This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
mktables: Rename variable
[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
KW
718 useful for debugging mktables, looking at diffs; but is slow,
719 memory intensive; resulting tables are usable but slow and
720 very large.
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 $v_version ge v6.0.0;
99870f4d
KW
771
772# The lists below are hashes, so the key is the item in the list, and the
773# value is the reason why it is in the list. This makes generation of
774# documentation easier.
775
776my %why_suppressed; # No file generated for these.
777
778# Files aren't generated for empty extraneous properties. This is arguable.
779# Extraneous properties generally come about because a property is no longer
780# used in a newer version of Unicode. If we generated a file without code
781# points, programs that used to work on that property will still execute
782# without errors. It just won't ever match (or will always match, with \P{}).
783# This means that the logic is now likely wrong. I (khw) think its better to
784# find this out by getting an error message. Just move them to the table
785# above to change this behavior
786my %why_suppress_if_empty_warn_if_not = (
787
788 # It is the only property that has ever officially been removed from the
789 # Standard. The database never contained any code points for it.
790 'Special_Case_Condition' => 'Obsolete',
791
792 # Apparently never official, but there were code points in some versions of
793 # old-style PropList.txt
794 'Non_Break' => 'Obsolete',
795);
796
797# These would normally go in the warn table just above, but they were changed
798# a long time before this program was written, so warnings about them are
799# moot.
800if ($v_version gt v3.2.0) {
801 push @tables_that_may_be_empty,
802 'Canonical_Combining_Class=Attached_Below_Left'
803}
804
5f7264c7 805# These are listed in the Property aliases file in 6.0, but Unihan is ignored
99870f4d
KW
806# unless explicitly added.
807if ($v_version ge v5.2.0) {
808 my $unihan = 'Unihan; remove from list if using Unihan';
ea25a9b2 809 foreach my $table (qw (
99870f4d
KW
810 kAccountingNumeric
811 kOtherNumeric
812 kPrimaryNumeric
813 kCompatibilityVariant
814 kIICore
815 kIRG_GSource
816 kIRG_HSource
817 kIRG_JSource
818 kIRG_KPSource
819 kIRG_MSource
820 kIRG_KSource
821 kIRG_TSource
822 kIRG_USource
823 kIRG_VSource
824 kRSUnicode
ea25a9b2 825 ))
99870f4d
KW
826 {
827 $why_suppress_if_empty_warn_if_not{$table} = $unihan;
828 }
ca12659b
NC
829}
830
272501f6
KW
831# Enum values for to_output_map() method in the Map_Table package.
832my $EXTERNAL_MAP = 1;
833my $INTERNAL_MAP = 2;
834
fcf1973c
KW
835# To override computed values for writing the map tables for these properties.
836# The default for enum map tables is to write them out, so that the Unicode
837# .txt files can be removed, but all the data to compute any property value
838# for any code point is available in a more compact form.
839my %global_to_output_map = (
840 # Needed by UCD.pm, but don't want to publicize that it exists, so won't
c12f2655
KW
841 # get stuck supporting it if things change. Since it is a STRING
842 # property, it normally would be listed in the pod, but INTERNAL_MAP
843 # suppresses that.
fcf1973c
KW
844 Unicode_1_Name => $INTERNAL_MAP,
845
846 Present_In => 0, # Suppress, as easily computed from Age
fcf1973c 847 Block => 0, # Suppress, as Blocks.txt is retained.
53d34b6c
KW
848
849 # Suppress, as mapping can be found instead from the
850 # Perl_Decomposition_Mapping file
851 Decomposition_Type => 0,
fcf1973c
KW
852);
853
99870f4d 854# Properties that this program ignores.
230e0c16
KW
855my @unimplemented_properties;
856
857# With this release, it is automatically handled if the Unihan db is
858# downloaded
859push @unimplemented_properties, 'Unicode_Radical_Stroke' if $v_version le v5.2.0;
d73e5302 860
99870f4d
KW
861# There are several types of obsolete properties defined by Unicode. These
862# must be hand-edited for every new Unicode release.
863my %why_deprecated; # Generates a deprecated warning message if used.
864my %why_stabilized; # Documentation only
865my %why_obsolete; # Documentation only
866
867{ # Closure
868 my $simple = 'Perl uses the more complete version of this property';
869 my $unihan = 'Unihan properties are by default not enabled in the Perl core. Instead use CPAN: Unicode::Unihan';
870
871 my $other_properties = 'other properties';
872 my $contributory = "Used by Unicode internally for generating $other_properties and not intended to be used stand-alone";
5d294d41 873 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
874
875 %why_deprecated = (
5f7264c7 876 'Grapheme_Link' => 'Deprecated by Unicode: Duplicates ccc=vr (Canonical_Combining_Class=Virama)',
99870f4d
KW
877 'Jamo_Short_Name' => $contributory,
878 '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',
879 'Other_Alphabetic' => $contributory,
880 'Other_Default_Ignorable_Code_Point' => $contributory,
881 'Other_Grapheme_Extend' => $contributory,
882 'Other_ID_Continue' => $contributory,
883 'Other_ID_Start' => $contributory,
884 'Other_Lowercase' => $contributory,
885 'Other_Math' => $contributory,
886 'Other_Uppercase' => $contributory,
e22aaf5c
KW
887 'Expands_On_NFC' => $why_no_expand,
888 'Expands_On_NFD' => $why_no_expand,
889 'Expands_On_NFKC' => $why_no_expand,
890 'Expands_On_NFKD' => $why_no_expand,
99870f4d
KW
891 );
892
893 %why_suppressed = (
5f7264c7 894 # There is a lib/unicore/Decomposition.pl (used by Normalize.pm) which
99870f4d
KW
895 # contains the same information, but without the algorithmically
896 # determinable Hangul syllables'. This file is not published, so it's
897 # existence is not noted in the comment.
e0b29447 898 'Decomposition_Mapping' => 'Accessible via Unicode::Normalize or Unicode::UCD::prop_invmap()',
99870f4d 899
3111abc0
KW
900 'Indic_Matra_Category' => "Provisional",
901 'Indic_Syllabic_Category' => "Provisional",
902
5f8d1a89
KW
903 # Don't suppress ISO_Comment, as otherwise special handling is needed
904 # to differentiate between it and gc=c, which can be written as 'isc',
905 # which is the same characters as ISO_Comment's short name.
99870f4d 906
fbb93542 907 'Name' => "Accessible via \\N{...} or 'use charnames;' or Unicode::UCD::prop_invmap()",
e0b29447
KW
908
909 'Simple_Case_Folding' => "$simple. Can access this through Unicode::UCD::casefold or Unicode::UCD::prop_invmap()",
910 'Simple_Lowercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
911 'Simple_Titlecase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
912 'Simple_Uppercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
99870f4d 913
5f7264c7 914 FC_NFKC_Closure => 'Supplanted in usage by NFKC_Casefold; otherwise not useful',
99870f4d
KW
915 );
916
1704a0ea
KW
917 foreach my $property (
918
919 # The following are suppressed because they were made contributory
920 # or deprecated by Unicode before Perl ever thought about
921 # supporting them.
922 'Jamo_Short_Name',
923 'Grapheme_Link',
924 'Expands_On_NFC',
925 'Expands_On_NFD',
926 'Expands_On_NFKC',
927 'Expands_On_NFKD',
928
929 # The following are suppressed because they have been marked
930 # as deprecated for a sufficient amount of time
931 'Other_Alphabetic',
932 'Other_Default_Ignorable_Code_Point',
933 'Other_Grapheme_Extend',
934 'Other_ID_Continue',
935 'Other_ID_Start',
936 'Other_Lowercase',
937 'Other_Math',
938 'Other_Uppercase',
e22aaf5c 939 ) {
99870f4d
KW
940 $why_suppressed{$property} = $why_deprecated{$property};
941 }
cf25bb62 942
99870f4d
KW
943 # Customize the message for all the 'Other_' properties
944 foreach my $property (keys %why_deprecated) {
945 next if (my $main_property = $property) !~ s/^Other_//;
946 $why_deprecated{$property} =~ s/$other_properties/the $main_property property (which should be used instead)/;
947 }
948}
949
950if ($v_version ge 4.0.0) {
951 $why_stabilized{'Hyphen'} = 'Use the Line_Break property instead; see www.unicode.org/reports/tr14';
5f7264c7
KW
952 if ($v_version ge 6.0.0) {
953 $why_deprecated{'Hyphen'} = 'Supplanted by Line_Break property values; see www.unicode.org/reports/tr14';
954 }
99870f4d 955}
5f7264c7 956if ($v_version ge 5.2.0 && $v_version lt 6.0.0) {
99870f4d 957 $why_obsolete{'ISO_Comment'} = 'Code points for it have been removed';
5f7264c7 958 if ($v_version ge 6.0.0) {
63f74647 959 $why_deprecated{'ISO_Comment'} = 'No longer needed for Unicode\'s internal chart generation; otherwise not useful, and code points for it have been removed';
5f7264c7 960 }
99870f4d
KW
961}
962
963# Probably obsolete forever
964if ($v_version ge v4.1.0) {
82aed44a
KW
965 $why_suppressed{'Script=Katakana_Or_Hiragana'} = 'Obsolete. All code points previously matched by this have been moved to "Script=Common".';
966}
967if ($v_version ge v6.0.0) {
2b352efd
KW
968 $why_suppressed{'Script=Katakana_Or_Hiragana'} .= ' Consider instead using "Script_Extensions=Katakana" or "Script_Extensions=Hiragana (or both)"';
969 $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
970}
971
972# This program can create files for enumerated-like properties, such as
973# 'Numeric_Type'. This file would be the same format as for a string
974# property, with a mapping from code point to its value, so you could look up,
975# for example, the script a code point is in. But no one so far wants this
976# mapping, or they have found another way to get it since this is a new
977# feature. So no file is generated except if it is in this list.
978my @output_mapped_properties = split "\n", <<END;
979END
980
c12f2655
KW
981# If you are using the Unihan database in a Unicode version before 5.2, you
982# need to add the properties that you want to extract from it to this table.
983# For your convenience, the properties in the 6.0 PropertyAliases.txt file are
984# listed, commented out
99870f4d
KW
985my @cjk_properties = split "\n", <<'END';
986#cjkAccountingNumeric; kAccountingNumeric
987#cjkOtherNumeric; kOtherNumeric
988#cjkPrimaryNumeric; kPrimaryNumeric
989#cjkCompatibilityVariant; kCompatibilityVariant
990#cjkIICore ; kIICore
991#cjkIRG_GSource; kIRG_GSource
992#cjkIRG_HSource; kIRG_HSource
993#cjkIRG_JSource; kIRG_JSource
994#cjkIRG_KPSource; kIRG_KPSource
995#cjkIRG_KSource; kIRG_KSource
996#cjkIRG_TSource; kIRG_TSource
997#cjkIRG_USource; kIRG_USource
998#cjkIRG_VSource; kIRG_VSource
999#cjkRSUnicode; kRSUnicode ; Unicode_Radical_Stroke; URS
1000END
1001
1002# Similarly for the property values. For your convenience, the lines in the
5f7264c7 1003# 6.0 PropertyAliases.txt file are listed. Just remove the first BUT NOT both
c12f2655 1004# '#' marks (for Unicode versions before 5.2)
99870f4d
KW
1005my @cjk_property_values = split "\n", <<'END';
1006## @missing: 0000..10FFFF; cjkAccountingNumeric; NaN
1007## @missing: 0000..10FFFF; cjkCompatibilityVariant; <code point>
1008## @missing: 0000..10FFFF; cjkIICore; <none>
1009## @missing: 0000..10FFFF; cjkIRG_GSource; <none>
1010## @missing: 0000..10FFFF; cjkIRG_HSource; <none>
1011## @missing: 0000..10FFFF; cjkIRG_JSource; <none>
1012## @missing: 0000..10FFFF; cjkIRG_KPSource; <none>
1013## @missing: 0000..10FFFF; cjkIRG_KSource; <none>
1014## @missing: 0000..10FFFF; cjkIRG_TSource; <none>
1015## @missing: 0000..10FFFF; cjkIRG_USource; <none>
1016## @missing: 0000..10FFFF; cjkIRG_VSource; <none>
1017## @missing: 0000..10FFFF; cjkOtherNumeric; NaN
1018## @missing: 0000..10FFFF; cjkPrimaryNumeric; NaN
1019## @missing: 0000..10FFFF; cjkRSUnicode; <none>
1020END
1021
1022# The input files don't list every code point. Those not listed are to be
1023# defaulted to some value. Below are hard-coded what those values are for
1024# non-binary properties as of 5.1. Starting in 5.0, there are
1025# machine-parsable comment lines in the files the give the defaults; so this
1026# list shouldn't have to be extended. The claim is that all missing entries
1027# for binary properties will default to 'N'. Unicode tried to change that in
1028# 5.2, but the beta period produced enough protest that they backed off.
1029#
1030# The defaults for the fields that appear in UnicodeData.txt in this hash must
1031# be in the form that it expects. The others may be synonyms.
1032my $CODE_POINT = '<code point>';
1033my %default_mapping = (
1034 Age => "Unassigned",
1035 # Bidi_Class => Complicated; set in code
1036 Bidi_Mirroring_Glyph => "",
1037 Block => 'No_Block',
1038 Canonical_Combining_Class => 0,
1039 Case_Folding => $CODE_POINT,
1040 Decomposition_Mapping => $CODE_POINT,
1041 Decomposition_Type => 'None',
1042 East_Asian_Width => "Neutral",
1043 FC_NFKC_Closure => $CODE_POINT,
1044 General_Category => 'Cn',
1045 Grapheme_Cluster_Break => 'Other',
1046 Hangul_Syllable_Type => 'NA',
1047 ISO_Comment => "",
1048 Jamo_Short_Name => "",
1049 Joining_Group => "No_Joining_Group",
1050 # Joining_Type => Complicated; set in code
1051 kIICore => 'N', # Is converted to binary
1052 #Line_Break => Complicated; set in code
1053 Lowercase_Mapping => $CODE_POINT,
1054 Name => "",
1055 Name_Alias => "",
1056 NFC_QC => 'Yes',
1057 NFD_QC => 'Yes',
1058 NFKC_QC => 'Yes',
1059 NFKD_QC => 'Yes',
1060 Numeric_Type => 'None',
1061 Numeric_Value => 'NaN',
1062 Script => ($v_version le 4.1.0) ? 'Common' : 'Unknown',
1063 Sentence_Break => 'Other',
1064 Simple_Case_Folding => $CODE_POINT,
1065 Simple_Lowercase_Mapping => $CODE_POINT,
1066 Simple_Titlecase_Mapping => $CODE_POINT,
1067 Simple_Uppercase_Mapping => $CODE_POINT,
1068 Titlecase_Mapping => $CODE_POINT,
1069 Unicode_1_Name => "",
1070 Unicode_Radical_Stroke => "",
1071 Uppercase_Mapping => $CODE_POINT,
1072 Word_Break => 'Other',
1073);
1074
1075# Below are files that Unicode furnishes, but this program ignores, and why
1076my %ignored_files = (
73ba1144
KW
1077 'CJKRadicals.txt' => 'Maps the kRSUnicode property values to corresponding code points',
1078 'Index.txt' => 'Alphabetical index of Unicode characters',
1079 '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',
1080 'NamesList.txt' => 'Annotated list of characters',
1081 'NormalizationCorrections.txt' => 'Documentation of corrections already incorporated into the Unicode data base',
1082 'Props.txt' => 'Only in very early releases; is a subset of F<PropList.txt> (which is used instead)',
1083 'ReadMe.txt' => 'Documentation',
1084 '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>',
1085 'EmojiSources.txt' => 'Maps certain Unicode code points to their legacy Japanese cell-phone values',
73ba1144
KW
1086 'auxiliary/WordBreakTest.html' => 'Documentation of validation tests',
1087 'auxiliary/SentenceBreakTest.html' => 'Documentation of validation tests',
1088 'auxiliary/GraphemeBreakTest.html' => 'Documentation of validation tests',
1089 'auxiliary/LineBreakTest.html' => 'Documentation of validation tests',
99870f4d
KW
1090);
1091
1fec9f60
KW
1092my %skipped_files; # List of files that we skip
1093
678f13d5 1094### End of externally interesting definitions, except for @input_file_objects
99870f4d
KW
1095
1096my $HEADER=<<"EOF";
1097# !!!!!!! DO NOT EDIT THIS FILE !!!!!!!
