This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
mktables: Generate native code-point tables
[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
c4019d52
KW
367# flag. This causes the tables to expand so there is one entry for each
368# non-algorithmically named code point giving, currently its name, and its
369# graphic representation if printable (and you have a font that knows about
370# it). This makes it easier to see what the particular code points are in
371# each output table. The tables are usable, but because they don't have
372# ranges (for the most part), a Perl using them will run slower. Non-named
373# code points are annotated with a description of their status, and contiguous
374# ones with the same description will be output as a range rather than
375# individually. Algorithmically named characters are also output as ranges,
376# except when there are just a few contiguous ones.
377#
99870f4d
KW
378# FUTURE ISSUES
379#
380# The program would break if Unicode were to change its names so that
381# interior white space, underscores, or dashes differences were significant
382# within property and property value names.
383#
384# It might be easier to use the xml versions of the UCD if this program ever
385# would need heavy revision, and the ability to handle old versions was not
386# required.
387#
388# There is the potential for name collisions, in that Perl has chosen names
389# that Unicode could decide it also likes. There have been such collisions in
390# the past, with mostly Perl deciding to adopt the Unicode definition of the
391# name. However in the 5.2 Unicode beta testing, there were a number of such
392# collisions, which were withdrawn before the final release, because of Perl's
393# and other's protests. These all involved new properties which began with
394# 'Is'. Based on the protests, Unicode is unlikely to try that again. Also,
395# many of the Perl-defined synonyms, like Any, Word, etc, are listed in a
396# Unicode document, so they are unlikely to be used by Unicode for another
397# purpose. However, they might try something beginning with 'In', or use any
398# of the other Perl-defined properties. This program will warn you of name
399# collisions, and refuse to generate tables with them, but manual intervention
400# will be required in this event. One scheme that could be implemented, if
401# necessary, would be to have this program generate another file, or add a
402# field to mktables.lst that gives the date of first definition of a property.
403# Each new release of Unicode would use that file as a basis for the next
404# iteration. And the Perl synonym addition code could sort based on the age
405# of the property, so older properties get priority, and newer ones that clash
406# would be refused; hence existing code would not be impacted, and some other
407# synonym would have to be used for the new property. This is ugly, and
408# manual intervention would certainly be easier to do in the short run; lets
409# hope it never comes to this.
678f13d5 410#
99870f4d
KW
411# A NOTE ON UNIHAN
412#
413# This program can generate tables from the Unihan database. But it doesn't
414# by default, letting the CPAN module Unicode::Unihan handle them. Prior to
415# version 5.2, this database was in a single file, Unihan.txt. In 5.2 the
416# database was split into 8 different files, all beginning with the letters
417# 'Unihan'. This program will read those file(s) if present, but it needs to
418# know which of the many properties in the file(s) should have tables created
419# for them. It will create tables for any properties listed in
420# PropertyAliases.txt and PropValueAliases.txt, plus any listed in the
421# @cjk_properties array and the @cjk_property_values array. Thus, if a
422# property you want is not in those files of the release you are building
423# against, you must add it to those two arrays. Starting in 4.0, the
424# Unicode_Radical_Stroke was listed in those files, so if the Unihan database
425# is present in the directory, a table will be generated for that property.
426# In 5.2, several more properties were added. For your convenience, the two
5f7264c7 427# arrays are initialized with all the 6.0 listed properties that are also in
99870f4d
KW
428# earlier releases. But these are commented out. You can just uncomment the
429# ones you want, or use them as a template for adding entries for other
430# properties.
431#
432# You may need to adjust the entries to suit your purposes. setup_unihan(),
433# and filter_unihan_line() are the functions where this is done. This program
434# already does some adjusting to make the lines look more like the rest of the
435# Unicode DB; You can see what that is in filter_unihan_line()
436#
437# There is a bug in the 3.2 data file in which some values for the
438# kPrimaryNumeric property have commas and an unexpected comment. A filter
439# could be added for these; or for a particular installation, the Unihan.txt
440# file could be edited to fix them.
99870f4d 441#
678f13d5
KW
442# HOW TO ADD A FILE TO BE PROCESSED
443#
444# A new file from Unicode needs to have an object constructed for it in
445# @input_file_objects, probably at the end or at the end of the extracted
446# ones. The program should warn you if its name will clash with others on
447# restrictive file systems, like DOS. If so, figure out a better name, and
448# add lines to the README.perl file giving that. If the file is a character
232ed87f 449# property, it should be in the format that Unicode has implicitly
678f13d5
KW
450# standardized for such files for the more recently introduced ones.
451# If so, the Input_file constructor for @input_file_objects can just be the
452# file name and release it first appeared in. If not, then it should be
453# possible to construct an each_line_handler() to massage the line into the
454# standardized form.
455#
456# For non-character properties, more code will be needed. You can look at
457# the existing entries for clues.
458#
459# UNICODE VERSIONS NOTES
460#
461# The Unicode UCD has had a number of errors in it over the versions. And
462# these remain, by policy, in the standard for that version. Therefore it is
463# risky to correct them, because code may be expecting the error. So this
464# program doesn't generally make changes, unless the error breaks the Perl
465# core. As an example, some versions of 2.1.x Jamo.txt have the wrong value
466# for U+1105, which causes real problems for the algorithms for Jamo
467# calculations, so it is changed here.
468#
469# But it isn't so clear cut as to what to do about concepts that are
470# introduced in a later release; should they extend back to earlier releases
471# where the concept just didn't exist? It was easier to do this than to not,
472# so that's what was done. For example, the default value for code points not
473# in the files for various properties was probably undefined until changed by
474# some version. No_Block for blocks is such an example. This program will
475# assign No_Block even in Unicode versions that didn't have it. This has the
476# benefit that code being written doesn't have to special case earlier
477# versions; and the detriment that it doesn't match the Standard precisely for
478# the affected versions.
479#
480# Here are some observations about some of the issues in early versions:
481#
232ed87f
KW
482# Prior to version 3.0, there were 3 character decompositions. These are not
483# handled by Unicode::Normalize, nor will it compile when presented a version
484# that has them. However, you can trivially get it to compile by simply
485# ignoring those decompositions, by changing the croak to a carp. At the time
486# of this writing, the line (in cpan/Unicode-Normalize/mkheader) reads
487#
488# croak("Weird Canonical Decomposition of U+$h");
489#
490# Simply change to a carp. It will compile, but will not know about any three
491# character decomposition.
492
493# The number of code points in \p{alpha=True} halved in 2.1.9. It turns out
494# that the reason is that the CJK block starting at 4E00 was removed from
495# PropList, and was not put back in until 3.1.0. The Perl extension (the
496# single property name \p{alpha}) has the correct values. But the compound
497# form is simply not generated until 3.1, as it can be argued that prior to
498# this release, this was not an official property. The comments for
499# filter_old_style_proplist() give more details.
678f13d5
KW
500#
501# Unicode introduced the synonym Space for White_Space in 4.1. Perl has
502# always had a \p{Space}. In release 3.2 only, they are not synonymous. The
503# reason is that 3.2 introduced U+205F=medium math space, which was not
504# classed as white space, but Perl figured out that it should have been. 4.0
505# reclassified it correctly.
506#
507# Another change between 3.2 and 4.0 is the CCC property value ATBL. In 3.2
232ed87f
KW
508# this was erroneously a synonym for 202 (it should be 200). In 4.0, ATB
509# became 202, and ATBL was left with no code points, as all the ones that
510# mapped to 202 stayed mapped to 202. Thus if your program used the numeric
511# name for the class, it would not have been affected, but if it used the
512# mnemonic, it would have been.
678f13d5
KW
513#
514# \p{Script=Hrkt} (Katakana_Or_Hiragana) came in 4.0.1. Before that code
515# points which eventually came to have this script property value, instead
516# mapped to "Unknown". But in the next release all these code points were
517# moved to \p{sc=common} instead.
99870f4d
KW
518#
519# The default for missing code points for BidiClass is complicated. Starting
520# in 3.1.1, the derived file DBidiClass.txt handles this, but this program
521# tries to do the best it can for earlier releases. It is done in
522# process_PropertyAliases()
523#
232ed87f
KW
524# In version 2.1.2, the entry in UnicodeData.txt:
525# 0275;LATIN SMALL LETTER BARRED O;Ll;0;L;;;;;N;;;;019F;
526# should instead be
527# 0275;LATIN SMALL LETTER BARRED O;Ll;0;L;;;;;N;;;019F;;019F
528# Without this change, there are casing problems for this character.
529#
99870f4d
KW
530##############################################################################
531
532my $UNDEF = ':UNDEF:'; # String to print out for undefined values in tracing
533 # and errors
534my $MAX_LINE_WIDTH = 78;
535
536# Debugging aid to skip most files so as to not be distracted by them when
537# concentrating on the ones being debugged. Add
538# non_skip => 1,
539# to the constructor for those files you want processed when you set this.
540# Files with a first version number of 0 are special: they are always
c12f2655
KW
541# processed regardless of the state of this flag. Generally, Jamo.txt and
542# UnicodeData.txt must not be skipped if you want this program to not die
543# before normal completion.
99870f4d
KW
544my $debug_skip = 0;
545
e9c4b4f8
KW
546
547# Normally these are suppressed.
548my $write_Unicode_deprecated_tables = 0;
549
99870f4d
KW
550# Set to 1 to enable tracing.
551our $to_trace = 0;
552
553{ # Closure for trace: debugging aid
554 my $print_caller = 1; # ? Include calling subroutine name
555 my $main_with_colon = 'main::';
556 my $main_colon_length = length($main_with_colon);
557
558 sub trace {
559 return unless $to_trace; # Do nothing if global flag not set
560
561 my @input = @_;
562
563 local $DB::trace = 0;
564 $DB::trace = 0; # Quiet 'used only once' message
565
566 my $line_number;
567
568 # Loop looking up the stack to get the first non-trace caller
569 my $caller_line;
570 my $caller_name;
571 my $i = 0;
572 do {
573 $line_number = $caller_line;
574 (my $pkg, my $file, $caller_line, my $caller) = caller $i++;
575 $caller = $main_with_colon unless defined $caller;
576
577 $caller_name = $caller;
578
579 # get rid of pkg
580 $caller_name =~ s/.*:://;
581 if (substr($caller_name, 0, $main_colon_length)
582 eq $main_with_colon)
583 {
584 $caller_name = substr($caller_name, $main_colon_length);
585 }
586
587 } until ($caller_name ne 'trace');
588
589 # If the stack was empty, we were called from the top level
590 $caller_name = 'main' if ($caller_name eq ""
591 || $caller_name eq 'trace');
592
593 my $output = "";
594 foreach my $string (@input) {
595 #print STDERR __LINE__, ": ", join ", ", @input, "\n";
596 if (ref $string eq 'ARRAY' || ref $string eq 'HASH') {
597 $output .= simple_dumper($string);
598 }
599 else {
600 $string = "$string" if ref $string;
601 $string = $UNDEF unless defined $string;
602 chomp $string;
603 $string = '""' if $string eq "";
604 $output .= " " if $output ne ""
605 && $string ne ""
606 && substr($output, -1, 1) ne " "
607 && substr($string, 0, 1) ne " ";
608 $output .= $string;
609 }
610 }
611
99f78760
KW
612 print STDERR sprintf "%4d: ", $line_number if defined $line_number;
613 print STDERR "$caller_name: " if $print_caller;
99870f4d
KW
614 print STDERR $output, "\n";
615 return;
616 }
617}
618
619# This is for a rarely used development feature that allows you to compare two
620# versions of the Unicode standard without having to deal with changes caused
c12f2655
KW
621# by the code points introduced in the later version. Change the 0 to a
622# string containing a SINGLE dotted Unicode release number (e.g. "2.1"). Only
623# code points introduced in that release and earlier will be used; later ones
624# are thrown away. You use the version number of the earliest one you want to
625# compare; then run this program on directory structures containing each
626# release, and compare the outputs. These outputs will therefore include only
627# the code points common to both releases, and you can see the changes caused
628# just by the underlying release semantic changes. For versions earlier than
629# 3.2, you must copy a version of DAge.txt into the directory.
630my $string_compare_versions = DEBUG && 0; # e.g., "2.1";
99870f4d
KW
631my $compare_versions = DEBUG
632 && $string_compare_versions
633 && pack "C*", split /\./, $string_compare_versions;
634
635sub uniques {
636 # Returns non-duplicated input values. From "Perl Best Practices:
637 # Encapsulated Cleverness". p. 455 in first edition.
638
639 my %seen;
0e407844
NC
640 # Arguably this breaks encapsulation, if the goal is to permit multiple
641 # distinct objects to stringify to the same value, and be interchangeable.
642 # However, for this program, no two objects stringify identically, and all
643 # lists passed to this function are either objects or strings. So this
644 # doesn't affect correctness, but it does give a couple of percent speedup.
645 no overloading;
99870f4d
KW
646 return grep { ! $seen{$_}++ } @_;
647}
648
649$0 = File::Spec->canonpath($0);
650
651my $make_test_script = 0; # ? Should we output a test script
6b5ab373 652my $make_norm_test_script = 0; # ? Should we output a normalization test script
99870f4d
KW
653my $write_unchanged_files = 0; # ? Should we update the output files even if
654 # we don't think they have changed
655my $use_directory = ""; # ? Should we chdir somewhere.
656my $pod_directory; # input directory to store the pod file.
657my $pod_file = 'perluniprops';
658my $t_path; # Path to the .t test file
659my $file_list = 'mktables.lst'; # File to store input and output file names.
660 # This is used to speed up the build, by not
661 # executing the main body of the program if
662 # nothing on the list has changed since the
663 # previous build
664my $make_list = 1; # ? Should we write $file_list. Set to always
665 # make a list so that when the pumpking is
666 # preparing a release, s/he won't have to do
667 # special things
668my $glob_list = 0; # ? Should we try to include unknown .txt files
669 # in the input.
bd9ebcfd
KW
670my $output_range_counts = $debugging_build; # ? Should we include the number
671 # of code points in ranges in
672 # the output
558712cf 673my $annotate = 0; # ? Should character names be in the output
9ef2b94f 674
99870f4d
KW
675# Verbosity levels; 0 is quiet
676my $NORMAL_VERBOSITY = 1;
677my $PROGRESS = 2;
678my $VERBOSE = 3;
679
680my $verbosity = $NORMAL_VERBOSITY;
681
682# Process arguments
683while (@ARGV) {
cf25bb62
JH
684 my $arg = shift @ARGV;
685 if ($arg eq '-v') {
99870f4d
KW
686 $verbosity = $VERBOSE;
687 }
688 elsif ($arg eq '-p') {
689 $verbosity = $PROGRESS;
690 $| = 1; # Flush buffers as we go.
691 }
692 elsif ($arg eq '-q') {
693 $verbosity = 0;
694 }
695 elsif ($arg eq '-w') {
696 $write_unchanged_files = 1; # update the files even if havent changed
697 }
698 elsif ($arg eq '-check') {
6ae7e459
YO
699 my $this = shift @ARGV;
700 my $ok = shift @ARGV;
701 if ($this ne $ok) {
702 print "Skipping as check params are not the same.\n";
703 exit(0);
704 }
00a8df5c 705 }
99870f4d
KW
706 elsif ($arg eq '-P' && defined ($pod_directory = shift)) {
707 -d $pod_directory or croak "Directory '$pod_directory' doesn't exist";
708 }
3df51b85
KW
709 elsif ($arg eq '-maketest' || ($arg eq '-T' && defined ($t_path = shift)))
710 {
99870f4d 711 $make_test_script = 1;
99870f4d 712 }
6b5ab373
KW
713 elsif ($arg eq '-makenormtest')
714 {
715 $make_norm_test_script = 1;
716 }
99870f4d
KW
717 elsif ($arg eq '-makelist') {
718 $make_list = 1;
719 }
720 elsif ($arg eq '-C' && defined ($use_directory = shift)) {
721 -d $use_directory or croak "Unknown directory '$use_directory'";
722 }
723 elsif ($arg eq '-L') {
724
725 # Existence not tested until have chdir'd
726 $file_list = shift;
727 }
728 elsif ($arg eq '-globlist') {
729 $glob_list = 1;
730 }
731 elsif ($arg eq '-c') {
732 $output_range_counts = ! $output_range_counts
733 }
b4a0206c 734 elsif ($arg eq '-annotate') {
558712cf 735 $annotate = 1;
bd9ebcfd
KW
736 $debugging_build = 1;
737 $output_range_counts = 1;
9ef2b94f 738 }
99870f4d
KW
739 else {
740 my $with_c = 'with';
741 $with_c .= 'out' if $output_range_counts; # Complements the state
742 croak <<END;
743usage: $0 [-c|-p|-q|-v|-w] [-C dir] [-L filelist] [ -P pod_dir ]
744 [ -T test_file_path ] [-globlist] [-makelist] [-maketest]
745 [-check A B ]
746 -c : Output comments $with_c number of code points in ranges
747 -q : Quiet Mode: Only output serious warnings.
748 -p : Set verbosity level to normal plus show progress.
749 -v : Set Verbosity level high: Show progress and non-serious
750 warnings
751 -w : Write files regardless
752 -C dir : Change to this directory before proceeding. All relative paths
753 except those specified by the -P and -T options will be done
754 with respect to this directory.
755 -P dir : Output $pod_file file to directory 'dir'.
3df51b85 756 -T path : Create a test script as 'path'; overrides -maketest
99870f4d
KW
757 -L filelist : Use alternate 'filelist' instead of standard one
758 -globlist : Take as input all non-Test *.txt files in current and sub
759 directories
3df51b85
KW
760 -maketest : Make test script 'TestProp.pl' in current (or -C directory),
761 overrides -T
99870f4d 762 -makelist : Rewrite the file list $file_list based on current setup
b4a0206c 763 -annotate : Output an annotation for each character in the table files;
c4019d52 764 useful for debugging mktables, looking at diffs; but is slow,
b318e5e5
KW
765 memory intensive; resulting tables are usable but are slow and
766 very large (and currently fail the Unicode::UCD.t tests).
