 last if ! ref $handle{$addr}; # Don't read unless is real file
2629 last if ! defined ($_ = readline $handle{$addr});
2630 }
2631 chomp;
2632 trace $_ if main::DEBUG && $to_trace;
2633
2634 # See if this line is the comment line that defines what property
2635 # value that code points that are not listed in the file should
2636 # have. The format or existence of these lines is not guaranteed
2637 # by Unicode since they are comments, but the documentation says
2638 # that this was added for machine-readability, so probably won't
2639 # change. This works starting in Unicode Version 5.0. They look
2640 # like:
2641 #
2642 # @missing: 0000..10FFFF; Not_Reordered
2643 # @missing: 0000..10FFFF; Decomposition_Mapping; <code point>
2644 # @missing: 0000..10FFFF; ; NaN
2645 #
2646 # Save the line for a later get_missings() call.
2647 if (/$missing_defaults_prefix/) {
2648 if ($has_missings_defaults{$addr} == $NO_DEFAULTS) {
2649 $self->carp_bad_line("Unexpected \@missing line. Assuming no missing entries");
2650 }
2651 elsif ($has_missings_defaults{$addr} == $NOT_IGNORED) {
2652 my @defaults = split /\s* ; \s*/x, $_;
2653
2654 # The first field is the @missing, which ends in a
2655 # semi-colon, so can safely shift.
2656 shift @defaults;
2657
2658 # Some of these lines may have empty field placeholders
2659 # which get in the way. An example is:
2660 # @missing: 0000..10FFFF; ; NaN
2661 # Remove them. Process starting from the top so the
2662 # splice doesn't affect things still to be looked at.
2663 for (my $i = @defaults - 1; $i >= 0; $i--) {
2664 next if $defaults[$i] ne "";
2665 splice @defaults, $i, 1;
2666 }
2667
2668 # What's left should be just the property (maybe) and the
2669 # default. Having only one element means it doesn't have
2670 # the property.
2671 my $default;
2672 my $property;
2673 if (@defaults >= 1) {
2674 if (@defaults == 1) {
2675 $default = $defaults[0];
2676 }
2677 else {
2678 $property = $defaults[0];
2679 $default = $defaults[1];
2680 }
2681 }
2682
2683 if (@defaults < 1
2684 || @defaults > 2
2685 || ($default =~ /^</
2686 && $default !~ /^<code *point>$/i
09f8d0ac
KW
2687 && $default !~ /^<none>$/i
2688 && $default !~ /^<script>$/i))
99870f4d
KW
2689 {
2690 $self->carp_bad_line("Unrecognized \@missing line: $_. Assuming no missing entries");
2691 }
2692 else {
2693
2694 # If the property is missing from the line, it should
2695 # be the one for the whole file
2696 $property = $property{$addr} if ! defined $property;
2697
2698 # Change <none> to the null string, which is what it
2699 # really means. If the default is the code point
2700 # itself, set it to <code point>, which is what
2701 # Unicode uses (but sometimes they've forgotten the
2702 # space)
2703 if ($default =~ /^<none>$/i) {
2704 $default = "";
2705 }
2706 elsif ($default =~ /^<code *point>$/i) {
2707 $default = $CODE_POINT;
2708 }
09f8d0ac
KW
2709 elsif ($default =~ /^<script>$/i) {
2710
2711 # Special case this one. Currently is from
2712 # ScriptExtensions.txt, and means for all unlisted
2713 # code points, use their Script property values.
2714 # For the code points not listed in that file, the
2715 # default value is 'Unknown'.
2716 $default = "Unknown";
2717 }
99870f4d
KW
2718
2719 # Store them as a sub-arrays with both components.
2720 push @{$missings{$addr}}, [ $default, $property ];
2721 }
2722 }
2723
2724 # There is nothing for the caller to process on this comment
2725 # line.
2726 next;
2727 }
2728
2729 # Remove comments and trailing space, and skip this line if the
2730 # result is empty
2731 s/#.*//;
2732 s/\s+$//;
2733 next if /^$/;
2734
2735 # Call any handlers for this line, and skip further processing of
2736 # the line if the handler sets the line to null.
2737 foreach my $sub_ref (@{$each_line_handler{$addr}}) {
2738 &{$sub_ref}($self);
2739 next LINE if /^$/;
2740 }
2741
2742 # Here the line is ok. return success.
2743 return 1;
2744 } # End of looping through lines.
2745
2746 # If there is an EOF handler, call it (only once) and if it generates
2747 # more lines to process go back in the loop to handle them.
2748 if ($eof_handler{$addr}) {
2749 &{$eof_handler{$addr}}($self);
2750 $eof_handler{$addr} = ""; # Currently only get one shot at it.
2751 goto LINE if $added_lines{$addr};
2752 }
2753
2754 # Return failure -- no more lines.
2755 return 0;
2756
2757 }
2758
74cd47d0
KW
2759 sub _next_line_with_remapped_range {
2760 my $self = shift;
2761 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2762
2763 # like _next_line(), but for use on non-ASCII platforms. It sets $_
2764 # to be the next logical input line, if any. Returns non-zero if such
2765 # a line exists. 'logical' means that any lines that have been added
2766 # via insert_lines() will be returned in $_ before the file is read
2767 # again.
