This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
mktables: Correct Unicode 6.1 omission
[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
KW
38
39##########################################################################
40#
41# mktables -- create the runtime Perl Unicode files (lib/unicore/.../*.pl),
42# from the Unicode database files (lib/unicore/.../*.txt), It also generates
43# a pod file and a .t file
44#
45# The structure of this file is:
46# First these introductory comments; then
47# code needed for everywhere, such as debugging stuff; then
48# code to handle input parameters; then
49# data structures likely to be of external interest (some of which depend on
50# the input parameters, so follows them; then
51# more data structures and subroutine and package (class) definitions; then
52# the small actual loop to process the input files and finish up; then
53# a __DATA__ section, for the .t tests
54#
5f7264c7 55# This program works on all releases of Unicode through at least 6.0. The
99870f4d
KW
56# outputs have been scrutinized most intently for release 5.1. The others
57# have been checked for somewhat more than just sanity. It can handle all
58# existing Unicode character properties in those releases.
59#
99870f4d
KW
60# This program is mostly about Unicode character (or code point) properties.
61# A property describes some attribute or quality of a code point, like if it
62# is lowercase or not, its name, what version of Unicode it was first defined
63# in, or what its uppercase equivalent is. Unicode deals with these disparate
64# possibilities by making all properties into mappings from each code point
65# into some corresponding value. In the case of it being lowercase or not,
66# the mapping is either to 'Y' or 'N' (or various synonyms thereof). Each
67# property maps each Unicode code point to a single value, called a "property
68# value". (Hence each Unicode property is a true mathematical function with
69# exactly one value per code point.)
70#
71# When using a property in a regular expression, what is desired isn't the
72# mapping of the code point to its property's value, but the reverse (or the
73# mathematical "inverse relation"): starting with the property value, "Does a
74# code point map to it?" These are written in a "compound" form:
75# \p{property=value}, e.g., \p{category=punctuation}. This program generates
76# files containing the lists of code points that map to each such regular
77# expression property value, one file per list
78#
79# There is also a single form shortcut that Perl adds for many of the commonly
80# used properties. This happens for all binary properties, plus script,
81# general_category, and block properties.
82#
83# Thus the outputs of this program are files. There are map files, mostly in
84# the 'To' directory; and there are list files for use in regular expression
85# matching, all in subdirectories of the 'lib' directory, with each
86# subdirectory being named for the property that the lists in it are for.
87# Bookkeeping, test, and documentation files are also generated.
88
89my $matches_directory = 'lib'; # Where match (\p{}) files go.
90my $map_directory = 'To'; # Where map files go.
91
92# DATA STRUCTURES
93#
94# The major data structures of this program are Property, of course, but also
95# Table. There are two kinds of tables, very similar to each other.
96# "Match_Table" is the data structure giving the list of code points that have
97# a particular property value, mentioned above. There is also a "Map_Table"
98# data structure which gives the property's mapping from code point to value.
99# There are two structures because the match tables need to be combined in
100# various ways, such as constructing unions, intersections, complements, etc.,
101# and the map ones don't. And there would be problems, perhaps subtle, if
102# a map table were inadvertently operated on in some of those ways.
103# The use of separate classes with operations defined on one but not the other
104# prevents accidentally confusing the two.
105#
106# At the heart of each table's data structure is a "Range_List", which is just
107# an ordered list of "Ranges", plus ancillary information, and methods to
108# operate on them. A Range is a compact way to store property information.
109# Each range has a starting code point, an ending code point, and a value that
110# is meant to apply to all the code points between the two end points,
111# inclusive. For a map table, this value is the property value for those
112# code points. Two such ranges could be written like this:
113# 0x41 .. 0x5A, 'Upper',
114# 0x61 .. 0x7A, 'Lower'
115#
116# Each range also has a type used as a convenience to classify the values.
117# Most ranges in this program will be Type 0, or normal, but there are some
118# ranges that have a non-zero type. These are used only in map tables, and
119# are for mappings that don't fit into the normal scheme of things. Mappings
120# that require a hash entry to communicate with utf8.c are one example;
121# another example is mappings for charnames.pm to use which indicate a name
122# that is algorithmically determinable from its code point (and vice-versa).
123# These are used to significantly compact these tables, instead of listing
124# each one of the tens of thousands individually.
125#
126# In a match table, the value of a range is irrelevant (and hence the type as
127# well, which will always be 0), and arbitrarily set to the null string.
128# Using the example above, there would be two match tables for those two
129# entries, one named Upper would contain the 0x41..0x5A range, and the other
130# named Lower would contain 0x61..0x7A.
131#
132# Actually, there are two types of range lists, "Range_Map" is the one
133# associated with map tables, and "Range_List" with match tables.
134# Again, this is so that methods can be defined on one and not the other so as
135# to prevent operating on them in incorrect ways.
136#
137# Eventually, most tables are written out to files to be read by utf8_heavy.pl
138# in the perl core. All tables could in theory be written, but some are
139# suppressed because there is no current practical use for them. It is easy
140# to change which get written by changing various lists that are near the top
141# of the actual code in this file. The table data structures contain enough
142# ancillary information to allow them to be treated as separate entities for
143# writing, such as the path to each one's file. There is a heading in each
144# map table that gives the format of its entries, and what the map is for all
145# the code points missing from it. (This allows tables to be more compact.)
678f13d5 146#
99870f4d
KW
147# The Property data structure contains one or more tables. All properties
148# contain a map table (except the $perl property which is a
149# pseudo-property containing only match tables), and any properties that
150# are usable in regular expression matches also contain various matching
151# tables, one for each value the property can have. A binary property can
152# have two values, True and False (or Y and N, which are preferred by Unicode
153# terminology). Thus each of these properties will have a map table that
154# takes every code point and maps it to Y or N (but having ranges cuts the
155# number of entries in that table way down), and two match tables, one
156# which has a list of all the code points that map to Y, and one for all the
157# code points that map to N. (For each of these, a third table is also
158# generated for the pseudo Perl property. It contains the identical code
159# points as the Y table, but can be written, not in the compound form, but in
160# a "single" form like \p{IsUppercase}.) Many properties are binary, but some
161# properties have several possible values, some have many, and properties like
162# Name have a different value for every named code point. Those will not,
163# unless the controlling lists are changed, have their match tables written
164# out. But all the ones which can be used in regular expression \p{} and \P{}
c12f2655
KW
165# constructs will. Prior to 5.14, generally a property would have either its
166# map table or its match tables written but not both. Again, what gets
167# written is controlled by lists which can easily be changed. Starting in
168# 5.14, advantage was taken of this, and all the map tables needed to
169# reconstruct the Unicode db are now written out, while suppressing the
170# Unicode .txt files that contain the data. Our tables are much more compact
171# than the .txt files, so a significant space savings was achieved.
172
173# Properties have a 'Type', like binary, or string, or enum depending on how
174# many match tables there are and the content of the maps. This 'Type' is
175# different than a range 'Type', so don't get confused by the two concepts
176# having the same name.
678f13d5 177#
99870f4d
KW
178# For information about the Unicode properties, see Unicode's UAX44 document:
179
180my $unicode_reference_url = 'http://www.unicode.org/reports/tr44/';
181
182# As stated earlier, this program will work on any release of Unicode so far.
183# Most obvious problems in earlier data have NOT been corrected except when
184# necessary to make Perl or this program work reasonably. For example, no
c12f2655
KW
185# folding information was given in early releases, so this program substitutes
186# lower case instead, just so that a regular expression with the /i option
187# will do something that actually gives the right results in many cases.
188# There are also a couple other corrections for version 1.1.5, commented at
189# the point they are made. As an example of corrections that weren't made
190# (but could be) is this statement from DerivedAge.txt: "The supplementary
191# private use code points and the non-character code points were assigned in
192# version 2.0, but not specifically listed in the UCD until versions 3.0 and
193# 3.1 respectively." (To be precise it was 3.0.1 not 3.0.0) More information
194# on Unicode version glitches is further down in these introductory comments.
99870f4d 195#
5f7264c7
KW
196# This program works on all non-provisional properties as of 6.0, though the
197# files for some are suppressed from apparent lack of demand for them. You
198# can change which are output by changing lists in this program.
678f13d5 199#
dc85bd38 200# The old version of mktables emphasized the term "Fuzzy" to mean Unicode's
99870f4d
KW
201# loose matchings rules (from Unicode TR18):
202#
203# The recommended names for UCD properties and property values are in
204# PropertyAliases.txt [Prop] and PropertyValueAliases.txt
205# [PropValue]. There are both abbreviated names and longer, more
206# descriptive names. It is strongly recommended that both names be
207# recognized, and that loose matching of property names be used,
208# whereby the case distinctions, whitespace, hyphens, and underbar
209# are ignored.
210# The program still allows Fuzzy to override its determination of if loose
211# matching should be used, but it isn't currently used, as it is no longer
212# needed; the calculations it makes are good enough.
678f13d5 213#
99870f4d
KW
214# SUMMARY OF HOW IT WORKS:
215#
216# Process arguments
217#
218# A list is constructed containing each input file that is to be processed
219#
220# Each file on the list is processed in a loop, using the associated handler
221# code for each:
222# The PropertyAliases.txt and PropValueAliases.txt files are processed
223# first. These files name the properties and property values.
224# Objects are created of all the property and property value names
225# that the rest of the input should expect, including all synonyms.
226# The other input files give mappings from properties to property
227# values. That is, they list code points and say what the mapping
228# is under the given property. Some files give the mappings for
229# just one property; and some for many. This program goes through
230# each file and populates the properties from them. Some properties
231# are listed in more than one file, and Unicode has set up a
232# precedence as to which has priority if there is a conflict. Thus
233# the order of processing matters, and this program handles the
234# conflict possibility by processing the overriding input files
235# last, so that if necessary they replace earlier values.
236# After this is all done, the program creates the property mappings not
237# furnished by Unicode, but derivable from what it does give.
238# The tables of code points that match each property value in each
239# property that is accessible by regular expressions are created.
240# The Perl-defined properties are created and populated. Many of these
241# require data determined from the earlier steps
242# Any Perl-defined synonyms are created, and name clashes between Perl
678f13d5 243# and Unicode are reconciled and warned about.
99870f4d
KW
244# All the properties are written to files
245# Any other files are written, and final warnings issued.
678f13d5 246#
99870f4d
KW
247# For clarity, a number of operators have been overloaded to work on tables:
248# ~ means invert (take all characters not in the set). The more
249# conventional '!' is not used because of the possibility of confusing
250# it with the actual boolean operation.
251# + means union
252# - means subtraction
253# & means intersection
254# The precedence of these is the order listed. Parentheses should be
255# copiously used. These are not a general scheme. The operations aren't
256# defined for a number of things, deliberately, to avoid getting into trouble.
257# Operations are done on references and affect the underlying structures, so
258# that the copy constructors for them have been overloaded to not return a new
259# clone, but the input object itself.
678f13d5 260#
99870f4d
KW
261# The bool operator is deliberately not overloaded to avoid confusion with
262# "should it mean if the object merely exists, or also is non-empty?".
99870f4d
KW
263#
264# WHY CERTAIN DESIGN DECISIONS WERE MADE
678f13d5
KW
265#
266# This program needs to be able to run under miniperl. Therefore, it uses a
267# minimum of other modules, and hence implements some things itself that could
268# be gotten from CPAN
269#
270# This program uses inputs published by the Unicode Consortium. These can
271# change incompatibly between releases without the Perl maintainers realizing
272# it. Therefore this program is now designed to try to flag these. It looks
273# at the directories where the inputs are, and flags any unrecognized files.
274# It keeps track of all the properties in the files it handles, and flags any
275# that it doesn't know how to handle. It also flags any input lines that
276# don't match the expected syntax, among other checks.
277#
278# It is also designed so if a new input file matches one of the known
279# templates, one hopefully just needs to add it to a list to have it
280# processed.
281#
282# As mentioned earlier, some properties are given in more than one file. In
283# particular, the files in the extracted directory are supposedly just
284# reformattings of the others. But they contain information not easily
285# derivable from the other files, including results for Unihan, which this
286# program doesn't ordinarily look at, and for unassigned code points. They
287# also have historically had errors or been incomplete. In an attempt to
288# create the best possible data, this program thus processes them first to
289# glean information missing from the other files; then processes those other
290# files to override any errors in the extracted ones. Much of the design was
291# driven by this need to store things and then possibly override them.
292#
293# It tries to keep fatal errors to a minimum, to generate something usable for
294# testing purposes. It always looks for files that could be inputs, and will
295# warn about any that it doesn't know how to handle (the -q option suppresses
296# the warning).
99870f4d 297#
678f13d5
KW
298# Why is there more than one type of range?
299# This simplified things. There are some very specialized code points that
300# have to be handled specially for output, such as Hangul syllable names.
301# By creating a range type (done late in the development process), it
302# allowed this to be stored with the range, and overridden by other input.
303# Originally these were stored in another data structure, and it became a
304# mess trying to decide if a second file that was for the same property was
305# overriding the earlier one or not.
306#
307# Why are there two kinds of tables, match and map?
308# (And there is a base class shared by the two as well.) As stated above,
309# they actually are for different things. Development proceeded much more
310# smoothly when I (khw) realized the distinction. Map tables are used to
311# give the property value for every code point (actually every code point
312# that doesn't map to a default value). Match tables are used for regular
313# expression matches, and are essentially the inverse mapping. Separating
314# the two allows more specialized methods, and error checks so that one
315# can't just take the intersection of two map tables, for example, as that
316# is nonsensical.
99870f4d 317#
23e33b60
KW
318# DEBUGGING
319#
678f13d5
KW
320# This program is written so it will run under miniperl. Occasionally changes
321# will cause an error where the backtrace doesn't work well under miniperl.
322# To diagnose the problem, you can instead run it under regular perl, if you
323# have one compiled.
324#
325# There is a good trace facility. To enable it, first sub DEBUG must be set
326# to return true. Then a line like
327#
328# local $to_trace = 1 if main::DEBUG;
329#
330# can be added to enable tracing in its lexical scope or until you insert
331# another line:
332#
333# local $to_trace = 0 if main::DEBUG;
334#
335# then use a line like "trace $a, @b, %c, ...;
336#
337# Some of the more complex subroutines already have trace statements in them.
338# Permanent trace statements should be like:
339#
340# trace ... if main::DEBUG && $to_trace;
341#
342# If there is just one or a few files that you're debugging, you can easily
343# cause most everything else to be skipped. Change the line
344#
345# my $debug_skip = 0;
346#
347# to 1, and every file whose object is in @input_file_objects and doesn't have
348# a, 'non_skip => 1,' in its constructor will be skipped.
349#
b4a0206c 350# To compare the output tables, it may be useful to specify the -annotate
c4019d52
KW
351# flag. This causes the tables to expand so there is one entry for each
352# non-algorithmically named code point giving, currently its name, and its
353# graphic representation if printable (and you have a font that knows about
354# it). This makes it easier to see what the particular code points are in
355# each output table. The tables are usable, but because they don't have
356# ranges (for the most part), a Perl using them will run slower. Non-named
357# code points are annotated with a description of their status, and contiguous
358# ones with the same description will be output as a range rather than
359# individually. Algorithmically named characters are also output as ranges,
360# except when there are just a few contiguous ones.
361#
99870f4d
KW
362# FUTURE ISSUES
363#
364# The program would break if Unicode were to change its names so that
365# interior white space, underscores, or dashes differences were significant
366# within property and property value names.
367#
368# It might be easier to use the xml versions of the UCD if this program ever
369# would need heavy revision, and the ability to handle old versions was not
370# required.
371#
372# There is the potential for name collisions, in that Perl has chosen names
373# that Unicode could decide it also likes. There have been such collisions in
374# the past, with mostly Perl deciding to adopt the Unicode definition of the
375# name. However in the 5.2 Unicode beta testing, there were a number of such
376# collisions, which were withdrawn before the final release, because of Perl's
377# and other's protests. These all involved new properties which began with
378# 'Is'. Based on the protests, Unicode is unlikely to try that again. Also,
379# many of the Perl-defined synonyms, like Any, Word, etc, are listed in a
380# Unicode document, so they are unlikely to be used by Unicode for another
381# purpose. However, they might try something beginning with 'In', or use any
382# of the other Perl-defined properties. This program will warn you of name
383# collisions, and refuse to generate tables with them, but manual intervention
384# will be required in this event. One scheme that could be implemented, if
385# necessary, would be to have this program generate another file, or add a
386# field to mktables.lst that gives the date of first definition of a property.
387# Each new release of Unicode would use that file as a basis for the next
388# iteration. And the Perl synonym addition code could sort based on the age
389# of the property, so older properties get priority, and newer ones that clash
390# would be refused; hence existing code would not be impacted, and some other
391# synonym would have to be used for the new property. This is ugly, and
392# manual intervention would certainly be easier to do in the short run; lets
393# hope it never comes to this.
678f13d5 394#
99870f4d
KW
395# A NOTE ON UNIHAN
396#
397# This program can generate tables from the Unihan database. But it doesn't
398# by default, letting the CPAN module Unicode::Unihan handle them. Prior to
399# version 5.2, this database was in a single file, Unihan.txt. In 5.2 the
400# database was split into 8 different files, all beginning with the letters
401# 'Unihan'. This program will read those file(s) if present, but it needs to
402# know which of the many properties in the file(s) should have tables created
403# for them. It will create tables for any properties listed in
404# PropertyAliases.txt and PropValueAliases.txt, plus any listed in the
405# @cjk_properties array and the @cjk_property_values array. Thus, if a
406# property you want is not in those files of the release you are building
407# against, you must add it to those two arrays. Starting in 4.0, the
408# Unicode_Radical_Stroke was listed in those files, so if the Unihan database
409# is present in the directory, a table will be generated for that property.
410# In 5.2, several more properties were added. For your convenience, the two
5f7264c7 411# arrays are initialized with all the 6.0 listed properties that are also in
99870f4d
KW
412# earlier releases. But these are commented out. You can just uncomment the
413# ones you want, or use them as a template for adding entries for other
414# properties.
415#
416# You may need to adjust the entries to suit your purposes. setup_unihan(),
417# and filter_unihan_line() are the functions where this is done. This program
418# already does some adjusting to make the lines look more like the rest of the
419# Unicode DB; You can see what that is in filter_unihan_line()
420#
421# There is a bug in the 3.2 data file in which some values for the
422# kPrimaryNumeric property have commas and an unexpected comment. A filter
423# could be added for these; or for a particular installation, the Unihan.txt
424# file could be edited to fix them.
99870f4d 425#
678f13d5
KW
426# HOW TO ADD A FILE TO BE PROCESSED
427#
428# A new file from Unicode needs to have an object constructed for it in
429# @input_file_objects, probably at the end or at the end of the extracted
430# ones. The program should warn you if its name will clash with others on
431# restrictive file systems, like DOS. If so, figure out a better name, and
432# add lines to the README.perl file giving that. If the file is a character
433# property, it should be in the format that Unicode has by default
434# standardized for such files for the more recently introduced ones.
435# If so, the Input_file constructor for @input_file_objects can just be the
436# file name and release it first appeared in. If not, then it should be
437# possible to construct an each_line_handler() to massage the line into the
438# standardized form.
439#
440# For non-character properties, more code will be needed. You can look at
441# the existing entries for clues.
442#
443# UNICODE VERSIONS NOTES
444#
445# The Unicode UCD has had a number of errors in it over the versions. And
446# these remain, by policy, in the standard for that version. Therefore it is
447# risky to correct them, because code may be expecting the error. So this
448# program doesn't generally make changes, unless the error breaks the Perl
449# core. As an example, some versions of 2.1.x Jamo.txt have the wrong value
450# for U+1105, which causes real problems for the algorithms for Jamo
451# calculations, so it is changed here.
452#
453# But it isn't so clear cut as to what to do about concepts that are
454# introduced in a later release; should they extend back to earlier releases
455# where the concept just didn't exist? It was easier to do this than to not,
456# so that's what was done. For example, the default value for code points not
457# in the files for various properties was probably undefined until changed by
458# some version. No_Block for blocks is such an example. This program will
459# assign No_Block even in Unicode versions that didn't have it. This has the
460# benefit that code being written doesn't have to special case earlier
461# versions; and the detriment that it doesn't match the Standard precisely for
462# the affected versions.
