This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
podcheck.t: white-space only: indent properly
[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
232ed87f 43# a pod file and .t files, depending on option parameters.
99870f4d
KW
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#
232ed87f
KW
55# This program works on all releases of Unicode so far. The outputs have been
56# scrutinized most intently for release 5.1. The others have been checked for
57# somewhat more than just sanity. It can handle all non-provisional Unicode
58# character properties in those releases.
99870f4d 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
232ed87f
KW
68# value". (Some more recently defined properties, map a code point to a set
69# of values.)
99870f4d
KW
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
232ed87f 122# that is algorithmically determinable from its code point (and the reverse).
99870f4d
KW
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.
232ed87f
KW
134# Again, this is so that methods can be defined on one and not the others so
135# as to prevent operating on them in incorrect ways.
99870f4d
KW
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
232ed87f 157# code points that map to N. (For each binary property, a third table is also
99870f4d 158# generated for the pseudo Perl property. It contains the identical code
232ed87f
KW
159# points as the Y table, but can be written in regular expressions, not in the
160# compound form, but in a "single" form like \p{IsUppercase}.) Many
161# properties are binary, but some properties have several possible values,
162# some have many, and properties like Name have a different value for every
163# named code point. Those will not, unless the controlling lists are changed,
164# have their match tables written out. But all the ones which can be used in
165# regular expression \p{} and \P{} constructs will. Prior to 5.14, generally
166# a property would have either its map table or its match tables written but
167# not both. Again, what gets written is controlled by lists which can easily
168# be changed. Starting in 5.14, advantage was taken of this, and all the map
169# tables needed to reconstruct the Unicode db are now written out, while
170# suppressing the Unicode .txt files that contain the data. Our tables are
171# much more compact than the .txt files, so a significant space savings was
172# achieved. Also, tables are not written out that are trivially derivable
173# from tables that do get written. So, there typically is no file containing
174# the code points not matched by a binary property (the table for \P{} versus
175# lowercase \p{}), since you just need to invert the True table to get the
176# False table.
177
178# Properties have a 'Type', like 'binary', or 'string', or 'enum' depending on
179# how many match tables there are and the content of the maps. This 'Type' is
c12f2655
KW
180# different than a range 'Type', so don't get confused by the two concepts
181# having the same name.
678f13d5 182#
99870f4d
KW
183# For information about the Unicode properties, see Unicode's UAX44 document:
184
185my $unicode_reference_url = 'http://www.unicode.org/reports/tr44/';
186
187# As stated earlier, this program will work on any release of Unicode so far.
188# Most obvious problems in earlier data have NOT been corrected except when
be864b6c 189# necessary to make Perl or this program work reasonably, and to keep out
232ed87f
KW
190# potential security issues. For example, no folding information was given in
191# early releases, so this program substitutes lower case instead, just so that
192# a regular expression with the /i option will do something that actually
193# gives the right results in many cases. There are also a couple other
194# corrections for version 1.1.5, commented at the point they are made. As an
195# example of corrections that weren't made (but could be) is this statement
196# from DerivedAge.txt: "The supplementary private use code points and the
197# non-character code points were assigned in version 2.0, but not specifically
198# listed in the UCD until versions 3.0 and 3.1 respectively." (To be precise
199# it was 3.0.1 not 3.0.0) More information on Unicode version glitches is
200# further down in these introductory comments.
99870f4d 201#
232ed87f
KW
202# This program works on all non-provisional properties as of the current
203# Unicode release, though the files for some are suppressed for various
204# reasons. You can change which are output by changing lists in this program.
678f13d5 205#
dc85bd38 206# The old version of mktables emphasized the term "Fuzzy" to mean Unicode's
99870f4d
KW
207# loose matchings rules (from Unicode TR18):
208#
209# The recommended names for UCD properties and property values are in
210# PropertyAliases.txt [Prop] and PropertyValueAliases.txt
211# [PropValue]. There are both abbreviated names and longer, more
212# descriptive names. It is strongly recommended that both names be
213# recognized, and that loose matching of property names be used,
214# whereby the case distinctions, whitespace, hyphens, and underbar
215# are ignored.
232ed87f 216#
99870f4d
KW
217# The program still allows Fuzzy to override its determination of if loose
218# matching should be used, but it isn't currently used, as it is no longer
219# needed; the calculations it makes are good enough.
