This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
mktables: Change member name for clarity
[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' => $why_no_expand,
99870f4d
KW
891 );
892
893 %why_suppressed = (
5f7264c7 894 # There is a lib/unicore/Decomposition.pl (used by Normalize.pm) which
99870f4d
KW
895 # contains the same information, but without the algorithmically
896 # determinable Hangul syllables'. This file is not published, so it's
897 # existence is not noted in the comment.
e0b29447 898 'Decomposition_Mapping' => 'Accessible via Unicode::Normalize or Unicode::UCD::prop_invmap()',
99870f4d 899
5f8d1a89
KW
900 # Don't suppress ISO_Comment, as otherwise special handling is needed
901 # to differentiate between it and gc=c, which can be written as 'isc',
902 # which is the same characters as ISO_Comment's short name.
99870f4d 903
e0b29447
KW
904 'Name' => "Accessible via 'use charnames;' or Unicode::UCD::prop_invmap()",
905
906 'Simple_Case_Folding' => "$simple. Can access this through Unicode::UCD::casefold or Unicode::UCD::prop_invmap()",
907 'Simple_Lowercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
908 'Simple_Titlecase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
909 'Simple_Uppercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
99870f4d 910
5f7264c7 911 FC_NFKC_Closure => 'Supplanted in usage by NFKC_Casefold; otherwise not useful',
99870f4d
KW
912 );
913
914 # The following are suppressed because they were made contributory or
915 # deprecated by Unicode before Perl ever thought about supporting them.
e22aaf5c
KW
916 foreach my $property ('Jamo_Short_Name',
917 'Grapheme_Link',
918 'Expands_On_NFC',
919 'Expands_On_NFD',
920 'Expands_On_NFKC',
921 'Expands_On_NFKD'
922 ) {
99870f4d
KW
923 $why_suppressed{$property} = $why_deprecated{$property};
924 }
cf25bb62 925
99870f4d
KW
926 # Customize the message for all the 'Other_' properties
927 foreach my $property (keys %why_deprecated) {
928 next if (my $main_property = $property) !~ s/^Other_//;
929 $why_deprecated{$property} =~ s/$other_properties/the $main_property property (which should be used instead)/;
930 }
931}
932
933if ($v_version ge 4.0.0) {
934 $why_stabilized{'Hyphen'} = 'Use the Line_Break property instead; see www.unicode.org/reports/tr14';
5f7264c7
KW
935 if ($v_version ge 6.0.0) {
936 $why_deprecated{'Hyphen'} = 'Supplanted by Line_Break property values; see www.unicode.org/reports/tr14';
937 }
99870f4d 938}
5f7264c7 939if ($v_version ge 5.2.0 && $v_version lt 6.0.0) {
99870f4d 940 $why_obsolete{'ISO_Comment'} = 'Code points for it have been removed';
5f7264c7 941 if ($v_version ge 6.0.0) {
63f74647 942 $why_deprecated{'ISO_Comment'} = 'No longer needed for Unicode\'s internal chart generation; otherwise not useful, and code points for it have been removed';
5f7264c7 943 }
99870f4d
KW
944}
945
946# Probably obsolete forever
947if ($v_version ge v4.1.0) {
82aed44a
KW
948 $why_suppressed{'Script=Katakana_Or_Hiragana'} = 'Obsolete. All code points previously matched by this have been moved to "Script=Common".';
949}
950if ($v_version ge v6.0.0) {
2b352efd
KW
951 $why_suppressed{'Script=Katakana_Or_Hiragana'} .= ' Consider instead using "Script_Extensions=Katakana" or "Script_Extensions=Hiragana (or both)"';
952 $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
953}
954
955# This program can create files for enumerated-like properties, such as
956# 'Numeric_Type'. This file would be the same format as for a string
957# property, with a mapping from code point to its value, so you could look up,
958# for example, the script a code point is in. But no one so far wants this
959# mapping, or they have found another way to get it since this is a new
960# feature. So no file is generated except if it is in this list.
961my @output_mapped_properties = split "\n", <<END;
962END
963
c12f2655
KW
964# If you are using the Unihan database in a Unicode version before 5.2, you
965# need to add the properties that you want to extract from it to this table.
966# For your convenience, the properties in the 6.0 PropertyAliases.txt file are
967# listed, commented out
99870f4d
KW
968my @cjk_properties = split "\n", <<'END';
969#cjkAccountingNumeric; kAccountingNumeric
970#cjkOtherNumeric; kOtherNumeric
971#cjkPrimaryNumeric; kPrimaryNumeric
972#cjkCompatibilityVariant; kCompatibilityVariant
973#cjkIICore ; kIICore
974#cjkIRG_GSource; kIRG_GSource
975#cjkIRG_HSource; kIRG_HSource
976#cjkIRG_JSource; kIRG_JSource
977#cjkIRG_KPSource; kIRG_KPSource
978#cjkIRG_KSource; kIRG_KSource
979#cjkIRG_TSource; kIRG_TSource
980#cjkIRG_USource; kIRG_USource
981#cjkIRG_VSource; kIRG_VSource
982#cjkRSUnicode; kRSUnicode ; Unicode_Radical_Stroke; URS
983END
984
985# Similarly for the property values. For your convenience, the lines in the
5f7264c7 986# 6.0 PropertyAliases.txt file are listed. Just remove the first BUT NOT both
c12f2655 987# '#' marks (for Unicode versions before 5.2)
99870f4d
KW
988my @cjk_property_values = split "\n", <<'END';
989## @missing: 0000..10FFFF; cjkAccountingNumeric; NaN
990## @missing: 0000..10FFFF; cjkCompatibilityVariant; <code point>
991## @missing: 0000..10FFFF; cjkIICore; <none>
992## @missing: 0000..10FFFF; cjkIRG_GSource; <none>
993## @missing: 0000..10FFFF; cjkIRG_HSource; <none>
994## @missing: 0000..10FFFF; cjkIRG_JSource; <none>
995## @missing: 0000..10FFFF; cjkIRG_KPSource; <none>
996## @missing: 0000..10FFFF; cjkIRG_KSource; <none>
997## @missing: 0000..10FFFF; cjkIRG_TSource; <none>
998## @missing: 0000..10FFFF; cjkIRG_USource; <none>
999## @missing: 0000..10FFFF; cjkIRG_VSource; <none>
1000## @missing: 0000..10FFFF; cjkOtherNumeric; NaN
1001## @missing: 0000..10FFFF; cjkPrimaryNumeric; NaN
1002## @missing: 0000..10FFFF; cjkRSUnicode; <none>
1003END
1004
1005# The input files don't list every code point. Those not listed are to be
1006# defaulted to some value. Below are hard-coded what those values are for
1007# non-binary properties as of 5.1. Starting in 5.0, there are
1008# machine-parsable comment lines in the files the give the defaults; so this
1009# list shouldn't have to be extended. The claim is that all missing entries
1010# for binary properties will default to 'N'. Unicode tried to change that in
1011# 5.2, but the beta period produced enough protest that they backed off.
1012#
1013# The defaults for the fields that appear in UnicodeData.txt in this hash must
1014# be in the form that it expects. The others may be synonyms.
1015my $CODE_POINT = '<code point>';
1016my %default_mapping = (
1017 Age => "Unassigned",
1018 # Bidi_Class => Complicated; set in code
1019 Bidi_Mirroring_Glyph => "",
1020 Block => 'No_Block',
1021 Canonical_Combining_Class => 0,
1022 Case_Folding => $CODE_POINT,
1023 Decomposition_Mapping => $CODE_POINT,
1024 Decomposition_Type => 'None',
1025 East_Asian_Width => "Neutral",
1026 FC_NFKC_Closure => $CODE_POINT,
1027 General_Category => 'Cn',
1028 Grapheme_Cluster_Break => 'Other',
1029 Hangul_Syllable_Type => 'NA',
1030 ISO_Comment => "",
1031 Jamo_Short_Name => "",
1032 Joining_Group => "No_Joining_Group",
1033 # Joining_Type => Complicated; set in code
1034 kIICore => 'N', # Is converted to binary
1035 #Line_Break => Complicated; set in code
1036 Lowercase_Mapping => $CODE_POINT,
1037 Name => "",
1038 Name_Alias => "",
1039 NFC_QC => 'Yes',
1040 NFD_QC => 'Yes',
1041 NFKC_QC => 'Yes',
1042 NFKD_QC => 'Yes',
1043 Numeric_Type => 'None',
1044 Numeric_Value => 'NaN',
1045 Script => ($v_version le 4.1.0) ? 'Common' : 'Unknown',
1046 Sentence_Break => 'Other',
1047 Simple_Case_Folding => $CODE_POINT,
1048 Simple_Lowercase_Mapping => $CODE_POINT,
1049 Simple_Titlecase_Mapping => $CODE_POINT,
1050 Simple_Uppercase_Mapping => $CODE_POINT,
1051 Titlecase_Mapping => $CODE_POINT,
1052 Unicode_1_Name => "",
1053 Unicode_Radical_Stroke => "",
1054 Uppercase_Mapping => $CODE_POINT,
1055 Word_Break => 'Other',
1056);
1057
1058# Below are files that Unicode furnishes, but this program ignores, and why
1059my %ignored_files = (
73ba1144
KW
1060 'CJKRadicals.txt' => 'Maps the kRSUnicode property values to corresponding code points',
1061 'Index.txt' => 'Alphabetical index of Unicode characters',
1062 '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',
1063 'NamesList.txt' => 'Annotated list of characters',
1064 'NormalizationCorrections.txt' => 'Documentation of corrections already incorporated into the Unicode data base',
1065 'Props.txt' => 'Only in very early releases; is a subset of F<PropList.txt> (which is used instead)',
1066 'ReadMe.txt' => 'Documentation',
1067 '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>',
1068 'EmojiSources.txt' => 'Maps certain Unicode code points to their legacy Japanese cell-phone values',
1069 'IndicMatraCategory.txt' => 'Provisional; for the analysis and processing of Indic scripts',
1070 'IndicSyllabicCategory.txt' => 'Provisional; for the analysis and processing of Indic scripts',
1071 'auxiliary/WordBreakTest.html' => 'Documentation of validation tests',
1072 'auxiliary/SentenceBreakTest.html' => 'Documentation of validation tests',
1073 'auxiliary/GraphemeBreakTest.html' => 'Documentation of validation tests',
1074 'auxiliary/LineBreakTest.html' => 'Documentation of validation tests',
99870f4d
KW
1075);
1076
678f13d5 1077### End of externally interesting definitions, except for @input_file_objects
99870f4d
KW
1078
1079my $HEADER=<<"EOF";
1080# !!!!!!! DO NOT EDIT THIS FILE !!!!!!!
3df51b85
KW
1081# This file is machine-generated by $0 from the Unicode
1082# database, Version $string_version. Any changes made here will be lost!
cf25bb62
JH
1083EOF
1084
126c3d4e 1085my $INTERNAL_ONLY_HEADER = <<"EOF";
99870f4d
KW
1086
1087# !!!!!!! INTERNAL PERL USE ONLY !!!!!!!
fac53429
KW
1088# This file is for internal use by core Perl only. The format and even the
1089# name or existence of this file are subject to change without notice. Don't
1090# use it directly.
99870f4d
KW
1091EOF
1092
1093my $DEVELOPMENT_ONLY=<<"EOF";
1094# !!!!!!! DEVELOPMENT USE ONLY !!!!!!!
1095# This file contains information artificially constrained to code points
1096# present in Unicode release $string_compare_versions.
1097# IT CANNOT BE RELIED ON. It is for use during development only and should
23e33b60 1098# not be used for production.
b6922eda
KW
1099
1100EOF
1101
6189eadc
KW
1102my $MAX_UNICODE_CODEPOINT_STRING = "10FFFF";
1103my $MAX_UNICODE_CODEPOINT = hex $MAX_UNICODE_CODEPOINT_STRING;
1104my $MAX_UNICODE_CODEPOINTS = $MAX_UNICODE_CODEPOINT + 1;
99870f4d
KW
1105
1106# Matches legal code point. 4-6 hex numbers, If there are 6, the first
1107# two must be 10; if there are 5, the first must not be a 0. Written this way
92199589
KW
1108# to decrease backtracking. The first regex allows the code point to be at
1109# the end of a word, but to work properly, the word shouldn't end with a valid
1110# hex character. The second one won't match a code point at the end of a
1111# word, and doesn't have the run-on issue
8c32d378
KW
1112my $run_on_code_point_re =
1113 qr/ (?: 10[0-9A-F]{4} | [1-9A-F][0-9A-F]{4} | [0-9A-F]{4} ) \b/x;
1114my $code_point_re = qr/\b$run_on_code_point_re/;
99870f4d
KW
1115
1116# This matches the beginning of the line in the Unicode db files that give the
1117# defaults for code points not listed (i.e., missing) in the file. The code
1118# depends on this ending with a semi-colon, so it can assume it is a valid
1119# field when the line is split() by semi-colons
1120my $missing_defaults_prefix =
6189eadc 1121 qr/^#\s+\@missing:\s+0000\.\.$MAX_UNICODE_CODEPOINT_STRING\s*;/;
99870f4d
KW
1122
1123# Property types. Unicode has more types, but these are sufficient for our
1124# purposes.
1125my $UNKNOWN = -1; # initialized to illegal value
1126my $NON_STRING = 1; # Either binary or enum
1127my $BINARY = 2;
06f26c45
KW
1128my $FORCED_BINARY = 3; # Not a binary property, but, besides its normal
1129 # tables, additional true and false tables are
1130 # generated so that false is anything matching the
1131 # default value, and true is everything else.
1132my $ENUM = 4; # Include catalog
1133my $STRING = 5; # Anything else: string or misc
99870f4d
KW
1134
1135# Some input files have lines that give default values for code points not
1136# contained in the file. Sometimes these should be ignored.
1137my $NO_DEFAULTS = 0; # Must evaluate to false
1138my $NOT_IGNORED = 1;
1139my $IGNORED = 2;
1140
1141# Range types. Each range has a type. Most ranges are type 0, for normal,
1142# and will appear in the main body of the tables in the output files, but
1143# there are other types of ranges as well, listed below, that are specially
1144# handled. There are pseudo-types as well that will never be stored as a
1145# type, but will affect the calculation of the type.
1146
1147# 0 is for normal, non-specials
1148my $MULTI_CP = 1; # Sequence of more than code point
1149my $HANGUL_SYLLABLE = 2;
1150my $CP_IN_NAME = 3; # The NAME contains the code point appended to it.
1151my $NULL = 4; # The map is to the null string; utf8.c can't
1152 # handle these, nor is there an accepted syntax
1153 # for them in \p{} constructs
f86864ac 1154my $COMPUTE_NO_MULTI_CP = 5; # Pseudo-type; means that ranges that would
99870f4d
KW
1155 # otherwise be $MULTI_CP type are instead type 0
1156
1157# process_generic_property_file() can accept certain overrides in its input.
1158# Each of these must begin AND end with $CMD_DELIM.
1159my $CMD_DELIM = "\a";
1160my $REPLACE_CMD = 'replace'; # Override the Replace
1161my $MAP_TYPE_CMD = 'map_type'; # Override the Type
1162
1163my $NO = 0;
1164my $YES = 1;
1165
1166# Values for the Replace argument to add_range.
1167# $NO # Don't replace; add only the code points not
1168 # already present.