463#
464# Here are some observations about some of the issues in early versions:
465#
6426c51b 466# The number of code points in \p{alpha} halved in 2.1.9. It turns out that
678f13d5
KW
467# the reason is that the CJK block starting at 4E00 was removed from PropList,
468# and was not put back in until 3.1.0
469#
470# Unicode introduced the synonym Space for White_Space in 4.1. Perl has
471# always had a \p{Space}. In release 3.2 only, they are not synonymous. The
472# reason is that 3.2 introduced U+205F=medium math space, which was not
473# classed as white space, but Perl figured out that it should have been. 4.0
474# reclassified it correctly.
475#
476# Another change between 3.2 and 4.0 is the CCC property value ATBL. In 3.2
477# this was erroneously a synonym for 202. In 4.0, ATB became 202, and ATBL
478# was left with no code points, as all the ones that mapped to 202 stayed
479# mapped to 202. Thus if your program used the numeric name for the class,
480# it would not have been affected, but if it used the mnemonic, it would have
481# been.
482#
483# \p{Script=Hrkt} (Katakana_Or_Hiragana) came in 4.0.1. Before that code
484# points which eventually came to have this script property value, instead
485# mapped to "Unknown". But in the next release all these code points were
486# moved to \p{sc=common} instead.
99870f4d
KW
487#
488# The default for missing code points for BidiClass is complicated. Starting
489# in 3.1.1, the derived file DBidiClass.txt handles this, but this program
490# tries to do the best it can for earlier releases. It is done in
491# process_PropertyAliases()
492#
493##############################################################################
494
495my $UNDEF = ':UNDEF:'; # String to print out for undefined values in tracing
496 # and errors
497my $MAX_LINE_WIDTH = 78;
498
499# Debugging aid to skip most files so as to not be distracted by them when
500# concentrating on the ones being debugged. Add
501# non_skip => 1,
502# to the constructor for those files you want processed when you set this.
503# Files with a first version number of 0 are special: they are always
c12f2655
KW
504# processed regardless of the state of this flag. Generally, Jamo.txt and
505# UnicodeData.txt must not be skipped if you want this program to not die
506# before normal completion.
99870f4d
KW
507my $debug_skip = 0;
508
509# Set to 1 to enable tracing.
510our $to_trace = 0;
511
512{ # Closure for trace: debugging aid
513 my $print_caller = 1; # ? Include calling subroutine name
514 my $main_with_colon = 'main::';
515 my $main_colon_length = length($main_with_colon);
516
517 sub trace {
518 return unless $to_trace; # Do nothing if global flag not set
519
520 my @input = @_;
521
522 local $DB::trace = 0;
523 $DB::trace = 0; # Quiet 'used only once' message
524
525 my $line_number;
526
527 # Loop looking up the stack to get the first non-trace caller
528 my $caller_line;
529 my $caller_name;
530 my $i = 0;
531 do {
532 $line_number = $caller_line;
533 (my $pkg, my $file, $caller_line, my $caller) = caller $i++;
534 $caller = $main_with_colon unless defined $caller;
535
536 $caller_name = $caller;
537
538 # get rid of pkg
539 $caller_name =~ s/.*:://;
540 if (substr($caller_name, 0, $main_colon_length)
541 eq $main_with_colon)
542 {
543 $caller_name = substr($caller_name, $main_colon_length);
544 }
545
546 } until ($caller_name ne 'trace');
547
548 # If the stack was empty, we were called from the top level
549 $caller_name = 'main' if ($caller_name eq ""
550 || $caller_name eq 'trace');
551
552 my $output = "";
553 foreach my $string (@input) {
554 #print STDERR __LINE__, ": ", join ", ", @input, "\n";
555 if (ref $string eq 'ARRAY' || ref $string eq 'HASH') {
556 $output .= simple_dumper($string);
557 }
558 else {
559 $string = "$string" if ref $string;
560 $string = $UNDEF unless defined $string;
561 chomp $string;
562 $string = '""' if $string eq "";
563 $output .= " " if $output ne ""
564 && $string ne ""
565 && substr($output, -1, 1) ne " "
566 && substr($string, 0, 1) ne " ";
567 $output .= $string;
568 }
569 }
570
99f78760
KW
571 print STDERR sprintf "%4d: ", $line_number if defined $line_number;
572 print STDERR "$caller_name: " if $print_caller;
99870f4d
KW
573 print STDERR $output, "\n";
574 return;
575 }
576}
577
578# This is for a rarely used development feature that allows you to compare two
579# versions of the Unicode standard without having to deal with changes caused
c12f2655
KW
580# by the code points introduced in the later version. Change the 0 to a
581# string containing a SINGLE dotted Unicode release number (e.g. "2.1"). Only
582# code points introduced in that release and earlier will be used; later ones
583# are thrown away. You use the version number of the earliest one you want to
584# compare; then run this program on directory structures containing each
585# release, and compare the outputs. These outputs will therefore include only
586# the code points common to both releases, and you can see the changes caused
587# just by the underlying release semantic changes. For versions earlier than
588# 3.2, you must copy a version of DAge.txt into the directory.
589my $string_compare_versions = DEBUG && 0; # e.g., "2.1";
99870f4d
KW
590my $compare_versions = DEBUG
591 && $string_compare_versions
592 && pack "C*", split /\./, $string_compare_versions;
593
594sub uniques {
595 # Returns non-duplicated input values. From "Perl Best Practices:
596 # Encapsulated Cleverness". p. 455 in first edition.
597
598 my %seen;
0e407844
NC
599 # Arguably this breaks encapsulation, if the goal is to permit multiple
600 # distinct objects to stringify to the same value, and be interchangeable.
601 # However, for this program, no two objects stringify identically, and all
602 # lists passed to this function are either objects or strings. So this
603 # doesn't affect correctness, but it does give a couple of percent speedup.
604 no overloading;
99870f4d
KW
605 return grep { ! $seen{$_}++ } @_;
606}
607
608$0 = File::Spec->canonpath($0);
609
610my $make_test_script = 0; # ? Should we output a test script
611my $write_unchanged_files = 0; # ? Should we update the output files even if
612 # we don't think they have changed
613my $use_directory = ""; # ? Should we chdir somewhere.
614my $pod_directory; # input directory to store the pod file.
615my $pod_file = 'perluniprops';
616my $t_path; # Path to the .t test file
617my $file_list = 'mktables.lst'; # File to store input and output file names.
618 # This is used to speed up the build, by not
619 # executing the main body of the program if
620 # nothing on the list has changed since the
621 # previous build
622my $make_list = 1; # ? Should we write $file_list. Set to always
623 # make a list so that when the pumpking is
624 # preparing a release, s/he won't have to do
625 # special things
626my $glob_list = 0; # ? Should we try to include unknown .txt files
627 # in the input.
bd9ebcfd
KW
628my $output_range_counts = $debugging_build; # ? Should we include the number
629 # of code points in ranges in
630 # the output
558712cf 631my $annotate = 0; # ? Should character names be in the output
9ef2b94f 632
99870f4d
KW
633# Verbosity levels; 0 is quiet
634my $NORMAL_VERBOSITY = 1;
635my $PROGRESS = 2;
636my $VERBOSE = 3;
637
638my $verbosity = $NORMAL_VERBOSITY;
639
640# Process arguments
641while (@ARGV) {
cf25bb62
JH
642 my $arg = shift @ARGV;
643 if ($arg eq '-v') {
99870f4d
KW
644 $verbosity = $VERBOSE;
645 }
646 elsif ($arg eq '-p') {
647 $verbosity = $PROGRESS;
648 $| = 1; # Flush buffers as we go.
649 }
650 elsif ($arg eq '-q') {
651 $verbosity = 0;
652 }
653 elsif ($arg eq '-w') {
654 $write_unchanged_files = 1; # update the files even if havent changed
655 }
656 elsif ($arg eq '-check') {
6ae7e459
YO
657 my $this = shift @ARGV;
658 my $ok = shift @ARGV;
659 if ($this ne $ok) {
660 print "Skipping as check params are not the same.\n";
661 exit(0);
662 }
00a8df5c 663 }
99870f4d
KW
664 elsif ($arg eq '-P' && defined ($pod_directory = shift)) {
665 -d $pod_directory or croak "Directory '$pod_directory' doesn't exist";
666 }
3df51b85
KW
667 elsif ($arg eq '-maketest' || ($arg eq '-T' && defined ($t_path = shift)))
668 {
99870f4d 669 $make_test_script = 1;
99870f4d
KW
670 }
671 elsif ($arg eq '-makelist') {
672 $make_list = 1;
673 }
674 elsif ($arg eq '-C' && defined ($use_directory = shift)) {
675 -d $use_directory or croak "Unknown directory '$use_directory'";
676 }
677 elsif ($arg eq '-L') {
678
679 # Existence not tested until have chdir'd
680 $file_list = shift;
681 }
682 elsif ($arg eq '-globlist') {
683 $glob_list = 1;
684 }
685 elsif ($arg eq '-c') {
686 $output_range_counts = ! $output_range_counts
687 }
b4a0206c 688 elsif ($arg eq '-annotate') {
558712cf 689 $annotate = 1;
bd9ebcfd
KW
690 $debugging_build = 1;
691 $output_range_counts = 1;
9ef2b94f 692 }
99870f4d
KW
693 else {
694 my $with_c = 'with';
695 $with_c .= 'out' if $output_range_counts; # Complements the state
696 croak <<END;
697usage: $0 [-c|-p|-q|-v|-w] [-C dir] [-L filelist] [ -P pod_dir ]
698 [ -T test_file_path ] [-globlist] [-makelist] [-maketest]
699 [-check A B ]
700 -c : Output comments $with_c number of code points in ranges
701 -q : Quiet Mode: Only output serious warnings.
702 -p : Set verbosity level to normal plus show progress.
703 -v : Set Verbosity level high: Show progress and non-serious
704 warnings
705 -w : Write files regardless
706 -C dir : Change to this directory before proceeding. All relative paths
707 except those specified by the -P and -T options will be done
708 with respect to this directory.
709 -P dir : Output $pod_file file to directory 'dir'.
3df51b85 710 -T path : Create a test script as 'path'; overrides -maketest
99870f4d
KW
711 -L filelist : Use alternate 'filelist' instead of standard one
712 -globlist : Take as input all non-Test *.txt files in current and sub
713 directories
3df51b85
KW
714 -maketest : Make test script 'TestProp.pl' in current (or -C directory),
715 overrides -T
99870f4d 716 -makelist : Rewrite the file list $file_list based on current setup
b4a0206c 717 -annotate : Output an annotation for each character in the table files;
c4019d52 718 useful for debugging mktables, looking at diffs; but is slow,
b318e5e5
KW
719 memory intensive; resulting tables are usable but are slow and
720 very large (and currently fail the Unicode::UCD.t tests).
99870f4d
KW
721 -check A B : Executes $0 only if A and B are the same
722END
723 }
724}
725
726# Stores the most-recently changed file. If none have changed, can skip the
727# build
aeab6150 728my $most_recent = (stat $0)[9]; # Do this before the chdir!
99870f4d
KW
729
730# Change directories now, because need to read 'version' early.
731if ($use_directory) {
3df51b85 732 if ($pod_directory && ! File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d
KW
733 $pod_directory = File::Spec->rel2abs($pod_directory);
734 }
3df51b85 735 if ($t_path && ! File::Spec->file_name_is_absolute($t_path)) {
99870f4d 736 $t_path = File::Spec->rel2abs($t_path);
00a8df5c 737 }
99870f4d 738 chdir $use_directory or croak "Failed to chdir to '$use_directory':$!";
3df51b85 739 if ($pod_directory && File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d 740 $pod_directory = File::Spec->abs2rel($pod_directory);
02b1aeec 741 }
3df51b85 742 if ($t_path && File::Spec->file_name_is_absolute($t_path)) {
99870f4d 743 $t_path = File::Spec->abs2rel($t_path);
02b1aeec 744 }
00a8df5c
YO
745}
746
99870f4d
KW
747# Get Unicode version into regular and v-string. This is done now because
748# various tables below get populated based on it. These tables are populated
749# here to be near the top of the file, and so easily seeable by those needing
750# to modify things.
751open my $VERSION, "<", "version"
752 or croak "$0: can't open required file 'version': $!\n";
753my $string_version = <$VERSION>;
754close $VERSION;
755chomp $string_version;
756my $v_version = pack "C*", split /\./, $string_version; # v string
757
758# The following are the complete names of properties with property values that
759# are known to not match any code points in some versions of Unicode, but that
760# may change in the future so they should be matchable, hence an empty file is
761# generated for them.
762my @tables_that_may_be_empty = (
763 'Joining_Type=Left_Joining',
764 );
765push @tables_that_may_be_empty, 'Script=Common' if $v_version le v4.0.1;
766push @tables_that_may_be_empty, 'Title' if $v_version lt v2.0.0;
767push @tables_that_may_be_empty, 'Script=Katakana_Or_Hiragana'
768 if $v_version ge v4.1.0;
82aed44a
KW
769push @tables_that_may_be_empty, 'Script_Extensions=Katakana_Or_Hiragana'
770 if $v_version ge v6.0.0;
f583b44c
KW
771push @tables_that_may_be_empty, 'Grapheme_Cluster_Break=Prepend'
772 if $v_version ge v6.1.0;
99870f4d
KW
773
774# The lists below are hashes, so the key is the item in the list, and the
775# value is the reason why it is in the list. This makes generation of
776# documentation easier.
777
778my %why_suppressed; # No file generated for these.
779
780# Files aren't generated for empty extraneous properties. This is arguable.
781# Extraneous properties generally come about because a property is no longer
782# used in a newer version of Unicode. If we generated a file without code
783# points, programs that used to work on that property will still execute
784# without errors. It just won't ever match (or will always match, with \P{}).
785# This means that the logic is now likely wrong. I (khw) think its better to
786# find this out by getting an error message. Just move them to the table
787# above to change this behavior
788my %why_suppress_if_empty_warn_if_not = (
789
790 # It is the only property that has ever officially been removed from the
791 # Standard. The database never contained any code points for it.
792 'Special_Case_Condition' => 'Obsolete',
793
794 # Apparently never official, but there were code points in some versions of
795 # old-style PropList.txt
796 'Non_Break' => 'Obsolete',
797);
798
799# These would normally go in the warn table just above, but they were changed
800# a long time before this program was written, so warnings about them are
801# moot.
802if ($v_version gt v3.2.0) {
803 push @tables_that_may_be_empty,
804 'Canonical_Combining_Class=Attached_Below_Left'
805}
806
5f7264c7 807# These are listed in the Property aliases file in 6.0, but Unihan is ignored
99870f4d
KW
808# unless explicitly added.
809if ($v_version ge v5.2.0) {
810 my $unihan = 'Unihan; remove from list if using Unihan';
ea25a9b2 811 foreach my $table (qw (
99870f4d
KW
812 kAccountingNumeric
813 kOtherNumeric
814 kPrimaryNumeric
815 kCompatibilityVariant
816 kIICore
817 kIRG_GSource
818 kIRG_HSource
819 kIRG_JSource
820 kIRG_KPSource
821 kIRG_MSource
822 kIRG_KSource
823 kIRG_TSource
824 kIRG_USource
825 kIRG_VSource
826 kRSUnicode
ea25a9b2 827 ))
99870f4d
KW
828 {
829 $why_suppress_if_empty_warn_if_not{$table} = $unihan;
830 }
ca12659b
NC
831}
832
272501f6
KW
833# Enum values for to_output_map() method in the Map_Table package.
834my $EXTERNAL_MAP = 1;
835my $INTERNAL_MAP = 2;
ce712c88 836my $OUTPUT_ADJUSTED = 3;
272501f6 837
fcf1973c
KW
838# To override computed values for writing the map tables for these properties.
839# The default for enum map tables is to write them out, so that the Unicode
840# .txt files can be removed, but all the data to compute any property value
841# for any code point is available in a more compact form.
842my %global_to_output_map = (
843 # Needed by UCD.pm, but don't want to publicize that it exists, so won't
c12f2655
KW
844 # get stuck supporting it if things change. Since it is a STRING
845 # property, it normally would be listed in the pod, but INTERNAL_MAP
846 # suppresses that.
fcf1973c
KW
847 Unicode_1_Name => $INTERNAL_MAP,
848
849 Present_In => 0, # Suppress, as easily computed from Age
fcf1973c 850 Block => 0, # Suppress, as Blocks.txt is retained.
53d34b6c
KW
851
852 # Suppress, as mapping can be found instead from the
853 # Perl_Decomposition_Mapping file
854 Decomposition_Type => 0,
fcf1973c
KW
855);
856
99870f4d 857# Properties that this program ignores.
230e0c16
KW
858my @unimplemented_properties;
859
860# With this release, it is automatically handled if the Unihan db is
861# downloaded
862push @unimplemented_properties, 'Unicode_Radical_Stroke' if $v_version le v5.2.0;
d73e5302 863
99870f4d
KW
864# There are several types of obsolete properties defined by Unicode. These
865# must be hand-edited for every new Unicode release.
866my %why_deprecated; # Generates a deprecated warning message if used.
867my %why_stabilized; # Documentation only
868my %why_obsolete; # Documentation only
869
870{ # Closure
871 my $simple = 'Perl uses the more complete version of this property';
872 my $unihan = 'Unihan properties are by default not enabled in the Perl core. Instead use CPAN: Unicode::Unihan';
873
874 my $other_properties = 'other properties';
875 my $contributory = "Used by Unicode internally for generating $other_properties and not intended to be used stand-alone";
5d294d41 876 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
877
878 %why_deprecated = (
5f7264c7 879 'Grapheme_Link' => 'Deprecated by Unicode: Duplicates ccc=vr (Canonical_Combining_Class=Virama)',
99870f4d
KW
880 'Jamo_Short_Name' => $contributory,
881 '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',
882 'Other_Alphabetic' => $contributory,
883 'Other_Default_Ignorable_Code_Point' => $contributory,
884 'Other_Grapheme_Extend' => $contributory,
885 'Other_ID_Continue' => $contributory,
886 'Other_ID_Start' => $contributory,
887 'Other_Lowercase' => $contributory,
888 'Other_Math' => $contributory,
889 'Other_Uppercase' => $contributory,
e22aaf5c
KW
890 'Expands_On_NFC' => $why_no_expand,
891 'Expands_On_NFD' => $why_no_expand,
892 'Expands_On_NFKC' => $why_no_expand,
893 'Expands_On_NFKD' => $why_no_expand,
99870f4d
KW
894 );
895
896 %why_suppressed = (
5f7264c7 897 # There is a lib/unicore/Decomposition.pl (used by Normalize.pm) which
99870f4d
KW
898 # contains the same information, but without the algorithmically
899 # determinable Hangul syllables'. This file is not published, so it's
900 # existence is not noted in the comment.
e0b29447 901 'Decomposition_Mapping' => 'Accessible via Unicode::Normalize or Unicode::UCD::prop_invmap()',
99870f4d 902
3111abc0
KW
903 'Indic_Matra_Category' => "Provisional",
904 'Indic_Syllabic_Category' => "Provisional",
905
5f8d1a89
KW
906 # Don't suppress ISO_Comment, as otherwise special handling is needed
907 # to differentiate between it and gc=c, which can be written as 'isc',
908 # which is the same characters as ISO_Comment's short name.
99870f4d 909
fbb93542 910 'Name' => "Accessible via \\N{...} or 'use charnames;' or Unicode::UCD::prop_invmap()",
e0b29447
KW
911
912 'Simple_Case_Folding' => "$simple. Can access this through Unicode::UCD::casefold or Unicode::UCD::prop_invmap()",
913 'Simple_Lowercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
914 'Simple_Titlecase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
915 'Simple_Uppercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
99870f4d 916
5f7264c7 917 FC_NFKC_Closure => 'Supplanted in usage by NFKC_Casefold; otherwise not useful',
99870f4d
KW
918 );
919
1704a0ea
KW
920 foreach my $property (
921
922 # The following are suppressed because they were made contributory
923 # or deprecated by Unicode before Perl ever thought about
924 # supporting them.