2768 #
2769 # The difference from _next_line() is that this remaps the Unicode
2770 # code points in the input to those of the native platform. Each
2771 # input line contains a single code point, or a single contiguous
2772 # range of them This routine splits each range into its individual
2773 # code points and caches them. It returns the cached values,
2774 # translated into their native equivalents, one at a time, for each
2775 # call, before reading the next line. Since native values can only be
2776 # a single byte wide, no translation is needed for code points above
2777 # 0xFF, and ranges that are entirely above that number are not split.
2778 # If an input line contains the range 254-1000, it would be split into
2779 # three elements: 254, 255, and 256-1000. (The downstream table
2780 # insertion code will sort and coalesce the individual code points
2781 # into appropriate ranges.)
2782
2783 my $addr = do { no overloading; pack 'J', $self; };
2784
2785 while (1) {
2786
2787 # Look in cache before reading the next line. Return any cached
2788 # value, translated
2789 my $inserted = shift @{$remapped_lines{$addr}};
2790 if (defined $inserted) {
2791 trace $inserted if main::DEBUG && $to_trace;
2792 $_ = $inserted =~ s/^ ( \d+ ) /sprintf("%04X", utf8::unicode_to_native($1))/xer;
2793 trace $_ if main::DEBUG && $to_trace;
2794 return 1;
2795 }
2796
2797 # Get the next line.
2798 return 0 unless _next_line($self);
2799
2800 # If there is a special handler for it, return the line,
2801 # untranslated. This should happen only for files that are
2802 # special, not being code-point related, such as property names.
2803 return 1 if $handler{$addr}
2804 != \&main::process_generic_property_file;
2805
2806 my ($range, $property_name, $map, @remainder)
2807 = split /\s*;\s*/, $_, -1; # -1 => retain trailing null fields
2808
2809 if (@remainder
2810 || ! defined $property_name
2811 || $range !~ /^ ($code_point_re) (?:\.\. ($code_point_re) )? $/x)
2812 {
2813 Carp::my_carp_bug("Unrecognized input line '$_'. Ignored");
2814 }
2815
2816 my $low = hex $1;
2817 my $high = (defined $2) ? hex $2 : $low;
2818
2819 # If the input maps the range to another code point, remap the
2820 # target if it is between 0 and 255.
2821 my $tail;
2822 if (defined $map) {
2823 $map =~ s/\b 00 ( [0-9A-F]{2} ) \b/sprintf("%04X", utf8::unicode_to_native(hex $1))/gxe;
2824 $tail = "$property_name; $map";
2825 $_ = "$range; $tail";
2826 }
2827 else {
2828 $tail = $property_name;
2829 }
2830
2831 # If entire range is above 255, just return it, unchanged (except
2832 # any mapped-to code point, already changed above)
2833 return 1 if $low > 255;
2834
2835 # Cache an entry for every code point < 255. For those in the
2836 # range above 255, return a dummy entry for just that portion of
2837 # the range. Note that this will be out-of-order, but that is not
2838 # a problem.
2839 foreach my $code_point ($low .. $high) {
2840 if ($code_point > 255) {
2841 $_ = sprintf "%04X..%04X; $tail", $code_point, $high;
2842 return 1;
2843 }
2844 push @{$remapped_lines{$addr}}, "$code_point; $tail";
2845 }
2846 } # End of looping through lines.
2847
2848 # NOTREACHED
2849 }
2850
99870f4d
KW
2851# Not currently used, not fully tested.
2852# sub peek {
2853# # Non-destructive look-ahead one non-adjusted, non-comment, non-blank
2854# # record. Not callable from an each_line_handler(), nor does it call
2855# # an each_line_handler() on the line.
2856#
2857# my $self = shift;
ffe43484 2858# my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2859#
2860# foreach my $inserted_ref (@{$added_lines{$addr}}) {
2861# my ($adjusted, $line) = @{$inserted_ref};
2862# next if $adjusted;
2863#
2864# # Remove comments and trailing space, and return a non-empty
2865# # resulting line
2866# $line =~ s/#.*//;
2867# $line =~ s/\s+$//;
2868# return $line if $line ne "";
2869# }
2870#
2871# return if ! ref $handle{$addr}; # Don't read unless is real file
2872# while (1) { # Loop until find non-comment, non-empty line
2873# local $to_trace = 1 if main::DEBUG;
2874# trace $_ if main::DEBUG && $to_trace;
2875# return if ! defined (my $line = readline $handle{$addr});
2876# chomp $line;
2877# push @{$added_lines{$addr}}, [ 0, $line ];
2878#
2879# $line =~ s/#.*//;
2880# $line =~ s/\s+$//;
2881# return $line if $line ne "";
2882# }
2883#
2884# return;
2885# }
2886
2887
2888 sub insert_lines {
2889 # Lines can be inserted so that it looks like they were in the input
2890 # file at the place it was when this routine is called. See also
2891 # insert_adjusted_lines(). Lines inserted via this routine go through
2892 # any each_line_handler()
2893
2894 my $self = shift;
2895
2896 # Each inserted line is an array, with the first element being 0 to
2897 # indicate that this line hasn't been adjusted, and needs to be
2898 # processed.