1169my $IF_NOT_EQUIVALENT = 1; # Replace only under certain conditions; details in
1170 # the comments at the subroutine definition.
1171my $UNCONDITIONALLY = 2; # Replace without conditions.
1172my $MULTIPLE = 4; # Don't replace, but add a duplicate record if
1173 # already there
56343c78 1174my $CROAK = 5; # Die with an error if is already there
99870f4d
KW
1175
1176# Flags to give property statuses. The phrases are to remind maintainers that
1177# if the flag is changed, the indefinite article referring to it in the
1178# documentation may need to be as well.
1179my $NORMAL = "";
99870f4d
KW
1180my $DEPRECATED = 'D';
1181my $a_bold_deprecated = "a 'B<$DEPRECATED>'";
1182my $A_bold_deprecated = "A 'B<$DEPRECATED>'";
1183my $DISCOURAGED = 'X';
1184my $a_bold_discouraged = "an 'B<$DISCOURAGED>'";
1185my $A_bold_discouraged = "An 'B<$DISCOURAGED>'";
1186my $STRICTER = 'T';
1187my $a_bold_stricter = "a 'B<$STRICTER>'";
1188my $A_bold_stricter = "A 'B<$STRICTER>'";
1189my $STABILIZED = 'S';
1190my $a_bold_stabilized = "an 'B<$STABILIZED>'";
1191my $A_bold_stabilized = "An 'B<$STABILIZED>'";
1192my $OBSOLETE = 'O';
1193my $a_bold_obsolete = "an 'B<$OBSOLETE>'";
1194my $A_bold_obsolete = "An 'B<$OBSOLETE>'";
1195
1196my %status_past_participles = (
1197 $DISCOURAGED => 'discouraged',
99870f4d
KW
1198 $STABILIZED => 'stabilized',
1199 $OBSOLETE => 'obsolete',
37e2e78e 1200 $DEPRECATED => 'deprecated',
99870f4d
KW
1201);
1202
395dfc19
KW
1203# Table fates. These are somewhat ordered, so that fates < $MAP_PROXIED should be
1204# externally documented.
301ba948 1205my $ORDINARY = 0; # The normal fate.
395dfc19
KW
1206my $MAP_PROXIED = 1; # The map table for the property isn't written out,
1207 # but there is a file written that can be used to
1208 # reconstruct this table
301ba948
KW
1209my $SUPPRESSED = 3; # The file for this table is not written out.
1210my $INTERNAL_ONLY = 4; # The file for this table is written out, but it is
1211 # for Perl's internal use only
1212my $PLACEHOLDER = 5; # A property that is defined as a placeholder in a
1213 # Unicode version that doesn't have it, but we need it
1214 # to be defined, if empty, to have things work.
1215 # Implies no pod entry generated
1216
f5817e0a
KW
1217# The format of the values of the tables:
1218my $EMPTY_FORMAT = "";
99870f4d
KW
1219my $BINARY_FORMAT = 'b';
1220my $DECIMAL_FORMAT = 'd';
1221my $FLOAT_FORMAT = 'f';
1222my $INTEGER_FORMAT = 'i';
1223my $HEX_FORMAT = 'x';
1224my $RATIONAL_FORMAT = 'r';
1225my $STRING_FORMAT = 's';
a14f3cb1 1226my $DECOMP_STRING_FORMAT = 'c';
c3ff2976 1227my $STRING_WHITE_SPACE_LIST = 'sw';
99870f4d
KW
1228
1229my %map_table_formats = (
1230 $BINARY_FORMAT => 'binary',
1231 $DECIMAL_FORMAT => 'single decimal digit',
1232 $FLOAT_FORMAT => 'floating point number',
1233 $INTEGER_FORMAT => 'integer',
add63c13 1234 $HEX_FORMAT => 'non-negative hex whole number; a code point',
99870f4d 1235 $RATIONAL_FORMAT => 'rational: an integer or a fraction',
1a9d544b 1236 $STRING_FORMAT => 'string',
92f9d56c 1237 $DECOMP_STRING_FORMAT => 'Perl\'s internal (Normalize.pm) decomposition mapping',
c3ff2976 1238 $STRING_WHITE_SPACE_LIST => 'string, but some elements are interpreted as a list; white space occurs only as list item separators'
99870f4d
KW
1239);
1240
1241# Unicode didn't put such derived files in a separate directory at first.
1242my $EXTRACTED_DIR = (-d 'extracted') ? 'extracted' : "";
1243my $EXTRACTED = ($EXTRACTED_DIR) ? "$EXTRACTED_DIR/" : "";
1244my $AUXILIARY = 'auxiliary';
1245
1246# Hashes that will eventually go into Heavy.pl for the use of utf8_heavy.pl
9e4a1e86 1247# and into UCD.pl for the use of UCD.pm
99870f4d
KW
1248my %loose_to_file_of; # loosely maps table names to their respective
1249 # files
1250my %stricter_to_file_of; # same; but for stricter mapping.
315bfd4e 1251my %loose_property_to_file_of; # Maps a loose property name to its map file
89cf10cc
KW
1252my %file_to_swash_name; # Maps the file name to its corresponding key name
1253 # in the hash %utf8::SwashInfo
99870f4d
KW
1254my %nv_floating_to_rational; # maps numeric values floating point numbers to
1255 # their rational equivalent
c12f2655
KW
1256my %loose_property_name_of; # Loosely maps (non_string) property names to
1257 # standard form
86a52d1e 1258my %string_property_loose_to_name; # Same, for string properties.
c15fda25
KW
1259my %loose_defaults; # keys are of form "prop=value", where 'prop' is
1260 # the property name in standard loose form, and
1261 # 'value' is the default value for that property,
1262 # also in standard loose form.
9e4a1e86
KW
1263my %loose_to_standard_value; # loosely maps table names to the canonical
1264 # alias for them
2df7880f
KW
1265my %ambiguous_names; # keys are alias names (in standard form) that
1266 # have more than one possible meaning.
5d1df013
KW
1267my %prop_aliases; # Keys are standard property name; values are each
1268 # one's aliases
1e863613
KW
1269my %prop_value_aliases; # Keys of top level are standard property name;
1270 # values are keys to another hash, Each one is
1271 # one of the property's values, in standard form.
1272 # The values are that prop-val's aliases.
2df7880f 1273my %ucd_pod; # Holds entries that will go into the UCD section of the pod
99870f4d 1274
d867ccfb
KW
1275# Most properties are immune to caseless matching, otherwise you would get
1276# nonsensical results, as properties are a function of a code point, not
1277# everything that is caselessly equivalent to that code point. For example,
1278# Changes_When_Case_Folded('s') should be false, whereas caselessly it would
1279# be true because 's' and 'S' are equivalent caselessly. However,
1280# traditionally, [:upper:] and [:lower:] are equivalent caselessly, so we
1281# extend that concept to those very few properties that are like this. Each
1282# such property will match the full range caselessly. They are hard-coded in
1283# the program; it's not worth trying to make it general as it's extremely
1284# unlikely that they will ever change.
1285my %caseless_equivalent_to;
1286
99870f4d
KW
1287# These constants names and values were taken from the Unicode standard,
1288# version 5.1, section 3.12. They are used in conjunction with Hangul
6e5a209b
KW
1289# syllables. The '_string' versions are so generated tables can retain the
1290# hex format, which is the more familiar value
1291my $SBase_string = "0xAC00";
1292my $SBase = CORE::hex $SBase_string;
1293my $LBase_string = "0x1100";
1294my $LBase = CORE::hex $LBase_string;
1295my $VBase_string = "0x1161";
1296my $VBase = CORE::hex $VBase_string;
1297my $TBase_string = "0x11A7";
1298my $TBase = CORE::hex $TBase_string;
99870f4d
KW
1299my $SCount = 11172;
1300my $LCount = 19;
1301my $VCount = 21;
1302my $TCount = 28;
1303my $NCount = $VCount * $TCount;
1304
1305# For Hangul syllables; These store the numbers from Jamo.txt in conjunction
1306# with the above published constants.
1307my %Jamo;
1308my %Jamo_L; # Leading consonants
1309my %Jamo_V; # Vowels
1310my %Jamo_T; # Trailing consonants
1311
bb1dd3da
KW
1312# For code points whose name contains its ordinal as a '-ABCD' suffix.
1313# The key is the base name of the code point, and the value is an
1314# array giving all the ranges that use this base name. Each range
1315# is actually a hash giving the 'low' and 'high' values of it.
1316my %names_ending_in_code_point;
1317my %loose_names_ending_in_code_point; # Same as above, but has blanks, dashes
1318 # removed from the names
1319# Inverse mapping. The list of ranges that have these kinds of
1320# names. Each element contains the low, high, and base names in an
1321# anonymous hash.
1322my @code_points_ending_in_code_point;
1323
1324# Boolean: does this Unicode version have the hangul syllables, and are we
1325# writing out a table for them?
1326my $has_hangul_syllables = 0;
1327
1328# Does this Unicode version have code points whose names end in their
1329# respective code points, and are we writing out a table for them? 0 for no;
1330# otherwise points to first property that a table is needed for them, so that
1331# if multiple tables are needed, we don't create duplicates
1332my $needing_code_points_ending_in_code_point = 0;
1333
37e2e78e 1334my @backslash_X_tests; # List of tests read in for testing \X
99870f4d
KW
1335my @unhandled_properties; # Will contain a list of properties found in
1336 # the input that we didn't process.
f86864ac 1337my @match_properties; # Properties that have match tables, to be
99870f4d
KW
1338 # listed in the pod
1339my @map_properties; # Properties that get map files written
1340my @named_sequences; # NamedSequences.txt contents.
1341my %potential_files; # Generated list of all .txt files in the directory
1342 # structure so we can warn if something is being
1343 # ignored.
1344my @files_actually_output; # List of files we generated.
1345my @more_Names; # Some code point names are compound; this is used
1346 # to store the extra components of them.
1347my $MIN_FRACTION_LENGTH = 3; # How many digits of a floating point number at
1348 # the minimum before we consider it equivalent to a
1349 # candidate rational
1350my $MAX_FLOATING_SLOP = 10 ** - $MIN_FRACTION_LENGTH; # And in floating terms
1351
1352# These store references to certain commonly used property objects
1353my $gc;
1354my $perl;
1355my $block;
3e20195b
KW
1356my $perl_charname;
1357my $print;
7fc6cb55 1358my $Any;
359523e2 1359my $script;
99870f4d
KW
1360
1361# Are there conflicting names because of beginning with 'In_', or 'Is_'
1362my $has_In_conflicts = 0;
1363my $has_Is_conflicts = 0;
1364
1365sub internal_file_to_platform ($) {
1366 # Convert our file paths which have '/' separators to those of the
1367 # platform.
1368
1369 my $file = shift;
1370 return undef unless defined $file;
1371
1372 return File::Spec->join(split '/', $file);
d07a55ed 1373}
5beb625e 1374
99870f4d
KW
1375sub file_exists ($) { # platform independent '-e'. This program internally
1376 # uses slash as a path separator.
1377 my $file = shift;
1378 return 0 if ! defined $file;
1379 return -e internal_file_to_platform($file);
1380}
5beb625e 1381
99870f4d 1382sub objaddr($) {
23e33b60
KW
1383 # Returns the address of the blessed input object.
1384 # It doesn't check for blessedness because that would do a string eval
1385 # every call, and the program is structured so that this is never called
1386 # for a non-blessed object.
99870f4d 1387
23e33b60 1388 no overloading; # If overloaded, numifying below won't work.
99870f4d
KW
1389
1390 # Numifying a ref gives its address.
051df77b 1391 return pack 'J', $_[0];
99870f4d
KW
1392}
1393
558712cf 1394# These are used only if $annotate is true.
c4019d52
KW
1395# The entire range of Unicode characters is examined to populate these
1396# after all the input has been processed. But most can be skipped, as they
1397# have the same descriptive phrases, such as being unassigned
1398my @viacode; # Contains the 1 million character names
1399my @printable; # boolean: And are those characters printable?
1400my @annotate_char_type; # Contains a type of those characters, specifically
1401 # for the purposes of annotation.
1402my $annotate_ranges; # A map of ranges of code points that have the same
98dc9551 1403 # name for the purposes of annotation. They map to the
c4019d52
KW
1404 # upper edge of the range, so that the end point can
1405 # be immediately found. This is used to skip ahead to
1406 # the end of a range, and avoid processing each
1407 # individual code point in it.
1408my $unassigned_sans_noncharacters; # A Range_List of the unassigned
1409 # characters, but excluding those which are
1410 # also noncharacter code points
1411
1412# The annotation types are an extension of the regular range types, though
1413# some of the latter are folded into one. Make the new types negative to
1414# avoid conflicting with the regular types
1415my $SURROGATE_TYPE = -1;
1416my $UNASSIGNED_TYPE = -2;
1417my $PRIVATE_USE_TYPE = -3;
1418my $NONCHARACTER_TYPE = -4;
1419my $CONTROL_TYPE = -5;
1420my $UNKNOWN_TYPE = -6; # Used only if there is a bug in this program
1421
1422sub populate_char_info ($) {
558712cf 1423 # Used only with the $annotate option. Populates the arrays with the
c4019d52
KW
1424 # input code point's info that are needed for outputting more detailed
1425 # comments. If calling context wants a return, it is the end point of
1426 # any contiguous range of characters that share essentially the same info
1427
1428 my $i = shift;
1429 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1430
1431 $viacode[$i] = $perl_charname->value_of($i) || "";
1432
1433 # A character is generally printable if Unicode says it is,
1434 # but below we make sure that most Unicode general category 'C' types
1435 # aren't.
1436 $printable[$i] = $print->contains($i);
1437
1438 $annotate_char_type[$i] = $perl_charname->type_of($i) || 0;
1439
1440 # Only these two regular types are treated specially for annotations
1441 # purposes
1442 $annotate_char_type[$i] = 0 if $annotate_char_type[$i] != $CP_IN_NAME
1443 && $annotate_char_type[$i] != $HANGUL_SYLLABLE;
1444
1445 # Give a generic name to all code points that don't have a real name.
1446 # We output ranges, if applicable, for these. Also calculate the end
1447 # point of the range.
1448 my $end;
1449 if (! $viacode[$i]) {
1450 if ($gc-> table('Surrogate')->contains($i)) {
1451 $viacode[$i] = 'Surrogate';
1452 $annotate_char_type[$i] = $SURROGATE_TYPE;
1453 $printable[$i] = 0;
1454 $end = $gc->table('Surrogate')->containing_range($i)->end;
1455 }
1456 elsif ($gc-> table('Private_use')->contains($i)) {
1457 $viacode[$i] = 'Private Use';
1458 $annotate_char_type[$i] = $PRIVATE_USE_TYPE;
1459 $printable[$i] = 0;
1460 $end = $gc->table('Private_Use')->containing_range($i)->end;
1461 }
1462 elsif (Property::property_ref('Noncharacter_Code_Point')-> table('Y')->
1463 contains($i))
1464 {
1465 $viacode[$i] = 'Noncharacter';
1466 $annotate_char_type[$i] = $NONCHARACTER_TYPE;
1467 $printable[$i] = 0;
1468 $end = property_ref('Noncharacter_Code_Point')->table('Y')->
1469 containing_range($i)->end;
1470 }
1471 elsif ($gc-> table('Control')->contains($i)) {
1472 $viacode[$i] = 'Control';
1473 $annotate_char_type[$i] = $CONTROL_TYPE;
1474 $printable[$i] = 0;
1475 $end = 0x81 if $i == 0x80; # Hard-code this one known case
1476 }
1477 elsif ($gc-> table('Unassigned')->contains($i)) {
1478 $viacode[$i] = 'Unassigned, block=' . $block-> value_of($i);
1479 $annotate_char_type[$i] = $UNASSIGNED_TYPE;
1480 $printable[$i] = 0;
1481
1482 # Because we name the unassigned by the blocks they are in, it
1483 # can't go past the end of that block, and it also can't go past
1484 # the unassigned range it is in. The special table makes sure
1485 # that the non-characters, which are unassigned, are separated
1486 # out.
