This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
mktables: Use chr() instead of pack()
[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#
491# Simply change to a carp. It will compile, but will not know about any three
492# character decomposition.
493
494# The number of code points in \p{alpha=True} halved in 2.1.9. It turns out
495# that the reason is that the CJK block starting at 4E00 was removed from
496# PropList, and was not put back in until 3.1.0. The Perl extension (the
497# single property name \p{alpha}) has the correct values. But the compound
498# form is simply not generated until 3.1, as it can be argued that prior to
499# this release, this was not an official property. The comments for
500# filter_old_style_proplist() give more details.
678f13d5
KW
501#
502# Unicode introduced the synonym Space for White_Space in 4.1. Perl has
503# always had a \p{Space}. In release 3.2 only, they are not synonymous. The
504# reason is that 3.2 introduced U+205F=medium math space, which was not
505# classed as white space, but Perl figured out that it should have been. 4.0
506# reclassified it correctly.
507#
508# Another change between 3.2 and 4.0 is the CCC property value ATBL. In 3.2
232ed87f
KW
509# this was erroneously a synonym for 202 (it should be 200). In 4.0, ATB
510# became 202, and ATBL was left with no code points, as all the ones that
511# mapped to 202 stayed mapped to 202. Thus if your program used the numeric
512# name for the class, it would not have been affected, but if it used the
513# mnemonic, it would have been.
678f13d5
KW
514#
515# \p{Script=Hrkt} (Katakana_Or_Hiragana) came in 4.0.1. Before that code
516# points which eventually came to have this script property value, instead
517# mapped to "Unknown". But in the next release all these code points were
518# moved to \p{sc=common} instead.
99870f4d
KW
519#
520# The default for missing code points for BidiClass is complicated. Starting
521# in 3.1.1, the derived file DBidiClass.txt handles this, but this program
522# tries to do the best it can for earlier releases. It is done in
523# process_PropertyAliases()
524#
232ed87f
KW
525# In version 2.1.2, the entry in UnicodeData.txt:
526# 0275;LATIN SMALL LETTER BARRED O;Ll;0;L;;;;;N;;;;019F;
527# should instead be
528# 0275;LATIN SMALL LETTER BARRED O;Ll;0;L;;;;;N;;;019F;;019F
529# Without this change, there are casing problems for this character.
530#
7803ad2d
KW
531# Search for $string_compare_versions to see how to compare changes to
532# properties between Unicode versions
533#
99870f4d
KW
534##############################################################################
535
536my $UNDEF = ':UNDEF:'; # String to print out for undefined values in tracing
537 # and errors
538my $MAX_LINE_WIDTH = 78;
539
540# Debugging aid to skip most files so as to not be distracted by them when
541# concentrating on the ones being debugged. Add
542# non_skip => 1,
543# to the constructor for those files you want processed when you set this.
544# Files with a first version number of 0 are special: they are always
c12f2655
KW
545# processed regardless of the state of this flag. Generally, Jamo.txt and
546# UnicodeData.txt must not be skipped if you want this program to not die
547# before normal completion.
99870f4d
KW
548my $debug_skip = 0;
549
e9c4b4f8
KW
550
551# Normally these are suppressed.
552my $write_Unicode_deprecated_tables = 0;
553
99870f4d
KW
554# Set to 1 to enable tracing.
555our $to_trace = 0;
556
557{ # Closure for trace: debugging aid
558 my $print_caller = 1; # ? Include calling subroutine name
559 my $main_with_colon = 'main::';
560 my $main_colon_length = length($main_with_colon);
561
562 sub trace {
563 return unless $to_trace; # Do nothing if global flag not set
564
565 my @input = @_;
566
567 local $DB::trace = 0;
568 $DB::trace = 0; # Quiet 'used only once' message
569
570 my $line_number;
571
572 # Loop looking up the stack to get the first non-trace caller
573 my $caller_line;
574 my $caller_name;
575 my $i = 0;
576 do {
577 $line_number = $caller_line;
578 (my $pkg, my $file, $caller_line, my $caller) = caller $i++;
579 $caller = $main_with_colon unless defined $caller;
580
581 $caller_name = $caller;
582
583 # get rid of pkg
584 $caller_name =~ s/.*:://;
585 if (substr($caller_name, 0, $main_colon_length)
586 eq $main_with_colon)
587 {
588 $caller_name = substr($caller_name, $main_colon_length);
589 }
590
591 } until ($caller_name ne 'trace');
592
593 # If the stack was empty, we were called from the top level
594 $caller_name = 'main' if ($caller_name eq ""
595 || $caller_name eq 'trace');
596
597 my $output = "";
598 foreach my $string (@input) {
599 #print STDERR __LINE__, ": ", join ", ", @input, "\n";
600 if (ref $string eq 'ARRAY' || ref $string eq 'HASH') {
601 $output .= simple_dumper($string);
602 }
603 else {
604 $string = "$string" if ref $string;
605 $string = $UNDEF unless defined $string;
606 chomp $string;
607 $string = '""' if $string eq "";
608 $output .= " " if $output ne ""
609 && $string ne ""
610 && substr($output, -1, 1) ne " "
611 && substr($string, 0, 1) ne " ";
612 $output .= $string;
613 }
614 }
615
99f78760
KW
616 print STDERR sprintf "%4d: ", $line_number if defined $line_number;
617 print STDERR "$caller_name: " if $print_caller;
99870f4d
KW
618 print STDERR $output, "\n";
619 return;
620 }
621}
622
623# This is for a rarely used development feature that allows you to compare two
624# versions of the Unicode standard without having to deal with changes caused
c12f2655
KW
625# by the code points introduced in the later version. Change the 0 to a
626# string containing a SINGLE dotted Unicode release number (e.g. "2.1"). Only
627# code points introduced in that release and earlier will be used; later ones
628# are thrown away. You use the version number of the earliest one you want to
629# compare; then run this program on directory structures containing each
630# release, and compare the outputs. These outputs will therefore include only
631# the code points common to both releases, and you can see the changes caused
632# just by the underlying release semantic changes. For versions earlier than
633# 3.2, you must copy a version of DAge.txt into the directory.
634my $string_compare_versions = DEBUG && 0; # e.g., "2.1";
99870f4d
KW
635my $compare_versions = DEBUG
636 && $string_compare_versions
637 && pack "C*", split /\./, $string_compare_versions;
638
639sub uniques {
640 # Returns non-duplicated input values. From "Perl Best Practices:
641 # Encapsulated Cleverness". p. 455 in first edition.
642
643 my %seen;
0e407844
NC
644 # Arguably this breaks encapsulation, if the goal is to permit multiple
645 # distinct objects to stringify to the same value, and be interchangeable.
646 # However, for this program, no two objects stringify identically, and all
647 # lists passed to this function are either objects or strings. So this
648 # doesn't affect correctness, but it does give a couple of percent speedup.
649 no overloading;
99870f4d
KW
650 return grep { ! $seen{$_}++ } @_;
651}
652
653$0 = File::Spec->canonpath($0);
654
655my $make_test_script = 0; # ? Should we output a test script
6b5ab373 656my $make_norm_test_script = 0; # ? Should we output a normalization test script
99870f4d
KW
657my $write_unchanged_files = 0; # ? Should we update the output files even if
658 # we don't think they have changed
659my $use_directory = ""; # ? Should we chdir somewhere.
660my $pod_directory; # input directory to store the pod file.
661my $pod_file = 'perluniprops';
662my $t_path; # Path to the .t test file
663my $file_list = 'mktables.lst'; # File to store input and output file names.
664 # This is used to speed up the build, by not
665 # executing the main body of the program if
666 # nothing on the list has changed since the
667 # previous build
668my $make_list = 1; # ? Should we write $file_list. Set to always
669 # make a list so that when the pumpking is
670 # preparing a release, s/he won't have to do
671 # special things
672my $glob_list = 0; # ? Should we try to include unknown .txt files
673 # in the input.
bd9ebcfd
KW
674my $output_range_counts = $debugging_build; # ? Should we include the number
675 # of code points in ranges in
676 # the output
558712cf 677my $annotate = 0; # ? Should character names be in the output
9ef2b94f 678
99870f4d
KW
679# Verbosity levels; 0 is quiet
680my $NORMAL_VERBOSITY = 1;
681my $PROGRESS = 2;
682my $VERBOSE = 3;
683
684my $verbosity = $NORMAL_VERBOSITY;
685
0458fbc1
KW
686# Stored in mktables.lst so that if this program is called with different
687# options, will regenerate even if the files otherwise look like they're
688# up-to-date.
689my $command_line_arguments = join " ", @ARGV;
690
99870f4d
KW
691# Process arguments
692while (@ARGV) {
cf25bb62
JH
693 my $arg = shift @ARGV;
694 if ($arg eq '-v') {
99870f4d
KW
695 $verbosity = $VERBOSE;
696 }
697 elsif ($arg eq '-p') {
698 $verbosity = $PROGRESS;
699 $| = 1; # Flush buffers as we go.
700 }
701 elsif ($arg eq '-q') {
702 $verbosity = 0;
703 }
704 elsif ($arg eq '-w') {
705 $write_unchanged_files = 1; # update the files even if havent changed
706 }
707 elsif ($arg eq '-check') {
6ae7e459
YO
708 my $this = shift @ARGV;
709 my $ok = shift @ARGV;
710 if ($this ne $ok) {
711 print "Skipping as check params are not the same.\n";
712 exit(0);
713 }
00a8df5c 714 }
99870f4d
KW
715 elsif ($arg eq '-P' && defined ($pod_directory = shift)) {
716 -d $pod_directory or croak "Directory '$pod_directory' doesn't exist";
717 }
3df51b85
KW
718 elsif ($arg eq '-maketest' || ($arg eq '-T' && defined ($t_path = shift)))
719 {
99870f4d 720 $make_test_script = 1;
99870f4d 721 }
6b5ab373
KW
722 elsif ($arg eq '-makenormtest')
723 {
724 $make_norm_test_script = 1;
725 }
99870f4d
KW
726 elsif ($arg eq '-makelist') {
727 $make_list = 1;
728 }
729 elsif ($arg eq '-C' && defined ($use_directory = shift)) {
730 -d $use_directory or croak "Unknown directory '$use_directory'";
731 }
732 elsif ($arg eq '-L') {
733
734 # Existence not tested until have chdir'd
735 $file_list = shift;
736 }
737 elsif ($arg eq '-globlist') {
738 $glob_list = 1;
739 }
740 elsif ($arg eq '-c') {
741 $output_range_counts = ! $output_range_counts
742 }
b4a0206c 743 elsif ($arg eq '-annotate') {
558712cf 744 $annotate = 1;
bd9ebcfd
KW
745 $debugging_build = 1;
746 $output_range_counts = 1;
9ef2b94f 747 }
99870f4d
KW
748 else {
749 my $with_c = 'with';
750 $with_c .= 'out' if $output_range_counts; # Complements the state
751 croak <<END;
752usage: $0 [-c|-p|-q|-v|-w] [-C dir] [-L filelist] [ -P pod_dir ]
753 [ -T test_file_path ] [-globlist] [-makelist] [-maketest]
754 [-check A B ]
755 -c : Output comments $with_c number of code points in ranges
756 -q : Quiet Mode: Only output serious warnings.
757 -p : Set verbosity level to normal plus show progress.
758 -v : Set Verbosity level high: Show progress and non-serious
759 warnings
760 -w : Write files regardless
761 -C dir : Change to this directory before proceeding. All relative paths
762 except those specified by the -P and -T options will be done
763 with respect to this directory.
764 -P dir : Output $pod_file file to directory 'dir'.
3df51b85 765 -T path : Create a test script as 'path'; overrides -maketest
99870f4d
KW
766 -L filelist : Use alternate 'filelist' instead of standard one
767 -globlist : Take as input all non-Test *.txt files in current and sub
768 directories
3df51b85
KW
769 -maketest : Make test script 'TestProp.pl' in current (or -C directory),
770 overrides -T
99870f4d 771 -makelist : Rewrite the file list $file_list based on current setup
b4a0206c 772 -annotate : Output an annotation for each character in the table files;
97a8a595
KW
773 useful for debugging mktables, looking at diffs; but is slow
774 and memory intensive
99870f4d
KW
775 -check A B : Executes $0 only if A and B are the same
776END
777 }
778}
779
780# Stores the most-recently changed file. If none have changed, can skip the
781# build
aeab6150 782my $most_recent = (stat $0)[9]; # Do this before the chdir!
99870f4d
KW
783
784# Change directories now, because need to read 'version' early.
785if ($use_directory) {
3df51b85 786 if ($pod_directory && ! File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d
KW
787 $pod_directory = File::Spec->rel2abs($pod_directory);
788 }
3df51b85 789 if ($t_path && ! File::Spec->file_name_is_absolute($t_path)) {
99870f4d 790 $t_path = File::Spec->rel2abs($t_path);
00a8df5c 791 }
99870f4d 792 chdir $use_directory or croak "Failed to chdir to '$use_directory':$!";
3df51b85 793 if ($pod_directory && File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d 794 $pod_directory = File::Spec->abs2rel($pod_directory);
02b1aeec 795 }
3df51b85 796 if ($t_path && File::Spec->file_name_is_absolute($t_path)) {
99870f4d 797 $t_path = File::Spec->abs2rel($t_path);
02b1aeec 798 }
00a8df5c
YO
799}
800
99870f4d
KW
801# Get Unicode version into regular and v-string. This is done now because
802# various tables below get populated based on it. These tables are populated
803# here to be near the top of the file, and so easily seeable by those needing
804# to modify things.
805open my $VERSION, "<", "version"
806 or croak "$0: can't open required file 'version': $!\n";
807my $string_version = <$VERSION>;
808close $VERSION;
809chomp $string_version;
810my $v_version = pack "C*", split /\./, $string_version; # v string
811
812# The following are the complete names of properties with property values that
813# are known to not match any code points in some versions of Unicode, but that
814# may change in the future so they should be matchable, hence an empty file is
815# generated for them.
a9c9e371
KW
816my @tables_that_may_be_empty;
817push @tables_that_may_be_empty, 'Joining_Type=Left_Joining'
818 if $v_version lt v6.3.0;
99870f4d
KW
819push @tables_that_may_be_empty, 'Script=Common' if $v_version le v4.0.1;
820push @tables_that_may_be_empty, 'Title' if $v_version lt v2.0.0;
821push @tables_that_may_be_empty, 'Script=Katakana_Or_Hiragana'
822 if $v_version ge v4.1.0;
82aed44a
KW
823push @tables_that_may_be_empty, 'Script_Extensions=Katakana_Or_Hiragana'
824 if $v_version ge v6.0.0;
f583b44c
KW
825push @tables_that_may_be_empty, 'Grapheme_Cluster_Break=Prepend'
826 if $v_version ge v6.1.0;
1e958ea9
KW
827push @tables_that_may_be_empty, 'Canonical_Combining_Class=CCC133'
828 if $v_version ge v6.2.0;
99870f4d
KW
829
830# The lists below are hashes, so the key is the item in the list, and the
831# value is the reason why it is in the list. This makes generation of
832# documentation easier.
833
834my %why_suppressed; # No file generated for these.
835
836# Files aren't generated for empty extraneous properties. This is arguable.
837# Extraneous properties generally come about because a property is no longer
838# used in a newer version of Unicode. If we generated a file without code
839# points, programs that used to work on that property will still execute
840# without errors. It just won't ever match (or will always match, with \P{}).
841# This means that the logic is now likely wrong. I (khw) think its better to
842# find this out by getting an error message. Just move them to the table
843# above to change this behavior
844my %why_suppress_if_empty_warn_if_not = (
845
846 # It is the only property that has ever officially been removed from the
847 # Standard. The database never contained any code points for it.