f998e60c 2899 no overloading;
051df77b 2900 push @{$added_lines{pack 'J', $self}}, map { [ 0, $_ ] } @_;
99870f4d
KW
2901 return;
2902 }
2903
2904 sub insert_adjusted_lines {
2905 # Lines can be inserted so that it looks like they were in the input
2906 # file at the place it was when this routine is called. See also
2907 # insert_lines(). Lines inserted via this routine are already fully
2908 # adjusted, ready to be processed; each_line_handler()s handlers will
2909 # not be called. This means this is not a completely general
2910 # facility, as only the last each_line_handler on the stack should
2911 # call this. It could be made more general, by passing to each of the
2912 # line_handlers their position on the stack, which they would pass on
2913 # to this routine, and that would replace the boolean first element in
2914 # the anonymous array pushed here, so that the next_line routine could
2915 # use that to call only those handlers whose index is after it on the
2916 # stack. But this is overkill for what is needed now.
2917
2918 my $self = shift;
2919 trace $_[0] if main::DEBUG && $to_trace;
2920
2921 # Each inserted line is an array, with the first element being 1 to
2922 # indicate that this line has been adjusted
f998e60c 2923 no overloading;
051df77b 2924 push @{$added_lines{pack 'J', $self}}, map { [ 1, $_ ] } @_;
99870f4d
KW
2925 return;
2926 }
2927
2928 sub get_missings {
2929 # Returns the stored up @missings lines' values, and clears the list.
2930 # The values are in an array, consisting of the default in the first
2931 # element, and the property in the 2nd. However, since these lines
2932 # can be stacked up, the return is an array of all these arrays.
2933
2934 my $self = shift;
2935 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2936
ffe43484 2937 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2938
2939 # If not accepting a list return, just return the first one.
2940 return shift @{$missings{$addr}} unless wantarray;
2941
2942 my @return = @{$missings{$addr}};
2943 undef @{$missings{$addr}};
2944 return @return;
2945 }
2946
2947 sub _insert_property_into_line {
2948 # Add a property field to $_, if this file requires it.
2949
f998e60c 2950 my $self = shift;
ffe43484 2951 my $addr = do { no overloading; pack 'J', $self; };
f998e60c 2952 my $property = $property{$addr};
99870f4d
KW
2953 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2954
2955 $_ =~ s/(;|$)/; $property$1/;
2956 return;
2957 }
2958
2959 sub carp_bad_line {
2960 # Output consistent error messages, using either a generic one, or the
2961 # one given by the optional parameter. To avoid gazillions of the
2962 # same message in case the syntax of a file is way off, this routine
2963 # only outputs the first instance of each message, incrementing a
2964 # count so the totals can be output at the end of the file.
2965
2966 my $self = shift;
2967 my $message = shift;
2968 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2969
ffe43484 2970 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2971
2972 $message = 'Unexpected line' unless $message;
2973
2974 # No trailing punctuation so as to fit with our addenda.
2975 $message =~ s/[.:;,]$//;
2976
2977 # If haven't seen this exact message before, output it now. Otherwise
2978 # increment the count of how many times it has occurred
2979 unless ($errors{$addr}->{$message}) {
2980 Carp::my_carp("$message in '$_' in "
f998e60c 2981 . $file{$addr}
99870f4d
KW
2982 . " at line $.. Skipping this line;");
2983 $errors{$addr}->{$message} = 1;
2984 }
2985 else {
2986 $errors{$addr}->{$message}++;
2987 }
2988
2989 # Clear the line to prevent any further (meaningful) processing of it.
2990 $_ = "";
2991
2992 return;
2993 }
2994} # End closure
2995
2996package Multi_Default;
2997
2998# Certain properties in early versions of Unicode had more than one possible
2999# default for code points missing from the files. In these cases, one
3000# default applies to everything left over after all the others are applied,
3001# and for each of the others, there is a description of which class of code
3002# points applies to it. This object helps implement this by storing the
3003# defaults, and for all but that final default, an eval string that generates
3004# the class that it applies to.
3005
3006
3007{ # Closure
3008
3009 main::setup_package();
3010
3011 my %class_defaults;
3012 # The defaults structure for the classes
3013 main::set_access('class_defaults', \%class_defaults);
3014
3015 my %other_default;
3016 # The default that applies to everything left over.
3017 main::set_access('other_default', \%other_default, 'r');
3018
3019
3020 sub new {
3021 # The constructor is called with default => eval pairs, terminated by
3022 # the left-over default. e.g.
3023 # Multi_Default->new(
3024 # 'T' => '$gc->table("Mn") + $gc->table("Cf") - 0x200C
3025 # - 0x200D',
3026 # 'R' => 'some other expression that evaluates to code points',
3027 # .
3028 # .
3029 # .
3030 # 'U'));
3031
3032 my $class = shift;
3033
3034 my $self = bless \do{my $anonymous_scalar}, $class;
ffe43484 3035 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3036
3037 while (@_ > 1) {
3038 my $default = shift;
3039 my $eval = shift;
3040 $class_defaults{$addr}->{$default} = $eval;
3041 }
3042
3043 $other_default{$addr} = shift;
3044
3045 return $self;
3046 }
3047
3048 sub get_next_defaults {
3049 # Iterates and returns the next class of defaults.