3df51b85
KW
1098# This file is machine-generated by $0 from the Unicode
1099# database, Version $string_version. Any changes made here will be lost!
cf25bb62
JH
1100EOF
1101
126c3d4e 1102my $INTERNAL_ONLY_HEADER = <<"EOF";
99870f4d
KW
1103
1104# !!!!!!! INTERNAL PERL USE ONLY !!!!!!!
fac53429
KW
1105# This file is for internal use by core Perl only. The format and even the
1106# name or existence of this file are subject to change without notice. Don't
1107# use it directly.
99870f4d
KW
1108EOF
1109
1110my $DEVELOPMENT_ONLY=<<"EOF";
1111# !!!!!!! DEVELOPMENT USE ONLY !!!!!!!
1112# This file contains information artificially constrained to code points
1113# present in Unicode release $string_compare_versions.
1114# IT CANNOT BE RELIED ON. It is for use during development only and should
23e33b60 1115# not be used for production.
b6922eda
KW
1116
1117EOF
1118
6189eadc
KW
1119my $MAX_UNICODE_CODEPOINT_STRING = "10FFFF";
1120my $MAX_UNICODE_CODEPOINT = hex $MAX_UNICODE_CODEPOINT_STRING;
1121my $MAX_UNICODE_CODEPOINTS = $MAX_UNICODE_CODEPOINT + 1;
99870f4d
KW
1122
1123# Matches legal code point. 4-6 hex numbers, If there are 6, the first
1124# two must be 10; if there are 5, the first must not be a 0. Written this way
92199589
KW
1125# to decrease backtracking. The first regex allows the code point to be at
1126# the end of a word, but to work properly, the word shouldn't end with a valid
1127# hex character. The second one won't match a code point at the end of a
1128# word, and doesn't have the run-on issue
8c32d378
KW
1129my $run_on_code_point_re =
1130 qr/ (?: 10[0-9A-F]{4} | [1-9A-F][0-9A-F]{4} | [0-9A-F]{4} ) \b/x;
1131my $code_point_re = qr/\b$run_on_code_point_re/;
99870f4d
KW
1132
1133# This matches the beginning of the line in the Unicode db files that give the
1134# defaults for code points not listed (i.e., missing) in the file. The code
1135# depends on this ending with a semi-colon, so it can assume it is a valid
1136# field when the line is split() by semi-colons
1137my $missing_defaults_prefix =
6189eadc 1138 qr/^#\s+\@missing:\s+0000\.\.$MAX_UNICODE_CODEPOINT_STRING\s*;/;
99870f4d
KW
1139
1140# Property types. Unicode has more types, but these are sufficient for our
1141# purposes.
1142my $UNKNOWN = -1; # initialized to illegal value
1143my $NON_STRING = 1; # Either binary or enum
1144my $BINARY = 2;
06f26c45
KW
1145my $FORCED_BINARY = 3; # Not a binary property, but, besides its normal
1146 # tables, additional true and false tables are
1147 # generated so that false is anything matching the
1148 # default value, and true is everything else.
1149my $ENUM = 4; # Include catalog
1150my $STRING = 5; # Anything else: string or misc
99870f4d
KW
1151
1152# Some input files have lines that give default values for code points not
1153# contained in the file. Sometimes these should be ignored.
1154my $NO_DEFAULTS = 0; # Must evaluate to false
1155my $NOT_IGNORED = 1;
1156my $IGNORED = 2;
1157
1158# Range types. Each range has a type. Most ranges are type 0, for normal,
1159# and will appear in the main body of the tables in the output files, but
1160# there are other types of ranges as well, listed below, that are specially
1161# handled. There are pseudo-types as well that will never be stored as a
1162# type, but will affect the calculation of the type.
1163
1164# 0 is for normal, non-specials
1165my $MULTI_CP = 1; # Sequence of more than code point
1166my $HANGUL_SYLLABLE = 2;
1167my $CP_IN_NAME = 3; # The NAME contains the code point appended to it.
1168my $NULL = 4; # The map is to the null string; utf8.c can't
1169 # handle these, nor is there an accepted syntax
1170 # for them in \p{} constructs
f86864ac 1171my $COMPUTE_NO_MULTI_CP = 5; # Pseudo-type; means that ranges that would
99870f4d
KW
1172 # otherwise be $MULTI_CP type are instead type 0
1173
1174# process_generic_property_file() can accept certain overrides in its input.
1175# Each of these must begin AND end with $CMD_DELIM.
1176my $CMD_DELIM = "\a";
1177my $REPLACE_CMD = 'replace'; # Override the Replace
1178my $MAP_TYPE_CMD = 'map_type'; # Override the Type
1179
1180my $NO = 0;
1181my $YES = 1;
1182
1183# Values for the Replace argument to add_range.
1184# $NO # Don't replace; add only the code points not
1185 # already present.
1186my $IF_NOT_EQUIVALENT = 1; # Replace only under certain conditions; details in
1187 # the comments at the subroutine definition.
1188my $UNCONDITIONALLY = 2; # Replace without conditions.
9470941f 1189my $MULTIPLE_BEFORE = 4; # Don't replace, but add a duplicate record if
99870f4d 1190 # already there
56343c78 1191my $CROAK = 5; # Die with an error if is already there
99870f4d
KW
1192
1193# Flags to give property statuses. The phrases are to remind maintainers that
1194# if the flag is changed, the indefinite article referring to it in the
1195# documentation may need to be as well.
1196my $NORMAL = "";
99870f4d
KW
1197my $DEPRECATED = 'D';
1198my $a_bold_deprecated = "a 'B<$DEPRECATED>'";
1199my $A_bold_deprecated = "A 'B<$DEPRECATED>'";
1200my $DISCOURAGED = 'X';
1201my $a_bold_discouraged = "an 'B<$DISCOURAGED>'";
1202my $A_bold_discouraged = "An 'B<$DISCOURAGED>'";
1203my $STRICTER = 'T';
1204my $a_bold_stricter = "a 'B<$STRICTER>'";
1205my $A_bold_stricter = "A 'B<$STRICTER>'";
1206my $STABILIZED = 'S';
1207my $a_bold_stabilized = "an 'B<$STABILIZED>'";
1208my $A_bold_stabilized = "An 'B<$STABILIZED>'";
1209my $OBSOLETE = 'O';
1210my $a_bold_obsolete = "an 'B<$OBSOLETE>'";
1211my $A_bold_obsolete = "An 'B<$OBSOLETE>'";
1212
1213my %status_past_participles = (
1214 $DISCOURAGED => 'discouraged',
99870f4d
KW
1215 $STABILIZED => 'stabilized',
1216 $OBSOLETE => 'obsolete',
37e2e78e 1217 $DEPRECATED => 'deprecated',
99870f4d
KW
1218);
1219
395dfc19
KW
1220# Table fates. These are somewhat ordered, so that fates < $MAP_PROXIED should be
1221# externally documented.
301ba948 1222my $ORDINARY = 0; # The normal fate.
395dfc19
KW
1223my $MAP_PROXIED = 1; # The map table for the property isn't written out,
1224 # but there is a file written that can be used to
1225 # reconstruct this table
301ba948
KW
1226my $SUPPRESSED = 3; # The file for this table is not written out.
1227my $INTERNAL_ONLY = 4; # The file for this table is written out, but it is
1228 # for Perl's internal use only
1229my $PLACEHOLDER = 5; # A property that is defined as a placeholder in a
1230 # Unicode version that doesn't have it, but we need it
1231 # to be defined, if empty, to have things work.
1232 # Implies no pod entry generated
1233
f5817e0a
KW
1234# The format of the values of the tables:
1235my $EMPTY_FORMAT = "";
99870f4d
KW
1236my $BINARY_FORMAT = 'b';
1237my $DECIMAL_FORMAT = 'd';
1238my $FLOAT_FORMAT = 'f';
1239my $INTEGER_FORMAT = 'i';
1240my $HEX_FORMAT = 'x';
1241my $RATIONAL_FORMAT = 'r';
1242my $STRING_FORMAT = 's';
a14f3cb1 1243my $DECOMP_STRING_FORMAT = 'c';
c3ff2976 1244my $STRING_WHITE_SPACE_LIST = 'sw';
99870f4d
KW
1245
1246my %map_table_formats = (
1247 $BINARY_FORMAT => 'binary',
1248 $DECIMAL_FORMAT => 'single decimal digit',
1249 $FLOAT_FORMAT => 'floating point number',
1250 $INTEGER_FORMAT => 'integer',
add63c13 1251 $HEX_FORMAT => 'non-negative hex whole number; a code point',
99870f4d 1252 $RATIONAL_FORMAT => 'rational: an integer or a fraction',
1a9d544b 1253 $STRING_FORMAT => 'string',
92f9d56c 1254 $DECOMP_STRING_FORMAT => 'Perl\'s internal (Normalize.pm) decomposition mapping',
c3ff2976 1255 $STRING_WHITE_SPACE_LIST => 'string, but some elements are interpreted as a list; white space occurs only as list item separators'
99870f4d
KW
1256);
1257
1258# Unicode didn't put such derived files in a separate directory at first.
1259my $EXTRACTED_DIR = (-d 'extracted') ? 'extracted' : "";
1260my $EXTRACTED = ($EXTRACTED_DIR) ? "$EXTRACTED_DIR/" : "";
1261my $AUXILIARY = 'auxiliary';
1262
1263# Hashes that will eventually go into Heavy.pl for the use of utf8_heavy.pl
9e4a1e86 1264# and into UCD.pl for the use of UCD.pm
99870f4d
KW
1265my %loose_to_file_of; # loosely maps table names to their respective
1266 # files
1267my %stricter_to_file_of; # same; but for stricter mapping.
315bfd4e 1268my %loose_property_to_file_of; # Maps a loose property name to its map file
89cf10cc
KW
1269my %file_to_swash_name; # Maps the file name to its corresponding key name
1270 # in the hash %utf8::SwashInfo
99870f4d
KW
1271my %nv_floating_to_rational; # maps numeric values floating point numbers to
1272 # their rational equivalent
c12f2655
KW
1273my %loose_property_name_of; # Loosely maps (non_string) property names to
1274 # standard form
86a52d1e 1275my %string_property_loose_to_name; # Same, for string properties.
c15fda25
KW
1276my %loose_defaults; # keys are of form "prop=value", where 'prop' is
1277 # the property name in standard loose form, and
1278 # 'value' is the default value for that property,
1279 # also in standard loose form.
9e4a1e86
KW
1280my %loose_to_standard_value; # loosely maps table names to the canonical
1281 # alias for them
2df7880f
KW
1282my %ambiguous_names; # keys are alias names (in standard form) that
1283 # have more than one possible meaning.
5d1df013
KW
1284my %prop_aliases; # Keys are standard property name; values are each
1285 # one's aliases
1e863613
KW
1286my %prop_value_aliases; # Keys of top level are standard property name;
1287 # values are keys to another hash, Each one is
1288 # one of the property's values, in standard form.
1289 # The values are that prop-val's aliases.
2df7880f 1290my %ucd_pod; # Holds entries that will go into the UCD section of the pod
99870f4d 1291
d867ccfb
KW
1292# Most properties are immune to caseless matching, otherwise you would get
1293# nonsensical results, as properties are a function of a code point, not
1294# everything that is caselessly equivalent to that code point. For example,
1295# Changes_When_Case_Folded('s') should be false, whereas caselessly it would
1296# be true because 's' and 'S' are equivalent caselessly. However,
1297# traditionally, [:upper:] and [:lower:] are equivalent caselessly, so we
1298# extend that concept to those very few properties that are like this. Each
1299# such property will match the full range caselessly. They are hard-coded in
1300# the program; it's not worth trying to make it general as it's extremely
1301# unlikely that they will ever change.
1302my %caseless_equivalent_to;
1303
99870f4d
KW
1304# These constants names and values were taken from the Unicode standard,
1305# version 5.1, section 3.12. They are used in conjunction with Hangul
6e5a209b
KW
1306# syllables. The '_string' versions are so generated tables can retain the
1307# hex format, which is the more familiar value
1308my $SBase_string = "0xAC00";
1309my $SBase = CORE::hex $SBase_string;
1310my $LBase_string = "0x1100";
1311my $LBase = CORE::hex $LBase_string;
1312my $VBase_string = "0x1161";
1313my $VBase = CORE::hex $VBase_string;
1314my $TBase_string = "0x11A7";
1315my $TBase = CORE::hex $TBase_string;
99870f4d
KW
1316my $SCount = 11172;
1317my $LCount = 19;
1318my $VCount = 21;
1319my $TCount = 28;
1320my $NCount = $VCount * $TCount;
1321
1322# For Hangul syllables; These store the numbers from Jamo.txt in conjunction
1323# with the above published constants.
1324my %Jamo;
1325my %Jamo_L; # Leading consonants
1326my %Jamo_V; # Vowels
1327my %Jamo_T; # Trailing consonants
1328
bb1dd3da
KW
1329# For code points whose name contains its ordinal as a '-ABCD' suffix.
1330# The key is the base name of the code point, and the value is an
1331# array giving all the ranges that use this base name. Each range
1332# is actually a hash giving the 'low' and 'high' values of it.
1333my %names_ending_in_code_point;
1334my %loose_names_ending_in_code_point; # Same as above, but has blanks, dashes
1335 # removed from the names
1336# Inverse mapping. The list of ranges that have these kinds of
1337# names. Each element contains the low, high, and base names in an
1338# anonymous hash.
1339my @code_points_ending_in_code_point;
1340
1341# Boolean: does this Unicode version have the hangul syllables, and are we
1342# writing out a table for them?
1343my $has_hangul_syllables = 0;
1344
1345# Does this Unicode version have code points whose names end in their
1346# respective code points, and are we writing out a table for them? 0 for no;
1347# otherwise points to first property that a table is needed for them, so that
1348# if multiple tables are needed, we don't create duplicates
1349my $needing_code_points_ending_in_code_point = 0;
1350
37e2e78e 1351my @backslash_X_tests; # List of tests read in for testing \X
99870f4d
KW
1352my @unhandled_properties; # Will contain a list of properties found in
1353 # the input that we didn't process.
f86864ac 1354my @match_properties; # Properties that have match tables, to be
99870f4d
KW
1355 # listed in the pod
1356my @map_properties; # Properties that get map files written
1357my @named_sequences; # NamedSequences.txt contents.
1358my %potential_files; # Generated list of all .txt files in the directory
1359 # structure so we can warn if something is being
1360 # ignored.
1361my @files_actually_output; # List of files we generated.
1362my @more_Names; # Some code point names are compound; this is used
1363 # to store the extra components of them.
1364my $MIN_FRACTION_LENGTH = 3; # How many digits of a floating point number at
1365 # the minimum before we consider it equivalent to a
1366 # candidate rational
1367my $MAX_FLOATING_SLOP = 10 ** - $MIN_FRACTION_LENGTH; # And in floating terms
1368
1369# These store references to certain commonly used property objects
1370my $gc;
1371my $perl;
1372my $block;
3e20195b
KW
1373my $perl_charname;
1374my $print;
7fc6cb55 1375my $Any;
359523e2 1376my $script;
99870f4d
KW
1377
1378# Are there conflicting names because of beginning with 'In_', or 'Is_'
1379my $has_In_conflicts = 0;
1380my $has_Is_conflicts = 0;
1381
1382sub internal_file_to_platform ($) {
1383 # Convert our file paths which have '/' separators to those of the
1384 # platform.