848 'Special_Case_Condition' => 'Obsolete',
849
850 # Apparently never official, but there were code points in some versions of
851 # old-style PropList.txt
852 'Non_Break' => 'Obsolete',
853);
854
855# These would normally go in the warn table just above, but they were changed
856# a long time before this program was written, so warnings about them are
857# moot.
858if ($v_version gt v3.2.0) {
859 push @tables_that_may_be_empty,
860 'Canonical_Combining_Class=Attached_Below_Left'
861}
862
5f7264c7 863# These are listed in the Property aliases file in 6.0, but Unihan is ignored
99870f4d
KW
864# unless explicitly added.
865if ($v_version ge v5.2.0) {
866 my $unihan = 'Unihan; remove from list if using Unihan';
ea25a9b2 867 foreach my $table (qw (
99870f4d
KW
868 kAccountingNumeric
869 kOtherNumeric
870 kPrimaryNumeric
871 kCompatibilityVariant
872 kIICore
873 kIRG_GSource
874 kIRG_HSource
875 kIRG_JSource
876 kIRG_KPSource
877 kIRG_MSource
878 kIRG_KSource
879 kIRG_TSource
880 kIRG_USource
881 kIRG_VSource
882 kRSUnicode
ea25a9b2 883 ))
99870f4d
KW
884 {
885 $why_suppress_if_empty_warn_if_not{$table} = $unihan;
886 }
ca12659b
NC
887}
888
272501f6
KW
889# Enum values for to_output_map() method in the Map_Table package.
890my $EXTERNAL_MAP = 1;
891my $INTERNAL_MAP = 2;
ce712c88 892my $OUTPUT_ADJUSTED = 3;
272501f6 893
fcf1973c
KW
894# To override computed values for writing the map tables for these properties.
895# The default for enum map tables is to write them out, so that the Unicode
896# .txt files can be removed, but all the data to compute any property value
897# for any code point is available in a more compact form.
898my %global_to_output_map = (
899 # Needed by UCD.pm, but don't want to publicize that it exists, so won't
c12f2655
KW
900 # get stuck supporting it if things change. Since it is a STRING
901 # property, it normally would be listed in the pod, but INTERNAL_MAP
902 # suppresses that.
fcf1973c
KW
903 Unicode_1_Name => $INTERNAL_MAP,
904
905 Present_In => 0, # Suppress, as easily computed from Age
74cd47d0
KW
906 Block => (NON_ASCII_PLATFORM) ? 1 : 0, # Suppress, as Blocks.txt is
907 # retained, but needed for
908 # non-ASCII
53d34b6c
KW
909
910 # Suppress, as mapping can be found instead from the
911 # Perl_Decomposition_Mapping file
912 Decomposition_Type => 0,
fcf1973c
KW
913);
914
99870f4d 915# Properties that this program ignores.
230e0c16
KW
916my @unimplemented_properties;
917
918# With this release, it is automatically handled if the Unihan db is
919# downloaded
920push @unimplemented_properties, 'Unicode_Radical_Stroke' if $v_version le v5.2.0;
d73e5302 921
99870f4d
KW
922# There are several types of obsolete properties defined by Unicode. These
923# must be hand-edited for every new Unicode release.
924my %why_deprecated; # Generates a deprecated warning message if used.
925my %why_stabilized; # Documentation only
926my %why_obsolete; # Documentation only
927
928{ # Closure
8364f9eb 929 my $simple = 'Perl uses the more complete version';
99870f4d
KW
930 my $unihan = 'Unihan properties are by default not enabled in the Perl core. Instead use CPAN: Unicode::Unihan';
931
932 my $other_properties = 'other properties';
933 my $contributory = "Used by Unicode internally for generating $other_properties and not intended to be used stand-alone";
5d294d41 934 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
935
936 %why_deprecated = (
5f7264c7 937 'Grapheme_Link' => 'Deprecated by Unicode: Duplicates ccc=vr (Canonical_Combining_Class=Virama)',
99870f4d
KW
938 'Jamo_Short_Name' => $contributory,
939 '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',
940 'Other_Alphabetic' => $contributory,
941 'Other_Default_Ignorable_Code_Point' => $contributory,
942 'Other_Grapheme_Extend' => $contributory,
943 'Other_ID_Continue' => $contributory,
944 'Other_ID_Start' => $contributory,
945 'Other_Lowercase' => $contributory,
946 'Other_Math' => $contributory,
947 'Other_Uppercase' => $contributory,
e22aaf5c
KW
948 'Expands_On_NFC' => $why_no_expand,
949 'Expands_On_NFD' => $why_no_expand,
950 'Expands_On_NFKC' => $why_no_expand,
951 'Expands_On_NFKD' => $why_no_expand,
99870f4d
KW
952 );
953
954 %why_suppressed = (
5f7264c7 955 # There is a lib/unicore/Decomposition.pl (used by Normalize.pm) which
99870f4d
KW
956 # contains the same information, but without the algorithmically
957 # determinable Hangul syllables'. This file is not published, so it's
958 # existence is not noted in the comment.
e0b29447 959 'Decomposition_Mapping' => 'Accessible via Unicode::Normalize or Unicode::UCD::prop_invmap()',
99870f4d 960
3111abc0
KW
961 'Indic_Matra_Category' => "Provisional",
962 'Indic_Syllabic_Category' => "Provisional",
963
5f8d1a89
KW
964 # Don't suppress ISO_Comment, as otherwise special handling is needed
965 # to differentiate between it and gc=c, which can be written as 'isc',
966 # which is the same characters as ISO_Comment's short name.
99870f4d 967
fbb93542 968 'Name' => "Accessible via \\N{...} or 'use charnames;' or Unicode::UCD::prop_invmap()",
e0b29447
KW
969
970 'Simple_Case_Folding' => "$simple. Can access this through Unicode::UCD::casefold or Unicode::UCD::prop_invmap()",
971 'Simple_Lowercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
972 'Simple_Titlecase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
973 'Simple_Uppercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
99870f4d 974
5f7264c7 975 FC_NFKC_Closure => '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
6189eadc
KW
1202my $MAX_UNICODE_CODEPOINT_STRING = "10FFFF";
1203my $MAX_UNICODE_CODEPOINT = hex $MAX_UNICODE_CODEPOINT_STRING;
1204my $MAX_UNICODE_CODEPOINTS = $MAX_UNICODE_CODEPOINT + 1;
99870f4d 1205
2d88a86a
KW
1206# We work with above-Unicode code points, up to UV_MAX. But when you get
1207# that high, above IV_MAX, some operations don't work, and you can easily get
1208# overflow. Therefore for internal use, we use a much smaller number,
1209# translating it to UV_MAX only for output. The exact number is immaterial
1210# (all Unicode code points are treated exactly the same), but the algorithm
1211# requires it to be at least 2 * $MAX_UNICODE_CODEPOINTS + 1;
1212my $MAX_WORKING_CODEPOINTS= $MAX_UNICODE_CODEPOINT * 8;
1213my $MAX_WORKING_CODEPOINT = $MAX_WORKING_CODEPOINTS - 1;
1214my $MAX_WORKING_CODEPOINT_STRING = sprintf("%X", $MAX_WORKING_CODEPOINT);
1215
1216my $MAX_PLATFORM_CODEPOINT = ~0;
1217
99870f4d
KW
1218# Matches legal code point. 4-6 hex numbers, If there are 6, the first
1219# two must be 10; if there are 5, the first must not be a 0. Written this way
92199589
KW
1220# to decrease backtracking. The first regex allows the code point to be at
1221# the end of a word, but to work properly, the word shouldn't end with a valid
1222# hex character. The second one won't match a code point at the end of a
1223# word, and doesn't have the run-on issue
8c32d378
KW
1224my $run_on_code_point_re =
1225 qr/ (?: 10[0-9A-F]{4} | [1-9A-F][0-9A-F]{4} | [0-9A-F]{4} ) \b/x;
1226my $code_point_re = qr/\b$run_on_code_point_re/;
99870f4d
KW
1227
1228# This matches the beginning of the line in the Unicode db files that give the
1229# defaults for code points not listed (i.e., missing) in the file. The code
1230# depends on this ending with a semi-colon, so it can assume it is a valid
1231# field when the line is split() by semi-colons
1232my $missing_defaults_prefix =
6189eadc 1233 qr/^#\s+\@missing:\s+0000\.\.$MAX_UNICODE_CODEPOINT_STRING\s*;/;
99870f4d
KW
1234
1235# Property types. Unicode has more types, but these are sufficient for our
1236# purposes.
1237my $UNKNOWN = -1; # initialized to illegal value
1238my $NON_STRING = 1; # Either binary or enum
1239my $BINARY = 2;
06f26c45
KW
1240my $FORCED_BINARY = 3; # Not a binary property, but, besides its normal
1241 # tables, additional true and false tables are
1242 # generated so that false is anything matching the
1243 # default value, and true is everything else.
1244my $ENUM = 4; # Include catalog
1245my $STRING = 5; # Anything else: string or misc
99870f4d
KW
1246
1247# Some input files have lines that give default values for code points not
1248# contained in the file. Sometimes these should be ignored.
1249my $NO_DEFAULTS = 0; # Must evaluate to false
1250my $NOT_IGNORED = 1;
1251my $IGNORED = 2;
1252
1253# Range types. Each range has a type. Most ranges are type 0, for normal,
1254# and will appear in the main body of the tables in the output files, but
1255# there are other types of ranges as well, listed below, that are specially
1256# handled. There are pseudo-types as well that will never be stored as a
1257# type, but will affect the calculation of the type.
1258
1259# 0 is for normal, non-specials
1260my $MULTI_CP = 1; # Sequence of more than code point
1261my $HANGUL_SYLLABLE = 2;
1262my $CP_IN_NAME = 3; # The NAME contains the code point appended to it.
1263my $NULL = 4; # The map is to the null string; utf8.c can't
1264 # handle these, nor is there an accepted syntax
1265 # for them in \p{} constructs
f86864ac 1266my $COMPUTE_NO_MULTI_CP = 5; # Pseudo-type; means that ranges that would
99870f4d
KW
1267 # otherwise be $MULTI_CP type are instead type 0
1268
1269# process_generic_property_file() can accept certain overrides in its input.
1270# Each of these must begin AND end with $CMD_DELIM.
1271my $CMD_DELIM = "\a";
1272my $REPLACE_CMD = 'replace'; # Override the Replace
1273my $MAP_TYPE_CMD = 'map_type'; # Override the Type
1274
1275my $NO = 0;
1276my $YES = 1;
1277
1278# Values for the Replace argument to add_range.
1279# $NO # Don't replace; add only the code points not
1280 # already present.
1281my $IF_NOT_EQUIVALENT = 1; # Replace only under certain conditions; details in
1282 # the comments at the subroutine definition.
1283my $UNCONDITIONALLY = 2; # Replace without conditions.
9470941f 1284my $MULTIPLE_BEFORE = 4; # Don't replace, but add a duplicate record if
99870f4d 1285 # already there
7f4b1e25
KW
1286my $MULTIPLE_AFTER = 5; # Don't replace, but add a duplicate record if
1287 # already there
1288my $CROAK = 6; # Die with an error if is already there
99870f4d
KW
1289
1290# Flags to give property statuses. The phrases are to remind maintainers that
1291# if the flag is changed, the indefinite article referring to it in the
1292# documentation may need to be as well.
1293my $NORMAL = "";
99870f4d
KW
1294my $DEPRECATED = 'D';
1295my $a_bold_deprecated = "a 'B<$DEPRECATED>'";
1296my $A_bold_deprecated = "A 'B<$DEPRECATED>'";
1297my $DISCOURAGED = 'X';
1298my $a_bold_discouraged = "an 'B<$DISCOURAGED>'";
1299my $A_bold_discouraged = "An 'B<$DISCOURAGED>'";
1300my $STRICTER = 'T';
1301my $a_bold_stricter = "a 'B<$STRICTER>'";
1302my $A_bold_stricter = "A 'B<$STRICTER>'";
1303my $STABILIZED = 'S';
1304my $a_bold_stabilized = "an 'B<$STABILIZED>'";
1305my $A_bold_stabilized = "An 'B<$STABILIZED>'";
1306my $OBSOLETE = 'O';
1307my $a_bold_obsolete = "an 'B<$OBSOLETE>'";
1308my $A_bold_obsolete = "An 'B<$OBSOLETE>'";
1309
1310my %status_past_participles = (
1311 $DISCOURAGED => 'discouraged',
99870f4d
KW
1312 $STABILIZED => 'stabilized',
1313 $OBSOLETE => 'obsolete',
37e2e78e 1314 $DEPRECATED => 'deprecated',
99870f4d
KW
1315);
1316
395dfc19
KW
1317# Table fates. These are somewhat ordered, so that fates < $MAP_PROXIED should be
1318# externally documented.
301ba948 1319my $ORDINARY = 0; # The normal fate.
395dfc19
KW
1320my $MAP_PROXIED = 1; # The map table for the property isn't written out,
1321 # but there is a file written that can be used to
1322 # reconstruct this table
3cdaf629 1323my $INTERNAL_ONLY = 2; # The file for this table is written out, but it is
301ba948 1324 # for Perl's internal use only
277b7b16
KW
1325my $LEGACY_ONLY = 3; # Like $INTERNAL_ONLY, but not actually used by Perl.
1326 # Is for backwards compatibility for applications that
1327 # read the file directly, so it's format is
1328 # unchangeable.
1329my $SUPPRESSED = 4; # The file for this table is not written out, and as a
3cdaf629
KW
1330 # result, we don't bother to do many computations on
1331 # it.
277b7b16 1332my $PLACEHOLDER = 5; # Like $SUPPRESSED, but we go through all the
3cdaf629
KW
1333 # computations anyway, as the values are needed for
1334 # things to work. This happens when we have Perl
1335 # extensions that depend on Unicode tables that
1336 # wouldn't normally be in a given Unicode version.
301ba948 1337
f5817e0a
KW
1338# The format of the values of the tables:
1339my $EMPTY_FORMAT = "";
99870f4d
KW
1340my $BINARY_FORMAT = 'b';
1341my $DECIMAL_FORMAT = 'd';
1342my $FLOAT_FORMAT = 'f';
1343my $INTEGER_FORMAT = 'i';
1344my $HEX_FORMAT = 'x';
1345my $RATIONAL_FORMAT = 'r';
1346my $STRING_FORMAT = 's';
d11155ec 1347my $ADJUST_FORMAT = 'a';
24303724 1348my $HEX_ADJUST_FORMAT = 'ax';
a14f3cb1 1349my $DECOMP_STRING_FORMAT = 'c';
c3ff2976 1350my $STRING_WHITE_SPACE_LIST = 'sw';
99870f4d
KW
1351
1352my %map_table_formats = (
1353 $BINARY_FORMAT => 'binary',
1354 $DECIMAL_FORMAT => 'single decimal digit',
1355 $FLOAT_FORMAT => 'floating point number',
1356 $INTEGER_FORMAT => 'integer',
add63c13 1357 $HEX_FORMAT => 'non-negative hex whole number; a code point',
99870f4d 1358 $RATIONAL_FORMAT => 'rational: an integer or a fraction',
1a9d544b 1359 $STRING_FORMAT => 'string',
d11155ec 1360 $ADJUST_FORMAT => 'some entries need adjustment',
24303724 1361 $HEX_ADJUST_FORMAT => 'mapped value in hex; some entries need adjustment',
92f9d56c 1362 $DECOMP_STRING_FORMAT => 'Perl\'s internal (Normalize.pm) decomposition mapping',
c3ff2976 1363 $STRING_WHITE_SPACE_LIST => 'string, but some elements are interpreted as a list; white space occurs only as list item separators'
99870f4d
KW
1364);
1365
1366# Unicode didn't put such derived files in a separate directory at first.