925 'Jamo_Short_Name',
926 'Grapheme_Link',
927 'Expands_On_NFC',
928 'Expands_On_NFD',
929 'Expands_On_NFKC',
930 'Expands_On_NFKD',
931
932 # The following are suppressed because they have been marked
933 # as deprecated for a sufficient amount of time
934 'Other_Alphabetic',
935 'Other_Default_Ignorable_Code_Point',
936 'Other_Grapheme_Extend',
937 'Other_ID_Continue',
938 'Other_ID_Start',
939 'Other_Lowercase',
940 'Other_Math',
941 'Other_Uppercase',
e22aaf5c 942 ) {
99870f4d
KW
943 $why_suppressed{$property} = $why_deprecated{$property};
944 }
cf25bb62 945
99870f4d
KW
946 # Customize the message for all the 'Other_' properties
947 foreach my $property (keys %why_deprecated) {
948 next if (my $main_property = $property) !~ s/^Other_//;
949 $why_deprecated{$property} =~ s/$other_properties/the $main_property property (which should be used instead)/;
950 }
951}
952
953if ($v_version ge 4.0.0) {
954 $why_stabilized{'Hyphen'} = 'Use the Line_Break property instead; see www.unicode.org/reports/tr14';
5f7264c7
KW
955 if ($v_version ge 6.0.0) {
956 $why_deprecated{'Hyphen'} = 'Supplanted by Line_Break property values; see www.unicode.org/reports/tr14';
957 }
99870f4d 958}
5f7264c7 959if ($v_version ge 5.2.0 && $v_version lt 6.0.0) {
99870f4d 960 $why_obsolete{'ISO_Comment'} = 'Code points for it have been removed';
5f7264c7 961 if ($v_version ge 6.0.0) {
63f74647 962 $why_deprecated{'ISO_Comment'} = 'No longer needed for Unicode\'s internal chart generation; otherwise not useful, and code points for it have been removed';
5f7264c7 963 }
99870f4d
KW
964}
965
966# Probably obsolete forever
967if ($v_version ge v4.1.0) {
82aed44a
KW
968 $why_suppressed{'Script=Katakana_Or_Hiragana'} = 'Obsolete. All code points previously matched by this have been moved to "Script=Common".';
969}
970if ($v_version ge v6.0.0) {
2b352efd
KW
971 $why_suppressed{'Script=Katakana_Or_Hiragana'} .= ' Consider instead using "Script_Extensions=Katakana" or "Script_Extensions=Hiragana (or both)"';
972 $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
973}
974
975# This program can create files for enumerated-like properties, such as
976# 'Numeric_Type'. This file would be the same format as for a string
977# property, with a mapping from code point to its value, so you could look up,
978# for example, the script a code point is in. But no one so far wants this
979# mapping, or they have found another way to get it since this is a new
980# feature. So no file is generated except if it is in this list.
981my @output_mapped_properties = split "\n", <<END;
982END
983
c12f2655
KW
984# If you are using the Unihan database in a Unicode version before 5.2, you
985# need to add the properties that you want to extract from it to this table.
986# For your convenience, the properties in the 6.0 PropertyAliases.txt file are
987# listed, commented out
99870f4d
KW
988my @cjk_properties = split "\n", <<'END';
989#cjkAccountingNumeric; kAccountingNumeric
990#cjkOtherNumeric; kOtherNumeric
991#cjkPrimaryNumeric; kPrimaryNumeric
992#cjkCompatibilityVariant; kCompatibilityVariant
993#cjkIICore ; kIICore
994#cjkIRG_GSource; kIRG_GSource
995#cjkIRG_HSource; kIRG_HSource
996#cjkIRG_JSource; kIRG_JSource
997#cjkIRG_KPSource; kIRG_KPSource
998#cjkIRG_KSource; kIRG_KSource
999#cjkIRG_TSource; kIRG_TSource
1000#cjkIRG_USource; kIRG_USource
1001#cjkIRG_VSource; kIRG_VSource
1002#cjkRSUnicode; kRSUnicode ; Unicode_Radical_Stroke; URS
1003END
1004
1005# Similarly for the property values. For your convenience, the lines in the
5f7264c7 1006# 6.0 PropertyAliases.txt file are listed. Just remove the first BUT NOT both
c12f2655 1007# '#' marks (for Unicode versions before 5.2)
99870f4d
KW
1008my @cjk_property_values = split "\n", <<'END';
1009## @missing: 0000..10FFFF; cjkAccountingNumeric; NaN
1010## @missing: 0000..10FFFF; cjkCompatibilityVariant; <code point>
1011## @missing: 0000..10FFFF; cjkIICore; <none>
1012## @missing: 0000..10FFFF; cjkIRG_GSource; <none>
1013## @missing: 0000..10FFFF; cjkIRG_HSource; <none>
1014## @missing: 0000..10FFFF; cjkIRG_JSource; <none>
1015## @missing: 0000..10FFFF; cjkIRG_KPSource; <none>
1016## @missing: 0000..10FFFF; cjkIRG_KSource; <none>
1017## @missing: 0000..10FFFF; cjkIRG_TSource; <none>
1018## @missing: 0000..10FFFF; cjkIRG_USource; <none>
1019## @missing: 0000..10FFFF; cjkIRG_VSource; <none>
1020## @missing: 0000..10FFFF; cjkOtherNumeric; NaN
1021## @missing: 0000..10FFFF; cjkPrimaryNumeric; NaN
1022## @missing: 0000..10FFFF; cjkRSUnicode; <none>
1023END
1024
1025# The input files don't list every code point. Those not listed are to be
1026# defaulted to some value. Below are hard-coded what those values are for
1027# non-binary properties as of 5.1. Starting in 5.0, there are
1028# machine-parsable comment lines in the files the give the defaults; so this
1029# list shouldn't have to be extended. The claim is that all missing entries
1030# for binary properties will default to 'N'. Unicode tried to change that in
1031# 5.2, but the beta period produced enough protest that they backed off.
1032#
1033# The defaults for the fields that appear in UnicodeData.txt in this hash must
1034# be in the form that it expects. The others may be synonyms.
1035my $CODE_POINT = '<code point>';
1036my %default_mapping = (
1037 Age => "Unassigned",
1038 # Bidi_Class => Complicated; set in code
1039 Bidi_Mirroring_Glyph => "",
1040 Block => 'No_Block',
1041 Canonical_Combining_Class => 0,
1042 Case_Folding => $CODE_POINT,
1043 Decomposition_Mapping => $CODE_POINT,
1044 Decomposition_Type => 'None',
1045 East_Asian_Width => "Neutral",
1046 FC_NFKC_Closure => $CODE_POINT,
1047 General_Category => 'Cn',
1048 Grapheme_Cluster_Break => 'Other',
1049 Hangul_Syllable_Type => 'NA',
1050 ISO_Comment => "",
1051 Jamo_Short_Name => "",
1052 Joining_Group => "No_Joining_Group",
1053 # Joining_Type => Complicated; set in code
1054 kIICore => 'N', # Is converted to binary
1055 #Line_Break => Complicated; set in code
1056 Lowercase_Mapping => $CODE_POINT,
1057 Name => "",
1058 Name_Alias => "",
1059 NFC_QC => 'Yes',
1060 NFD_QC => 'Yes',
1061 NFKC_QC => 'Yes',
1062 NFKD_QC => 'Yes',
1063 Numeric_Type => 'None',
1064 Numeric_Value => 'NaN',
1065 Script => ($v_version le 4.1.0) ? 'Common' : 'Unknown',
1066 Sentence_Break => 'Other',
1067 Simple_Case_Folding => $CODE_POINT,
1068 Simple_Lowercase_Mapping => $CODE_POINT,
1069 Simple_Titlecase_Mapping => $CODE_POINT,
1070 Simple_Uppercase_Mapping => $CODE_POINT,
1071 Titlecase_Mapping => $CODE_POINT,
1072 Unicode_1_Name => "",
1073 Unicode_Radical_Stroke => "",
1074 Uppercase_Mapping => $CODE_POINT,
1075 Word_Break => 'Other',
1076);
1077
1078# Below are files that Unicode furnishes, but this program ignores, and why
1079my %ignored_files = (
73ba1144
KW
1080 'CJKRadicals.txt' => 'Maps the kRSUnicode property values to corresponding code points',
1081 'Index.txt' => 'Alphabetical index of Unicode characters',
1082 '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',
1083 'NamesList.txt' => 'Annotated list of characters',
1084 'NormalizationCorrections.txt' => 'Documentation of corrections already incorporated into the Unicode data base',
1085 'Props.txt' => 'Only in very early releases; is a subset of F<PropList.txt> (which is used instead)',
1086 'ReadMe.txt' => 'Documentation',
1087 '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>',
1088 'EmojiSources.txt' => 'Maps certain Unicode code points to their legacy Japanese cell-phone values',
73ba1144
KW
1089 'auxiliary/WordBreakTest.html' => 'Documentation of validation tests',
1090 'auxiliary/SentenceBreakTest.html' => 'Documentation of validation tests',
1091 'auxiliary/GraphemeBreakTest.html' => 'Documentation of validation tests',
1092 'auxiliary/LineBreakTest.html' => 'Documentation of validation tests',
99870f4d
KW
1093);
1094
1fec9f60
KW
1095my %skipped_files; # List of files that we skip
1096
678f13d5 1097### End of externally interesting definitions, except for @input_file_objects
99870f4d
KW
1098
1099my $HEADER=<<"EOF";
1100# !!!!!!! DO NOT EDIT THIS FILE !!!!!!!
3df51b85
KW
1101# This file is machine-generated by $0 from the Unicode
1102# database, Version $string_version. Any changes made here will be lost!
cf25bb62
JH
1103EOF
1104
126c3d4e 1105my $INTERNAL_ONLY_HEADER = <<"EOF";
99870f4d
KW
1106
1107# !!!!!!! INTERNAL PERL USE ONLY !!!!!!!
fac53429
KW
1108# This file is for internal use by core Perl only. The format and even the
1109# name or existence of this file are subject to change without notice. Don't
1110# use it directly.
99870f4d
KW
1111EOF
1112
1113my $DEVELOPMENT_ONLY=<<"EOF";
1114# !!!!!!! DEVELOPMENT USE ONLY !!!!!!!
1115# This file contains information artificially constrained to code points
1116# present in Unicode release $string_compare_versions.
1117# IT CANNOT BE RELIED ON. It is for use during development only and should
23e33b60 1118# not be used for production.
b6922eda
KW
1119
1120EOF
1121
6189eadc
KW
1122my $MAX_UNICODE_CODEPOINT_STRING = "10FFFF";
1123my $MAX_UNICODE_CODEPOINT = hex $MAX_UNICODE_CODEPOINT_STRING;
1124my $MAX_UNICODE_CODEPOINTS = $MAX_UNICODE_CODEPOINT + 1;
99870f4d
KW
1125
1126# Matches legal code point. 4-6 hex numbers, If there are 6, the first
1127# two must be 10; if there are 5, the first must not be a 0. Written this way
92199589
KW
1128# to decrease backtracking. The first regex allows the code point to be at
1129# the end of a word, but to work properly, the word shouldn't end with a valid
1130# hex character. The second one won't match a code point at the end of a
1131# word, and doesn't have the run-on issue
8c32d378
KW
1132my $run_on_code_point_re =
1133 qr/ (?: 10[0-9A-F]{4} | [1-9A-F][0-9A-F]{4} | [0-9A-F]{4} ) \b/x;
1134my $code_point_re = qr/\b$run_on_code_point_re/;
99870f4d
KW
1135
1136# This matches the beginning of the line in the Unicode db files that give the
1137# defaults for code points not listed (i.e., missing) in the file. The code
1138# depends on this ending with a semi-colon, so it can assume it is a valid
1139# field when the line is split() by semi-colons
1140my $missing_defaults_prefix =
6189eadc 1141 qr/^#\s+\@missing:\s+0000\.\.$MAX_UNICODE_CODEPOINT_STRING\s*;/;
99870f4d
KW
1142
1143# Property types. Unicode has more types, but these are sufficient for our
1144# purposes.
1145my $UNKNOWN = -1; # initialized to illegal value
1146my $NON_STRING = 1; # Either binary or enum
1147my $BINARY = 2;
06f26c45
KW
1148my $FORCED_BINARY = 3; # Not a binary property, but, besides its normal
1149 # tables, additional true and false tables are
1150 # generated so that false is anything matching the
1151 # default value, and true is everything else.
1152my $ENUM = 4; # Include catalog
1153my $STRING = 5; # Anything else: string or misc
99870f4d
KW
1154
1155# Some input files have lines that give default values for code points not
1156# contained in the file. Sometimes these should be ignored.
1157my $NO_DEFAULTS = 0; # Must evaluate to false
1158my $NOT_IGNORED = 1;
1159my $IGNORED = 2;
1160
1161# Range types. Each range has a type. Most ranges are type 0, for normal,
1162# and will appear in the main body of the tables in the output files, but
1163# there are other types of ranges as well, listed below, that are specially
1164# handled. There are pseudo-types as well that will never be stored as a
1165# type, but will affect the calculation of the type.
1166
1167# 0 is for normal, non-specials
1168my $MULTI_CP = 1; # Sequence of more than code point
1169my $HANGUL_SYLLABLE = 2;
1170my $CP_IN_NAME = 3; # The NAME contains the code point appended to it.
1171my $NULL = 4; # The map is to the null string; utf8.c can't
1172 # handle these, nor is there an accepted syntax
1173 # for them in \p{} constructs
f86864ac 1174my $COMPUTE_NO_MULTI_CP = 5; # Pseudo-type; means that ranges that would
99870f4d
KW
1175 # otherwise be $MULTI_CP type are instead type 0
1176
1177# process_generic_property_file() can accept certain overrides in its input.
1178# Each of these must begin AND end with $CMD_DELIM.
1179my $CMD_DELIM = "\a";
1180my $REPLACE_CMD = 'replace'; # Override the Replace
1181my $MAP_TYPE_CMD = 'map_type'; # Override the Type
1182
1183my $NO = 0;
1184my $YES = 1;
1185
1186# Values for the Replace argument to add_range.
1187# $NO # Don't replace; add only the code points not
1188 # already present.
1189my $IF_NOT_EQUIVALENT = 1; # Replace only under certain conditions; details in
1190 # the comments at the subroutine definition.
1191my $UNCONDITIONALLY = 2; # Replace without conditions.
9470941f 1192my $MULTIPLE_BEFORE = 4; # Don't replace, but add a duplicate record if
99870f4d 1193 # already there
7f4b1e25
KW
1194my $MULTIPLE_AFTER = 5; # Don't replace, but add a duplicate record if
1195 # already there
1196my $CROAK = 6; # Die with an error if is already there
99870f4d
KW
1197
1198# Flags to give property statuses. The phrases are to remind maintainers that
1199# if the flag is changed, the indefinite article referring to it in the
1200# documentation may need to be as well.
1201my $NORMAL = "";
99870f4d
KW
1202my $DEPRECATED = 'D';
1203my $a_bold_deprecated = "a 'B<$DEPRECATED>'";
1204my $A_bold_deprecated = "A 'B<$DEPRECATED>'";
1205my $DISCOURAGED = 'X';
1206my $a_bold_discouraged = "an 'B<$DISCOURAGED>'";
1207my $A_bold_discouraged = "An 'B<$DISCOURAGED>'";
1208my $STRICTER = 'T';
1209my $a_bold_stricter = "a 'B<$STRICTER>'";
1210my $A_bold_stricter = "A 'B<$STRICTER>'";
1211my $STABILIZED = 'S';
1212my $a_bold_stabilized = "an 'B<$STABILIZED>'";
1213my $A_bold_stabilized = "An 'B<$STABILIZED>'";
1214my $OBSOLETE = 'O';
1215my $a_bold_obsolete = "an 'B<$OBSOLETE>'";
1216my $A_bold_obsolete = "An 'B<$OBSOLETE>'";
1217
1218my %status_past_participles = (
1219 $DISCOURAGED => 'discouraged',
99870f4d
KW
1220 $STABILIZED => 'stabilized',
1221 $OBSOLETE => 'obsolete',
37e2e78e 1222 $DEPRECATED => 'deprecated',
99870f4d
KW
1223);
1224
395dfc19
KW
1225# Table fates. These are somewhat ordered, so that fates < $MAP_PROXIED should be
1226# externally documented.
301ba948 1227my $ORDINARY = 0; # The normal fate.
395dfc19
KW
1228my $MAP_PROXIED = 1; # The map table for the property isn't written out,
1229 # but there is a file written that can be used to
1230 # reconstruct this table
301ba948
KW
1231my $SUPPRESSED = 3; # The file for this table is not written out.
1232my $INTERNAL_ONLY = 4; # The file for this table is written out, but it is
1233 # for Perl's internal use only
1234my $PLACEHOLDER = 5; # A property that is defined as a placeholder in a
1235 # Unicode version that doesn't have it, but we need it
1236 # to be defined, if empty, to have things work.
1237 # Implies no pod entry generated
1238
f5817e0a
KW
1239# The format of the values of the tables:
1240my $EMPTY_FORMAT = "";
99870f4d
KW
1241my $BINARY_FORMAT = 'b';
1242my $DECIMAL_FORMAT = 'd';
1243my $FLOAT_FORMAT = 'f';
1244my $INTEGER_FORMAT = 'i';
1245my $HEX_FORMAT = 'x';
1246my $RATIONAL_FORMAT = 'r';
1247my $STRING_FORMAT = 's';
d11155ec 1248my $ADJUST_FORMAT = 'a';
a14f3cb1 1249my $DECOMP_STRING_FORMAT = 'c';
c3ff2976 1250my $STRING_WHITE_SPACE_LIST = 'sw';
99870f4d
KW
1251
1252my %map_table_formats = (
1253 $BINARY_FORMAT => 'binary',
1254 $DECIMAL_FORMAT => 'single decimal digit',
1255 $FLOAT_FORMAT => 'floating point number',
1256 $INTEGER_FORMAT => 'integer',
add63c13 1257 $HEX_FORMAT => 'non-negative hex whole number; a code point',
99870f4d 1258 $RATIONAL_FORMAT => 'rational: an integer or a fraction',
1a9d544b 1259 $STRING_FORMAT => 'string',
d11155ec 1260 $ADJUST_FORMAT => 'some entries need adjustment',
92f9d56c 1261 $DECOMP_STRING_FORMAT => 'Perl\'s internal (Normalize.pm) decomposition mapping',
c3ff2976 1262 $STRING_WHITE_SPACE_LIST => 'string, but some elements are interpreted as a list; white space occurs only as list item separators'
99870f4d
KW
1263);
1264
1265# Unicode didn't put such derived files in a separate directory at first.
1266my $EXTRACTED_DIR = (-d 'extracted') ? 'extracted' : "";
1267my $EXTRACTED = ($EXTRACTED_DIR) ? "$EXTRACTED_DIR/" : "";
1268my $AUXILIARY = 'auxiliary';
1269
1270# Hashes that will eventually go into Heavy.pl for the use of utf8_heavy.pl
9e4a1e86 1271# and into UCD.pl for the use of UCD.pm
99870f4d
KW
1272my %loose_to_file_of; # loosely maps table names to their respective
1273 # files
1274my %stricter_to_file_of; # same; but for stricter mapping.
315bfd4e 1275my %loose_property_to_file_of; # Maps a loose property name to its map file
89cf10cc
KW
1276my %file_to_swash_name; # Maps the file name to its corresponding key name
1277 # in the hash %utf8::SwashInfo
99870f4d
KW
1278my %nv_floating_to_rational; # maps numeric values floating point numbers to
1279 # their rational equivalent
c12f2655
KW
1280my %loose_property_name_of; # Loosely maps (non_string) property names to
1281 # standard form
86a52d1e 1282my %string_property_loose_to_name; # Same, for string properties.
c15fda25
KW
1283my %loose_defaults; # keys are of form "prop=value", where 'prop' is
1284 # the property name in standard loose form, and
1285 # 'value' is the default value for that property,
1286 # also in standard loose form.