1385
1386 my $file = shift;
1387 return undef unless defined $file;
1388
1389 return File::Spec->join(split '/', $file);
d07a55ed 1390}
5beb625e 1391
99870f4d
KW
1392sub file_exists ($) { # platform independent '-e'. This program internally
1393 # uses slash as a path separator.
1394 my $file = shift;
1395 return 0 if ! defined $file;
1396 return -e internal_file_to_platform($file);
1397}
5beb625e 1398
99870f4d 1399sub objaddr($) {
23e33b60
KW
1400 # Returns the address of the blessed input object.
1401 # It doesn't check for blessedness because that would do a string eval
1402 # every call, and the program is structured so that this is never called
1403 # for a non-blessed object.
99870f4d 1404
23e33b60 1405 no overloading; # If overloaded, numifying below won't work.
99870f4d
KW
1406
1407 # Numifying a ref gives its address.
051df77b 1408 return pack 'J', $_[0];
99870f4d
KW
1409}
1410
558712cf 1411# These are used only if $annotate is true.
c4019d52
KW
1412# The entire range of Unicode characters is examined to populate these
1413# after all the input has been processed. But most can be skipped, as they
1414# have the same descriptive phrases, such as being unassigned
1415my @viacode; # Contains the 1 million character names
1416my @printable; # boolean: And are those characters printable?
1417my @annotate_char_type; # Contains a type of those characters, specifically
1418 # for the purposes of annotation.
1419my $annotate_ranges; # A map of ranges of code points that have the same
98dc9551 1420 # name for the purposes of annotation. They map to the
c4019d52
KW
1421 # upper edge of the range, so that the end point can
1422 # be immediately found. This is used to skip ahead to
1423 # the end of a range, and avoid processing each
1424 # individual code point in it.
1425my $unassigned_sans_noncharacters; # A Range_List of the unassigned
1426 # characters, but excluding those which are
1427 # also noncharacter code points
1428
1429# The annotation types are an extension of the regular range types, though
1430# some of the latter are folded into one. Make the new types negative to
1431# avoid conflicting with the regular types
1432my $SURROGATE_TYPE = -1;
1433my $UNASSIGNED_TYPE = -2;
1434my $PRIVATE_USE_TYPE = -3;
1435my $NONCHARACTER_TYPE = -4;
1436my $CONTROL_TYPE = -5;
1437my $UNKNOWN_TYPE = -6; # Used only if there is a bug in this program
1438
1439sub populate_char_info ($) {
558712cf 1440 # Used only with the $annotate option. Populates the arrays with the
c4019d52
KW
1441 # input code point's info that are needed for outputting more detailed
1442 # comments. If calling context wants a return, it is the end point of
1443 # any contiguous range of characters that share essentially the same info
1444
1445 my $i = shift;
1446 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1447
1448 $viacode[$i] = $perl_charname->value_of($i) || "";
1449
1450 # A character is generally printable if Unicode says it is,
1451 # but below we make sure that most Unicode general category 'C' types
1452 # aren't.
1453 $printable[$i] = $print->contains($i);
1454
1455 $annotate_char_type[$i] = $perl_charname->type_of($i) || 0;
1456
1457 # Only these two regular types are treated specially for annotations
1458 # purposes
1459 $annotate_char_type[$i] = 0 if $annotate_char_type[$i] != $CP_IN_NAME
1460 && $annotate_char_type[$i] != $HANGUL_SYLLABLE;
1461
1462 # Give a generic name to all code points that don't have a real name.
1463 # We output ranges, if applicable, for these. Also calculate the end
1464 # point of the range.
1465 my $end;
1466 if (! $viacode[$i]) {
1467 if ($gc-> table('Surrogate')->contains($i)) {
1468 $viacode[$i] = 'Surrogate';
1469 $annotate_char_type[$i] = $SURROGATE_TYPE;
1470 $printable[$i] = 0;
1471 $end = $gc->table('Surrogate')->containing_range($i)->end;
1472 }
1473 elsif ($gc-> table('Private_use')->contains($i)) {
1474 $viacode[$i] = 'Private Use';
1475 $annotate_char_type[$i] = $PRIVATE_USE_TYPE;
1476 $printable[$i] = 0;
1477 $end = $gc->table('Private_Use')->containing_range($i)->end;
1478 }
1479 elsif (Property::property_ref('Noncharacter_Code_Point')-> table('Y')->
1480 contains($i))
1481 {
1482 $viacode[$i] = 'Noncharacter';
1483 $annotate_char_type[$i] = $NONCHARACTER_TYPE;
1484 $printable[$i] = 0;
1485 $end = property_ref('Noncharacter_Code_Point')->table('Y')->
1486 containing_range($i)->end;
1487 }
1488 elsif ($gc-> table('Control')->contains($i)) {
1489 $viacode[$i] = 'Control';
1490 $annotate_char_type[$i] = $CONTROL_TYPE;
1491 $printable[$i] = 0;
1492 $end = 0x81 if $i == 0x80; # Hard-code this one known case
1493 }
1494 elsif ($gc-> table('Unassigned')->contains($i)) {
1495 $viacode[$i] = 'Unassigned, block=' . $block-> value_of($i);
1496 $annotate_char_type[$i] = $UNASSIGNED_TYPE;
1497 $printable[$i] = 0;
1498
1499 # Because we name the unassigned by the blocks they are in, it
1500 # can't go past the end of that block, and it also can't go past
1501 # the unassigned range it is in. The special table makes sure
1502 # that the non-characters, which are unassigned, are separated
1503 # out.
1504 $end = min($block->containing_range($i)->end,
1505 $unassigned_sans_noncharacters-> containing_range($i)->
1506 end);
13ca76ff
KW
1507 }
1508 else {
1509 Carp::my_carp_bug("Can't figure out how to annotate "
1510 . sprintf("U+%04X", $i)
1511 . ". Proceeding anyway.");
c4019d52
KW
1512 $viacode[$i] = 'UNKNOWN';
1513 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1514 $printable[$i] = 0;
1515 }
1516 }
1517
1518 # Here, has a name, but if it's one in which the code point number is
1519 # appended to the name, do that.
1520 elsif ($annotate_char_type[$i] == $CP_IN_NAME) {
1521 $viacode[$i] .= sprintf("-%04X", $i);
1522 $end = $perl_charname->containing_range($i)->end;
1523 }
1524
1525 # And here, has a name, but if it's a hangul syllable one, replace it with
1526 # the correct name from the Unicode algorithm
1527 elsif ($annotate_char_type[$i] == $HANGUL_SYLLABLE) {
1528 use integer;
1529 my $SIndex = $i - $SBase;
1530 my $L = $LBase + $SIndex / $NCount;
1531 my $V = $VBase + ($SIndex % $NCount) / $TCount;
1532 my $T = $TBase + $SIndex % $TCount;
1533 $viacode[$i] = "HANGUL SYLLABLE $Jamo{$L}$Jamo{$V}";
1534 $viacode[$i] .= $Jamo{$T} if $T != $TBase;
1535 $end = $perl_charname->containing_range($i)->end;
1536 }
1537
1538 return if ! defined wantarray;
1539 return $i if ! defined $end; # If not a range, return the input
1540
1541 # Save this whole range so can find the end point quickly
1542 $annotate_ranges->add_map($i, $end, $end);
1543
1544 return $end;
1545}
1546
23e33b60
KW
1547# Commented code below should work on Perl 5.8.
1548## This 'require' doesn't necessarily work in miniperl, and even if it does,
1549## the native perl version of it (which is what would operate under miniperl)
1550## is extremely slow, as it does a string eval every call.
1551#my $has_fast_scalar_util = $\18 !~ /miniperl/
1552# && defined eval "require Scalar::Util";
1553#
1554#sub objaddr($) {
1555# # Returns the address of the blessed input object. Uses the XS version if
1556# # available. It doesn't check for blessedness because that would do a
1557# # string eval every call, and the program is structured so that this is
1558# # never called for a non-blessed object.
1559#
1560# return Scalar::Util::refaddr($_[0]) if $has_fast_scalar_util;
1561#
1562# # Check at least that is a ref.
1563# my $pkg = ref($_[0]) or return undef;
1564#
1565# # Change to a fake package to defeat any overloaded stringify
1566# bless $_[0], 'main::Fake';
1567#
1568# # Numifying a ref gives its address.
051df77b 1569# my $addr = pack 'J', $_[0];
23e33b60
KW
1570#
1571# # Return to original class
1572# bless $_[0], $pkg;
1573# return $addr;
1574#}
1575
99870f4d
KW
1576sub max ($$) {
1577 my $a = shift;
1578 my $b = shift;
1579 return $a if $a >= $b;
1580 return $b;
1581}
1582
1583sub min ($$) {
1584 my $a = shift;
1585 my $b = shift;
1586 return $a if $a <= $b;
1587 return $b;
1588}
1589
1590sub clarify_number ($) {
1591 # This returns the input number with underscores inserted every 3 digits
1592 # in large (5 digits or more) numbers. Input must be entirely digits, not
1593 # checked.
1594
1595 my $number = shift;
1596 my $pos = length($number) - 3;
1597 return $number if $pos <= 1;
1598 while ($pos > 0) {
1599 substr($number, $pos, 0) = '_';
1600 $pos -= 3;
5beb625e 1601 }
99870f4d 1602 return $number;
99598c8c
JH
1603}
1604
12ac2576 1605
99870f4d 1606package Carp;
7ebf06b3 1607
99870f4d
KW
1608# These routines give a uniform treatment of messages in this program. They
1609# are placed in the Carp package to cause the stack trace to not include them,
1610# although an alternative would be to use another package and set @CARP_NOT
1611# for it.
12ac2576 1612
99870f4d 1613our $Verbose = 1 if main::DEBUG; # Useful info when debugging
12ac2576 1614
99f78760
KW
1615# This is a work-around suggested by Nicholas Clark to fix a problem with Carp
1616# and overload trying to load Scalar:Util under miniperl. See
1617# http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2009-11/msg01057.html
1618undef $overload::VERSION;
1619
99870f4d
KW
1620sub my_carp {
1621 my $message = shift || "";
1622 my $nofold = shift || 0;
7ebf06b3 1623
99870f4d
KW
1624 if ($message) {
1625 $message = main::join_lines($message);
1626 $message =~ s/^$0: *//; # Remove initial program name
1627 $message =~ s/[.;,]+$//; # Remove certain ending punctuation
1628 $message = "\n$0: $message;";
12ac2576 1629
99870f4d
KW
1630 # Fold the message with program name, semi-colon end punctuation
1631 # (which looks good with the message that carp appends to it), and a
1632 # hanging indent for continuation lines.
1633 $message = main::simple_fold($message, "", 4) unless $nofold;
1634 $message =~ s/\n$//; # Remove the trailing nl so what carp
1635 # appends is to the same line
1636 }
12ac2576 1637
99870f4d 1638 return $message if defined wantarray; # If a caller just wants the msg
12ac2576 1639
99870f4d
KW
1640 carp $message;
1641 return;
1642}
7ebf06b3 1643
99870f4d
KW
1644sub my_carp_bug {
1645 # This is called when it is clear that the problem is caused by a bug in
1646 # this program.
7ebf06b3 1647
99870f4d
KW
1648 my $message = shift;
1649 $message =~ s/^$0: *//;
1650 $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");
1651 carp $message;
1652 return;
1653}
7ebf06b3 1654
99870f4d
KW
1655sub carp_too_few_args {
1656 if (@_ != 2) {
1657 my_carp_bug("Wrong number of arguments: to 'carp_too_few_arguments'. No action taken.");
1658 return;
12ac2576 1659 }
7ebf06b3 1660
99870f4d
KW
1661 my $args_ref = shift;
1662 my $count = shift;
7ebf06b3 1663
99870f4d
KW
1664 my_carp_bug("Need at least $count arguments to "
1665 . (caller 1)[3]
1666 . ". Instead got: '"
1667 . join ', ', @$args_ref
1668 . "'. No action taken.");
1669 return;
12ac2576
JP
1670}
1671
99870f4d
KW
1672sub carp_extra_args {
1673 my $args_ref = shift;
1674 my_carp_bug("Too many arguments to 'carp_extra_args': (" . join(', ', @_) . "); Extras ignored.") if @_;
12ac2576 1675
99870f4d
KW
1676 unless (ref $args_ref) {
1677 my_carp_bug("Argument to 'carp_extra_args' ($args_ref) must be a ref. Not checking arguments.");
1678 return;
1679 }
1680 my ($package, $file, $line) = caller;
1681 my $subroutine = (caller 1)[3];
cf25bb62 1682
99870f4d
KW
1683 my $list;
1684 if (ref $args_ref eq 'HASH') {
1685 foreach my $key (keys %$args_ref) {
1686 $args_ref->{$key} = $UNDEF unless defined $args_ref->{$key};
cf25bb62 1687 }
99870f4d 1688 $list = join ', ', each %{$args_ref};
cf25bb62 1689 }
99870f4d
KW
1690 elsif (ref $args_ref eq 'ARRAY') {
1691 foreach my $arg (@$args_ref) {
1692 $arg = $UNDEF unless defined $arg;
1693 }
1694 $list = join ', ', @$args_ref;
1695 }
1696 else {
1697 my_carp_bug("Can't cope with ref "
1698 . ref($args_ref)
1699 . " . argument to 'carp_extra_args'. Not checking arguments.");
1700 return;
1701 }
1702
1703 my_carp_bug("Unrecognized parameters in options: '$list' to $subroutine. Skipped.");
1704 return;
d73e5302
JH
1705}
1706
99870f4d
KW
1707package main;
1708
1709{ # Closure
1710
1711 # This program uses the inside-out method for objects, as recommended in
1712 # "Perl Best Practices". This closure aids in generating those. There
1713 # are two routines. setup_package() is called once per package to set
1714 # things up, and then set_access() is called for each hash representing a
1715 # field in the object. These routines arrange for the object to be
1716 # properly destroyed when no longer used, and for standard accessor
1717 # functions to be generated. If you need more complex accessors, just
1718 # write your own and leave those accesses out of the call to set_access().
1719 # More details below.
1720
1721 my %constructor_fields; # fields that are to be used in constructors; see
1722 # below
1723
1724 # The values of this hash will be the package names as keys to other
1725 # hashes containing the name of each field in the package as keys, and
1726 # references to their respective hashes as values.
1727 my %package_fields;
1728
1729 sub setup_package {
1730 # Sets up the package, creating standard DESTROY and dump methods
1731 # (unless already defined). The dump method is used in debugging by
1732 # simple_dumper().
1733 # The optional parameters are:
1734 # a) a reference to a hash, that gets populated by later
1735 # set_access() calls with one of the accesses being
1736 # 'constructor'. The caller can then refer to this, but it is
1737 # not otherwise used by these two routines.
1738 # b) a reference to a callback routine to call during destruction
1739 # of the object, before any fields are actually destroyed
1740
1741 my %args = @_;
1742 my $constructor_ref = delete $args{'Constructor_Fields'};
1743 my $destroy_callback = delete $args{'Destroy_Callback'};
1744 Carp::carp_extra_args(\@_) if main::DEBUG && %args;
1745
1746 my %fields;
1747 my $package = (caller)[0];
1748
1749 $package_fields{$package} = \%fields;
1750 $constructor_fields{$package} = $constructor_ref;
1751
1752 unless ($package->can('DESTROY')) {
1753 my $destroy_name = "${package}::DESTROY";
1754 no strict "refs";
1755
1756 # Use typeglob to give the anonymous subroutine the name we want
1757 *$destroy_name = sub {
1758 my $self = shift;
ffe43484 1759 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1760
1761 $self->$destroy_callback if $destroy_callback;
1762 foreach my $field (keys %{$package_fields{$package}}) {
1763 #print STDERR __LINE__, ": Destroying ", ref $self, " ", sprintf("%04X", $addr), ": ", $field, "\n";
1764 delete $package_fields{$package}{$field}{$addr};
1765 }
1766 return;
1767 }
1768 }
1769
1770 unless ($package->can('dump')) {
1771 my $dump_name = "${package}::dump";
1772 no strict "refs";
1773 *$dump_name = sub {
1774 my $self = shift;
1775 return dump_inside_out($self, $package_fields{$package}, @_);
1776 }
1777 }
1778 return;
1779 }
1780
1781 sub set_access {
1782 # Arrange for the input field to be garbage collected when no longer
1783 # needed. Also, creates standard accessor functions for the field
1784 # based on the optional parameters-- none if none of these parameters:
1785 # 'addable' creates an 'add_NAME()' accessor function.