1367my $EXTRACTED_DIR = (-d 'extracted') ? 'extracted' : "";
1368my $EXTRACTED = ($EXTRACTED_DIR) ? "$EXTRACTED_DIR/" : "";
1369my $AUXILIARY = 'auxiliary';
1370
3854b4b8
KW
1371# Hashes and arrays that will eventually go into Heavy.pl for the use of
1372# utf8_heavy.pl and into UCD.pl for the use of UCD.pm
99870f4d
KW
1373my %loose_to_file_of; # loosely maps table names to their respective
1374 # files
1375my %stricter_to_file_of; # same; but for stricter mapping.
315bfd4e 1376my %loose_property_to_file_of; # Maps a loose property name to its map file
3854b4b8
KW
1377my @inline_definitions = "V0"; # Each element gives a definition of a unique
1378 # inversion list. When a definition is inlined,
1379 # its value in the hash it's in (one of the two
1380 # defined just above) will include an index into
1381 # this array. The 0th element is initialized to
1382 # the definition for a zero length invwersion list
89cf10cc
KW
1383my %file_to_swash_name; # Maps the file name to its corresponding key name
1384 # in the hash %utf8::SwashInfo
99870f4d
KW
1385my %nv_floating_to_rational; # maps numeric values floating point numbers to
1386 # their rational equivalent
c12f2655
KW
1387my %loose_property_name_of; # Loosely maps (non_string) property names to
1388 # standard form
86a52d1e 1389my %string_property_loose_to_name; # Same, for string properties.
c15fda25
KW
1390my %loose_defaults; # keys are of form "prop=value", where 'prop' is
1391 # the property name in standard loose form, and
1392 # 'value' is the default value for that property,
1393 # also in standard loose form.
9e4a1e86
KW
1394my %loose_to_standard_value; # loosely maps table names to the canonical
1395 # alias for them
2df7880f
KW
1396my %ambiguous_names; # keys are alias names (in standard form) that
1397 # have more than one possible meaning.
5d1df013
KW
1398my %prop_aliases; # Keys are standard property name; values are each
1399 # one's aliases
1e863613
KW
1400my %prop_value_aliases; # Keys of top level are standard property name;
1401 # values are keys to another hash, Each one is
1402 # one of the property's values, in standard form.
1403 # The values are that prop-val's aliases.
2df7880f 1404my %ucd_pod; # Holds entries that will go into the UCD section of the pod
99870f4d 1405
d867ccfb
KW
1406# Most properties are immune to caseless matching, otherwise you would get
1407# nonsensical results, as properties are a function of a code point, not
1408# everything that is caselessly equivalent to that code point. For example,
1409# Changes_When_Case_Folded('s') should be false, whereas caselessly it would
1410# be true because 's' and 'S' are equivalent caselessly. However,
1411# traditionally, [:upper:] and [:lower:] are equivalent caselessly, so we
1412# extend that concept to those very few properties that are like this. Each
1413# such property will match the full range caselessly. They are hard-coded in
1414# the program; it's not worth trying to make it general as it's extremely
1415# unlikely that they will ever change.
1416my %caseless_equivalent_to;
1417
99870f4d
KW
1418# These constants names and values were taken from the Unicode standard,
1419# version 5.1, section 3.12. They are used in conjunction with Hangul
6e5a209b
KW
1420# syllables. The '_string' versions are so generated tables can retain the
1421# hex format, which is the more familiar value
1422my $SBase_string = "0xAC00";
1423my $SBase = CORE::hex $SBase_string;
1424my $LBase_string = "0x1100";
1425my $LBase = CORE::hex $LBase_string;
1426my $VBase_string = "0x1161";
1427my $VBase = CORE::hex $VBase_string;
1428my $TBase_string = "0x11A7";
1429my $TBase = CORE::hex $TBase_string;
99870f4d
KW
1430my $SCount = 11172;
1431my $LCount = 19;
1432my $VCount = 21;
1433my $TCount = 28;
1434my $NCount = $VCount * $TCount;
1435
1436# For Hangul syllables; These store the numbers from Jamo.txt in conjunction
1437# with the above published constants.
1438my %Jamo;
1439my %Jamo_L; # Leading consonants
1440my %Jamo_V; # Vowels
1441my %Jamo_T; # Trailing consonants
1442
bb1dd3da
KW
1443# For code points whose name contains its ordinal as a '-ABCD' suffix.
1444# The key is the base name of the code point, and the value is an
1445# array giving all the ranges that use this base name. Each range
1446# is actually a hash giving the 'low' and 'high' values of it.
1447my %names_ending_in_code_point;
1448my %loose_names_ending_in_code_point; # Same as above, but has blanks, dashes
1449 # removed from the names
1450# Inverse mapping. The list of ranges that have these kinds of
1451# names. Each element contains the low, high, and base names in an
1452# anonymous hash.
1453my @code_points_ending_in_code_point;
1454
6b5ab373
KW
1455# To hold Unicode's normalization test suite
1456my @normalization_tests;
1457
bb1dd3da
KW
1458# Boolean: does this Unicode version have the hangul syllables, and are we
1459# writing out a table for them?
1460my $has_hangul_syllables = 0;
1461
1462# Does this Unicode version have code points whose names end in their
1463# respective code points, and are we writing out a table for them? 0 for no;
1464# otherwise points to first property that a table is needed for them, so that
1465# if multiple tables are needed, we don't create duplicates
1466my $needing_code_points_ending_in_code_point = 0;
1467
37e2e78e 1468my @backslash_X_tests; # List of tests read in for testing \X
99870f4d
KW
1469my @unhandled_properties; # Will contain a list of properties found in
1470 # the input that we didn't process.
f86864ac 1471my @match_properties; # Properties that have match tables, to be
99870f4d
KW
1472 # listed in the pod
1473my @map_properties; # Properties that get map files written
1474my @named_sequences; # NamedSequences.txt contents.
1475my %potential_files; # Generated list of all .txt files in the directory
1476 # structure so we can warn if something is being
1477 # ignored.
1478my @files_actually_output; # List of files we generated.
1479my @more_Names; # Some code point names are compound; this is used
1480 # to store the extra components of them.
1481my $MIN_FRACTION_LENGTH = 3; # How many digits of a floating point number at
1482 # the minimum before we consider it equivalent to a
1483 # candidate rational
1484my $MAX_FLOATING_SLOP = 10 ** - $MIN_FRACTION_LENGTH; # And in floating terms
1485
1486# These store references to certain commonly used property objects
3c88a801 1487my $ccc;
99870f4d
KW
1488my $gc;
1489my $perl;
1490my $block;
3e20195b
KW
1491my $perl_charname;
1492my $print;
ac7dbdb6 1493my $All;
359523e2 1494my $script;
99870f4d
KW
1495
1496# Are there conflicting names because of beginning with 'In_', or 'Is_'
1497my $has_In_conflicts = 0;
1498my $has_Is_conflicts = 0;
1499
1500sub internal_file_to_platform ($) {
1501 # Convert our file paths which have '/' separators to those of the
1502 # platform.
1503
1504 my $file = shift;
1505 return undef unless defined $file;
1506
1507 return File::Spec->join(split '/', $file);
d07a55ed 1508}
5beb625e 1509
99870f4d
KW
1510sub file_exists ($) { # platform independent '-e'. This program internally
1511 # uses slash as a path separator.
1512 my $file = shift;
1513 return 0 if ! defined $file;
1514 return -e internal_file_to_platform($file);
1515}
5beb625e 1516
99870f4d 1517sub objaddr($) {
23e33b60
KW
1518 # Returns the address of the blessed input object.
1519 # It doesn't check for blessedness because that would do a string eval
1520 # every call, and the program is structured so that this is never called
1521 # for a non-blessed object.
99870f4d 1522
23e33b60 1523 no overloading; # If overloaded, numifying below won't work.
99870f4d
KW
1524
1525 # Numifying a ref gives its address.
051df77b 1526 return pack 'J', $_[0];
99870f4d
KW
1527}
1528
558712cf 1529# These are used only if $annotate is true.
c4019d52
KW
1530# The entire range of Unicode characters is examined to populate these
1531# after all the input has been processed. But most can be skipped, as they
1532# have the same descriptive phrases, such as being unassigned
1533my @viacode; # Contains the 1 million character names
1534my @printable; # boolean: And are those characters printable?
1535my @annotate_char_type; # Contains a type of those characters, specifically
1536 # for the purposes of annotation.
1537my $annotate_ranges; # A map of ranges of code points that have the same
98dc9551 1538 # name for the purposes of annotation. They map to the
c4019d52
KW
1539 # upper edge of the range, so that the end point can
1540 # be immediately found. This is used to skip ahead to
1541 # the end of a range, and avoid processing each
1542 # individual code point in it.
1543my $unassigned_sans_noncharacters; # A Range_List of the unassigned
1544 # characters, but excluding those which are
1545 # also noncharacter code points
1546
1547# The annotation types are an extension of the regular range types, though
1548# some of the latter are folded into one. Make the new types negative to
1549# avoid conflicting with the regular types
1550my $SURROGATE_TYPE = -1;
1551my $UNASSIGNED_TYPE = -2;
1552my $PRIVATE_USE_TYPE = -3;
1553my $NONCHARACTER_TYPE = -4;
1554my $CONTROL_TYPE = -5;
2d88a86a
KW
1555my $ABOVE_UNICODE_TYPE = -6;
1556my $UNKNOWN_TYPE = -7; # Used only if there is a bug in this program
c4019d52
KW
1557
1558sub populate_char_info ($) {
558712cf 1559 # Used only with the $annotate option. Populates the arrays with the
c4019d52
KW
1560 # input code point's info that are needed for outputting more detailed
1561 # comments. If calling context wants a return, it is the end point of
1562 # any contiguous range of characters that share essentially the same info
1563
1564 my $i = shift;
1565 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1566
1567 $viacode[$i] = $perl_charname->value_of($i) || "";
1568
1569 # A character is generally printable if Unicode says it is,
1570 # but below we make sure that most Unicode general category 'C' types
1571 # aren't.
1572 $printable[$i] = $print->contains($i);
1573
1574 $annotate_char_type[$i] = $perl_charname->type_of($i) || 0;
1575
1576 # Only these two regular types are treated specially for annotations
1577 # purposes
1578 $annotate_char_type[$i] = 0 if $annotate_char_type[$i] != $CP_IN_NAME
1579 && $annotate_char_type[$i] != $HANGUL_SYLLABLE;
1580
1581 # Give a generic name to all code points that don't have a real name.
1582 # We output ranges, if applicable, for these. Also calculate the end
1583 # point of the range.
1584 my $end;
1585 if (! $viacode[$i]) {
1d025d66 1586 my $nonchar;
2d88a86a
KW
1587 if ($i > $MAX_UNICODE_CODEPOINT) {
1588 $viacode[$i] = 'Above-Unicode';
1589 $annotate_char_type[$i] = $ABOVE_UNICODE_TYPE;
1590 $printable[$i] = 0;
1591 $end = $MAX_WORKING_CODEPOINT;
1592 }
1593 elsif ($gc-> table('Private_use')->contains($i)) {
c4019d52
KW
1594 $viacode[$i] = 'Private Use';
1595 $annotate_char_type[$i] = $PRIVATE_USE_TYPE;
1596 $printable[$i] = 0;
1597 $end = $gc->table('Private_Use')->containing_range($i)->end;
1598 }
1d025d66
KW
1599 elsif ((defined ($nonchar =
1600 Property::property_ref('Noncharacter_Code_Point'))
1601 && $nonchar->table('Y')->contains($i)))
c4019d52
KW
1602 {
1603 $viacode[$i] = 'Noncharacter';
1604 $annotate_char_type[$i] = $NONCHARACTER_TYPE;
1605 $printable[$i] = 0;
1606 $end = property_ref('Noncharacter_Code_Point')->table('Y')->
1607 containing_range($i)->end;
1608 }
1609 elsif ($gc-> table('Control')->contains($i)) {
c71dea7f 1610 $viacode[$i] = property_ref('Name_Alias')->value_of($i) || 'Control';
c4019d52
KW
1611 $annotate_char_type[$i] = $CONTROL_TYPE;
1612 $printable[$i] = 0;
c4019d52
KW
1613 }
1614 elsif ($gc-> table('Unassigned')->contains($i)) {
c4019d52
KW
1615 $annotate_char_type[$i] = $UNASSIGNED_TYPE;
1616 $printable[$i] = 0;
1d025d66
KW
1617 if ($v_version lt v2.0.0) { # No blocks in earliest releases
1618 $viacode[$i] = 'Unassigned';
1619 $end = $gc-> table('Unassigned')->containing_range($i)->end;
1620 }
1621 else {
1622 $viacode[$i] = 'Unassigned, block=' . $block-> value_of($i);
c4019d52 1623
bf06c733
KW
1624 # Because we name the unassigned by the blocks they are in, it
1625 # can't go past the end of that block, and it also can't go
1626 # past the unassigned range it is in. The special table makes
1627 # sure that the non-characters, which are unassigned, are
1628 # separated out.
1629 $end = min($block->containing_range($i)->end,
1630 $unassigned_sans_noncharacters->
1631 containing_range($i)->end);
1d025d66
KW
1632 }
1633 }
1634 elsif ($v_version lt v2.0.0) { # No surrogates in earliest releases
1635 $viacode[$i] = $gc->value_of($i);
1636 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1637 $printable[$i] = 0;
1638 }
1639 elsif ($gc-> table('Surrogate')->contains($i)) {
1640 $viacode[$i] = 'Surrogate';
1641 $annotate_char_type[$i] = $SURROGATE_TYPE;
1642 $printable[$i] = 0;
1643 $end = $gc->table('Surrogate')->containing_range($i)->end;
13ca76ff
KW
1644 }
1645 else {
1646 Carp::my_carp_bug("Can't figure out how to annotate "
1647 . sprintf("U+%04X", $i)
1648 . ". Proceeding anyway.");
c4019d52
KW
1649 $viacode[$i] = 'UNKNOWN';
1650 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1651 $printable[$i] = 0;
1652 }
1653 }
1654
1655 # Here, has a name, but if it's one in which the code point number is
1656 # appended to the name, do that.
1657 elsif ($annotate_char_type[$i] == $CP_IN_NAME) {
1658 $viacode[$i] .= sprintf("-%04X", $i);
1659 $end = $perl_charname->containing_range($i)->end;
1660 }
1661
1662 # And here, has a name, but if it's a hangul syllable one, replace it with
1663 # the correct name from the Unicode algorithm
1664 elsif ($annotate_char_type[$i] == $HANGUL_SYLLABLE) {
1665 use integer;
1666 my $SIndex = $i - $SBase;
1667 my $L = $LBase + $SIndex / $NCount;
1668 my $V = $VBase + ($SIndex % $NCount) / $TCount;
1669 my $T = $TBase + $SIndex % $TCount;
1670 $viacode[$i] = "HANGUL SYLLABLE $Jamo{$L}$Jamo{$V}";
1671 $viacode[$i] .= $Jamo{$T} if $T != $TBase;
1672 $end = $perl_charname->containing_range($i)->end;
1673 }
1674
1675 return if ! defined wantarray;
1676 return $i if ! defined $end; # If not a range, return the input
1677
1678 # Save this whole range so can find the end point quickly
1679 $annotate_ranges->add_map($i, $end, $end);
1680
1681 return $end;
1682}
1683
23e33b60
KW
1684# Commented code below should work on Perl 5.8.
1685## This 'require' doesn't necessarily work in miniperl, and even if it does,
1686## the native perl version of it (which is what would operate under miniperl)
1687## is extremely slow, as it does a string eval every call.