1487 $end = min($block->containing_range($i)->end,
1488 $unassigned_sans_noncharacters-> containing_range($i)->
1489 end);
13ca76ff
KW
1490 }
1491 else {
1492 Carp::my_carp_bug("Can't figure out how to annotate "
1493 . sprintf("U+%04X", $i)
1494 . ". Proceeding anyway.");
c4019d52
KW
1495 $viacode[$i] = 'UNKNOWN';
1496 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1497 $printable[$i] = 0;
1498 }
1499 }
1500
1501 # Here, has a name, but if it's one in which the code point number is
1502 # appended to the name, do that.
1503 elsif ($annotate_char_type[$i] == $CP_IN_NAME) {
1504 $viacode[$i] .= sprintf("-%04X", $i);
1505 $end = $perl_charname->containing_range($i)->end;
1506 }
1507
1508 # And here, has a name, but if it's a hangul syllable one, replace it with
1509 # the correct name from the Unicode algorithm
1510 elsif ($annotate_char_type[$i] == $HANGUL_SYLLABLE) {
1511 use integer;
1512 my $SIndex = $i - $SBase;
1513 my $L = $LBase + $SIndex / $NCount;
1514 my $V = $VBase + ($SIndex % $NCount) / $TCount;
1515 my $T = $TBase + $SIndex % $TCount;
1516 $viacode[$i] = "HANGUL SYLLABLE $Jamo{$L}$Jamo{$V}";
1517 $viacode[$i] .= $Jamo{$T} if $T != $TBase;
1518 $end = $perl_charname->containing_range($i)->end;
1519 }
1520
1521 return if ! defined wantarray;
1522 return $i if ! defined $end; # If not a range, return the input
1523
1524 # Save this whole range so can find the end point quickly
1525 $annotate_ranges->add_map($i, $end, $end);
1526
1527 return $end;
1528}
1529
23e33b60
KW
1530# Commented code below should work on Perl 5.8.
1531## This 'require' doesn't necessarily work in miniperl, and even if it does,
1532## the native perl version of it (which is what would operate under miniperl)
1533## is extremely slow, as it does a string eval every call.
1534#my $has_fast_scalar_util = $\18 !~ /miniperl/
1535# && defined eval "require Scalar::Util";
1536#
1537#sub objaddr($) {
1538# # Returns the address of the blessed input object. Uses the XS version if
1539# # available. It doesn't check for blessedness because that would do a
1540# # string eval every call, and the program is structured so that this is
1541# # never called for a non-blessed object.
1542#
1543# return Scalar::Util::refaddr($_[0]) if $has_fast_scalar_util;
1544#
1545# # Check at least that is a ref.
1546# my $pkg = ref($_[0]) or return undef;
1547#
1548# # Change to a fake package to defeat any overloaded stringify
1549# bless $_[0], 'main::Fake';
1550#
1551# # Numifying a ref gives its address.
051df77b 1552# my $addr = pack 'J', $_[0];
23e33b60
KW
1553#
1554# # Return to original class
1555# bless $_[0], $pkg;
1556# return $addr;
1557#}
1558
99870f4d
KW
1559sub max ($$) {
1560 my $a = shift;
1561 my $b = shift;
1562 return $a if $a >= $b;
1563 return $b;
1564}
1565
1566sub min ($$) {
1567 my $a = shift;
1568 my $b = shift;
1569 return $a if $a <= $b;
1570 return $b;
1571}
1572
1573sub clarify_number ($) {
1574 # This returns the input number with underscores inserted every 3 digits
1575 # in large (5 digits or more) numbers. Input must be entirely digits, not
1576 # checked.
1577
1578 my $number = shift;
1579 my $pos = length($number) - 3;
1580 return $number if $pos <= 1;
1581 while ($pos > 0) {
1582 substr($number, $pos, 0) = '_';
1583 $pos -= 3;
5beb625e 1584 }
99870f4d 1585 return $number;
99598c8c
JH
1586}
1587
12ac2576 1588
99870f4d 1589package Carp;
7ebf06b3 1590
99870f4d
KW
1591# These routines give a uniform treatment of messages in this program. They
1592# are placed in the Carp package to cause the stack trace to not include them,
1593# although an alternative would be to use another package and set @CARP_NOT
1594# for it.
12ac2576 1595
99870f4d 1596our $Verbose = 1 if main::DEBUG; # Useful info when debugging
12ac2576 1597
99f78760
KW
1598# This is a work-around suggested by Nicholas Clark to fix a problem with Carp
1599# and overload trying to load Scalar:Util under miniperl. See
1600# http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2009-11/msg01057.html
1601undef $overload::VERSION;
1602
99870f4d
KW
1603sub my_carp {
1604 my $message = shift || "";
1605 my $nofold = shift || 0;
7ebf06b3 1606
99870f4d
KW
1607 if ($message) {
1608 $message = main::join_lines($message);
1609 $message =~ s/^$0: *//; # Remove initial program name
1610 $message =~ s/[.;,]+$//; # Remove certain ending punctuation
1611 $message = "\n$0: $message;";
12ac2576 1612
99870f4d
KW
1613 # Fold the message with program name, semi-colon end punctuation
1614 # (which looks good with the message that carp appends to it), and a
1615 # hanging indent for continuation lines.
1616 $message = main::simple_fold($message, "", 4) unless $nofold;
1617 $message =~ s/\n$//; # Remove the trailing nl so what carp
1618 # appends is to the same line
1619 }
12ac2576 1620
99870f4d 1621 return $message if defined wantarray; # If a caller just wants the msg
12ac2576 1622
99870f4d
KW
1623 carp $message;
1624 return;
1625}
7ebf06b3 1626
99870f4d
KW
1627sub my_carp_bug {
1628 # This is called when it is clear that the problem is caused by a bug in
1629 # this program.
7ebf06b3 1630
99870f4d
KW
1631 my $message = shift;
1632 $message =~ s/^$0: *//;
1633 $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");
1634 carp $message;
1635 return;
1636}
7ebf06b3 1637
99870f4d
KW
1638sub carp_too_few_args {
1639 if (@_ != 2) {
1640 my_carp_bug("Wrong number of arguments: to 'carp_too_few_arguments'. No action taken.");
1641 return;
12ac2576 1642 }
7ebf06b3 1643
99870f4d
KW
1644 my $args_ref = shift;
1645 my $count = shift;
7ebf06b3 1646
99870f4d
KW
1647 my_carp_bug("Need at least $count arguments to "
1648 . (caller 1)[3]
1649 . ". Instead got: '"
1650 . join ', ', @$args_ref
1651 . "'. No action taken.");
1652 return;
12ac2576
JP
1653}
1654
99870f4d
KW
1655sub carp_extra_args {
1656 my $args_ref = shift;
1657 my_carp_bug("Too many arguments to 'carp_extra_args': (" . join(', ', @_) . "); Extras ignored.") if @_;
12ac2576 1658
99870f4d
KW
1659 unless (ref $args_ref) {
1660 my_carp_bug("Argument to 'carp_extra_args' ($args_ref) must be a ref. Not checking arguments.");
1661 return;
1662 }
1663 my ($package, $file, $line) = caller;
1664 my $subroutine = (caller 1)[3];
cf25bb62 1665
99870f4d
KW
1666 my $list;
1667 if (ref $args_ref eq 'HASH') {
1668 foreach my $key (keys %$args_ref) {
1669 $args_ref->{$key} = $UNDEF unless defined $args_ref->{$key};
cf25bb62 1670 }
99870f4d 1671 $list = join ', ', each %{$args_ref};
cf25bb62 1672 }
99870f4d
KW
1673 elsif (ref $args_ref eq 'ARRAY') {
1674 foreach my $arg (@$args_ref) {
1675 $arg = $UNDEF unless defined $arg;
1676 }
1677 $list = join ', ', @$args_ref;
1678 }
1679 else {
1680 my_carp_bug("Can't cope with ref "
1681 . ref($args_ref)
1682 . " . argument to 'carp_extra_args'. Not checking arguments.");
1683 return;
1684 }
1685
1686 my_carp_bug("Unrecognized parameters in options: '$list' to $subroutine. Skipped.");
1687 return;
d73e5302
JH
1688}
1689
99870f4d
KW
1690package main;
1691
1692{ # Closure
1693
1694 # This program uses the inside-out method for objects, as recommended in
1695 # "Perl Best Practices". This closure aids in generating those. There
1696 # are two routines. setup_package() is called once per package to set
1697 # things up, and then set_access() is called for each hash representing a
1698 # field in the object. These routines arrange for the object to be
1699 # properly destroyed when no longer used, and for standard accessor
1700 # functions to be generated. If you need more complex accessors, just
1701 # write your own and leave those accesses out of the call to set_access().
1702 # More details below.
1703
1704 my %constructor_fields; # fields that are to be used in constructors; see
1705 # below
1706
1707 # The values of this hash will be the package names as keys to other
1708 # hashes containing the name of each field in the package as keys, and
1709 # references to their respective hashes as values.
1710 my %package_fields;
1711
1712 sub setup_package {
1713 # Sets up the package, creating standard DESTROY and dump methods
1714 # (unless already defined). The dump method is used in debugging by
1715 # simple_dumper().
1716 # The optional parameters are:
1717 # a) a reference to a hash, that gets populated by later
1718 # set_access() calls with one of the accesses being
1719 # 'constructor'. The caller can then refer to this, but it is
1720 # not otherwise used by these two routines.
1721 # b) a reference to a callback routine to call during destruction
1722 # of the object, before any fields are actually destroyed
1723
1724 my %args = @_;
1725 my $constructor_ref = delete $args{'Constructor_Fields'};
1726 my $destroy_callback = delete $args{'Destroy_Callback'};
1727 Carp::carp_extra_args(\@_) if main::DEBUG && %args;
1728
1729 my %fields;
1730 my $package = (caller)[0];
1731
1732 $package_fields{$package} = \%fields;
1733 $constructor_fields{$package} = $constructor_ref;
1734
1735 unless ($package->can('DESTROY')) {
1736 my $destroy_name = "${package}::DESTROY";
1737 no strict "refs";
1738
1739 # Use typeglob to give the anonymous subroutine the name we want
1740 *$destroy_name = sub {
1741 my $self = shift;
ffe43484 1742 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1743
1744 $self->$destroy_callback if $destroy_callback;
1745 foreach my $field (keys %{$package_fields{$package}}) {
1746 #print STDERR __LINE__, ": Destroying ", ref $self, " ", sprintf("%04X", $addr), ": ", $field, "\n";
1747 delete $package_fields{$package}{$field}{$addr};
1748 }
1749 return;
1750 }
1751 }
1752
1753 unless ($package->can('dump')) {
1754 my $dump_name = "${package}::dump";
1755 no strict "refs";
1756 *$dump_name = sub {
1757 my $self = shift;
1758 return dump_inside_out($self, $package_fields{$package}, @_);
1759 }
1760 }
1761 return;
1762 }
1763
1764 sub set_access {
1765 # Arrange for the input field to be garbage collected when no longer
1766 # needed. Also, creates standard accessor functions for the field
1767 # based on the optional parameters-- none if none of these parameters:
1768 # 'addable' creates an 'add_NAME()' accessor function.
1769 # 'readable' or 'readable_array' creates a 'NAME()' accessor
1770 # function.
1771 # 'settable' creates a 'set_NAME()' accessor function.
1772 # 'constructor' doesn't create an accessor function, but adds the
1773 # field to the hash that was previously passed to
1774 # setup_package();
1775 # Any of the accesses can be abbreviated down, so that 'a', 'ad',
1776 # 'add' etc. all mean 'addable'.
1777 # The read accessor function will work on both array and scalar
1778 # values. If another accessor in the parameter list is 'a', the read
1779 # access assumes an array. You can also force it to be array access
1780 # by specifying 'readable_array' instead of 'readable'
1781 #
1782 # A sort-of 'protected' access can be set-up by preceding the addable,
1783 # readable or settable with some initial portion of 'protected_' (but,
1784 # the underscore is required), like 'p_a', 'pro_set', etc. The
1785 # "protection" is only by convention. All that happens is that the
1786 # accessor functions' names begin with an underscore. So instead of
1787 # calling set_foo, the call is _set_foo. (Real protection could be
c1739a4a 1788 # accomplished by having a new subroutine, end_package, called at the
99870f4d
KW
1789 # end of each package, and then storing the __LINE__ ranges and
1790 # checking them on every accessor. But that is way overkill.)
1791
1792 # We create anonymous subroutines as the accessors and then use
1793 # typeglobs to assign them to the proper package and name
1794
1795 my $name = shift; # Name of the field
1796 my $field = shift; # Reference to the inside-out hash containing the
1797 # field
1798
1799 my $package = (caller)[0];
1800
1801 if (! exists $package_fields{$package}) {
1802 croak "$0: Must call 'setup_package' before 'set_access'";
1803 }
d73e5302 1804
99870f4d
KW
1805 # Stash the field so DESTROY can get it.
1806 $package_fields{$package}{$name} = $field;
cf25bb62 1807
99870f4d
KW
1808 # Remaining arguments are the accessors. For each...
1809 foreach my $access (@_) {
1810 my $access = lc $access;
cf25bb62 1811
99870f4d 1812 my $protected = "";
cf25bb62 1813
99870f4d
KW
1814 # Match the input as far as it goes.
1815 if ($access =~ /^(p[^_]*)_/) {
1816 $protected = $1;
1817 if (substr('protected_', 0, length $protected)
1818 eq $protected)
1819 {
1820
1821 # Add 1 for the underscore not included in $protected
1822 $access = substr($access, length($protected) + 1);
1823 $protected = '_';
1824 }
1825 else {
1826 $protected = "";
1827 }
1828 }
1829
1830 if (substr('addable', 0, length $access) eq $access) {
1831 my $subname = "${package}::${protected}add_$name";
1832 no strict "refs";
1833
1834 # add_ accessor. Don't add if already there, which we
1835 # determine using 'eq' for scalars and '==' otherwise.