3050 my $self = shift;
3051 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3052
ffe43484 3053 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3054
3055 return each %{$class_defaults{$addr}};
3056 }
3057}
3058
3059package Alias;
3060
3061# An alias is one of the names that a table goes by. This class defines them
3062# including some attributes. Everything is currently setup in the
3063# constructor.
3064
3065
3066{ # Closure
3067
3068 main::setup_package();
3069
3070 my %name;
3071 main::set_access('name', \%name, 'r');
3072
3073 my %loose_match;
c12f2655 3074 # Should this name match loosely or not.
99870f4d
KW
3075 main::set_access('loose_match', \%loose_match, 'r');
3076
33e96e72
KW
3077 my %make_re_pod_entry;
3078 # Some aliases should not get their own entries in the re section of the
3079 # pod, because they are covered by a wild-card, and some we want to
3080 # discourage use of. Binary
f82fe4ba 3081 main::set_access('make_re_pod_entry', \%make_re_pod_entry, 'r', 's');
99870f4d 3082
fd1e3e84
KW
3083 my %ucd;
3084 # Is this documented to be accessible via Unicode::UCD
3085 main::set_access('ucd', \%ucd, 'r', 's');
3086
99870f4d
KW
3087 my %status;
3088 # Aliases have a status, like deprecated, or even suppressed (which means
3089 # they don't appear in documentation). Enum
3090 main::set_access('status', \%status, 'r');
3091
0eac1e20 3092 my %ok_as_filename;
99870f4d
KW
3093 # Similarly, some aliases should not be considered as usable ones for
3094 # external use, such as file names, or we don't want documentation to
3095 # recommend them. Boolean
0eac1e20 3096 main::set_access('ok_as_filename', \%ok_as_filename, 'r');
99870f4d
KW
3097
3098 sub new {
3099 my $class = shift;
3100
3101 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 3102 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3103
3104 $name{$addr} = shift;
3105 $loose_match{$addr} = shift;
33e96e72 3106 $make_re_pod_entry{$addr} = shift;
0eac1e20 3107 $ok_as_filename{$addr} = shift;
99870f4d 3108 $status{$addr} = shift;
fd1e3e84 3109 $ucd{$addr} = shift;
99870f4d
KW
3110
3111 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3112
3113 # Null names are never ok externally
0eac1e20 3114 $ok_as_filename{$addr} = 0 if $name{$addr} eq "";
99870f4d
KW
3115
3116 return $self;
3117 }
3118}
3119
3120package Range;
3121
3122# A range is the basic unit for storing code points, and is described in the
3123# comments at the beginning of the program. Each range has a starting code
3124# point; an ending code point (not less than the starting one); a value
3125# that applies to every code point in between the two end-points, inclusive;
3126# and an enum type that applies to the value. The type is for the user's
3127# convenience, and has no meaning here, except that a non-zero type is
3128# considered to not obey the normal Unicode rules for having standard forms.
3129#
3130# The same structure is used for both map and match tables, even though in the
3131# latter, the value (and hence type) is irrelevant and could be used as a
3132# comment. In map tables, the value is what all the code points in the range
3133# map to. Type 0 values have the standardized version of the value stored as
3134# well, so as to not have to recalculate it a lot.
3135
3136sub trace { return main::trace(@_); }
3137
3138{ # Closure
3139
3140 main::setup_package();
3141
3142 my %start;
3143 main::set_access('start', \%start, 'r', 's');
3144
3145 my %end;
3146 main::set_access('end', \%end, 'r', 's');
3147
3148 my %value;
3149 main::set_access('value', \%value, 'r');
3150
3151 my %type;
3152 main::set_access('type', \%type, 'r');
3153
3154 my %standard_form;
3155 # The value in internal standard form. Defined only if the type is 0.
3156 main::set_access('standard_form', \%standard_form);
3157
3158 # Note that if these fields change, the dump() method should as well
3159
3160 sub new {
3161 return Carp::carp_too_few_args(\@_, 3) if main::DEBUG && @_ < 3;
3162 my $class = shift;
3163
3164 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 3165 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3166
3167 $start{$addr} = shift;
3168 $end{$addr} = shift;
3169
3170 my %args = @_;
3171
3172 my $value = delete $args{'Value'}; # Can be 0
3173 $value = "" unless defined $value;
3174 $value{$addr} = $value;
3175
3176 $type{$addr} = delete $args{'Type'} || 0;
3177
3178 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3179
99870f4d
KW
3180 return $self;
3181 }
3182
3183 use overload
3184 fallback => 0,
3185 qw("") => "_operator_stringify",
3186 "." => \&main::_operator_dot,
1285127e 3187 ".=" => \&main::_operator_dot_equal,
99870f4d
KW
3188 ;
3189
3190 sub _operator_stringify {
3191 my $self = shift;
ffe43484 3192 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3193
3194 # Output it like '0041..0065 (value)'
3195 my $return = sprintf("%04X", $start{$addr})
3196 . '..'
3197 . sprintf("%04X", $end{$addr});
3198 my $value = $value{$addr};
3199 my $type = $type{$addr};
3200 $return .= ' (';
3201 $return .= "$value";
3202 $return .= ", Type=$type" if $type != 0;
3203 $return .= ')';
3204
3205 return $return;
3206 }
3207
3208 sub standard_form {
c292d35a
NC
3209 # Calculate the standard form only if needed, and cache the result.