9e4a1e86
KW
1287my %loose_to_standard_value; # loosely maps table names to the canonical
1288 # alias for them
2df7880f
KW
1289my %ambiguous_names; # keys are alias names (in standard form) that
1290 # have more than one possible meaning.
5d1df013
KW
1291my %prop_aliases; # Keys are standard property name; values are each
1292 # one's aliases
1e863613
KW
1293my %prop_value_aliases; # Keys of top level are standard property name;
1294 # values are keys to another hash, Each one is
1295 # one of the property's values, in standard form.
1296 # The values are that prop-val's aliases.
2df7880f 1297my %ucd_pod; # Holds entries that will go into the UCD section of the pod
99870f4d 1298
d867ccfb
KW
1299# Most properties are immune to caseless matching, otherwise you would get
1300# nonsensical results, as properties are a function of a code point, not
1301# everything that is caselessly equivalent to that code point. For example,
1302# Changes_When_Case_Folded('s') should be false, whereas caselessly it would
1303# be true because 's' and 'S' are equivalent caselessly. However,
1304# traditionally, [:upper:] and [:lower:] are equivalent caselessly, so we
1305# extend that concept to those very few properties that are like this. Each
1306# such property will match the full range caselessly. They are hard-coded in
1307# the program; it's not worth trying to make it general as it's extremely
1308# unlikely that they will ever change.
1309my %caseless_equivalent_to;
1310
99870f4d
KW
1311# These constants names and values were taken from the Unicode standard,
1312# version 5.1, section 3.12. They are used in conjunction with Hangul
6e5a209b
KW
1313# syllables. The '_string' versions are so generated tables can retain the
1314# hex format, which is the more familiar value
1315my $SBase_string = "0xAC00";
1316my $SBase = CORE::hex $SBase_string;
1317my $LBase_string = "0x1100";
1318my $LBase = CORE::hex $LBase_string;
1319my $VBase_string = "0x1161";
1320my $VBase = CORE::hex $VBase_string;
1321my $TBase_string = "0x11A7";
1322my $TBase = CORE::hex $TBase_string;
99870f4d
KW
1323my $SCount = 11172;
1324my $LCount = 19;
1325my $VCount = 21;
1326my $TCount = 28;
1327my $NCount = $VCount * $TCount;
1328
1329# For Hangul syllables; These store the numbers from Jamo.txt in conjunction
1330# with the above published constants.
1331my %Jamo;
1332my %Jamo_L; # Leading consonants
1333my %Jamo_V; # Vowels
1334my %Jamo_T; # Trailing consonants
1335
bb1dd3da
KW
1336# For code points whose name contains its ordinal as a '-ABCD' suffix.
1337# The key is the base name of the code point, and the value is an
1338# array giving all the ranges that use this base name. Each range
1339# is actually a hash giving the 'low' and 'high' values of it.
1340my %names_ending_in_code_point;
1341my %loose_names_ending_in_code_point; # Same as above, but has blanks, dashes
1342 # removed from the names
1343# Inverse mapping. The list of ranges that have these kinds of
1344# names. Each element contains the low, high, and base names in an
1345# anonymous hash.
1346my @code_points_ending_in_code_point;
1347
1348# Boolean: does this Unicode version have the hangul syllables, and are we
1349# writing out a table for them?
1350my $has_hangul_syllables = 0;
1351
1352# Does this Unicode version have code points whose names end in their
1353# respective code points, and are we writing out a table for them? 0 for no;
1354# otherwise points to first property that a table is needed for them, so that
1355# if multiple tables are needed, we don't create duplicates
1356my $needing_code_points_ending_in_code_point = 0;
1357
37e2e78e 1358my @backslash_X_tests; # List of tests read in for testing \X
99870f4d
KW
1359my @unhandled_properties; # Will contain a list of properties found in
1360 # the input that we didn't process.
f86864ac 1361my @match_properties; # Properties that have match tables, to be
99870f4d
KW
1362 # listed in the pod
1363my @map_properties; # Properties that get map files written
1364my @named_sequences; # NamedSequences.txt contents.
1365my %potential_files; # Generated list of all .txt files in the directory
1366 # structure so we can warn if something is being
1367 # ignored.
1368my @files_actually_output; # List of files we generated.
1369my @more_Names; # Some code point names are compound; this is used
1370 # to store the extra components of them.
1371my $MIN_FRACTION_LENGTH = 3; # How many digits of a floating point number at
1372 # the minimum before we consider it equivalent to a
1373 # candidate rational
1374my $MAX_FLOATING_SLOP = 10 ** - $MIN_FRACTION_LENGTH; # And in floating terms
1375
1376# These store references to certain commonly used property objects
1377my $gc;
1378my $perl;
1379my $block;
3e20195b
KW
1380my $perl_charname;
1381my $print;
7fc6cb55 1382my $Any;
359523e2 1383my $script;
99870f4d
KW
1384
1385# Are there conflicting names because of beginning with 'In_', or 'Is_'
1386my $has_In_conflicts = 0;
1387my $has_Is_conflicts = 0;
1388
1389sub internal_file_to_platform ($) {
1390 # Convert our file paths which have '/' separators to those of the
1391 # platform.
1392
1393 my $file = shift;
1394 return undef unless defined $file;
1395
1396 return File::Spec->join(split '/', $file);
d07a55ed 1397}
5beb625e 1398
99870f4d
KW
1399sub file_exists ($) { # platform independent '-e'. This program internally
1400 # uses slash as a path separator.
1401 my $file = shift;
1402 return 0 if ! defined $file;
1403 return -e internal_file_to_platform($file);
1404}
5beb625e 1405
99870f4d 1406sub objaddr($) {
23e33b60
KW
1407 # Returns the address of the blessed input object.
1408 # It doesn't check for blessedness because that would do a string eval
1409 # every call, and the program is structured so that this is never called
1410 # for a non-blessed object.
99870f4d 1411
23e33b60 1412 no overloading; # If overloaded, numifying below won't work.
99870f4d
KW
1413
1414 # Numifying a ref gives its address.
051df77b 1415 return pack 'J', $_[0];
99870f4d
KW
1416}
1417
558712cf 1418# These are used only if $annotate is true.
c4019d52
KW
1419# The entire range of Unicode characters is examined to populate these
1420# after all the input has been processed. But most can be skipped, as they
1421# have the same descriptive phrases, such as being unassigned
1422my @viacode; # Contains the 1 million character names
1423my @printable; # boolean: And are those characters printable?
1424my @annotate_char_type; # Contains a type of those characters, specifically
1425 # for the purposes of annotation.
1426my $annotate_ranges; # A map of ranges of code points that have the same
98dc9551 1427 # name for the purposes of annotation. They map to the
c4019d52
KW
1428 # upper edge of the range, so that the end point can
1429 # be immediately found. This is used to skip ahead to
1430 # the end of a range, and avoid processing each
1431 # individual code point in it.
1432my $unassigned_sans_noncharacters; # A Range_List of the unassigned
1433 # characters, but excluding those which are
1434 # also noncharacter code points
1435
1436# The annotation types are an extension of the regular range types, though
1437# some of the latter are folded into one. Make the new types negative to
1438# avoid conflicting with the regular types
1439my $SURROGATE_TYPE = -1;
1440my $UNASSIGNED_TYPE = -2;
1441my $PRIVATE_USE_TYPE = -3;
1442my $NONCHARACTER_TYPE = -4;
1443my $CONTROL_TYPE = -5;
1444my $UNKNOWN_TYPE = -6; # Used only if there is a bug in this program
1445
1446sub populate_char_info ($) {
558712cf 1447 # Used only with the $annotate option. Populates the arrays with the
c4019d52
KW
1448 # input code point's info that are needed for outputting more detailed
1449 # comments. If calling context wants a return, it is the end point of
1450 # any contiguous range of characters that share essentially the same info
1451
1452 my $i = shift;
1453 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1454
1455 $viacode[$i] = $perl_charname->value_of($i) || "";
1456
1457 # A character is generally printable if Unicode says it is,
1458 # but below we make sure that most Unicode general category 'C' types
1459 # aren't.
1460 $printable[$i] = $print->contains($i);
1461
1462 $annotate_char_type[$i] = $perl_charname->type_of($i) || 0;
1463
1464 # Only these two regular types are treated specially for annotations
1465 # purposes
1466 $annotate_char_type[$i] = 0 if $annotate_char_type[$i] != $CP_IN_NAME
1467 && $annotate_char_type[$i] != $HANGUL_SYLLABLE;
1468
1469 # Give a generic name to all code points that don't have a real name.
1470 # We output ranges, if applicable, for these. Also calculate the end
1471 # point of the range.
1472 my $end;
1473 if (! $viacode[$i]) {
1474 if ($gc-> table('Surrogate')->contains($i)) {
1475 $viacode[$i] = 'Surrogate';
1476 $annotate_char_type[$i] = $SURROGATE_TYPE;
1477 $printable[$i] = 0;
1478 $end = $gc->table('Surrogate')->containing_range($i)->end;
1479 }
1480 elsif ($gc-> table('Private_use')->contains($i)) {
1481 $viacode[$i] = 'Private Use';
1482 $annotate_char_type[$i] = $PRIVATE_USE_TYPE;
1483 $printable[$i] = 0;
1484 $end = $gc->table('Private_Use')->containing_range($i)->end;
1485 }
1486 elsif (Property::property_ref('Noncharacter_Code_Point')-> table('Y')->
1487 contains($i))
1488 {
1489 $viacode[$i] = 'Noncharacter';
1490 $annotate_char_type[$i] = $NONCHARACTER_TYPE;
1491 $printable[$i] = 0;
1492 $end = property_ref('Noncharacter_Code_Point')->table('Y')->
1493 containing_range($i)->end;
1494 }
1495 elsif ($gc-> table('Control')->contains($i)) {
1496 $viacode[$i] = 'Control';
1497 $annotate_char_type[$i] = $CONTROL_TYPE;
1498 $printable[$i] = 0;
1499 $end = 0x81 if $i == 0x80; # Hard-code this one known case
1500 }
1501 elsif ($gc-> table('Unassigned')->contains($i)) {
1502 $viacode[$i] = 'Unassigned, block=' . $block-> value_of($i);
1503 $annotate_char_type[$i] = $UNASSIGNED_TYPE;
1504 $printable[$i] = 0;
1505
1506 # Because we name the unassigned by the blocks they are in, it
1507 # can't go past the end of that block, and it also can't go past
1508 # the unassigned range it is in. The special table makes sure
1509 # that the non-characters, which are unassigned, are separated
1510 # out.
1511 $end = min($block->containing_range($i)->end,
1512 $unassigned_sans_noncharacters-> containing_range($i)->
1513 end);
13ca76ff
KW
1514 }
1515 else {
1516 Carp::my_carp_bug("Can't figure out how to annotate "
1517 . sprintf("U+%04X", $i)
1518 . ". Proceeding anyway.");
c4019d52
KW
1519 $viacode[$i] = 'UNKNOWN';
1520 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1521 $printable[$i] = 0;
1522 }
1523 }
1524
1525 # Here, has a name, but if it's one in which the code point number is
1526 # appended to the name, do that.
1527 elsif ($annotate_char_type[$i] == $CP_IN_NAME) {
1528 $viacode[$i] .= sprintf("-%04X", $i);
1529 $end = $perl_charname->containing_range($i)->end;
1530 }
1531
1532 # And here, has a name, but if it's a hangul syllable one, replace it with
1533 # the correct name from the Unicode algorithm
1534 elsif ($annotate_char_type[$i] == $HANGUL_SYLLABLE) {
1535 use integer;
1536 my $SIndex = $i - $SBase;
1537 my $L = $LBase + $SIndex / $NCount;
1538 my $V = $VBase + ($SIndex % $NCount) / $TCount;
1539 my $T = $TBase + $SIndex % $TCount;
1540 $viacode[$i] = "HANGUL SYLLABLE $Jamo{$L}$Jamo{$V}";
1541 $viacode[$i] .= $Jamo{$T} if $T != $TBase;
1542 $end = $perl_charname->containing_range($i)->end;
1543 }
1544
1545 return if ! defined wantarray;
1546 return $i if ! defined $end; # If not a range, return the input
1547
1548 # Save this whole range so can find the end point quickly
1549 $annotate_ranges->add_map($i, $end, $end);
1550
1551 return $end;
1552}
1553
23e33b60
KW
1554# Commented code below should work on Perl 5.8.
1555## This 'require' doesn't necessarily work in miniperl, and even if it does,
1556## the native perl version of it (which is what would operate under miniperl)
1557## is extremely slow, as it does a string eval every call.
1558#my $has_fast_scalar_util = $\18 !~ /miniperl/
1559# && defined eval "require Scalar::Util";
1560#
1561#sub objaddr($) {
1562# # Returns the address of the blessed input object. Uses the XS version if
1563# # available. It doesn't check for blessedness because that would do a
1564# # string eval every call, and the program is structured so that this is
1565# # never called for a non-blessed object.
1566#
1567# return Scalar::Util::refaddr($_[0]) if $has_fast_scalar_util;
1568#
1569# # Check at least that is a ref.
1570# my $pkg = ref($_[0]) or return undef;
1571#
1572# # Change to a fake package to defeat any overloaded stringify
1573# bless $_[0], 'main::Fake';
1574#
1575# # Numifying a ref gives its address.
051df77b 1576# my $addr = pack 'J', $_[0];
23e33b60
KW
1577#
1578# # Return to original class
1579# bless $_[0], $pkg;
1580# return $addr;
1581#}
1582
99870f4d
KW
1583sub max ($$) {
1584 my $a = shift;
1585 my $b = shift;
1586 return $a if $a >= $b;
1587 return $b;
1588}
1589
1590sub min ($$) {
1591 my $a = shift;
1592 my $b = shift;
1593 return $a if $a <= $b;
1594 return $b;
1595}
1596
1597sub clarify_number ($) {
1598 # This returns the input number with underscores inserted every 3 digits
1599 # in large (5 digits or more) numbers. Input must be entirely digits, not
1600 # checked.
1601
1602 my $number = shift;
1603 my $pos = length($number) - 3;
1604 return $number if $pos <= 1;
1605 while ($pos > 0) {
1606 substr($number, $pos, 0) = '_';
1607 $pos -= 3;
5beb625e 1608 }
99870f4d 1609 return $number;
99598c8c
JH
1610}
1611
12ac2576 1612
99870f4d 1613package Carp;
7ebf06b3 1614
99870f4d
KW
1615# These routines give a uniform treatment of messages in this program. They
1616# are placed in the Carp package to cause the stack trace to not include them,
1617# although an alternative would be to use another package and set @CARP_NOT
1618# for it.
12ac2576 1619
99870f4d 1620our $Verbose = 1 if main::DEBUG; # Useful info when debugging
12ac2576 1621
99f78760
KW
1622# This is a work-around suggested by Nicholas Clark to fix a problem with Carp
1623# and overload trying to load Scalar:Util under miniperl. See
1624# http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2009-11/msg01057.html
1625undef $overload::VERSION;
1626
99870f4d
KW
1627sub my_carp {
1628 my $message = shift || "";
1629 my $nofold = shift || 0;
7ebf06b3 1630
99870f4d
KW
1631 if ($message) {
1632 $message = main::join_lines($message);
1633 $message =~ s/^$0: *//; # Remove initial program name
1634 $message =~ s/[.;,]+$//; # Remove certain ending punctuation
1635 $message = "\n$0: $message;";
12ac2576 1636
99870f4d
KW
1637 # Fold the message with program name, semi-colon end punctuation
1638 # (which looks good with the message that carp appends to it), and a
1639 # hanging indent for continuation lines.
1640 $message = main::simple_fold($message, "", 4) unless $nofold;
1641 $message =~ s/\n$//; # Remove the trailing nl so what carp
1642 # appends is to the same line
1643 }
12ac2576 1644
99870f4d 1645 return $message if defined wantarray; # If a caller just wants the msg
12ac2576 1646
99870f4d
KW
1647 carp $message;
1648 return;
1649}
7ebf06b3 1650
99870f4d
KW
1651sub my_carp_bug {
1652 # This is called when it is clear that the problem is caused by a bug in
1653 # this program.
7ebf06b3 1654
99870f4d
KW
1655 my $message = shift;
1656 $message =~ s/^$0: *//;
1657 $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");
1658 carp $message;
1659 return;
1660}
7ebf06b3 1661
99870f4d
KW
1662sub carp_too_few_args {
1663 if (@_ != 2) {
1664 my_carp_bug("Wrong number of arguments: to 'carp_too_few_arguments'. No action taken.");
1665 return;
12ac2576 1666 }
7ebf06b3 1667
99870f4d
KW
1668 my $args_ref = shift;
1669 my $count = shift;
7ebf06b3 1670
99870f4d
KW
1671 my_carp_bug("Need at least $count arguments to "
1672 . (caller 1)[3]
1673 . ". Instead got: '"
1674 . join ', ', @$args_ref
1675 . "'. No action taken.");
1676 return;
12ac2576
JP
1677}
1678
99870f4d
KW
1679sub carp_extra_args {
1680 my $args_ref = shift;
1681 my_carp_bug("Too many arguments to 'carp_extra_args': (" . join(', ', @_) . "); Extras ignored.") if @_;
12ac2576 1682
99870f4d
KW
1683 unless (ref $args_ref) {
1684 my_carp_bug("Argument to 'carp_extra_args' ($args_ref) must be a ref. Not checking arguments.");
1685 return;
1686 }
1687 my ($package, $file, $line) = caller;
1688 my $subroutine = (caller 1)[3];
cf25bb62 1689
99870f4d
KW
1690 my $list;
1691 if (ref $args_ref eq 'HASH') {
1692 foreach my $key (keys %$args_ref) {
1693 $args_ref->{$key} = $UNDEF unless defined $args_ref->{$key};
cf25bb62 1694 }
99870f4d 1695 $list = join ', ', each %{$args_ref};
cf25bb62 1696 }
99870f4d
KW
1697 elsif (ref $args_ref eq 'ARRAY') {
1698 foreach my $arg (@$args_ref) {
1699 $arg = $UNDEF unless defined $arg;
1700 }
1701 $list = join ', ', @$args_ref;
1702 }
1703 else {
1704 my_carp_bug("Can't cope with ref "
1705 . ref($args_ref)
1706 . " . argument to 'carp_extra_args'. Not checking arguments.");
1707 return;
1708 }
1709
1710 my_carp_bug("Unrecognized parameters in options: '$list' to $subroutine. Skipped.");
1711 return;
d73e5302
JH
1712}
1713
99870f4d
KW
1714package main;
1715
1716{ # Closure
1717
1718 # This program uses the inside-out method for objects, as recommended in
1719 # "Perl Best Practices". This closure aids in generating those. There
1720 # are two routines. setup_package() is called once per package to set
1721 # things up, and then set_access() is called for each hash representing a
1722 # field in the object. These routines arrange for the object to be
1723 # properly destroyed when no longer used, and for standard accessor
1724 # functions to be generated. If you need more complex accessors, just
1725 # write your own and leave those accesses out of the call to set_access().
1726 # More details below.
1727
1728 my %constructor_fields; # fields that are to be used in constructors; see
1729 # below
1730
1731 # The values of this hash will be the package names as keys to other
1732 # hashes containing the name of each field in the package as keys, and
1733 # references to their respective hashes as values.
1734 my %package_fields;
1735
1736 sub setup_package {
1737 # Sets up the package, creating standard DESTROY and dump methods
1738 # (unless already defined). The dump method is used in debugging by
1739 # simple_dumper().
1740 # The optional parameters are:
1741 # a) a reference to a hash, that gets populated by later
1742 # set_access() calls with one of the accesses being
1743 # 'constructor'. The caller can then refer to this, but it is
1744 # not otherwise used by these two routines.