1836 *$subname = sub {
1837 use strict "refs";
1838 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
1839 my $self = shift;
1840 my $value = shift;
ffe43484 1841 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1842 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1843 if (ref $value) {
f998e60c 1844 return if grep { $value == $_ } @{$field->{$addr}};
99870f4d
KW
1845 }
1846 else {
f998e60c 1847 return if grep { $value eq $_ } @{$field->{$addr}};
99870f4d 1848 }
f998e60c 1849 push @{$field->{$addr}}, $value;
99870f4d
KW
1850 return;
1851 }
1852 }
1853 elsif (substr('constructor', 0, length $access) eq $access) {
1854 if ($protected) {
1855 Carp::my_carp_bug("Can't set-up 'protected' constructors")
1856 }
1857 else {
1858 $constructor_fields{$package}{$name} = $field;
1859 }
1860 }
1861 elsif (substr('readable_array', 0, length $access) eq $access) {
1862
1863 # Here has read access. If one of the other parameters for
1864 # access is array, or this one specifies array (by being more
1865 # than just 'readable_'), then create a subroutine that
1866 # assumes the data is an array. Otherwise just a scalar
1867 my $subname = "${package}::${protected}$name";
1868 if (grep { /^a/i } @_
1869 or length($access) > length('readable_'))
1870 {
1871 no strict "refs";
1872 *$subname = sub {
1873 use strict "refs";
23e33b60 1874 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
ffe43484 1875 my $addr = do { no overloading; pack 'J', $_[0]; };
99870f4d
KW
1876 if (ref $field->{$addr} ne 'ARRAY') {
1877 my $type = ref $field->{$addr};
1878 $type = 'scalar' unless $type;
1879 Carp::my_carp_bug("Trying to read $name as an array when it is a $type. Big problems.");
1880 return;
1881 }
1882 return scalar @{$field->{$addr}} unless wantarray;
1883
1884 # Make a copy; had problems with caller modifying the
1885 # original otherwise
1886 my @return = @{$field->{$addr}};
1887 return @return;
1888 }
1889 }
1890 else {
1891
1892 # Here not an array value, a simpler function.
1893 no strict "refs";
1894 *$subname = sub {
1895 use strict "refs";
23e33b60 1896 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
f998e60c 1897 no overloading;
051df77b 1898 return $field->{pack 'J', $_[0]};
99870f4d
KW
1899 }
1900 }
1901 }
1902 elsif (substr('settable', 0, length $access) eq $access) {
1903 my $subname = "${package}::${protected}set_$name";
1904 no strict "refs";
1905 *$subname = sub {
1906 use strict "refs";
23e33b60
KW
1907 if (main::DEBUG) {
1908 return Carp::carp_too_few_args(\@_, 2) if @_ < 2;
1909 Carp::carp_extra_args(\@_) if @_ > 2;
1910 }
1911 # $self is $_[0]; $value is $_[1]
f998e60c 1912 no overloading;
051df77b 1913 $field->{pack 'J', $_[0]} = $_[1];
99870f4d
KW
1914 return;
1915 }
1916 }
1917 else {
1918 Carp::my_carp_bug("Unknown accessor type $access. No accessor set.");
1919 }
cf25bb62 1920 }
99870f4d 1921 return;
cf25bb62 1922 }
99870f4d
KW
1923}
1924
1925package Input_file;
1926
1927# All input files use this object, which stores various attributes about them,
1928# and provides for convenient, uniform handling. The run method wraps the
1929# processing. It handles all the bookkeeping of opening, reading, and closing
1930# the file, returning only significant input lines.
1931#
1932# Each object gets a handler which processes the body of the file, and is
1933# called by run(). Most should use the generic, default handler, which has
1934# code scrubbed to handle things you might not expect. A handler should
1935# basically be a while(next_line()) {...} loop.
1936#
1937# You can also set up handlers to
1938# 1) call before the first line is read for pre processing
1939# 2) call to adjust each line of the input before the main handler gets them
1940# 3) call upon EOF before the main handler exits its loop
1941# 4) call at the end for post processing
1942#
1943# $_ is used to store the input line, and is to be filtered by the
1944# each_line_handler()s. So, if the format of the line is not in the desired
1945# format for the main handler, these are used to do that adjusting. They can
1946# be stacked (by enclosing them in an [ anonymous array ] in the constructor,
1947# so the $_ output of one is used as the input to the next. None of the other
1948# handlers are stackable, but could easily be changed to be so.
1949#
1950# Most of the handlers can call insert_lines() or insert_adjusted_lines()
1951# which insert the parameters as lines to be processed before the next input
1952# file line is read. This allows the EOF handler to flush buffers, for
1953# example. The difference between the two routines is that the lines inserted
1954# by insert_lines() are subjected to the each_line_handler()s. (So if you
1955# called it from such a handler, you would get infinite recursion.) Lines
1956# inserted by insert_adjusted_lines() go directly to the main handler without
1957# any adjustments. If the post-processing handler calls any of these, there
1958# will be no effect. Some error checking for these conditions could be added,
1959# but it hasn't been done.
1960#
1961# carp_bad_line() should be called to warn of bad input lines, which clears $_
1962# to prevent further processing of the line. This routine will output the
1963# message as a warning once, and then keep a count of the lines that have the
1964# same message, and output that count at the end of the file's processing.
1965# This keeps the number of messages down to a manageable amount.
1966#
1967# get_missings() should be called to retrieve any @missing input lines.
1968# Messages will be raised if this isn't done if the options aren't to ignore
1969# missings.
1970
1971sub trace { return main::trace(@_); }
1972
99870f4d
KW
1973{ # Closure
1974 # Keep track of fields that are to be put into the constructor.
1975 my %constructor_fields;
1976
1977 main::setup_package(Constructor_Fields => \%constructor_fields);
1978
1979 my %file; # Input file name, required
1980 main::set_access('file', \%file, qw{ c r });
1981
1982 my %first_released; # Unicode version file was first released in, required
1983 main::set_access('first_released', \%first_released, qw{ c r });
1984
1985 my %handler; # Subroutine to process the input file, defaults to
1986 # 'process_generic_property_file'
1987 main::set_access('handler', \%handler, qw{ c });
1988
1989 my %property;
1990 # name of property this file is for. defaults to none, meaning not
1991 # applicable, or is otherwise determinable, for example, from each line.
1992 main::set_access('property', \%property, qw{ c });
1993
1994 my %optional;
1995 # If this is true, the file is optional. If not present, no warning is
1996 # output. If it is present, the string given by this parameter is
1997 # evaluated, and if false the file is not processed.
1998 main::set_access('optional', \%optional, 'c', 'r');
1999
2000 my %non_skip;
2001 # This is used for debugging, to skip processing of all but a few input
2002 # files. Add 'non_skip => 1' to the constructor for those files you want
2003 # processed when you set the $debug_skip global.
2004 main::set_access('non_skip', \%non_skip, 'c');
2005
37e2e78e 2006 my %skip;
09ca89ce
KW
2007 # This is used to skip processing of this input file semi-permanently,
2008 # when it evaluates to true. The value should be the reason the file is
2009 # being skipped. It is used for files that we aren't planning to process
2010 # anytime soon, but want to allow to be in the directory and not raise a
2011 # message that we are not handling. Mostly for test files. This is in
2012 # contrast to the non_skip element, which is supposed to be used very
2013 # temporarily for debugging. Sets 'optional' to 1. Also, files that we
2014 # pretty much will never look at can be placed in the global
2015 # %ignored_files instead. Ones used here will be added to that list.
37e2e78e
KW
2016 main::set_access('skip', \%skip, 'c');
2017
99870f4d
KW
2018 my %each_line_handler;
2019 # list of subroutines to look at and filter each non-comment line in the
2020 # file. defaults to none. The subroutines are called in order, each is
2021 # to adjust $_ for the next one, and the final one adjusts it for
2022 # 'handler'
2023 main::set_access('each_line_handler', \%each_line_handler, 'c');
2024
2025 my %has_missings_defaults;
2026 # ? Are there lines in the file giving default values for code points
2027 # missing from it?. Defaults to NO_DEFAULTS. Otherwise NOT_IGNORED is
2028 # the norm, but IGNORED means it has such lines, but the handler doesn't
2029 # use them. Having these three states allows us to catch changes to the
2030 # UCD that this program should track
2031 main::set_access('has_missings_defaults',
2032 \%has_missings_defaults, qw{ c r });
2033
2034 my %pre_handler;
2035 # Subroutine to call before doing anything else in the file. If undef, no
2036 # such handler is called.
2037 main::set_access('pre_handler', \%pre_handler, qw{ c });
2038
2039 my %eof_handler;
2040 # Subroutine to call upon getting an EOF on the input file, but before
2041 # that is returned to the main handler. This is to allow buffers to be
2042 # flushed. The handler is expected to call insert_lines() or
2043 # insert_adjusted() with the buffered material
2044 main::set_access('eof_handler', \%eof_handler, qw{ c r });
2045
2046 my %post_handler;
2047 # Subroutine to call after all the lines of the file are read in and
2048 # processed. If undef, no such handler is called.
2049 main::set_access('post_handler', \%post_handler, qw{ c });
2050
2051 my %progress_message;
2052 # Message to print to display progress in lieu of the standard one
2053 main::set_access('progress_message', \%progress_message, qw{ c });
2054
2055 my %handle;
2056 # cache open file handle, internal. Is undef if file hasn't been
2057 # processed at all, empty if has;
2058 main::set_access('handle', \%handle);
2059
2060 my %added_lines;
2061 # cache of lines added virtually to the file, internal
2062 main::set_access('added_lines', \%added_lines);
2063
2064 my %errors;
2065 # cache of errors found, internal
2066 main::set_access('errors', \%errors);
2067
2068 my %missings;
2069 # storage of '@missing' defaults lines
2070 main::set_access('missings', \%missings);
2071
2072 sub new {
2073 my $class = shift;
2074
2075 my $self = bless \do{ my $anonymous_scalar }, $class;
ffe43484 2076 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2077
2078 # Set defaults
2079 $handler{$addr} = \&main::process_generic_property_file;
2080 $non_skip{$addr} = 0;
37e2e78e 2081 $skip{$addr} = 0;
99870f4d
KW
2082 $has_missings_defaults{$addr} = $NO_DEFAULTS;
2083 $handle{$addr} = undef;
2084 $added_lines{$addr} = [ ];
2085 $each_line_handler{$addr} = [ ];
2086 $errors{$addr} = { };
2087 $missings{$addr} = [ ];
2088
2089 # Two positional parameters.
99f78760 2090 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
99870f4d
KW
2091 $file{$addr} = main::internal_file_to_platform(shift);
2092 $first_released{$addr} = shift;
2093
2094 # The rest of the arguments are key => value pairs
2095 # %constructor_fields has been set up earlier to list all possible
2096 # ones. Either set or push, depending on how the default has been set
2097 # up just above.
2098 my %args = @_;
2099 foreach my $key (keys %args) {
2100 my $argument = $args{$key};
2101
2102 # Note that the fields are the lower case of the constructor keys
2103 my $hash = $constructor_fields{lc $key};
2104 if (! defined $hash) {
2105 Carp::my_carp_bug("Unrecognized parameters '$key => $argument' to new() for $self. Skipped");
2106 next;
2107 }
2108 if (ref $hash->{$addr} eq 'ARRAY') {
2109 if (ref $argument eq 'ARRAY') {
2110 foreach my $argument (@{$argument}) {
2111 next if ! defined $argument;
2112 push @{$hash->{$addr}}, $argument;
2113 }
2114 }
2115 else {
2116 push @{$hash->{$addr}}, $argument if defined $argument;
2117 }
2118 }
2119 else {
2120 $hash->{$addr} = $argument;
2121 }
2122 delete $args{$key};
2123 };
2124
2125 # If the file has a property for it, it means that the property is not
2126 # listed in the file's entries. So add a handler to the list of line
2127 # handlers to insert the property name into the lines, to provide a
2128 # uniform interface to the final processing subroutine.
2129 # the final code doesn't have to worry about that.
2130 if ($property{$addr}) {
2131 push @{$each_line_handler{$addr}}, \&_insert_property_into_line;
2132 }
2133
2134 if ($non_skip{$addr} && ! $debug_skip && $verbosity) {
2135 print "Warning: " . __PACKAGE__ . " constructor for $file{$addr} has useless 'non_skip' in it\n";
a3a8c5f0 2136 }
99870f4d 2137
09ca89ce
KW
2138 # If skipping, set to optional, and add to list of ignored files,
2139 # including its reason
2140 if ($skip{$addr}) {
2141 $optional{$addr} = 1;
2142 $ignored_files{$file{$addr}} = $skip{$addr}
2143 }
37e2e78e 2144
99870f4d 2145 return $self;
d73e5302
JH
2146 }
2147
cf25bb62 2148
99870f4d
KW
2149 use overload
2150 fallback => 0,
2151 qw("") => "_operator_stringify",
2152 "." => \&main::_operator_dot,
2153 ;
cf25bb62 2154
99870f4d
KW
2155 sub _operator_stringify {
2156 my $self = shift;
cf25bb62 2157
99870f4d 2158 return __PACKAGE__ . " object for " . $self->file;
d73e5302 2159 }
d73e5302 2160
99870f4d
KW
2161 # flag to make sure extracted files are processed early
2162 my $seen_non_extracted_non_age = 0;
d73e5302 2163
99870f4d
KW
2164 sub run {
2165 # Process the input object $self. This opens and closes the file and
2166 # calls all the handlers for it. Currently, this can only be called
2167 # once per file, as it destroy's the EOF handler
d73e5302 2168
99870f4d
KW
2169 my $self = shift;
2170 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
b6922eda 2171
ffe43484 2172 my $addr = do { no overloading; pack 'J', $self; };
b6922eda 2173
99870f4d 2174 my $file = $file{$addr};
d73e5302 2175
99870f4d
KW
2176 # Don't process if not expecting this file (because released later
2177 # than this Unicode version), and isn't there. This means if someone
2178 # copies it into an earlier version's directory, we will go ahead and
2179 # process it.
2180 return if $first_released{$addr} gt $v_version && ! -e $file;
2181
2182 # If in debugging mode and this file doesn't have the non-skip
2183 # flag set, and isn't one of the critical files, skip it.
2184 if ($debug_skip
2185 && $first_released{$addr} ne v0
2186 && ! $non_skip{$addr})
2187 {
2188 print "Skipping $file in debugging\n" if $verbosity;
2189 return;
2190 }
2191
2192 # File could be optional
37e2e78e 2193 if ($optional{$addr}) {
99870f4d
KW
2194 return unless -e $file;
2195 my $result = eval $optional{$addr};
2196 if (! defined $result) {
2197 Carp::my_carp_bug("Got '$@' when tried to eval $optional{$addr}. $file Skipped.");
2198 return;
2199 }
2200 if (! $result) {
2201 if ($verbosity) {
2202 print STDERR "Skipping processing input file '$file' because '$optional{$addr}' is not true\n";
2203 }
2204 return;
2205 }
2206 }
2207
2208 if (! defined $file || ! -e $file) {
2209
2210 # If the file doesn't exist, see if have internal data for it
2211 # (based on first_released being 0).
2212 if ($first_released{$addr} eq v0) {
2213 $handle{$addr} = 'pretend_is_open';
2214 }
2215 else {
2216 if (! $optional{$addr} # File could be optional
2217 && $v_version ge $first_released{$addr})
2218 {
2219 print STDERR "Skipping processing input file '$file' because not found\n" if $v_version ge $first_released{$addr};
2220 }
2221 return;
2222 }
2223 }
2224 else {
2225
37e2e78e
KW
2226 # Here, the file exists. Some platforms may change the case of
2227 # its name
99870f4d 2228 if ($seen_non_extracted_non_age) {
517956bf 2229 if ($file =~ /$EXTRACTED/i) {
99870f4d 2230 Carp::my_carp_bug(join_lines(<<END
99f78760 2231$file should be processed just after the 'Prop...Alias' files, and before
99870f4d
KW
2232anything not in the $EXTRACTED_DIR directory. Proceeding, but the results may
2233have subtle problems
2234END
2235 ));
2236 }
2237 }
2238 elsif ($EXTRACTED_DIR
2239 && $first_released{$addr} ne v0
517956bf
CB
2240 && $file !~ /$EXTRACTED/i
2241 && lc($file) ne 'dage.txt')
99870f4d
KW
2242 {
2243 # We don't set this (by the 'if' above) if we have no
2244 # extracted directory, so if running on an early version,
2245 # this test won't work. Not worth worrying about.