3210 # The standard form is the value itself if the type is special.
3211 # This represents a considerable CPU and memory saving - at the time
3212 # of writing there are 368676 non-special objects, but the standard
3213 # form is only requested for 22047 of them - ie about 6%.
99870f4d
KW
3214
3215 my $self = shift;
3216 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3217
ffe43484 3218 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3219
3220 return $standard_form{$addr} if defined $standard_form{$addr};
c292d35a
NC
3221
3222 my $value = $value{$addr};
3223 return $value if $type{$addr};
3224 return $standard_form{$addr} = main::standardize($value);
99870f4d
KW
3225 }
3226
3227 sub dump {
3228 # Human, not machine readable. For machine readable, comment out this
3229 # entire routine and let the standard one take effect.
3230 my $self = shift;
3231 my $indent = shift;
3232 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3233
ffe43484 3234 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3235
3236 my $return = $indent
3237 . sprintf("%04X", $start{$addr})
3238 . '..'
3239 . sprintf("%04X", $end{$addr})
3240 . " '$value{$addr}';";
3241 if (! defined $standard_form{$addr}) {
3242 $return .= "(type=$type{$addr})";
3243 }
3244 elsif ($standard_form{$addr} ne $value{$addr}) {
3245 $return .= "(standard '$standard_form{$addr}')";
3246 }
3247 return $return;
3248 }
3249} # End closure
3250
3251package _Range_List_Base;
3252
3253# Base class for range lists. A range list is simply an ordered list of
3254# ranges, so that the ranges with the lowest starting numbers are first in it.
3255#
3256# When a new range is added that is adjacent to an existing range that has the
3257# same value and type, it merges with it to form a larger range.
3258#
3259# Ranges generally do not overlap, except that there can be multiple entries
3260# of single code point ranges. This is because of NameAliases.txt.
3261#
3262# In this program, there is a standard value such that if two different
3263# values, have the same standard value, they are considered equivalent. This
3264# value was chosen so that it gives correct results on Unicode data
3265
3266# There are a number of methods to manipulate range lists, and some operators
3267# are overloaded to handle them.
3268
99870f4d
KW
3269sub trace { return main::trace(@_); }
3270
3271{ # Closure
3272
3273 our $addr;
3274
5b348b71
KW
3275 # Max is initialized to a negative value that isn't adjacent to 0, for
3276 # simpler tests
3277 my $max_init = -2;
3278
99870f4d
KW
3279 main::setup_package();
3280
3281 my %ranges;
3282 # The list of ranges
3283 main::set_access('ranges', \%ranges, 'readable_array');
3284
3285 my %max;
3286 # The highest code point in the list. This was originally a method, but
3287 # actual measurements said it was used a lot.
3288 main::set_access('max', \%max, 'r');
3289
3290 my %each_range_iterator;
3291 # Iterator position for each_range()
3292 main::set_access('each_range_iterator', \%each_range_iterator);
3293
3294 my %owner_name_of;
3295 # Name of parent this is attached to, if any. Solely for better error
3296 # messages.
3297 main::set_access('owner_name_of', \%owner_name_of, 'p_r');
3298
3299 my %_search_ranges_cache;
3300 # A cache of the previous result from _search_ranges(), for better
3301 # performance
3302 main::set_access('_search_ranges_cache', \%_search_ranges_cache);
3303
3304 sub new {
3305 my $class = shift;
3306 my %args = @_;
3307
3308 # Optional initialization data for the range list.
3309 my $initialize = delete $args{'Initialize'};
3310
3311 my $self;
3312
3313 # Use _union() to initialize. _union() returns an object of this
3314 # class, which means that it will call this constructor recursively.
3315 # But it won't have this $initialize parameter so that it won't
3316 # infinitely loop on this.
3317 return _union($class, $initialize, %args) if defined $initialize;
3318
3319 $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 3320 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3321
3322 # Optional parent object, only for debug info.
3323 $owner_name_of{$addr} = delete $args{'Owner'};
3324 $owner_name_of{$addr} = "" if ! defined $owner_name_of{$addr};
3325
3326 # Stringify, in case it is an object.
3327 $owner_name_of{$addr} = "$owner_name_of{$addr}";
3328
3329 # This is used only for error messages, and so a colon is added
3330 $owner_name_of{$addr} .= ": " if $owner_name_of{$addr} ne "";
3331
3332 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3333
5b348b71 3334 $max{$addr} = $max_init;
99870f4d
KW
3335
3336 $_search_ranges_cache{$addr} = 0;
3337 $ranges{$addr} = [];
3338
3339 return $self;
3340 }
3341
3342 use overload
3343 fallback => 0,
3344 qw("") => "_operator_stringify",
3345 "." => \&main::_operator_dot,
1285127e 3346 ".=" => \&main::_operator_dot_equal,
99870f4d
KW
3347 ;
3348
3349 sub _operator_stringify {
3350 my $self = shift;
ffe43484 3351 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3352
3353 return "Range_List attached to '$owner_name_of{$addr}'"
3354 if $owner_name_of{$addr};
3355 return "anonymous Range_List " . \$self;
3356 }
3357
3358 sub _union {
3359 # Returns the union of the input code points. It can be called as
3360 # either a constructor or a method. If called as a method, the result
3361 # will be a new() instance of the calling object, containing the union
3362 # of that object with the other parameter's code points; if called as
d59563d0 3363 # a constructor, the first parameter gives the class that the new object
99870f4d
KW
3364 # should be, and the second parameter gives the code points to go into
3365 # it.