7e017d6d 1688#my $has_fast_scalar_util = $^X !~ /miniperl/
23e33b60
KW
1689# && defined eval "require Scalar::Util";
1690#
1691#sub objaddr($) {
1692# # Returns the address of the blessed input object. Uses the XS version if
1693# # available. It doesn't check for blessedness because that would do a
1694# # string eval every call, and the program is structured so that this is
1695# # never called for a non-blessed object.
1696#
1697# return Scalar::Util::refaddr($_[0]) if $has_fast_scalar_util;
1698#
1699# # Check at least that is a ref.
1700# my $pkg = ref($_[0]) or return undef;
1701#
1702# # Change to a fake package to defeat any overloaded stringify
1703# bless $_[0], 'main::Fake';
1704#
1705# # Numifying a ref gives its address.
051df77b 1706# my $addr = pack 'J', $_[0];
23e33b60
KW
1707#
1708# # Return to original class
1709# bless $_[0], $pkg;
1710# return $addr;
1711#}
1712
99870f4d
KW
1713sub max ($$) {
1714 my $a = shift;
1715 my $b = shift;
1716 return $a if $a >= $b;
1717 return $b;
1718}
1719
1720sub min ($$) {
1721 my $a = shift;
1722 my $b = shift;
1723 return $a if $a <= $b;
1724 return $b;
1725}
1726
1727sub clarify_number ($) {
1728 # This returns the input number with underscores inserted every 3 digits
1729 # in large (5 digits or more) numbers. Input must be entirely digits, not
1730 # checked.
1731
1732 my $number = shift;
1733 my $pos = length($number) - 3;
1734 return $number if $pos <= 1;
1735 while ($pos > 0) {
1736 substr($number, $pos, 0) = '_';
1737 $pos -= 3;
5beb625e 1738 }
99870f4d 1739 return $number;
99598c8c
JH
1740}
1741
731cb813
KW
1742sub clarify_code_point_count ($) {
1743 # This is like clarify_number(), but the input is assumed to be a count of
1744 # code points, rather than a generic number.
1745
2d88a86a
KW
1746 my $append = "";
1747
1748 my $number = shift;
1749 if ($number > $MAX_UNICODE_CODEPOINTS) {
1750 $number -= ($MAX_WORKING_CODEPOINTS - $MAX_UNICODE_CODEPOINTS);
1751 return "All above-Unicode code points" if $number == 0;
1752 $append = " + all above-Unicode code points";
1753 }
1754 return clarify_number($number) . $append;
731cb813 1755}
12ac2576 1756
99870f4d 1757package Carp;
7ebf06b3 1758
99870f4d
KW
1759# These routines give a uniform treatment of messages in this program. They
1760# are placed in the Carp package to cause the stack trace to not include them,
1761# although an alternative would be to use another package and set @CARP_NOT
1762# for it.
12ac2576 1763
99870f4d 1764our $Verbose = 1 if main::DEBUG; # Useful info when debugging
12ac2576 1765
99f78760
KW
1766# This is a work-around suggested by Nicholas Clark to fix a problem with Carp
1767# and overload trying to load Scalar:Util under miniperl. See
1768# http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2009-11/msg01057.html
1769undef $overload::VERSION;
1770
99870f4d
KW
1771sub my_carp {
1772 my $message = shift || "";
1773 my $nofold = shift || 0;
7ebf06b3 1774
99870f4d
KW
1775 if ($message) {
1776 $message = main::join_lines($message);
1777 $message =~ s/^$0: *//; # Remove initial program name
1778 $message =~ s/[.;,]+$//; # Remove certain ending punctuation
1779 $message = "\n$0: $message;";
12ac2576 1780
99870f4d
KW
1781 # Fold the message with program name, semi-colon end punctuation
1782 # (which looks good with the message that carp appends to it), and a
1783 # hanging indent for continuation lines.
1784 $message = main::simple_fold($message, "", 4) unless $nofold;
1785 $message =~ s/\n$//; # Remove the trailing nl so what carp
1786 # appends is to the same line
1787 }
12ac2576 1788
99870f4d 1789 return $message if defined wantarray; # If a caller just wants the msg
12ac2576 1790
99870f4d
KW
1791 carp $message;
1792 return;
1793}
7ebf06b3 1794
99870f4d
KW
1795sub my_carp_bug {
1796 # This is called when it is clear that the problem is caused by a bug in
1797 # this program.
7ebf06b3 1798
99870f4d
KW
1799 my $message = shift;
1800 $message =~ s/^$0: *//;
1801 $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");
1802 carp $message;
1803 return;
1804}
7ebf06b3 1805
99870f4d
KW
1806sub carp_too_few_args {
1807 if (@_ != 2) {
1808 my_carp_bug("Wrong number of arguments: to 'carp_too_few_arguments'. No action taken.");
1809 return;
12ac2576 1810 }
7ebf06b3 1811
99870f4d
KW
1812 my $args_ref = shift;
1813 my $count = shift;
7ebf06b3 1814
99870f4d
KW
1815 my_carp_bug("Need at least $count arguments to "
1816 . (caller 1)[3]
1817 . ". Instead got: '"
1818 . join ', ', @$args_ref
1819 . "'. No action taken.");
1820 return;
12ac2576
JP
1821}
1822
99870f4d
KW
1823sub carp_extra_args {
1824 my $args_ref = shift;
1825 my_carp_bug("Too many arguments to 'carp_extra_args': (" . join(', ', @_) . "); Extras ignored.") if @_;
12ac2576 1826
99870f4d
KW
1827 unless (ref $args_ref) {
1828 my_carp_bug("Argument to 'carp_extra_args' ($args_ref) must be a ref. Not checking arguments.");
1829 return;
1830 }
1831 my ($package, $file, $line) = caller;
1832 my $subroutine = (caller 1)[3];
cf25bb62 1833
99870f4d
KW
1834 my $list;
1835 if (ref $args_ref eq 'HASH') {
1836 foreach my $key (keys %$args_ref) {
1837 $args_ref->{$key} = $UNDEF unless defined $args_ref->{$key};
cf25bb62 1838 }
99870f4d 1839 $list = join ', ', each %{$args_ref};
cf25bb62 1840 }
99870f4d
KW
1841 elsif (ref $args_ref eq 'ARRAY') {
1842 foreach my $arg (@$args_ref) {
1843 $arg = $UNDEF unless defined $arg;
1844 }
1845 $list = join ', ', @$args_ref;
1846 }
1847 else {
1848 my_carp_bug("Can't cope with ref "
1849 . ref($args_ref)
1850 . " . argument to 'carp_extra_args'. Not checking arguments.");
1851 return;
1852 }
1853
1854 my_carp_bug("Unrecognized parameters in options: '$list' to $subroutine. Skipped.");
1855 return;
d73e5302
JH
1856}
1857
99870f4d
KW
1858package main;
1859
1860{ # Closure
1861
1862 # This program uses the inside-out method for objects, as recommended in
1b6b3fa9
KW
1863 # "Perl Best Practices". (This is the best solution still, since this has
1864 # to run under miniperl.) This closure aids in generating those. There
99870f4d
KW
1865 # are two routines. setup_package() is called once per package to set
1866 # things up, and then set_access() is called for each hash representing a
1867 # field in the object. These routines arrange for the object to be
1868 # properly destroyed when no longer used, and for standard accessor
1869 # functions to be generated. If you need more complex accessors, just
1870 # write your own and leave those accesses out of the call to set_access().
1871 # More details below.
1872
1873 my %constructor_fields; # fields that are to be used in constructors; see
1874 # below
1875
1876 # The values of this hash will be the package names as keys to other
1877 # hashes containing the name of each field in the package as keys, and
1878 # references to their respective hashes as values.
1879 my %package_fields;
1880
1881 sub setup_package {
1882 # Sets up the package, creating standard DESTROY and dump methods
1883 # (unless already defined). The dump method is used in debugging by
1884 # simple_dumper().
1885 # The optional parameters are:
1886 # a) a reference to a hash, that gets populated by later
1887 # set_access() calls with one of the accesses being
1888 # 'constructor'. The caller can then refer to this, but it is
1889 # not otherwise used by these two routines.
1890 # b) a reference to a callback routine to call during destruction
1891 # of the object, before any fields are actually destroyed
1892
1893 my %args = @_;
1894 my $constructor_ref = delete $args{'Constructor_Fields'};
1895 my $destroy_callback = delete $args{'Destroy_Callback'};
1896 Carp::carp_extra_args(\@_) if main::DEBUG && %args;
1897
1898 my %fields;
1899 my $package = (caller)[0];
1900
1901 $package_fields{$package} = \%fields;
1902 $constructor_fields{$package} = $constructor_ref;
1903
1904 unless ($package->can('DESTROY')) {
1905 my $destroy_name = "${package}::DESTROY";
1906 no strict "refs";
1907
1908 # Use typeglob to give the anonymous subroutine the name we want
1909 *$destroy_name = sub {
1910 my $self = shift;
ffe43484 1911 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1912
1913 $self->$destroy_callback if $destroy_callback;
1914 foreach my $field (keys %{$package_fields{$package}}) {
1915 #print STDERR __LINE__, ": Destroying ", ref $self, " ", sprintf("%04X", $addr), ": ", $field, "\n";
1916 delete $package_fields{$package}{$field}{$addr};
1917 }
1918 return;
1919 }
1920 }
1921
1922 unless ($package->can('dump')) {
1923 my $dump_name = "${package}::dump";
1924 no strict "refs";
1925 *$dump_name = sub {
1926 my $self = shift;
1927 return dump_inside_out($self, $package_fields{$package}, @_);
1928 }
1929 }
1930 return;
1931 }
1932
1933 sub set_access {
1934 # Arrange for the input field to be garbage collected when no longer
1935 # needed. Also, creates standard accessor functions for the field
1936 # based on the optional parameters-- none if none of these parameters:
1937 # 'addable' creates an 'add_NAME()' accessor function.
1938 # 'readable' or 'readable_array' creates a 'NAME()' accessor
1939 # function.
1940 # 'settable' creates a 'set_NAME()' accessor function.
1941 # 'constructor' doesn't create an accessor function, but adds the
1942 # field to the hash that was previously passed to
1943 # setup_package();
1944 # Any of the accesses can be abbreviated down, so that 'a', 'ad',
1945 # 'add' etc. all mean 'addable'.
1946 # The read accessor function will work on both array and scalar
1947 # values. If another accessor in the parameter list is 'a', the read
1948 # access assumes an array. You can also force it to be array access
1949 # by specifying 'readable_array' instead of 'readable'
1950 #
1951 # A sort-of 'protected' access can be set-up by preceding the addable,
1952 # readable or settable with some initial portion of 'protected_' (but,
1953 # the underscore is required), like 'p_a', 'pro_set', etc. The
1954 # "protection" is only by convention. All that happens is that the
1955 # accessor functions' names begin with an underscore. So instead of
1956 # calling set_foo, the call is _set_foo. (Real protection could be
c1739a4a 1957 # accomplished by having a new subroutine, end_package, called at the
99870f4d
KW
1958 # end of each package, and then storing the __LINE__ ranges and
1959 # checking them on every accessor. But that is way overkill.)
1960
1961 # We create anonymous subroutines as the accessors and then use
1962 # typeglobs to assign them to the proper package and name
1963
1964 my $name = shift; # Name of the field
1965 my $field = shift; # Reference to the inside-out hash containing the
1966 # field
1967
1968 my $package = (caller)[0];
1969
1970 if (! exists $package_fields{$package}) {
1971 croak "$0: Must call 'setup_package' before 'set_access'";
1972 }
d73e5302 1973
99870f4d
KW
1974 # Stash the field so DESTROY can get it.
1975 $package_fields{$package}{$name} = $field;
cf25bb62 1976
99870f4d
KW
1977 # Remaining arguments are the accessors. For each...
1978 foreach my $access (@_) {
1979 my $access = lc $access;
cf25bb62 1980
99870f4d 1981 my $protected = "";
cf25bb62 1982
99870f4d
KW
1983 # Match the input as far as it goes.
1984 if ($access =~ /^(p[^_]*)_/) {
1985 $protected = $1;
1986 if (substr('protected_', 0, length $protected)
1987 eq $protected)
1988 {
1989
1990 # Add 1 for the underscore not included in $protected
1991 $access = substr($access, length($protected) + 1);
1992 $protected = '_';
1993 }
1994 else {
1995 $protected = "";
1996 }
1997 }
1998
1999 if (substr('addable', 0, length $access) eq $access) {
2000 my $subname = "${package}::${protected}add_$name";
2001 no strict "refs";
2002
2003 # add_ accessor. Don't add if already there, which we
2004 # determine using 'eq' for scalars and '==' otherwise.
2005 *$subname = sub {
2006 use strict "refs";
2007 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
2008 my $self = shift;
2009 my $value = shift;
ffe43484 2010 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2011 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2012 if (ref $value) {
f998e60c 2013 return if grep { $value == $_ } @{$field->{$addr}};
99870f4d
KW
2014 }
2015 else {
f998e60c 2016 return if grep { $value eq $_ } @{$field->{$addr}};
99870f4d 2017 }
f998e60c 2018 push @{$field->{$addr}}, $value;
99870f4d
KW
2019 return;
2020 }
2021 }
2022 elsif (substr('constructor', 0, length $access) eq $access) {
2023 if ($protected) {
2024 Carp::my_carp_bug("Can't set-up 'protected' constructors")
2025 }
2026 else {
2027 $constructor_fields{$package}{$name} = $field;
2028 }
2029 }
2030 elsif (substr('readable_array', 0, length $access) eq $access) {
2031
2032 # Here has read access. If one of the other parameters for
2033 # access is array, or this one specifies array (by being more
2034 # than just 'readable_'), then create a subroutine that
2035 # assumes the data is an array. Otherwise just a scalar
2036 my $subname = "${package}::${protected}$name";
2037 if (grep { /^a/i } @_
2038 or length($access) > length('readable_'))
2039 {
2040 no strict "refs";
2041 *$subname = sub {
2042 use strict "refs";
23e33b60 2043 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
ffe43484 2044 my $addr = do { no overloading; pack 'J', $_[0]; };
99870f4d
KW
2045 if (ref $field->{$addr} ne 'ARRAY') {
2046 my $type = ref $field->{$addr};
2047 $type = 'scalar' unless $type;
2048 Carp::my_carp_bug("Trying to read $name as an array when it is a $type. Big problems.");
2049 return;
2050 }
2051 return scalar @{$field->{$addr}} unless wantarray;
2052
2053 # Make a copy; had problems with caller modifying the
2054 # original otherwise
2055 my @return = @{$field->{$addr}};
2056 return @return;
2057 }
2058 }
2059 else {
2060
2061 # Here not an array value, a simpler function.
2062 no strict "refs";
2063 *$subname = sub {
2064 use strict "refs";
23e33b60 2065 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
f998e60c 2066 no overloading;
051df77b 2067 return $field->{pack 'J', $_[0]};
99870f4d
KW
2068 }
2069 }
2070 }
2071 elsif (substr('settable', 0, length $access) eq $access) {
2072 my $subname = "${package}::${protected}set_$name";
2073 no strict "refs";
2074 *$subname = sub {
2075 use strict "refs";
23e33b60
KW
2076 if (main::DEBUG) {
2077 return Carp::carp_too_few_args(\@_, 2) if @_ < 2;
2078 Carp::carp_extra_args(\@_) if @_ > 2;
2079 }
2080 # $self is $_[0]; $value is $_[1]
f998e60c 2081 no overloading;
051df77b 2082 $field->{pack 'J', $_[0]} = $_[1];
99870f4d
KW
2083 return;
2084 }
2085 }
2086 else {
2087 Carp::my_carp_bug("Unknown accessor type $access. No accessor set.");
2088 }
cf25bb62 2089 }
99870f4d 2090 return;
cf25bb62 2091 }
99870f4d
KW
2092}
2093
2094package Input_file;
2095
2096# All input files use this object, which stores various attributes about them,
2097# and provides for convenient, uniform handling. The run method wraps the
2098# processing. It handles all the bookkeeping of opening, reading, and closing
2099# the file, returning only significant input lines.