1786 # 'readable' or 'readable_array' creates a 'NAME()' accessor
1787 # function.
1788 # 'settable' creates a 'set_NAME()' accessor function.
1789 # 'constructor' doesn't create an accessor function, but adds the
1790 # field to the hash that was previously passed to
1791 # setup_package();
1792 # Any of the accesses can be abbreviated down, so that 'a', 'ad',
1793 # 'add' etc. all mean 'addable'.
1794 # The read accessor function will work on both array and scalar
1795 # values. If another accessor in the parameter list is 'a', the read
1796 # access assumes an array. You can also force it to be array access
1797 # by specifying 'readable_array' instead of 'readable'
1798 #
1799 # A sort-of 'protected' access can be set-up by preceding the addable,
1800 # readable or settable with some initial portion of 'protected_' (but,
1801 # the underscore is required), like 'p_a', 'pro_set', etc. The
1802 # "protection" is only by convention. All that happens is that the
1803 # accessor functions' names begin with an underscore. So instead of
1804 # calling set_foo, the call is _set_foo. (Real protection could be
c1739a4a 1805 # accomplished by having a new subroutine, end_package, called at the
99870f4d
KW
1806 # end of each package, and then storing the __LINE__ ranges and
1807 # checking them on every accessor. But that is way overkill.)
1808
1809 # We create anonymous subroutines as the accessors and then use
1810 # typeglobs to assign them to the proper package and name
1811
1812 my $name = shift; # Name of the field
1813 my $field = shift; # Reference to the inside-out hash containing the
1814 # field
1815
1816 my $package = (caller)[0];
1817
1818 if (! exists $package_fields{$package}) {
1819 croak "$0: Must call 'setup_package' before 'set_access'";
1820 }
d73e5302 1821
99870f4d
KW
1822 # Stash the field so DESTROY can get it.
1823 $package_fields{$package}{$name} = $field;
cf25bb62 1824
99870f4d
KW
1825 # Remaining arguments are the accessors. For each...
1826 foreach my $access (@_) {
1827 my $access = lc $access;
cf25bb62 1828
99870f4d 1829 my $protected = "";
cf25bb62 1830
99870f4d
KW
1831 # Match the input as far as it goes.
1832 if ($access =~ /^(p[^_]*)_/) {
1833 $protected = $1;
1834 if (substr('protected_', 0, length $protected)
1835 eq $protected)
1836 {
1837
1838 # Add 1 for the underscore not included in $protected
1839 $access = substr($access, length($protected) + 1);
1840 $protected = '_';
1841 }
1842 else {
1843 $protected = "";
1844 }
1845 }
1846
1847 if (substr('addable', 0, length $access) eq $access) {
1848 my $subname = "${package}::${protected}add_$name";
1849 no strict "refs";
1850
1851 # add_ accessor. Don't add if already there, which we
1852 # determine using 'eq' for scalars and '==' otherwise.
1853 *$subname = sub {
1854 use strict "refs";
1855 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
1856 my $self = shift;
1857 my $value = shift;
ffe43484 1858 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1859 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1860 if (ref $value) {
f998e60c 1861 return if grep { $value == $_ } @{$field->{$addr}};
99870f4d
KW
1862 }
1863 else {
f998e60c 1864 return if grep { $value eq $_ } @{$field->{$addr}};
99870f4d 1865 }
f998e60c 1866 push @{$field->{$addr}}, $value;
99870f4d
KW
1867 return;
1868 }
1869 }
1870 elsif (substr('constructor', 0, length $access) eq $access) {
1871 if ($protected) {
1872 Carp::my_carp_bug("Can't set-up 'protected' constructors")
1873 }
1874 else {
1875 $constructor_fields{$package}{$name} = $field;
1876 }
1877 }
1878 elsif (substr('readable_array', 0, length $access) eq $access) {
1879
1880 # Here has read access. If one of the other parameters for
1881 # access is array, or this one specifies array (by being more
1882 # than just 'readable_'), then create a subroutine that
1883 # assumes the data is an array. Otherwise just a scalar
1884 my $subname = "${package}::${protected}$name";
1885 if (grep { /^a/i } @_
1886 or length($access) > length('readable_'))
1887 {
1888 no strict "refs";
1889 *$subname = sub {
1890 use strict "refs";
23e33b60 1891 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
ffe43484 1892 my $addr = do { no overloading; pack 'J', $_[0]; };
99870f4d
KW
1893 if (ref $field->{$addr} ne 'ARRAY') {
1894 my $type = ref $field->{$addr};
1895 $type = 'scalar' unless $type;
1896 Carp::my_carp_bug("Trying to read $name as an array when it is a $type. Big problems.");
1897 return;
1898 }
1899 return scalar @{$field->{$addr}} unless wantarray;
1900
1901 # Make a copy; had problems with caller modifying the
1902 # original otherwise
1903 my @return = @{$field->{$addr}};
1904 return @return;
1905 }
1906 }
1907 else {
1908
1909 # Here not an array value, a simpler function.
1910 no strict "refs";
1911 *$subname = sub {
1912 use strict "refs";
23e33b60 1913 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
f998e60c 1914 no overloading;
051df77b 1915 return $field->{pack 'J', $_[0]};
99870f4d
KW
1916 }
1917 }
1918 }
1919 elsif (substr('settable', 0, length $access) eq $access) {
1920 my $subname = "${package}::${protected}set_$name";
1921 no strict "refs";
1922 *$subname = sub {
1923 use strict "refs";
23e33b60
KW
1924 if (main::DEBUG) {
1925 return Carp::carp_too_few_args(\@_, 2) if @_ < 2;
1926 Carp::carp_extra_args(\@_) if @_ > 2;
1927 }
1928 # $self is $_[0]; $value is $_[1]
f998e60c 1929 no overloading;
051df77b 1930 $field->{pack 'J', $_[0]} = $_[1];
99870f4d
KW
1931 return;
1932 }
1933 }
1934 else {
1935 Carp::my_carp_bug("Unknown accessor type $access. No accessor set.");
1936 }
cf25bb62 1937 }
99870f4d 1938 return;
cf25bb62 1939 }
99870f4d
KW
1940}
1941
1942package Input_file;
1943
1944# All input files use this object, which stores various attributes about them,
1945# and provides for convenient, uniform handling. The run method wraps the
1946# processing. It handles all the bookkeeping of opening, reading, and closing
1947# the file, returning only significant input lines.
1948#
1949# Each object gets a handler which processes the body of the file, and is
1950# called by run(). Most should use the generic, default handler, which has
1951# code scrubbed to handle things you might not expect. A handler should
1952# basically be a while(next_line()) {...} loop.
1953#
1954# You can also set up handlers to
1955# 1) call before the first line is read for pre processing
1956# 2) call to adjust each line of the input before the main handler gets them
1957# 3) call upon EOF before the main handler exits its loop
1958# 4) call at the end for post processing
1959#
1960# $_ is used to store the input line, and is to be filtered by the
1961# each_line_handler()s. So, if the format of the line is not in the desired
1962# format for the main handler, these are used to do that adjusting. They can
1963# be stacked (by enclosing them in an [ anonymous array ] in the constructor,
1964# so the $_ output of one is used as the input to the next. None of the other
1965# handlers are stackable, but could easily be changed to be so.
1966#
1967# Most of the handlers can call insert_lines() or insert_adjusted_lines()
1968# which insert the parameters as lines to be processed before the next input
1969# file line is read. This allows the EOF handler to flush buffers, for
1970# example. The difference between the two routines is that the lines inserted
1971# by insert_lines() are subjected to the each_line_handler()s. (So if you
1972# called it from such a handler, you would get infinite recursion.) Lines
1973# inserted by insert_adjusted_lines() go directly to the main handler without
1974# any adjustments. If the post-processing handler calls any of these, there
1975# will be no effect. Some error checking for these conditions could be added,
1976# but it hasn't been done.
1977#
1978# carp_bad_line() should be called to warn of bad input lines, which clears $_
1979# to prevent further processing of the line. This routine will output the
1980# message as a warning once, and then keep a count of the lines that have the
1981# same message, and output that count at the end of the file's processing.
1982# This keeps the number of messages down to a manageable amount.
1983#
1984# get_missings() should be called to retrieve any @missing input lines.
1985# Messages will be raised if this isn't done if the options aren't to ignore
1986# missings.
1987
1988sub trace { return main::trace(@_); }
1989
99870f4d
KW
1990{ # Closure
1991 # Keep track of fields that are to be put into the constructor.
1992 my %constructor_fields;
1993
1994 main::setup_package(Constructor_Fields => \%constructor_fields);
1995
1996 my %file; # Input file name, required
1997 main::set_access('file', \%file, qw{ c r });
1998
1999 my %first_released; # Unicode version file was first released in, required
2000 main::set_access('first_released', \%first_released, qw{ c r });
2001
2002 my %handler; # Subroutine to process the input file, defaults to
2003 # 'process_generic_property_file'
2004 main::set_access('handler', \%handler, qw{ c });
2005
2006 my %property;
2007 # name of property this file is for. defaults to none, meaning not
2008 # applicable, or is otherwise determinable, for example, from each line.
2009 main::set_access('property', \%property, qw{ c });
2010
2011 my %optional;
2012 # If this is true, the file is optional. If not present, no warning is
2013 # output. If it is present, the string given by this parameter is
2014 # evaluated, and if false the file is not processed.
2015 main::set_access('optional', \%optional, 'c', 'r');
2016
2017 my %non_skip;
2018 # This is used for debugging, to skip processing of all but a few input
2019 # files. Add 'non_skip => 1' to the constructor for those files you want
2020 # processed when you set the $debug_skip global.
2021 main::set_access('non_skip', \%non_skip, 'c');
2022
37e2e78e 2023 my %skip;
09ca89ce
KW
2024 # This is used to skip processing of this input file semi-permanently,
2025 # when it evaluates to true. The value should be the reason the file is
2026 # being skipped. It is used for files that we aren't planning to process
2027 # anytime soon, but want to allow to be in the directory and not raise a
2028 # message that we are not handling. Mostly for test files. This is in
2029 # contrast to the non_skip element, which is supposed to be used very
2030 # temporarily for debugging. Sets 'optional' to 1. Also, files that we
2031 # pretty much will never look at can be placed in the global
1fec9f60 2032 # %ignored_files instead. Ones used here will be added to %skipped files
37e2e78e
KW
2033 main::set_access('skip', \%skip, 'c');
2034
99870f4d
KW
2035 my %each_line_handler;
2036 # list of subroutines to look at and filter each non-comment line in the
2037 # file. defaults to none. The subroutines are called in order, each is
2038 # to adjust $_ for the next one, and the final one adjusts it for
2039 # 'handler'
2040 main::set_access('each_line_handler', \%each_line_handler, 'c');
2041
2042 my %has_missings_defaults;
2043 # ? Are there lines in the file giving default values for code points
2044 # missing from it?. Defaults to NO_DEFAULTS. Otherwise NOT_IGNORED is
2045 # the norm, but IGNORED means it has such lines, but the handler doesn't
2046 # use them. Having these three states allows us to catch changes to the
2047 # UCD that this program should track
2048 main::set_access('has_missings_defaults',
2049 \%has_missings_defaults, qw{ c r });
2050
2051 my %pre_handler;
2052 # Subroutine to call before doing anything else in the file. If undef, no
2053 # such handler is called.
2054 main::set_access('pre_handler', \%pre_handler, qw{ c });
2055
2056 my %eof_handler;
2057 # Subroutine to call upon getting an EOF on the input file, but before
2058 # that is returned to the main handler. This is to allow buffers to be
2059 # flushed. The handler is expected to call insert_lines() or
2060 # insert_adjusted() with the buffered material
2061 main::set_access('eof_handler', \%eof_handler, qw{ c r });
2062
2063 my %post_handler;
2064 # Subroutine to call after all the lines of the file are read in and
2065 # processed. If undef, no such handler is called.
2066 main::set_access('post_handler', \%post_handler, qw{ c });
2067
2068 my %progress_message;
2069 # Message to print to display progress in lieu of the standard one
2070 main::set_access('progress_message', \%progress_message, qw{ c });
2071
2072 my %handle;
2073 # cache open file handle, internal. Is undef if file hasn't been
2074 # processed at all, empty if has;
2075 main::set_access('handle', \%handle);
2076
2077 my %added_lines;
2078 # cache of lines added virtually to the file, internal
2079 main::set_access('added_lines', \%added_lines);
2080
2081 my %errors;
2082 # cache of errors found, internal
2083 main::set_access('errors', \%errors);
2084
2085 my %missings;
2086 # storage of '@missing' defaults lines
2087 main::set_access('missings', \%missings);
2088
2089 sub new {
2090 my $class = shift;
2091
2092 my $self = bless \do{ my $anonymous_scalar }, $class;
ffe43484 2093 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2094
2095 # Set defaults
2096 $handler{$addr} = \&main::process_generic_property_file;
2097 $non_skip{$addr} = 0;
37e2e78e 2098 $skip{$addr} = 0;
99870f4d
KW
2099 $has_missings_defaults{$addr} = $NO_DEFAULTS;
2100 $handle{$addr} = undef;
2101 $added_lines{$addr} = [ ];
2102 $each_line_handler{$addr} = [ ];
2103 $errors{$addr} = { };
2104 $missings{$addr} = [ ];
2105
2106 # Two positional parameters.
99f78760 2107 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
99870f4d
KW
2108 $file{$addr} = main::internal_file_to_platform(shift);
2109 $first_released{$addr} = shift;
2110
2111 # The rest of the arguments are key => value pairs
2112 # %constructor_fields has been set up earlier to list all possible
2113 # ones. Either set or push, depending on how the default has been set
2114 # up just above.
2115 my %args = @_;
2116 foreach my $key (keys %args) {
2117 my $argument = $args{$key};
2118
2119 # Note that the fields are the lower case of the constructor keys
2120 my $hash = $constructor_fields{lc $key};
2121 if (! defined $hash) {
2122 Carp::my_carp_bug("Unrecognized parameters '$key => $argument' to new() for $self. Skipped");
2123 next;
2124 }
2125 if (ref $hash->{$addr} eq 'ARRAY') {
2126 if (ref $argument eq 'ARRAY') {
2127 foreach my $argument (@{$argument}) {
2128 next if ! defined $argument;
2129 push @{$hash->{$addr}}, $argument;
2130 }
2131 }
2132 else {
2133 push @{$hash->{$addr}}, $argument if defined $argument;
2134 }
2135 }
2136 else {
2137 $hash->{$addr} = $argument;
2138 }
2139 delete $args{$key};
2140 };
2141
2142 # If the file has a property for it, it means that the property is not
2143 # listed in the file's entries. So add a handler to the list of line
2144 # handlers to insert the property name into the lines, to provide a
2145 # uniform interface to the final processing subroutine.
2146 # the final code doesn't have to worry about that.