3366 # In either case, there are two parameters looked at by this routine;
3367 # any additional parameters are passed to the new() constructor.
3368 #
3369 # The code points can come in the form of some object that contains
3370 # ranges, and has a conventionally named method to access them; or
3371 # they can be an array of individual code points (as integers); or
3372 # just a single code point.
3373 #
3374 # If they are ranges, this routine doesn't make any effort to preserve
3198cc57
KW
3375 # the range values and types of one input over the other. Therefore
3376 # this base class should not allow _union to be called from other than
99870f4d
KW
3377 # initialization code, so as to prevent two tables from being added
3378 # together where the range values matter. The general form of this
3379 # routine therefore belongs in a derived class, but it was moved here
3380 # to avoid duplication of code. The failure to overload this in this
3381 # class keeps it safe.
3198cc57
KW
3382 #
3383 # It does make the effort during initialization to accept tables with
3384 # multiple values for the same code point, and to preserve the order
3385 # of these. If there is only one input range or range set, it doesn't
3386 # sort (as it should already be sorted to the desired order), and will
3387 # accept multiple values per code point. Otherwise it will merge
3388 # multiple values into a single one.
99870f4d
KW
3389
3390 my $self;
3391 my @args; # Arguments to pass to the constructor
3392
3393 my $class = shift;
3394
3395 # If a method call, will start the union with the object itself, and
3396 # the class of the new object will be the same as self.
3397 if (ref $class) {
3398 $self = $class;
3399 $class = ref $self;
3400 push @args, $self;
3401 }
3402
3403 # Add the other required parameter.
3404 push @args, shift;
3405 # Rest of parameters are passed on to the constructor
3406
3407 # Accumulate all records from both lists.
3408 my @records;
3198cc57 3409 my $input_count = 0;
99870f4d
KW
3410 for my $arg (@args) {
3411 #local $to_trace = 0 if main::DEBUG;
3412 trace "argument = $arg" if main::DEBUG && $to_trace;
3413 if (! defined $arg) {
3414 my $message = "";
3415 if (defined $self) {
f998e60c 3416 no overloading;
051df77b 3417 $message .= $owner_name_of{pack 'J', $self};
99870f4d 3418 }
ada6088e 3419 Carp::my_carp_bug($message . "Undefined argument to _union. No union done.");
99870f4d
KW
3420 return;
3421 }
3198cc57 3422
99870f4d
KW
3423 $arg = [ $arg ] if ! ref $arg;
3424 my $type = ref $arg;
3425 if ($type eq 'ARRAY') {
3426 foreach my $element (@$arg) {
3427 push @records, Range->new($element, $element);
3198cc57 3428 $input_count++;
99870f4d
KW
3429 }
3430 }
3431 elsif ($arg->isa('Range')) {
3432 push @records, $arg;
3198cc57 3433 $input_count++;
99870f4d
KW
3434 }
3435 elsif ($arg->can('ranges')) {
3436 push @records, $arg->ranges;
3198cc57 3437 $input_count++;
99870f4d
KW
3438 }
3439 else {
3440 my $message = "";
3441 if (defined $self) {
f998e60c 3442 no overloading;
051df77b 3443 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
3444 }
3445 Carp::my_carp_bug($message . "Cannot take the union of a $type. No union done.");
3446 return;
3447 }
3448 }
3449
3450 # Sort with the range containing the lowest ordinal first, but if
3451 # two ranges start at the same code point, sort with the bigger range
3452 # of the two first, because it takes fewer cycles.
3198cc57
KW
3453 if ($input_count > 1) {
3454 @records = sort { ($a->start <=> $b->start)
99870f4d
KW
3455 or
3456 # if b is shorter than a, b->end will be
3457 # less than a->end, and we want to select
3458 # a, so want to return -1
3459 ($b->end <=> $a->end)
3460 } @records;
3198cc57 3461 }
99870f4d
KW
3462
3463 my $new = $class->new(@_);
3464
3465 # Fold in records so long as they add new information.
3466 for my $set (@records) {
3467 my $start = $set->start;
3468 my $end = $set->end;
d59563d0 3469 my $value = $set->value;
3198cc57 3470 my $type = $set->type;
99870f4d 3471 if ($start > $new->max) {
3198cc57 3472 $new->_add_delete('+', $start, $end, $value, Type => $type);
99870f4d
KW
3473 }
3474 elsif ($end > $new->max) {
3198cc57
KW
3475 $new->_add_delete('+', $new->max +1, $end, $value,
3476 Type => $type);
3477 }
3478 elsif ($input_count == 1) {
3479 # Here, overlaps existing range, but is from a single input,
3480 # so preserve the multiple values from that input.