99870f4d
KW
767 -check A B : Executes $0 only if A and B are the same
768END
769 }
770}
771
772# Stores the most-recently changed file. If none have changed, can skip the
773# build
aeab6150 774my $most_recent = (stat $0)[9]; # Do this before the chdir!
99870f4d
KW
775
776# Change directories now, because need to read 'version' early.
777if ($use_directory) {
3df51b85 778 if ($pod_directory && ! File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d
KW
779 $pod_directory = File::Spec->rel2abs($pod_directory);
780 }
3df51b85 781 if ($t_path && ! File::Spec->file_name_is_absolute($t_path)) {
99870f4d 782 $t_path = File::Spec->rel2abs($t_path);
00a8df5c 783 }
99870f4d 784 chdir $use_directory or croak "Failed to chdir to '$use_directory':$!";
3df51b85 785 if ($pod_directory && File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d 786 $pod_directory = File::Spec->abs2rel($pod_directory);
02b1aeec 787 }
3df51b85 788 if ($t_path && File::Spec->file_name_is_absolute($t_path)) {
99870f4d 789 $t_path = File::Spec->abs2rel($t_path);
02b1aeec 790 }
00a8df5c
YO
791}
792
99870f4d
KW
793# Get Unicode version into regular and v-string. This is done now because
794# various tables below get populated based on it. These tables are populated
795# here to be near the top of the file, and so easily seeable by those needing
796# to modify things.
797open my $VERSION, "<", "version"
798 or croak "$0: can't open required file 'version': $!\n";
799my $string_version = <$VERSION>;
800close $VERSION;
801chomp $string_version;
802my $v_version = pack "C*", split /\./, $string_version; # v string
803
804# The following are the complete names of properties with property values that
805# are known to not match any code points in some versions of Unicode, but that
806# may change in the future so they should be matchable, hence an empty file is
807# generated for them.
808my @tables_that_may_be_empty = (
809 'Joining_Type=Left_Joining',
810 );
811push @tables_that_may_be_empty, 'Script=Common' if $v_version le v4.0.1;
812push @tables_that_may_be_empty, 'Title' if $v_version lt v2.0.0;
813push @tables_that_may_be_empty, 'Script=Katakana_Or_Hiragana'
814 if $v_version ge v4.1.0;
82aed44a
KW
815push @tables_that_may_be_empty, 'Script_Extensions=Katakana_Or_Hiragana'
816 if $v_version ge v6.0.0;
f583b44c
KW
817push @tables_that_may_be_empty, 'Grapheme_Cluster_Break=Prepend'
818 if $v_version ge v6.1.0;
1e958ea9
KW
819push @tables_that_may_be_empty, 'Canonical_Combining_Class=CCC133'
820 if $v_version ge v6.2.0;
99870f4d
KW
821
822# The lists below are hashes, so the key is the item in the list, and the
823# value is the reason why it is in the list. This makes generation of
824# documentation easier.
825
826my %why_suppressed; # No file generated for these.
827
828# Files aren't generated for empty extraneous properties. This is arguable.
829# Extraneous properties generally come about because a property is no longer
830# used in a newer version of Unicode. If we generated a file without code
831# points, programs that used to work on that property will still execute
832# without errors. It just won't ever match (or will always match, with \P{}).
833# This means that the logic is now likely wrong. I (khw) think its better to
834# find this out by getting an error message. Just move them to the table
835# above to change this behavior
836my %why_suppress_if_empty_warn_if_not = (
837
838 # It is the only property that has ever officially been removed from the
839 # Standard. The database never contained any code points for it.
840 'Special_Case_Condition' => 'Obsolete',
841
842 # Apparently never official, but there were code points in some versions of
843 # old-style PropList.txt
844 'Non_Break' => 'Obsolete',
845);
846
847# These would normally go in the warn table just above, but they were changed
848# a long time before this program was written, so warnings about them are
849# moot.
850if ($v_version gt v3.2.0) {
851 push @tables_that_may_be_empty,
852 'Canonical_Combining_Class=Attached_Below_Left'
853}
854
5f7264c7 855# These are listed in the Property aliases file in 6.0, but Unihan is ignored
99870f4d
KW
856# unless explicitly added.
857if ($v_version ge v5.2.0) {
858 my $unihan = 'Unihan; remove from list if using Unihan';
ea25a9b2 859 foreach my $table (qw (
99870f4d
KW
860 kAccountingNumeric
861 kOtherNumeric
862 kPrimaryNumeric
863 kCompatibilityVariant
864 kIICore
865 kIRG_GSource
866 kIRG_HSource
867 kIRG_JSource
868 kIRG_KPSource
869 kIRG_MSource
870 kIRG_KSource
871 kIRG_TSource
872 kIRG_USource
873 kIRG_VSource
874 kRSUnicode
ea25a9b2 875 ))
99870f4d
KW
876 {
877 $why_suppress_if_empty_warn_if_not{$table} = $unihan;
878 }
ca12659b
NC
879}
880
272501f6
KW
881# Enum values for to_output_map() method in the Map_Table package.
882my $EXTERNAL_MAP = 1;
883my $INTERNAL_MAP = 2;
ce712c88 884my $OUTPUT_ADJUSTED = 3;
272501f6 885
fcf1973c
KW
886# To override computed values for writing the map tables for these properties.
887# The default for enum map tables is to write them out, so that the Unicode
888# .txt files can be removed, but all the data to compute any property value
889# for any code point is available in a more compact form.
890my %global_to_output_map = (
891 # Needed by UCD.pm, but don't want to publicize that it exists, so won't
c12f2655
KW
892 # get stuck supporting it if things change. Since it is a STRING
893 # property, it normally would be listed in the pod, but INTERNAL_MAP
894 # suppresses that.
fcf1973c
KW
895 Unicode_1_Name => $INTERNAL_MAP,
896
897 Present_In => 0, # Suppress, as easily computed from Age
74cd47d0
KW
898 Block => (NON_ASCII_PLATFORM) ? 1 : 0, # Suppress, as Blocks.txt is
899 # retained, but needed for
900 # non-ASCII
53d34b6c
KW
901
902 # Suppress, as mapping can be found instead from the
903 # Perl_Decomposition_Mapping file
904 Decomposition_Type => 0,
fcf1973c
KW
905);
906
99870f4d 907# Properties that this program ignores.
230e0c16
KW
908my @unimplemented_properties;
909
910# With this release, it is automatically handled if the Unihan db is
911# downloaded
912push @unimplemented_properties, 'Unicode_Radical_Stroke' if $v_version le v5.2.0;
d73e5302 913
99870f4d
KW
914# There are several types of obsolete properties defined by Unicode. These
915# must be hand-edited for every new Unicode release.
916my %why_deprecated; # Generates a deprecated warning message if used.
917my %why_stabilized; # Documentation only
918my %why_obsolete; # Documentation only
919
920{ # Closure
921 my $simple = 'Perl uses the more complete version of this property';
922 my $unihan = 'Unihan properties are by default not enabled in the Perl core. Instead use CPAN: Unicode::Unihan';
923
924 my $other_properties = 'other properties';
925 my $contributory = "Used by Unicode internally for generating $other_properties and not intended to be used stand-alone";
5d294d41 926 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
927
928 %why_deprecated = (
5f7264c7 929 'Grapheme_Link' => 'Deprecated by Unicode: Duplicates ccc=vr (Canonical_Combining_Class=Virama)',
99870f4d
KW
930 'Jamo_Short_Name' => $contributory,
931 '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',
932 'Other_Alphabetic' => $contributory,
933 'Other_Default_Ignorable_Code_Point' => $contributory,
934 'Other_Grapheme_Extend' => $contributory,
935 'Other_ID_Continue' => $contributory,
936 'Other_ID_Start' => $contributory,
937 'Other_Lowercase' => $contributory,
938 'Other_Math' => $contributory,
939 'Other_Uppercase' => $contributory,
e22aaf5c
KW
940 'Expands_On_NFC' => $why_no_expand,
941 'Expands_On_NFD' => $why_no_expand,
942 'Expands_On_NFKC' => $why_no_expand,
943 'Expands_On_NFKD' => $why_no_expand,
99870f4d
KW
944 );
945
946 %why_suppressed = (
5f7264c7 947 # There is a lib/unicore/Decomposition.pl (used by Normalize.pm) which
99870f4d
KW
948 # contains the same information, but without the algorithmically
949 # determinable Hangul syllables'. This file is not published, so it's
950 # existence is not noted in the comment.
e0b29447 951 'Decomposition_Mapping' => 'Accessible via Unicode::Normalize or Unicode::UCD::prop_invmap()',
99870f4d 952
3111abc0
KW
953 'Indic_Matra_Category' => "Provisional",
954 'Indic_Syllabic_Category' => "Provisional",
955
5f8d1a89
KW
956 # Don't suppress ISO_Comment, as otherwise special handling is needed
957 # to differentiate between it and gc=c, which can be written as 'isc',
958 # which is the same characters as ISO_Comment's short name.
99870f4d 959
fbb93542 960 'Name' => "Accessible via \\N{...} or 'use charnames;' or Unicode::UCD::prop_invmap()",
e0b29447
KW
961
962 'Simple_Case_Folding' => "$simple. Can access this through Unicode::UCD::casefold or Unicode::UCD::prop_invmap()",
963 'Simple_Lowercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
964 'Simple_Titlecase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
965 'Simple_Uppercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
99870f4d 966
5f7264c7 967 FC_NFKC_Closure => 'Supplanted in usage by NFKC_Casefold; otherwise not useful',
99870f4d
KW
968 );
969
1704a0ea
KW
970 foreach my $property (
971
972 # The following are suppressed because they were made contributory
973 # or deprecated by Unicode before Perl ever thought about
974 # supporting them.
975 'Jamo_Short_Name',
976 'Grapheme_Link',
977 'Expands_On_NFC',
978 'Expands_On_NFD',
979 'Expands_On_NFKC',
980 'Expands_On_NFKD',
981
982 # The following are suppressed because they have been marked
983 # as deprecated for a sufficient amount of time
984 'Other_Alphabetic',
985 'Other_Default_Ignorable_Code_Point',
986 'Other_Grapheme_Extend',
987 'Other_ID_Continue',
988 'Other_ID_Start',
989 'Other_Lowercase',
990 'Other_Math',
991 'Other_Uppercase',
e22aaf5c 992 ) {
99870f4d
KW
993 $why_suppressed{$property} = $why_deprecated{$property};
994 }
cf25bb62 995
99870f4d
KW
996 # Customize the message for all the 'Other_' properties
997 foreach my $property (keys %why_deprecated) {
998 next if (my $main_property = $property) !~ s/^Other_//;
999 $why_deprecated{$property} =~ s/$other_properties/the $main_property property (which should be used instead)/;
1000 }
1001}
1002
e9c4b4f8
KW
1003if ($write_Unicode_deprecated_tables) {
1004 foreach my $property (keys %why_suppressed) {
1005 delete $why_suppressed{$property} if $property =~
1006 / ^ Other | Grapheme /x;
1007 }
1008}
1009
99870f4d
KW
1010if ($v_version ge 4.0.0) {
1011 $why_stabilized{'Hyphen'} = 'Use the Line_Break property instead; see www.unicode.org/reports/tr14';
5f7264c7
KW
1012 if ($v_version ge 6.0.0) {
1013 $why_deprecated{'Hyphen'} = 'Supplanted by Line_Break property values; see www.unicode.org/reports/tr14';
1014 }
99870f4d 1015}
5f7264c7 1016if ($v_version ge 5.2.0 && $v_version lt 6.0.0) {
99870f4d 1017 $why_obsolete{'ISO_Comment'} = 'Code points for it have been removed';
5f7264c7 1018 if ($v_version ge 6.0.0) {
63f74647 1019 $why_deprecated{'ISO_Comment'} = 'No longer needed for Unicode\'s internal chart generation; otherwise not useful, and code points for it have been removed';
5f7264c7 1020 }
99870f4d
KW
1021}
1022
1023# Probably obsolete forever
1024if ($v_version ge v4.1.0) {
82aed44a
KW
1025 $why_suppressed{'Script=Katakana_Or_Hiragana'} = 'Obsolete. All code points previously matched by this have been moved to "Script=Common".';
1026}
1027if ($v_version ge v6.0.0) {
caa75395 1028 $why_suppressed{'Script=Katakana_Or_Hiragana'} .= ' Consider instead using "Script_Extensions=Katakana" or "Script_Extensions=Hiragana" (or both)';
2b352efd 1029 $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
1030}
1031
1032# This program can create files for enumerated-like properties, such as
1033# 'Numeric_Type'. This file would be the same format as for a string
1034# property, with a mapping from code point to its value, so you could look up,
1035# for example, the script a code point is in. But no one so far wants this
1036# mapping, or they have found another way to get it since this is a new
1037# feature. So no file is generated except if it is in this list.
1038my @output_mapped_properties = split "\n", <<END;
1039END
1040
c12f2655
KW
1041# If you are using the Unihan database in a Unicode version before 5.2, you
1042# need to add the properties that you want to extract from it to this table.
1043# For your convenience, the properties in the 6.0 PropertyAliases.txt file are
1044# listed, commented out
99870f4d
KW
1045my @cjk_properties = split "\n", <<'END';
1046#cjkAccountingNumeric; kAccountingNumeric
1047#cjkOtherNumeric; kOtherNumeric
1048#cjkPrimaryNumeric; kPrimaryNumeric
1049#cjkCompatibilityVariant; kCompatibilityVariant
1050#cjkIICore ; kIICore
1051#cjkIRG_GSource; kIRG_GSource
1052#cjkIRG_HSource; kIRG_HSource
1053#cjkIRG_JSource; kIRG_JSource
1054#cjkIRG_KPSource; kIRG_KPSource
1055#cjkIRG_KSource; kIRG_KSource
1056#cjkIRG_TSource; kIRG_TSource
1057#cjkIRG_USource; kIRG_USource
1058#cjkIRG_VSource; kIRG_VSource
1059#cjkRSUnicode; kRSUnicode ; Unicode_Radical_Stroke; URS
1060END
1061
1062# Similarly for the property values. For your convenience, the lines in the
5f7264c7 1063# 6.0 PropertyAliases.txt file are listed. Just remove the first BUT NOT both
c12f2655 1064# '#' marks (for Unicode versions before 5.2)
99870f4d
KW
1065my @cjk_property_values = split "\n", <<'END';
1066## @missing: 0000..10FFFF; cjkAccountingNumeric; NaN
1067## @missing: 0000..10FFFF; cjkCompatibilityVariant; <code point>
1068## @missing: 0000..10FFFF; cjkIICore; <none>
1069## @missing: 0000..10FFFF; cjkIRG_GSource; <none>
1070## @missing: 0000..10FFFF; cjkIRG_HSource; <none>
1071## @missing: 0000..10FFFF; cjkIRG_JSource; <none>
1072## @missing: 0000..10FFFF; cjkIRG_KPSource; <none>
1073## @missing: 0000..10FFFF; cjkIRG_KSource; <none>
1074## @missing: 0000..10FFFF; cjkIRG_TSource; <none>
1075## @missing: 0000..10FFFF; cjkIRG_USource; <none>
1076## @missing: 0000..10FFFF; cjkIRG_VSource; <none>
1077## @missing: 0000..10FFFF; cjkOtherNumeric; NaN
1078## @missing: 0000..10FFFF; cjkPrimaryNumeric; NaN
1079## @missing: 0000..10FFFF; cjkRSUnicode; <none>
1080END
1081
1082# The input files don't list every code point. Those not listed are to be
1083# defaulted to some value. Below are hard-coded what those values are for
1084# non-binary properties as of 5.1. Starting in 5.0, there are
caa75395 1085# machine-parsable comment lines in the files that give the defaults; so this
99870f4d
KW
1086# list shouldn't have to be extended. The claim is that all missing entries
1087# for binary properties will default to 'N'. Unicode tried to change that in
1088# 5.2, but the beta period produced enough protest that they backed off.
1089#
1090# The defaults for the fields that appear in UnicodeData.txt in this hash must
1091# be in the form that it expects. The others may be synonyms.
1092my $CODE_POINT = '<code point>';
1093my %default_mapping = (
1094 Age => "Unassigned",
1095 # Bidi_Class => Complicated; set in code
1096 Bidi_Mirroring_Glyph => "",
1097 Block => 'No_Block',
1098 Canonical_Combining_Class => 0,
1099 Case_Folding => $CODE_POINT,
1100 Decomposition_Mapping => $CODE_POINT,
1101 Decomposition_Type => 'None',
1102 East_Asian_Width => "Neutral",
1103 FC_NFKC_Closure => $CODE_POINT,
1104 General_Category => 'Cn',
1105 Grapheme_Cluster_Break => 'Other',
1106 Hangul_Syllable_Type => 'NA',
1107 ISO_Comment => "",
1108 Jamo_Short_Name => "",
1109 Joining_Group => "No_Joining_Group",
1110 # Joining_Type => Complicated; set in code
1111 kIICore => 'N', # Is converted to binary
1112 #Line_Break => Complicated; set in code
1113 Lowercase_Mapping => $CODE_POINT,
1114 Name => "",
1115 Name_Alias => "",
1116 NFC_QC => 'Yes',
1117 NFD_QC => 'Yes',
1118 NFKC_QC => 'Yes',
1119 NFKD_QC => 'Yes',
1120 Numeric_Type => 'None',
1121 Numeric_Value => 'NaN',
1122 Script => ($v_version le 4.1.0) ? 'Common' : 'Unknown',
1123 Sentence_Break => 'Other',
1124 Simple_Case_Folding => $CODE_POINT,
1125 Simple_Lowercase_Mapping => $CODE_POINT,
1126 Simple_Titlecase_Mapping => $CODE_POINT,
1127 Simple_Uppercase_Mapping => $CODE_POINT,
1128 Titlecase_Mapping => $CODE_POINT,
1129 Unicode_1_Name => "",
1130 Unicode_Radical_Stroke => "",
1131 Uppercase_Mapping => $CODE_POINT,
1132 Word_Break => 'Other',
1133);
1134
232ed87f
KW
1135# Below are files that Unicode furnishes, but this program ignores, and why.