2100#
2101# Each object gets a handler which processes the body of the file, and is
74cd47d0
KW
2102# called by run(). All character property files must use the generic,
2103# default handler, which has code scrubbed to handle things you might not
2104# expect, including automatic EBCDIC handling. For files that don't deal with
2105# mapping code points to a property value, such as test files,
2106# PropertyAliases, PropValueAliases, and named sequences, you can override the
2107# handler to be a custom one. Such a handler should basically be a
2108# while(next_line()) {...} loop.
99870f4d
KW
2109#
2110# You can also set up handlers to
537124e4 2111# 1) call before the first line is read, for pre processing
83b68635
KW
2112# 2) call to adjust each line of the input before the main handler gets
2113# them. This can be automatically generated, if appropriately simple
2114# enough, by specifiying a Properties parameter in the constructor.
99870f4d 2115# 3) call upon EOF before the main handler exits its loop
537124e4 2116# 4) call at the end, for post processing
99870f4d
KW
2117#
2118# $_ is used to store the input line, and is to be filtered by the
2119# each_line_handler()s. So, if the format of the line is not in the desired
2120# format for the main handler, these are used to do that adjusting. They can
2121# be stacked (by enclosing them in an [ anonymous array ] in the constructor,
2122# so the $_ output of one is used as the input to the next. None of the other
2123# handlers are stackable, but could easily be changed to be so.
2124#
2125# Most of the handlers can call insert_lines() or insert_adjusted_lines()
2126# which insert the parameters as lines to be processed before the next input
2127# file line is read. This allows the EOF handler to flush buffers, for
2128# example. The difference between the two routines is that the lines inserted
2129# by insert_lines() are subjected to the each_line_handler()s. (So if you
2130# called it from such a handler, you would get infinite recursion.) Lines
2131# inserted by insert_adjusted_lines() go directly to the main handler without
2132# any adjustments. If the post-processing handler calls any of these, there
2133# will be no effect. Some error checking for these conditions could be added,
2134# but it hasn't been done.
2135#
2136# carp_bad_line() should be called to warn of bad input lines, which clears $_
2137# to prevent further processing of the line. This routine will output the
2138# message as a warning once, and then keep a count of the lines that have the
2139# same message, and output that count at the end of the file's processing.
2140# This keeps the number of messages down to a manageable amount.
2141#
2142# get_missings() should be called to retrieve any @missing input lines.
2143# Messages will be raised if this isn't done if the options aren't to ignore
2144# missings.
2145
2146sub trace { return main::trace(@_); }
2147
99870f4d
KW
2148{ # Closure
2149 # Keep track of fields that are to be put into the constructor.
2150 my %constructor_fields;
2151
2152 main::setup_package(Constructor_Fields => \%constructor_fields);
2153
2154 my %file; # Input file name, required
2155 main::set_access('file', \%file, qw{ c r });
2156
2157 my %first_released; # Unicode version file was first released in, required
2158 main::set_access('first_released', \%first_released, qw{ c r });
2159
2160 my %handler; # Subroutine to process the input file, defaults to
2161 # 'process_generic_property_file'
2162 main::set_access('handler', \%handler, qw{ c });
2163
2164 my %property;
2165 # name of property this file is for. defaults to none, meaning not
2166 # applicable, or is otherwise determinable, for example, from each line.
696609bf 2167 main::set_access('property', \%property, qw{ c r });
99870f4d
KW
2168
2169 my %optional;
2170 # If this is true, the file is optional. If not present, no warning is
2171 # output. If it is present, the string given by this parameter is
2172 # evaluated, and if false the file is not processed.
2173 main::set_access('optional', \%optional, 'c', 'r');
2174
2175 my %non_skip;
2176 # This is used for debugging, to skip processing of all but a few input
2177 # files. Add 'non_skip => 1' to the constructor for those files you want
2178 # processed when you set the $debug_skip global.
2179 main::set_access('non_skip', \%non_skip, 'c');
2180
37e2e78e 2181 my %skip;
09ca89ce
KW
2182 # This is used to skip processing of this input file semi-permanently,
2183 # when it evaluates to true. The value should be the reason the file is
2184 # being skipped. It is used for files that we aren't planning to process
2185 # anytime soon, but want to allow to be in the directory and not raise a
2186 # message that we are not handling. Mostly for test files. This is in
2187 # contrast to the non_skip element, which is supposed to be used very
2188 # temporarily for debugging. Sets 'optional' to 1. Also, files that we
2189 # pretty much will never look at can be placed in the global
1fec9f60 2190 # %ignored_files instead. Ones used here will be added to %skipped files
37e2e78e
KW
2191 main::set_access('skip', \%skip, 'c');
2192
99870f4d
KW
2193 my %each_line_handler;
2194 # list of subroutines to look at and filter each non-comment line in the
2195 # file. defaults to none. The subroutines are called in order, each is
2196 # to adjust $_ for the next one, and the final one adjusts it for
2197 # 'handler'
2198 main::set_access('each_line_handler', \%each_line_handler, 'c');
2199
83b68635
KW
2200 my %properties; # Optional ordered list of the properties that occur in each
2201 # meaningful line of the input file. If present, an appropriate
2202 # each_line_handler() is automatically generated and pushed onto the stack
2203 # of such handlers. This is useful when a file contains multiple
2204 # proerties per line, but no other special considerations are necessary.
2205 # The special value "<ignored>" means to discard the corresponding input
2206 # field.
2207 # Any @missing lines in the file should also match this syntax; no such
2208 # files exist as of 6.3. But if it happens in a future release, the code
2209 # could be expanded to properly parse them.
2210 main::set_access('properties', \%properties, qw{ c r });
2211
99870f4d
KW
2212 my %has_missings_defaults;
2213 # ? Are there lines in the file giving default values for code points
2214 # missing from it?. Defaults to NO_DEFAULTS. Otherwise NOT_IGNORED is
2215 # the norm, but IGNORED means it has such lines, but the handler doesn't
2216 # use them. Having these three states allows us to catch changes to the
83b68635
KW
2217 # UCD that this program should track. XXX This could be expanded to
2218 # specify the syntax for such lines, like %properties above.
99870f4d
KW
2219 main::set_access('has_missings_defaults',
2220 \%has_missings_defaults, qw{ c r });
2221
2222 my %pre_handler;
2223 # Subroutine to call before doing anything else in the file. If undef, no
2224 # such handler is called.
2225 main::set_access('pre_handler', \%pre_handler, qw{ c });
2226
2227 my %eof_handler;
2228 # Subroutine to call upon getting an EOF on the input file, but before
2229 # that is returned to the main handler. This is to allow buffers to be
2230 # flushed. The handler is expected to call insert_lines() or
2231 # insert_adjusted() with the buffered material
2232 main::set_access('eof_handler', \%eof_handler, qw{ c r });
2233
2234 my %post_handler;
2235 # Subroutine to call after all the lines of the file are read in and
2236 # processed. If undef, no such handler is called.
2237 main::set_access('post_handler', \%post_handler, qw{ c });
2238
2239 my %progress_message;
2240 # Message to print to display progress in lieu of the standard one
2241 main::set_access('progress_message', \%progress_message, qw{ c });
2242
2243 my %handle;
2244 # cache open file handle, internal. Is undef if file hasn't been
2245 # processed at all, empty if has;
2246 main::set_access('handle', \%handle);
2247
2248 my %added_lines;
2249 # cache of lines added virtually to the file, internal
2250 main::set_access('added_lines', \%added_lines);
2251
74cd47d0
KW
2252 my %remapped_lines;
2253 # cache of lines added virtually to the file, internal
2254 main::set_access('remapped_lines', \%remapped_lines);
2255
99870f4d
KW
2256 my %errors;
2257 # cache of errors found, internal
2258 main::set_access('errors', \%errors);
2259
2260 my %missings;
2261 # storage of '@missing' defaults lines
2262 main::set_access('missings', \%missings);
2263
74cd47d0
KW
2264 sub _next_line;
2265 sub _next_line_with_remapped_range;
2266
99870f4d
KW
2267 sub new {
2268 my $class = shift;
2269
2270 my $self = bless \do{ my $anonymous_scalar }, $class;
ffe43484 2271 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2272
2273 # Set defaults
2274 $handler{$addr} = \&main::process_generic_property_file;
2275 $non_skip{$addr} = 0;
37e2e78e 2276 $skip{$addr} = 0;
99870f4d
KW
2277 $has_missings_defaults{$addr} = $NO_DEFAULTS;
2278 $handle{$addr} = undef;
2279 $added_lines{$addr} = [ ];
74cd47d0 2280 $remapped_lines{$addr} = [ ];
99870f4d
KW
2281 $each_line_handler{$addr} = [ ];
2282 $errors{$addr} = { };
2283 $missings{$addr} = [ ];
2284
2285 # Two positional parameters.
99f78760 2286 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
99870f4d
KW
2287 $file{$addr} = main::internal_file_to_platform(shift);
2288 $first_released{$addr} = shift;
2289
2290 # The rest of the arguments are key => value pairs
2291 # %constructor_fields has been set up earlier to list all possible
2292 # ones. Either set or push, depending on how the default has been set
2293 # up just above.
2294 my %args = @_;
2295 foreach my $key (keys %args) {
2296 my $argument = $args{$key};
2297
2298 # Note that the fields are the lower case of the constructor keys
2299 my $hash = $constructor_fields{lc $key};
2300 if (! defined $hash) {
2301 Carp::my_carp_bug("Unrecognized parameters '$key => $argument' to new() for $self. Skipped");
2302 next;
2303 }
2304 if (ref $hash->{$addr} eq 'ARRAY') {
2305 if (ref $argument eq 'ARRAY') {
2306 foreach my $argument (@{$argument}) {
2307 next if ! defined $argument;
2308 push @{$hash->{$addr}}, $argument;
2309 }
2310 }
2311 else {
2312 push @{$hash->{$addr}}, $argument if defined $argument;
2313 }
2314 }
2315 else {
2316 $hash->{$addr} = $argument;
2317 }
2318 delete $args{$key};
2319 };
2320
2321 # If the file has a property for it, it means that the property is not
2322 # listed in the file's entries. So add a handler to the list of line
2323 # handlers to insert the property name into the lines, to provide a
2324 # uniform interface to the final processing subroutine.
2325 # the final code doesn't have to worry about that.
2326 if ($property{$addr}) {
2327 push @{$each_line_handler{$addr}}, \&_insert_property_into_line;
2328 }
2329
2330 if ($non_skip{$addr} && ! $debug_skip && $verbosity) {
2331 print "Warning: " . __PACKAGE__ . " constructor for $file{$addr} has useless 'non_skip' in it\n";
a3a8c5f0 2332 }
99870f4d 2333
09ca89ce
KW
2334 # If skipping, set to optional, and add to list of ignored files,
2335 # including its reason
2336 if ($skip{$addr}) {
2337 $optional{$addr} = 1;
1fec9f60 2338 $skipped_files{$file{$addr}} = $skip{$addr}
09ca89ce 2339 }
83b68635
KW
2340 elsif ($properties{$addr}) {
2341
2342 # Add a handler for each line in the input so that it creates a
2343 # separate input line for each property in those input lines, thus
2344 # making them suitable for process_generic_property_file().
2345
2346 push @{$each_line_handler{$addr}},
2347 sub {
2348 my $file = shift;
2349 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2350
2351 my @fields = split /\s*;\s*/, $_, -1;
2352
2353 if (@fields - 1 > @{$properties{$addr}}) {
2354 $file->carp_bad_line('Extra fields');
2355 $_ = "";
2356 return;
2357 }
2358 my $range = shift @fields; # 0th element is always the
2359 # range
2360
2361 # The next fields in the input line correspond
2362 # respectively to the stored properties.
2363 for my $i (0 .. @{$properties{$addr}} - 1) {
2364 my $property_name = $properties{$addr}[$i];
2365 next if $property_name eq '<ignored>';
2366 $file->insert_adjusted_lines(
2367 "$range; $property_name; $fields[$i]");
2368 }
2369 $_ = "";
2370
2371 return;
2372 };
2373 }
37e2e78e 2374
74cd47d0
KW
2375 { # On non-ascii platforms, we use a special handler
2376 no strict;
2377 no warnings 'once';
2378 *next_line = (main::NON_ASCII_PLATFORM)
2379 ? *_next_line_with_remapped_range
2380 : *_next_line;
2381 }
2382
99870f4d 2383 return $self;
d73e5302
JH
2384 }
2385
cf25bb62 2386
99870f4d
KW
2387 use overload
2388 fallback => 0,
2389 qw("") => "_operator_stringify",
2390 "." => \&main::_operator_dot,
1285127e 2391 ".=" => \&main::_operator_dot_equal,
99870f4d 2392 ;
cf25bb62 2393
99870f4d
KW
2394 sub _operator_stringify {
2395 my $self = shift;
cf25bb62 2396
99870f4d 2397 return __PACKAGE__ . " object for " . $self->file;
d73e5302 2398 }
d73e5302 2399
99870f4d
KW
2400 # flag to make sure extracted files are processed early
2401 my $seen_non_extracted_non_age = 0;
d73e5302 2402
99870f4d
KW
2403 sub run {
2404 # Process the input object $self. This opens and closes the file and
2405 # calls all the handlers for it. Currently, this can only be called
2406 # once per file, as it destroy's the EOF handler
d73e5302 2407
99870f4d
KW
2408 my $self = shift;
2409 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
b6922eda 2410
ffe43484 2411 my $addr = do { no overloading; pack 'J', $self; };
b6922eda 2412
99870f4d 2413 my $file = $file{$addr};
d73e5302 2414
99870f4d
KW
2415 # Don't process if not expecting this file (because released later
2416 # than this Unicode version), and isn't there. This means if someone
2417 # copies it into an earlier version's directory, we will go ahead and
2418 # process it.
2419 return if $first_released{$addr} gt $v_version && ! -e $file;
2420
2421 # If in debugging mode and this file doesn't have the non-skip
2422 # flag set, and isn't one of the critical files, skip it.
2423 if ($debug_skip
2424 && $first_released{$addr} ne v0
2425 && ! $non_skip{$addr})
2426 {
2427 print "Skipping $file in debugging\n" if $verbosity;
2428 return;
2429 }
2430
2431 # File could be optional
37e2e78e 2432 if ($optional{$addr}) {
99870f4d
KW
2433 return unless -e $file;
2434 my $result = eval $optional{$addr};
2435 if (! defined $result) {
2436 Carp::my_carp_bug("Got '$@' when tried to eval $optional{$addr}. $file Skipped.");
2437 return;
2438 }
2439 if (! $result) {
2440 if ($verbosity) {
2441 print STDERR "Skipping processing input file '$file' because '$optional{$addr}' is not true\n";
2442 }
2443 return;
2444 }
2445 }
2446
2447 if (! defined $file || ! -e $file) {
2448
2449 # If the file doesn't exist, see if have internal data for it
2450 # (based on first_released being 0).