2147 if ($property{$addr}) {
2148 push @{$each_line_handler{$addr}}, \&_insert_property_into_line;
2149 }
2150
2151 if ($non_skip{$addr} && ! $debug_skip && $verbosity) {
2152 print "Warning: " . __PACKAGE__ . " constructor for $file{$addr} has useless 'non_skip' in it\n";
a3a8c5f0 2153 }
99870f4d 2154
09ca89ce
KW
2155 # If skipping, set to optional, and add to list of ignored files,
2156 # including its reason
2157 if ($skip{$addr}) {
2158 $optional{$addr} = 1;
1fec9f60 2159 $skipped_files{$file{$addr}} = $skip{$addr}
09ca89ce 2160 }
37e2e78e 2161
99870f4d 2162 return $self;
d73e5302
JH
2163 }
2164
cf25bb62 2165
99870f4d
KW
2166 use overload
2167 fallback => 0,
2168 qw("") => "_operator_stringify",
2169 "." => \&main::_operator_dot,
2170 ;
cf25bb62 2171
99870f4d
KW
2172 sub _operator_stringify {
2173 my $self = shift;
cf25bb62 2174
99870f4d 2175 return __PACKAGE__ . " object for " . $self->file;
d73e5302 2176 }
d73e5302 2177
99870f4d
KW
2178 # flag to make sure extracted files are processed early
2179 my $seen_non_extracted_non_age = 0;
d73e5302 2180
99870f4d
KW
2181 sub run {
2182 # Process the input object $self. This opens and closes the file and
2183 # calls all the handlers for it. Currently, this can only be called
2184 # once per file, as it destroy's the EOF handler
d73e5302 2185
99870f4d
KW
2186 my $self = shift;
2187 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
b6922eda 2188
ffe43484 2189 my $addr = do { no overloading; pack 'J', $self; };
b6922eda 2190
99870f4d 2191 my $file = $file{$addr};
d73e5302 2192
99870f4d
KW
2193 # Don't process if not expecting this file (because released later
2194 # than this Unicode version), and isn't there. This means if someone
2195 # copies it into an earlier version's directory, we will go ahead and
2196 # process it.
2197 return if $first_released{$addr} gt $v_version && ! -e $file;
2198
2199 # If in debugging mode and this file doesn't have the non-skip
2200 # flag set, and isn't one of the critical files, skip it.
2201 if ($debug_skip
2202 && $first_released{$addr} ne v0
2203 && ! $non_skip{$addr})
2204 {
2205 print "Skipping $file in debugging\n" if $verbosity;
2206 return;
2207 }
2208
2209 # File could be optional
37e2e78e 2210 if ($optional{$addr}) {
99870f4d
KW
2211 return unless -e $file;
2212 my $result = eval $optional{$addr};
2213 if (! defined $result) {
2214 Carp::my_carp_bug("Got '$@' when tried to eval $optional{$addr}. $file Skipped.");
2215 return;
2216 }
2217 if (! $result) {
2218 if ($verbosity) {
2219 print STDERR "Skipping processing input file '$file' because '$optional{$addr}' is not true\n";
2220 }
2221 return;
2222 }
2223 }
2224
2225 if (! defined $file || ! -e $file) {
2226
2227 # If the file doesn't exist, see if have internal data for it
2228 # (based on first_released being 0).
2229 if ($first_released{$addr} eq v0) {
2230 $handle{$addr} = 'pretend_is_open';
2231 }
2232 else {
2233 if (! $optional{$addr} # File could be optional
2234 && $v_version ge $first_released{$addr})
2235 {
2236 print STDERR "Skipping processing input file '$file' because not found\n" if $v_version ge $first_released{$addr};
2237 }
2238 return;
2239 }
2240 }
2241 else {
2242
37e2e78e
KW
2243 # Here, the file exists. Some platforms may change the case of
2244 # its name
99870f4d 2245 if ($seen_non_extracted_non_age) {
517956bf 2246 if ($file =~ /$EXTRACTED/i) {
99870f4d 2247 Carp::my_carp_bug(join_lines(<<END
99f78760 2248$file should be processed just after the 'Prop...Alias' files, and before
99870f4d
KW
2249anything not in the $EXTRACTED_DIR directory. Proceeding, but the results may
2250have subtle problems
2251END
2252 ));
2253 }
2254 }
2255 elsif ($EXTRACTED_DIR
2256 && $first_released{$addr} ne v0
517956bf
CB
2257 && $file !~ /$EXTRACTED/i
2258 && lc($file) ne 'dage.txt')
99870f4d
KW
2259 {
2260 # We don't set this (by the 'if' above) if we have no
2261 # extracted directory, so if running on an early version,
2262 # this test won't work. Not worth worrying about.
2263 $seen_non_extracted_non_age = 1;
2264 }
2265
2266 # And mark the file as having being processed, and warn if it
2267 # isn't a file we are expecting. As we process the files,
2268 # they are deleted from the hash, so any that remain at the
2269 # end of the program are files that we didn't process.
517956bf 2270 my $fkey = File::Spec->rel2abs($file);
faf3cf6b
KW
2271 my $expecting = delete $potential_files{lc($fkey)};
2272
678f13d5
KW
2273 Carp::my_carp("Was not expecting '$file'.") if
2274 ! $expecting
99870f4d
KW
2275 && ! defined $handle{$addr};
2276
37e2e78e
KW
2277 # Having deleted from expected files, we can quit if not to do
2278 # anything. Don't print progress unless really want verbosity
2279 if ($skip{$addr}) {
2280 print "Skipping $file.\n" if $verbosity >= $VERBOSE;
2281 return;
2282 }
2283
99870f4d
KW
2284 # Open the file, converting the slashes used in this program
2285 # into the proper form for the OS
2286 my $file_handle;
2287 if (not open $file_handle, "<", $file) {
2288 Carp::my_carp("Can't open $file. Skipping: $!");
2289 return 0;
2290 }
2291 $handle{$addr} = $file_handle; # Cache the open file handle
2292 }
2293
2294 if ($verbosity >= $PROGRESS) {
2295 if ($progress_message{$addr}) {
2296 print "$progress_message{$addr}\n";
2297 }
2298 else {
2299 # If using a virtual file, say so.
2300 print "Processing ", (-e $file)
2301 ? $file
2302 : "substitute $file",
2303 "\n";
2304 }
2305 }
2306
2307
2308 # Call any special handler for before the file.
2309 &{$pre_handler{$addr}}($self) if $pre_handler{$addr};
2310
2311 # Then the main handler
2312 &{$handler{$addr}}($self);
2313
2314 # Then any special post-file handler.
2315 &{$post_handler{$addr}}($self) if $post_handler{$addr};
2316
2317 # If any errors have been accumulated, output the counts (as the first
2318 # error message in each class was output when it was encountered).
2319 if ($errors{$addr}) {
2320 my $total = 0;
2321 my $types = 0;
2322 foreach my $error (keys %{$errors{$addr}}) {
2323 $total += $errors{$addr}->{$error};
2324 delete $errors{$addr}->{$error};
2325 $types++;
2326 }
2327 if ($total > 1) {
2328 my $message
2329 = "A total of $total lines had errors in $file. ";
2330
2331 $message .= ($types == 1)
2332 ? '(Only the first one was displayed.)'
2333 : '(Only the first of each type was displayed.)';
2334 Carp::my_carp($message);
2335 }
2336 }
2337
2338 if (@{$missings{$addr}}) {
2339 Carp::my_carp_bug("Handler for $file didn't look at all the \@missing lines. Generated tables likely are wrong");
2340 }
2341
2342 # If a real file handle, close it.
2343 close $handle{$addr} or Carp::my_carp("Can't close $file: $!") if
2344 ref $handle{$addr};
2345 $handle{$addr} = ""; # Uses empty to indicate that has already seen
2346 # the file, as opposed to undef
2347 return;
2348 }
2349
2350 sub next_line {
2351 # Sets $_ to be the next logical input line, if any. Returns non-zero
2352 # if such a line exists. 'logical' means that any lines that have
2353 # been added via insert_lines() will be returned in $_ before the file
2354 # is read again.
2355
2356 my $self = shift;
2357 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2358
ffe43484 2359 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2360
2361 # Here the file is open (or if the handle is not a ref, is an open
2362 # 'virtual' file). Get the next line; any inserted lines get priority
2363 # over the file itself.
2364 my $adjusted;
2365
2366 LINE:
2367 while (1) { # Loop until find non-comment, non-empty line
2368 #local $to_trace = 1 if main::DEBUG;
2369 my $inserted_ref = shift @{$added_lines{$addr}};
2370 if (defined $inserted_ref) {
2371 ($adjusted, $_) = @{$inserted_ref};
2372 trace $adjusted, $_ if main::DEBUG && $to_trace;
2373 return 1 if $adjusted;
2374 }
2375 else {
2376 last if ! ref $handle{$addr}; # Don't read unless is real file
2377 last if ! defined ($_ = readline $handle{$addr});
2378 }
2379 chomp;
2380 trace $_ if main::DEBUG && $to_trace;
2381
2382 # See if this line is the comment line that defines what property
2383 # value that code points that are not listed in the file should
2384 # have. The format or existence of these lines is not guaranteed
2385 # by Unicode since they are comments, but the documentation says
2386 # that this was added for machine-readability, so probably won't
2387 # change. This works starting in Unicode Version 5.0. They look
2388 # like:
2389 #
2390 # @missing: 0000..10FFFF; Not_Reordered
2391 # @missing: 0000..10FFFF; Decomposition_Mapping; <code point>
2392 # @missing: 0000..10FFFF; ; NaN
2393 #
2394 # Save the line for a later get_missings() call.
2395 if (/$missing_defaults_prefix/) {
2396 if ($has_missings_defaults{$addr} == $NO_DEFAULTS) {
2397 $self->carp_bad_line("Unexpected \@missing line. Assuming no missing entries");
2398 }
2399 elsif ($has_missings_defaults{$addr} == $NOT_IGNORED) {
2400 my @defaults = split /\s* ; \s*/x, $_;
2401
2402 # The first field is the @missing, which ends in a
2403 # semi-colon, so can safely shift.
2404 shift @defaults;
2405
2406 # Some of these lines may have empty field placeholders
2407 # which get in the way. An example is:
2408 # @missing: 0000..10FFFF; ; NaN
2409 # Remove them. Process starting from the top so the
2410 # splice doesn't affect things still to be looked at.
2411 for (my $i = @defaults - 1; $i >= 0; $i--) {
2412 next if $defaults[$i] ne "";
2413 splice @defaults, $i, 1;
2414 }
2415
2416 # What's left should be just the property (maybe) and the
2417 # default. Having only one element means it doesn't have
2418 # the property.
2419 my $default;
2420 my $property;
2421 if (@defaults >= 1) {
2422 if (@defaults == 1) {
2423 $default = $defaults[0];
2424 }
2425 else {
2426 $property = $defaults[0];
2427 $default = $defaults[1];
2428 }
2429 }
2430
2431 if (@defaults < 1
2432 || @defaults > 2
2433 || ($default =~ /^</
2434 && $default !~ /^<code *point>$/i
2435 && $default !~ /^<none>$/i))
2436 {
2437 $self->carp_bad_line("Unrecognized \@missing line: $_. Assuming no missing entries");
2438 }
2439 else {
2440
2441 # If the property is missing from the line, it should
2442 # be the one for the whole file
2443 $property = $property{$addr} if ! defined $property;
2444
2445 # Change <none> to the null string, which is what it
2446 # really means. If the default is the code point
2447 # itself, set it to <code point>, which is what
2448 # Unicode uses (but sometimes they've forgotten the
2449 # space)
2450 if ($default =~ /^<none>$/i) {
2451 $default = "";
2452 }
2453 elsif ($default =~ /^<code *point>$/i) {
2454 $default = $CODE_POINT;
2455 }
2456
2457 # Store them as a sub-arrays with both components.
2458 push @{$missings{$addr}}, [ $default, $property ];
2459 }
2460 }
2461
2462 # There is nothing for the caller to process on this comment
2463 # line.
2464 next;
2465 }
2466
2467 # Remove comments and trailing space, and skip this line if the
2468 # result is empty
2469 s/#.*//;
2470 s/\s+$//;
2471 next if /^$/;
2472
2473 # Call any handlers for this line, and skip further processing of
2474 # the line if the handler sets the line to null.
2475 foreach my $sub_ref (@{$each_line_handler{$addr}}) {
2476 &{$sub_ref}($self);
2477 next LINE if /^$/;
2478 }
2479
2480 # Here the line is ok. return success.
2481 return 1;
2482 } # End of looping through lines.
2483
2484 # If there is an EOF handler, call it (only once) and if it generates
2485 # more lines to process go back in the loop to handle them.
2486 if ($eof_handler{$addr}) {
2487 &{$eof_handler{$addr}}($self);
2488 $eof_handler{$addr} = ""; # Currently only get one shot at it.
2489 goto LINE if $added_lines{$addr};
2490 }
2491
2492 # Return failure -- no more lines.
2493 return 0;
2494
2495 }
2496
2497# Not currently used, not fully tested.
2498# sub peek {
2499# # Non-destructive look-ahead one non-adjusted, non-comment, non-blank
2500# # record. Not callable from an each_line_handler(), nor does it call
2501# # an each_line_handler() on the line.
2502#
2503# my $self = shift;
ffe43484 2504# my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2505#
2506# foreach my $inserted_ref (@{$added_lines{$addr}}) {
2507# my ($adjusted, $line) = @{$inserted_ref};
2508# next if $adjusted;
2509#
2510# # Remove comments and trailing space, and return a non-empty
2511# # resulting line
2512# $line =~ s/#.*//;
2513# $line =~ s/\s+$//;
2514# return $line if $line ne "";
2515# }
2516#
2517# return if ! ref $handle{$addr}; # Don't read unless is real file
2518# while (1) { # Loop until find non-comment, non-empty line
2519# local $to_trace = 1 if main::DEBUG;
2520# trace $_ if main::DEBUG && $to_trace;
2521# return if ! defined (my $line = readline $handle{$addr});
2522# chomp $line;
2523# push @{$added_lines{$addr}}, [ 0, $line ];
2524#
2525# $line =~ s/#.*//;
2526# $line =~ s/\s+$//;
2527# return $line if $line ne "";
2528# }
2529#
2530# return;
2531# }
2532
2533
2534 sub insert_lines {
2535 # Lines can be inserted so that it looks like they were in the input
2536 # file at the place it was when this routine is called. See also
2537 # insert_adjusted_lines(). Lines inserted via this routine go through
2538 # any each_line_handler()
2539
2540 my $self = shift;
2541
2542 # Each inserted line is an array, with the first element being 0 to
2543 # indicate that this line hasn't been adjusted, and needs to be
2544 # processed.
f998e60c 2545 no overloading;
051df77b 2546 push @{$added_lines{pack 'J', $self}}, map { [ 0, $_ ] } @_;
99870f4d
KW
2547 return;
2548 }
2549
2550 sub insert_adjusted_lines {
2551 # Lines can be inserted so that it looks like they were in the input
2552 # file at the place it was when this routine is called. See also
2553 # insert_lines(). Lines inserted via this routine are already fully
2554 # adjusted, ready to be processed; each_line_handler()s handlers will
2555 # not be called. This means this is not a completely general
2556 # facility, as only the last each_line_handler on the stack should
2557 # call this. It could be made more general, by passing to each of the
2558 # line_handlers their position on the stack, which they would pass on
2559 # to this routine, and that would replace the boolean first element in
2560 # the anonymous array pushed here, so that the next_line routine could
2561 # use that to call only those handlers whose index is after it on the
2562 # stack. But this is overkill for what is needed now.
2563
2564 my $self = shift;
2565 trace $_[0] if main::DEBUG && $to_trace;
2566
2567 # Each inserted line is an array, with the first element being 1 to
2568 # indicate that this line has been adjusted
f998e60c 2569 no overloading;
051df77b 2570 push @{$added_lines{pack 'J', $self}}, map { [ 1, $_ ] } @_;
99870f4d
KW
2571 return;
2572 }
2573
2574 sub get_missings {
2575 # Returns the stored up @missings lines' values, and clears the list.
2576 # The values are in an array, consisting of the default in the first
2577 # element, and the property in the 2nd. However, since these lines
2578 # can be stacked up, the return is an array of all these arrays.
2579
2580 my $self = shift;
2581 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2582
ffe43484 2583 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2584
2585 # If not accepting a list return, just return the first one.