1136# NormalizationCorrections.txt requires some more explanation. It documents
1137# the cumulative fixes to erroneous normalizations in earlier Unicode
1138# versions. Its main purpose is so that someone running on an earlier version
1139# can use this file to override what got published in that earlier release.
1140# It would be easy for mktables to read and handle this file. But all the
1141# corrections in it should already be in the other files for the release it
1142# is. To get it to actually mean something useful, someone would have to be
1143# using an earlier Unicode release, and copy it to the files for that release
1144# and recomplile. So far there has been no demand to do that, so this hasn't
1145# been implemented.
99870f4d 1146my %ignored_files = (
73ba1144
KW
1147 'CJKRadicals.txt' => 'Maps the kRSUnicode property values to corresponding code points',
1148 'Index.txt' => 'Alphabetical index of Unicode characters',
1149 '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',
1150 'NamesList.txt' => 'Annotated list of characters',
1151 'NormalizationCorrections.txt' => 'Documentation of corrections already incorporated into the Unicode data base',
1152 'Props.txt' => 'Only in very early releases; is a subset of F<PropList.txt> (which is used instead)',
1153 'ReadMe.txt' => 'Documentation',
1154 '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>',
1155 'EmojiSources.txt' => 'Maps certain Unicode code points to their legacy Japanese cell-phone values',
8ee2793f
KW
1156 'USourceData.txt' => 'Documentation of status and cross reference of proposals for encoding by Unicode of Unihan characters',
1157 'USourceData.pdf' => 'Documentation of status and cross reference of proposals for encoding by Unicode of Unihan characters',
73ba1144
KW
1158 'auxiliary/WordBreakTest.html' => 'Documentation of validation tests',
1159 'auxiliary/SentenceBreakTest.html' => 'Documentation of validation tests',
1160 'auxiliary/GraphemeBreakTest.html' => 'Documentation of validation tests',
1161 'auxiliary/LineBreakTest.html' => 'Documentation of validation tests',
99870f4d
KW
1162);
1163
1fec9f60
KW
1164my %skipped_files; # List of files that we skip
1165
678f13d5 1166### End of externally interesting definitions, except for @input_file_objects
99870f4d
KW
1167
1168my $HEADER=<<"EOF";
1169# !!!!!!! DO NOT EDIT THIS FILE !!!!!!!
3df51b85
KW
1170# This file is machine-generated by $0 from the Unicode
1171# database, Version $string_version. Any changes made here will be lost!
cf25bb62
JH
1172EOF
1173
126c3d4e 1174my $INTERNAL_ONLY_HEADER = <<"EOF";
99870f4d
KW
1175
1176# !!!!!!! INTERNAL PERL USE ONLY !!!!!!!
fac53429
KW
1177# This file is for internal use by core Perl only. The format and even the
1178# name or existence of this file are subject to change without notice. Don't
1179# use it directly.
99870f4d
KW
1180EOF
1181
1182my $DEVELOPMENT_ONLY=<<"EOF";
1183# !!!!!!! DEVELOPMENT USE ONLY !!!!!!!
1184# This file contains information artificially constrained to code points
1185# present in Unicode release $string_compare_versions.
1186# IT CANNOT BE RELIED ON. It is for use during development only and should
23e33b60 1187# not be used for production.
b6922eda
KW
1188
1189EOF
1190
6189eadc
KW
1191my $MAX_UNICODE_CODEPOINT_STRING = "10FFFF";
1192my $MAX_UNICODE_CODEPOINT = hex $MAX_UNICODE_CODEPOINT_STRING;
1193my $MAX_UNICODE_CODEPOINTS = $MAX_UNICODE_CODEPOINT + 1;
99870f4d
KW
1194
1195# Matches legal code point. 4-6 hex numbers, If there are 6, the first
1196# two must be 10; if there are 5, the first must not be a 0. Written this way
92199589
KW
1197# to decrease backtracking. The first regex allows the code point to be at
1198# the end of a word, but to work properly, the word shouldn't end with a valid
1199# hex character. The second one won't match a code point at the end of a
1200# word, and doesn't have the run-on issue
8c32d378
KW
1201my $run_on_code_point_re =
1202 qr/ (?: 10[0-9A-F]{4} | [1-9A-F][0-9A-F]{4} | [0-9A-F]{4} ) \b/x;
1203my $code_point_re = qr/\b$run_on_code_point_re/;
99870f4d
KW
1204
1205# This matches the beginning of the line in the Unicode db files that give the
1206# defaults for code points not listed (i.e., missing) in the file. The code
1207# depends on this ending with a semi-colon, so it can assume it is a valid
1208# field when the line is split() by semi-colons
1209my $missing_defaults_prefix =
6189eadc 1210 qr/^#\s+\@missing:\s+0000\.\.$MAX_UNICODE_CODEPOINT_STRING\s*;/;
99870f4d
KW
1211
1212# Property types. Unicode has more types, but these are sufficient for our
1213# purposes.
1214my $UNKNOWN = -1; # initialized to illegal value
1215my $NON_STRING = 1; # Either binary or enum
1216my $BINARY = 2;
06f26c45
KW
1217my $FORCED_BINARY = 3; # Not a binary property, but, besides its normal
1218 # tables, additional true and false tables are
1219 # generated so that false is anything matching the
1220 # default value, and true is everything else.
1221my $ENUM = 4; # Include catalog
1222my $STRING = 5; # Anything else: string or misc
99870f4d
KW
1223
1224# Some input files have lines that give default values for code points not
1225# contained in the file. Sometimes these should be ignored.
1226my $NO_DEFAULTS = 0; # Must evaluate to false
1227my $NOT_IGNORED = 1;
1228my $IGNORED = 2;
1229
1230# Range types. Each range has a type. Most ranges are type 0, for normal,
1231# and will appear in the main body of the tables in the output files, but
1232# there are other types of ranges as well, listed below, that are specially
1233# handled. There are pseudo-types as well that will never be stored as a
1234# type, but will affect the calculation of the type.
1235
1236# 0 is for normal, non-specials
1237my $MULTI_CP = 1; # Sequence of more than code point
1238my $HANGUL_SYLLABLE = 2;
1239my $CP_IN_NAME = 3; # The NAME contains the code point appended to it.
1240my $NULL = 4; # The map is to the null string; utf8.c can't
1241 # handle these, nor is there an accepted syntax
1242 # for them in \p{} constructs
f86864ac 1243my $COMPUTE_NO_MULTI_CP = 5; # Pseudo-type; means that ranges that would
99870f4d
KW
1244 # otherwise be $MULTI_CP type are instead type 0
1245
1246# process_generic_property_file() can accept certain overrides in its input.
1247# Each of these must begin AND end with $CMD_DELIM.
1248my $CMD_DELIM = "\a";
1249my $REPLACE_CMD = 'replace'; # Override the Replace
1250my $MAP_TYPE_CMD = 'map_type'; # Override the Type
1251
1252my $NO = 0;
1253my $YES = 1;
1254
1255# Values for the Replace argument to add_range.
1256# $NO # Don't replace; add only the code points not
1257 # already present.
1258my $IF_NOT_EQUIVALENT = 1; # Replace only under certain conditions; details in
1259 # the comments at the subroutine definition.
1260my $UNCONDITIONALLY = 2; # Replace without conditions.
9470941f 1261my $MULTIPLE_BEFORE = 4; # Don't replace, but add a duplicate record if
99870f4d 1262 # already there
7f4b1e25
KW
1263my $MULTIPLE_AFTER = 5; # Don't replace, but add a duplicate record if
1264 # already there
1265my $CROAK = 6; # Die with an error if is already there
99870f4d
KW
1266
1267# Flags to give property statuses. The phrases are to remind maintainers that
1268# if the flag is changed, the indefinite article referring to it in the
1269# documentation may need to be as well.
1270my $NORMAL = "";
99870f4d
KW
1271my $DEPRECATED = 'D';
1272my $a_bold_deprecated = "a 'B<$DEPRECATED>'";
1273my $A_bold_deprecated = "A 'B<$DEPRECATED>'";
1274my $DISCOURAGED = 'X';
1275my $a_bold_discouraged = "an 'B<$DISCOURAGED>'";
1276my $A_bold_discouraged = "An 'B<$DISCOURAGED>'";
1277my $STRICTER = 'T';
1278my $a_bold_stricter = "a 'B<$STRICTER>'";
1279my $A_bold_stricter = "A 'B<$STRICTER>'";
1280my $STABILIZED = 'S';
1281my $a_bold_stabilized = "an 'B<$STABILIZED>'";
1282my $A_bold_stabilized = "An 'B<$STABILIZED>'";
1283my $OBSOLETE = 'O';
1284my $a_bold_obsolete = "an 'B<$OBSOLETE>'";
1285my $A_bold_obsolete = "An 'B<$OBSOLETE>'";
1286
1287my %status_past_participles = (
1288 $DISCOURAGED => 'discouraged',
99870f4d
KW
1289 $STABILIZED => 'stabilized',
1290 $OBSOLETE => 'obsolete',
37e2e78e 1291 $DEPRECATED => 'deprecated',
99870f4d
KW
1292);
1293
395dfc19
KW
1294# Table fates. These are somewhat ordered, so that fates < $MAP_PROXIED should be
1295# externally documented.
301ba948 1296my $ORDINARY = 0; # The normal fate.
395dfc19
KW
1297my $MAP_PROXIED = 1; # The map table for the property isn't written out,
1298 # but there is a file written that can be used to
1299 # reconstruct this table
3cdaf629 1300my $INTERNAL_ONLY = 2; # The file for this table is written out, but it is
301ba948 1301 # for Perl's internal use only
3cdaf629
KW
1302my $SUPPRESSED = 3; # The file for this table is not written out, and as a
1303 # result, we don't bother to do many computations on
1304 # it.
1305my $PLACEHOLDER = 4; # Like $SUPPRESSED, but we go through all the
1306 # computations anyway, as the values are needed for
1307 # things to work. This happens when we have Perl
1308 # extensions that depend on Unicode tables that
1309 # wouldn't normally be in a given Unicode version.
301ba948 1310
f5817e0a
KW
1311# The format of the values of the tables:
1312my $EMPTY_FORMAT = "";
99870f4d
KW
1313my $BINARY_FORMAT = 'b';
1314my $DECIMAL_FORMAT = 'd';
1315my $FLOAT_FORMAT = 'f';
1316my $INTEGER_FORMAT = 'i';
1317my $HEX_FORMAT = 'x';
1318my $RATIONAL_FORMAT = 'r';
1319my $STRING_FORMAT = 's';
d11155ec 1320my $ADJUST_FORMAT = 'a';
a14f3cb1 1321my $DECOMP_STRING_FORMAT = 'c';
c3ff2976 1322my $STRING_WHITE_SPACE_LIST = 'sw';
99870f4d
KW
1323
1324my %map_table_formats = (
1325 $BINARY_FORMAT => 'binary',
1326 $DECIMAL_FORMAT => 'single decimal digit',
1327 $FLOAT_FORMAT => 'floating point number',
1328 $INTEGER_FORMAT => 'integer',
add63c13 1329 $HEX_FORMAT => 'non-negative hex whole number; a code point',
99870f4d 1330 $RATIONAL_FORMAT => 'rational: an integer or a fraction',
1a9d544b 1331 $STRING_FORMAT => 'string',
d11155ec 1332 $ADJUST_FORMAT => 'some entries need adjustment',
92f9d56c 1333 $DECOMP_STRING_FORMAT => 'Perl\'s internal (Normalize.pm) decomposition mapping',
c3ff2976 1334 $STRING_WHITE_SPACE_LIST => 'string, but some elements are interpreted as a list; white space occurs only as list item separators'
99870f4d
KW
1335);
1336
1337# Unicode didn't put such derived files in a separate directory at first.
1338my $EXTRACTED_DIR = (-d 'extracted') ? 'extracted' : "";
1339my $EXTRACTED = ($EXTRACTED_DIR) ? "$EXTRACTED_DIR/" : "";
1340my $AUXILIARY = 'auxiliary';
1341
1342# Hashes that will eventually go into Heavy.pl for the use of utf8_heavy.pl
9e4a1e86 1343# and into UCD.pl for the use of UCD.pm
99870f4d
KW
1344my %loose_to_file_of; # loosely maps table names to their respective
1345 # files
1346my %stricter_to_file_of; # same; but for stricter mapping.
315bfd4e 1347my %loose_property_to_file_of; # Maps a loose property name to its map file
89cf10cc
KW
1348my %file_to_swash_name; # Maps the file name to its corresponding key name
1349 # in the hash %utf8::SwashInfo
99870f4d
KW
1350my %nv_floating_to_rational; # maps numeric values floating point numbers to
1351 # their rational equivalent
c12f2655
KW
1352my %loose_property_name_of; # Loosely maps (non_string) property names to
1353 # standard form
86a52d1e 1354my %string_property_loose_to_name; # Same, for string properties.
c15fda25
KW
1355my %loose_defaults; # keys are of form "prop=value", where 'prop' is
1356 # the property name in standard loose form, and
1357 # 'value' is the default value for that property,
1358 # also in standard loose form.
9e4a1e86
KW
1359my %loose_to_standard_value; # loosely maps table names to the canonical
1360 # alias for them
2df7880f
KW
1361my %ambiguous_names; # keys are alias names (in standard form) that
1362 # have more than one possible meaning.
5d1df013
KW
1363my %prop_aliases; # Keys are standard property name; values are each
1364 # one's aliases
1e863613
KW
1365my %prop_value_aliases; # Keys of top level are standard property name;
1366 # values are keys to another hash, Each one is
1367 # one of the property's values, in standard form.
1368 # The values are that prop-val's aliases.
2df7880f 1369my %ucd_pod; # Holds entries that will go into the UCD section of the pod
99870f4d 1370
d867ccfb
KW
1371# Most properties are immune to caseless matching, otherwise you would get
1372# nonsensical results, as properties are a function of a code point, not
1373# everything that is caselessly equivalent to that code point. For example,
1374# Changes_When_Case_Folded('s') should be false, whereas caselessly it would
1375# be true because 's' and 'S' are equivalent caselessly. However,
1376# traditionally, [:upper:] and [:lower:] are equivalent caselessly, so we
1377# extend that concept to those very few properties that are like this. Each
1378# such property will match the full range caselessly. They are hard-coded in
1379# the program; it's not worth trying to make it general as it's extremely
1380# unlikely that they will ever change.
1381my %caseless_equivalent_to;
1382
99870f4d
KW
1383# These constants names and values were taken from the Unicode standard,
1384# version 5.1, section 3.12. They are used in conjunction with Hangul
6e5a209b
KW
1385# syllables. The '_string' versions are so generated tables can retain the
1386# hex format, which is the more familiar value
1387my $SBase_string = "0xAC00";
1388my $SBase = CORE::hex $SBase_string;
1389my $LBase_string = "0x1100";
1390my $LBase = CORE::hex $LBase_string;
1391my $VBase_string = "0x1161";
1392my $VBase = CORE::hex $VBase_string;
1393my $TBase_string = "0x11A7";
1394my $TBase = CORE::hex $TBase_string;
99870f4d
KW
1395my $SCount = 11172;
1396my $LCount = 19;
1397my $VCount = 21;
1398my $TCount = 28;
1399my $NCount = $VCount * $TCount;
1400
1401# For Hangul syllables; These store the numbers from Jamo.txt in conjunction
1402# with the above published constants.
1403my %Jamo;
1404my %Jamo_L; # Leading consonants
1405my %Jamo_V; # Vowels
1406my %Jamo_T; # Trailing consonants
1407
bb1dd3da
KW
1408# For code points whose name contains its ordinal as a '-ABCD' suffix.
1409# The key is the base name of the code point, and the value is an
1410# array giving all the ranges that use this base name. Each range
1411# is actually a hash giving the 'low' and 'high' values of it.
1412my %names_ending_in_code_point;
1413my %loose_names_ending_in_code_point; # Same as above, but has blanks, dashes
1414 # removed from the names
1415# Inverse mapping. The list of ranges that have these kinds of
1416# names. Each element contains the low, high, and base names in an
1417# anonymous hash.
1418my @code_points_ending_in_code_point;
1419
6b5ab373
KW
1420# To hold Unicode's normalization test suite
1421my @normalization_tests;
1422
bb1dd3da
KW
1423# Boolean: does this Unicode version have the hangul syllables, and are we
1424# writing out a table for them?
1425my $has_hangul_syllables = 0;
1426
1427# Does this Unicode version have code points whose names end in their
1428# respective code points, and are we writing out a table for them? 0 for no;
1429# otherwise points to first property that a table is needed for them, so that
1430# if multiple tables are needed, we don't create duplicates
1431my $needing_code_points_ending_in_code_point = 0;
1432
37e2e78e 1433my @backslash_X_tests; # List of tests read in for testing \X
99870f4d
KW
1434my @unhandled_properties; # Will contain a list of properties found in
1435 # the input that we didn't process.
f86864ac 1436my @match_properties; # Properties that have match tables, to be
99870f4d
KW
1437 # listed in the pod
1438my @map_properties; # Properties that get map files written
1439my @named_sequences; # NamedSequences.txt contents.
1440my %potential_files; # Generated list of all .txt files in the directory
1441 # structure so we can warn if something is being
1442 # ignored.
1443my @files_actually_output; # List of files we generated.
1444my @more_Names; # Some code point names are compound; this is used
1445 # to store the extra components of them.
1446my $MIN_FRACTION_LENGTH = 3; # How many digits of a floating point number at
1447 # the minimum before we consider it equivalent to a
1448 # candidate rational
1449my $MAX_FLOATING_SLOP = 10 ** - $MIN_FRACTION_LENGTH; # And in floating terms
1450
1451# These store references to certain commonly used property objects
1452my $gc;
1453my $perl;
1454my $block;
3e20195b
KW
1455my $perl_charname;
1456my $print;
7fc6cb55 1457my $Any;
359523e2 1458my $script;
99870f4d
KW
1459
1460# Are there conflicting names because of beginning with 'In_', or 'Is_'
1461my $has_In_conflicts = 0;
1462my $has_Is_conflicts = 0;
1463
1464sub internal_file_to_platform ($) {
1465 # Convert our file paths which have '/' separators to those of the
1466 # platform.