3481 $new->_add_delete('+', $start, $end, $value, Type => $type,
3482 Replace => $MULTIPLE_AFTER);
99870f4d
KW
3483 }
3484 }
3485
3486 return $new;
3487 }
3488
3489 sub range_count { # Return the number of ranges in the range list
3490 my $self = shift;
3491 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3492
f998e60c 3493 no overloading;
051df77b 3494 return scalar @{$ranges{pack 'J', $self}};
99870f4d
KW
3495 }
3496
3497 sub min {
3498 # Returns the minimum code point currently in the range list, or if
3499 # the range list is empty, 2 beyond the max possible. This is a
3500 # method because used so rarely, that not worth saving between calls,
3501 # and having to worry about changing it as ranges are added and
3502 # deleted.
3503
3504 my $self = shift;
3505 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3506
ffe43484 3507 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3508
3509 # If the range list is empty, return a large value that isn't adjacent
3510 # to any that could be in the range list, for simpler tests
2d88a86a 3511 return $MAX_WORKING_CODEPOINT + 2 unless scalar @{$ranges{$addr}};
99870f4d
KW
3512 return $ranges{$addr}->[0]->start;
3513 }
3514
3515 sub contains {
3516 # Boolean: Is argument in the range list? If so returns $i such that:
3517 # range[$i]->end < $codepoint <= range[$i+1]->end
3518 # which is one beyond what you want; this is so that the 0th range
3519 # doesn't return false
3520 my $self = shift;
3521 my $codepoint = shift;
3522 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3523
99870f4d
KW
3524 my $i = $self->_search_ranges($codepoint);
3525 return 0 unless defined $i;
3526
3527 # The search returns $i, such that
3528 # range[$i-1]->end < $codepoint <= range[$i]->end
3529 # So is in the table if and only iff it is at least the start position
3530 # of range $i.
f998e60c 3531 no overloading;
051df77b 3532 return 0 if $ranges{pack 'J', $self}->[$i]->start > $codepoint;
99870f4d
KW
3533 return $i + 1;
3534 }
3535
2f7a8815
KW
3536 sub containing_range {
3537 # Returns the range object that contains the code point, undef if none
3538
3539 my $self = shift;
3540 my $codepoint = shift;
3541 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3542
3543 my $i = $self->contains($codepoint);
3544 return unless $i;
3545
3546 # contains() returns 1 beyond where we should look
3547 no overloading;
3548 return $ranges{pack 'J', $self}->[$i-1];
3549 }
3550
99870f4d
KW
3551 sub value_of {
3552 # Returns the value associated with the code point, undef if none
3553
3554 my $self = shift;
3555 my $codepoint = shift;
3556 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3557
d69c231b
KW
3558 my $range = $self->containing_range($codepoint);
3559 return unless defined $range;
99870f4d 3560
d69c231b 3561 return $range->value;
99870f4d
KW
3562 }
3563
0a9dbafc
KW
3564 sub type_of {
3565 # Returns the type of the range containing the code point, undef if
3566 # the code point is not in the table
3567
3568 my $self = shift;
3569 my $codepoint = shift;
3570 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3571
3572 my $range = $self->containing_range($codepoint);
3573 return unless defined $range;
3574
3575 return $range->type;
3576 }
3577
99870f4d
KW
3578 sub _search_ranges {
3579 # Find the range in the list which contains a code point, or where it
3580 # should go if were to add it. That is, it returns $i, such that:
3581 # range[$i-1]->end < $codepoint <= range[$i]->end
3582 # Returns undef if no such $i is possible (e.g. at end of table), or
3583 # if there is an error.
3584
3585 my $self = shift;
3586 my $code_point = shift;
3587 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3588
ffe43484 3589 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3590
3591 return if $code_point > $max{$addr};
3592 my $r = $ranges{$addr}; # The current list of ranges
3593 my $range_list_size = scalar @$r;
3594 my $i;
3595
3596 use integer; # want integer division
3597
3598 # Use the cached result as the starting guess for this one, because,
3599 # an experiment on 5.1 showed that 90% of the time the cache was the
3600 # same as the result on the next call (and 7% it was one less).
3601 $i = $_search_ranges_cache{$addr};
3602 $i = 0 if $i >= $range_list_size; # Reset if no longer valid (prob.
3603 # from an intervening deletion
3604 #local $to_trace = 1 if main::DEBUG;
3605 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);
3606 return $i if $code_point <= $r->[$i]->end
3607 && ($i == 0 || $r->[$i-1]->end < $code_point);
3608
3609 # Here the cache doesn't yield the correct $i. Try adding 1.
3610 if ($i < $range_list_size - 1
3611 && $r->[$i]->end < $code_point &&
3612 $code_point <= $r->[$i+1]->end)
3613 {
3614 $i++;
3615 trace "next \$i is correct: $i" if main::DEBUG && $to_trace;
3616 $_search_ranges_cache{$addr} = $i;
3617 return $i;
3618 }
3619
3620 # Here, adding 1 also didn't work. We do a binary search to
3621 # find the correct position, starting with current $i
3622 my $lower = 0;
3623 my $upper = $range_list_size - 1;
3624 while (1) {
3625 trace "top of loop i=$i:", sprintf("%04X", $r->[$lower]->start), "[$lower] .. ", sprintf("%04X", $r->[$i]->start), "[$i] .. ", sprintf("%04X", $r->[$upper]->start), "[$upper]" if main::DEBUG && $to_trace;
3626
3627 if ($code_point <= $r->[$i]->end) {
3628
3629 # Here we have met the upper constraint. We can quit if we
3630 # also meet the lower one.