2246 $seen_non_extracted_non_age = 1;
2247 }
2248
2249 # And mark the file as having being processed, and warn if it
2250 # isn't a file we are expecting. As we process the files,
2251 # they are deleted from the hash, so any that remain at the
2252 # end of the program are files that we didn't process.
517956bf
CB
2253 my $fkey = File::Spec->rel2abs($file);
2254 my $expecting = delete $potential_files{$fkey};
2255 $expecting = delete $potential_files{lc($fkey)} unless defined $expecting;
678f13d5
KW
2256 Carp::my_carp("Was not expecting '$file'.") if
2257 ! $expecting
99870f4d
KW
2258 && ! defined $handle{$addr};
2259
37e2e78e
KW
2260 # Having deleted from expected files, we can quit if not to do
2261 # anything. Don't print progress unless really want verbosity
2262 if ($skip{$addr}) {
2263 print "Skipping $file.\n" if $verbosity >= $VERBOSE;
2264 return;
2265 }
2266
99870f4d
KW
2267 # Open the file, converting the slashes used in this program
2268 # into the proper form for the OS
2269 my $file_handle;
2270 if (not open $file_handle, "<", $file) {
2271 Carp::my_carp("Can't open $file. Skipping: $!");
2272 return 0;
2273 }
2274 $handle{$addr} = $file_handle; # Cache the open file handle
2275 }
2276
2277 if ($verbosity >= $PROGRESS) {
2278 if ($progress_message{$addr}) {
2279 print "$progress_message{$addr}\n";
2280 }
2281 else {
2282 # If using a virtual file, say so.
2283 print "Processing ", (-e $file)
2284 ? $file
2285 : "substitute $file",
2286 "\n";
2287 }
2288 }
2289
2290
2291 # Call any special handler for before the file.
2292 &{$pre_handler{$addr}}($self) if $pre_handler{$addr};
2293
2294 # Then the main handler
2295 &{$handler{$addr}}($self);
2296
2297 # Then any special post-file handler.
2298 &{$post_handler{$addr}}($self) if $post_handler{$addr};
2299
2300 # If any errors have been accumulated, output the counts (as the first
2301 # error message in each class was output when it was encountered).
2302 if ($errors{$addr}) {
2303 my $total = 0;
2304 my $types = 0;
2305 foreach my $error (keys %{$errors{$addr}}) {
2306 $total += $errors{$addr}->{$error};
2307 delete $errors{$addr}->{$error};
2308 $types++;
2309 }
2310 if ($total > 1) {
2311 my $message
2312 = "A total of $total lines had errors in $file. ";
2313
2314 $message .= ($types == 1)
2315 ? '(Only the first one was displayed.)'
2316 : '(Only the first of each type was displayed.)';
2317 Carp::my_carp($message);
2318 }
2319 }
2320
2321 if (@{$missings{$addr}}) {
2322 Carp::my_carp_bug("Handler for $file didn't look at all the \@missing lines. Generated tables likely are wrong");
2323 }
2324
2325 # If a real file handle, close it.
2326 close $handle{$addr} or Carp::my_carp("Can't close $file: $!") if
2327 ref $handle{$addr};
2328 $handle{$addr} = ""; # Uses empty to indicate that has already seen
2329 # the file, as opposed to undef
2330 return;
2331 }
2332
2333 sub next_line {
2334 # Sets $_ to be the next logical input line, if any. Returns non-zero
2335 # if such a line exists. 'logical' means that any lines that have
2336 # been added via insert_lines() will be returned in $_ before the file
2337 # is read again.
2338
2339 my $self = shift;
2340 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2341
ffe43484 2342 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2343
2344 # Here the file is open (or if the handle is not a ref, is an open
2345 # 'virtual' file). Get the next line; any inserted lines get priority
2346 # over the file itself.
2347 my $adjusted;
2348
2349 LINE:
2350 while (1) { # Loop until find non-comment, non-empty line
2351 #local $to_trace = 1 if main::DEBUG;
2352 my $inserted_ref = shift @{$added_lines{$addr}};
2353 if (defined $inserted_ref) {
2354 ($adjusted, $_) = @{$inserted_ref};
2355 trace $adjusted, $_ if main::DEBUG && $to_trace;
2356 return 1 if $adjusted;
2357 }
2358 else {
2359 last if ! ref $handle{$addr}; # Don't read unless is real file
2360 last if ! defined ($_ = readline $handle{$addr});
2361 }
2362 chomp;
2363 trace $_ if main::DEBUG && $to_trace;
2364
2365 # See if this line is the comment line that defines what property
2366 # value that code points that are not listed in the file should
2367 # have. The format or existence of these lines is not guaranteed
2368 # by Unicode since they are comments, but the documentation says
2369 # that this was added for machine-readability, so probably won't
2370 # change. This works starting in Unicode Version 5.0. They look
2371 # like:
2372 #
2373 # @missing: 0000..10FFFF; Not_Reordered
2374 # @missing: 0000..10FFFF; Decomposition_Mapping; <code point>
2375 # @missing: 0000..10FFFF; ; NaN
2376 #
2377 # Save the line for a later get_missings() call.
2378 if (/$missing_defaults_prefix/) {
2379 if ($has_missings_defaults{$addr} == $NO_DEFAULTS) {
2380 $self->carp_bad_line("Unexpected \@missing line. Assuming no missing entries");
2381 }
2382 elsif ($has_missings_defaults{$addr} == $NOT_IGNORED) {
2383 my @defaults = split /\s* ; \s*/x, $_;
2384
2385 # The first field is the @missing, which ends in a
2386 # semi-colon, so can safely shift.
2387 shift @defaults;
2388
2389 # Some of these lines may have empty field placeholders
2390 # which get in the way. An example is:
2391 # @missing: 0000..10FFFF; ; NaN
2392 # Remove them. Process starting from the top so the
2393 # splice doesn't affect things still to be looked at.
2394 for (my $i = @defaults - 1; $i >= 0; $i--) {
2395 next if $defaults[$i] ne "";
2396 splice @defaults, $i, 1;
2397 }
2398
2399 # What's left should be just the property (maybe) and the
2400 # default. Having only one element means it doesn't have
2401 # the property.
2402 my $default;
2403 my $property;
2404 if (@defaults >= 1) {
2405 if (@defaults == 1) {
2406 $default = $defaults[0];
2407 }
2408 else {
2409 $property = $defaults[0];
2410 $default = $defaults[1];
2411 }
2412 }
2413
2414 if (@defaults < 1
2415 || @defaults > 2
2416 || ($default =~ /^</
2417 && $default !~ /^<code *point>$/i
2418 && $default !~ /^<none>$/i))
2419 {
2420 $self->carp_bad_line("Unrecognized \@missing line: $_. Assuming no missing entries");
2421 }
2422 else {
2423
2424 # If the property is missing from the line, it should
2425 # be the one for the whole file
2426 $property = $property{$addr} if ! defined $property;
2427
2428 # Change <none> to the null string, which is what it
2429 # really means. If the default is the code point
2430 # itself, set it to <code point>, which is what
2431 # Unicode uses (but sometimes they've forgotten the
2432 # space)
2433 if ($default =~ /^<none>$/i) {
2434 $default = "";
2435 }
2436 elsif ($default =~ /^<code *point>$/i) {
2437 $default = $CODE_POINT;
2438 }
2439
2440 # Store them as a sub-arrays with both components.
2441 push @{$missings{$addr}}, [ $default, $property ];
2442 }
2443 }
2444
2445 # There is nothing for the caller to process on this comment
2446 # line.
2447 next;
2448 }
2449
2450 # Remove comments and trailing space, and skip this line if the
2451 # result is empty
2452 s/#.*//;
2453 s/\s+$//;
2454 next if /^$/;
2455
2456 # Call any handlers for this line, and skip further processing of
2457 # the line if the handler sets the line to null.
2458 foreach my $sub_ref (@{$each_line_handler{$addr}}) {
2459 &{$sub_ref}($self);
2460 next LINE if /^$/;
2461 }
2462
2463 # Here the line is ok. return success.
2464 return 1;
2465 } # End of looping through lines.
2466
2467 # If there is an EOF handler, call it (only once) and if it generates
2468 # more lines to process go back in the loop to handle them.
2469 if ($eof_handler{$addr}) {
2470 &{$eof_handler{$addr}}($self);
2471 $eof_handler{$addr} = ""; # Currently only get one shot at it.
2472 goto LINE if $added_lines{$addr};
2473 }
2474
2475 # Return failure -- no more lines.
2476 return 0;
2477
2478 }
2479
2480# Not currently used, not fully tested.
2481# sub peek {
2482# # Non-destructive look-ahead one non-adjusted, non-comment, non-blank
2483# # record. Not callable from an each_line_handler(), nor does it call
2484# # an each_line_handler() on the line.
2485#
2486# my $self = shift;
ffe43484 2487# my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2488#
2489# foreach my $inserted_ref (@{$added_lines{$addr}}) {
2490# my ($adjusted, $line) = @{$inserted_ref};
2491# next if $adjusted;
2492#
2493# # Remove comments and trailing space, and return a non-empty
2494# # resulting line
2495# $line =~ s/#.*//;
2496# $line =~ s/\s+$//;
2497# return $line if $line ne "";
2498# }
2499#
2500# return if ! ref $handle{$addr}; # Don't read unless is real file
2501# while (1) { # Loop until find non-comment, non-empty line
2502# local $to_trace = 1 if main::DEBUG;
2503# trace $_ if main::DEBUG && $to_trace;
2504# return if ! defined (my $line = readline $handle{$addr});
2505# chomp $line;
2506# push @{$added_lines{$addr}}, [ 0, $line ];
2507#
2508# $line =~ s/#.*//;
2509# $line =~ s/\s+$//;
2510# return $line if $line ne "";
2511# }
2512#
2513# return;
2514# }
2515
2516
2517 sub insert_lines {
2518 # Lines can be inserted so that it looks like they were in the input
2519 # file at the place it was when this routine is called. See also
2520 # insert_adjusted_lines(). Lines inserted via this routine go through
2521 # any each_line_handler()
2522
2523 my $self = shift;
2524
2525 # Each inserted line is an array, with the first element being 0 to
2526 # indicate that this line hasn't been adjusted, and needs to be
2527 # processed.
f998e60c 2528 no overloading;
051df77b 2529 push @{$added_lines{pack 'J', $self}}, map { [ 0, $_ ] } @_;
99870f4d
KW
2530 return;
2531 }
2532
2533 sub insert_adjusted_lines {
2534 # Lines can be inserted so that it looks like they were in the input
2535 # file at the place it was when this routine is called. See also
2536 # insert_lines(). Lines inserted via this routine are already fully
2537 # adjusted, ready to be processed; each_line_handler()s handlers will
2538 # not be called. This means this is not a completely general
2539 # facility, as only the last each_line_handler on the stack should
2540 # call this. It could be made more general, by passing to each of the
2541 # line_handlers their position on the stack, which they would pass on
2542 # to this routine, and that would replace the boolean first element in
2543 # the anonymous array pushed here, so that the next_line routine could
2544 # use that to call only those handlers whose index is after it on the
2545 # stack. But this is overkill for what is needed now.
2546
2547 my $self = shift;
2548 trace $_[0] if main::DEBUG && $to_trace;
2549
2550 # Each inserted line is an array, with the first element being 1 to
2551 # indicate that this line has been adjusted
f998e60c 2552 no overloading;
051df77b 2553 push @{$added_lines{pack 'J', $self}}, map { [ 1, $_ ] } @_;
99870f4d
KW
2554 return;
2555 }
2556
2557 sub get_missings {
2558 # Returns the stored up @missings lines' values, and clears the list.
2559 # The values are in an array, consisting of the default in the first
2560 # element, and the property in the 2nd. However, since these lines
2561 # can be stacked up, the return is an array of all these arrays.
2562
2563 my $self = shift;
2564 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2565
ffe43484 2566 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2567
2568 # If not accepting a list return, just return the first one.
2569 return shift @{$missings{$addr}} unless wantarray;
2570
2571 my @return = @{$missings{$addr}};
2572 undef @{$missings{$addr}};
2573 return @return;
2574 }
2575
2576 sub _insert_property_into_line {
2577 # Add a property field to $_, if this file requires it.
2578
f998e60c 2579 my $self = shift;
ffe43484 2580 my $addr = do { no overloading; pack 'J', $self; };
f998e60c 2581 my $property = $property{$addr};
99870f4d
KW
2582 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2583
2584 $_ =~ s/(;|$)/; $property$1/;
2585 return;
2586 }
2587
2588 sub carp_bad_line {
2589 # Output consistent error messages, using either a generic one, or the
2590 # one given by the optional parameter. To avoid gazillions of the
2591 # same message in case the syntax of a file is way off, this routine
2592 # only outputs the first instance of each message, incrementing a
2593 # count so the totals can be output at the end of the file.
2594
2595 my $self = shift;
2596 my $message = shift;
2597 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2598
ffe43484 2599 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2600
2601 $message = 'Unexpected line' unless $message;
2602
2603 # No trailing punctuation so as to fit with our addenda.
2604 $message =~ s/[.:;,]$//;
2605
2606 # If haven't seen this exact message before, output it now. Otherwise
2607 # increment the count of how many times it has occurred
2608 unless ($errors{$addr}->{$message}) {
2609 Carp::my_carp("$message in '$_' in "
f998e60c 2610 . $file{$addr}
99870f4d
KW
2611 . " at line $.. Skipping this line;");
2612 $errors{$addr}->{$message} = 1;
2613 }
2614 else {
2615 $errors{$addr}->{$message}++;
2616 }
2617
2618 # Clear the line to prevent any further (meaningful) processing of it.
2619 $_ = "";
2620
2621 return;
2622 }
2623} # End closure
2624
2625package Multi_Default;
2626
2627# Certain properties in early versions of Unicode had more than one possible
2628# default for code points missing from the files. In these cases, one
2629# default applies to everything left over after all the others are applied,
2630# and for each of the others, there is a description of which class of code
2631# points applies to it. This object helps implement this by storing the
2632# defaults, and for all but that final default, an eval string that generates
2633# the class that it applies to.
2634
2635
2636{ # Closure
2637
2638 main::setup_package();
2639
2640 my %class_defaults;
2641 # The defaults structure for the classes
2642 main::set_access('class_defaults', \%class_defaults);
2643
2644 my %other_default;
2645 # The default that applies to everything left over.
2646 main::set_access('other_default', \%other_default, 'r');
2647
2648
2649 sub new {
2650 # The constructor is called with default => eval pairs, terminated by
2651 # the left-over default. e.g.
2652 # Multi_Default->new(
2653 # 'T' => '$gc->table("Mn") + $gc->table("Cf") - 0x200C
2654 # - 0x200D',
2655 # 'R' => 'some other expression that evaluates to code points',
2656 # .