1467
1468 my $file = shift;
1469 return undef unless defined $file;
1470
1471 return File::Spec->join(split '/', $file);
d07a55ed 1472}
5beb625e 1473
99870f4d
KW
1474sub file_exists ($) { # platform independent '-e'. This program internally
1475 # uses slash as a path separator.
1476 my $file = shift;
1477 return 0 if ! defined $file;
1478 return -e internal_file_to_platform($file);
1479}
5beb625e 1480
99870f4d 1481sub objaddr($) {
23e33b60
KW
1482 # Returns the address of the blessed input object.
1483 # It doesn't check for blessedness because that would do a string eval
1484 # every call, and the program is structured so that this is never called
1485 # for a non-blessed object.
99870f4d 1486
23e33b60 1487 no overloading; # If overloaded, numifying below won't work.
99870f4d
KW
1488
1489 # Numifying a ref gives its address.
051df77b 1490 return pack 'J', $_[0];
99870f4d
KW
1491}
1492
558712cf 1493# These are used only if $annotate is true.
c4019d52
KW
1494# The entire range of Unicode characters is examined to populate these
1495# after all the input has been processed. But most can be skipped, as they
1496# have the same descriptive phrases, such as being unassigned
1497my @viacode; # Contains the 1 million character names
1498my @printable; # boolean: And are those characters printable?
1499my @annotate_char_type; # Contains a type of those characters, specifically
1500 # for the purposes of annotation.
1501my $annotate_ranges; # A map of ranges of code points that have the same
98dc9551 1502 # name for the purposes of annotation. They map to the
c4019d52
KW
1503 # upper edge of the range, so that the end point can
1504 # be immediately found. This is used to skip ahead to
1505 # the end of a range, and avoid processing each
1506 # individual code point in it.
1507my $unassigned_sans_noncharacters; # A Range_List of the unassigned
1508 # characters, but excluding those which are
1509 # also noncharacter code points
1510
1511# The annotation types are an extension of the regular range types, though
1512# some of the latter are folded into one. Make the new types negative to
1513# avoid conflicting with the regular types
1514my $SURROGATE_TYPE = -1;
1515my $UNASSIGNED_TYPE = -2;
1516my $PRIVATE_USE_TYPE = -3;
1517my $NONCHARACTER_TYPE = -4;
1518my $CONTROL_TYPE = -5;
1519my $UNKNOWN_TYPE = -6; # Used only if there is a bug in this program
1520
1521sub populate_char_info ($) {
558712cf 1522 # Used only with the $annotate option. Populates the arrays with the
c4019d52
KW
1523 # input code point's info that are needed for outputting more detailed
1524 # comments. If calling context wants a return, it is the end point of
1525 # any contiguous range of characters that share essentially the same info
1526
1527 my $i = shift;
1528 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1529
1530 $viacode[$i] = $perl_charname->value_of($i) || "";
1531
1532 # A character is generally printable if Unicode says it is,
1533 # but below we make sure that most Unicode general category 'C' types
1534 # aren't.
1535 $printable[$i] = $print->contains($i);
1536
1537 $annotate_char_type[$i] = $perl_charname->type_of($i) || 0;
1538
1539 # Only these two regular types are treated specially for annotations
1540 # purposes
1541 $annotate_char_type[$i] = 0 if $annotate_char_type[$i] != $CP_IN_NAME
1542 && $annotate_char_type[$i] != $HANGUL_SYLLABLE;
1543
1544 # Give a generic name to all code points that don't have a real name.
1545 # We output ranges, if applicable, for these. Also calculate the end
1546 # point of the range.
1547 my $end;
1548 if (! $viacode[$i]) {
1d025d66
KW
1549 my $nonchar;
1550 if ($gc-> table('Private_use')->contains($i)) {
c4019d52
KW
1551 $viacode[$i] = 'Private Use';
1552 $annotate_char_type[$i] = $PRIVATE_USE_TYPE;
1553 $printable[$i] = 0;
1554 $end = $gc->table('Private_Use')->containing_range($i)->end;
1555 }
1d025d66
KW
1556 elsif ((defined ($nonchar =
1557 Property::property_ref('Noncharacter_Code_Point'))
1558 && $nonchar->table('Y')->contains($i)))
c4019d52
KW
1559 {
1560 $viacode[$i] = 'Noncharacter';
1561 $annotate_char_type[$i] = $NONCHARACTER_TYPE;
1562 $printable[$i] = 0;
1563 $end = property_ref('Noncharacter_Code_Point')->table('Y')->
1564 containing_range($i)->end;
1565 }
1566 elsif ($gc-> table('Control')->contains($i)) {
c71dea7f 1567 $viacode[$i] = property_ref('Name_Alias')->value_of($i) || 'Control';
c4019d52
KW
1568 $annotate_char_type[$i] = $CONTROL_TYPE;
1569 $printable[$i] = 0;
c4019d52
KW
1570 }
1571 elsif ($gc-> table('Unassigned')->contains($i)) {
c4019d52
KW
1572 $annotate_char_type[$i] = $UNASSIGNED_TYPE;
1573 $printable[$i] = 0;
1d025d66
KW
1574 if ($v_version lt v2.0.0) { # No blocks in earliest releases
1575 $viacode[$i] = 'Unassigned';
1576 $end = $gc-> table('Unassigned')->containing_range($i)->end;
1577 }
1578 else {
1579 $viacode[$i] = 'Unassigned, block=' . $block-> value_of($i);
c4019d52 1580
bf06c733
KW
1581 # Because we name the unassigned by the blocks they are in, it
1582 # can't go past the end of that block, and it also can't go
1583 # past the unassigned range it is in. The special table makes
1584 # sure that the non-characters, which are unassigned, are
1585 # separated out.
1586 $end = min($block->containing_range($i)->end,
1587 $unassigned_sans_noncharacters->
1588 containing_range($i)->end);
1d025d66
KW
1589 }
1590 }
1591 elsif ($v_version lt v2.0.0) { # No surrogates in earliest releases
1592 $viacode[$i] = $gc->value_of($i);
1593 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1594 $printable[$i] = 0;
1595 }
1596 elsif ($gc-> table('Surrogate')->contains($i)) {
1597 $viacode[$i] = 'Surrogate';
1598 $annotate_char_type[$i] = $SURROGATE_TYPE;
1599 $printable[$i] = 0;
1600 $end = $gc->table('Surrogate')->containing_range($i)->end;
13ca76ff
KW
1601 }
1602 else {
1603 Carp::my_carp_bug("Can't figure out how to annotate "
1604 . sprintf("U+%04X", $i)
1605 . ". Proceeding anyway.");
c4019d52
KW
1606 $viacode[$i] = 'UNKNOWN';
1607 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1608 $printable[$i] = 0;
1609 }
1610 }
1611
1612 # Here, has a name, but if it's one in which the code point number is
1613 # appended to the name, do that.
1614 elsif ($annotate_char_type[$i] == $CP_IN_NAME) {
1615 $viacode[$i] .= sprintf("-%04X", $i);
1616 $end = $perl_charname->containing_range($i)->end;
1617 }
1618
1619 # And here, has a name, but if it's a hangul syllable one, replace it with
1620 # the correct name from the Unicode algorithm
1621 elsif ($annotate_char_type[$i] == $HANGUL_SYLLABLE) {
1622 use integer;
1623 my $SIndex = $i - $SBase;
1624 my $L = $LBase + $SIndex / $NCount;
1625 my $V = $VBase + ($SIndex % $NCount) / $TCount;
1626 my $T = $TBase + $SIndex % $TCount;
1627 $viacode[$i] = "HANGUL SYLLABLE $Jamo{$L}$Jamo{$V}";
1628 $viacode[$i] .= $Jamo{$T} if $T != $TBase;
1629 $end = $perl_charname->containing_range($i)->end;
1630 }
1631
1632 return if ! defined wantarray;
1633 return $i if ! defined $end; # If not a range, return the input
1634
1635 # Save this whole range so can find the end point quickly
1636 $annotate_ranges->add_map($i, $end, $end);
1637
1638 return $end;
1639}
1640
23e33b60
KW
1641# Commented code below should work on Perl 5.8.
1642## This 'require' doesn't necessarily work in miniperl, and even if it does,
1643## the native perl version of it (which is what would operate under miniperl)
1644## is extremely slow, as it does a string eval every call.
7e017d6d 1645#my $has_fast_scalar_util = $^X !~ /miniperl/
23e33b60
KW
1646# && defined eval "require Scalar::Util";
1647#
1648#sub objaddr($) {
1649# # Returns the address of the blessed input object. Uses the XS version if
1650# # available. It doesn't check for blessedness because that would do a
1651# # string eval every call, and the program is structured so that this is
1652# # never called for a non-blessed object.
1653#
1654# return Scalar::Util::refaddr($_[0]) if $has_fast_scalar_util;
1655#
1656# # Check at least that is a ref.
1657# my $pkg = ref($_[0]) or return undef;
1658#
1659# # Change to a fake package to defeat any overloaded stringify
1660# bless $_[0], 'main::Fake';
1661#
1662# # Numifying a ref gives its address.
051df77b 1663# my $addr = pack 'J', $_[0];
23e33b60
KW
1664#
1665# # Return to original class
1666# bless $_[0], $pkg;
1667# return $addr;
1668#}
1669
99870f4d
KW
1670sub max ($$) {
1671 my $a = shift;
1672 my $b = shift;
1673 return $a if $a >= $b;
1674 return $b;
1675}
1676
1677sub min ($$) {
1678 my $a = shift;
1679 my $b = shift;
1680 return $a if $a <= $b;
1681 return $b;
1682}
1683
1684sub clarify_number ($) {
1685 # This returns the input number with underscores inserted every 3 digits
1686 # in large (5 digits or more) numbers. Input must be entirely digits, not
1687 # checked.
1688
1689 my $number = shift;
1690 my $pos = length($number) - 3;
1691 return $number if $pos <= 1;
1692 while ($pos > 0) {
1693 substr($number, $pos, 0) = '_';
1694 $pos -= 3;
5beb625e 1695 }
99870f4d 1696 return $number;
99598c8c
JH
1697}
1698
12ac2576 1699
99870f4d 1700package Carp;
7ebf06b3 1701
99870f4d
KW
1702# These routines give a uniform treatment of messages in this program. They
1703# are placed in the Carp package to cause the stack trace to not include them,
1704# although an alternative would be to use another package and set @CARP_NOT
1705# for it.
12ac2576 1706
99870f4d 1707our $Verbose = 1 if main::DEBUG; # Useful info when debugging
12ac2576 1708
99f78760
KW
1709# This is a work-around suggested by Nicholas Clark to fix a problem with Carp
1710# and overload trying to load Scalar:Util under miniperl. See
1711# http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2009-11/msg01057.html
1712undef $overload::VERSION;
1713
99870f4d
KW
1714sub my_carp {
1715 my $message = shift || "";
1716 my $nofold = shift || 0;
7ebf06b3 1717
99870f4d
KW
1718 if ($message) {
1719 $message = main::join_lines($message);
1720 $message =~ s/^$0: *//; # Remove initial program name
1721 $message =~ s/[.;,]+$//; # Remove certain ending punctuation
1722 $message = "\n$0: $message;";
12ac2576 1723
99870f4d
KW
1724 # Fold the message with program name, semi-colon end punctuation
1725 # (which looks good with the message that carp appends to it), and a
1726 # hanging indent for continuation lines.
1727 $message = main::simple_fold($message, "", 4) unless $nofold;
1728 $message =~ s/\n$//; # Remove the trailing nl so what carp
1729 # appends is to the same line
1730 }
12ac2576 1731
99870f4d 1732 return $message if defined wantarray; # If a caller just wants the msg
12ac2576 1733
99870f4d
KW
1734 carp $message;
1735 return;
1736}
7ebf06b3 1737
99870f4d
KW
1738sub my_carp_bug {
1739 # This is called when it is clear that the problem is caused by a bug in
1740 # this program.
7ebf06b3 1741
99870f4d
KW
1742 my $message = shift;
1743 $message =~ s/^$0: *//;
1744 $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");
1745 carp $message;
1746 return;
1747}
7ebf06b3 1748
99870f4d
KW
1749sub carp_too_few_args {
1750 if (@_ != 2) {
1751 my_carp_bug("Wrong number of arguments: to 'carp_too_few_arguments'. No action taken.");
1752 return;
12ac2576 1753 }
7ebf06b3 1754
99870f4d
KW
1755 my $args_ref = shift;
1756 my $count = shift;
7ebf06b3 1757
99870f4d
KW
1758 my_carp_bug("Need at least $count arguments to "
1759 . (caller 1)[3]
1760 . ". Instead got: '"
1761 . join ', ', @$args_ref
1762 . "'. No action taken.");
1763 return;
12ac2576
JP
1764}
1765
99870f4d
KW
1766sub carp_extra_args {
1767 my $args_ref = shift;
1768 my_carp_bug("Too many arguments to 'carp_extra_args': (" . join(', ', @_) . "); Extras ignored.") if @_;
12ac2576 1769
99870f4d
KW
1770 unless (ref $args_ref) {
1771 my_carp_bug("Argument to 'carp_extra_args' ($args_ref) must be a ref. Not checking arguments.");
1772 return;
1773 }
1774 my ($package, $file, $line) = caller;
1775 my $subroutine = (caller 1)[3];
cf25bb62 1776
99870f4d
KW
1777 my $list;
1778 if (ref $args_ref eq 'HASH') {
1779 foreach my $key (keys %$args_ref) {
1780 $args_ref->{$key} = $UNDEF unless defined $args_ref->{$key};
cf25bb62 1781 }
99870f4d 1782 $list = join ', ', each %{$args_ref};
cf25bb62 1783 }
99870f4d
KW
1784 elsif (ref $args_ref eq 'ARRAY') {
1785 foreach my $arg (@$args_ref) {
1786 $arg = $UNDEF unless defined $arg;
1787 }
1788 $list = join ', ', @$args_ref;
1789 }
1790 else {
1791 my_carp_bug("Can't cope with ref "
1792 . ref($args_ref)
1793 . " . argument to 'carp_extra_args'. Not checking arguments.");
1794 return;
1795 }
1796
1797 my_carp_bug("Unrecognized parameters in options: '$list' to $subroutine. Skipped.");
1798 return;
d73e5302
JH
1799}
1800
99870f4d
KW
1801package main;
1802
1803{ # Closure
1804
1805 # This program uses the inside-out method for objects, as recommended in
1806 # "Perl Best Practices". This closure aids in generating those. There
1807 # are two routines. setup_package() is called once per package to set
1808 # things up, and then set_access() is called for each hash representing a
1809 # field in the object. These routines arrange for the object to be
1810 # properly destroyed when no longer used, and for standard accessor
1811 # functions to be generated. If you need more complex accessors, just
1812 # write your own and leave those accesses out of the call to set_access().
1813 # More details below.
1814
1815 my %constructor_fields; # fields that are to be used in constructors; see
1816 # below
1817
1818 # The values of this hash will be the package names as keys to other
1819 # hashes containing the name of each field in the package as keys, and
1820 # references to their respective hashes as values.
1821 my %package_fields;
1822
1823 sub setup_package {
1824 # Sets up the package, creating standard DESTROY and dump methods
1825 # (unless already defined). The dump method is used in debugging by
1826 # simple_dumper().
1827 # The optional parameters are:
1828 # a) a reference to a hash, that gets populated by later
1829 # set_access() calls with one of the accesses being
1830 # 'constructor'. The caller can then refer to this, but it is
1831 # not otherwise used by these two routines.
1832 # b) a reference to a callback routine to call during destruction
1833 # of the object, before any fields are actually destroyed
1834
1835 my %args = @_;
1836 my $constructor_ref = delete $args{'Constructor_Fields'};
1837 my $destroy_callback = delete $args{'Destroy_Callback'};
1838 Carp::carp_extra_args(\@_) if main::DEBUG && %args;
1839
1840 my %fields;
1841 my $package = (caller)[0];
1842
1843 $package_fields{$package} = \%fields;
1844 $constructor_fields{$package} = $constructor_ref;
1845
1846 unless ($package->can('DESTROY')) {
1847 my $destroy_name = "${package}::DESTROY";
1848 no strict "refs";
1849
1850 # Use typeglob to give the anonymous subroutine the name we want
1851 *$destroy_name = sub {
1852 my $self = shift;
ffe43484 1853 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1854
1855 $self->$destroy_callback if $destroy_callback;
1856 foreach my $field (keys %{$package_fields{$package}}) {
1857 #print STDERR __LINE__, ": Destroying ", ref $self, " ", sprintf("%04X", $addr), ": ", $field, "\n";
1858 delete $package_fields{$package}{$field}{$addr};
1859 }
1860 return;
1861 }
1862 }
1863
1864 unless ($package->can('dump')) {
1865 my $dump_name = "${package}::dump";
1866 no strict "refs";
1867 *$dump_name = sub {
1868 my $self = shift;
1869 return dump_inside_out($self, $package_fields{$package}, @_);
1870 }
1871 }
1872 return;
1873 }
1874
1875 sub set_access {
1876 # Arrange for the input field to be garbage collected when no longer
1877 # needed. Also, creates standard accessor functions for the field
1878 # based on the optional parameters-- none if none of these parameters:
1879 # 'addable' creates an 'add_NAME()' accessor function.
1880 # 'readable' or 'readable_array' creates a 'NAME()' accessor
1881 # function.
1882 # 'settable' creates a 'set_NAME()' accessor function.
1883 # 'constructor' doesn't create an accessor function, but adds the
1884 # field to the hash that was previously passed to
1885 # setup_package();
1886 # Any of the accesses can be abbreviated down, so that 'a', 'ad',
1887 # 'add' etc. all mean 'addable'.
1888 # The read accessor function will work on both array and scalar
1889 # values. If another accessor in the parameter list is 'a', the read
1890 # access assumes an array. You can also force it to be array access
1891 # by specifying 'readable_array' instead of 'readable'
1892 #
1893 # A sort-of 'protected' access can be set-up by preceding the addable,
1894 # readable or settable with some initial portion of 'protected_' (but,
1895 # the underscore is required), like 'p_a', 'pro_set', etc. The
1896 # "protection" is only by convention. All that happens is that the
1897 # accessor functions' names begin with an underscore. So instead of
1898 # calling set_foo, the call is _set_foo. (Real protection could be
c1739a4a 1899 # accomplished by having a new subroutine, end_package, called at the
99870f4d
KW
1900 # end of each package, and then storing the __LINE__ ranges and
1901 # checking them on every accessor. But that is way overkill.)