2451 if ($first_released{$addr} eq v0) {
2452 $handle{$addr} = 'pretend_is_open';
2453 }
2454 else {
2455 if (! $optional{$addr} # File could be optional
2456 && $v_version ge $first_released{$addr})
2457 {
2458 print STDERR "Skipping processing input file '$file' because not found\n" if $v_version ge $first_released{$addr};
2459 }
2460 return;
2461 }
2462 }
2463 else {
2464
37e2e78e
KW
2465 # Here, the file exists. Some platforms may change the case of
2466 # its name
99870f4d 2467 if ($seen_non_extracted_non_age) {
517956bf 2468 if ($file =~ /$EXTRACTED/i) {
1675ea0d 2469 Carp::my_carp_bug(main::join_lines(<<END
99f78760 2470$file should be processed just after the 'Prop...Alias' files, and before
99870f4d
KW
2471anything not in the $EXTRACTED_DIR directory. Proceeding, but the results may
2472have subtle problems
2473END
2474 ));
2475 }
2476 }
2477 elsif ($EXTRACTED_DIR
2478 && $first_released{$addr} ne v0
517956bf
CB
2479 && $file !~ /$EXTRACTED/i
2480 && lc($file) ne 'dage.txt')
99870f4d
KW
2481 {
2482 # We don't set this (by the 'if' above) if we have no
2483 # extracted directory, so if running on an early version,
2484 # this test won't work. Not worth worrying about.
2485 $seen_non_extracted_non_age = 1;
2486 }
2487
2488 # And mark the file as having being processed, and warn if it
2489 # isn't a file we are expecting. As we process the files,
2490 # they are deleted from the hash, so any that remain at the
2491 # end of the program are files that we didn't process.
517956bf 2492 my $fkey = File::Spec->rel2abs($file);
faf3cf6b
KW
2493 my $expecting = delete $potential_files{lc($fkey)};
2494
678f13d5
KW
2495 Carp::my_carp("Was not expecting '$file'.") if
2496 ! $expecting
99870f4d
KW
2497 && ! defined $handle{$addr};
2498
37e2e78e
KW
2499 # Having deleted from expected files, we can quit if not to do
2500 # anything. Don't print progress unless really want verbosity
2501 if ($skip{$addr}) {
2502 print "Skipping $file.\n" if $verbosity >= $VERBOSE;
2503 return;
2504 }
2505
99870f4d
KW
2506 # Open the file, converting the slashes used in this program
2507 # into the proper form for the OS
2508 my $file_handle;
2509 if (not open $file_handle, "<", $file) {
2510 Carp::my_carp("Can't open $file. Skipping: $!");
2511 return 0;
2512 }
2513 $handle{$addr} = $file_handle; # Cache the open file handle
9e65c3f4 2514
ebdc9bc9
KW
2515 if ($v_version ge v3.2.0
2516 && lc($file) ne 'unicodedata.txt'
2517
2518 # Unihan files used another format until v7
2519 && ($v_version ge v7.0.0 || $file !~ /^Unihan/i))
2520 {
9e65c3f4
KW
2521 $_ = <$file_handle>;
2522 if ($_ !~ / - $string_version \. /x) {
2523 chomp;
2524 $_ =~ s/^#\s*//;
2525 die Carp::my_carp("File '$file' is version '$_'. It should be version $string_version");
2526 }
2527 }
99870f4d
KW
2528 }
2529
2530 if ($verbosity >= $PROGRESS) {
2531 if ($progress_message{$addr}) {
2532 print "$progress_message{$addr}\n";
2533 }
2534 else {
2535 # If using a virtual file, say so.
2536 print "Processing ", (-e $file)
2537 ? $file
2538 : "substitute $file",
2539 "\n";
2540 }
2541 }
2542
2543
2544 # Call any special handler for before the file.
2545 &{$pre_handler{$addr}}($self) if $pre_handler{$addr};
2546
2547 # Then the main handler
2548 &{$handler{$addr}}($self);
2549
2550 # Then any special post-file handler.
2551 &{$post_handler{$addr}}($self) if $post_handler{$addr};
2552
2553 # If any errors have been accumulated, output the counts (as the first
2554 # error message in each class was output when it was encountered).
2555 if ($errors{$addr}) {
2556 my $total = 0;
2557 my $types = 0;
2558 foreach my $error (keys %{$errors{$addr}}) {
2559 $total += $errors{$addr}->{$error};
2560 delete $errors{$addr}->{$error};
2561 $types++;
2562 }
2563 if ($total > 1) {
2564 my $message
2565 = "A total of $total lines had errors in $file. ";
2566
2567 $message .= ($types == 1)
2568 ? '(Only the first one was displayed.)'
2569 : '(Only the first of each type was displayed.)';
2570 Carp::my_carp($message);
2571 }
2572 }
2573
2574 if (@{$missings{$addr}}) {
2575 Carp::my_carp_bug("Handler for $file didn't look at all the \@missing lines. Generated tables likely are wrong");
2576 }
2577
2578 # If a real file handle, close it.
2579 close $handle{$addr} or Carp::my_carp("Can't close $file: $!") if
2580 ref $handle{$addr};
2581 $handle{$addr} = ""; # Uses empty to indicate that has already seen
2582 # the file, as opposed to undef
2583 return;
2584 }
2585
74cd47d0 2586 sub _next_line {
99870f4d
KW
2587 # Sets $_ to be the next logical input line, if any. Returns non-zero
2588 # if such a line exists. 'logical' means that any lines that have
2589 # been added via insert_lines() will be returned in $_ before the file
2590 # is read again.
2591
2592 my $self = shift;
2593 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2594
ffe43484 2595 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2596
2597 # Here the file is open (or if the handle is not a ref, is an open
2598 # 'virtual' file). Get the next line; any inserted lines get priority
2599 # over the file itself.
2600 my $adjusted;
2601
2602 LINE:
2603 while (1) { # Loop until find non-comment, non-empty line
2604 #local $to_trace = 1 if main::DEBUG;
2605 my $inserted_ref = shift @{$added_lines{$addr}};
2606 if (defined $inserted_ref) {
2607 ($adjusted, $_) = @{$inserted_ref};
2608 trace $adjusted, $_ if main::DEBUG && $to_trace;
2609 return 1 if $adjusted;
2610 }
2611 else {
2612 last if ! ref $handle{$addr}; # Don't read unless is real file
2613 last if ! defined ($_ = readline $handle{$addr});
2614 }
2615 chomp;
2616 trace $_ if main::DEBUG && $to_trace;
2617
2618 # See if this line is the comment line that defines what property
2619 # value that code points that are not listed in the file should
2620 # have. The format or existence of these lines is not guaranteed
2621 # by Unicode since they are comments, but the documentation says
2622 # that this was added for machine-readability, so probably won't
2623 # change. This works starting in Unicode Version 5.0. They look
2624 # like:
2625 #
2626 # @missing: 0000..10FFFF; Not_Reordered
2627 # @missing: 0000..10FFFF; Decomposition_Mapping; <code point>
2628 # @missing: 0000..10FFFF; ; NaN
2629 #
2630 # Save the line for a later get_missings() call.
2631 if (/$missing_defaults_prefix/) {
2632 if ($has_missings_defaults{$addr} == $NO_DEFAULTS) {
2633 $self->carp_bad_line("Unexpected \@missing line. Assuming no missing entries");
2634 }
2635 elsif ($has_missings_defaults{$addr} == $NOT_IGNORED) {
2636 my @defaults = split /\s* ; \s*/x, $_;
2637
2638 # The first field is the @missing, which ends in a
2639 # semi-colon, so can safely shift.
2640 shift @defaults;
2641
2642 # Some of these lines may have empty field placeholders
2643 # which get in the way. An example is:
2644 # @missing: 0000..10FFFF; ; NaN
2645 # Remove them. Process starting from the top so the
2646 # splice doesn't affect things still to be looked at.
2647 for (my $i = @defaults - 1; $i >= 0; $i--) {
2648 next if $defaults[$i] ne "";
2649 splice @defaults, $i, 1;
2650 }
2651
2652 # What's left should be just the property (maybe) and the
2653 # default. Having only one element means it doesn't have
2654 # the property.
2655 my $default;
2656 my $property;
2657 if (@defaults >= 1) {
2658 if (@defaults == 1) {
2659 $default = $defaults[0];
2660 }
2661 else {
2662 $property = $defaults[0];
2663 $default = $defaults[1];
2664 }
2665 }
2666
2667 if (@defaults < 1
2668 || @defaults > 2
2669 || ($default =~ /^</
2670 && $default !~ /^<code *point>$/i
09f8d0ac
KW
2671 && $default !~ /^<none>$/i
2672 && $default !~ /^<script>$/i))
99870f4d
KW
2673 {
2674 $self->carp_bad_line("Unrecognized \@missing line: $_. Assuming no missing entries");
2675 }
2676 else {
2677
2678 # If the property is missing from the line, it should
2679 # be the one for the whole file
2680 $property = $property{$addr} if ! defined $property;
2681
2682 # Change <none> to the null string, which is what it
2683 # really means. If the default is the code point
2684 # itself, set it to <code point>, which is what
2685 # Unicode uses (but sometimes they've forgotten the
2686 # space)
2687 if ($default =~ /^<none>$/i) {
2688 $default = "";
2689 }
2690 elsif ($default =~ /^<code *point>$/i) {
2691 $default = $CODE_POINT;
2692 }
09f8d0ac
KW
2693 elsif ($default =~ /^<script>$/i) {
2694
2695 # Special case this one. Currently is from
2696 # ScriptExtensions.txt, and means for all unlisted
2697 # code points, use their Script property values.
2698 # For the code points not listed in that file, the
2699 # default value is 'Unknown'.
2700 $default = "Unknown";
2701 }
99870f4d
KW
2702
2703 # Store them as a sub-arrays with both components.
2704 push @{$missings{$addr}}, [ $default, $property ];
2705 }
2706 }
2707
2708 # There is nothing for the caller to process on this comment
2709 # line.
2710 next;
2711 }
2712
2713 # Remove comments and trailing space, and skip this line if the
2714 # result is empty
2715 s/#.*//;
2716 s/\s+$//;
2717 next if /^$/;
2718
2719 # Call any handlers for this line, and skip further processing of
2720 # the line if the handler sets the line to null.
2721 foreach my $sub_ref (@{$each_line_handler{$addr}}) {
2722 &{$sub_ref}($self);
2723 next LINE if /^$/;
2724 }
2725
2726 # Here the line is ok. return success.
2727 return 1;
2728 } # End of looping through lines.
2729
2730 # If there is an EOF handler, call it (only once) and if it generates
2731 # more lines to process go back in the loop to handle them.
2732 if ($eof_handler{$addr}) {
2733 &{$eof_handler{$addr}}($self);
2734 $eof_handler{$addr} = ""; # Currently only get one shot at it.
2735 goto LINE if $added_lines{$addr};
2736 }
2737
2738 # Return failure -- no more lines.
2739 return 0;
2740
2741 }
2742
74cd47d0
KW
2743 sub _next_line_with_remapped_range {
2744 my $self = shift;
2745 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2746
2747 # like _next_line(), but for use on non-ASCII platforms. It sets $_
2748 # to be the next logical input line, if any. Returns non-zero if such
2749 # a line exists. 'logical' means that any lines that have been added
2750 # via insert_lines() will be returned in $_ before the file is read
2751 # again.
2752 #
2753 # The difference from _next_line() is that this remaps the Unicode
2754 # code points in the input to those of the native platform. Each
2755 # input line contains a single code point, or a single contiguous
2756 # range of them This routine splits each range into its individual
2757 # code points and caches them. It returns the cached values,
2758 # translated into their native equivalents, one at a time, for each
2759 # call, before reading the next line. Since native values can only be
2760 # a single byte wide, no translation is needed for code points above
2761 # 0xFF, and ranges that are entirely above that number are not split.
2762 # If an input line contains the range 254-1000, it would be split into
2763 # three elements: 254, 255, and 256-1000. (The downstream table
2764 # insertion code will sort and coalesce the individual code points
2765 # into appropriate ranges.)
2766
2767 my $addr = do { no overloading; pack 'J', $self; };
2768
2769 while (1) {
2770
2771 # Look in cache before reading the next line. Return any cached
2772 # value, translated
2773 my $inserted = shift @{$remapped_lines{$addr}};
2774 if (defined $inserted) {
2775 trace $inserted if main::DEBUG && $to_trace;
2776 $_ = $inserted =~ s/^ ( \d+ ) /sprintf("%04X", utf8::unicode_to_native($1))/xer;
2777 trace $_ if main::DEBUG && $to_trace;
2778 return 1;
2779 }
2780
2781 # Get the next line.
2782 return 0 unless _next_line($self);
2783
2784 # If there is a special handler for it, return the line,
2785 # untranslated. This should happen only for files that are
2786 # special, not being code-point related, such as property names.
2787 return 1 if $handler{$addr}
2788 != \&main::process_generic_property_file;
2789
2790 my ($range, $property_name, $map, @remainder)
2791 = split /\s*;\s*/, $_, -1; # -1 => retain trailing null fields
2792
2793 if (@remainder
2794 || ! defined $property_name
2795 || $range !~ /^ ($code_point_re) (?:\.\. ($code_point_re) )? $/x)
2796 {
2797 Carp::my_carp_bug("Unrecognized input line '$_'. Ignored");
2798 }
2799
2800 my $low = hex $1;
2801 my $high = (defined $2) ? hex $2 : $low;
2802
2803 # If the input maps the range to another code point, remap the
2804 # target if it is between 0 and 255.
2805 my $tail;
2806 if (defined $map) {
2807 $map =~ s/\b 00 ( [0-9A-F]{2} ) \b/sprintf("%04X", utf8::unicode_to_native(hex $1))/gxe;
2808 $tail = "$property_name; $map";
2809 $_ = "$range; $tail";
2810 }
2811 else {
2812 $tail = $property_name;
2813 }
2814
2815 # If entire range is above 255, just return it, unchanged (except
2816 # any mapped-to code point, already changed above)
2817 return 1 if $low > 255;
2818
2819 # Cache an entry for every code point < 255. For those in the
2820 # range above 255, return a dummy entry for just that portion of
2821 # the range. Note that this will be out-of-order, but that is not
2822 # a problem.
2823 foreach my $code_point ($low .. $high) {
2824 if ($code_point > 255) {
2825 $_ = sprintf "%04X..%04X; $tail", $code_point, $high;
2826 return 1;
2827 }
2828 push @{$remapped_lines{$addr}}, "$code_point; $tail";
2829 }
2830 } # End of looping through lines.
2831
2832 # NOTREACHED
2833 }
2834
99870f4d
KW
2835# Not currently used, not fully tested.
2836# sub peek {
2837# # Non-destructive look-ahead one non-adjusted, non-comment, non-blank
2838# # record. Not callable from an each_line_handler(), nor does it call
2839# # an each_line_handler() on the line.
2840#
2841# my $self = shift;
ffe43484 2842# my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2843#
2844# foreach my $inserted_ref (@{$added_lines{$addr}}) {
2845# my ($adjusted, $line) = @{$inserted_ref};
2846# next if $adjusted;
2847#
2848# # Remove comments and trailing space, and return a non-empty
2849# # resulting line
2850# $line =~ s/#.*//;
2851# $line =~ s/\s+$//;
2852# return $line if $line ne "";
2853# }
2854#
2855# return if ! ref $handle{$addr}; # Don't read unless is real file
2856# while (1) { # Loop until find non-comment, non-empty line
2857# local $to_trace = 1 if main::DEBUG;
2858# trace $_ if main::DEBUG && $to_trace;
2859# return if ! defined (my $line = readline $handle{$addr});
2860# chomp $line;
2861# push @{$added_lines{$addr}}, [ 0, $line ];
2862#
2863# $line =~ s/#.*//;
2864# $line =~ s/\s+$//;
2865# return $line if $line ne "";
2866# }
2867#
2868# return;
2869# }
2870
2871
2872 sub insert_lines {
2873 # Lines can be inserted so that it looks like they were in the input
2874 # file at the place it was when this routine is called. See also
2875 # insert_adjusted_lines(). Lines inserted via this routine go through
2876 # any each_line_handler()
2877
2878 my $self = shift;
2879
2880 # Each inserted line is an array, with the first element being 0 to
2881 # indicate that this line hasn't been adjusted, and needs to be
2882 # processed.
f998e60c 2883 no overloading;
051df77b 2884 push @{$added_lines{pack 'J', $self}}, map { [ 0, $_ ] } @_;
99870f4d
KW
2885 return;
2886 }
2887
2888 sub insert_adjusted_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_lines(). Lines inserted via this routine are already fully
2892 # adjusted, ready to be processed; each_line_handler()s handlers will
2893 # not be called. This means this is not a completely general
2894 # facility, as only the last each_line_handler on the stack should
2895 # call this. It could be made more general, by passing to each of the
2896 # line_handlers their position on the stack, which they would pass on
2897 # to this routine, and that would replace the boolean first element in
2898 # the anonymous array pushed here, so that the next_line routine could
2899 # use that to call only those handlers whose index is after it on the
2900 # stack. But this is overkill for what is needed now.