1745 # b) a reference to a callback routine to call during destruction
1746 # of the object, before any fields are actually destroyed
1747
1748 my %args = @_;
1749 my $constructor_ref = delete $args{'Constructor_Fields'};
1750 my $destroy_callback = delete $args{'Destroy_Callback'};
1751 Carp::carp_extra_args(\@_) if main::DEBUG && %args;
1752
1753 my %fields;
1754 my $package = (caller)[0];
1755
1756 $package_fields{$package} = \%fields;
1757 $constructor_fields{$package} = $constructor_ref;
1758
1759 unless ($package->can('DESTROY')) {
1760 my $destroy_name = "${package}::DESTROY";
1761 no strict "refs";
1762
1763 # Use typeglob to give the anonymous subroutine the name we want
1764 *$destroy_name = sub {
1765 my $self = shift;
ffe43484 1766 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1767
1768 $self->$destroy_callback if $destroy_callback;
1769 foreach my $field (keys %{$package_fields{$package}}) {
1770 #print STDERR __LINE__, ": Destroying ", ref $self, " ", sprintf("%04X", $addr), ": ", $field, "\n";
1771 delete $package_fields{$package}{$field}{$addr};
1772 }
1773 return;
1774 }
1775 }
1776
1777 unless ($package->can('dump')) {
1778 my $dump_name = "${package}::dump";
1779 no strict "refs";
1780 *$dump_name = sub {
1781 my $self = shift;
1782 return dump_inside_out($self, $package_fields{$package}, @_);
1783 }
1784 }
1785 return;
1786 }
1787
1788 sub set_access {
1789 # Arrange for the input field to be garbage collected when no longer
1790 # needed. Also, creates standard accessor functions for the field
1791 # based on the optional parameters-- none if none of these parameters:
1792 # 'addable' creates an 'add_NAME()' accessor function.
1793 # 'readable' or 'readable_array' creates a 'NAME()' accessor
1794 # function.
1795 # 'settable' creates a 'set_NAME()' accessor function.
1796 # 'constructor' doesn't create an accessor function, but adds the
1797 # field to the hash that was previously passed to
1798 # setup_package();
1799 # Any of the accesses can be abbreviated down, so that 'a', 'ad',
1800 # 'add' etc. all mean 'addable'.
1801 # The read accessor function will work on both array and scalar
1802 # values. If another accessor in the parameter list is 'a', the read
1803 # access assumes an array. You can also force it to be array access
1804 # by specifying 'readable_array' instead of 'readable'
1805 #
1806 # A sort-of 'protected' access can be set-up by preceding the addable,
1807 # readable or settable with some initial portion of 'protected_' (but,
1808 # the underscore is required), like 'p_a', 'pro_set', etc. The
1809 # "protection" is only by convention. All that happens is that the
1810 # accessor functions' names begin with an underscore. So instead of
1811 # calling set_foo, the call is _set_foo. (Real protection could be
c1739a4a 1812 # accomplished by having a new subroutine, end_package, called at the
99870f4d
KW
1813 # end of each package, and then storing the __LINE__ ranges and
1814 # checking them on every accessor. But that is way overkill.)
1815
1816 # We create anonymous subroutines as the accessors and then use
1817 # typeglobs to assign them to the proper package and name
1818
1819 my $name = shift; # Name of the field
1820 my $field = shift; # Reference to the inside-out hash containing the
1821 # field
1822
1823 my $package = (caller)[0];
1824
1825 if (! exists $package_fields{$package}) {
1826 croak "$0: Must call 'setup_package' before 'set_access'";
1827 }
d73e5302 1828
99870f4d
KW
1829 # Stash the field so DESTROY can get it.
1830 $package_fields{$package}{$name} = $field;
cf25bb62 1831
99870f4d
KW
1832 # Remaining arguments are the accessors. For each...
1833 foreach my $access (@_) {
1834 my $access = lc $access;
cf25bb62 1835
99870f4d 1836 my $protected = "";
cf25bb62 1837
99870f4d
KW
1838 # Match the input as far as it goes.
1839 if ($access =~ /^(p[^_]*)_/) {
1840 $protected = $1;
1841 if (substr('protected_', 0, length $protected)
1842 eq $protected)
1843 {
1844
1845 # Add 1 for the underscore not included in $protected
1846 $access = substr($access, length($protected) + 1);
1847 $protected = '_';
1848 }
1849 else {
1850 $protected = "";
1851 }
1852 }
1853
1854 if (substr('addable', 0, length $access) eq $access) {
1855 my $subname = "${package}::${protected}add_$name";
1856 no strict "refs";
1857
1858 # add_ accessor. Don't add if already there, which we
1859 # determine using 'eq' for scalars and '==' otherwise.
1860 *$subname = sub {
1861 use strict "refs";
1862 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
1863 my $self = shift;
1864 my $value = shift;
ffe43484 1865 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1866 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1867 if (ref $value) {
f998e60c 1868 return if grep { $value == $_ } @{$field->{$addr}};
99870f4d
KW
1869 }
1870 else {
f998e60c 1871 return if grep { $value eq $_ } @{$field->{$addr}};
99870f4d 1872 }
f998e60c 1873 push @{$field->{$addr}}, $value;
99870f4d
KW
1874 return;
1875 }
1876 }
1877 elsif (substr('constructor', 0, length $access) eq $access) {
1878 if ($protected) {
1879 Carp::my_carp_bug("Can't set-up 'protected' constructors")
1880 }
1881 else {
1882 $constructor_fields{$package}{$name} = $field;
1883 }
1884 }
1885 elsif (substr('readable_array', 0, length $access) eq $access) {
1886
1887 # Here has read access. If one of the other parameters for
1888 # access is array, or this one specifies array (by being more
1889 # than just 'readable_'), then create a subroutine that
1890 # assumes the data is an array. Otherwise just a scalar
1891 my $subname = "${package}::${protected}$name";
1892 if (grep { /^a/i } @_
1893 or length($access) > length('readable_'))
1894 {
1895 no strict "refs";
1896 *$subname = sub {
1897 use strict "refs";
23e33b60 1898 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
ffe43484 1899 my $addr = do { no overloading; pack 'J', $_[0]; };
99870f4d
KW
1900 if (ref $field->{$addr} ne 'ARRAY') {
1901 my $type = ref $field->{$addr};
1902 $type = 'scalar' unless $type;
1903 Carp::my_carp_bug("Trying to read $name as an array when it is a $type. Big problems.");
1904 return;
1905 }
1906 return scalar @{$field->{$addr}} unless wantarray;
1907
1908 # Make a copy; had problems with caller modifying the
1909 # original otherwise
1910 my @return = @{$field->{$addr}};
1911 return @return;
1912 }
1913 }
1914 else {
1915
1916 # Here not an array value, a simpler function.
1917 no strict "refs";
1918 *$subname = sub {
1919 use strict "refs";
23e33b60 1920 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
f998e60c 1921 no overloading;
051df77b 1922 return $field->{pack 'J', $_[0]};
99870f4d
KW
1923 }
1924 }
1925 }
1926 elsif (substr('settable', 0, length $access) eq $access) {
1927 my $subname = "${package}::${protected}set_$name";
1928 no strict "refs";
1929 *$subname = sub {
1930 use strict "refs";
23e33b60
KW
1931 if (main::DEBUG) {
1932 return Carp::carp_too_few_args(\@_, 2) if @_ < 2;
1933 Carp::carp_extra_args(\@_) if @_ > 2;
1934 }
1935 # $self is $_[0]; $value is $_[1]
f998e60c 1936 no overloading;
051df77b 1937 $field->{pack 'J', $_[0]} = $_[1];
99870f4d
KW
1938 return;
1939 }
1940 }
1941 else {
1942 Carp::my_carp_bug("Unknown accessor type $access. No accessor set.");
1943 }
cf25bb62 1944 }
99870f4d 1945 return;
cf25bb62 1946 }
99870f4d
KW
1947}
1948
1949package Input_file;
1950
1951# All input files use this object, which stores various attributes about them,
1952# and provides for convenient, uniform handling. The run method wraps the
1953# processing. It handles all the bookkeeping of opening, reading, and closing
1954# the file, returning only significant input lines.
1955#
1956# Each object gets a handler which processes the body of the file, and is
1957# called by run(). Most should use the generic, default handler, which has
1958# code scrubbed to handle things you might not expect. A handler should
1959# basically be a while(next_line()) {...} loop.
1960#
1961# You can also set up handlers to
1962# 1) call before the first line is read for pre processing
1963# 2) call to adjust each line of the input before the main handler gets them
1964# 3) call upon EOF before the main handler exits its loop
1965# 4) call at the end for post processing
1966#
1967# $_ is used to store the input line, and is to be filtered by the
1968# each_line_handler()s. So, if the format of the line is not in the desired
1969# format for the main handler, these are used to do that adjusting. They can
1970# be stacked (by enclosing them in an [ anonymous array ] in the constructor,
1971# so the $_ output of one is used as the input to the next. None of the other
1972# handlers are stackable, but could easily be changed to be so.
1973#
1974# Most of the handlers can call insert_lines() or insert_adjusted_lines()
1975# which insert the parameters as lines to be processed before the next input
1976# file line is read. This allows the EOF handler to flush buffers, for
1977# example. The difference between the two routines is that the lines inserted
1978# by insert_lines() are subjected to the each_line_handler()s. (So if you
1979# called it from such a handler, you would get infinite recursion.) Lines
1980# inserted by insert_adjusted_lines() go directly to the main handler without
1981# any adjustments. If the post-processing handler calls any of these, there
1982# will be no effect. Some error checking for these conditions could be added,
1983# but it hasn't been done.
1984#
1985# carp_bad_line() should be called to warn of bad input lines, which clears $_
1986# to prevent further processing of the line. This routine will output the
1987# message as a warning once, and then keep a count of the lines that have the
1988# same message, and output that count at the end of the file's processing.
1989# This keeps the number of messages down to a manageable amount.
1990#
1991# get_missings() should be called to retrieve any @missing input lines.
1992# Messages will be raised if this isn't done if the options aren't to ignore
1993# missings.
1994
1995sub trace { return main::trace(@_); }
1996
99870f4d
KW
1997{ # Closure
1998 # Keep track of fields that are to be put into the constructor.
1999 my %constructor_fields;
2000
2001 main::setup_package(Constructor_Fields => \%constructor_fields);
2002
2003 my %file; # Input file name, required
2004 main::set_access('file', \%file, qw{ c r });
2005
2006 my %first_released; # Unicode version file was first released in, required
2007 main::set_access('first_released', \%first_released, qw{ c r });
2008
2009 my %handler; # Subroutine to process the input file, defaults to
2010 # 'process_generic_property_file'
2011 main::set_access('handler', \%handler, qw{ c });
2012
2013 my %property;
2014 # name of property this file is for. defaults to none, meaning not
2015 # applicable, or is otherwise determinable, for example, from each line.
2016 main::set_access('property', \%property, qw{ c });
2017
2018 my %optional;
2019 # If this is true, the file is optional. If not present, no warning is
2020 # output. If it is present, the string given by this parameter is
2021 # evaluated, and if false the file is not processed.
2022 main::set_access('optional', \%optional, 'c', 'r');
2023
2024 my %non_skip;
2025 # This is used for debugging, to skip processing of all but a few input
2026 # files. Add 'non_skip => 1' to the constructor for those files you want
2027 # processed when you set the $debug_skip global.
2028 main::set_access('non_skip', \%non_skip, 'c');
2029
37e2e78e 2030 my %skip;
09ca89ce
KW
2031 # This is used to skip processing of this input file semi-permanently,
2032 # when it evaluates to true. The value should be the reason the file is
2033 # being skipped. It is used for files that we aren't planning to process
2034 # anytime soon, but want to allow to be in the directory and not raise a
2035 # message that we are not handling. Mostly for test files. This is in
2036 # contrast to the non_skip element, which is supposed to be used very
2037 # temporarily for debugging. Sets 'optional' to 1. Also, files that we
2038 # pretty much will never look at can be placed in the global
1fec9f60 2039 # %ignored_files instead. Ones used here will be added to %skipped files
37e2e78e
KW
2040 main::set_access('skip', \%skip, 'c');
2041
99870f4d
KW
2042 my %each_line_handler;
2043 # list of subroutines to look at and filter each non-comment line in the
2044 # file. defaults to none. The subroutines are called in order, each is
2045 # to adjust $_ for the next one, and the final one adjusts it for
2046 # 'handler'
2047 main::set_access('each_line_handler', \%each_line_handler, 'c');
2048
2049 my %has_missings_defaults;
2050 # ? Are there lines in the file giving default values for code points
2051 # missing from it?. Defaults to NO_DEFAULTS. Otherwise NOT_IGNORED is
2052 # the norm, but IGNORED means it has such lines, but the handler doesn't
2053 # use them. Having these three states allows us to catch changes to the
2054 # UCD that this program should track
2055 main::set_access('has_missings_defaults',
2056 \%has_missings_defaults, qw{ c r });
2057
2058 my %pre_handler;
2059 # Subroutine to call before doing anything else in the file. If undef, no
2060 # such handler is called.
2061 main::set_access('pre_handler', \%pre_handler, qw{ c });
2062
2063 my %eof_handler;
2064 # Subroutine to call upon getting an EOF on the input file, but before
2065 # that is returned to the main handler. This is to allow buffers to be
2066 # flushed. The handler is expected to call insert_lines() or
2067 # insert_adjusted() with the buffered material
2068 main::set_access('eof_handler', \%eof_handler, qw{ c r });
2069
2070 my %post_handler;
2071 # Subroutine to call after all the lines of the file are read in and
2072 # processed. If undef, no such handler is called.
2073 main::set_access('post_handler', \%post_handler, qw{ c });
2074
2075 my %progress_message;
2076 # Message to print to display progress in lieu of the standard one
2077 main::set_access('progress_message', \%progress_message, qw{ c });
2078
2079 my %handle;
2080 # cache open file handle, internal. Is undef if file hasn't been
2081 # processed at all, empty if has;
2082 main::set_access('handle', \%handle);
2083
2084 my %added_lines;
2085 # cache of lines added virtually to the file, internal
2086 main::set_access('added_lines', \%added_lines);
2087
2088 my %errors;
2089 # cache of errors found, internal
2090 main::set_access('errors', \%errors);
2091
2092 my %missings;
2093 # storage of '@missing' defaults lines
2094 main::set_access('missings', \%missings);
2095
2096 sub new {
2097 my $class = shift;
2098
2099 my $self = bless \do{ my $anonymous_scalar }, $class;
ffe43484 2100 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2101
2102 # Set defaults
2103 $handler{$addr} = \&main::process_generic_property_file;
2104 $non_skip{$addr} = 0;
37e2e78e 2105 $skip{$addr} = 0;
99870f4d
KW
2106 $has_missings_defaults{$addr} = $NO_DEFAULTS;
2107 $handle{$addr} = undef;
2108 $added_lines{$addr} = [ ];
2109 $each_line_handler{$addr} = [ ];
2110 $errors{$addr} = { };
2111 $missings{$addr} = [ ];
2112
2113 # Two positional parameters.
99f78760 2114 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
99870f4d
KW
2115 $file{$addr} = main::internal_file_to_platform(shift);
2116 $first_released{$addr} = shift;
2117
2118 # The rest of the arguments are key => value pairs
2119 # %constructor_fields has been set up earlier to list all possible
2120 # ones. Either set or push, depending on how the default has been set
2121 # up just above.
2122 my %args = @_;
2123 foreach my $key (keys %args) {
2124 my $argument = $args{$key};
2125
2126 # Note that the fields are the lower case of the constructor keys
2127 my $hash = $constructor_fields{lc $key};
2128 if (! defined $hash) {
2129 Carp::my_carp_bug("Unrecognized parameters '$key => $argument' to new() for $self. Skipped");
2130 next;
2131 }
2132 if (ref $hash->{$addr} eq 'ARRAY') {
2133 if (ref $argument eq 'ARRAY') {
2134 foreach my $argument (@{$argument}) {
2135 next if ! defined $argument;
2136 push @{$hash->{$addr}}, $argument;
2137 }
2138 }
2139 else {
2140 push @{$hash->{$addr}}, $argument if defined $argument;
2141 }
2142 }
2143 else {
2144 $hash->{$addr} = $argument;
2145 }
2146 delete $args{$key};
2147 };
2148
2149 # If the file has a property for it, it means that the property is not
2150 # listed in the file's entries. So add a handler to the list of line
2151 # handlers to insert the property name into the lines, to provide a
2152 # uniform interface to the final processing subroutine.
2153 # the final code doesn't have to worry about that.
2154 if ($property{$addr}) {
2155 push @{$each_line_handler{$addr}}, \&_insert_property_into_line;
2156 }
2157
2158 if ($non_skip{$addr} && ! $debug_skip && $verbosity) {
2159 print "Warning: " . __PACKAGE__ . " constructor for $file{$addr} has useless 'non_skip' in it\n";
a3a8c5f0 2160 }
99870f4d 2161
09ca89ce
KW
2162 # If skipping, set to optional, and add to list of ignored files,
2163 # including its reason
2164 if ($skip{$addr}) {
2165 $optional{$addr} = 1;
1fec9f60 2166 $skipped_files{$file{$addr}} = $skip{$addr}
09ca89ce 2167 }
37e2e78e 2168
99870f4d 2169 return $self;
d73e5302
JH
2170 }
2171
cf25bb62 2172
99870f4d
KW
2173 use overload
2174 fallback => 0,
2175 qw("") => "_operator_stringify",
2176 "." => \&main::_operator_dot,
2177 ;
cf25bb62 2178
99870f4d
KW
2179 sub _operator_stringify {
2180 my $self = shift;
cf25bb62 2181
99870f4d 2182 return __PACKAGE__ . " object for " . $self->file;
d73e5302 2183 }
d73e5302 2184
99870f4d
KW
2185 # flag to make sure extracted files are processed early
2186 my $seen_non_extracted_non_age = 0;
d73e5302 2187
99870f4d
KW
2188 sub run {
2189 # Process the input object $self. This opens and closes the file and
2190 # calls all the handlers for it. Currently, this can only be called
2191 # once per file, as it destroy's the EOF handler
d73e5302 2192
99870f4d
KW
2193 my $self = shift;
2194 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
b6922eda 2195
ffe43484 2196 my $addr = do { no overloading; pack 'J', $self; };
b6922eda 2197
99870f4d 2198 my $file = $file{$addr};
d73e5302 2199
99870f4d
KW
2200 # Don't process if not expecting this file (because released later
2201 # than this Unicode version), and isn't there. This means if someone
2202 # copies it into an earlier version's directory, we will go ahead and
2203 # process it.
2204 return if $first_released{$addr} gt $v_version && ! -e $file;
2205
2206 # If in debugging mode and this file doesn't have the non-skip
2207 # flag set, and isn't one of the critical files, skip it.
2208 if ($debug_skip
2209 && $first_released{$addr} ne v0
2210 && ! $non_skip{$addr})
2211 {
2212 print "Skipping $file in debugging\n" if $verbosity;
2213 return;
2214 }
2215
2216 # File could be optional
37e2e78e 2217 if ($optional{$addr}) {
99870f4d
KW
2218 return unless -e $file;
2219 my $result = eval $optional{$addr};
2220 if (! defined $result) {
2221 Carp::my_carp_bug("Got '$@' when tried to eval $optional{$addr}. $file Skipped.");
2222 return;
2223 }
2224 if (! $result) {
2225 if ($verbosity) {
2226 print STDERR "Skipping processing input file '$file' because '$optional{$addr}' is not true\n";
2227 }
2228 return;
2229 }
2230 }
2231
2232 if (! defined $file || ! -e $file) {
2233
2234 # If the file doesn't exist, see if have internal data for it
2235 # (based on first_released being 0).