1902
1903 # We create anonymous subroutines as the accessors and then use
1904 # typeglobs to assign them to the proper package and name
1905
1906 my $name = shift; # Name of the field
1907 my $field = shift; # Reference to the inside-out hash containing the
1908 # field
1909
1910 my $package = (caller)[0];
1911
1912 if (! exists $package_fields{$package}) {
1913 croak "$0: Must call 'setup_package' before 'set_access'";
1914 }
d73e5302 1915
99870f4d
KW
1916 # Stash the field so DESTROY can get it.
1917 $package_fields{$package}{$name} = $field;
cf25bb62 1918
99870f4d
KW
1919 # Remaining arguments are the accessors. For each...
1920 foreach my $access (@_) {
1921 my $access = lc $access;
cf25bb62 1922
99870f4d 1923 my $protected = "";
cf25bb62 1924
99870f4d
KW
1925 # Match the input as far as it goes.
1926 if ($access =~ /^(p[^_]*)_/) {
1927 $protected = $1;
1928 if (substr('protected_', 0, length $protected)
1929 eq $protected)
1930 {
1931
1932 # Add 1 for the underscore not included in $protected
1933 $access = substr($access, length($protected) + 1);
1934 $protected = '_';
1935 }
1936 else {
1937 $protected = "";
1938 }
1939 }
1940
1941 if (substr('addable', 0, length $access) eq $access) {
1942 my $subname = "${package}::${protected}add_$name";
1943 no strict "refs";
1944
1945 # add_ accessor. Don't add if already there, which we
1946 # determine using 'eq' for scalars and '==' otherwise.
1947 *$subname = sub {
1948 use strict "refs";
1949 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
1950 my $self = shift;
1951 my $value = shift;
ffe43484 1952 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1953 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1954 if (ref $value) {
f998e60c 1955 return if grep { $value == $_ } @{$field->{$addr}};
99870f4d
KW
1956 }
1957 else {
f998e60c 1958 return if grep { $value eq $_ } @{$field->{$addr}};
99870f4d 1959 }
f998e60c 1960 push @{$field->{$addr}}, $value;
99870f4d
KW
1961 return;
1962 }
1963 }
1964 elsif (substr('constructor', 0, length $access) eq $access) {
1965 if ($protected) {
1966 Carp::my_carp_bug("Can't set-up 'protected' constructors")
1967 }
1968 else {
1969 $constructor_fields{$package}{$name} = $field;
1970 }
1971 }
1972 elsif (substr('readable_array', 0, length $access) eq $access) {
1973
1974 # Here has read access. If one of the other parameters for
1975 # access is array, or this one specifies array (by being more
1976 # than just 'readable_'), then create a subroutine that
1977 # assumes the data is an array. Otherwise just a scalar
1978 my $subname = "${package}::${protected}$name";
1979 if (grep { /^a/i } @_
1980 or length($access) > length('readable_'))
1981 {
1982 no strict "refs";
1983 *$subname = sub {
1984 use strict "refs";
23e33b60 1985 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
ffe43484 1986 my $addr = do { no overloading; pack 'J', $_[0]; };
99870f4d
KW
1987 if (ref $field->{$addr} ne 'ARRAY') {
1988 my $type = ref $field->{$addr};
1989 $type = 'scalar' unless $type;
1990 Carp::my_carp_bug("Trying to read $name as an array when it is a $type. Big problems.");
1991 return;
1992 }
1993 return scalar @{$field->{$addr}} unless wantarray;
1994
1995 # Make a copy; had problems with caller modifying the
1996 # original otherwise
1997 my @return = @{$field->{$addr}};
1998 return @return;
1999 }
2000 }
2001 else {
2002
2003 # Here not an array value, a simpler function.
2004 no strict "refs";
2005 *$subname = sub {
2006 use strict "refs";
23e33b60 2007 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
f998e60c 2008 no overloading;
051df77b 2009 return $field->{pack 'J', $_[0]};
99870f4d
KW
2010 }
2011 }
2012 }
2013 elsif (substr('settable', 0, length $access) eq $access) {
2014 my $subname = "${package}::${protected}set_$name";
2015 no strict "refs";
2016 *$subname = sub {
2017 use strict "refs";
23e33b60
KW
2018 if (main::DEBUG) {
2019 return Carp::carp_too_few_args(\@_, 2) if @_ < 2;
2020 Carp::carp_extra_args(\@_) if @_ > 2;
2021 }
2022 # $self is $_[0]; $value is $_[1]
f998e60c 2023 no overloading;
051df77b 2024 $field->{pack 'J', $_[0]} = $_[1];
99870f4d
KW
2025 return;
2026 }
2027 }
2028 else {
2029 Carp::my_carp_bug("Unknown accessor type $access. No accessor set.");
2030 }
cf25bb62 2031 }
99870f4d 2032 return;
cf25bb62 2033 }
99870f4d
KW
2034}
2035
2036package Input_file;
2037
2038# All input files use this object, which stores various attributes about them,
2039# and provides for convenient, uniform handling. The run method wraps the
2040# processing. It handles all the bookkeeping of opening, reading, and closing
2041# the file, returning only significant input lines.
2042#
2043# Each object gets a handler which processes the body of the file, and is
74cd47d0
KW
2044# called by run(). All character property files must use the generic,
2045# default handler, which has code scrubbed to handle things you might not
2046# expect, including automatic EBCDIC handling. For files that don't deal with
2047# mapping code points to a property value, such as test files,
2048# PropertyAliases, PropValueAliases, and named sequences, you can override the
2049# handler to be a custom one. Such a handler should basically be a
2050# while(next_line()) {...} loop.
99870f4d
KW
2051#
2052# You can also set up handlers to
537124e4 2053# 1) call before the first line is read, for pre processing
99870f4d
KW
2054# 2) call to adjust each line of the input before the main handler gets them
2055# 3) call upon EOF before the main handler exits its loop
537124e4 2056# 4) call at the end, for post processing
99870f4d
KW
2057#
2058# $_ is used to store the input line, and is to be filtered by the
2059# each_line_handler()s. So, if the format of the line is not in the desired
2060# format for the main handler, these are used to do that adjusting. They can
2061# be stacked (by enclosing them in an [ anonymous array ] in the constructor,
2062# so the $_ output of one is used as the input to the next. None of the other
2063# handlers are stackable, but could easily be changed to be so.
2064#
2065# Most of the handlers can call insert_lines() or insert_adjusted_lines()
2066# which insert the parameters as lines to be processed before the next input
2067# file line is read. This allows the EOF handler to flush buffers, for
2068# example. The difference between the two routines is that the lines inserted
2069# by insert_lines() are subjected to the each_line_handler()s. (So if you
2070# called it from such a handler, you would get infinite recursion.) Lines
2071# inserted by insert_adjusted_lines() go directly to the main handler without
2072# any adjustments. If the post-processing handler calls any of these, there
2073# will be no effect. Some error checking for these conditions could be added,
2074# but it hasn't been done.
2075#
2076# carp_bad_line() should be called to warn of bad input lines, which clears $_
2077# to prevent further processing of the line. This routine will output the
2078# message as a warning once, and then keep a count of the lines that have the
2079# same message, and output that count at the end of the file's processing.
2080# This keeps the number of messages down to a manageable amount.
2081#
2082# get_missings() should be called to retrieve any @missing input lines.
2083# Messages will be raised if this isn't done if the options aren't to ignore
2084# missings.
2085
2086sub trace { return main::trace(@_); }
2087
99870f4d
KW
2088{ # Closure
2089 # Keep track of fields that are to be put into the constructor.
2090 my %constructor_fields;
2091
2092 main::setup_package(Constructor_Fields => \%constructor_fields);
2093
2094 my %file; # Input file name, required
2095 main::set_access('file', \%file, qw{ c r });
2096
2097 my %first_released; # Unicode version file was first released in, required
2098 main::set_access('first_released', \%first_released, qw{ c r });
2099
2100 my %handler; # Subroutine to process the input file, defaults to
2101 # 'process_generic_property_file'
2102 main::set_access('handler', \%handler, qw{ c });
2103
2104 my %property;
2105 # name of property this file is for. defaults to none, meaning not
2106 # applicable, or is otherwise determinable, for example, from each line.
696609bf 2107 main::set_access('property', \%property, qw{ c r });
99870f4d
KW
2108
2109 my %optional;
2110 # If this is true, the file is optional. If not present, no warning is
2111 # output. If it is present, the string given by this parameter is
2112 # evaluated, and if false the file is not processed.
2113 main::set_access('optional', \%optional, 'c', 'r');
2114
2115 my %non_skip;
2116 # This is used for debugging, to skip processing of all but a few input
2117 # files. Add 'non_skip => 1' to the constructor for those files you want
2118 # processed when you set the $debug_skip global.
2119 main::set_access('non_skip', \%non_skip, 'c');
2120
37e2e78e 2121 my %skip;
09ca89ce
KW
2122 # This is used to skip processing of this input file semi-permanently,
2123 # when it evaluates to true. The value should be the reason the file is
2124 # being skipped. It is used for files that we aren't planning to process
2125 # anytime soon, but want to allow to be in the directory and not raise a
2126 # message that we are not handling. Mostly for test files. This is in
2127 # contrast to the non_skip element, which is supposed to be used very
2128 # temporarily for debugging. Sets 'optional' to 1. Also, files that we
2129 # pretty much will never look at can be placed in the global
1fec9f60 2130 # %ignored_files instead. Ones used here will be added to %skipped files
37e2e78e
KW
2131 main::set_access('skip', \%skip, 'c');
2132
99870f4d
KW
2133 my %each_line_handler;
2134 # list of subroutines to look at and filter each non-comment line in the
2135 # file. defaults to none. The subroutines are called in order, each is
2136 # to adjust $_ for the next one, and the final one adjusts it for
2137 # 'handler'
2138 main::set_access('each_line_handler', \%each_line_handler, 'c');
2139
2140 my %has_missings_defaults;
2141 # ? Are there lines in the file giving default values for code points
2142 # missing from it?. Defaults to NO_DEFAULTS. Otherwise NOT_IGNORED is
2143 # the norm, but IGNORED means it has such lines, but the handler doesn't
2144 # use them. Having these three states allows us to catch changes to the
2145 # UCD that this program should track
2146 main::set_access('has_missings_defaults',
2147 \%has_missings_defaults, qw{ c r });
2148
2149 my %pre_handler;
2150 # Subroutine to call before doing anything else in the file. If undef, no
2151 # such handler is called.
2152 main::set_access('pre_handler', \%pre_handler, qw{ c });
2153
2154 my %eof_handler;
2155 # Subroutine to call upon getting an EOF on the input file, but before
2156 # that is returned to the main handler. This is to allow buffers to be
2157 # flushed. The handler is expected to call insert_lines() or
2158 # insert_adjusted() with the buffered material
2159 main::set_access('eof_handler', \%eof_handler, qw{ c r });
2160
2161 my %post_handler;
2162 # Subroutine to call after all the lines of the file are read in and
2163 # processed. If undef, no such handler is called.
2164 main::set_access('post_handler', \%post_handler, qw{ c });
2165
2166 my %progress_message;
2167 # Message to print to display progress in lieu of the standard one
2168 main::set_access('progress_message', \%progress_message, qw{ c });
2169
2170 my %handle;
2171 # cache open file handle, internal. Is undef if file hasn't been
2172 # processed at all, empty if has;
2173 main::set_access('handle', \%handle);
2174
2175 my %added_lines;
2176 # cache of lines added virtually to the file, internal
2177 main::set_access('added_lines', \%added_lines);
2178
74cd47d0
KW
2179 my %remapped_lines;
2180 # cache of lines added virtually to the file, internal
2181 main::set_access('remapped_lines', \%remapped_lines);
2182
99870f4d
KW
2183 my %errors;
2184 # cache of errors found, internal
2185 main::set_access('errors', \%errors);
2186
2187 my %missings;
2188 # storage of '@missing' defaults lines
2189 main::set_access('missings', \%missings);
2190
74cd47d0
KW
2191 sub _next_line;
2192 sub _next_line_with_remapped_range;
2193
99870f4d
KW
2194 sub new {
2195 my $class = shift;
2196
2197 my $self = bless \do{ my $anonymous_scalar }, $class;
ffe43484 2198 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2199
2200 # Set defaults
2201 $handler{$addr} = \&main::process_generic_property_file;
2202 $non_skip{$addr} = 0;
37e2e78e 2203 $skip{$addr} = 0;
99870f4d
KW
2204 $has_missings_defaults{$addr} = $NO_DEFAULTS;
2205 $handle{$addr} = undef;
2206 $added_lines{$addr} = [ ];
74cd47d0 2207 $remapped_lines{$addr} = [ ];
99870f4d
KW
2208 $each_line_handler{$addr} = [ ];
2209 $errors{$addr} = { };
2210 $missings{$addr} = [ ];
2211
2212 # Two positional parameters.
99f78760 2213 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
99870f4d
KW
2214 $file{$addr} = main::internal_file_to_platform(shift);
2215 $first_released{$addr} = shift;
2216
2217 # The rest of the arguments are key => value pairs
2218 # %constructor_fields has been set up earlier to list all possible
2219 # ones. Either set or push, depending on how the default has been set
2220 # up just above.
2221 my %args = @_;
2222 foreach my $key (keys %args) {
2223 my $argument = $args{$key};
2224
2225 # Note that the fields are the lower case of the constructor keys
2226 my $hash = $constructor_fields{lc $key};
2227 if (! defined $hash) {
2228 Carp::my_carp_bug("Unrecognized parameters '$key => $argument' to new() for $self. Skipped");
2229 next;
2230 }
2231 if (ref $hash->{$addr} eq 'ARRAY') {
2232 if (ref $argument eq 'ARRAY') {
2233 foreach my $argument (@{$argument}) {
2234 next if ! defined $argument;
2235 push @{$hash->{$addr}}, $argument;
2236 }
2237 }
2238 else {
2239 push @{$hash->{$addr}}, $argument if defined $argument;
2240 }
2241 }
2242 else {
2243 $hash->{$addr} = $argument;
2244 }
2245 delete $args{$key};
2246 };
2247
2248 # If the file has a property for it, it means that the property is not
2249 # listed in the file's entries. So add a handler to the list of line
2250 # handlers to insert the property name into the lines, to provide a
2251 # uniform interface to the final processing subroutine.
2252 # the final code doesn't have to worry about that.
2253 if ($property{$addr}) {
2254 push @{$each_line_handler{$addr}}, \&_insert_property_into_line;
2255 }
2256
2257 if ($non_skip{$addr} && ! $debug_skip && $verbosity) {
2258 print "Warning: " . __PACKAGE__ . " constructor for $file{$addr} has useless 'non_skip' in it\n";
a3a8c5f0 2259 }
99870f4d 2260
09ca89ce
KW
2261 # If skipping, set to optional, and add to list of ignored files,
2262 # including its reason
2263 if ($skip{$addr}) {
2264 $optional{$addr} = 1;
1fec9f60 2265 $skipped_files{$file{$addr}} = $skip{$addr}
09ca89ce 2266 }
37e2e78e 2267
74cd47d0
KW
2268 { # On non-ascii platforms, we use a special handler
2269 no strict;
2270 no warnings 'once';
2271 *next_line = (main::NON_ASCII_PLATFORM)
2272 ? *_next_line_with_remapped_range
2273 : *_next_line;
2274 }
2275
99870f4d 2276 return $self;
d73e5302
JH
2277 }
2278
cf25bb62 2279
99870f4d
KW
2280 use overload
2281 fallback => 0,
2282 qw("") => "_operator_stringify",
2283 "." => \&main::_operator_dot,
1285127e 2284 ".=" => \&main::_operator_dot_equal,
99870f4d 2285 ;
cf25bb62 2286
99870f4d
KW
2287 sub _operator_stringify {
2288 my $self = shift;
cf25bb62 2289
99870f4d 2290 return __PACKAGE__ . " object for " . $self->file;
d73e5302 2291 }
d73e5302 2292
99870f4d
KW
2293 # flag to make sure extracted files are processed early
2294 my $seen_non_extracted_non_age = 0;
d73e5302 2295
99870f4d
KW
2296 sub run {
2297 # Process the input object $self. This opens and closes the file and
2298 # calls all the handlers for it. Currently, this can only be called
2299 # once per file, as it destroy's the EOF handler
d73e5302 2300
99870f4d
KW
2301 my $self = shift;
2302 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
b6922eda 2303
ffe43484 2304 my $addr = do { no overloading; pack 'J', $self; };
b6922eda 2305
99870f4d 2306 my $file = $file{$addr};
d73e5302 2307
99870f4d
KW
2308 # Don't process if not expecting this file (because released later
2309 # than this Unicode version), and isn't there. This means if someone
2310 # copies it into an earlier version's directory, we will go ahead and
2311 # process it.
2312 return if $first_released{$addr} gt $v_version && ! -e $file;
2313
2314 # If in debugging mode and this file doesn't have the non-skip
2315 # flag set, and isn't one of the critical files, skip it.
2316 if ($debug_skip
2317 && $first_released{$addr} ne v0
2318 && ! $non_skip{$addr})
2319 {
2320 print "Skipping $file in debugging\n" if $verbosity;
2321 return;
2322 }
2323
2324 # File could be optional
37e2e78e 2325 if ($optional{$addr}) {
99870f4d
KW
2326 return unless -e $file;
2327 my $result = eval $optional{$addr};
2328 if (! defined $result) {
2329 Carp::my_carp_bug("Got '$@' when tried to eval $optional{$addr}. $file Skipped.");
2330 return;
2331 }
2332 if (! $result) {
2333 if ($verbosity) {
2334 print STDERR "Skipping processing input file '$file' because '$optional{$addr}' is not true\n";
2335 }
2336 return;
2337 }
2338 }
2339
2340 if (! defined $file || ! -e $file) {
2341
2342 # If the file doesn't exist, see if have internal data for it
2343 # (based on first_released being 0).