2901
2902 my $self = shift;
2903 trace $_[0] if main::DEBUG && $to_trace;
2904
2905 # Each inserted line is an array, with the first element being 1 to
2906 # indicate that this line has been adjusted
f998e60c 2907 no overloading;
051df77b 2908 push @{$added_lines{pack 'J', $self}}, map { [ 1, $_ ] } @_;
99870f4d
KW
2909 return;
2910 }
2911
2912 sub get_missings {
2913 # Returns the stored up @missings lines' values, and clears the list.
2914 # The values are in an array, consisting of the default in the first
2915 # element, and the property in the 2nd. However, since these lines
2916 # can be stacked up, the return is an array of all these arrays.
2917
2918 my $self = shift;
2919 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2920
ffe43484 2921 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2922
2923 # If not accepting a list return, just return the first one.
2924 return shift @{$missings{$addr}} unless wantarray;
2925
2926 my @return = @{$missings{$addr}};
2927 undef @{$missings{$addr}};
2928 return @return;
2929 }
2930
2931 sub _insert_property_into_line {
2932 # Add a property field to $_, if this file requires it.
2933
f998e60c 2934 my $self = shift;
ffe43484 2935 my $addr = do { no overloading; pack 'J', $self; };
f998e60c 2936 my $property = $property{$addr};
99870f4d
KW
2937 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2938
2939 $_ =~ s/(;|$)/; $property$1/;
2940 return;
2941 }
2942
2943 sub carp_bad_line {
2944 # Output consistent error messages, using either a generic one, or the
2945 # one given by the optional parameter. To avoid gazillions of the
2946 # same message in case the syntax of a file is way off, this routine
2947 # only outputs the first instance of each message, incrementing a
2948 # count so the totals can be output at the end of the file.
2949
2950 my $self = shift;
2951 my $message = shift;
2952 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2953
ffe43484 2954 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2955
2956 $message = 'Unexpected line' unless $message;
2957
2958 # No trailing punctuation so as to fit with our addenda.
2959 $message =~ s/[.:;,]$//;
2960
2961 # If haven't seen this exact message before, output it now. Otherwise
2962 # increment the count of how many times it has occurred
2963 unless ($errors{$addr}->{$message}) {
2964 Carp::my_carp("$message in '$_' in "
f998e60c 2965 . $file{$addr}
99870f4d
KW
2966 . " at line $.. Skipping this line;");
2967 $errors{$addr}->{$message} = 1;
2968 }
2969 else {
2970 $errors{$addr}->{$message}++;
2971 }
2972
2973 # Clear the line to prevent any further (meaningful) processing of it.
2974 $_ = "";
2975
2976 return;
2977 }
2978} # End closure
2979
2980package Multi_Default;
2981
2982# Certain properties in early versions of Unicode had more than one possible
2983# default for code points missing from the files. In these cases, one
2984# default applies to everything left over after all the others are applied,
2985# and for each of the others, there is a description of which class of code
2986# points applies to it. This object helps implement this by storing the
2987# defaults, and for all but that final default, an eval string that generates
2988# the class that it applies to.
2989
2990
2991{ # Closure
2992
2993 main::setup_package();
2994
2995 my %class_defaults;
2996 # The defaults structure for the classes
2997 main::set_access('class_defaults', \%class_defaults);
2998
2999 my %other_default;
3000 # The default that applies to everything left over.
3001 main::set_access('other_default', \%other_default, 'r');
3002
3003
3004 sub new {
3005 # The constructor is called with default => eval pairs, terminated by
3006 # the left-over default. e.g.
3007 # Multi_Default->new(
3008 # 'T' => '$gc->table("Mn") + $gc->table("Cf") - 0x200C
3009 # - 0x200D',
3010 # 'R' => 'some other expression that evaluates to code points',
3011 # .
3012 # .
3013 # .
3014 # 'U'));
3015
3016 my $class = shift;
3017
3018 my $self = bless \do{my $anonymous_scalar}, $class;
ffe43484 3019 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3020
3021 while (@_ > 1) {
3022 my $default = shift;
3023 my $eval = shift;
3024 $class_defaults{$addr}->{$default} = $eval;
3025 }
3026
3027 $other_default{$addr} = shift;
3028
3029 return $self;
3030 }
3031
3032 sub get_next_defaults {
3033 # Iterates and returns the next class of defaults.
3034 my $self = shift;
3035 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3036
ffe43484 3037 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3038
3039 return each %{$class_defaults{$addr}};
3040 }
3041}
3042
3043package Alias;
3044
3045# An alias is one of the names that a table goes by. This class defines them
3046# including some attributes. Everything is currently setup in the
3047# constructor.
3048
3049
3050{ # Closure
3051
3052 main::setup_package();
3053
3054 my %name;
3055 main::set_access('name', \%name, 'r');
3056
3057 my %loose_match;
c12f2655 3058 # Should this name match loosely or not.
99870f4d
KW
3059 main::set_access('loose_match', \%loose_match, 'r');
3060
33e96e72
KW
3061 my %make_re_pod_entry;
3062 # Some aliases should not get their own entries in the re section of the
3063 # pod, because they are covered by a wild-card, and some we want to
3064 # discourage use of. Binary
f82fe4ba 3065 main::set_access('make_re_pod_entry', \%make_re_pod_entry, 'r', 's');
99870f4d 3066
fd1e3e84
KW
3067 my %ucd;
3068 # Is this documented to be accessible via Unicode::UCD
3069 main::set_access('ucd', \%ucd, 'r', 's');
3070
99870f4d
KW
3071 my %status;
3072 # Aliases have a status, like deprecated, or even suppressed (which means
3073 # they don't appear in documentation). Enum
3074 main::set_access('status', \%status, 'r');
3075
0eac1e20 3076 my %ok_as_filename;
99870f4d
KW
3077 # Similarly, some aliases should not be considered as usable ones for
3078 # external use, such as file names, or we don't want documentation to
3079 # recommend them. Boolean
0eac1e20 3080 main::set_access('ok_as_filename', \%ok_as_filename, 'r');
99870f4d
KW
3081
3082 sub new {
3083 my $class = shift;
3084
3085 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 3086 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3087
3088 $name{$addr} = shift;
3089 $loose_match{$addr} = shift;
33e96e72 3090 $make_re_pod_entry{$addr} = shift;
0eac1e20 3091 $ok_as_filename{$addr} = shift;
99870f4d 3092 $status{$addr} = shift;
fd1e3e84 3093 $ucd{$addr} = shift;
99870f4d
KW
3094
3095 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3096
3097 # Null names are never ok externally
0eac1e20 3098 $ok_as_filename{$addr} = 0 if $name{$addr} eq "";
99870f4d
KW
3099
3100 return $self;
3101 }
3102}
3103
3104package Range;
3105
3106# A range is the basic unit for storing code points, and is described in the
3107# comments at the beginning of the program. Each range has a starting code
3108# point; an ending code point (not less than the starting one); a value
3109# that applies to every code point in between the two end-points, inclusive;
3110# and an enum type that applies to the value. The type is for the user's
3111# convenience, and has no meaning here, except that a non-zero type is
3112# considered to not obey the normal Unicode rules for having standard forms.
3113#
3114# The same structure is used for both map and match tables, even though in the
3115# latter, the value (and hence type) is irrelevant and could be used as a
3116# comment. In map tables, the value is what all the code points in the range
3117# map to. Type 0 values have the standardized version of the value stored as
3118# well, so as to not have to recalculate it a lot.
3119
3120sub trace { return main::trace(@_); }
3121
3122{ # Closure
3123
3124 main::setup_package();
3125
3126 my %start;
3127 main::set_access('start', \%start, 'r', 's');
3128
3129 my %end;
3130 main::set_access('end', \%end, 'r', 's');
3131
3132 my %value;
3133 main::set_access('value', \%value, 'r');
3134
3135 my %type;
3136 main::set_access('type', \%type, 'r');
3137
3138 my %standard_form;
3139 # The value in internal standard form. Defined only if the type is 0.
3140 main::set_access('standard_form', \%standard_form);
3141
3142 # Note that if these fields change, the dump() method should as well
3143
3144 sub new {
3145 return Carp::carp_too_few_args(\@_, 3) if main::DEBUG && @_ < 3;
3146 my $class = shift;
3147
3148 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 3149 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3150
3151 $start{$addr} = shift;
3152 $end{$addr} = shift;
3153
3154 my %args = @_;
3155
3156 my $value = delete $args{'Value'}; # Can be 0
3157 $value = "" unless defined $value;
3158 $value{$addr} = $value;
3159
3160 $type{$addr} = delete $args{'Type'} || 0;
3161
3162 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3163
99870f4d
KW
3164 return $self;
3165 }
3166
3167 use overload
3168 fallback => 0,
3169 qw("") => "_operator_stringify",
3170 "." => \&main::_operator_dot,
1285127e 3171 ".=" => \&main::_operator_dot_equal,
99870f4d
KW
3172 ;
3173
3174 sub _operator_stringify {
3175 my $self = shift;
ffe43484 3176 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3177
3178 # Output it like '0041..0065 (value)'
3179 my $return = sprintf("%04X", $start{$addr})
3180 . '..'
3181 . sprintf("%04X", $end{$addr});
3182 my $value = $value{$addr};
3183 my $type = $type{$addr};
3184 $return .= ' (';
3185 $return .= "$value";
3186 $return .= ", Type=$type" if $type != 0;
3187 $return .= ')';
3188
3189 return $return;
3190 }
3191
3192 sub standard_form {
c292d35a
NC
3193 # Calculate the standard form only if needed, and cache the result.
3194 # The standard form is the value itself if the type is special.
3195 # This represents a considerable CPU and memory saving - at the time
3196 # of writing there are 368676 non-special objects, but the standard
3197 # form is only requested for 22047 of them - ie about 6%.
99870f4d
KW
3198
3199 my $self = shift;
3200 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3201
ffe43484 3202 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3203
3204 return $standard_form{$addr} if defined $standard_form{$addr};
c292d35a
NC
3205
3206 my $value = $value{$addr};
3207 return $value if $type{$addr};
3208 return $standard_form{$addr} = main::standardize($value);
99870f4d
KW
3209 }
3210
3211 sub dump {
3212 # Human, not machine readable. For machine readable, comment out this
3213 # entire routine and let the standard one take effect.
3214 my $self = shift;
3215 my $indent = shift;
3216 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3217
ffe43484 3218 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3219
3220 my $return = $indent
3221 . sprintf("%04X", $start{$addr})
3222 . '..'
3223 . sprintf("%04X", $end{$addr})
3224 . " '$value{$addr}';";
3225 if (! defined $standard_form{$addr}) {
3226 $return .= "(type=$type{$addr})";
3227 }
3228 elsif ($standard_form{$addr} ne $value{$addr}) {
3229 $return .= "(standard '$standard_form{$addr}')";
3230 }
3231 return $return;
3232 }
3233} # End closure
3234
3235package _Range_List_Base;
3236
3237# Base class for range lists. A range list is simply an ordered list of
3238# ranges, so that the ranges with the lowest starting numbers are first in it.
3239#
3240# When a new range is added that is adjacent to an existing range that has the
3241# same value and type, it merges with it to form a larger range.
3242#
3243# Ranges generally do not overlap, except that there can be multiple entries
3244# of single code point ranges. This is because of NameAliases.txt.
3245#
3246# In this program, there is a standard value such that if two different
3247# values, have the same standard value, they are considered equivalent. This
3248# value was chosen so that it gives correct results on Unicode data
3249
3250# There are a number of methods to manipulate range lists, and some operators
3251# are overloaded to handle them.
3252
99870f4d
KW
3253sub trace { return main::trace(@_); }
3254
3255{ # Closure
3256
3257 our $addr;
3258
5b348b71
KW
3259 # Max is initialized to a negative value that isn't adjacent to 0, for
3260 # simpler tests
3261 my $max_init = -2;
3262
99870f4d
KW
3263 main::setup_package();
3264
3265 my %ranges;
3266 # The list of ranges
3267 main::set_access('ranges', \%ranges, 'readable_array');
3268
3269 my %max;
3270 # The highest code point in the list. This was originally a method, but
3271 # actual measurements said it was used a lot.
3272 main::set_access('max', \%max, 'r');
3273
3274 my %each_range_iterator;
3275 # Iterator position for each_range()
3276 main::set_access('each_range_iterator', \%each_range_iterator);
3277
3278 my %owner_name_of;
3279 # Name of parent this is attached to, if any. Solely for better error
3280 # messages.
3281 main::set_access('owner_name_of', \%owner_name_of, 'p_r');
3282
3283 my %_search_ranges_cache;
3284 # A cache of the previous result from _search_ranges(), for better
3285 # performance
3286 main::set_access('_search_ranges_cache', \%_search_ranges_cache);
3287
3288 sub new {
3289 my $class = shift;
3290 my %args = @_;
3291
3292 # Optional initialization data for the range list.
3293 my $initialize = delete $args{'Initialize'};
3294
3295 my $self;
3296
3297 # Use _union() to initialize. _union() returns an object of this
3298 # class, which means that it will call this constructor recursively.
3299 # But it won't have this $initialize parameter so that it won't
3300 # infinitely loop on this.
3301 return _union($class, $initialize, %args) if defined $initialize;
3302
3303 $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 3304 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3305
3306 # Optional parent object, only for debug info.
3307 $owner_name_of{$addr} = delete $args{'Owner'};
3308 $owner_name_of{$addr} = "" if ! defined $owner_name_of{$addr};
3309
3310 # Stringify, in case it is an object.
3311 $owner_name_of{$addr} = "$owner_name_of{$addr}";
3312
3313 # This is used only for error messages, and so a colon is added
3314 $owner_name_of{$addr} .= ": " if $owner_name_of{$addr} ne "";
3315
3316 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3317
5b348b71 3318 $max{$addr} = $max_init;
99870f4d
KW
3319
3320 $_search_ranges_cache{$addr} = 0;
3321 $ranges{$addr} = [];
3322
3323 return $self;
3324 }
3325
3326 use overload
3327 fallback => 0,
3328 qw("") => "_operator_stringify",
3329 "." => \&main::_operator_dot,
1285127e 3330 ".=" => \&main::_operator_dot_equal,
99870f4d
KW
3331 ;
3332
3333 sub _operator_stringify {
3334 my $self = shift;
ffe43484 3335 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3336
3337 return "Range_List attached to '$owner_name_of{$addr}'"
3338 if $owner_name_of{$addr};
3339 return "anonymous Range_List " . \$self;
3340 }
3341
3342 sub _union {
3343 # Returns the union of the input code points. It can be called as
3344 # either a constructor or a method. If called as a method, the result
3345 # will be a new() instance of the calling object, containing the union
3346 # of that object with the other parameter's code points; if called as
d59563d0 3347 # a constructor, the first parameter gives the class that the new object
99870f4d
KW
3348 # should be, and the second parameter gives the code points to go into
3349 # it.
3350 # In either case, there are two parameters looked at by this routine;
3351 # any additional parameters are passed to the new() constructor.