2657 # .
2658 # .
2659 # 'U'));
2660
2661 my $class = shift;
2662
2663 my $self = bless \do{my $anonymous_scalar}, $class;
ffe43484 2664 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2665
2666 while (@_ > 1) {
2667 my $default = shift;
2668 my $eval = shift;
2669 $class_defaults{$addr}->{$default} = $eval;
2670 }
2671
2672 $other_default{$addr} = shift;
2673
2674 return $self;
2675 }
2676
2677 sub get_next_defaults {
2678 # Iterates and returns the next class of defaults.
2679 my $self = shift;
2680 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2681
ffe43484 2682 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2683
2684 return each %{$class_defaults{$addr}};
2685 }
2686}
2687
2688package Alias;
2689
2690# An alias is one of the names that a table goes by. This class defines them
2691# including some attributes. Everything is currently setup in the
2692# constructor.
2693
2694
2695{ # Closure
2696
2697 main::setup_package();
2698
2699 my %name;
2700 main::set_access('name', \%name, 'r');
2701
2702 my %loose_match;
c12f2655 2703 # Should this name match loosely or not.
99870f4d
KW
2704 main::set_access('loose_match', \%loose_match, 'r');
2705
33e96e72
KW
2706 my %make_re_pod_entry;
2707 # Some aliases should not get their own entries in the re section of the
2708 # pod, because they are covered by a wild-card, and some we want to
2709 # discourage use of. Binary
f82fe4ba 2710 main::set_access('make_re_pod_entry', \%make_re_pod_entry, 'r', 's');
99870f4d 2711
fd1e3e84
KW
2712 my %ucd;
2713 # Is this documented to be accessible via Unicode::UCD
2714 main::set_access('ucd', \%ucd, 'r', 's');
2715
99870f4d
KW
2716 my %status;
2717 # Aliases have a status, like deprecated, or even suppressed (which means
2718 # they don't appear in documentation). Enum
2719 main::set_access('status', \%status, 'r');
2720
0eac1e20 2721 my %ok_as_filename;
99870f4d
KW
2722 # Similarly, some aliases should not be considered as usable ones for
2723 # external use, such as file names, or we don't want documentation to
2724 # recommend them. Boolean
0eac1e20 2725 main::set_access('ok_as_filename', \%ok_as_filename, 'r');
99870f4d
KW
2726
2727 sub new {
2728 my $class = shift;
2729
2730 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2731 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2732
2733 $name{$addr} = shift;
2734 $loose_match{$addr} = shift;
33e96e72 2735 $make_re_pod_entry{$addr} = shift;
0eac1e20 2736 $ok_as_filename{$addr} = shift;
99870f4d 2737 $status{$addr} = shift;
fd1e3e84 2738 $ucd{$addr} = shift;
99870f4d
KW
2739
2740 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2741
2742 # Null names are never ok externally
0eac1e20 2743 $ok_as_filename{$addr} = 0 if $name{$addr} eq "";
99870f4d
KW
2744
2745 return $self;
2746 }
2747}
2748
2749package Range;
2750
2751# A range is the basic unit for storing code points, and is described in the
2752# comments at the beginning of the program. Each range has a starting code
2753# point; an ending code point (not less than the starting one); a value
2754# that applies to every code point in between the two end-points, inclusive;
2755# and an enum type that applies to the value. The type is for the user's
2756# convenience, and has no meaning here, except that a non-zero type is
2757# considered to not obey the normal Unicode rules for having standard forms.
2758#
2759# The same structure is used for both map and match tables, even though in the
2760# latter, the value (and hence type) is irrelevant and could be used as a
2761# comment. In map tables, the value is what all the code points in the range
2762# map to. Type 0 values have the standardized version of the value stored as
2763# well, so as to not have to recalculate it a lot.
2764
2765sub trace { return main::trace(@_); }
2766
2767{ # Closure
2768
2769 main::setup_package();
2770
2771 my %start;
2772 main::set_access('start', \%start, 'r', 's');
2773
2774 my %end;
2775 main::set_access('end', \%end, 'r', 's');
2776
2777 my %value;
2778 main::set_access('value', \%value, 'r');
2779
2780 my %type;
2781 main::set_access('type', \%type, 'r');
2782
2783 my %standard_form;
2784 # The value in internal standard form. Defined only if the type is 0.
2785 main::set_access('standard_form', \%standard_form);
2786
2787 # Note that if these fields change, the dump() method should as well
2788
2789 sub new {
2790 return Carp::carp_too_few_args(\@_, 3) if main::DEBUG && @_ < 3;
2791 my $class = shift;
2792
2793 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2794 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2795
2796 $start{$addr} = shift;
2797 $end{$addr} = shift;
2798
2799 my %args = @_;
2800
2801 my $value = delete $args{'Value'}; # Can be 0
2802 $value = "" unless defined $value;
2803 $value{$addr} = $value;
2804
2805 $type{$addr} = delete $args{'Type'} || 0;
2806
2807 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2808
2809 if (! $type{$addr}) {
2810 $standard_form{$addr} = main::standardize($value);
2811 }
2812
2813 return $self;
2814 }
2815
2816 use overload
2817 fallback => 0,
2818 qw("") => "_operator_stringify",
2819 "." => \&main::_operator_dot,
2820 ;
2821
2822 sub _operator_stringify {
2823 my $self = shift;
ffe43484 2824 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2825
2826 # Output it like '0041..0065 (value)'
2827 my $return = sprintf("%04X", $start{$addr})
2828 . '..'
2829 . sprintf("%04X", $end{$addr});
2830 my $value = $value{$addr};
2831 my $type = $type{$addr};
2832 $return .= ' (';
2833 $return .= "$value";
2834 $return .= ", Type=$type" if $type != 0;
2835 $return .= ')';
2836
2837 return $return;
2838 }
2839
2840 sub standard_form {
2841 # The standard form is the value itself if the standard form is
2842 # undefined (that is if the value is special)
2843
2844 my $self = shift;
2845 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2846
ffe43484 2847 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2848
2849 return $standard_form{$addr} if defined $standard_form{$addr};
2850 return $value{$addr};
2851 }
2852
2853 sub dump {
2854 # Human, not machine readable. For machine readable, comment out this
2855 # entire routine and let the standard one take effect.
2856 my $self = shift;
2857 my $indent = shift;
2858 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2859
ffe43484 2860 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2861
2862 my $return = $indent
2863 . sprintf("%04X", $start{$addr})
2864 . '..'
2865 . sprintf("%04X", $end{$addr})
2866 . " '$value{$addr}';";
2867 if (! defined $standard_form{$addr}) {
2868 $return .= "(type=$type{$addr})";
2869 }
2870 elsif ($standard_form{$addr} ne $value{$addr}) {
2871 $return .= "(standard '$standard_form{$addr}')";
2872 }
2873 return $return;
2874 }
2875} # End closure
2876
2877package _Range_List_Base;
2878
2879# Base class for range lists. A range list is simply an ordered list of
2880# ranges, so that the ranges with the lowest starting numbers are first in it.
2881#
2882# When a new range is added that is adjacent to an existing range that has the
2883# same value and type, it merges with it to form a larger range.
2884#
2885# Ranges generally do not overlap, except that there can be multiple entries
2886# of single code point ranges. This is because of NameAliases.txt.
2887#
2888# In this program, there is a standard value such that if two different
2889# values, have the same standard value, they are considered equivalent. This
2890# value was chosen so that it gives correct results on Unicode data
2891
2892# There are a number of methods to manipulate range lists, and some operators
2893# are overloaded to handle them.
2894
99870f4d
KW
2895sub trace { return main::trace(@_); }
2896
2897{ # Closure
2898
2899 our $addr;
2900
2901 main::setup_package();
2902
2903 my %ranges;
2904 # The list of ranges
2905 main::set_access('ranges', \%ranges, 'readable_array');
2906
2907 my %max;
2908 # The highest code point in the list. This was originally a method, but
2909 # actual measurements said it was used a lot.
2910 main::set_access('max', \%max, 'r');
2911
2912 my %each_range_iterator;
2913 # Iterator position for each_range()
2914 main::set_access('each_range_iterator', \%each_range_iterator);
2915
2916 my %owner_name_of;
2917 # Name of parent this is attached to, if any. Solely for better error
2918 # messages.
2919 main::set_access('owner_name_of', \%owner_name_of, 'p_r');
2920
2921 my %_search_ranges_cache;
2922 # A cache of the previous result from _search_ranges(), for better
2923 # performance
2924 main::set_access('_search_ranges_cache', \%_search_ranges_cache);
2925
2926 sub new {
2927 my $class = shift;
2928 my %args = @_;
2929
2930 # Optional initialization data for the range list.
2931 my $initialize = delete $args{'Initialize'};
2932
2933 my $self;
2934
2935 # Use _union() to initialize. _union() returns an object of this
2936 # class, which means that it will call this constructor recursively.
2937 # But it won't have this $initialize parameter so that it won't
2938 # infinitely loop on this.
2939 return _union($class, $initialize, %args) if defined $initialize;
2940
2941 $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2942 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2943
2944 # Optional parent object, only for debug info.
2945 $owner_name_of{$addr} = delete $args{'Owner'};
2946 $owner_name_of{$addr} = "" if ! defined $owner_name_of{$addr};
2947
2948 # Stringify, in case it is an object.
2949 $owner_name_of{$addr} = "$owner_name_of{$addr}";
2950
2951 # This is used only for error messages, and so a colon is added
2952 $owner_name_of{$addr} .= ": " if $owner_name_of{$addr} ne "";
2953
2954 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2955
2956 # Max is initialized to a negative value that isn't adjacent to 0,
2957 # for simpler tests
2958 $max{$addr} = -2;
2959
2960 $_search_ranges_cache{$addr} = 0;
2961 $ranges{$addr} = [];
2962
2963 return $self;
2964 }
2965
2966 use overload
2967 fallback => 0,
2968 qw("") => "_operator_stringify",
2969 "." => \&main::_operator_dot,
2970 ;
2971
2972 sub _operator_stringify {
2973 my $self = shift;
ffe43484 2974 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2975
2976 return "Range_List attached to '$owner_name_of{$addr}'"
2977 if $owner_name_of{$addr};
2978 return "anonymous Range_List " . \$self;
2979 }
2980
2981 sub _union {
2982 # Returns the union of the input code points. It can be called as
2983 # either a constructor or a method. If called as a method, the result
2984 # will be a new() instance of the calling object, containing the union
2985 # of that object with the other parameter's code points; if called as
2986 # a constructor, the first parameter gives the class the new object
2987 # should be, and the second parameter gives the code points to go into
2988 # it.
2989 # In either case, there are two parameters looked at by this routine;
2990 # any additional parameters are passed to the new() constructor.
2991 #
2992 # The code points can come in the form of some object that contains
2993 # ranges, and has a conventionally named method to access them; or
2994 # they can be an array of individual code points (as integers); or
2995 # just a single code point.
2996 #
2997 # If they are ranges, this routine doesn't make any effort to preserve
2998 # the range values of one input over the other. Therefore this base
2999 # class should not allow _union to be called from other than
3000 # initialization code, so as to prevent two tables from being added
3001 # together where the range values matter. The general form of this
3002 # routine therefore belongs in a derived class, but it was moved here
3003 # to avoid duplication of code. The failure to overload this in this
3004 # class keeps it safe.
3005 #
3006
3007 my $self;
3008 my @args; # Arguments to pass to the constructor
3009
3010 my $class = shift;
3011
3012 # If a method call, will start the union with the object itself, and
3013 # the class of the new object will be the same as self.
3014 if (ref $class) {
3015 $self = $class;
3016 $class = ref $self;
3017 push @args, $self;
3018 }
3019
3020 # Add the other required parameter.
3021 push @args, shift;
3022 # Rest of parameters are passed on to the constructor
3023
3024 # Accumulate all records from both lists.
3025 my @records;
3026 for my $arg (@args) {
3027 #local $to_trace = 0 if main::DEBUG;
3028 trace "argument = $arg" if main::DEBUG && $to_trace;
3029 if (! defined $arg) {
3030 my $message = "";
3031 if (defined $self) {
f998e60c 3032 no overloading;
051df77b 3033 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
3034 }
3035 Carp::my_carp_bug($message .= "Undefined argument to _union. No union done.");
3036 return;
3037 }
3038 $arg = [ $arg ] if ! ref $arg;
3039 my $type = ref $arg;
3040 if ($type eq 'ARRAY') {
3041 foreach my $element (@$arg) {
3042 push @records, Range->new($element, $element);
3043 }
3044 }
3045 elsif ($arg->isa('Range')) {
3046 push @records, $arg;
3047 }
3048 elsif ($arg->can('ranges')) {
3049 push @records, $arg->ranges;
3050 }
3051 else {
3052 my $message = "";
3053 if (defined $self) {
f998e60c 3054 no overloading;
051df77b 3055 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
3056 }
3057 Carp::my_carp_bug($message . "Cannot take the union of a $type. No union done.");
3058 return;
3059 }
3060 }
3061
3062 # Sort with the range containing the lowest ordinal first, but if
3063 # two ranges start at the same code point, sort with the bigger range
3064 # of the two first, because it takes fewer cycles.
3065 @records = sort { ($a->start <=> $b->start)
3066 or
3067 # if b is shorter than a, b->end will be
3068 # less than a->end, and we want to select
3069 # a, so want to return -1
3070 ($b->end <=> $a->end)
3071 } @records;
3072
3073 my $new = $class->new(@_);
3074
3075 # Fold in records so long as they add new information.
3076 for my $set (@records) {
3077 my $start = $set->start;
3078 my $end = $set->end;
3079 my $value = $set->value;
3080 if ($start > $new->max) {
3081 $new->_add_delete('+', $start, $end, $value);
3082 }
3083 elsif ($end > $new->max) {
3084 $new->_add_delete('+', $new->max +1, $end, $value);
3085 }
3086 }
3087
3088 return $new;
3089 }
3090
3091 sub range_count { # Return the number of ranges in the range list
3092 my $self = shift;
3093 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3094
f998e60c 3095 no overloading;
051df77b 3096 return scalar @{$ranges{pack 'J', $self}};
99870f4d
KW
3097 }
3098
3099 sub min {
3100 # Returns the minimum code point currently in the range list, or if
3101 # the range list is empty, 2 beyond the max possible. This is a
3102 # method because used so rarely, that not worth saving between calls,
3103 # and having to worry about changing it as ranges are added and
3104 # deleted.