2344 if ($first_released{$addr} eq v0) {
2345 $handle{$addr} = 'pretend_is_open';
2346 }
2347 else {
2348 if (! $optional{$addr} # File could be optional
2349 && $v_version ge $first_released{$addr})
2350 {
2351 print STDERR "Skipping processing input file '$file' because not found\n" if $v_version ge $first_released{$addr};
2352 }
2353 return;
2354 }
2355 }
2356 else {
2357
37e2e78e
KW
2358 # Here, the file exists. Some platforms may change the case of
2359 # its name
99870f4d 2360 if ($seen_non_extracted_non_age) {
517956bf 2361 if ($file =~ /$EXTRACTED/i) {
1675ea0d 2362 Carp::my_carp_bug(main::join_lines(<<END
99f78760 2363$file should be processed just after the 'Prop...Alias' files, and before
99870f4d
KW
2364anything not in the $EXTRACTED_DIR directory. Proceeding, but the results may
2365have subtle problems
2366END
2367 ));
2368 }
2369 }
2370 elsif ($EXTRACTED_DIR
2371 && $first_released{$addr} ne v0
517956bf
CB
2372 && $file !~ /$EXTRACTED/i
2373 && lc($file) ne 'dage.txt')
99870f4d
KW
2374 {
2375 # We don't set this (by the 'if' above) if we have no
2376 # extracted directory, so if running on an early version,
2377 # this test won't work. Not worth worrying about.
2378 $seen_non_extracted_non_age = 1;
2379 }
2380
2381 # And mark the file as having being processed, and warn if it
2382 # isn't a file we are expecting. As we process the files,
2383 # they are deleted from the hash, so any that remain at the
2384 # end of the program are files that we didn't process.
517956bf 2385 my $fkey = File::Spec->rel2abs($file);
faf3cf6b
KW
2386 my $expecting = delete $potential_files{lc($fkey)};
2387
678f13d5
KW
2388 Carp::my_carp("Was not expecting '$file'.") if
2389 ! $expecting
99870f4d
KW
2390 && ! defined $handle{$addr};
2391
37e2e78e
KW
2392 # Having deleted from expected files, we can quit if not to do
2393 # anything. Don't print progress unless really want verbosity
2394 if ($skip{$addr}) {
2395 print "Skipping $file.\n" if $verbosity >= $VERBOSE;
2396 return;
2397 }
2398
99870f4d
KW
2399 # Open the file, converting the slashes used in this program
2400 # into the proper form for the OS
2401 my $file_handle;
2402 if (not open $file_handle, "<", $file) {
2403 Carp::my_carp("Can't open $file. Skipping: $!");
2404 return 0;
2405 }
2406 $handle{$addr} = $file_handle; # Cache the open file handle
2407 }
2408
2409 if ($verbosity >= $PROGRESS) {
2410 if ($progress_message{$addr}) {
2411 print "$progress_message{$addr}\n";
2412 }
2413 else {
2414 # If using a virtual file, say so.
2415 print "Processing ", (-e $file)
2416 ? $file
2417 : "substitute $file",
2418 "\n";
2419 }
2420 }
2421
2422
2423 # Call any special handler for before the file.
2424 &{$pre_handler{$addr}}($self) if $pre_handler{$addr};
2425
2426 # Then the main handler
2427 &{$handler{$addr}}($self);
2428
2429 # Then any special post-file handler.
2430 &{$post_handler{$addr}}($self) if $post_handler{$addr};
2431
2432 # If any errors have been accumulated, output the counts (as the first
2433 # error message in each class was output when it was encountered).
2434 if ($errors{$addr}) {
2435 my $total = 0;
2436 my $types = 0;
2437 foreach my $error (keys %{$errors{$addr}}) {
2438 $total += $errors{$addr}->{$error};
2439 delete $errors{$addr}->{$error};
2440 $types++;
2441 }
2442 if ($total > 1) {
2443 my $message
2444 = "A total of $total lines had errors in $file. ";
2445
2446 $message .= ($types == 1)
2447 ? '(Only the first one was displayed.)'
2448 : '(Only the first of each type was displayed.)';
2449 Carp::my_carp($message);
2450 }
2451 }
2452
2453 if (@{$missings{$addr}}) {
2454 Carp::my_carp_bug("Handler for $file didn't look at all the \@missing lines. Generated tables likely are wrong");
2455 }
2456
2457 # If a real file handle, close it.
2458 close $handle{$addr} or Carp::my_carp("Can't close $file: $!") if
2459 ref $handle{$addr};
2460 $handle{$addr} = ""; # Uses empty to indicate that has already seen
2461 # the file, as opposed to undef
2462 return;
2463 }
2464
74cd47d0 2465 sub _next_line {
99870f4d
KW
2466 # Sets $_ to be the next logical input line, if any. Returns non-zero
2467 # if such a line exists. 'logical' means that any lines that have
2468 # been added via insert_lines() will be returned in $_ before the file
2469 # is read again.
2470
2471 my $self = shift;
2472 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2473
ffe43484 2474 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2475
2476 # Here the file is open (or if the handle is not a ref, is an open
2477 # 'virtual' file). Get the next line; any inserted lines get priority
2478 # over the file itself.
2479 my $adjusted;
2480
2481 LINE:
2482 while (1) { # Loop until find non-comment, non-empty line
2483 #local $to_trace = 1 if main::DEBUG;
2484 my $inserted_ref = shift @{$added_lines{$addr}};
2485 if (defined $inserted_ref) {
2486 ($adjusted, $_) = @{$inserted_ref};
2487 trace $adjusted, $_ if main::DEBUG && $to_trace;
2488 return 1 if $adjusted;
2489 }
2490 else {
2491 last if ! ref $handle{$addr}; # Don't read unless is real file
2492 last if ! defined ($_ = readline $handle{$addr});
2493 }
2494 chomp;
2495 trace $_ if main::DEBUG && $to_trace;
2496
2497 # See if this line is the comment line that defines what property
2498 # value that code points that are not listed in the file should
2499 # have. The format or existence of these lines is not guaranteed
2500 # by Unicode since they are comments, but the documentation says
2501 # that this was added for machine-readability, so probably won't
2502 # change. This works starting in Unicode Version 5.0. They look
2503 # like:
2504 #
2505 # @missing: 0000..10FFFF; Not_Reordered
2506 # @missing: 0000..10FFFF; Decomposition_Mapping; <code point>
2507 # @missing: 0000..10FFFF; ; NaN
2508 #
2509 # Save the line for a later get_missings() call.
2510 if (/$missing_defaults_prefix/) {
2511 if ($has_missings_defaults{$addr} == $NO_DEFAULTS) {
2512 $self->carp_bad_line("Unexpected \@missing line. Assuming no missing entries");
2513 }
2514 elsif ($has_missings_defaults{$addr} == $NOT_IGNORED) {
2515 my @defaults = split /\s* ; \s*/x, $_;
2516
2517 # The first field is the @missing, which ends in a
2518 # semi-colon, so can safely shift.
2519 shift @defaults;
2520
2521 # Some of these lines may have empty field placeholders
2522 # which get in the way. An example is:
2523 # @missing: 0000..10FFFF; ; NaN
2524 # Remove them. Process starting from the top so the
2525 # splice doesn't affect things still to be looked at.
2526 for (my $i = @defaults - 1; $i >= 0; $i--) {
2527 next if $defaults[$i] ne "";
2528 splice @defaults, $i, 1;
2529 }
2530
2531 # What's left should be just the property (maybe) and the
2532 # default. Having only one element means it doesn't have
2533 # the property.
2534 my $default;
2535 my $property;
2536 if (@defaults >= 1) {
2537 if (@defaults == 1) {
2538 $default = $defaults[0];
2539 }
2540 else {
2541 $property = $defaults[0];
2542 $default = $defaults[1];
2543 }
2544 }
2545
2546 if (@defaults < 1
2547 || @defaults > 2
2548 || ($default =~ /^</
2549 && $default !~ /^<code *point>$/i
09f8d0ac
KW
2550 && $default !~ /^<none>$/i
2551 && $default !~ /^<script>$/i))
99870f4d
KW
2552 {
2553 $self->carp_bad_line("Unrecognized \@missing line: $_. Assuming no missing entries");
2554 }
2555 else {
2556
2557 # If the property is missing from the line, it should
2558 # be the one for the whole file
2559 $property = $property{$addr} if ! defined $property;
2560
2561 # Change <none> to the null string, which is what it
2562 # really means. If the default is the code point
2563 # itself, set it to <code point>, which is what
2564 # Unicode uses (but sometimes they've forgotten the
2565 # space)
2566 if ($default =~ /^<none>$/i) {
2567 $default = "";
2568 }
2569 elsif ($default =~ /^<code *point>$/i) {
2570 $default = $CODE_POINT;
2571 }
09f8d0ac
KW
2572 elsif ($default =~ /^<script>$/i) {
2573
2574 # Special case this one. Currently is from
2575 # ScriptExtensions.txt, and means for all unlisted
2576 # code points, use their Script property values.
2577 # For the code points not listed in that file, the
2578 # default value is 'Unknown'.
2579 $default = "Unknown";
2580 }
99870f4d
KW
2581
2582 # Store them as a sub-arrays with both components.
2583 push @{$missings{$addr}}, [ $default, $property ];
2584 }
2585 }
2586
2587 # There is nothing for the caller to process on this comment
2588 # line.
2589 next;
2590 }
2591
2592 # Remove comments and trailing space, and skip this line if the
2593 # result is empty
2594 s/#.*//;
2595 s/\s+$//;
2596 next if /^$/;
2597
2598 # Call any handlers for this line, and skip further processing of
2599 # the line if the handler sets the line to null.
2600 foreach my $sub_ref (@{$each_line_handler{$addr}}) {
2601 &{$sub_ref}($self);
2602 next LINE if /^$/;
2603 }
2604
2605 # Here the line is ok. return success.
2606 return 1;
2607 } # End of looping through lines.
2608
2609 # If there is an EOF handler, call it (only once) and if it generates
2610 # more lines to process go back in the loop to handle them.
2611 if ($eof_handler{$addr}) {
2612 &{$eof_handler{$addr}}($self);
2613 $eof_handler{$addr} = ""; # Currently only get one shot at it.
2614 goto LINE if $added_lines{$addr};
2615 }
2616
2617 # Return failure -- no more lines.
2618 return 0;
2619
2620 }
2621
74cd47d0
KW
2622 sub _next_line_with_remapped_range {
2623 my $self = shift;
2624 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2625
2626 # like _next_line(), but for use on non-ASCII platforms. It sets $_
2627 # to be the next logical input line, if any. Returns non-zero if such
2628 # a line exists. 'logical' means that any lines that have been added
2629 # via insert_lines() will be returned in $_ before the file is read
2630 # again.
2631 #
2632 # The difference from _next_line() is that this remaps the Unicode
2633 # code points in the input to those of the native platform. Each
2634 # input line contains a single code point, or a single contiguous
2635 # range of them This routine splits each range into its individual
2636 # code points and caches them. It returns the cached values,
2637 # translated into their native equivalents, one at a time, for each
2638 # call, before reading the next line. Since native values can only be
2639 # a single byte wide, no translation is needed for code points above
2640 # 0xFF, and ranges that are entirely above that number are not split.
2641 # If an input line contains the range 254-1000, it would be split into
2642 # three elements: 254, 255, and 256-1000. (The downstream table
2643 # insertion code will sort and coalesce the individual code points
2644 # into appropriate ranges.)
2645
2646 my $addr = do { no overloading; pack 'J', $self; };
2647
2648 while (1) {
2649
2650 # Look in cache before reading the next line. Return any cached
2651 # value, translated
2652 my $inserted = shift @{$remapped_lines{$addr}};
2653 if (defined $inserted) {
2654 trace $inserted if main::DEBUG && $to_trace;
2655 $_ = $inserted =~ s/^ ( \d+ ) /sprintf("%04X", utf8::unicode_to_native($1))/xer;
2656 trace $_ if main::DEBUG && $to_trace;
2657 return 1;
2658 }
2659
2660 # Get the next line.
2661 return 0 unless _next_line($self);
2662
2663 # If there is a special handler for it, return the line,
2664 # untranslated. This should happen only for files that are
2665 # special, not being code-point related, such as property names.
2666 return 1 if $handler{$addr}
2667 != \&main::process_generic_property_file;
2668
2669 my ($range, $property_name, $map, @remainder)
2670 = split /\s*;\s*/, $_, -1; # -1 => retain trailing null fields
2671
2672 if (@remainder
2673 || ! defined $property_name
2674 || $range !~ /^ ($code_point_re) (?:\.\. ($code_point_re) )? $/x)
2675 {
2676 Carp::my_carp_bug("Unrecognized input line '$_'. Ignored");
2677 }
2678
2679 my $low = hex $1;
2680 my $high = (defined $2) ? hex $2 : $low;
2681
2682 # If the input maps the range to another code point, remap the
2683 # target if it is between 0 and 255.
2684 my $tail;
2685 if (defined $map) {
2686 $map =~ s/\b 00 ( [0-9A-F]{2} ) \b/sprintf("%04X", utf8::unicode_to_native(hex $1))/gxe;
2687 $tail = "$property_name; $map";
2688 $_ = "$range; $tail";
2689 }
2690 else {
2691 $tail = $property_name;
2692 }
2693
2694 # If entire range is above 255, just return it, unchanged (except
2695 # any mapped-to code point, already changed above)
2696 return 1 if $low > 255;
2697
2698 # Cache an entry for every code point < 255. For those in the
2699 # range above 255, return a dummy entry for just that portion of
2700 # the range. Note that this will be out-of-order, but that is not
2701 # a problem.
2702 foreach my $code_point ($low .. $high) {
2703 if ($code_point > 255) {
2704 $_ = sprintf "%04X..%04X; $tail", $code_point, $high;
2705 return 1;
2706 }
2707 push @{$remapped_lines{$addr}}, "$code_point; $tail";
2708 }
2709 } # End of looping through lines.
2710
2711 # NOTREACHED
2712 }
2713
99870f4d
KW
2714# Not currently used, not fully tested.
2715# sub peek {
2716# # Non-destructive look-ahead one non-adjusted, non-comment, non-blank
2717# # record. Not callable from an each_line_handler(), nor does it call
2718# # an each_line_handler() on the line.
2719#
2720# my $self = shift;
ffe43484 2721# my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2722#
2723# foreach my $inserted_ref (@{$added_lines{$addr}}) {
2724# my ($adjusted, $line) = @{$inserted_ref};
2725# next if $adjusted;
2726#
2727# # Remove comments and trailing space, and return a non-empty
2728# # resulting line
2729# $line =~ s/#.*//;
2730# $line =~ s/\s+$//;
2731# return $line if $line ne "";
2732# }
2733#
2734# return if ! ref $handle{$addr}; # Don't read unless is real file
2735# while (1) { # Loop until find non-comment, non-empty line
2736# local $to_trace = 1 if main::DEBUG;
2737# trace $_ if main::DEBUG && $to_trace;
2738# return if ! defined (my $line = readline $handle{$addr});
2739# chomp $line;
2740# push @{$added_lines{$addr}}, [ 0, $line ];
2741#
2742# $line =~ s/#.*//;
2743# $line =~ s/\s+$//;
2744# return $line if $line ne "";
2745# }
2746#
2747# return;
2748# }
2749
2750
2751 sub insert_lines {
2752 # Lines can be inserted so that it looks like they were in the input
2753 # file at the place it was when this routine is called. See also
2754 # insert_adjusted_lines(). Lines inserted via this routine go through
2755 # any each_line_handler()
2756
2757 my $self = shift;
2758
2759 # Each inserted line is an array, with the first element being 0 to
2760 # indicate that this line hasn't been adjusted, and needs to be
2761 # processed.
f998e60c 2762 no overloading;
051df77b 2763 push @{$added_lines{pack 'J', $self}}, map { [ 0, $_ ] } @_;
99870f4d
KW
2764 return;
2765 }
2766
2767 sub insert_adjusted_lines {
2768 # Lines can be inserted so that it looks like they were in the input
2769 # file at the place it was when this routine is called. See also
2770 # insert_lines(). Lines inserted via this routine are already fully
2771 # adjusted, ready to be processed; each_line_handler()s handlers will
2772 # not be called. This means this is not a completely general
2773 # facility, as only the last each_line_handler on the stack should
2774 # call this. It could be made more general, by passing to each of the
2775 # line_handlers their position on the stack, which they would pass on
2776 # to this routine, and that would replace the boolean first element in
2777 # the anonymous array pushed here, so that the next_line routine could
2778 # use that to call only those handlers whose index is after it on the
2779 # stack. But this is overkill for what is needed now.
2780
2781 my $self = shift;
2782 trace $_[0] if main::DEBUG && $to_trace;
2783
2784 # Each inserted line is an array, with the first element being 1 to
2785 # indicate that this line has been adjusted
f998e60c 2786 no overloading;
051df77b 2787 push @{$added_lines{pack 'J', $self}}, map { [ 1, $_ ] } @_;
99870f4d
KW
2788 return;
2789 }
2790
2791 sub get_missings {
2792 # Returns the stored up @missings lines' values, and clears the list.
2793 # The values are in an array, consisting of the default in the first
2794 # element, and the property in the 2nd. However, since these lines
2795 # can be stacked up, the return is an array of all these arrays.
2796
2797 my $self = shift;
2798 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2799
ffe43484 2800 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2801
2802 # If not accepting a list return, just return the first one.
2803 return shift @{$missings{$addr}} unless wantarray;
2804
2805 my @return = @{$missings{$addr}};
2806 undef @{$missings{$addr}};
2807 return @return;
2808 }
2809
2810 sub _insert_property_into_line {
2811 # Add a property field to $_, if this file requires it.