2236 if ($first_released{$addr} eq v0) {
2237 $handle{$addr} = 'pretend_is_open';
2238 }
2239 else {
2240 if (! $optional{$addr} # File could be optional
2241 && $v_version ge $first_released{$addr})
2242 {
2243 print STDERR "Skipping processing input file '$file' because not found\n" if $v_version ge $first_released{$addr};
2244 }
2245 return;
2246 }
2247 }
2248 else {
2249
37e2e78e
KW
2250 # Here, the file exists. Some platforms may change the case of
2251 # its name
99870f4d 2252 if ($seen_non_extracted_non_age) {
517956bf 2253 if ($file =~ /$EXTRACTED/i) {
1675ea0d 2254 Carp::my_carp_bug(main::join_lines(<<END
99f78760 2255$file should be processed just after the 'Prop...Alias' files, and before
99870f4d
KW
2256anything not in the $EXTRACTED_DIR directory. Proceeding, but the results may
2257have subtle problems
2258END
2259 ));
2260 }
2261 }
2262 elsif ($EXTRACTED_DIR
2263 && $first_released{$addr} ne v0
517956bf
CB
2264 && $file !~ /$EXTRACTED/i
2265 && lc($file) ne 'dage.txt')
99870f4d
KW
2266 {
2267 # We don't set this (by the 'if' above) if we have no
2268 # extracted directory, so if running on an early version,
2269 # this test won't work. Not worth worrying about.
2270 $seen_non_extracted_non_age = 1;
2271 }
2272
2273 # And mark the file as having being processed, and warn if it
2274 # isn't a file we are expecting. As we process the files,
2275 # they are deleted from the hash, so any that remain at the
2276 # end of the program are files that we didn't process.
517956bf 2277 my $fkey = File::Spec->rel2abs($file);
faf3cf6b
KW
2278 my $expecting = delete $potential_files{lc($fkey)};
2279
678f13d5
KW
2280 Carp::my_carp("Was not expecting '$file'.") if
2281 ! $expecting
99870f4d
KW
2282 && ! defined $handle{$addr};
2283
37e2e78e
KW
2284 # Having deleted from expected files, we can quit if not to do
2285 # anything. Don't print progress unless really want verbosity
2286 if ($skip{$addr}) {
2287 print "Skipping $file.\n" if $verbosity >= $VERBOSE;
2288 return;
2289 }
2290
99870f4d
KW
2291 # Open the file, converting the slashes used in this program
2292 # into the proper form for the OS
2293 my $file_handle;
2294 if (not open $file_handle, "<", $file) {
2295 Carp::my_carp("Can't open $file. Skipping: $!");
2296 return 0;
2297 }
2298 $handle{$addr} = $file_handle; # Cache the open file handle
2299 }
2300
2301 if ($verbosity >= $PROGRESS) {
2302 if ($progress_message{$addr}) {
2303 print "$progress_message{$addr}\n";
2304 }
2305 else {
2306 # If using a virtual file, say so.
2307 print "Processing ", (-e $file)
2308 ? $file
2309 : "substitute $file",
2310 "\n";
2311 }
2312 }
2313
2314
2315 # Call any special handler for before the file.
2316 &{$pre_handler{$addr}}($self) if $pre_handler{$addr};
2317
2318 # Then the main handler
2319 &{$handler{$addr}}($self);
2320
2321 # Then any special post-file handler.
2322 &{$post_handler{$addr}}($self) if $post_handler{$addr};
2323
2324 # If any errors have been accumulated, output the counts (as the first
2325 # error message in each class was output when it was encountered).
2326 if ($errors{$addr}) {
2327 my $total = 0;
2328 my $types = 0;
2329 foreach my $error (keys %{$errors{$addr}}) {
2330 $total += $errors{$addr}->{$error};
2331 delete $errors{$addr}->{$error};
2332 $types++;
2333 }
2334 if ($total > 1) {
2335 my $message
2336 = "A total of $total lines had errors in $file. ";
2337
2338 $message .= ($types == 1)
2339 ? '(Only the first one was displayed.)'
2340 : '(Only the first of each type was displayed.)';
2341 Carp::my_carp($message);
2342 }
2343 }
2344
2345 if (@{$missings{$addr}}) {
2346 Carp::my_carp_bug("Handler for $file didn't look at all the \@missing lines. Generated tables likely are wrong");
2347 }
2348
2349 # If a real file handle, close it.
2350 close $handle{$addr} or Carp::my_carp("Can't close $file: $!") if
2351 ref $handle{$addr};
2352 $handle{$addr} = ""; # Uses empty to indicate that has already seen
2353 # the file, as opposed to undef
2354 return;
2355 }
2356
2357 sub next_line {
2358 # Sets $_ to be the next logical input line, if any. Returns non-zero
2359 # if such a line exists. 'logical' means that any lines that have
2360 # been added via insert_lines() will be returned in $_ before the file
2361 # is read again.
2362
2363 my $self = shift;
2364 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2365
ffe43484 2366 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2367
2368 # Here the file is open (or if the handle is not a ref, is an open
2369 # 'virtual' file). Get the next line; any inserted lines get priority
2370 # over the file itself.
2371 my $adjusted;
2372
2373 LINE:
2374 while (1) { # Loop until find non-comment, non-empty line
2375 #local $to_trace = 1 if main::DEBUG;
2376 my $inserted_ref = shift @{$added_lines{$addr}};
2377 if (defined $inserted_ref) {
2378 ($adjusted, $_) = @{$inserted_ref};
2379 trace $adjusted, $_ if main::DEBUG && $to_trace;
2380 return 1 if $adjusted;
2381 }
2382 else {
2383 last if ! ref $handle{$addr}; # Don't read unless is real file
2384 last if ! defined ($_ = readline $handle{$addr});
2385 }
2386 chomp;
2387 trace $_ if main::DEBUG && $to_trace;
2388
2389 # See if this line is the comment line that defines what property
2390 # value that code points that are not listed in the file should
2391 # have. The format or existence of these lines is not guaranteed
2392 # by Unicode since they are comments, but the documentation says
2393 # that this was added for machine-readability, so probably won't
2394 # change. This works starting in Unicode Version 5.0. They look
2395 # like:
2396 #
2397 # @missing: 0000..10FFFF; Not_Reordered
2398 # @missing: 0000..10FFFF; Decomposition_Mapping; <code point>
2399 # @missing: 0000..10FFFF; ; NaN
2400 #
2401 # Save the line for a later get_missings() call.
2402 if (/$missing_defaults_prefix/) {
2403 if ($has_missings_defaults{$addr} == $NO_DEFAULTS) {
2404 $self->carp_bad_line("Unexpected \@missing line. Assuming no missing entries");
2405 }
2406 elsif ($has_missings_defaults{$addr} == $NOT_IGNORED) {
2407 my @defaults = split /\s* ; \s*/x, $_;
2408
2409 # The first field is the @missing, which ends in a
2410 # semi-colon, so can safely shift.
2411 shift @defaults;
2412
2413 # Some of these lines may have empty field placeholders
2414 # which get in the way. An example is:
2415 # @missing: 0000..10FFFF; ; NaN
2416 # Remove them. Process starting from the top so the
2417 # splice doesn't affect things still to be looked at.
2418 for (my $i = @defaults - 1; $i >= 0; $i--) {
2419 next if $defaults[$i] ne "";
2420 splice @defaults, $i, 1;
2421 }
2422
2423 # What's left should be just the property (maybe) and the
2424 # default. Having only one element means it doesn't have
2425 # the property.
2426 my $default;
2427 my $property;
2428 if (@defaults >= 1) {
2429 if (@defaults == 1) {
2430 $default = $defaults[0];
2431 }
2432 else {
2433 $property = $defaults[0];
2434 $default = $defaults[1];
2435 }
2436 }
2437
2438 if (@defaults < 1
2439 || @defaults > 2
2440 || ($default =~ /^</
2441 && $default !~ /^<code *point>$/i
09f8d0ac
KW
2442 && $default !~ /^<none>$/i
2443 && $default !~ /^<script>$/i))
99870f4d
KW
2444 {
2445 $self->carp_bad_line("Unrecognized \@missing line: $_. Assuming no missing entries");
2446 }
2447 else {
2448
2449 # If the property is missing from the line, it should
2450 # be the one for the whole file
2451 $property = $property{$addr} if ! defined $property;
2452
2453 # Change <none> to the null string, which is what it
2454 # really means. If the default is the code point
2455 # itself, set it to <code point>, which is what
2456 # Unicode uses (but sometimes they've forgotten the
2457 # space)
2458 if ($default =~ /^<none>$/i) {
2459 $default = "";
2460 }
2461 elsif ($default =~ /^<code *point>$/i) {
2462 $default = $CODE_POINT;
2463 }
09f8d0ac
KW
2464 elsif ($default =~ /^<script>$/i) {
2465
2466 # Special case this one. Currently is from
2467 # ScriptExtensions.txt, and means for all unlisted
2468 # code points, use their Script property values.
2469 # For the code points not listed in that file, the
2470 # default value is 'Unknown'.
2471 $default = "Unknown";
2472 }
99870f4d
KW
2473
2474 # Store them as a sub-arrays with both components.
2475 push @{$missings{$addr}}, [ $default, $property ];
2476 }
2477 }
2478
2479 # There is nothing for the caller to process on this comment
2480 # line.
2481 next;
2482 }
2483
2484 # Remove comments and trailing space, and skip this line if the
2485 # result is empty
2486 s/#.*//;
2487 s/\s+$//;
2488 next if /^$/;
2489
2490 # Call any handlers for this line, and skip further processing of
2491 # the line if the handler sets the line to null.
2492 foreach my $sub_ref (@{$each_line_handler{$addr}}) {
2493 &{$sub_ref}($self);
2494 next LINE if /^$/;
2495 }
2496
2497 # Here the line is ok. return success.
2498 return 1;
2499 } # End of looping through lines.
2500
2501 # If there is an EOF handler, call it (only once) and if it generates
2502 # more lines to process go back in the loop to handle them.
2503 if ($eof_handler{$addr}) {
2504 &{$eof_handler{$addr}}($self);
2505 $eof_handler{$addr} = ""; # Currently only get one shot at it.
2506 goto LINE if $added_lines{$addr};
2507 }
2508
2509 # Return failure -- no more lines.
2510 return 0;
2511
2512 }
2513
2514# Not currently used, not fully tested.
2515# sub peek {
2516# # Non-destructive look-ahead one non-adjusted, non-comment, non-blank
2517# # record. Not callable from an each_line_handler(), nor does it call
2518# # an each_line_handler() on the line.
2519#
2520# my $self = shift;
ffe43484 2521# my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2522#
2523# foreach my $inserted_ref (@{$added_lines{$addr}}) {
2524# my ($adjusted, $line) = @{$inserted_ref};
2525# next if $adjusted;
2526#
2527# # Remove comments and trailing space, and return a non-empty
2528# # resulting line
2529# $line =~ s/#.*//;
2530# $line =~ s/\s+$//;
2531# return $line if $line ne "";
2532# }
2533#
2534# return if ! ref $handle{$addr}; # Don't read unless is real file
2535# while (1) { # Loop until find non-comment, non-empty line
2536# local $to_trace = 1 if main::DEBUG;
2537# trace $_ if main::DEBUG && $to_trace;
2538# return if ! defined (my $line = readline $handle{$addr});
2539# chomp $line;
2540# push @{$added_lines{$addr}}, [ 0, $line ];
2541#
2542# $line =~ s/#.*//;
2543# $line =~ s/\s+$//;
2544# return $line if $line ne "";
2545# }
2546#
2547# return;
2548# }
2549
2550
2551 sub insert_lines {
2552 # Lines can be inserted so that it looks like they were in the input
2553 # file at the place it was when this routine is called. See also
2554 # insert_adjusted_lines(). Lines inserted via this routine go through
2555 # any each_line_handler()
2556
2557 my $self = shift;
2558
2559 # Each inserted line is an array, with the first element being 0 to
2560 # indicate that this line hasn't been adjusted, and needs to be
2561 # processed.
f998e60c 2562 no overloading;
051df77b 2563 push @{$added_lines{pack 'J', $self}}, map { [ 0, $_ ] } @_;
99870f4d
KW
2564 return;
2565 }
2566
2567 sub insert_adjusted_lines {
2568 # Lines can be inserted so that it looks like they were in the input
2569 # file at the place it was when this routine is called. See also
2570 # insert_lines(). Lines inserted via this routine are already fully
2571 # adjusted, ready to be processed; each_line_handler()s handlers will
2572 # not be called. This means this is not a completely general
2573 # facility, as only the last each_line_handler on the stack should
2574 # call this. It could be made more general, by passing to each of the
2575 # line_handlers their position on the stack, which they would pass on
2576 # to this routine, and that would replace the boolean first element in
2577 # the anonymous array pushed here, so that the next_line routine could
2578 # use that to call only those handlers whose index is after it on the
2579 # stack. But this is overkill for what is needed now.
2580
2581 my $self = shift;
2582 trace $_[0] if main::DEBUG && $to_trace;
2583
2584 # Each inserted line is an array, with the first element being 1 to
2585 # indicate that this line has been adjusted
f998e60c 2586 no overloading;
051df77b 2587 push @{$added_lines{pack 'J', $self}}, map { [ 1, $_ ] } @_;
99870f4d
KW
2588 return;
2589 }
2590
2591 sub get_missings {
2592 # Returns the stored up @missings lines' values, and clears the list.
2593 # The values are in an array, consisting of the default in the first
2594 # element, and the property in the 2nd. However, since these lines
2595 # can be stacked up, the return is an array of all these arrays.
2596
2597 my $self = shift;
2598 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2599
ffe43484 2600 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2601
2602 # If not accepting a list return, just return the first one.
2603 return shift @{$missings{$addr}} unless wantarray;
2604
2605 my @return = @{$missings{$addr}};
2606 undef @{$missings{$addr}};
2607 return @return;
2608 }
2609
2610 sub _insert_property_into_line {
2611 # Add a property field to $_, if this file requires it.
2612
f998e60c 2613 my $self = shift;
ffe43484 2614 my $addr = do { no overloading; pack 'J', $self; };
f998e60c 2615 my $property = $property{$addr};
99870f4d
KW
2616 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2617
2618 $_ =~ s/(;|$)/; $property$1/;
2619 return;
2620 }
2621
2622 sub carp_bad_line {
2623 # Output consistent error messages, using either a generic one, or the
2624 # one given by the optional parameter. To avoid gazillions of the
2625 # same message in case the syntax of a file is way off, this routine
2626 # only outputs the first instance of each message, incrementing a
2627 # count so the totals can be output at the end of the file.
2628
2629 my $self = shift;
2630 my $message = shift;
2631 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2632
ffe43484 2633 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2634
2635 $message = 'Unexpected line' unless $message;
2636
2637 # No trailing punctuation so as to fit with our addenda.
2638 $message =~ s/[.:;,]$//;
2639
2640 # If haven't seen this exact message before, output it now. Otherwise
2641 # increment the count of how many times it has occurred
2642 unless ($errors{$addr}->{$message}) {
2643 Carp::my_carp("$message in '$_' in "
f998e60c 2644 . $file{$addr}
99870f4d
KW
2645 . " at line $.. Skipping this line;");
2646 $errors{$addr}->{$message} = 1;
2647 }
2648 else {
2649 $errors{$addr}->{$message}++;
2650 }
2651
2652 # Clear the line to prevent any further (meaningful) processing of it.
2653 $_ = "";
2654
2655 return;
2656 }
2657} # End closure
2658
2659package Multi_Default;
2660
2661# Certain properties in early versions of Unicode had more than one possible
2662# default for code points missing from the files. In these cases, one
2663# default applies to everything left over after all the others are applied,
2664# and for each of the others, there is a description of which class of code
2665# points applies to it. This object helps implement this by storing the
2666# defaults, and for all but that final default, an eval string that generates
2667# the class that it applies to.
2668
2669
2670{ # Closure
2671
2672 main::setup_package();
2673
2674 my %class_defaults;
2675 # The defaults structure for the classes
2676 main::set_access('class_defaults', \%class_defaults);
2677
2678 my %other_default;
2679 # The default that applies to everything left over.
2680 main::set_access('other_default', \%other_default, 'r');
2681
2682
2683 sub new {
2684 # The constructor is called with default => eval pairs, terminated by
2685 # the left-over default. e.g.
2686 # Multi_Default->new(
2687 # 'T' => '$gc->table("Mn") + $gc->table("Cf") - 0x200C
2688 # - 0x200D',
2689 # 'R' => 'some other expression that evaluates to code points',
2690 # .
2691 # .
2692 # .
2693 # 'U'));
2694
2695 my $class = shift;
2696
2697 my $self = bless \do{my $anonymous_scalar}, $class;
ffe43484 2698 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2699
2700 while (@_ > 1) {
2701 my $default = shift;
2702 my $eval = shift;
2703 $class_defaults{$addr}->{$default} = $eval;
2704 }
2705
2706 $other_default{$addr} = shift;
2707
2708 return $self;
2709 }
2710
2711 sub get_next_defaults {
2712 # Iterates and returns the next class of defaults.
2713 my $self = shift;
2714 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2715
ffe43484 2716 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2717
2718 return each %{$class_defaults{$addr}};
2719 }
2720}
2721
2722package Alias;
2723
2724# An alias is one of the names that a table goes by. This class defines them
2725# including some attributes. Everything is currently setup in the
2726# constructor.
2727
2728
2729{ # Closure
2730
2731 main::setup_package();
2732
2733 my %name;
2734 main::set_access('name', \%name, 'r');
2735
2736 my %loose_match;
c12f2655 2737 # Should this name match loosely or not.
99870f4d
KW
2738 main::set_access('loose_match', \%loose_match, 'r');
2739
33e96e72
KW
2740 my %make_re_pod_entry;
2741 # Some aliases should not get their own entries in the re section of the
2742 # pod, because they are covered by a wild-card, and some we want to
2743 # discourage use of. Binary
f82fe4ba 2744 main::set_access('make_re_pod_entry', \%make_re_pod_entry, 'r', 's');
99870f4d 2745
fd1e3e84
KW
2746 my %ucd;
2747 # Is this documented to be accessible via Unicode::UCD
2748 main::set_access('ucd', \%ucd, 'r', 's');
2749
99870f4d
KW
2750 my %status;
2751 # Aliases have a status, like deprecated, or even suppressed (which means
2752 # they don't appear in documentation). Enum
2753 main::set_access('status', \%status, 'r');
2754
0eac1e20 2755 my %ok_as_filename;
99870f4d
KW
2756 # Similarly, some aliases should not be considered as usable ones for
2757 # external use, such as file names, or we don't want documentation to
2758 # recommend them. Boolean
0eac1e20 2759 main::set_access('ok_as_filename', \%ok_as_filename, 'r');
99870f4d
KW
2760
2761 sub new {
2762 my $class = shift;
2763
2764 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2765 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2766
2767 $name{$addr} = shift;
2768 $loose_match{$addr} = shift;
33e96e72 2769 $make_re_pod_entry{$addr} = shift;
0eac1e20 2770 $ok_as_filename{$addr} = shift;
99870f4d 2771 $status{$addr} = shift;
fd1e3e84 2772 $ucd{$addr} = shift;
99870f4d
KW
2773
2774 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2775
2776 # Null names are never ok externally
0eac1e20 2777 $ok_as_filename{$addr} = 0 if $name{$addr} eq "";
99870f4d
KW
2778
2779 return $self;
2780 }
2781}
2782
2783package Range;
2784
2785# A range is the basic unit for storing code points, and is described in the
2786# comments at the beginning of the program. Each range has a starting code
2787# point; an ending code point (not less than the starting one); a value
2788# that applies to every code point in between the two end-points, inclusive;
2789# and an enum type that applies to the value. The type is for the user's
2790# convenience, and has no meaning here, except that a non-zero type is
2791# considered to not obey the normal Unicode rules for having standard forms.
2792#
2793# The same structure is used for both map and match tables, even though in the
2794# latter, the value (and hence type) is irrelevant and could be used as a
2795# comment. In map tables, the value is what all the code points in the range
2796# map to. Type 0 values have the standardized version of the value stored as
2797# well, so as to not have to recalculate it a lot.