2586 return shift @{$missings{$addr}} unless wantarray;
2587
2588 my @return = @{$missings{$addr}};
2589 undef @{$missings{$addr}};
2590 return @return;
2591 }
2592
2593 sub _insert_property_into_line {
2594 # Add a property field to $_, if this file requires it.
2595
f998e60c 2596 my $self = shift;
ffe43484 2597 my $addr = do { no overloading; pack 'J', $self; };
f998e60c 2598 my $property = $property{$addr};
99870f4d
KW
2599 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2600
2601 $_ =~ s/(;|$)/; $property$1/;
2602 return;
2603 }
2604
2605 sub carp_bad_line {
2606 # Output consistent error messages, using either a generic one, or the
2607 # one given by the optional parameter. To avoid gazillions of the
2608 # same message in case the syntax of a file is way off, this routine
2609 # only outputs the first instance of each message, incrementing a
2610 # count so the totals can be output at the end of the file.
2611
2612 my $self = shift;
2613 my $message = shift;
2614 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2615
ffe43484 2616 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2617
2618 $message = 'Unexpected line' unless $message;
2619
2620 # No trailing punctuation so as to fit with our addenda.
2621 $message =~ s/[.:;,]$//;
2622
2623 # If haven't seen this exact message before, output it now. Otherwise
2624 # increment the count of how many times it has occurred
2625 unless ($errors{$addr}->{$message}) {
2626 Carp::my_carp("$message in '$_' in "
f998e60c 2627 . $file{$addr}
99870f4d
KW
2628 . " at line $.. Skipping this line;");
2629 $errors{$addr}->{$message} = 1;
2630 }
2631 else {
2632 $errors{$addr}->{$message}++;
2633 }
2634
2635 # Clear the line to prevent any further (meaningful) processing of it.
2636 $_ = "";
2637
2638 return;
2639 }
2640} # End closure
2641
2642package Multi_Default;
2643
2644# Certain properties in early versions of Unicode had more than one possible
2645# default for code points missing from the files. In these cases, one
2646# default applies to everything left over after all the others are applied,
2647# and for each of the others, there is a description of which class of code
2648# points applies to it. This object helps implement this by storing the
2649# defaults, and for all but that final default, an eval string that generates
2650# the class that it applies to.
2651
2652
2653{ # Closure
2654
2655 main::setup_package();
2656
2657 my %class_defaults;
2658 # The defaults structure for the classes
2659 main::set_access('class_defaults', \%class_defaults);
2660
2661 my %other_default;
2662 # The default that applies to everything left over.
2663 main::set_access('other_default', \%other_default, 'r');
2664
2665
2666 sub new {
2667 # The constructor is called with default => eval pairs, terminated by
2668 # the left-over default. e.g.
2669 # Multi_Default->new(
2670 # 'T' => '$gc->table("Mn") + $gc->table("Cf") - 0x200C
2671 # - 0x200D',
2672 # 'R' => 'some other expression that evaluates to code points',
2673 # .
2674 # .
2675 # .
2676 # 'U'));
2677
2678 my $class = shift;
2679
2680 my $self = bless \do{my $anonymous_scalar}, $class;
ffe43484 2681 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2682
2683 while (@_ > 1) {
2684 my $default = shift;
2685 my $eval = shift;
2686 $class_defaults{$addr}->{$default} = $eval;
2687 }
2688
2689 $other_default{$addr} = shift;
2690
2691 return $self;
2692 }
2693
2694 sub get_next_defaults {
2695 # Iterates and returns the next class of defaults.
2696 my $self = shift;
2697 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2698
ffe43484 2699 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2700
2701 return each %{$class_defaults{$addr}};
2702 }
2703}
2704
2705package Alias;
2706
2707# An alias is one of the names that a table goes by. This class defines them
2708# including some attributes. Everything is currently setup in the
2709# constructor.
2710
2711
2712{ # Closure
2713
2714 main::setup_package();
2715
2716 my %name;
2717 main::set_access('name', \%name, 'r');
2718
2719 my %loose_match;
c12f2655 2720 # Should this name match loosely or not.
99870f4d
KW
2721 main::set_access('loose_match', \%loose_match, 'r');
2722
33e96e72
KW
2723 my %make_re_pod_entry;
2724 # Some aliases should not get their own entries in the re section of the
2725 # pod, because they are covered by a wild-card, and some we want to
2726 # discourage use of. Binary
f82fe4ba 2727 main::set_access('make_re_pod_entry', \%make_re_pod_entry, 'r', 's');
99870f4d 2728
fd1e3e84
KW
2729 my %ucd;
2730 # Is this documented to be accessible via Unicode::UCD
2731 main::set_access('ucd', \%ucd, 'r', 's');
2732
99870f4d
KW
2733 my %status;
2734 # Aliases have a status, like deprecated, or even suppressed (which means
2735 # they don't appear in documentation). Enum
2736 main::set_access('status', \%status, 'r');
2737
0eac1e20 2738 my %ok_as_filename;
99870f4d
KW
2739 # Similarly, some aliases should not be considered as usable ones for
2740 # external use, such as file names, or we don't want documentation to
2741 # recommend them. Boolean
0eac1e20 2742 main::set_access('ok_as_filename', \%ok_as_filename, 'r');
99870f4d
KW
2743
2744 sub new {
2745 my $class = shift;
2746
2747 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2748 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2749
2750 $name{$addr} = shift;
2751 $loose_match{$addr} = shift;
33e96e72 2752 $make_re_pod_entry{$addr} = shift;
0eac1e20 2753 $ok_as_filename{$addr} = shift;
99870f4d 2754 $status{$addr} = shift;
fd1e3e84 2755 $ucd{$addr} = shift;
99870f4d
KW
2756
2757 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2758
2759 # Null names are never ok externally
0eac1e20 2760 $ok_as_filename{$addr} = 0 if $name{$addr} eq "";
99870f4d
KW
2761
2762 return $self;
2763 }
2764}
2765
2766package Range;
2767
2768# A range is the basic unit for storing code points, and is described in the
2769# comments at the beginning of the program. Each range has a starting code
2770# point; an ending code point (not less than the starting one); a value
2771# that applies to every code point in between the two end-points, inclusive;
2772# and an enum type that applies to the value. The type is for the user's
2773# convenience, and has no meaning here, except that a non-zero type is
2774# considered to not obey the normal Unicode rules for having standard forms.
2775#
2776# The same structure is used for both map and match tables, even though in the
2777# latter, the value (and hence type) is irrelevant and could be used as a
2778# comment. In map tables, the value is what all the code points in the range
2779# map to. Type 0 values have the standardized version of the value stored as
2780# well, so as to not have to recalculate it a lot.
2781
2782sub trace { return main::trace(@_); }
2783
2784{ # Closure
2785
2786 main::setup_package();
2787
2788 my %start;
2789 main::set_access('start', \%start, 'r', 's');
2790
2791 my %end;
2792 main::set_access('end', \%end, 'r', 's');
2793
2794 my %value;
2795 main::set_access('value', \%value, 'r');
2796
2797 my %type;
2798 main::set_access('type', \%type, 'r');
2799
2800 my %standard_form;
2801 # The value in internal standard form. Defined only if the type is 0.
2802 main::set_access('standard_form', \%standard_form);
2803
2804 # Note that if these fields change, the dump() method should as well
2805
2806 sub new {
2807 return Carp::carp_too_few_args(\@_, 3) if main::DEBUG && @_ < 3;
2808 my $class = shift;
2809
2810 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2811 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2812
2813 $start{$addr} = shift;
2814 $end{$addr} = shift;
2815
2816 my %args = @_;
2817
2818 my $value = delete $args{'Value'}; # Can be 0
2819 $value = "" unless defined $value;
2820 $value{$addr} = $value;
2821
2822 $type{$addr} = delete $args{'Type'} || 0;
2823
2824 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2825
2826 if (! $type{$addr}) {
2827 $standard_form{$addr} = main::standardize($value);
2828 }
2829
2830 return $self;
2831 }
2832
2833 use overload
2834 fallback => 0,
2835 qw("") => "_operator_stringify",
2836 "." => \&main::_operator_dot,
2837 ;
2838
2839 sub _operator_stringify {
2840 my $self = shift;
ffe43484 2841 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2842
2843 # Output it like '0041..0065 (value)'
2844 my $return = sprintf("%04X", $start{$addr})
2845 . '..'
2846 . sprintf("%04X", $end{$addr});
2847 my $value = $value{$addr};
2848 my $type = $type{$addr};
2849 $return .= ' (';
2850 $return .= "$value";
2851 $return .= ", Type=$type" if $type != 0;
2852 $return .= ')';
2853
2854 return $return;
2855 }
2856
2857 sub standard_form {
2858 # The standard form is the value itself if the standard form is
2859 # undefined (that is if the value is special)
2860
2861 my $self = shift;
2862 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2863
ffe43484 2864 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2865
2866 return $standard_form{$addr} if defined $standard_form{$addr};
2867 return $value{$addr};
2868 }
2869
2870 sub dump {
2871 # Human, not machine readable. For machine readable, comment out this
2872 # entire routine and let the standard one take effect.
2873 my $self = shift;
2874 my $indent = shift;
2875 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2876
ffe43484 2877 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2878
2879 my $return = $indent
2880 . sprintf("%04X", $start{$addr})
2881 . '..'
2882 . sprintf("%04X", $end{$addr})
2883 . " '$value{$addr}';";
2884 if (! defined $standard_form{$addr}) {
2885 $return .= "(type=$type{$addr})";
2886 }
2887 elsif ($standard_form{$addr} ne $value{$addr}) {
2888 $return .= "(standard '$standard_form{$addr}')";
2889 }
2890 return $return;
2891 }
2892} # End closure
2893
2894package _Range_List_Base;
2895
2896# Base class for range lists. A range list is simply an ordered list of
2897# ranges, so that the ranges with the lowest starting numbers are first in it.
2898#
2899# When a new range is added that is adjacent to an existing range that has the
2900# same value and type, it merges with it to form a larger range.
2901#
2902# Ranges generally do not overlap, except that there can be multiple entries
2903# of single code point ranges. This is because of NameAliases.txt.
2904#
2905# In this program, there is a standard value such that if two different
2906# values, have the same standard value, they are considered equivalent. This
2907# value was chosen so that it gives correct results on Unicode data
2908
2909# There are a number of methods to manipulate range lists, and some operators
2910# are overloaded to handle them.
2911
99870f4d
KW
2912sub trace { return main::trace(@_); }
2913
2914{ # Closure
2915
2916 our $addr;
2917
2918 main::setup_package();
2919
2920 my %ranges;
2921 # The list of ranges
2922 main::set_access('ranges', \%ranges, 'readable_array');
2923
2924 my %max;
2925 # The highest code point in the list. This was originally a method, but
2926 # actual measurements said it was used a lot.
2927 main::set_access('max', \%max, 'r');
2928
2929 my %each_range_iterator;
2930 # Iterator position for each_range()
2931 main::set_access('each_range_iterator', \%each_range_iterator);
2932
2933 my %owner_name_of;
2934 # Name of parent this is attached to, if any. Solely for better error
2935 # messages.
2936 main::set_access('owner_name_of', \%owner_name_of, 'p_r');
2937
2938 my %_search_ranges_cache;
2939 # A cache of the previous result from _search_ranges(), for better
2940 # performance
2941 main::set_access('_search_ranges_cache', \%_search_ranges_cache);
2942
2943 sub new {
2944 my $class = shift;
2945 my %args = @_;
2946
2947 # Optional initialization data for the range list.
2948 my $initialize = delete $args{'Initialize'};
2949
2950 my $self;
2951
2952 # Use _union() to initialize. _union() returns an object of this
2953 # class, which means that it will call this constructor recursively.
2954 # But it won't have this $initialize parameter so that it won't
2955 # infinitely loop on this.
2956 return _union($class, $initialize, %args) if defined $initialize;
2957
2958 $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2959 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2960
2961 # Optional parent object, only for debug info.
2962 $owner_name_of{$addr} = delete $args{'Owner'};
2963 $owner_name_of{$addr} = "" if ! defined $owner_name_of{$addr};
2964
2965 # Stringify, in case it is an object.
2966 $owner_name_of{$addr} = "$owner_name_of{$addr}";
2967
2968 # This is used only for error messages, and so a colon is added
2969 $owner_name_of{$addr} .= ": " if $owner_name_of{$addr} ne "";
2970
2971 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2972
2973 # Max is initialized to a negative value that isn't adjacent to 0,
2974 # for simpler tests
2975 $max{$addr} = -2;
2976
2977 $_search_ranges_cache{$addr} = 0;
2978 $ranges{$addr} = [];
2979
2980 return $self;
2981 }
2982
2983 use overload
2984 fallback => 0,
2985 qw("") => "_operator_stringify",
2986 "." => \&main::_operator_dot,
2987 ;
2988
2989 sub _operator_stringify {
2990 my $self = shift;
ffe43484 2991 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2992
2993 return "Range_List attached to '$owner_name_of{$addr}'"
2994 if $owner_name_of{$addr};
2995 return "anonymous Range_List " . \$self;
2996 }
2997
2998 sub _union {
2999 # Returns the union of the input code points. It can be called as
3000 # either a constructor or a method. If called as a method, the result
3001 # will be a new() instance of the calling object, containing the union
3002 # of that object with the other parameter's code points; if called as
3003 # a constructor, the first parameter gives the class the new object
3004 # should be, and the second parameter gives the code points to go into
3005 # it.
3006 # In either case, there are two parameters looked at by this routine;
3007 # any additional parameters are passed to the new() constructor.
3008 #
3009 # The code points can come in the form of some object that contains
3010 # ranges, and has a conventionally named method to access them; or
3011 # they can be an array of individual code points (as integers); or
3012 # just a single code point.
3013 #
3014 # If they are ranges, this routine doesn't make any effort to preserve
3015 # the range values of one input over the other. Therefore this base
3016 # class should not allow _union to be called from other than
3017 # initialization code, so as to prevent two tables from being added
3018 # together where the range values matter. The general form of this
3019 # routine therefore belongs in a derived class, but it was moved here
3020 # to avoid duplication of code. The failure to overload this in this
3021 # class keeps it safe.
3022 #
3023
3024 my $self;
3025 my @args; # Arguments to pass to the constructor
3026
3027 my $class = shift;
3028
3029 # If a method call, will start the union with the object itself, and
3030 # the class of the new object will be the same as self.
3031 if (ref $class) {
3032 $self = $class;
3033 $class = ref $self;
3034 push @args, $self;
3035 }
3036
3037 # Add the other required parameter.
3038 push @args, shift;
3039 # Rest of parameters are passed on to the constructor
3040
3041 # Accumulate all records from both lists.
3042 my @records;
3043 for my $arg (@args) {
3044 #local $to_trace = 0 if main::DEBUG;
3045 trace "argument = $arg" if main::DEBUG && $to_trace;
3046 if (! defined $arg) {
3047 my $message = "";
3048 if (defined $self) {
f998e60c 3049 no overloading;
051df77b 3050 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
3051 }
3052 Carp::my_carp_bug($message .= "Undefined argument to _union. No union done.");
3053 return;
3054 }
3055 $arg = [ $arg ] if ! ref $arg;
3056 my $type = ref $arg;
3057 if ($type eq 'ARRAY') {
3058 foreach my $element (@$arg) {
3059 push @records, Range->new($element, $element);
3060 }
3061 }
3062 elsif ($arg->isa('Range')) {
3063 push @records, $arg;
3064 }
3065 elsif ($arg->can('ranges')) {
3066 push @records, $arg->ranges;
3067 }
3068 else {
3069 my $message = "";
3070 if (defined $self) {
f998e60c 3071 no overloading;
051df77b 3072 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
3073 }
3074 Carp::my_carp_bug($message . "Cannot take the union of a $type. No union done.");
3075 return;
3076 }
3077 }
3078
3079 # Sort with the range containing the lowest ordinal first, but if
3080 # two ranges start at the same code point, sort with the bigger range
3081 # of the two first, because it takes fewer cycles.