2812
f998e60c 2813 my $self = shift;
ffe43484 2814 my $addr = do { no overloading; pack 'J', $self; };
f998e60c 2815 my $property = $property{$addr};
99870f4d
KW
2816 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2817
2818 $_ =~ s/(;|$)/; $property$1/;
2819 return;
2820 }
2821
2822 sub carp_bad_line {
2823 # Output consistent error messages, using either a generic one, or the
2824 # one given by the optional parameter. To avoid gazillions of the
2825 # same message in case the syntax of a file is way off, this routine
2826 # only outputs the first instance of each message, incrementing a
2827 # count so the totals can be output at the end of the file.
2828
2829 my $self = shift;
2830 my $message = shift;
2831 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2832
ffe43484 2833 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2834
2835 $message = 'Unexpected line' unless $message;
2836
2837 # No trailing punctuation so as to fit with our addenda.
2838 $message =~ s/[.:;,]$//;
2839
2840 # If haven't seen this exact message before, output it now. Otherwise
2841 # increment the count of how many times it has occurred
2842 unless ($errors{$addr}->{$message}) {
2843 Carp::my_carp("$message in '$_' in "
f998e60c 2844 . $file{$addr}
99870f4d
KW
2845 . " at line $.. Skipping this line;");
2846 $errors{$addr}->{$message} = 1;
2847 }
2848 else {
2849 $errors{$addr}->{$message}++;
2850 }
2851
2852 # Clear the line to prevent any further (meaningful) processing of it.
2853 $_ = "";
2854
2855 return;
2856 }
2857} # End closure
2858
2859package Multi_Default;
2860
2861# Certain properties in early versions of Unicode had more than one possible
2862# default for code points missing from the files. In these cases, one
2863# default applies to everything left over after all the others are applied,
2864# and for each of the others, there is a description of which class of code
2865# points applies to it. This object helps implement this by storing the
2866# defaults, and for all but that final default, an eval string that generates
2867# the class that it applies to.
2868
2869
2870{ # Closure
2871
2872 main::setup_package();
2873
2874 my %class_defaults;
2875 # The defaults structure for the classes
2876 main::set_access('class_defaults', \%class_defaults);
2877
2878 my %other_default;
2879 # The default that applies to everything left over.
2880 main::set_access('other_default', \%other_default, 'r');
2881
2882
2883 sub new {
2884 # The constructor is called with default => eval pairs, terminated by
2885 # the left-over default. e.g.
2886 # Multi_Default->new(
2887 # 'T' => '$gc->table("Mn") + $gc->table("Cf") - 0x200C
2888 # - 0x200D',
2889 # 'R' => 'some other expression that evaluates to code points',
2890 # .
2891 # .
2892 # .
2893 # 'U'));
2894
2895 my $class = shift;
2896
2897 my $self = bless \do{my $anonymous_scalar}, $class;
ffe43484 2898 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2899
2900 while (@_ > 1) {
2901 my $default = shift;
2902 my $eval = shift;
2903 $class_defaults{$addr}->{$default} = $eval;
2904 }
2905
2906 $other_default{$addr} = shift;
2907
2908 return $self;
2909 }
2910
2911 sub get_next_defaults {
2912 # Iterates and returns the next class of defaults.
2913 my $self = shift;
2914 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2915
ffe43484 2916 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2917
2918 return each %{$class_defaults{$addr}};
2919 }
2920}
2921
2922package Alias;
2923
2924# An alias is one of the names that a table goes by. This class defines them
2925# including some attributes. Everything is currently setup in the
2926# constructor.
2927
2928
2929{ # Closure
2930
2931 main::setup_package();
2932
2933 my %name;
2934 main::set_access('name', \%name, 'r');
2935
2936 my %loose_match;
c12f2655 2937 # Should this name match loosely or not.
99870f4d
KW
2938 main::set_access('loose_match', \%loose_match, 'r');
2939
33e96e72
KW
2940 my %make_re_pod_entry;
2941 # Some aliases should not get their own entries in the re section of the
2942 # pod, because they are covered by a wild-card, and some we want to
2943 # discourage use of. Binary
f82fe4ba 2944 main::set_access('make_re_pod_entry', \%make_re_pod_entry, 'r', 's');
99870f4d 2945
fd1e3e84
KW
2946 my %ucd;
2947 # Is this documented to be accessible via Unicode::UCD
2948 main::set_access('ucd', \%ucd, 'r', 's');
2949
99870f4d
KW
2950 my %status;
2951 # Aliases have a status, like deprecated, or even suppressed (which means
2952 # they don't appear in documentation). Enum
2953 main::set_access('status', \%status, 'r');
2954
0eac1e20 2955 my %ok_as_filename;
99870f4d
KW
2956 # Similarly, some aliases should not be considered as usable ones for
2957 # external use, such as file names, or we don't want documentation to
2958 # recommend them. Boolean
0eac1e20 2959 main::set_access('ok_as_filename', \%ok_as_filename, 'r');
99870f4d
KW
2960
2961 sub new {
2962 my $class = shift;
2963
2964 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2965 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2966
2967 $name{$addr} = shift;
2968 $loose_match{$addr} = shift;
33e96e72 2969 $make_re_pod_entry{$addr} = shift;
0eac1e20 2970 $ok_as_filename{$addr} = shift;
99870f4d 2971 $status{$addr} = shift;
fd1e3e84 2972 $ucd{$addr} = shift;
99870f4d
KW
2973
2974 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2975
2976 # Null names are never ok externally
0eac1e20 2977 $ok_as_filename{$addr} = 0 if $name{$addr} eq "";
99870f4d
KW
2978
2979 return $self;
2980 }
2981}
2982
2983package Range;
2984
2985# A range is the basic unit for storing code points, and is described in the
2986# comments at the beginning of the program. Each range has a starting code
2987# point; an ending code point (not less than the starting one); a value
2988# that applies to every code point in between the two end-points, inclusive;
2989# and an enum type that applies to the value. The type is for the user's
2990# convenience, and has no meaning here, except that a non-zero type is
2991# considered to not obey the normal Unicode rules for having standard forms.
2992#
2993# The same structure is used for both map and match tables, even though in the
2994# latter, the value (and hence type) is irrelevant and could be used as a
2995# comment. In map tables, the value is what all the code points in the range
2996# map to. Type 0 values have the standardized version of the value stored as
2997# well, so as to not have to recalculate it a lot.
2998
2999sub trace { return main::trace(@_); }
3000
3001{ # Closure
3002
3003 main::setup_package();
3004
3005 my %start;
3006 main::set_access('start', \%start, 'r', 's');
3007
3008 my %end;
3009 main::set_access('end', \%end, 'r', 's');
3010
3011 my %value;
3012 main::set_access('value', \%value, 'r');
3013
3014 my %type;
3015 main::set_access('type', \%type, 'r');
3016
3017 my %standard_form;
3018 # The value in internal standard form. Defined only if the type is 0.
3019 main::set_access('standard_form', \%standard_form);
3020
3021 # Note that if these fields change, the dump() method should as well
3022
3023 sub new {
3024 return Carp::carp_too_few_args(\@_, 3) if main::DEBUG && @_ < 3;
3025 my $class = shift;
3026
3027 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 3028 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3029
3030 $start{$addr} = shift;
3031 $end{$addr} = shift;
3032
3033 my %args = @_;
3034
3035 my $value = delete $args{'Value'}; # Can be 0
3036 $value = "" unless defined $value;
3037 $value{$addr} = $value;
3038
3039 $type{$addr} = delete $args{'Type'} || 0;
3040
3041 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3042
99870f4d
KW
3043 return $self;
3044 }
3045
3046 use overload
3047 fallback => 0,
3048 qw("") => "_operator_stringify",
3049 "." => \&main::_operator_dot,
1285127e 3050 ".=" => \&main::_operator_dot_equal,
99870f4d
KW
3051 ;
3052
3053 sub _operator_stringify {
3054 my $self = shift;
ffe43484 3055 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3056
3057 # Output it like '0041..0065 (value)'
3058 my $return = sprintf("%04X", $start{$addr})
3059 . '..'
3060 . sprintf("%04X", $end{$addr});
3061 my $value = $value{$addr};
3062 my $type = $type{$addr};
3063 $return .= ' (';
3064 $return .= "$value";
3065 $return .= ", Type=$type" if $type != 0;
3066 $return .= ')';
3067
3068 return $return;
3069 }
3070
3071 sub standard_form {
c292d35a
NC
3072 # Calculate the standard form only if needed, and cache the result.
3073 # The standard form is the value itself if the type is special.
3074 # This represents a considerable CPU and memory saving - at the time
3075 # of writing there are 368676 non-special objects, but the standard
3076 # form is only requested for 22047 of them - ie about 6%.
99870f4d
KW
3077
3078 my $self = shift;
3079 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3080
ffe43484 3081 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3082
3083 return $standard_form{$addr} if defined $standard_form{$addr};
c292d35a
NC
3084
3085 my $value = $value{$addr};
3086 return $value if $type{$addr};
3087 return $standard_form{$addr} = main::standardize($value);
99870f4d
KW
3088 }
3089
3090 sub dump {
3091 # Human, not machine readable. For machine readable, comment out this
3092 # entire routine and let the standard one take effect.
3093 my $self = shift;
3094 my $indent = shift;
3095 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3096
ffe43484 3097 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3098
3099 my $return = $indent
3100 . sprintf("%04X", $start{$addr})
3101 . '..'
3102 . sprintf("%04X", $end{$addr})
3103 . " '$value{$addr}';";
3104 if (! defined $standard_form{$addr}) {
3105 $return .= "(type=$type{$addr})";
3106 }
3107 elsif ($standard_form{$addr} ne $value{$addr}) {
3108 $return .= "(standard '$standard_form{$addr}')";
3109 }
3110 return $return;
3111 }
3112} # End closure
3113
3114package _Range_List_Base;
3115
3116# Base class for range lists. A range list is simply an ordered list of
3117# ranges, so that the ranges with the lowest starting numbers are first in it.
3118#
3119# When a new range is added that is adjacent to an existing range that has the
3120# same value and type, it merges with it to form a larger range.
3121#
3122# Ranges generally do not overlap, except that there can be multiple entries
3123# of single code point ranges. This is because of NameAliases.txt.
3124#
3125# In this program, there is a standard value such that if two different
3126# values, have the same standard value, they are considered equivalent. This
3127# value was chosen so that it gives correct results on Unicode data
3128
3129# There are a number of methods to manipulate range lists, and some operators
3130# are overloaded to handle them.
3131
99870f4d
KW
3132sub trace { return main::trace(@_); }
3133
3134{ # Closure
3135
3136 our $addr;
3137
5b348b71
KW
3138 # Max is initialized to a negative value that isn't adjacent to 0, for
3139 # simpler tests
3140 my $max_init = -2;
3141
99870f4d
KW
3142 main::setup_package();
3143
3144 my %ranges;
3145 # The list of ranges
3146 main::set_access('ranges', \%ranges, 'readable_array');
3147
3148 my %max;
3149 # The highest code point in the list. This was originally a method, but
3150 # actual measurements said it was used a lot.
3151 main::set_access('max', \%max, 'r');
3152
3153 my %each_range_iterator;
3154 # Iterator position for each_range()
3155 main::set_access('each_range_iterator', \%each_range_iterator);
3156
3157 my %owner_name_of;
3158 # Name of parent this is attached to, if any. Solely for better error
3159 # messages.
3160 main::set_access('owner_name_of', \%owner_name_of, 'p_r');
3161
3162 my %_search_ranges_cache;
3163 # A cache of the previous result from _search_ranges(), for better
3164 # performance
3165 main::set_access('_search_ranges_cache', \%_search_ranges_cache);
3166
3167 sub new {
3168 my $class = shift;
3169 my %args = @_;
3170
3171 # Optional initialization data for the range list.
3172 my $initialize = delete $args{'Initialize'};
3173
3174 my $self;
3175
3176 # Use _union() to initialize. _union() returns an object of this
3177 # class, which means that it will call this constructor recursively.
3178 # But it won't have this $initialize parameter so that it won't
3179 # infinitely loop on this.
3180 return _union($class, $initialize, %args) if defined $initialize;
3181
3182 $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 3183 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3184
3185 # Optional parent object, only for debug info.
3186 $owner_name_of{$addr} = delete $args{'Owner'};
3187 $owner_name_of{$addr} = "" if ! defined $owner_name_of{$addr};
3188
3189 # Stringify, in case it is an object.
3190 $owner_name_of{$addr} = "$owner_name_of{$addr}";
3191
3192 # This is used only for error messages, and so a colon is added
3193 $owner_name_of{$addr} .= ": " if $owner_name_of{$addr} ne "";
3194
3195 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3196
5b348b71 3197 $max{$addr} = $max_init;
99870f4d
KW
3198
3199 $_search_ranges_cache{$addr} = 0;
3200 $ranges{$addr} = [];
3201
3202 return $self;
3203 }
3204
3205 use overload
3206 fallback => 0,
3207 qw("") => "_operator_stringify",
3208 "." => \&main::_operator_dot,
1285127e 3209 ".=" => \&main::_operator_dot_equal,
99870f4d
KW
3210 ;
3211
3212 sub _operator_stringify {
3213 my $self = shift;
ffe43484 3214 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3215
3216 return "Range_List attached to '$owner_name_of{$addr}'"
3217 if $owner_name_of{$addr};
3218 return "anonymous Range_List " . \$self;
3219 }
3220
3221 sub _union {
3222 # Returns the union of the input code points. It can be called as
3223 # either a constructor or a method. If called as a method, the result
3224 # will be a new() instance of the calling object, containing the union
3225 # of that object with the other parameter's code points; if called as
d59563d0 3226 # a constructor, the first parameter gives the class that the new object
99870f4d
KW
3227 # should be, and the second parameter gives the code points to go into
3228 # it.
3229 # In either case, there are two parameters looked at by this routine;
3230 # any additional parameters are passed to the new() constructor.
3231 #
3232 # The code points can come in the form of some object that contains
3233 # ranges, and has a conventionally named method to access them; or
3234 # they can be an array of individual code points (as integers); or
3235 # just a single code point.
3236 #
3237 # If they are ranges, this routine doesn't make any effort to preserve
3198cc57
KW
3238 # the range values and types of one input over the other. Therefore
3239 # this base class should not allow _union to be called from other than
99870f4d
KW
3240 # initialization code, so as to prevent two tables from being added
3241 # together where the range values matter. The general form of this
3242 # routine therefore belongs in a derived class, but it was moved here
3243 # to avoid duplication of code. The failure to overload this in this
3244 # class keeps it safe.
3198cc57
KW
3245 #
3246 # It does make the effort during initialization to accept tables with
3247 # multiple values for the same code point, and to preserve the order
3248 # of these. If there is only one input range or range set, it doesn't
3249 # sort (as it should already be sorted to the desired order), and will
3250 # accept multiple values per code point. Otherwise it will merge
3251 # multiple values into a single one.
99870f4d
KW
3252
3253 my $self;
3254 my @args; # Arguments to pass to the constructor
3255
3256 my $class = shift;
3257
3258 # If a method call, will start the union with the object itself, and
3259 # the class of the new object will be the same as self.
3260 if (ref $class) {
3261 $self = $class;
3262 $class = ref $self;
3263 push @args, $self;
3264 }
3265
3266 # Add the other required parameter.
3267 push @args, shift;
3268 # Rest of parameters are passed on to the constructor
3269
3270 # Accumulate all records from both lists.
3271 my @records;
3198cc57 3272 my $input_count = 0;
99870f4d
KW
3273 for my $arg (@args) {
3274 #local $to_trace = 0 if main::DEBUG;
3275 trace "argument = $arg" if main::DEBUG && $to_trace;
3276 if (! defined $arg) {
3277 my $message = "";
3278 if (defined $self) {
f998e60c 3279 no overloading;
051df77b 3280 $message .= $owner_name_of{pack 'J', $self};
99870f4d 3281 }
ada6088e 3282 Carp::my_carp_bug($message . "Undefined argument to _union. No union done.");
99870f4d
KW
3283 return;
3284 }
3198cc57 3285
99870f4d
KW
3286 $arg = [ $arg ] if ! ref $arg;
3287 my $type = ref $arg;
3288 if ($type eq 'ARRAY') {
3289 foreach my $element (@$arg) {
3290 push @records, Range->new($element, $element);
3198cc57 3291 $input_count++;
99870f4d
KW
3292 }
3293 }
3294 elsif ($arg->isa('Range')) {
3295 push @records, $arg;
3198cc57 3296 $input_count++;
99870f4d
KW
3297 }
3298 elsif ($arg->can('ranges')) {
3299 push @records, $arg->ranges;
3198cc57 3300 $input_count++;
99870f4d
KW
3301 }
3302 else {
3303 my $message = "";
3304 if (defined $self) {
f998e60c 3305 no overloading;
051df77b 3306 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
3307 }
3308 Carp::my_carp_bug($message . "Cannot take the union of a $type. No union done.");
3309 return;
3310 }
3311 }
3312
3313 # Sort with the range containing the lowest ordinal first, but if
3314 # two ranges start at the same code point, sort with the bigger range
3315 # of the two first, because it takes fewer cycles.
3198cc57
KW
3316 if ($input_count > 1) {
3317 @records = sort { ($a->start <=> $b->start)
99870f4d
KW
3318 or
3319 # if b is shorter than a, b->end will be
3320 # less than a->end, and we want to select
3321 # a, so want to return -1
3322 ($b->end <=> $a->end)
3323 } @records;
3198cc57 3324 }
99870f4d
KW
3325
3326 my $new = $class->new(@_);
3327
3328 # Fold in records so long as they add new information.
3329 for my $set (@records) {
3330 my $start = $set->start;
3331 my $end = $set->end;
d59563d0 3332 my $value = $set->value;
3198cc57 3333 my $type = $set->type;
99870f4d 3334 if ($start > $new->max) {
3198cc57 3335 $new->_add_delete('+', $start, $end, $value, Type => $type);
99870f4d
KW
3336 }
3337 elsif ($end > $new->max) {
3198cc57
KW
3338 $new->_add_delete('+', $new->max +1, $end, $value,
3339 Type => $type);
3340 }
3341 elsif ($input_count == 1) {
3342 # Here, overlaps existing range, but is from a single input,
3343 # so preserve the multiple values from that input.