3105
3106 my $self = shift;
3107 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3108
ffe43484 3109 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3110
3111 # If the range list is empty, return a large value that isn't adjacent
3112 # to any that could be in the range list, for simpler tests
6189eadc 3113 return $MAX_UNICODE_CODEPOINT + 2 unless scalar @{$ranges{$addr}};
99870f4d
KW
3114 return $ranges{$addr}->[0]->start;
3115 }
3116
3117 sub contains {
3118 # Boolean: Is argument in the range list? If so returns $i such that:
3119 # range[$i]->end < $codepoint <= range[$i+1]->end
3120 # which is one beyond what you want; this is so that the 0th range
3121 # doesn't return false
3122 my $self = shift;
3123 my $codepoint = shift;
3124 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3125
99870f4d
KW
3126 my $i = $self->_search_ranges($codepoint);
3127 return 0 unless defined $i;
3128
3129 # The search returns $i, such that
3130 # range[$i-1]->end < $codepoint <= range[$i]->end
3131 # So is in the table if and only iff it is at least the start position
3132 # of range $i.
f998e60c 3133 no overloading;
051df77b 3134 return 0 if $ranges{pack 'J', $self}->[$i]->start > $codepoint;
99870f4d
KW
3135 return $i + 1;
3136 }
3137
2f7a8815
KW
3138 sub containing_range {
3139 # Returns the range object that contains the code point, undef if none
3140
3141 my $self = shift;
3142 my $codepoint = shift;
3143 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3144
3145 my $i = $self->contains($codepoint);
3146 return unless $i;
3147
3148 # contains() returns 1 beyond where we should look
3149 no overloading;
3150 return $ranges{pack 'J', $self}->[$i-1];
3151 }
3152
99870f4d
KW
3153 sub value_of {
3154 # Returns the value associated with the code point, undef if none
3155
3156 my $self = shift;
3157 my $codepoint = shift;
3158 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3159
d69c231b
KW
3160 my $range = $self->containing_range($codepoint);
3161 return unless defined $range;
99870f4d 3162
d69c231b 3163 return $range->value;
99870f4d
KW
3164 }
3165
0a9dbafc
KW
3166 sub type_of {
3167 # Returns the type of the range containing the code point, undef if
3168 # the code point is not in the table
3169
3170 my $self = shift;
3171 my $codepoint = shift;
3172 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3173
3174 my $range = $self->containing_range($codepoint);
3175 return unless defined $range;
3176
3177 return $range->type;
3178 }
3179
99870f4d
KW
3180 sub _search_ranges {
3181 # Find the range in the list which contains a code point, or where it
3182 # should go if were to add it. That is, it returns $i, such that:
3183 # range[$i-1]->end < $codepoint <= range[$i]->end
3184 # Returns undef if no such $i is possible (e.g. at end of table), or
3185 # if there is an error.
3186
3187 my $self = shift;
3188 my $code_point = shift;
3189 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3190
ffe43484 3191 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3192
3193 return if $code_point > $max{$addr};
3194 my $r = $ranges{$addr}; # The current list of ranges
3195 my $range_list_size = scalar @$r;
3196 my $i;
3197
3198 use integer; # want integer division
3199
3200 # Use the cached result as the starting guess for this one, because,
3201 # an experiment on 5.1 showed that 90% of the time the cache was the
3202 # same as the result on the next call (and 7% it was one less).
3203 $i = $_search_ranges_cache{$addr};
3204 $i = 0 if $i >= $range_list_size; # Reset if no longer valid (prob.
3205 # from an intervening deletion
3206 #local $to_trace = 1 if main::DEBUG;
3207 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);
3208 return $i if $code_point <= $r->[$i]->end
3209 && ($i == 0 || $r->[$i-1]->end < $code_point);
3210
3211 # Here the cache doesn't yield the correct $i. Try adding 1.
3212 if ($i < $range_list_size - 1
3213 && $r->[$i]->end < $code_point &&
3214 $code_point <= $r->[$i+1]->end)
3215 {
3216 $i++;
3217 trace "next \$i is correct: $i" if main::DEBUG && $to_trace;
3218 $_search_ranges_cache{$addr} = $i;
3219 return $i;
3220 }
3221
3222 # Here, adding 1 also didn't work. We do a binary search to
3223 # find the correct position, starting with current $i
3224 my $lower = 0;
3225 my $upper = $range_list_size - 1;
3226 while (1) {
3227 trace "top of loop i=$i:", sprintf("%04X", $r->[$lower]->start), "[$lower] .. ", sprintf("%04X", $r->[$i]->start), "[$i] .. ", sprintf("%04X", $r->[$upper]->start), "[$upper]" if main::DEBUG && $to_trace;
3228
3229 if ($code_point <= $r->[$i]->end) {
3230
3231 # Here we have met the upper constraint. We can quit if we
3232 # also meet the lower one.
3233 last if $i == 0 || $r->[$i-1]->end < $code_point;
3234
3235 $upper = $i; # Still too high.
3236
3237 }
3238 else {
3239
3240 # Here, $r[$i]->end < $code_point, so look higher up.
3241 $lower = $i;
3242 }
3243
3244 # Split search domain in half to try again.
3245 my $temp = ($upper + $lower) / 2;
3246
3247 # No point in continuing unless $i changes for next time
3248 # in the loop.
3249 if ($temp == $i) {
3250
3251 # We can't reach the highest element because of the averaging.
3252 # So if one below the upper edge, force it there and try one
3253 # more time.
3254 if ($i == $range_list_size - 2) {
3255
3256 trace "Forcing to upper edge" if main::DEBUG && $to_trace;
3257 $i = $range_list_size - 1;
3258
3259 # Change $lower as well so if fails next time through,
3260 # taking the average will yield the same $i, and we will
3261 # quit with the error message just below.
3262 $lower = $i;
3263 next;
3264 }
3265 Carp::my_carp_bug("$owner_name_of{$addr}Can't find where the range ought to go. No action taken.");
3266 return;
3267 }
3268 $i = $temp;
3269 } # End of while loop
3270
3271 if (main::DEBUG && $to_trace) {
3272 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i;
3273 trace "i= [ $i ]", $r->[$i];
3274 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < $range_list_size - 1;
3275 }
3276
3277 # Here we have found the offset. Cache it as a starting point for the
3278 # next call.
3279 $_search_ranges_cache{$addr} = $i;
3280 return $i;
3281 }
3282
3283 sub _add_delete {
3284 # Add, replace or delete ranges to or from a list. The $type
3285 # parameter gives which:
3286 # '+' => insert or replace a range, returning a list of any changed
3287 # ranges.
3288 # '-' => delete a range, returning a list of any deleted ranges.
3289 #
3290 # The next three parameters give respectively the start, end, and
3291 # value associated with the range. 'value' should be null unless the
3292 # operation is '+';
3293 #
3294 # The range list is kept sorted so that the range with the lowest
3295 # starting position is first in the list, and generally, adjacent
c1739a4a 3296 # ranges with the same values are merged into a single larger one (see
99870f4d
KW
3297 # exceptions below).
3298 #
c1739a4a 3299 # There are more parameters; all are key => value pairs:
99870f4d
KW
3300 # Type gives the type of the value. It is only valid for '+'.
3301 # All ranges have types; if this parameter is omitted, 0 is
3302 # assumed. Ranges with type 0 are assumed to obey the
3303 # Unicode rules for casing, etc; ranges with other types are
3304 # not. Otherwise, the type is arbitrary, for the caller's
3305 # convenience, and looked at only by this routine to keep
3306 # adjacent ranges of different types from being merged into
3307 # a single larger range, and when Replace =>
3308 # $IF_NOT_EQUIVALENT is specified (see just below).
3309 # Replace determines what to do if the range list already contains
3310 # ranges which coincide with all or portions of the input
3311 # range. It is only valid for '+':
3312 # => $NO means that the new value is not to replace
3313 # any existing ones, but any empty gaps of the
3314 # range list coinciding with the input range
3315 # will be filled in with the new value.
3316 # => $UNCONDITIONALLY means to replace the existing values with
3317 # this one unconditionally. However, if the
3318 # new and old values are identical, the
3319 # replacement is skipped to save cycles
3320 # => $IF_NOT_EQUIVALENT means to replace the existing values
3321 # with this one if they are not equivalent.
3322 # Ranges are equivalent if their types are the
c1739a4a 3323 # same, and they are the same string; or if
99870f4d
KW
3324 # both are type 0 ranges, if their Unicode
3325 # standard forms are identical. In this last
3326 # case, the routine chooses the more "modern"
3327 # one to use. This is because some of the
3328 # older files are formatted with values that
3329 # are, for example, ALL CAPs, whereas the
3330 # derived files have a more modern style,
3331 # which looks better. By looking for this
3332 # style when the pre-existing and replacement
3333 # standard forms are the same, we can move to
3334 # the modern style
3335 # => $MULTIPLE means that if this range duplicates an
3336 # existing one, but has a different value,
3337 # don't replace the existing one, but insert
3338 # this, one so that the same range can occur
53d84487
KW
3339 # multiple times. They are stored LIFO, so
3340 # that the final one inserted is the first one
3341 # returned in an ordered search of the table.
99870f4d
KW
3342 # => anything else is the same as => $IF_NOT_EQUIVALENT
3343 #
c1739a4a
KW
3344 # "same value" means identical for non-type-0 ranges, and it means
3345 # having the same standard forms for type-0 ranges.
99870f4d
KW
3346
3347 return Carp::carp_too_few_args(\@_, 5) if main::DEBUG && @_ < 5;
3348
3349 my $self = shift;
3350 my $operation = shift; # '+' for add/replace; '-' for delete;
3351 my $start = shift;
3352 my $end = shift;
3353 my $value = shift;
3354
3355 my %args = @_;
3356
3357 $value = "" if not defined $value; # warning: $value can be "0"
3358
3359 my $replace = delete $args{'Replace'};
3360 $replace = $IF_NOT_EQUIVALENT unless defined $replace;
3361
3362 my $type = delete $args{'Type'};
3363 $type = 0 unless defined $type;
3364
3365 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3366
ffe43484 3367 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3368
3369 if ($operation ne '+' && $operation ne '-') {
3370 Carp::my_carp_bug("$owner_name_of{$addr}First parameter to _add_delete must be '+' or '-'. No action taken.");
3371 return;
3372 }
3373 unless (defined $start && defined $end) {
3374 Carp::my_carp_bug("$owner_name_of{$addr}Undefined start and/or end to _add_delete. No action taken.");
3375 return;
3376 }
3377 unless ($end >= $start) {
3378 Carp::my_carp_bug("$owner_name_of{$addr}End of range (" . sprintf("%04X", $end) . ") must not be before start (" . sprintf("%04X", $start) . "). No action taken.");
3379 return;
3380 }
3381 #local $to_trace = 1 if main::DEBUG;
3382
3383 if ($operation eq '-') {
3384 if ($replace != $IF_NOT_EQUIVALENT) {
3385 Carp::my_carp_bug("$owner_name_of{$addr}Replace => \$IF_NOT_EQUIVALENT is required when deleting a range from a range list. Assuming Replace => \$IF_NOT_EQUIVALENT.");
3386 $replace = $IF_NOT_EQUIVALENT;
3387 }
3388 if ($type) {
3389 Carp::my_carp_bug("$owner_name_of{$addr}Type => 0 is required when deleting a range from a range list. Assuming Type => 0.");
3390 $type = 0;
3391 }
3392 if ($value ne "") {
3393 Carp::my_carp_bug("$owner_name_of{$addr}Value => \"\" is required when deleting a range from a range list. Assuming Value => \"\".");
3394 $value = "";
3395 }
3396 }
3397
3398 my $r = $ranges{$addr}; # The current list of ranges
3399 my $range_list_size = scalar @$r; # And its size
3400 my $max = $max{$addr}; # The current high code point in
3401 # the list of ranges
3402
3403 # Do a special case requiring fewer machine cycles when the new range
3404 # starts after the current highest point. The Unicode input data is
3405 # structured so this is common.
3406 if ($start > $max) {
3407
3408 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) type=$type" if main::DEBUG && $to_trace;
3409 return if $operation eq '-'; # Deleting a non-existing range is a
3410 # no-op
3411
3412 # If the new range doesn't logically extend the current final one
3413 # in the range list, create a new range at the end of the range
3414 # list. (max cleverly is initialized to a negative number not
3415 # adjacent to 0 if the range list is empty, so even adding a range
3416 # to an empty range list starting at 0 will have this 'if'
3417 # succeed.)
3418 if ($start > $max + 1 # non-adjacent means can't extend.
3419 || @{$r}[-1]->value ne $value # values differ, can't extend.
3420 || @{$r}[-1]->type != $type # types differ, can't extend.
3421 ) {
3422 push @$r, Range->new($start, $end,
3423 Value => $value,
3424 Type => $type);
3425 }
3426 else {
3427
3428 # Here, the new range starts just after the current highest in
3429 # the range list, and they have the same type and value.
3430 # Extend the current range to incorporate the new one.
3431 @{$r}[-1]->set_end($end);
3432 }
3433
3434 # This becomes the new maximum.
3435 $max{$addr} = $end;
3436
3437 return;
3438 }
3439 #local $to_trace = 0 if main::DEBUG;
3440
3441 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) replace=$replace" if main::DEBUG && $to_trace;
3442
3443 # Here, the input range isn't after the whole rest of the range list.
3444 # Most likely 'splice' will be needed. The rest of the routine finds
3445 # the needed splice parameters, and if necessary, does the splice.
3446 # First, find the offset parameter needed by the splice function for
3447 # the input range. Note that the input range may span multiple
3448 # existing ones, but we'll worry about that later. For now, just find
3449 # the beginning. If the input range is to be inserted starting in a
3450 # position not currently in the range list, it must (obviously) come
3451 # just after the range below it, and just before the range above it.
3452 # Slightly less obviously, it will occupy the position currently
3453 # occupied by the range that is to come after it. More formally, we
3454 # are looking for the position, $i, in the array of ranges, such that:
3455 #
3456 # r[$i-1]->start <= r[$i-1]->end < $start < r[$i]->start <= r[$i]->end
3457 #
3458 # (The ordered relationships within existing ranges are also shown in
3459 # the equation above). However, if the start of the input range is
3460 # within an existing range, the splice offset should point to that
3461 # existing range's position in the list; that is $i satisfies a
3462 # somewhat different equation, namely:
3463 #
3464 #r[$i-1]->start <= r[$i-1]->end < r[$i]->start <= $start <= r[$i]->end
3465 #
3466 # More briefly, $start can come before or after r[$i]->start, and at
3467 # this point, we don't know which it will be. However, these
3468 # two equations share these constraints:
3469 #
3470 # r[$i-1]->end < $start <= r[$i]->end
3471 #
3472 # And that is good enough to find $i.
3473
3474 my $i = $self->_search_ranges($start);
3475 if (! defined $i) {
3476 Carp::my_carp_bug("Searching $self for range beginning with $start unexpectedly returned undefined. Operation '$operation' not performed");
3477 return;
3478 }
3479
3480 # The search function returns $i such that:
3481 #
3482 # r[$i-1]->end < $start <= r[$i]->end
3483 #
3484 # That means that $i points to the first range in the range list
3485 # that could possibly be affected by this operation. We still don't
3486 # know if the start of the input range is within r[$i], or if it
3487 # points to empty space between r[$i-1] and r[$i].
3488 trace "[$i] is the beginning splice point. Existing range there is ", $r->[$i] if main::DEBUG && $to_trace;
3489
3490 # Special case the insertion of data that is not to replace any
3491 # existing data.
3492 if ($replace == $NO) { # If $NO, has to be operation '+'
3493 #local $to_trace = 1 if main::DEBUG;
3494 trace "Doesn't replace" if main::DEBUG && $to_trace;
3495
3496 # Here, the new range is to take effect only on those code points
3497 # that aren't already in an existing range. This can be done by
3498 # looking through the existing range list and finding the gaps in
3499 # the ranges that this new range affects, and then calling this
3500 # function recursively on each of those gaps, leaving untouched
3501 # anything already in the list. Gather up a list of the changed
3502 # gaps first so that changes to the internal state as new ranges
3503 # are added won't be a problem.