2798
2799sub trace { return main::trace(@_); }
2800
2801{ # Closure
2802
2803 main::setup_package();
2804
2805 my %start;
2806 main::set_access('start', \%start, 'r', 's');
2807
2808 my %end;
2809 main::set_access('end', \%end, 'r', 's');
2810
2811 my %value;
2812 main::set_access('value', \%value, 'r');
2813
2814 my %type;
2815 main::set_access('type', \%type, 'r');
2816
2817 my %standard_form;
2818 # The value in internal standard form. Defined only if the type is 0.
2819 main::set_access('standard_form', \%standard_form);
2820
2821 # Note that if these fields change, the dump() method should as well
2822
2823 sub new {
2824 return Carp::carp_too_few_args(\@_, 3) if main::DEBUG && @_ < 3;
2825 my $class = shift;
2826
2827 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2828 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2829
2830 $start{$addr} = shift;
2831 $end{$addr} = shift;
2832
2833 my %args = @_;
2834
2835 my $value = delete $args{'Value'}; # Can be 0
2836 $value = "" unless defined $value;
2837 $value{$addr} = $value;
2838
2839 $type{$addr} = delete $args{'Type'} || 0;
2840
2841 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2842
2843 if (! $type{$addr}) {
2844 $standard_form{$addr} = main::standardize($value);
2845 }
2846
2847 return $self;
2848 }
2849
2850 use overload
2851 fallback => 0,
2852 qw("") => "_operator_stringify",
2853 "." => \&main::_operator_dot,
2854 ;
2855
2856 sub _operator_stringify {
2857 my $self = shift;
ffe43484 2858 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2859
2860 # Output it like '0041..0065 (value)'
2861 my $return = sprintf("%04X", $start{$addr})
2862 . '..'
2863 . sprintf("%04X", $end{$addr});
2864 my $value = $value{$addr};
2865 my $type = $type{$addr};
2866 $return .= ' (';
2867 $return .= "$value";
2868 $return .= ", Type=$type" if $type != 0;
2869 $return .= ')';
2870
2871 return $return;
2872 }
2873
2874 sub standard_form {
2875 # The standard form is the value itself if the standard form is
2876 # undefined (that is if the value is special)
2877
2878 my $self = shift;
2879 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2880
ffe43484 2881 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2882
2883 return $standard_form{$addr} if defined $standard_form{$addr};
2884 return $value{$addr};
2885 }
2886
2887 sub dump {
2888 # Human, not machine readable. For machine readable, comment out this
2889 # entire routine and let the standard one take effect.
2890 my $self = shift;
2891 my $indent = shift;
2892 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2893
ffe43484 2894 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2895
2896 my $return = $indent
2897 . sprintf("%04X", $start{$addr})
2898 . '..'
2899 . sprintf("%04X", $end{$addr})
2900 . " '$value{$addr}';";
2901 if (! defined $standard_form{$addr}) {
2902 $return .= "(type=$type{$addr})";
2903 }
2904 elsif ($standard_form{$addr} ne $value{$addr}) {
2905 $return .= "(standard '$standard_form{$addr}')";
2906 }
2907 return $return;
2908 }
2909} # End closure
2910
2911package _Range_List_Base;
2912
2913# Base class for range lists. A range list is simply an ordered list of
2914# ranges, so that the ranges with the lowest starting numbers are first in it.
2915#
2916# When a new range is added that is adjacent to an existing range that has the
2917# same value and type, it merges with it to form a larger range.
2918#
2919# Ranges generally do not overlap, except that there can be multiple entries
2920# of single code point ranges. This is because of NameAliases.txt.
2921#
2922# In this program, there is a standard value such that if two different
2923# values, have the same standard value, they are considered equivalent. This
2924# value was chosen so that it gives correct results on Unicode data
2925
2926# There are a number of methods to manipulate range lists, and some operators
2927# are overloaded to handle them.
2928
99870f4d
KW
2929sub trace { return main::trace(@_); }
2930
2931{ # Closure
2932
2933 our $addr;
2934
2935 main::setup_package();
2936
2937 my %ranges;
2938 # The list of ranges
2939 main::set_access('ranges', \%ranges, 'readable_array');
2940
2941 my %max;
2942 # The highest code point in the list. This was originally a method, but
2943 # actual measurements said it was used a lot.
2944 main::set_access('max', \%max, 'r');
2945
2946 my %each_range_iterator;
2947 # Iterator position for each_range()
2948 main::set_access('each_range_iterator', \%each_range_iterator);
2949
2950 my %owner_name_of;
2951 # Name of parent this is attached to, if any. Solely for better error
2952 # messages.
2953 main::set_access('owner_name_of', \%owner_name_of, 'p_r');
2954
2955 my %_search_ranges_cache;
2956 # A cache of the previous result from _search_ranges(), for better
2957 # performance
2958 main::set_access('_search_ranges_cache', \%_search_ranges_cache);
2959
2960 sub new {
2961 my $class = shift;
2962 my %args = @_;
2963
2964 # Optional initialization data for the range list.
2965 my $initialize = delete $args{'Initialize'};
2966
2967 my $self;
2968
2969 # Use _union() to initialize. _union() returns an object of this
2970 # class, which means that it will call this constructor recursively.
2971 # But it won't have this $initialize parameter so that it won't
2972 # infinitely loop on this.
2973 return _union($class, $initialize, %args) if defined $initialize;
2974
2975 $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2976 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2977
2978 # Optional parent object, only for debug info.
2979 $owner_name_of{$addr} = delete $args{'Owner'};
2980 $owner_name_of{$addr} = "" if ! defined $owner_name_of{$addr};
2981
2982 # Stringify, in case it is an object.
2983 $owner_name_of{$addr} = "$owner_name_of{$addr}";
2984
2985 # This is used only for error messages, and so a colon is added
2986 $owner_name_of{$addr} .= ": " if $owner_name_of{$addr} ne "";
2987
2988 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2989
2990 # Max is initialized to a negative value that isn't adjacent to 0,
2991 # for simpler tests
2992 $max{$addr} = -2;
2993
2994 $_search_ranges_cache{$addr} = 0;
2995 $ranges{$addr} = [];
2996
2997 return $self;
2998 }
2999
3000 use overload
3001 fallback => 0,
3002 qw("") => "_operator_stringify",
3003 "." => \&main::_operator_dot,
3004 ;
3005
3006 sub _operator_stringify {
3007 my $self = shift;
ffe43484 3008 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3009
3010 return "Range_List attached to '$owner_name_of{$addr}'"
3011 if $owner_name_of{$addr};
3012 return "anonymous Range_List " . \$self;
3013 }
3014
3015 sub _union {
3016 # Returns the union of the input code points. It can be called as
3017 # either a constructor or a method. If called as a method, the result
3018 # will be a new() instance of the calling object, containing the union
3019 # of that object with the other parameter's code points; if called as
d59563d0 3020 # a constructor, the first parameter gives the class that the new object
99870f4d
KW
3021 # should be, and the second parameter gives the code points to go into
3022 # it.
3023 # In either case, there are two parameters looked at by this routine;
3024 # any additional parameters are passed to the new() constructor.
3025 #
3026 # The code points can come in the form of some object that contains
3027 # ranges, and has a conventionally named method to access them; or
3028 # they can be an array of individual code points (as integers); or
3029 # just a single code point.
3030 #
3031 # If they are ranges, this routine doesn't make any effort to preserve
3198cc57
KW
3032 # the range values and types of one input over the other. Therefore
3033 # this base class should not allow _union to be called from other than
99870f4d
KW
3034 # initialization code, so as to prevent two tables from being added
3035 # together where the range values matter. The general form of this
3036 # routine therefore belongs in a derived class, but it was moved here
3037 # to avoid duplication of code. The failure to overload this in this
3038 # class keeps it safe.
3198cc57
KW
3039 #
3040 # It does make the effort during initialization to accept tables with
3041 # multiple values for the same code point, and to preserve the order
3042 # of these. If there is only one input range or range set, it doesn't
3043 # sort (as it should already be sorted to the desired order), and will
3044 # accept multiple values per code point. Otherwise it will merge
3045 # multiple values into a single one.
99870f4d
KW
3046
3047 my $self;
3048 my @args; # Arguments to pass to the constructor
3049
3050 my $class = shift;
3051
3052 # If a method call, will start the union with the object itself, and
3053 # the class of the new object will be the same as self.
3054 if (ref $class) {
3055 $self = $class;
3056 $class = ref $self;
3057 push @args, $self;
3058 }
3059
3060 # Add the other required parameter.
3061 push @args, shift;
3062 # Rest of parameters are passed on to the constructor
3063
3064 # Accumulate all records from both lists.
3065 my @records;
3198cc57 3066 my $input_count = 0;
99870f4d
KW
3067 for my $arg (@args) {
3068 #local $to_trace = 0 if main::DEBUG;
3069 trace "argument = $arg" if main::DEBUG && $to_trace;
3070 if (! defined $arg) {
3071 my $message = "";
3072 if (defined $self) {
f998e60c 3073 no overloading;
051df77b 3074 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
3075 }
3076 Carp::my_carp_bug($message .= "Undefined argument to _union. No union done.");
3077 return;
3078 }
3198cc57 3079
99870f4d
KW
3080 $arg = [ $arg ] if ! ref $arg;
3081 my $type = ref $arg;
3082 if ($type eq 'ARRAY') {
3083 foreach my $element (@$arg) {
3084 push @records, Range->new($element, $element);
3198cc57 3085 $input_count++;
99870f4d
KW
3086 }
3087 }
3088 elsif ($arg->isa('Range')) {
3089 push @records, $arg;
3198cc57 3090 $input_count++;
99870f4d
KW
3091 }
3092 elsif ($arg->can('ranges')) {
3093 push @records, $arg->ranges;
3198cc57 3094 $input_count++;
99870f4d
KW
3095 }
3096 else {
3097 my $message = "";
3098 if (defined $self) {
f998e60c 3099 no overloading;
051df77b 3100 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
3101 }
3102 Carp::my_carp_bug($message . "Cannot take the union of a $type. No union done.");
3103 return;
3104 }
3105 }
3106
3107 # Sort with the range containing the lowest ordinal first, but if
3108 # two ranges start at the same code point, sort with the bigger range
3109 # of the two first, because it takes fewer cycles.
3198cc57
KW
3110 if ($input_count > 1) {
3111 @records = sort { ($a->start <=> $b->start)
99870f4d
KW
3112 or
3113 # if b is shorter than a, b->end will be
3114 # less than a->end, and we want to select
3115 # a, so want to return -1
3116 ($b->end <=> $a->end)
3117 } @records;
3198cc57 3118 }
99870f4d
KW
3119
3120 my $new = $class->new(@_);
3121
3122 # Fold in records so long as they add new information.
3123 for my $set (@records) {
3124 my $start = $set->start;
3125 my $end = $set->end;
d59563d0 3126 my $value = $set->value;
3198cc57 3127 my $type = $set->type;
99870f4d 3128 if ($start > $new->max) {
3198cc57 3129 $new->_add_delete('+', $start, $end, $value, Type => $type);
99870f4d
KW
3130 }
3131 elsif ($end > $new->max) {
3198cc57
KW
3132 $new->_add_delete('+', $new->max +1, $end, $value,
3133 Type => $type);
3134 }
3135 elsif ($input_count == 1) {
3136 # Here, overlaps existing range, but is from a single input,
3137 # so preserve the multiple values from that input.
3138 $new->_add_delete('+', $start, $end, $value, Type => $type,
3139 Replace => $MULTIPLE_AFTER);
99870f4d
KW
3140 }
3141 }
3142
3143 return $new;
3144 }
3145
3146 sub range_count { # Return the number of ranges in the range list
3147 my $self = shift;
3148 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3149
f998e60c 3150 no overloading;
051df77b 3151 return scalar @{$ranges{pack 'J', $self}};
99870f4d
KW
3152 }
3153
3154 sub min {
3155 # Returns the minimum code point currently in the range list, or if
3156 # the range list is empty, 2 beyond the max possible. This is a
3157 # method because used so rarely, that not worth saving between calls,
3158 # and having to worry about changing it as ranges are added and
3159 # deleted.
3160
3161 my $self = shift;
3162 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3163
ffe43484 3164 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3165
3166 # If the range list is empty, return a large value that isn't adjacent
3167 # to any that could be in the range list, for simpler tests
6189eadc 3168 return $MAX_UNICODE_CODEPOINT + 2 unless scalar @{$ranges{$addr}};
99870f4d
KW
3169 return $ranges{$addr}->[0]->start;
3170 }
3171
3172 sub contains {
3173 # Boolean: Is argument in the range list? If so returns $i such that:
3174 # range[$i]->end < $codepoint <= range[$i+1]->end
3175 # which is one beyond what you want; this is so that the 0th range
3176 # doesn't return false
3177 my $self = shift;
3178 my $codepoint = shift;
3179 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3180
99870f4d
KW
3181 my $i = $self->_search_ranges($codepoint);
3182 return 0 unless defined $i;
3183
3184 # The search returns $i, such that
3185 # range[$i-1]->end < $codepoint <= range[$i]->end
3186 # So is in the table if and only iff it is at least the start position
3187 # of range $i.
f998e60c 3188 no overloading;
051df77b 3189 return 0 if $ranges{pack 'J', $self}->[$i]->start > $codepoint;
99870f4d
KW
3190 return $i + 1;
3191 }
3192
2f7a8815
KW
3193 sub containing_range {
3194 # Returns the range object that contains the code point, undef if none
3195
3196 my $self = shift;
3197 my $codepoint = shift;
3198 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3199
3200 my $i = $self->contains($codepoint);
3201 return unless $i;
3202
3203 # contains() returns 1 beyond where we should look
3204 no overloading;
3205 return $ranges{pack 'J', $self}->[$i-1];
3206 }
3207
99870f4d
KW
3208 sub value_of {
3209 # Returns the value associated with the code point, undef if none
3210
3211 my $self = shift;
3212 my $codepoint = shift;
3213 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3214
d69c231b
KW
3215 my $range = $self->containing_range($codepoint);
3216 return unless defined $range;
99870f4d 3217
d69c231b 3218 return $range->value;
99870f4d
KW
3219 }
3220
0a9dbafc
KW
3221 sub type_of {
3222 # Returns the type of the range containing the code point, undef if
3223 # the code point is not in the table
3224
3225 my $self = shift;
3226 my $codepoint = shift;
3227 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3228
3229 my $range = $self->containing_range($codepoint);
3230 return unless defined $range;
3231
3232 return $range->type;
3233 }
3234
99870f4d
KW
3235 sub _search_ranges {
3236 # Find the range in the list which contains a code point, or where it
3237 # should go if were to add it. That is, it returns $i, such that:
3238 # range[$i-1]->end < $codepoint <= range[$i]->end
3239 # Returns undef if no such $i is possible (e.g. at end of table), or
3240 # if there is an error.
3241
3242 my $self = shift;
3243 my $code_point = shift;
3244 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3245
ffe43484 3246 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3247
3248 return if $code_point > $max{$addr};
3249 my $r = $ranges{$addr}; # The current list of ranges
3250 my $range_list_size = scalar @$r;
3251 my $i;
3252
3253 use integer; # want integer division
3254
3255 # Use the cached result as the starting guess for this one, because,
3256 # an experiment on 5.1 showed that 90% of the time the cache was the
3257 # same as the result on the next call (and 7% it was one less).
3258 $i = $_search_ranges_cache{$addr};
3259 $i = 0 if $i >= $range_list_size; # Reset if no longer valid (prob.
3260 # from an intervening deletion
3261 #local $to_trace = 1 if main::DEBUG;
3262 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);
3263 return $i if $code_point <= $r->[$i]->end
3264 && ($i == 0 || $r->[$i-1]->end < $code_point);
3265
3266 # Here the cache doesn't yield the correct $i. Try adding 1.
3267 if ($i < $range_list_size - 1
3268 && $r->[$i]->end < $code_point &&
3269 $code_point <= $r->[$i+1]->end)
3270 {
3271 $i++;
3272 trace "next \$i is correct: $i" if main::DEBUG && $to_trace;
3273 $_search_ranges_cache{$addr} = $i;
3274 return $i;
3275 }
3276
3277 # Here, adding 1 also didn't work. We do a binary search to
3278 # find the correct position, starting with current $i
3279 my $lower = 0;
3280 my $upper = $range_list_size - 1;
3281 while (1) {
3282 trace "top of loop i=$i:", sprintf("%04X", $r->[$lower]->start), "[$lower] .. ", sprintf("%04X", $r->[$i]->start), "[$i] .. ", sprintf("%04X", $r->[$upper]->start), "[$upper]" if main::DEBUG && $to_trace;
3283
3284 if ($code_point <= $r->[$i]->end) {
3285
3286 # Here we have met the upper constraint. We can quit if we
3287 # also meet the lower one.
3288 last if $i == 0 || $r->[$i-1]->end < $code_point;
3289
3290 $upper = $i; # Still too high.
3291
3292 }
3293 else {
3294
3295 # Here, $r[$i]->end < $code_point, so look higher up.
3296 $lower = $i;
3297 }
3298
3299 # Split search domain in half to try again.
3300 my $temp = ($upper + $lower) / 2;
3301
3302 # No point in continuing unless $i changes for next time
3303 # in the loop.
3304 if ($temp == $i) {
3305
3306 # We can't reach the highest element because of the averaging.
3307 # So if one below the upper edge, force it there and try one
3308 # more time.
3309 if ($i == $range_list_size - 2) {
3310
3311 trace "Forcing to upper edge" if main::DEBUG && $to_trace;
3312 $i = $range_list_size - 1;
3313
3314 # Change $lower as well so if fails next time through,
3315 # taking the average will yield the same $i, and we will
3316 # quit with the error message just below.
3317 $lower = $i;
3318 next;
3319 }
3320 Carp::my_carp_bug("$owner_name_of{$addr}Can't find where the range ought to go. No action taken.");
3321 return;
3322 }
3323 $i = $temp;
3324 } # End of while loop
3325
3326 if (main::DEBUG && $to_trace) {
3327 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i;
3328 trace "i= [ $i ]", $r->[$i];
3329 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < $range_list_size - 1;
3330 }
3331
3332 # Here we have found the offset. Cache it as a starting point for the
3333 # next call.
3334 $_search_ranges_cache{$addr} = $i;
3335 return $i;
3336 }
3337
3338 sub _add_delete {
3339 # Add, replace or delete ranges to or from a list. The $type
3340 # parameter gives which:
3341 # '+' => insert or replace a range, returning a list of any changed
3342 # ranges.
3343 # '-' => delete a range, returning a list of any deleted ranges.
3344 #
3345 # The next three parameters give respectively the start, end, and
3346 # value associated with the range. 'value' should be null unless the
3347 # operation is '+';
3348 #
3349 # The range list is kept sorted so that the range with the lowest
3350 # starting position is first in the list, and generally, adjacent
c1739a4a 3351 # ranges with the same values are merged into a single larger one (see
99870f4d
KW
3352 # exceptions below).
3353 #
c1739a4a 3354 # There are more parameters; all are key => value pairs:
99870f4d
KW
3355 # Type gives the type of the value. It is only valid for '+'.
3356 # All ranges have types; if this parameter is omitted, 0 is
3357 # assumed. Ranges with type 0 are assumed to obey the
3358 # Unicode rules for casing, etc; ranges with other types are
3359 # not. Otherwise, the type is arbitrary, for the caller's
3360 # convenience, and looked at only by this routine to keep
3361 # adjacent ranges of different types from being merged into
3362 # a single larger range, and when Replace =>
3363 # $IF_NOT_EQUIVALENT is specified (see just below).
3364 # Replace determines what to do if the range list already contains
3365 # ranges which coincide with all or portions of the input
3366 # range. It is only valid for '+':
3367 # => $NO means that the new value is not to replace
3368 # any existing ones, but any empty gaps of the
3369 # range list coinciding with the input range
3370 # will be filled in with the new value.
3371 # => $UNCONDITIONALLY means to replace the existing values with
3372 # this one unconditionally. However, if the
3373 # new and old values are identical, the
3374 # replacement is skipped to save cycles
3375 # => $IF_NOT_EQUIVALENT means to replace the existing values
d59563d0 3376 # (the default) with this one if they are not equivalent.