3082 @records = sort { ($a->start <=> $b->start)
3083 or
3084 # if b is shorter than a, b->end will be
3085 # less than a->end, and we want to select
3086 # a, so want to return -1
3087 ($b->end <=> $a->end)
3088 } @records;
3089
3090 my $new = $class->new(@_);
3091
3092 # Fold in records so long as they add new information.
3093 for my $set (@records) {
3094 my $start = $set->start;
3095 my $end = $set->end;
3096 my $value = $set->value;
3097 if ($start > $new->max) {
3098 $new->_add_delete('+', $start, $end, $value);
3099 }
3100 elsif ($end > $new->max) {
3101 $new->_add_delete('+', $new->max +1, $end, $value);
3102 }
3103 }
3104
3105 return $new;
3106 }
3107
3108 sub range_count { # Return the number of ranges in the range list
3109 my $self = shift;
3110 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3111
f998e60c 3112 no overloading;
051df77b 3113 return scalar @{$ranges{pack 'J', $self}};
99870f4d
KW
3114 }
3115
3116 sub min {
3117 # Returns the minimum code point currently in the range list, or if
3118 # the range list is empty, 2 beyond the max possible. This is a
3119 # method because used so rarely, that not worth saving between calls,
3120 # and having to worry about changing it as ranges are added and
3121 # deleted.
3122
3123 my $self = shift;
3124 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3125
ffe43484 3126 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3127
3128 # If the range list is empty, return a large value that isn't adjacent
3129 # to any that could be in the range list, for simpler tests
6189eadc 3130 return $MAX_UNICODE_CODEPOINT + 2 unless scalar @{$ranges{$addr}};
99870f4d
KW
3131 return $ranges{$addr}->[0]->start;
3132 }
3133
3134 sub contains {
3135 # Boolean: Is argument in the range list? If so returns $i such that:
3136 # range[$i]->end < $codepoint <= range[$i+1]->end
3137 # which is one beyond what you want; this is so that the 0th range
3138 # doesn't return false
3139 my $self = shift;
3140 my $codepoint = shift;
3141 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3142
99870f4d
KW
3143 my $i = $self->_search_ranges($codepoint);
3144 return 0 unless defined $i;
3145
3146 # The search returns $i, such that
3147 # range[$i-1]->end < $codepoint <= range[$i]->end
3148 # So is in the table if and only iff it is at least the start position
3149 # of range $i.
f998e60c 3150 no overloading;
051df77b 3151 return 0 if $ranges{pack 'J', $self}->[$i]->start > $codepoint;
99870f4d
KW
3152 return $i + 1;
3153 }
3154
2f7a8815
KW
3155 sub containing_range {
3156 # Returns the range object that contains the code point, undef if none
3157
3158 my $self = shift;
3159 my $codepoint = shift;
3160 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3161
3162 my $i = $self->contains($codepoint);
3163 return unless $i;
3164
3165 # contains() returns 1 beyond where we should look
3166 no overloading;
3167 return $ranges{pack 'J', $self}->[$i-1];
3168 }
3169
99870f4d
KW
3170 sub value_of {
3171 # Returns the value associated with the code point, undef if none
3172
3173 my $self = shift;
3174 my $codepoint = shift;
3175 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3176
d69c231b
KW
3177 my $range = $self->containing_range($codepoint);
3178 return unless defined $range;
99870f4d 3179
d69c231b 3180 return $range->value;
99870f4d
KW
3181 }
3182
0a9dbafc
KW
3183 sub type_of {
3184 # Returns the type of the range containing the code point, undef if
3185 # the code point is not in the table
3186
3187 my $self = shift;
3188 my $codepoint = shift;
3189 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3190
3191 my $range = $self->containing_range($codepoint);
3192 return unless defined $range;
3193
3194 return $range->type;
3195 }
3196
99870f4d
KW
3197 sub _search_ranges {
3198 # Find the range in the list which contains a code point, or where it
3199 # should go if were to add it. That is, it returns $i, such that:
3200 # range[$i-1]->end < $codepoint <= range[$i]->end
3201 # Returns undef if no such $i is possible (e.g. at end of table), or
3202 # if there is an error.
3203
3204 my $self = shift;
3205 my $code_point = shift;
3206 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3207
ffe43484 3208 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3209
3210 return if $code_point > $max{$addr};
3211 my $r = $ranges{$addr}; # The current list of ranges
3212 my $range_list_size = scalar @$r;
3213 my $i;
3214
3215 use integer; # want integer division
3216
3217 # Use the cached result as the starting guess for this one, because,
3218 # an experiment on 5.1 showed that 90% of the time the cache was the
3219 # same as the result on the next call (and 7% it was one less).
3220 $i = $_search_ranges_cache{$addr};
3221 $i = 0 if $i >= $range_list_size; # Reset if no longer valid (prob.
3222 # from an intervening deletion
3223 #local $to_trace = 1 if main::DEBUG;
3224 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);
3225 return $i if $code_point <= $r->[$i]->end
3226 && ($i == 0 || $r->[$i-1]->end < $code_point);
3227
3228 # Here the cache doesn't yield the correct $i. Try adding 1.
3229 if ($i < $range_list_size - 1
3230 && $r->[$i]->end < $code_point &&
3231 $code_point <= $r->[$i+1]->end)
3232 {
3233 $i++;
3234 trace "next \$i is correct: $i" if main::DEBUG && $to_trace;
3235 $_search_ranges_cache{$addr} = $i;
3236 return $i;
3237 }
3238
3239 # Here, adding 1 also didn't work. We do a binary search to
3240 # find the correct position, starting with current $i
3241 my $lower = 0;
3242 my $upper = $range_list_size - 1;
3243 while (1) {
3244 trace "top of loop i=$i:", sprintf("%04X", $r->[$lower]->start), "[$lower] .. ", sprintf("%04X", $r->[$i]->start), "[$i] .. ", sprintf("%04X", $r->[$upper]->start), "[$upper]" if main::DEBUG && $to_trace;
3245
3246 if ($code_point <= $r->[$i]->end) {
3247
3248 # Here we have met the upper constraint. We can quit if we
3249 # also meet the lower one.
3250 last if $i == 0 || $r->[$i-1]->end < $code_point;
3251
3252 $upper = $i; # Still too high.
3253
3254 }
3255 else {
3256
3257 # Here, $r[$i]->end < $code_point, so look higher up.
3258 $lower = $i;
3259 }
3260
3261 # Split search domain in half to try again.
3262 my $temp = ($upper + $lower) / 2;
3263
3264 # No point in continuing unless $i changes for next time
3265 # in the loop.
3266 if ($temp == $i) {
3267
3268 # We can't reach the highest element because of the averaging.
3269 # So if one below the upper edge, force it there and try one
3270 # more time.
3271 if ($i == $range_list_size - 2) {
3272
3273 trace "Forcing to upper edge" if main::DEBUG && $to_trace;
3274 $i = $range_list_size - 1;
3275
3276 # Change $lower as well so if fails next time through,
3277 # taking the average will yield the same $i, and we will
3278 # quit with the error message just below.
3279 $lower = $i;
3280 next;
3281 }
3282 Carp::my_carp_bug("$owner_name_of{$addr}Can't find where the range ought to go. No action taken.");
3283 return;
3284 }
3285 $i = $temp;
3286 } # End of while loop
3287
3288 if (main::DEBUG && $to_trace) {
3289 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i;
3290 trace "i= [ $i ]", $r->[$i];
3291 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < $range_list_size - 1;
3292 }
3293
3294 # Here we have found the offset. Cache it as a starting point for the
3295 # next call.
3296 $_search_ranges_cache{$addr} = $i;
3297 return $i;
3298 }
3299
3300 sub _add_delete {
3301 # Add, replace or delete ranges to or from a list. The $type
3302 # parameter gives which:
3303 # '+' => insert or replace a range, returning a list of any changed
3304 # ranges.
3305 # '-' => delete a range, returning a list of any deleted ranges.
3306 #
3307 # The next three parameters give respectively the start, end, and
3308 # value associated with the range. 'value' should be null unless the
3309 # operation is '+';
3310 #
3311 # The range list is kept sorted so that the range with the lowest
3312 # starting position is first in the list, and generally, adjacent
c1739a4a 3313 # ranges with the same values are merged into a single larger one (see
99870f4d
KW
3314 # exceptions below).
3315 #
c1739a4a 3316 # There are more parameters; all are key => value pairs:
99870f4d
KW
3317 # Type gives the type of the value. It is only valid for '+'.
3318 # All ranges have types; if this parameter is omitted, 0 is
3319 # assumed. Ranges with type 0 are assumed to obey the
3320 # Unicode rules for casing, etc; ranges with other types are
3321 # not. Otherwise, the type is arbitrary, for the caller's
3322 # convenience, and looked at only by this routine to keep
3323 # adjacent ranges of different types from being merged into
3324 # a single larger range, and when Replace =>
3325 # $IF_NOT_EQUIVALENT is specified (see just below).
3326 # Replace determines what to do if the range list already contains
3327 # ranges which coincide with all or portions of the input
3328 # range. It is only valid for '+':
3329 # => $NO means that the new value is not to replace
3330 # any existing ones, but any empty gaps of the
3331 # range list coinciding with the input range
3332 # will be filled in with the new value.
3333 # => $UNCONDITIONALLY means to replace the existing values with
3334 # this one unconditionally. However, if the
3335 # new and old values are identical, the
3336 # replacement is skipped to save cycles
3337 # => $IF_NOT_EQUIVALENT means to replace the existing values
3338 # with this one if they are not equivalent.
3339 # Ranges are equivalent if their types are the
c1739a4a 3340 # same, and they are the same string; or if
99870f4d
KW
3341 # both are type 0 ranges, if their Unicode
3342 # standard forms are identical. In this last
3343 # case, the routine chooses the more "modern"
3344 # one to use. This is because some of the
3345 # older files are formatted with values that
3346 # are, for example, ALL CAPs, whereas the
3347 # derived files have a more modern style,
3348 # which looks better. By looking for this
3349 # style when the pre-existing and replacement
3350 # standard forms are the same, we can move to
3351 # the modern style
9470941f 3352 # => $MULTIPLE_BEFORE means that if this range duplicates an
99870f4d
KW
3353 # existing one, but has a different value,
3354 # don't replace the existing one, but insert
3355 # this, one so that the same range can occur
53d84487
KW
3356 # multiple times. They are stored LIFO, so
3357 # that the final one inserted is the first one
3358 # returned in an ordered search of the table.
99870f4d
KW
3359 # => anything else is the same as => $IF_NOT_EQUIVALENT
3360 #
c1739a4a
KW
3361 # "same value" means identical for non-type-0 ranges, and it means
3362 # having the same standard forms for type-0 ranges.
99870f4d
KW
3363
3364 return Carp::carp_too_few_args(\@_, 5) if main::DEBUG && @_ < 5;
3365
3366 my $self = shift;
3367 my $operation = shift; # '+' for add/replace; '-' for delete;
3368 my $start = shift;
3369 my $end = shift;
3370 my $value = shift;
3371
3372 my %args = @_;
3373
3374 $value = "" if not defined $value; # warning: $value can be "0"
3375
3376 my $replace = delete $args{'Replace'};
3377 $replace = $IF_NOT_EQUIVALENT unless defined $replace;
3378
3379 my $type = delete $args{'Type'};
3380 $type = 0 unless defined $type;
3381
3382 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3383
ffe43484 3384 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3385
3386 if ($operation ne '+' && $operation ne '-') {
3387 Carp::my_carp_bug("$owner_name_of{$addr}First parameter to _add_delete must be '+' or '-'. No action taken.");
3388 return;
3389 }
3390 unless (defined $start && defined $end) {
3391 Carp::my_carp_bug("$owner_name_of{$addr}Undefined start and/or end to _add_delete. No action taken.");
3392 return;
3393 }
3394 unless ($end >= $start) {
3395 Carp::my_carp_bug("$owner_name_of{$addr}End of range (" . sprintf("%04X", $end) . ") must not be before start (" . sprintf("%04X", $start) . "). No action taken.");
3396 return;
3397 }
3398 #local $to_trace = 1 if main::DEBUG;
3399
3400 if ($operation eq '-') {
3401 if ($replace != $IF_NOT_EQUIVALENT) {
3402 Carp::my_carp_bug("$owner_name_of{$addr}Replace => \$IF_NOT_EQUIVALENT is required when deleting a range from a range list. Assuming Replace => \$IF_NOT_EQUIVALENT.");
3403 $replace = $IF_NOT_EQUIVALENT;
3404 }
3405 if ($type) {
3406 Carp::my_carp_bug("$owner_name_of{$addr}Type => 0 is required when deleting a range from a range list. Assuming Type => 0.");
3407 $type = 0;
3408 }
3409 if ($value ne "") {
3410 Carp::my_carp_bug("$owner_name_of{$addr}Value => \"\" is required when deleting a range from a range list. Assuming Value => \"\".");
3411 $value = "";
3412 }
3413 }
3414
3415 my $r = $ranges{$addr}; # The current list of ranges
3416 my $range_list_size = scalar @$r; # And its size
3417 my $max = $max{$addr}; # The current high code point in
3418 # the list of ranges
3419
3420 # Do a special case requiring fewer machine cycles when the new range
3421 # starts after the current highest point. The Unicode input data is
3422 # structured so this is common.
3423 if ($start > $max) {
3424
3425 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) type=$type" if main::DEBUG && $to_trace;
3426 return if $operation eq '-'; # Deleting a non-existing range is a
3427 # no-op
3428
3429 # If the new range doesn't logically extend the current final one
3430 # in the range list, create a new range at the end of the range
3431 # list. (max cleverly is initialized to a negative number not
3432 # adjacent to 0 if the range list is empty, so even adding a range
3433 # to an empty range list starting at 0 will have this 'if'
3434 # succeed.)
3435 if ($start > $max + 1 # non-adjacent means can't extend.
3436 || @{$r}[-1]->value ne $value # values differ, can't extend.
3437 || @{$r}[-1]->type != $type # types differ, can't extend.
3438 ) {
3439 push @$r, Range->new($start, $end,
3440 Value => $value,
3441 Type => $type);
3442 }
3443 else {
3444
3445 # Here, the new range starts just after the current highest in
3446 # the range list, and they have the same type and value.
3447 # Extend the current range to incorporate the new one.
3448 @{$r}[-1]->set_end($end);
3449 }
3450
3451 # This becomes the new maximum.
3452 $max{$addr} = $end;
3453
3454 return;
3455 }
3456 #local $to_trace = 0 if main::DEBUG;
3457
3458 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) replace=$replace" if main::DEBUG && $to_trace;
3459
3460 # Here, the input range isn't after the whole rest of the range list.
3461 # Most likely 'splice' will be needed. The rest of the routine finds
3462 # the needed splice parameters, and if necessary, does the splice.
3463 # First, find the offset parameter needed by the splice function for
3464 # the input range. Note that the input range may span multiple
3465 # existing ones, but we'll worry about that later. For now, just find
3466 # the beginning. If the input range is to be inserted starting in a
3467 # position not currently in the range list, it must (obviously) come
3468 # just after the range below it, and just before the range above it.
3469 # Slightly less obviously, it will occupy the position currently
3470 # occupied by the range that is to come after it. More formally, we
3471 # are looking for the position, $i, in the array of ranges, such that:
3472 #
3473 # r[$i-1]->start <= r[$i-1]->end < $start < r[$i]->start <= r[$i]->end
3474 #
3475 # (The ordered relationships within existing ranges are also shown in
3476 # the equation above). However, if the start of the input range is
3477 # within an existing range, the splice offset should point to that
3478 # existing range's position in the list; that is $i satisfies a
3479 # somewhat different equation, namely:
3480 #
3481 #r[$i-1]->start <= r[$i-1]->end < r[$i]->start <= $start <= r[$i]->end
3482 #
3483 # More briefly, $start can come before or after r[$i]->start, and at
3484 # this point, we don't know which it will be. However, these
3485 # two equations share these constraints:
3486 #
3487 # r[$i-1]->end < $start <= r[$i]->end
3488 #
3489 # And that is good enough to find $i.