3504 my @gap_list;
3505
3506 # First, if the starting point of the input range is outside an
3507 # existing one, there is a gap from there to the beginning of the
3508 # existing range -- add a span to fill the part that this new
3509 # range occupies
3510 if ($start < $r->[$i]->start) {
3511 push @gap_list, Range->new($start,
3512 main::min($end,
3513 $r->[$i]->start - 1),
3514 Type => $type);
3515 trace "gap before $r->[$i] [$i], will add", $gap_list[-1] if main::DEBUG && $to_trace;
3516 }
3517
3518 # Then look through the range list for other gaps until we reach
3519 # the highest range affected by the input one.
3520 my $j;
3521 for ($j = $i+1; $j < $range_list_size; $j++) {
3522 trace "j=[$j]", $r->[$j] if main::DEBUG && $to_trace;
3523 last if $end < $r->[$j]->start;
3524
3525 # If there is a gap between when this range starts and the
3526 # previous one ends, add a span to fill it. Note that just
3527 # because there are two ranges doesn't mean there is a
3528 # non-zero gap between them. It could be that they have
3529 # different values or types
3530 if ($r->[$j-1]->end + 1 != $r->[$j]->start) {
3531 push @gap_list,
3532 Range->new($r->[$j-1]->end + 1,
3533 $r->[$j]->start - 1,
3534 Type => $type);
3535 trace "gap between $r->[$j-1] and $r->[$j] [$j], will add: $gap_list[-1]" if main::DEBUG && $to_trace;
3536 }
3537 }
3538
3539 # Here, we have either found an existing range in the range list,
3540 # beyond the area affected by the input one, or we fell off the
3541 # end of the loop because the input range affects the whole rest
3542 # of the range list. In either case, $j is 1 higher than the
3543 # highest affected range. If $j == $i, it means that there are no
3544 # affected ranges, that the entire insertion is in the gap between
3545 # r[$i-1], and r[$i], which we already have taken care of before
3546 # the loop.
3547 # On the other hand, if there are affected ranges, it might be
3548 # that there is a gap that needs filling after the final such
3549 # range to the end of the input range
3550 if ($r->[$j-1]->end < $end) {
3551 push @gap_list, Range->new(main::max($start,
3552 $r->[$j-1]->end + 1),
3553 $end,
3554 Type => $type);
3555 trace "gap after $r->[$j-1], will add $gap_list[-1]" if main::DEBUG && $to_trace;
3556 }
3557
3558 # Call recursively to fill in all the gaps.
3559 foreach my $gap (@gap_list) {
3560 $self->_add_delete($operation,
3561 $gap->start,
3562 $gap->end,
3563 $value,
3564 Type => $type);
3565 }
3566
3567 return;
3568 }
3569
53d84487
KW
3570 # Here, we have taken care of the case where $replace is $NO.
3571 # Remember that here, r[$i-1]->end < $start <= r[$i]->end
3572 # If inserting a multiple record, this is where it goes, before the
3573 # first (if any) existing one. This implies an insertion, and no
3574 # change to any existing ranges. Note that $i can be -1 if this new
3575 # range doesn't actually duplicate any existing, and comes at the
3576 # beginning of the list.
3577 if ($replace == $MULTIPLE) {
3578
3579 if ($start != $end) {
3580 Carp::my_carp_bug("$owner_name_of{$addr}Can't cope with adding a multiple record when the range ($start..$end) contains more than one code point. No action taken.");
3581 return;
3582 }
3583
3584 # Don't add an exact duplicate, as it isn't really a multiple
3585 if ($end >= $r->[$i]->start) {
1f6798c4
KW
3586 my $existing_value = $r->[$i]->value;
3587 my $existing_type = $r->[$i]->type;
3588 return if $value eq $existing_value && $type eq $existing_type;
3589
3590 # If the multiple value is part of an existing range, we want
3591 # to split up that range, so that only the single code point
3592 # is affected. To do this, we first call ourselves
3593 # recursively to delete that code point from the table, having
3594 # preserved its current data above. Then we call ourselves
3595 # recursively again to add the new multiple, which we know by
3596 # the test just above is different than the current code
3597 # point's value, so it will become a range containing a single
3598 # code point: just itself. Finally, we add back in the
3599 # pre-existing code point, which will again be a single code
3600 # point range. Because 'i' likely will have changed as a
3601 # result of these operations, we can't just continue on, but
3602 # do this operation recursively as well.
53d84487 3603 if ($r->[$i]->start != $r->[$i]->end) {
1f6798c4
KW
3604 $self->_add_delete('-', $start, $end, "");
3605 $self->_add_delete('+', $start, $end, $value, Type => $type);
3606 return $self->_add_delete('+', $start, $end, $existing_value, Type => $existing_type, Replace => $MULTIPLE);
53d84487 3607 }
53d84487
KW
3608 }
3609
3610 trace "Adding multiple record at $i with $start..$end, $value" if main::DEBUG && $to_trace;
3611 my @return = splice @$r,
3612 $i,
3613 0,
3614 Range->new($start,
3615 $end,
3616 Value => $value,
3617 Type => $type);
3618 if (main::DEBUG && $to_trace) {
3619 trace "After splice:";
3620 trace 'i-2=[', $i-2, ']', $r->[$i-2] if $i >= 2;
3621 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i >= 1;
3622 trace "i =[", $i, "]", $r->[$i] if $i >= 0;
3623 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < @$r - 1;
3624 trace 'i+2=[', $i+2, ']', $r->[$i+2] if $i < @$r - 2;
3625 trace 'i+3=[', $i+3, ']', $r->[$i+3] if $i < @$r - 3;
3626 }
3627 return @return;
3628 }
3629
3630 # Here, we have taken care of $NO and $MULTIPLE replaces. This leaves
3631 # delete, insert, and replace either unconditionally or if not
3632 # equivalent. $i still points to the first potential affected range.
3633 # Now find the highest range affected, which will determine the length
3634 # parameter to splice. (The input range can span multiple existing
3635 # ones.) If this isn't a deletion, while we are looking through the
3636 # range list, see also if this is a replacement rather than a clean
3637 # insertion; that is if it will change the values of at least one
3638 # existing range. Start off assuming it is an insert, until find it
3639 # isn't.
3640 my $clean_insert = $operation eq '+';
99870f4d
KW
3641 my $j; # This will point to the highest affected range
3642
3643 # For non-zero types, the standard form is the value itself;
3644 my $standard_form = ($type) ? $value : main::standardize($value);
3645
3646 for ($j = $i; $j < $range_list_size; $j++) {
3647 trace "Looking for highest affected range; the one at $j is ", $r->[$j] if main::DEBUG && $to_trace;
3648
3649 # If find a range that it doesn't overlap into, we can stop
3650 # searching
3651 last if $end < $r->[$j]->start;
3652
969a34cc
KW
3653 # Here, overlaps the range at $j. If the values don't match,
3654 # and so far we think this is a clean insertion, it becomes a
3655 # non-clean insertion, i.e., a 'change' or 'replace' instead.
3656 if ($clean_insert) {
99870f4d 3657 if ($r->[$j]->standard_form ne $standard_form) {
969a34cc 3658 $clean_insert = 0;
56343c78
KW
3659 if ($replace == $CROAK) {
3660 main::croak("The range to add "
3661 . sprintf("%04X", $start)
3662 . '-'
3663 . sprintf("%04X", $end)
3664 . " with value '$value' overlaps an existing range $r->[$j]");
3665 }
99870f4d
KW
3666 }
3667 else {
3668
3669 # Here, the two values are essentially the same. If the
3670 # two are actually identical, replacing wouldn't change
3671 # anything so skip it.
3672 my $pre_existing = $r->[$j]->value;
3673 if ($pre_existing ne $value) {
3674
3675 # Here the new and old standardized values are the
3676 # same, but the non-standardized values aren't. If
3677 # replacing unconditionally, then replace
3678 if( $replace == $UNCONDITIONALLY) {
969a34cc 3679 $clean_insert = 0;
99870f4d
KW
3680 }
3681 else {
3682
3683 # Here, are replacing conditionally. Decide to
3684 # replace or not based on which appears to look
3685 # the "nicest". If one is mixed case and the
3686 # other isn't, choose the mixed case one.
3687 my $new_mixed = $value =~ /[A-Z]/
3688 && $value =~ /[a-z]/;
3689 my $old_mixed = $pre_existing =~ /[A-Z]/
3690 && $pre_existing =~ /[a-z]/;
3691
3692 if ($old_mixed != $new_mixed) {
969a34cc 3693 $clean_insert = 0 if $new_mixed;
99870f4d 3694 if (main::DEBUG && $to_trace) {
969a34cc
KW
3695 if ($clean_insert) {
3696 trace "Retaining $pre_existing over $value";
99870f4d
KW
3697 }
3698 else {
969a34cc 3699 trace "Replacing $pre_existing with $value";
99870f4d
KW
3700 }
3701 }
3702 }
3703 else {
3704
3705 # Here casing wasn't different between the two.
3706 # If one has hyphens or underscores and the
3707 # other doesn't, choose the one with the
3708 # punctuation.
3709 my $new_punct = $value =~ /[-_]/;
3710 my $old_punct = $pre_existing =~ /[-_]/;
3711
3712 if ($old_punct != $new_punct) {
969a34cc 3713 $clean_insert = 0 if $new_punct;
99870f4d 3714 if (main::DEBUG && $to_trace) {
969a34cc
KW
3715 if ($clean_insert) {
3716 trace "Retaining $pre_existing over $value";
99870f4d
KW
3717 }
3718 else {
969a34cc 3719 trace "Replacing $pre_existing with $value";
99870f4d
KW
3720 }
3721 }
3722 } # else existing one is just as "good";
3723 # retain it to save cycles.
3724 }
3725 }
3726 }
3727 }
3728 }
3729 } # End of loop looking for highest affected range.
3730
3731 # Here, $j points to one beyond the highest range that this insertion
3732 # affects (hence to beyond the range list if that range is the final
3733 # one in the range list).
3734
3735 # The splice length is all the affected ranges. Get it before
3736 # subtracting, for efficiency, so we don't have to later add 1.
3737 my $length = $j - $i;
3738
3739 $j--; # $j now points to the highest affected range.
3740 trace "Final affected range is $j: $r->[$j]" if main::DEBUG && $to_trace;
3741
99870f4d
KW
3742 # Here, have taken care of $NO and $MULTIPLE replaces.
3743 # $j points to the highest affected range. But it can be < $i or even
3744 # -1. These happen only if the insertion is entirely in the gap
3745 # between r[$i-1] and r[$i]. Here's why: j < i means that the j loop
3746 # above exited first time through with $end < $r->[$i]->start. (And
3747 # then we subtracted one from j) This implies also that $start <
3748 # $r->[$i]->start, but we know from above that $r->[$i-1]->end <
3749 # $start, so the entire input range is in the gap.
3750 if ($j < $i) {
3751
3752 # Here the entire input range is in the gap before $i.
3753
3754 if (main::DEBUG && $to_trace) {
3755 if ($i) {
3756 trace "Entire range is between $r->[$i-1] and $r->[$i]";
3757 }
3758 else {
3759 trace "Entire range is before $r->[$i]";
3760 }
3761 }
3762 return if $operation ne '+'; # Deletion of a non-existent range is
3763 # a no-op
3764 }
3765 else {
3766
969a34cc
KW
3767 # Here part of the input range is not in the gap before $i. Thus,
3768 # there is at least one affected one, and $j points to the highest
3769 # such one.
99870f4d
KW
3770
3771 # At this point, here is the situation:
3772 # This is not an insertion of a multiple, nor of tentative ($NO)
3773 # data.
3774 # $i points to the first element in the current range list that
3775 # may be affected by this operation. In fact, we know
3776 # that the range at $i is affected because we are in
3777 # the else branch of this 'if'
3778 # $j points to the highest affected range.
3779 # In other words,
3780 # r[$i-1]->end < $start <= r[$i]->end
3781 # And:
3782 # r[$i-1]->end < $start <= $end <= r[$j]->end
3783 #
3784 # Also:
969a34cc
KW
3785 # $clean_insert is a boolean which is set true if and only if
3786 # this is a "clean insertion", i.e., not a change nor a
3787 # deletion (multiple was handled above).
99870f4d
KW
3788
3789 # We now have enough information to decide if this call is a no-op
969a34cc
KW
3790 # or not. It is a no-op if this is an insertion of already
3791 # existing data.
99870f4d 3792
969a34cc 3793 if (main::DEBUG && $to_trace && $clean_insert
99870f4d
KW
3794 && $i == $j
3795 && $start >= $r->[$i]->start)
3796 {
3797 trace "no-op";
3798 }
969a34cc 3799 return if $clean_insert
99870f4d
KW
3800 && $i == $j # more than one affected range => not no-op
3801
3802 # Here, r[$i-1]->end < $start <= $end <= r[$i]->end
3803 # Further, $start and/or $end is >= r[$i]->start
3804 # The test below hence guarantees that
3805 # r[$i]->start < $start <= $end <= r[$i]->end
3806 # This means the input range is contained entirely in
3807 # the one at $i, so is a no-op
3808 && $start >= $r->[$i]->start;
3809 }
3810
3811 # Here, we know that some action will have to be taken. We have
3812 # calculated the offset and length (though adjustments may be needed)
3813 # for the splice. Now start constructing the replacement list.
3814 my @replacement;
3815 my $splice_start = $i;
3816
3817 my $extends_below;
3818 my $extends_above;
3819
3820 # See if should extend any adjacent ranges.
3821 if ($operation eq '-') { # Don't extend deletions
3822 $extends_below = $extends_above = 0;
3823 }
3824 else { # Here, should extend any adjacent ranges. See if there are
3825 # any.
3826 $extends_below = ($i > 0
3827 # can't extend unless adjacent
3828 && $r->[$i-1]->end == $start -1
3829 # can't extend unless are same standard value
3830 && $r->[$i-1]->standard_form eq $standard_form
3831 # can't extend unless share type
3832 && $r->[$i-1]->type == $type);
3833 $extends_above = ($j+1 < $range_list_size
3834 && $r->[$j+1]->start == $end +1
3835 && $r->[$j+1]->standard_form eq $standard_form
23822bda 3836 && $r->[$j+1]->type == $type);
99870f4d
KW
3837 }
3838 if ($extends_below && $extends_above) { # Adds to both
3839 $splice_start--; # start replace at element below
3840 $length += 2; # will replace on both sides
3841 trace "Extends both below and above ranges" if main::DEBUG && $to_trace;
3842
3843 # The result will fill in any gap, replacing both sides, and
3844 # create one large range.
3845 @replacement = Range->new($r->[$i-1]->start,
3846 $r->[$j+1]->end,
3847 Value => $value,
3848 Type => $type);
3849 }
3850 else {
3851
3852 # Here we know that the result won't just be the conglomeration of
3853 # a new range with both its adjacent neighbors. But it could
3854 # extend one of them.
3855
3856 if ($extends_below) {
3857
3858 # Here the new element adds to the one below, but not to the
3859 # one above. If inserting, and only to that one range, can
3860 # just change its ending to include the new one.
969a34cc 3861 if ($length == 0 && $clean_insert) {
99870f4d
KW
3862 $r->[$i-1]->set_end($end);
3863 trace "inserted range extends range to below so it is now $r->[$i-1]" if main::DEBUG && $to_trace;