3352 #
3353 # The code points can come in the form of some object that contains
3354 # ranges, and has a conventionally named method to access them; or
3355 # they can be an array of individual code points (as integers); or
3356 # just a single code point.
3357 #
3358 # If they are ranges, this routine doesn't make any effort to preserve
3198cc57
KW
3359 # the range values and types of one input over the other. Therefore
3360 # this base class should not allow _union to be called from other than
99870f4d
KW
3361 # initialization code, so as to prevent two tables from being added
3362 # together where the range values matter. The general form of this
3363 # routine therefore belongs in a derived class, but it was moved here
3364 # to avoid duplication of code. The failure to overload this in this
3365 # class keeps it safe.
3198cc57
KW
3366 #
3367 # It does make the effort during initialization to accept tables with
3368 # multiple values for the same code point, and to preserve the order
3369 # of these. If there is only one input range or range set, it doesn't
3370 # sort (as it should already be sorted to the desired order), and will
3371 # accept multiple values per code point. Otherwise it will merge
3372 # multiple values into a single one.
99870f4d
KW
3373
3374 my $self;
3375 my @args; # Arguments to pass to the constructor
3376
3377 my $class = shift;
3378
3379 # If a method call, will start the union with the object itself, and
3380 # the class of the new object will be the same as self.
3381 if (ref $class) {
3382 $self = $class;
3383 $class = ref $self;
3384 push @args, $self;
3385 }
3386
3387 # Add the other required parameter.
3388 push @args, shift;
3389 # Rest of parameters are passed on to the constructor
3390
3391 # Accumulate all records from both lists.
3392 my @records;
3198cc57 3393 my $input_count = 0;
99870f4d
KW
3394 for my $arg (@args) {
3395 #local $to_trace = 0 if main::DEBUG;
3396 trace "argument = $arg" if main::DEBUG && $to_trace;
3397 if (! defined $arg) {
3398 my $message = "";
3399 if (defined $self) {
f998e60c 3400 no overloading;
051df77b 3401 $message .= $owner_name_of{pack 'J', $self};
99870f4d 3402 }
ada6088e 3403 Carp::my_carp_bug($message . "Undefined argument to _union. No union done.");
99870f4d
KW
3404 return;
3405 }
3198cc57 3406
99870f4d
KW
3407 $arg = [ $arg ] if ! ref $arg;
3408 my $type = ref $arg;
3409 if ($type eq 'ARRAY') {
3410 foreach my $element (@$arg) {
3411 push @records, Range->new($element, $element);
3198cc57 3412 $input_count++;
99870f4d
KW
3413 }
3414 }
3415 elsif ($arg->isa('Range')) {
3416 push @records, $arg;
3198cc57 3417 $input_count++;
99870f4d
KW
3418 }
3419 elsif ($arg->can('ranges')) {
3420 push @records, $arg->ranges;
3198cc57 3421 $input_count++;
99870f4d
KW
3422 }
3423 else {
3424 my $message = "";
3425 if (defined $self) {
f998e60c 3426 no overloading;
051df77b 3427 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
3428 }
3429 Carp::my_carp_bug($message . "Cannot take the union of a $type. No union done.");
3430 return;
3431 }
3432 }
3433
3434 # Sort with the range containing the lowest ordinal first, but if
3435 # two ranges start at the same code point, sort with the bigger range
3436 # of the two first, because it takes fewer cycles.
3198cc57
KW
3437 if ($input_count > 1) {
3438 @records = sort { ($a->start <=> $b->start)
99870f4d
KW
3439 or
3440 # if b is shorter than a, b->end will be
3441 # less than a->end, and we want to select
3442 # a, so want to return -1
3443 ($b->end <=> $a->end)
3444 } @records;
3198cc57 3445 }
99870f4d
KW
3446
3447 my $new = $class->new(@_);
3448
3449 # Fold in records so long as they add new information.
3450 for my $set (@records) {
3451 my $start = $set->start;
3452 my $end = $set->end;
d59563d0 3453 my $value = $set->value;
3198cc57 3454 my $type = $set->type;
99870f4d 3455 if ($start > $new->max) {
3198cc57 3456 $new->_add_delete('+', $start, $end, $value, Type => $type);
99870f4d
KW
3457 }
3458 elsif ($end > $new->max) {
3198cc57
KW
3459 $new->_add_delete('+', $new->max +1, $end, $value,
3460 Type => $type);
3461 }
3462 elsif ($input_count == 1) {
3463 # Here, overlaps existing range, but is from a single input,
3464 # so preserve the multiple values from that input.
3465 $new->_add_delete('+', $start, $end, $value, Type => $type,
3466 Replace => $MULTIPLE_AFTER);
99870f4d
KW
3467 }
3468 }
3469
3470 return $new;
3471 }
3472
3473 sub range_count { # Return the number of ranges in the range list
3474 my $self = shift;
3475 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3476
f998e60c 3477 no overloading;
051df77b 3478 return scalar @{$ranges{pack 'J', $self}};
99870f4d
KW
3479 }
3480
3481 sub min {
3482 # Returns the minimum code point currently in the range list, or if
3483 # the range list is empty, 2 beyond the max possible. This is a
3484 # method because used so rarely, that not worth saving between calls,
3485 # and having to worry about changing it as ranges are added and
3486 # deleted.
3487
3488 my $self = shift;
3489 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3490
ffe43484 3491 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3492
3493 # If the range list is empty, return a large value that isn't adjacent
3494 # to any that could be in the range list, for simpler tests
2d88a86a 3495 return $MAX_WORKING_CODEPOINT + 2 unless scalar @{$ranges{$addr}};
99870f4d
KW
3496 return $ranges{$addr}->[0]->start;
3497 }
3498
3499 sub contains {
3500 # Boolean: Is argument in the range list? If so returns $i such that:
3501 # range[$i]->end < $codepoint <= range[$i+1]->end
3502 # which is one beyond what you want; this is so that the 0th range
3503 # doesn't return false
3504 my $self = shift;
3505 my $codepoint = shift;
3506 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3507
99870f4d
KW
3508 my $i = $self->_search_ranges($codepoint);
3509 return 0 unless defined $i;
3510
3511 # The search returns $i, such that
3512 # range[$i-1]->end < $codepoint <= range[$i]->end
3513 # So is in the table if and only iff it is at least the start position
3514 # of range $i.
f998e60c 3515 no overloading;
051df77b 3516 return 0 if $ranges{pack 'J', $self}->[$i]->start > $codepoint;
99870f4d
KW
3517 return $i + 1;
3518 }
3519
2f7a8815
KW
3520 sub containing_range {
3521 # Returns the range object that contains the code point, undef if none
3522
3523 my $self = shift;
3524 my $codepoint = shift;
3525 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3526
3527 my $i = $self->contains($codepoint);
3528 return unless $i;
3529
3530 # contains() returns 1 beyond where we should look
3531 no overloading;
3532 return $ranges{pack 'J', $self}->[$i-1];
3533 }
3534
99870f4d
KW
3535 sub value_of {
3536 # Returns the value associated with the code point, undef if none
3537
3538 my $self = shift;
3539 my $codepoint = shift;
3540 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3541
d69c231b
KW
3542 my $range = $self->containing_range($codepoint);
3543 return unless defined $range;
99870f4d 3544
d69c231b 3545 return $range->value;
99870f4d
KW
3546 }
3547
0a9dbafc
KW
3548 sub type_of {
3549 # Returns the type of the range containing the code point, undef if
3550 # the code point is not in the table
3551
3552 my $self = shift;
3553 my $codepoint = shift;
3554 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3555
3556 my $range = $self->containing_range($codepoint);
3557 return unless defined $range;
3558
3559 return $range->type;
3560 }
3561
99870f4d
KW
3562 sub _search_ranges {
3563 # Find the range in the list which contains a code point, or where it
3564 # should go if were to add it. That is, it returns $i, such that:
3565 # range[$i-1]->end < $codepoint <= range[$i]->end
3566 # Returns undef if no such $i is possible (e.g. at end of table), or
3567 # if there is an error.
3568
3569 my $self = shift;
3570 my $code_point = shift;
3571 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3572
ffe43484 3573 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3574
3575 return if $code_point > $max{$addr};
3576 my $r = $ranges{$addr}; # The current list of ranges
3577 my $range_list_size = scalar @$r;
3578 my $i;
3579
3580 use integer; # want integer division
3581
3582 # Use the cached result as the starting guess for this one, because,
3583 # an experiment on 5.1 showed that 90% of the time the cache was the
3584 # same as the result on the next call (and 7% it was one less).
3585 $i = $_search_ranges_cache{$addr};
3586 $i = 0 if $i >= $range_list_size; # Reset if no longer valid (prob.
3587 # from an intervening deletion
3588 #local $to_trace = 1 if main::DEBUG;
3589 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);
3590 return $i if $code_point <= $r->[$i]->end
3591 && ($i == 0 || $r->[$i-1]->end < $code_point);
3592
3593 # Here the cache doesn't yield the correct $i. Try adding 1.
3594 if ($i < $range_list_size - 1
3595 && $r->[$i]->end < $code_point &&
3596 $code_point <= $r->[$i+1]->end)
3597 {
3598 $i++;
3599 trace "next \$i is correct: $i" if main::DEBUG && $to_trace;
3600 $_search_ranges_cache{$addr} = $i;
3601 return $i;
3602 }
3603
3604 # Here, adding 1 also didn't work. We do a binary search to
3605 # find the correct position, starting with current $i
3606 my $lower = 0;
3607 my $upper = $range_list_size - 1;
3608 while (1) {
3609 trace "top of loop i=$i:", sprintf("%04X", $r->[$lower]->start), "[$lower] .. ", sprintf("%04X", $r->[$i]->start), "[$i] .. ", sprintf("%04X", $r->[$upper]->start), "[$upper]" if main::DEBUG && $to_trace;
3610
3611 if ($code_point <= $r->[$i]->end) {
3612
3613 # Here we have met the upper constraint. We can quit if we
3614 # also meet the lower one.
3615 last if $i == 0 || $r->[$i-1]->end < $code_point;
3616
3617 $upper = $i; # Still too high.
3618
3619 }
3620 else {
3621
3622 # Here, $r[$i]->end < $code_point, so look higher up.
3623 $lower = $i;
3624 }
3625
3626 # Split search domain in half to try again.
3627 my $temp = ($upper + $lower) / 2;
3628
3629 # No point in continuing unless $i changes for next time
3630 # in the loop.
3631 if ($temp == $i) {
3632
3633 # We can't reach the highest element because of the averaging.
3634 # So if one below the upper edge, force it there and try one
3635 # more time.
3636 if ($i == $range_list_size - 2) {
3637
3638 trace "Forcing to upper edge" if main::DEBUG && $to_trace;
3639 $i = $range_list_size - 1;
3640
3641 # Change $lower as well so if fails next time through,
3642 # taking the average will yield the same $i, and we will
3643 # quit with the error message just below.
3644 $lower = $i;
3645 next;
3646 }
3647 Carp::my_carp_bug("$owner_name_of{$addr}Can't find where the range ought to go. No action taken.");
3648 return;
3649 }
3650 $i = $temp;
3651 } # End of while loop
3652
3653 if (main::DEBUG && $to_trace) {
3654 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i;
3655 trace "i= [ $i ]", $r->[$i];
3656 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < $range_list_size - 1;
3657 }
3658
3659 # Here we have found the offset. Cache it as a starting point for the
3660 # next call.
3661 $_search_ranges_cache{$addr} = $i;
3662 return $i;
3663 }
3664
3665 sub _add_delete {
3666 # Add, replace or delete ranges to or from a list. The $type
3667 # parameter gives which:
3668 # '+' => insert or replace a range, returning a list of any changed
3669 # ranges.
3670 # '-' => delete a range, returning a list of any deleted ranges.
3671 #
3672 # The next three parameters give respectively the start, end, and
3673 # value associated with the range. 'value' should be null unless the
3674 # operation is '+';
3675 #
3676 # The range list is kept sorted so that the range with the lowest
3677 # starting position is first in the list, and generally, adjacent
c1739a4a 3678 # ranges with the same values are merged into a single larger one (see
99870f4d
KW
3679 # exceptions below).
3680 #
c1739a4a 3681 # There are more parameters; all are key => value pairs:
99870f4d
KW
3682 # Type gives the type of the value. It is only valid for '+'.
3683 # All ranges have types; if this parameter is omitted, 0 is
3684 # assumed. Ranges with type 0 are assumed to obey the
3685 # Unicode rules for casing, etc; ranges with other types are
3686 # not. Otherwise, the type is arbitrary, for the caller's
3687 # convenience, and looked at only by this routine to keep
3688 # adjacent ranges of different types from being merged into
3689 # a single larger range, and when Replace =>
3690 # $IF_NOT_EQUIVALENT is specified (see just below).
3691 # Replace determines what to do if the range list already contains
3692 # ranges which coincide with all or portions of the input
3693 # range. It is only valid for '+':
3694 # => $NO means that the new value is not to replace
3695 # any existing ones, but any empty gaps of the
3696 # range list coinciding with the input range
3697 # will be filled in with the new value.
3698 # => $UNCONDITIONALLY means to replace the existing values with
3699 # this one unconditionally. However, if the
3700 # new and old values are identical, the
3701 # replacement is skipped to save cycles
3702 # => $IF_NOT_EQUIVALENT means to replace the existing values
d59563d0 3703 # (the default) with this one if they are not equivalent.
99870f4d 3704 # Ranges are equivalent if their types are the
c1739a4a 3705 # same, and they are the same string; or if
99870f4d
KW
3706 # both are type 0 ranges, if their Unicode
3707 # standard forms are identical. In this last
3708 # case, the routine chooses the more "modern"
3709 # one to use. This is because some of the
3710 # older files are formatted with values that
3711 # are, for example, ALL CAPs, whereas the
3712 # derived files have a more modern style,
3713 # which looks better. By looking for this
3714 # style when the pre-existing and replacement
3715 # standard forms are the same, we can move to
3716 # the modern style
9470941f 3717 # => $MULTIPLE_BEFORE means that if this range duplicates an
99870f4d
KW
3718 # existing one, but has a different value,
3719 # don't replace the existing one, but insert
3720 # this, one so that the same range can occur
53d84487
KW
3721 # multiple times. They are stored LIFO, so
3722 # that the final one inserted is the first one
3723 # returned in an ordered search of the table.
6901521e
KW
3724 # If this is an exact duplicate, including the
3725 # value, the original will be moved to be
3726 # first, before any other duplicate ranges
3727 # with different values.
7f4b1e25
KW
3728 # => $MULTIPLE_AFTER is like $MULTIPLE_BEFORE, but is stored
3729 # FIFO, so that this one is inserted after all
6901521e
KW
3730 # others that currently exist. If this is an
3731 # exact duplicate, including value, of an
3732 # existing range, this one is discarded
3733 # (leaving the existing one in its original,
3734 # higher priority position
99870f4d
KW
3735 # => anything else is the same as => $IF_NOT_EQUIVALENT
3736 #
c1739a4a
KW
3737 # "same value" means identical for non-type-0 ranges, and it means
3738 # having the same standard forms for type-0 ranges.
99870f4d
KW
3739
3740 return Carp::carp_too_few_args(\@_, 5) if main::DEBUG && @_ < 5;
3741
3742 my $self = shift;
3743 my $operation = shift; # '+' for add/replace; '-' for delete;
3744 my $start = shift;
3745 my $end = shift;
3746 my $value = shift;
3747
3748 my %args = @_;
3749
3750 $value = "" if not defined $value; # warning: $value can be "0"
3751
3752 my $replace = delete $args{'Replace'};
3753 $replace = $IF_NOT_EQUIVALENT unless defined $replace;
3754
3755 my $type = delete $args{'Type'};
3756 $type = 0 unless defined