3631 last if $i == 0 || $r->[$i-1]->end < $code_point;
3632
3633 $upper = $i; # Still too high.
3634
3635 }
3636 else {
3637
3638 # Here, $r[$i]->end < $code_point, so look higher up.
3639 $lower = $i;
3640 }
3641
3642 # Split search domain in half to try again.
3643 my $temp = ($upper + $lower) / 2;
3644
3645 # No point in continuing unless $i changes for next time
3646 # in the loop.
3647 if ($temp == $i) {
3648
3649 # We can't reach the highest element because of the averaging.
3650 # So if one below the upper edge, force it there and try one
3651 # more time.
3652 if ($i == $range_list_size - 2) {
3653
3654 trace "Forcing to upper edge" if main::DEBUG && $to_trace;
3655 $i = $range_list_size - 1;
3656
3657 # Change $lower as well so if fails next time through,
3658 # taking the average will yield the same $i, and we will
3659 # quit with the error message just below.
3660 $lower = $i;
3661 next;
3662 }
3663 Carp::my_carp_bug("$owner_name_of{$addr}Can't find where the range ought to go. No action taken.");
3664 return;
3665 }
3666 $i = $temp;
3667 } # End of while loop
3668
3669 if (main::DEBUG && $to_trace) {
3670 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i;
3671 trace "i= [ $i ]", $r->[$i];
3672 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < $range_list_size - 1;
3673 }
3674
3675 # Here we have found the offset. Cache it as a starting point for the
3676 # next call.
3677 $_search_ranges_cache{$addr} = $i;
3678 return $i;
3679 }
3680
3681 sub _add_delete {
3682 # Add, replace or delete ranges to or from a list. The $type
3683 # parameter gives which:
3684 # '+' => insert or replace a range, returning a list of any changed
3685 # ranges.
3686 # '-' => delete a range, returning a list of any deleted ranges.
3687 #
3688 # The next three parameters give respectively the start, end, and
3689 # value associated with the range. 'value' should be null unless the
3690 # operation is '+';
3691 #
3692 # The range list is kept sorted so that the range with the lowest
3693 # starting position is first in the list, and generally, adjacent
c1739a4a 3694 # ranges with the same values are merged into a single larger one (see
99870f4d
KW
3695 # exceptions below).
3696 #
c1739a4a 3697 # There are more parameters; all are key => value pairs:
99870f4d
KW
3698 # Type gives the type of the value. It is only valid for '+'.
3699 # All ranges have types; if this parameter is omitted, 0 is
3700 # assumed. Ranges with type 0 are assumed to obey the
3701 # Unicode rules for casing, etc; ranges with other types are
3702 # not. Otherwise, the type is arbitrary, for the caller's
3703 # convenience, and looked at only by this routine to keep
3704 # adjacent ranges of different types from being merged into
3705 # a single larger range, and when Replace =>
3706 # $IF_NOT_EQUIVALENT is specified (see just below).
3707 # Replace determines what to do if the range list already contains
3708 # ranges which coincide with all or portions of the input
3709 # range. It is only valid for '+':
3710 # => $NO means that the new value is not to replace
3711 # any existing ones, but any empty gaps of the
3712 # range list coinciding with the input range
3713 # will be filled in with the new value.
3714 # => $UNCONDITIONALLY means to replace the existing values with
3715 # this one unconditionally. However, if the
3716 # new and old values are identical, the
3717 # replacement is skipped to save cycles
3718 # => $IF_NOT_EQUIVALENT means to replace the existing values
d59563d0 3719 # (the default) with this one if they are not equivalent.
99870f4d 3720 # Ranges are equivalent if their types are the
c1739a4a 3721 # same, and they are the same string; or if
99870f4d
KW
3722 # both are type 0 ranges, if their Unicode
3723 # standard forms are identical. In this last
3724 # case, the routine chooses the more "modern"
3725 # one to use. This is because some of the
3726 # older files are formatted with values that
3727 # are, for example, ALL CAPs, whereas the
3728 # derived files have a more modern style,
3729 # which looks better. By looking for this
3730 # style when the pre-existing and replacement
3731 # standard forms are the same, we can move to
3732 # the modern style
9470941f 3733 # => $MULTIPLE_BEFORE means that if this range duplicates an
99870f4d
KW
3734 # existing one, but has a different value,
3735 # don't replace the existing one, but insert
3736 # this, one so that the same range can occur
53d84487
KW
3737 # multiple times. They are stored LIFO, so
3738 # that the final one inserted is the first one
3739 # returned in an ordered search of the table.
6901521e
KW
3740 # If this is an exact duplicate, including the
3741 # value, the original will be moved to be
3742 # first, before any other duplicate ranges
3743 # with different values.
7f4b1e25
KW
3744 # => $MULTIPLE_AFTER is like $MULTIPLE_BEFORE, but is stored
3745 # FIFO, so that this one is inserted after all
6901521e
KW
3746 # others that currently exist. If this is an