3490
3491 my $i = $self->_search_ranges($start);
3492 if (! defined $i) {
3493 Carp::my_carp_bug("Searching $self for range beginning with $start unexpectedly returned undefined. Operation '$operation' not performed");
3494 return;
3495 }
3496
3497 # The search function returns $i such that:
3498 #
3499 # r[$i-1]->end < $start <= r[$i]->end
3500 #
3501 # That means that $i points to the first range in the range list
3502 # that could possibly be affected by this operation. We still don't
3503 # know if the start of the input range is within r[$i], or if it
3504 # points to empty space between r[$i-1] and r[$i].
3505 trace "[$i] is the beginning splice point. Existing range there is ", $r->[$i] if main::DEBUG && $to_trace;
3506
3507 # Special case the insertion of data that is not to replace any
3508 # existing data.
3509 if ($replace == $NO) { # If $NO, has to be operation '+'
3510 #local $to_trace = 1 if main::DEBUG;
3511 trace "Doesn't replace" if main::DEBUG && $to_trace;
3512
3513 # Here, the new range is to take effect only on those code points
3514 # that aren't already in an existing range. This can be done by
3515 # looking through the existing range list and finding the gaps in
3516 # the ranges that this new range affects, and then calling this
3517 # function recursively on each of those gaps, leaving untouched
3518 # anything already in the list. Gather up a list of the changed
3519 # gaps first so that changes to the internal state as new ranges
3520 # are added won't be a problem.
3521 my @gap_list;
3522
3523 # First, if the starting point of the input range is outside an
3524 # existing one, there is a gap from there to the beginning of the
3525 # existing range -- add a span to fill the part that this new
3526 # range occupies
3527 if ($start < $r->[$i]->start) {
3528 push @gap_list, Range->new($start,
3529 main::min($end,
3530 $r->[$i]->start - 1),
3531 Type => $type);
3532 trace "gap before $r->[$i] [$i], will add", $gap_list[-1] if main::DEBUG && $to_trace;
3533 }
3534
3535 # Then look through the range list for other gaps until we reach
3536 # the highest range affected by the input one.
3537 my $j;
3538 for ($j = $i+1; $j < $range_list_size; $j++) {
3539 trace "j=[$j]", $r->[$j] if main::DEBUG && $to_trace;
3540 last if $end < $r->[$j]->start;
3541
3542 # If there is a gap between when this range starts and the
3543 # previous one ends, add a span to fill it. Note that just
3544 # because there are two ranges doesn't mean there is a
3545 # non-zero gap between them. It could be that they have
3546 # different values or types
3547 if ($r->[$j-1]->end + 1 != $r->[$j]->start) {
3548 push @gap_list,
3549 Range->new($r->[$j-1]->end + 1,
3550 $r->[$j]->start - 1,
3551 Type => $type);
3552 trace "gap between $r->[$j-1] and $r->[$j] [$j], will add: $gap_list[-1]" if main::DEBUG && $to_trace;
3553 }
3554 }
3555
3556 # Here, we have either found an existing range in the range list,
3557 # beyond the area affected by the input one, or we fell off the
3558 # end of the loop because the input range affects the whole rest
3559 # of the range list. In either case, $j is 1 higher than the
3560 # highest affected range. If $j == $i, it means that there are no
3561 # affected ranges, that the entire insertion is in the gap between
3562 # r[$i-1], and r[$i], which we already have taken care of before
3563 # the loop.
3564 # On the other hand, if there are affected ranges, it might be
3565 # that there is a gap that needs filling after the final such
3566 # range to the end of the input range
3567 if ($r->[$j-1]->end < $end) {
3568 push @gap_list, Range->new(main::max($start,
3569 $r->[$j-1]->end + 1),
3570 $end,
3571 Type => $type);
3572 trace "gap after $r->[$j-1], will add $gap_list[-1]" if main::DEBUG && $to_trace;
3573 }
3574
3575 # Call recursively to fill in all the gaps.
3576 foreach my $gap (@gap_list) {
3577 $self->_add_delete($operation,
3578 $gap->start,
3579 $gap->end,
3580 $value,
3581 Type => $type);
3582 }
3583
3584 return;
3585 }
3586
53d84487
KW
3587 # Here, we have taken care of the case where $replace is $NO.
3588 # Remember that here, r[$i-1]->end < $start <= r[$i]->end
3589 # If inserting a multiple record, this is where it goes, before the
3590 # first (if any) existing one. This implies an insertion, and no
3591 # change to any existing ranges. Note that $i can be -1 if this new
3592 # range doesn't actually duplicate any existing, and comes at the
3593 # beginning of the list.
9470941f 3594 if ($replace == $MULTIPLE_BEFORE) {
53d84487
KW
3595
3596 if ($start != $end) {
3597 Carp::my_carp_bug("$owner_name_of{$addr}Can't cope with adding a multiple record when the range ($start..$end) contains more than one code point. No action taken.");
3598 return;
3599 }
3600
19155fcc 3601 # If the new code point is within a current range ...
53d84487 3602 if ($end >= $r->[$i]->start) {
19155fcc
KW
3603
3604 # Don't add an exact duplicate, as it isn't really a multiple
1f6798c4
KW
3605 my $existing_value = $r->[$i]->value;
3606 my $existing_type = $r->[$i]->type;
3607 return if $value eq $existing_value && $type eq $existing_type;
3608
3609 # If the multiple value is part of an existing range, we want
3610 # to split up that range, so that only the single code point
3611 # is affected. To do this, we first call ourselves
3612 # recursively to delete that code point from the table, having
3613 # preserved its current data above. Then we call ourselves
3614 # recursively again to add the new multiple, which we know by
3615 # the test just above is different than the current code
3616 # point's value, so it will become a range containing a single
3617 # code point: just itself. Finally, we add back in the
3618 # pre-existing code point, which will again be a single code
3619 # point range. Because 'i' likely will have changed as a
3620 # result of these operations, we can't just continue on, but
3621 # do this operation recursively as well.
53d84487 3622 if ($r->[$i]->start != $r->[$i]->end) {
1f6798c4
KW
3623 $self->_add_delete('-', $start, $end, "");
3624 $self->_add_delete('+', $start, $end, $value, Type => $type);
9470941f 3625 return $self->_add_delete('+', $start, $end, $existing_value, Type => $existing_type, Replace => $MULTIPLE_BEFORE);
53d84487 3626 }
53d84487
KW
3627 }
3628
3629 trace "Adding multiple record at $i with $start..$end, $value" if main::DEBUG && $to_trace;
3630 my @return = splice @$r,
3631 $i,
3632 0,
3633 Range->new($start,
3634 $end,
3635 Value => $value,
3636 Type => $type);
3637 if (main::DEBUG && $to_trace) {
3638 trace "After splice:";
3639 trace 'i-2=[', $i-2, ']', $r->[$i-2] if $i >= 2;
3640 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i >= 1;
3641 trace "i =[", $i, "]", $r->[$i] if $i >= 0;
3642 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < @$r - 1;
3643 trace 'i+2=[', $i+2, ']', $r->[$i+2] if $i < @$r - 2;
3644 trace 'i+3=[', $i+3, ']', $r->[$i+3] if $i < @$r - 3;
3645 }
3646 return @return;
3647 }
3648
3649 # Here, we have taken care of $NO and $MULTIPLE replaces. This leaves
3650 # delete, insert, and replace either unconditionally or if not
3651 # equivalent. $i still points to the first potential affected range.
3652 # Now find the highest range affected, which will determine the length
3653 # parameter to splice. (The input range can span multiple existing
3654 # ones.) If this isn't a deletion, while we are looking through the
3655 # range list, see also if this is a replacement rather than a clean
3656 # insertion; that is if it will change the values of at least one
3657 # existing range. Start off assuming it is an insert, until find it
3658 # isn't.
3659 my $clean_insert = $operation eq '+';
99870f4d
KW
3660 my $j; # This will point to the highest affected range
3661
3662 # For non-zero types, the standard form is the value itself;
3663 my $standard_form = ($type) ? $value : main::standardize($value);
3664
3665 for ($j = $i; $j < $range_list_size; $j++) {
3666 trace "Looking for highest affected range; the one at $j is ", $r->[$j] if main::DEBUG && $to_trace;
3667
3668 # If find a range that it doesn't overlap into, we can stop
3669 # searching
3670 last if $end < $r->[$j]->start;
3671
969a34cc
KW
3672 # Here, overlaps the range at $j. If the values don't match,
3673 # and so far we think this is a clean insertion, it becomes a
3674 # non-clean insertion, i.e., a 'change' or 'replace' instead.
3675 if ($clean_insert) {
99870f4d 3676 if ($r->[$j]->standard_form ne $standard_form) {
969a34cc 3677 $clean_insert = 0;
56343c78
KW
3678 if ($replace == $CROAK) {
3679 main::croak("The range to add "
3680 . sprintf("%04X", $start)
3681 . '-'
3682 . sprintf("%04X", $end)
3683 . " with value '$value' overlaps an existing range $r->[$j]");
3684 }
99870f4d
KW
3685 }
3686 else {
3687
3688 # Here, the two values are essentially the same. If the
3689 # two are actually identical, replacing wouldn't change
3690 # anything so skip it.
3691 my $pre_existing = $r->[$j]->value;
3692 if ($pre_existing ne $value) {
3693
3694 # Here the new and old standardized values are the
3695 # same, but the non-standardized values aren't. If
3696 # replacing unconditionally, then replace
3697 if( $replace == $UNCONDITIONALLY) {
969a34cc 3698 $clean_insert = 0;
99870f4d
KW
3699 }
3700 else {
3701
3702 # Here, are replacing conditionally. Decide to
3703 # replace or not based on which appears to look
3704 # the "nicest". If one is mixed case and the
3705 # other isn't, choose the mixed case one.
3706 my $new_mixed = $value =~ /[A-Z]/
3707 && $value =~ /[a-z]/;
3708 my $old_mixed = $pre_existing =~ /[A-Z]/
3709 && $pre_existing =~ /[a-z]/;
3710
3711 if ($old_mixed != $new_mixed) {
969a34cc 3712 $clean_insert = 0 if $new_mixed;
99870f4d 3713 if (main::DEBUG && $to_trace) {
969a34cc
KW
3714 if ($clean_insert) {
3715 trace "Retaining $pre_existing over $value";
99870f4d
KW
3716 }
3717 else {
969a34cc 3718 trace "Replacing $pre_existing with $value";
99870f4d
KW
3719 }
3720 }
3721 }
3722 else {
3723
3724 # Here casing wasn't different between the two.
3725 # If one has hyphens or underscores and the
3726 # other doesn't, choose the one with the
3727 # punctuation.
3728 my $new_punct = $value =~ /[-_]/;
3729 my $old_punct = $pre_existing =~ /[-_]/;
3730
3731 if ($old_punct != $new_punct) {
969a34cc 3732 $clean_insert = 0 if $new_punct;
99870f4d 3733 if (main::DEBUG && $to_trace) {
969a34cc
KW
3734 if ($clean_insert) {
3735 trace "Retaining $pre_existing over $value";
99870f4d
KW
3736 }
3737 else {
969a34cc 3738 trace "Replacing $pre_existing with $value";
99870f4d
KW
3739 }
3740 }
3741 } # else existing one is just as "good";
3742 # retain it to save cycles.
3743 }
3744 }
3745 }
3746 }
3747 }
3748 } # End of loop looking for highest affected range.
3749
3750 # Here, $j points to one beyond the highest range that this insertion
3751 # affects (hence to beyond the range list if that range is the final
3752 # one in the range list).
3753
3754 # The splice length is all the affected ranges. Get it before
3755 # subtracting, for efficiency, so we don't have to later add 1.
3756 my $length = $j - $i;
3757
3758 $j--; # $j now points to the highest affected range.
3759 trace "Final affected range is $j: $r->[$j]" if main::DEBUG && $to_trace;
3760
99870f4d
KW
3761 # Here, have taken care of $NO and $MULTIPLE replaces.
3762 # $j points to the highest affected range. But it can be < $i or even
3763 # -1. These happen only if the insertion is entirely in the gap
3764 # between r[$i-1] and r[$i]. Here's why: j < i means that the j loop
3765 # above exited first time through with $end < $r->[$i]->start. (And
3766 # then we subtracted one from j) This implies also that $start <
3767 # $r->[$i]->start, but we know from above that $r->[$i-1]->end <
3768 # $start, so the entire input range is in the gap.
3769 if ($j < $i) {
3770
3771 # Here the entire input range is in the gap before $i.
3772
3773 if (main::DEBUG && $to_trace) {
3774 if ($i) {
3775 trace "Entire range is between $r->[$i-1] and $r->[$i]";
3776 }
3777 else {
3778 trace "Entire range is before $r->[$i]";
3779 }
3780 }
3781 return if $operation ne '+'; # Deletion of a non-existent range is
3782 # a no-op
3783 }
3784 else {
3785
969a34cc
KW
3786 # Here part of the input range is not in the gap before $i. Thus,
3787 # there is at least one affected one, and $j points to the highest
3788 # such one.
99870f4d
KW
3789
3790 # At this point, here is the situation:
3791 # This is not an insertion of a multiple, nor of tentative ($NO)
3792 # data.
3793 # $i points to the first element in the current range list that
3794 # may be affected by this operation. In fact, we know
3795 # that the range at $i is affected because we are in
3796 # the else branch of this 'if'
3797 # $j points to the highest affected range.
3798 # In other words,
3799 # r[$i-1]->end < $start <= r[$i]->end
3800 # And:
3801 # r[$i-1]->end < $start <= $end <= r[$j]->end
3802 #
3803 # Also:
969a34cc
KW
3804 # $clean_insert is a boolean which is set true if and only if
3805 # this is a "clean insertion", i.e., not a change nor a
3806 # deletion (multiple was handled above).
99870f4d
KW
3807
3808 # We now have enough information to decide if this call is a no-op
969a34cc
KW
3809 # or not. It is a no-op if this is an insertion of already
3810 # existing data.
99870f4d 3811
969a34cc 3812 if (main::DEBUG && $to_trace && $clean_insert
99870f4d
KW
3813 && $i == $j
3814 && $start >= $r->[$i]->start)
3815 {
3816 trace "no-op";
3817 }
969a34cc 3818 return if $clean_insert
99870f4d
KW
3819 && $i == $j # more than one affected range => not no-op
3820
3821 # Here, r[$i-1]->end < $start <= $end <= r[$i]->end
3822 # Further, $start and/or $end is >= r[$i]->start
3823 # The test below hence guarantees that
3824 # r[$i]->start < $start <= $end <= r[$i]->end
3825 # This means the input range is contained entirely in
3826 # the one at $i, so is a no-op
3827 && $start >= $r->[$i]->start;
3828 }
3829
3830 # Here, we know that some action will have to be taken. We have
3831 # calculated the offset and length (though adjustments may be needed)
3832 # for the splice. Now start constructing the replacement list.
3833 my @replacement;
3834 my $splice_start = $i;
3835
3836 my $extends_below;
3837 my $extends_above;
3838
3839 # See if should extend any adjacent ranges.
3840 if ($operation eq '-') { # Don't extend deletions
3841 $extends_below = $extends_above = 0;
3842 }
3843 else { # Here, should extend any adjacent ranges. See if there are
3844 # any.
3845 $extends_below = ($i > 0
3846 # can't extend unless adjacent
3847 && $r->[$i-1]->end == $start -1
3848 # can't extend unless are same standard value
3849 && $r->[$i-1]->standard_form eq $standard_form
3850 # can't extend unless share type
3851 && $r->[$i-1]->type == $type);
3852 $extends_above = ($j+1 < $range_list_size
3853 && $r->[$j+1]->start == $end +1
3854 && $r->[$j+1]->standard_form eq $standard_form
23822bda 3855 && $r->[$j+1]->type == $type);
99870f4d
KW
3856 }
3857<