99870f4d 3377 # Ranges are equivalent if their types are the
c1739a4a 3378 # same, and they are the same string; or if
99870f4d
KW
3379 # both are type 0 ranges, if their Unicode
3380 # standard forms are identical. In this last
3381 # case, the routine chooses the more "modern"
3382 # one to use. This is because some of the
3383 # older files are formatted with values that
3384 # are, for example, ALL CAPs, whereas the
3385 # derived files have a more modern style,
3386 # which looks better. By looking for this
3387 # style when the pre-existing and replacement
3388 # standard forms are the same, we can move to
3389 # the modern style
9470941f 3390 # => $MULTIPLE_BEFORE means that if this range duplicates an
99870f4d
KW
3391 # existing one, but has a different value,
3392 # don't replace the existing one, but insert
3393 # this, one so that the same range can occur
53d84487
KW
3394 # multiple times. They are stored LIFO, so
3395 # that the final one inserted is the first one
3396 # returned in an ordered search of the table.
7f4b1e25
KW
3397 # => $MULTIPLE_AFTER is like $MULTIPLE_BEFORE, but is stored
3398 # FIFO, so that this one is inserted after all
3399 # others that currently exist.
99870f4d
KW
3400 # => anything else is the same as => $IF_NOT_EQUIVALENT
3401 #
c1739a4a
KW
3402 # "same value" means identical for non-type-0 ranges, and it means
3403 # having the same standard forms for type-0 ranges.
99870f4d
KW
3404
3405 return Carp::carp_too_few_args(\@_, 5) if main::DEBUG && @_ < 5;
3406
3407 my $self = shift;
3408 my $operation = shift; # '+' for add/replace; '-' for delete;
3409 my $start = shift;
3410 my $end = shift;
3411 my $value = shift;
3412
3413 my %args = @_;
3414
3415 $value = "" if not defined $value; # warning: $value can be "0"
3416
3417 my $replace = delete $args{'Replace'};
3418 $replace = $IF_NOT_EQUIVALENT unless defined $replace;
3419
3420 my $type = delete $args{'Type'};
3421 $type = 0 unless defined $type;
3422
3423 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3424
ffe43484 3425 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3426
3427 if ($operation ne '+' && $operation ne '-') {
3428 Carp::my_carp_bug("$owner_name_of{$addr}First parameter to _add_delete must be '+' or '-'. No action taken.");
3429 return;
3430 }
3431 unless (defined $start && defined $end) {
3432 Carp::my_carp_bug("$owner_name_of{$addr}Undefined start and/or end to _add_delete. No action taken.");
3433 return;
3434 }
3435 unless ($end >= $start) {
3436 Carp::my_carp_bug("$owner_name_of{$addr}End of range (" . sprintf("%04X", $end) . ") must not be before start (" . sprintf("%04X", $start) . "). No action taken.");
3437 return;
3438 }
3439 #local $to_trace = 1 if main::DEBUG;
3440
3441 if ($operation eq '-') {
3442 if ($replace != $IF_NOT_EQUIVALENT) {
3443 Carp::my_carp_bug("$owner_name_of{$addr}Replace => \$IF_NOT_EQUIVALENT is required when deleting a range from a range list. Assuming Replace => \$IF_NOT_EQUIVALENT.");
3444 $replace = $IF_NOT_EQUIVALENT;
3445 }
3446 if ($type) {
3447 Carp::my_carp_bug("$owner_name_of{$addr}Type => 0 is required when deleting a range from a range list. Assuming Type => 0.");
3448 $type = 0;
3449 }
3450 if ($value ne "") {
3451 Carp::my_carp_bug("$owner_name_of{$addr}Value => \"\" is required when deleting a range from a range list. Assuming Value => \"\".");
3452 $value = "";
3453 }
3454 }
3455
3456 my $r = $ranges{$addr}; # The current list of ranges
3457 my $range_list_size = scalar @$r; # And its size
3458 my $max = $max{$addr}; # The current high code point in
3459 # the list of ranges
3460
3461 # Do a special case requiring fewer machine cycles when the new range
3462 # starts after the current highest point. The Unicode input data is
3463 # structured so this is common.
3464 if ($start > $max) {
3465
3466 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) type=$type" if main::DEBUG && $to_trace;
3467 return if $operation eq '-'; # Deleting a non-existing range is a
3468 # no-op
3469
3470 # If the new range doesn't logically extend the current final one
3471 # in the range list, create a new range at the end of the range
3472 # list. (max cleverly is initialized to a negative number not
3473 # adjacent to 0 if the range list is empty, so even adding a range
3474 # to an empty range list starting at 0 will have this 'if'
3475 # succeed.)
3476 if ($start > $max + 1 # non-adjacent means can't extend.
3477 || @{$r}[-1]->value ne $value # values differ, can't extend.
3478 || @{$r}[-1]->type != $type # types differ, can't extend.
3479 ) {
3480 push @$r, Range->new($start, $end,
3481 Value => $value,
3482 Type => $type);
3483 }
3484 else {
3485
3486 # Here, the new range starts just after the current highest in
3487 # the range list, and they have the same type and value.
3488 # Extend the current range to incorporate the new one.
3489 @{$r}[-1]->set_end($end);
3490 }
3491
3492 # This becomes the new maximum.
3493 $max{$addr} = $end;
3494
3495 return;
3496 }
3497 #local $to_trace = 0 if main::DEBUG;
3498
3499 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) replace=$replace" if main::DEBUG && $to_trace;
3500
3501 # Here, the input range isn't after the whole rest of the range list.
3502 # Most likely 'splice' will be needed. The rest of the routine finds
3503 # the needed splice parameters, and if necessary, does the splice.
3504 # First, find the offset parameter needed by the splice function for
3505 # the input range. Note that the input range may span multiple
3506 # existing ones, but we'll worry about that later. For now, just find
3507 # the beginning. If the input range is to be inserted starting in a
3508 # position not currently in the range list, it must (obviously) come
3509 # just after the range below it, and just before the range above it.
3510 # Slightly less obviously, it will occupy the position currently
3511 # occupied by the range that is to come after it. More formally, we
3512 # are looking for the position, $i, in the array of ranges, such that:
3513 #
3514 # r[$i-1]->start <= r[$i-1]->end < $start < r[$i]->start <= r[$i]->end
3515 #
3516 # (The ordered relationships within existing ranges are also shown in
3517 # the equation above). However, if the start of the input range is
3518 # within an existing range, the splice offset should point to that
3519 # existing range's position in the list; that is $i satisfies a
3520 # somewhat different equation, namely:
3521 #
3522 #r[$i-1]->start <= r[$i-1]->end < r[$i]->start <= $start <= r[$i]->end
3523 #
3524 # More briefly, $start can come before or after r[$i]->start, and at
3525 # this point, we don't know which it will be. However, these
3526 # two equations share these constraints:
3527 #
3528 # r[$i-1]->end < $start <= r[$i]->end
3529 #
3530 # And that is good enough to find $i.
3531
3532 my $i = $self->_search_ranges($start);
3533 if (! defined $i) {
3534 Carp::my_carp_bug("Searching $self for range beginning with $start unexpectedly returned undefined. Operation '$operation' not performed");
3535 return;
3536 }
3537
3538 # The search function returns $i such that:
3539 #
3540 # r[$i-1]->end < $start <= r[$i]->end
3541 #
3542 # That means that $i points to the first range in the range list
3543 # that could possibly be affected by this operation. We still don't
3544 # know if the start of the input range is within r[$i], or if it
3545 # points to empty space between r[$i-1] and r[$i].
3546 trace "[$i] is the beginning splice point. Existing range there is ", $r->[$i] if main::DEBUG && $to_trace;
3547
3548 # Special case the insertion of data that is not to replace any
3549 # existing data.
3550 if ($replace == $NO) { # If $NO, has to be operation '+'
3551 #local $to_trace = 1 if main::DEBUG;
3552 trace "Doesn't replace" if main::DEBUG && $to_trace;
3553
3554 # Here, the new range is to take effect only on those code points
3555 # that aren't already in an existing range. This can be done by
3556 # looking through the existing range list and finding the gaps in
3557 # the ranges that this new range affects, and then calling this
3558 # function recursively on each of those gaps, leaving untouched
3559 # anything already in the list. Gather up a list of the changed
3560 # gaps first so that changes to the internal state as new ranges
3561 # are added won't be a problem.
3562 my @gap_list;
3563
3564 # First, if the starting point of the input range is outside an
3565 # existing one, there is a gap from there to the beginning of the
3566 # existing range -- add a span to fill the part that this new
3567 # range occupies
3568 if ($start < $r->[$i]->start) {
3569 push @gap_list, Range->new($start,
3570 main::min($end,
3571 $r->[$i]->start - 1),
3572 Type => $type);
3573 trace "gap before $r->[$i] [$i], will add", $gap_list[-1] if main::DEBUG && $to_trace;
3574 }
3575
3576 # Then look through the range list for other gaps until we reach
3577 # the highest range affected by the input one.
3578 my $j;
3579 for ($j = $i+1; $j < $range_list_size; $j++) {
3580 trace "j=[$j]", $r->[$j] if main::DEBUG && $to_trace;
3581 last if $end < $r->[$j]->start;
3582
3583 # If there is a gap between when this range starts and the
3584 # previous one ends, add a span to fill it. Note that just
3585 # because there are two ranges doesn't mean there is a
3586 # non-zero gap between them. It could be that they have
3587 # different values or types
3588 if ($r->[$j-1]->end + 1 != $r->[$j]->start) {
3589 push @gap_list,
3590 Range->new($r->[$j-1]->end + 1,
3591 $r->[$j]->start - 1,
3592 Type => $type);
3593 trace "gap between $r->[$j-1] and $r->[$j] [$j], will add: $gap_list[-1]" if main::DEBUG && $to_trace;
3594 }
3595 }
3596
3597 # Here, we have either found an existing range in the range list,
3598 # beyond the area affected by the input one, or we fell off the
3599 # end of the loop because the input range affects the whole rest
3600 # of the range list. In either case, $j is 1 higher than the
3601 # highest affected range. If $j == $i, it means that there are no
3602 # affected ranges, that the entire insertion is in the gap between
3603 # r[$i-1], and r[$i], which we already have taken care of before
3604 # the loop.
3605 # On the other hand, if there are affected ranges, it might be
3606 # that there is a gap that needs filling after the final such
3607 # range to the end of the input range
3608 if ($r->[$j-1]->end < $end) {
3609 push @gap_list, Range->new(main::max($start,
3610 $r->[$j-1]->end + 1),
3611 $end,
3612 Type => $type);
3613 trace "gap after $r->[$j-1], will add $gap_list[-1]" if main::DEBUG && $to_trace;
3614 }
3615
3616 # Call recursively to fill in all the gaps.
3617 foreach my $gap (@gap_list) {
3618 $self->_add_delete($operation,
3619 $gap->start,
3620 $gap->end,
3621 $value,
3622 Type => $type);
3623 }
3624
3625 return;
3626 }
3627
53d84487
KW
3628 # Here, we have taken care of the case where $replace is $NO.
3629 # Remember that here, r[$i-1]->end < $start <= r[$i]->end
3630 # If inserting a multiple record, this is where it goes, before the
7f4b1e25
KW
3631 # first (if any) existing one if inserting LIFO. (If this is to go
3632 # afterwards, FIFO, we below move the pointer to there.) These imply
3633 # an insertion, and no change to any existing ranges. Note that $i
3634 # can be -1 if this new range doesn't actually duplicate any existing,
3635 # and comes at the beginning of the list.
3636 if ($replace == $MULTIPLE_BEFORE || $replace == $MULTIPLE_AFTER) {
53d84487
KW
3637
3638 if ($start != $end) {
3639 Carp::my_carp_bug("$owner_name_of{$addr}Can't cope with adding a multiple record when the range ($start..$end) contains more than one code point. No action taken.");
3640 return;
3641 }
3642
19155fcc 3643 # If the new code point is within a current range ...
53d84487 3644 if ($end >= $r->[$i]->start) {
19155fcc
KW
3645
3646 # Don't add an exact duplicate, as it isn't really a multiple
1f6798c4
KW
3647 my $existing_value = $r->[$i]->value;
3648 my $existing_type = $r->[$i]->type;
3649 return if $value eq $existing_value && $type eq $existing_type;
3650
3651 # If the multiple value is part of an existing range, we want
3652 # to split up that range, so that only the single code point
3653 # is affected. To do this, we first call ourselves
3654 # recursively to delete that code point from the table, having
3655 # preserved its current data above. Then we call ourselves
3656 # recursively again to add the new multiple, which we know by
3657 # the test just above is different than the current code
3658 # point's value, so it will become a range containing a single
3659 # code point: just itself. Finally, we add back in the
3660 # pre-existing code point, which will again be a single code
3661 # point range. Because 'i' likely will have changed as a
3662 # result of these operations, we can't just continue on, but
7f4b1e25
KW
3663 # do this operation recursively as well. If we are inserting
3664 # LIFO, the pre-existing code point needs to go after the new
3665 # one, so use MULTIPLE_AFTER; and vice versa.
53d84487 3666 if ($r->[$i]->start != $r->[$i]->end) {
1f6798c4
KW
3667 $self->_add_delete('-', $start, $end, "");
3668 $self->_add_delete('+', $start, $end, $value, Type => $type);
7f4b1e25
KW
3669 return $self->_add_delete('+',
3670 $start, $end,
3671 $existing_value,
3672 Type => $existing_type,
3673 Replace => ($replace == $MULTIPLE_BEFORE)
3674 ? $MULTIPLE_AFTER
3675 : $MULTIPLE_BEFORE);
3676 }
3677 }
3678
3679 # If to place this new record after, move to beyond all existing
1722e378
KW
3680 # ones; but don't add this one if identical to any of them, as it
3681 # isn't really a multiple
7f4b1e25
KW
3682 if ($replace == $MULTIPLE_AFTER) {
3683 while ($i < @$r && $r->[$i]->start == $start) {
1722e378
KW
3684 return if $value eq $r->[$i]->value
3685 && $type eq $r->[$i]->type;
7f4b1e25 3686 $i++;
53d84487 3687 }
53d84487
KW
3688 }
3689
3690 trace "Adding multiple record at $i with $start..$end, $value" if main::DEBUG && $to_trace;
3691 my @return = splice @$r,
3692 $i,
3693 0,
3694 Range->new($start,
3695 $end,
3696 Value => $value,
3697 Type => $type);
3698 if (main::DEBUG && $to_trace) {
3699 trace "After splice:";
3700 trace 'i-2=[', $i-2, ']', $r->[$i-2] if $i >= 2;
3701 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i >= 1;
3702 trace "i =[", $i, "]", $r->[$i] if $i >= 0;
3703 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < @$r - 1;
3704 trace 'i+2=[', $i+2, ']', $r->[$i+2] if $i < @$r - 2;
3705 trace 'i+3=[', $i+3, ']', $r->[$i+3] if $i < @$r - 3;
3706 }
3707 return @return;
3708 }
3709
7f4b1e25
KW
3710 # Here, we have taken care of $NO and $MULTIPLE_foo replaces. This
3711 # leaves delete, insert, and replace either unconditionally or if not
53d84487
KW
3712 # equivalent. $i still points to the first potential affected range.
3713 # Now find the highest range affected, which will determine the length
3714 # parameter to splice. (The input range can span multiple existing
3715 # ones.) If this isn't a deletion, while we are looking through the
3716 # range list, see also if this is a replacement rather than a clean
3717 # insertion; that is if it will change the values of at least one
3718 # existing range. Start off assuming it is an insert, until find it
3719 # isn't.
3720 my $clean_insert = $operation eq '+';
99870f4d
KW
3721 my $j; # This will point to the highest affected range
3722
3723 # For non-zero types, the standard form is the value itself;
3724 my $standard_form = ($type) ? $value : main::standardize($value);
3725
3726 for ($j = $i; $j < $range_list_size; $j++) {
3727 trace "Looking for highest affected range; the one at $j is ", $r->[$j] if main::DEBUG && $to_trace;
3728
3729 # If find a range that it doesn't overlap into, we can stop
3730 # searching
3731 last if $end < $r->[$j]->start;
3732
969a34cc
KW
3733 # Here, overlaps the range at $j. If the values don't match,
3734 # and so far we think this is a clean insertion, it becomes a
3735 # non-clean insertion, i.e., a 'change' or 'replace' instead.
3736 if ($clean_insert) {
99870f4d 3737 if ($r->[$j]->standard_form ne $standard_form) {
969a34cc 3738 $clean_insert = 0;
56343c78
KW
3739 if ($replace == $CROAK) {
3740 main::croak("The range to add "
3741 . sprintf("%04X", $start)
3742 . '-'
3743 . sprintf("%04X", $end)
3744 . " with value '$value' overlaps an existing range $r->[$j]");
3745 }
99870f4d
KW
3746 }
3747 else {
3748
3749 # Here, the two values are essentially the same. If the
3750 # two are actually identical, replacing wouldn't change
3751 # anything so skip it.
3752 my $pre_existing = $r->[$j]->value;
3753 if ($pre_existing ne $value) {
3754
3755 # Here the new and old standardized values are the
3756 # same, but the non-standardized values aren't. If
3757 # replacing unconditionally, then replace
3758 if( $replace == $UNCONDITIONALLY) {
969a34cc 3759 $clean_insert = 0;
99870f4d
KW
3760 }
3761 else {
3762
3763 # Here, are replacing conditionally. Decide to
3764 # replace or not based on which appears to look
3765 # the "nicest". If one is mixed case and the
3766 # other isn't, choose the mixed case one.
3767 my $new_mixed = $value =~ /[A-Z]/
3768 && $value =~ /[a-z]/;
3769 my $old_mixed = $pre_existing =~ /[A-Z]/
3770 && $pre_existing =~ /[a-z]/;
3771
3772 if ($old_mixed != $new_mixed) {
969a34cc 3773 $clean_insert = 0 if $new_mixed;
99870f4d 3774 if (main::DEBUG && $to_trace) {
969a34cc
KW
3775 if ($clean_insert) {
3776 trace "Retaining $pre_existing over $value";
99870f4d
KW
3777 }
3778 else {
969a34cc 3779 trace "Replacing $pre_existing with $value";
99870f4d
KW
3780 }
3781 }
3782 }
3783 else {
3784
3785 # Here casing wasn't different between the two.
3786 # If one has hyphens or underscores and the
3787 # other doesn't, choose the one with the
3788 # punctuation.
3789 my $new_punct = $value =~ /[-_]/;
3790 my $old_punct = $pre_existing =~ /[-_]/;
3791
3792 if ($old_punct != $new_punct) {
969a34cc 3793 $clean_insert = 0 if $new_punct;
99870f4d 3794 if (main::DEBUG && $to_trace) {
969a34cc
KW
3795 if ($clean_insert) {
3796 trace "Retaining $pre_existing over $value";
99870f4d
KW
3797 }
3798 else {
969a34cc 3799 trace "Replacing $pre_existing with $value";
99870f4d
KW
3800 }
3801 }
3802 } # else existing one is just as "good";
3803 # retain it to save cycles.
3804 }
3805 }
3806 }
3807 }
3808 }
3809 } # End of loop looking for highest affected range.
3810
3811 # Here, $j points to one beyond the highest range that this insertion
3812 # affects (hence to beyond the range list if that range is the final
3813 # one in the range list).
3814
3815 # The splice length is all the affected ranges. Get it before
3816 # subtracting, for efficiency, so we don't have to later add 1.
3817 my $length = $j - $i;
3818
3819 $j--; # $j now points to the highest affected range.
3820 trace "Final affected range is $j: $r->[$j]" if main::DEBUG && $to_trace;
3821
7f4b1e25 3822 # Here, have taken care of $NO and $MULTIPLE_foo replaces.
99870f4d
KW
3823 # $j points to the highest affected range. But it can be < $i or even
3824 # -1. These happen only if the insertion is entirely in the gap
3825 # between r[$i-1] and r[$i]. Here's why: j < i means that the j loop
3826 # above exited first time through with $end < $r->[$i]->start. (And