3344 $new->_add_delete('+', $start, $end, $value, Type => $type,
3345 Replace => $MULTIPLE_AFTER);
99870f4d
KW
3346 }
3347 }
3348
3349 return $new;
3350 }
3351
3352 sub range_count { # Return the number of ranges in the range list
3353 my $self = shift;
3354 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3355
f998e60c 3356 no overloading;
051df77b 3357 return scalar @{$ranges{pack 'J', $self}};
99870f4d
KW
3358 }
3359
3360 sub min {
3361 # Returns the minimum code point currently in the range list, or if
3362 # the range list is empty, 2 beyond the max possible. This is a
3363 # method because used so rarely, that not worth saving between calls,
3364 # and having to worry about changing it as ranges are added and
3365 # deleted.
3366
3367 my $self = shift;
3368 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3369
ffe43484 3370 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3371
3372 # If the range list is empty, return a large value that isn't adjacent
3373 # to any that could be in the range list, for simpler tests
6189eadc 3374 return $MAX_UNICODE_CODEPOINT + 2 unless scalar @{$ranges{$addr}};
99870f4d
KW
3375 return $ranges{$addr}->[0]->start;
3376 }
3377
3378 sub contains {
3379 # Boolean: Is argument in the range list? If so returns $i such that:
3380 # range[$i]->end < $codepoint <= range[$i+1]->end
3381 # which is one beyond what you want; this is so that the 0th range
3382 # doesn't return false
3383 my $self = shift;
3384 my $codepoint = shift;
3385 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3386
99870f4d
KW
3387 my $i = $self->_search_ranges($codepoint);
3388 return 0 unless defined $i;
3389
3390 # The search returns $i, such that
3391 # range[$i-1]->end < $codepoint <= range[$i]->end
3392 # So is in the table if and only iff it is at least the start position
3393 # of range $i.
f998e60c 3394 no overloading;
051df77b 3395 return 0 if $ranges{pack 'J', $self}->[$i]->start > $codepoint;
99870f4d
KW
3396 return $i + 1;
3397 }
3398
2f7a8815
KW
3399 sub containing_range {
3400 # Returns the range object that contains the code point, undef if none
3401
3402 my $self = shift;
3403 my $codepoint = shift;
3404 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3405
3406 my $i = $self->contains($codepoint);
3407 return unless $i;
3408
3409 # contains() returns 1 beyond where we should look
3410 no overloading;
3411 return $ranges{pack 'J', $self}->[$i-1];
3412 }
3413
99870f4d
KW
3414 sub value_of {
3415 # Returns the value associated with the code point, undef if none
3416
3417 my $self = shift;
3418 my $codepoint = shift;
3419 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3420
d69c231b
KW
3421 my $range = $self->containing_range($codepoint);
3422 return unless defined $range;
99870f4d 3423
d69c231b 3424 return $range->value;
99870f4d
KW
3425 }
3426
0a9dbafc
KW
3427 sub type_of {
3428 # Returns the type of the range containing the code point, undef if
3429 # the code point is not in the table
3430
3431 my $self = shift;
3432 my $codepoint = shift;
3433 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3434
3435 my $range = $self->containing_range($codepoint);
3436 return unless defined $range;
3437
3438 return $range->type;
3439 }
3440
99870f4d
KW
3441 sub _search_ranges {
3442 # Find the range in the list which contains a code point, or where it
3443 # should go if were to add it. That is, it returns $i, such that:
3444 # range[$i-1]->end < $codepoint <= range[$i]->end
3445 # Returns undef if no such $i is possible (e.g. at end of table), or
3446 # if there is an error.
3447
3448 my $self = shift;
3449 my $code_point = shift;
3450 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3451
ffe43484 3452 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3453
3454 return if $code_point > $max{$addr};
3455 my $r = $ranges{$addr}; # The current list of ranges
3456 my $range_list_size = scalar @$r;
3457 my $i;
3458
3459 use integer; # want integer division
3460
3461 # Use the cached result as the starting guess for this one, because,
3462 # an experiment on 5.1 showed that 90% of the time the cache was the
3463 # same as the result on the next call (and 7% it was one less).
3464 $i = $_search_ranges_cache{$addr};
3465 $i = 0 if $i >= $range_list_size; # Reset if no longer valid (prob.
3466 # from an intervening deletion
3467 #local $to_trace = 1 if main::DEBUG;
3468 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);
3469 return $i if $code_point <= $r->[$i]->end
3470 && ($i == 0 || $r->[$i-1]->end < $code_point);
3471
3472 # Here the cache doesn't yield the correct $i. Try adding 1.
3473 if ($i < $range_list_size - 1
3474 && $r->[$i]->end < $code_point &&
3475 $code_point <= $r->[$i+1]->end)
3476 {
3477 $i++;
3478 trace "next \$i is correct: $i" if main::DEBUG && $to_trace;
3479 $_search_ranges_cache{$addr} = $i;
3480 return $i;
3481 }
3482
3483 # Here, adding 1 also didn't work. We do a binary search to
3484 # find the correct position, starting with current $i
3485 my $lower = 0;
3486 my $upper = $range_list_size - 1;
3487 while (1) {
3488 trace "top of loop i=$i:", sprintf("%04X", $r->[$lower]->start), "[$lower] .. ", sprintf("%04X", $r->[$i]->start), "[$i] .. ", sprintf("%04X", $r->[$upper]->start), "[$upper]" if main::DEBUG && $to_trace;
3489
3490 if ($code_point <= $r->[$i]->end) {
3491
3492 # Here we have met the upper constraint. We can quit if we
3493 # also meet the lower one.
3494 last if $i == 0 || $r->[$i-1]->end < $code_point;
3495
3496 $upper = $i; # Still too high.
3497
3498 }
3499 else {
3500
3501 # Here, $r[$i]->end < $code_point, so look higher up.
3502 $lower = $i;
3503 }
3504
3505 # Split search domain in half to try again.
3506 my $temp = ($upper + $lower) / 2;
3507
3508 # No point in continuing unless $i changes for next time
3509 # in the loop.
3510 if ($temp == $i) {
3511
3512 # We can't reach the highest element because of the averaging.
3513 # So if one below the upper edge, force it there and try one
3514 # more time.
3515 if ($i == $range_list_size - 2) {
3516
3517 trace "Forcing to upper edge" if main::DEBUG && $to_trace;
3518 $i = $range_list_size - 1;
3519
3520 # Change $lower as well so if fails next time through,
3521 # taking the average will yield the same $i, and we will
3522 # quit with the error message just below.
3523 $lower = $i;
3524 next;
3525 }
3526 Carp::my_carp_bug("$owner_name_of{$addr}Can't find where the range ought to go. No action taken.");
3527 return;
3528 }
3529 $i = $temp;
3530 } # End of while loop
3531
3532 if (main::DEBUG && $to_trace) {
3533 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i;
3534 trace "i= [ $i ]", $r->[$i];
3535 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < $range_list_size - 1;
3536 }
3537
3538 # Here we have found the offset. Cache it as a starting point for the
3539 # next call.
3540 $_search_ranges_cache{$addr} = $i;
3541 return $i;
3542 }
3543
3544 sub _add_delete {
3545 # Add, replace or delete ranges to or from a list. The $type
3546 # parameter gives which:
3547 # '+' => insert or replace a range, returning a list of any changed
3548 # ranges.
3549 # '-' => delete a range, returning a list of any deleted ranges.
3550 #
3551 # The next three parameters give respectively the start, end, and
3552 # value associated with the range. 'value' should be null unless the
3553 # operation is '+';
3554 #
3555 # The range list is kept sorted so that the range with the lowest
3556 # starting position is first in the list, and generally, adjacent
c1739a4a 3557 # ranges with the same values are merged into a single larger one (see
99870f4d
KW
3558 # exceptions below).
3559 #
c1739a4a 3560 # There are more parameters; all are key => value pairs:
99870f4d
KW
3561 # Type gives the type of the value. It is only valid for '+'.
3562 # All ranges have types; if this parameter is omitted, 0 is
3563 # assumed. Ranges with type 0 are assumed to obey the
3564 # Unicode rules for casing, etc; ranges with other types are
3565 # not. Otherwise, the type is arbitrary, for the caller's
3566 # convenience, and looked at only by this routine to keep
3567 # adjacent ranges of different types from being merged into
3568 # a single larger range, and when Replace =>
3569 # $IF_NOT_EQUIVALENT is specified (see just below).
3570 # Replace determines what to do if the range list already contains
3571 # ranges which coincide with all or portions of the input
3572 # range. It is only valid for '+':
3573 # => $NO means that the new value is not to replace
3574 # any existing ones, but any empty gaps of the
3575 # range list coinciding with the input range
3576 # will be filled in with the new value.
3577 # => $UNCONDITIONALLY means to replace the existing values with
3578 # this one unconditionally. However, if the
3579 # new and old values are identical, the
3580 # replacement is skipped to save cycles
3581 # => $IF_NOT_EQUIVALENT means to replace the existing values
d59563d0 3582 # (the default) with this one if they are not equivalent.
99870f4d 3583 # Ranges are equivalent if their types are the
c1739a4a 3584 # same, and they are the same string; or if
99870f4d
KW
3585 # both are type 0 ranges, if their Unicode
3586 # standard forms are identical. In this last
3587 # case, the routine chooses the more "modern"
3588 # one to use. This is because some of the
3589 # older files are formatted with values that
3590 # are, for example, ALL CAPs, whereas the
3591 # derived files have a more modern style,
3592 # which looks better. By looking for this
3593 # style when the pre-existing and replacement
3594 # standard forms are the same, we can move to
3595 # the modern style
9470941f 3596 # => $MULTIPLE_BEFORE means that if this range duplicates an
99870f4d
KW
3597 # existing one, but has a different value,
3598 # don't replace the existing one, but insert
3599 # this, one so that the same range can occur
53d84487
KW
3600 # multiple times. They are stored LIFO, so
3601 # that the final one inserted is the first one
3602 # returned in an ordered search of the table.
6901521e
KW
3603 # If this is an exact duplicate, including the
3604 # value, the original will be moved to be
3605 # first, before any other duplicate ranges
3606 # with different values.
7f4b1e25
KW
3607 # => $MULTIPLE_AFTER is like $MULTIPLE_BEFORE, but is stored
3608 # FIFO, so that this one is inserted after all
6901521e
KW
3609 # others that currently exist. If this is an
3610 # exact duplicate, including value, of an
3611 # existing range, this one is discarded
3612 # (leaving the existing one in its original,
3613 # higher priority position
99870f4d
KW
3614 # => anything else is the same as => $IF_NOT_EQUIVALENT
3615 #
c1739a4a
KW
3616 # "same value" means identical for non-type-0 ranges, and it means
3617 # having the same standard forms for type-0 ranges.
99870f4d
KW
3618
3619 return Carp::carp_too_few_args(\@_, 5) if main::DEBUG && @_ < 5;
3620
3621 my $self = shift;
3622 my $operation = shift; # '+' for add/replace; '-' for delete;
3623 my $start = shift;
3624 my $end = shift;
3625 my $value = shift;
3626
3627 my %args = @_;
3628
3629 $value = "" if not defined $value; # warning: $value can be "0"
3630
3631 my $replace = delete $args{'Replace'};
3632 $replace = $IF_NOT_EQUIVALENT unless defined $replace;
3633
3634 my $type = delete $args{'Type'};
3635 $type = 0 unless defined $type;
3636
3637 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3638
ffe43484 3639 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3640
3641 if ($operation ne '+' && $operation ne '-') {
3642 Carp::my_carp_bug("$owner_name_of{$addr}First parameter to _add_delete must be '+' or '-'. No action taken.");
3643 return;
3644 }
3645 unless (defined $start && defined $end) {
3646 Carp::my_carp_bug("$owner_name_of{$addr}Undefined start and/or end to _add_delete. No action taken.");
3647 return;
3648 }
3649 unless ($end >= $start) {
3650 Carp::my_carp_bug("$owner_name_of{$addr}End of range (" . sprintf("%04X", $end) . ") must not be before start (" . sprintf("%04X", $start) . "). No action taken.");
3651 return;
3652 }
556ca434
KW
3653 if ($end > $MAX_UNICODE_CODEPOINT && $operation eq '+') {
3654 Carp::my_carp("$owner_name_of{$addr}Warning: Range '" . sprintf("%04X..%04X", $start, $end) . ") is above the Unicode maximum of " . sprintf("%04X", $MAX_UNICODE_CODEPOINT) . ". Adding it anyway");
3655 }
99870f4d
KW
3656 #local $to_trace = 1 if main::DEBUG;
3657
3658 if ($operation eq '-') {
3659 if ($replace != $IF_NOT_EQUIVALENT) {
3660 Carp::my_carp_bug("$owner_name_of{$addr}Replace => \$IF_NOT_EQUIVALENT is required when deleting a range from a range list. Assuming Replace => \$IF_NOT_EQUIVALENT.");
3661 $replace = $IF_NOT_EQUIVALENT;
3662 }
3663 if ($type) {
3664 Carp::my_carp_bug("$owner_name_of{$addr}Type => 0 is required when deleting a range from a range list. Assuming Type => 0.");
3665 $type = 0;
3666 }
3667 if ($value ne "") {
3668 Carp::my_carp_bug("$owner_name_of{$addr}Value => \"\" is required when deleting a range from a range list. Assuming Value => \"\".");
3669 $value = "";
3670 }
3671 }
3672
3673 my $r = $ranges{$addr}; # The current list of ranges
3674 my $range_list_size = scalar @$r; # And its size
3675 my $max = $max{$addr}; # The current high code point in
3676 # the list of ranges
3677
3678 # Do a special case requiring fewer machine cycles when the new range
3679 # starts after the current highest point. The Unicode input data is
3680 # structured so this is common.
3681 if ($start > $max) {
3682
52d4d76a 3683 trace "$owner_name_of{$addr} $operation", sprintf("%04X..%04X (%s) type=%d; prev max=%04X", $start, $end, $value, $type, $max) if main::DEBUG && $to_trace;
99870f4d
KW
3684 return if $operation eq '-'; # Deleting a non-existing range is a
3685 # no-op
3686
3687 # If the new range doesn't logically extend the current final one
3688 # in the range list, create a new range at the end of the range
3689 # list. (max cleverly is initialized to a negative number not
3690 # adjacent to 0 if the range list is empty, so even adding a range
3691 # to an empty range list starting at 0 will have this 'if'
3692 # succeed.)
3693 if ($start > $max + 1 # non-adjacent means can't extend.
3694 || @{$r}[-1]->value ne $value # values differ, can't extend.
3695 || @{$r}[-1]->type != $type # types differ, can't extend.
3696 ) {
3697 push @$r, Range->new($start, $end,
3698 Value => $value,
3699 Type => $type);
3700 }
3701 else {
3702
3703 # Here, the new range starts just after the current highest in
3704 # the range list, and they have the same type and value.
3705 # Extend the current range to incorporate the new one.
3706 @{$r}[-1]->set_end($end);
3707 }
3708
3709 # This becomes the new maximum.
3710 $max{$addr} = $end;
3711
3712 return;
3713 }
3714 #local $to_trace = 0 if main::DEBUG;
3715
3716 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) replace=$replace" if main::DEBUG && $to_trace;
3717
3718 # Here, the input range isn't after the whole rest of the range list.
3719 # Most likely 'splice' will be needed. The rest of the routine finds
3720 # the needed splice parameters, and if necessary, does the splice.
3721 # First, find the offset parameter needed by the splice function for
3722 # the input range. Note that the input range may span multiple
3723 # existing ones, but we'll worry about that later. For now, just find
3724 # the beginning. If the input range is to be inserted starting in a
3725 # position not currently in the range list, it must (obviously) come
3726 # just after the range below it, and just before the range above it.
3727 # Slightly less obviously, it will occupy the position currently
3728 # occupied by the range that is to come after it. More formally, we
3729 # are looking for the position, $i, in the array of ranges, such that:
3730 #
3731 # r[$i-1]->start <= r[$i-1]->end < $start < r[$i]->start <= r[$i]->end
3732 #
3733 # (The ordered relationships within existing ranges are also shown in
3734 # the equation above). However, if the start of the input range is
3735 # within an existing range, the splice offset should point to that
3736 # existing range's position in the list; that is $i satisfies a
3737 # somewhat different equation, namely:
3738 #
3739 #r[$i-1]->start <= r[$i-1]->end < r[$i]->start <= $start <= r[$i]->end
3740 #
3741 # More briefly, $start can come before or after r[$i]->start, and at
3742 # this point, we don't know which it will be. However, these
3743 # two equations share these constraints:
3744 #
3745 # r[$i-1]->end < $start <= r[$i]->end
3746 #
3747 # And that is good enough to find $i.
3748
3749 my $i = $self->_search_ranges($start);
3750 if (! defined $i) {
3751 Carp::my_carp_bug("Searching $self for range beginning with $start unexpectedly returned undefined. Operation '$operation' not performed");
3752 return;
3753 }
3754
3755 # The search function returns $i such that:
3756 #
3757 # r[$i-1]->end < $start <= r[$i]->end
3758 #
3759 # That means that $i points to the first range in the range list
3760 # that could possibly be affected by this operation. We still don't
3761 # know if the start of the input range is within r[$i], or if it
3762 # points to empty space between r[$i-1] and r[$i].
3763 trace "[$i] is the beginning splice point. Existing range there is ", $r->[$i] if main::DEBUG && $to_trace;
3764
3765 # Special case the insertion of data that is not to replace any
3766 # existing data.
3767 if ($replace == $NO) { # If $NO, has to be operation '+'
3768 #local $to_trace = 1 if main::DEBUG;
3769 trace "Doesn't replace" if main::DEBUG && $to_trace;
3770
3771 # Here, the new range is to take effect only on those code points
3772 # that aren't already in an existing range. This can be done by
3773 # looking through the existing range list and finding the gaps in
3774 # the ranges that this new range affects, and then calling this
3775 # function recursively on each of those gaps, leaving untouched
3776 # anything already in the list. Gather up a list of the changed
3777 # gaps first so that changes to the internal state as new ranges
3778 # are added won't be a problem.
3779 my @gap_list;
3780
3781 # First, if the starting point of the input range is outside an