678f13d5 220#
99870f4d
KW
221# SUMMARY OF HOW IT WORKS:
222#
223# Process arguments
224#
225# A list is constructed containing each input file that is to be processed
226#
227# Each file on the list is processed in a loop, using the associated handler
228# code for each:
229# The PropertyAliases.txt and PropValueAliases.txt files are processed
230# first. These files name the properties and property values.
231# Objects are created of all the property and property value names
232# that the rest of the input should expect, including all synonyms.
233# The other input files give mappings from properties to property
234# values. That is, they list code points and say what the mapping
235# is under the given property. Some files give the mappings for
236# just one property; and some for many. This program goes through
232ed87f
KW
237# each file and populates the properties and their map tables from
238# them. Some properties are listed in more than one file, and
239# Unicode has set up a precedence as to which has priority if there
240# is a conflict. Thus the order of processing matters, and this
241# program handles the conflict possibility by processing the
242# overriding input files last, so that if necessary they replace
243# earlier values.
99870f4d
KW
244# After this is all done, the program creates the property mappings not
245# furnished by Unicode, but derivable from what it does give.
246# The tables of code points that match each property value in each
247# property that is accessible by regular expressions are created.
248# The Perl-defined properties are created and populated. Many of these
249# require data determined from the earlier steps
250# Any Perl-defined synonyms are created, and name clashes between Perl
678f13d5 251# and Unicode are reconciled and warned about.
99870f4d
KW
252# All the properties are written to files
253# Any other files are written, and final warnings issued.
678f13d5 254#
99870f4d
KW
255# For clarity, a number of operators have been overloaded to work on tables:
256# ~ means invert (take all characters not in the set). The more
257# conventional '!' is not used because of the possibility of confusing
258# it with the actual boolean operation.
259# + means union
260# - means subtraction
261# & means intersection
262# The precedence of these is the order listed. Parentheses should be
263# copiously used. These are not a general scheme. The operations aren't
264# defined for a number of things, deliberately, to avoid getting into trouble.
265# Operations are done on references and affect the underlying structures, so
266# that the copy constructors for them have been overloaded to not return a new
267# clone, but the input object itself.
678f13d5 268#
99870f4d
KW
269# The bool operator is deliberately not overloaded to avoid confusion with
270# "should it mean if the object merely exists, or also is non-empty?".
99870f4d
KW
271#
272# WHY CERTAIN DESIGN DECISIONS WERE MADE
678f13d5
KW
273#
274# This program needs to be able to run under miniperl. Therefore, it uses a
275# minimum of other modules, and hence implements some things itself that could
276# be gotten from CPAN
277#
278# This program uses inputs published by the Unicode Consortium. These can
279# change incompatibly between releases without the Perl maintainers realizing
280# it. Therefore this program is now designed to try to flag these. It looks
281# at the directories where the inputs are, and flags any unrecognized files.
282# It keeps track of all the properties in the files it handles, and flags any
283# that it doesn't know how to handle. It also flags any input lines that
284# don't match the expected syntax, among other checks.
285#
286# It is also designed so if a new input file matches one of the known
287# templates, one hopefully just needs to add it to a list to have it
288# processed.
289#
290# As mentioned earlier, some properties are given in more than one file. In
291# particular, the files in the extracted directory are supposedly just
292# reformattings of the others. But they contain information not easily
293# derivable from the other files, including results for Unihan, which this
294# program doesn't ordinarily look at, and for unassigned code points. They
295# also have historically had errors or been incomplete. In an attempt to
296# create the best possible data, this program thus processes them first to
297# glean information missing from the other files; then processes those other
298# files to override any errors in the extracted ones. Much of the design was
299# driven by this need to store things and then possibly override them.
300#
301# It tries to keep fatal errors to a minimum, to generate something usable for
302# testing purposes. It always looks for files that could be inputs, and will
303# warn about any that it doesn't know how to handle (the -q option suppresses
304# the warning).
99870f4d 305#
678f13d5
KW
306# Why is there more than one type of range?
307# This simplified things. There are some very specialized code points that
308# have to be handled specially for output, such as Hangul syllable names.
309# By creating a range type (done late in the development process), it
310# allowed this to be stored with the range, and overridden by other input.
311# Originally these were stored in another data structure, and it became a
312# mess trying to decide if a second file that was for the same property was
313# overriding the earlier one or not.
314#
315# Why are there two kinds of tables, match and map?
316# (And there is a base class shared by the two as well.) As stated above,
317# they actually are for different things. Development proceeded much more
318# smoothly when I (khw) realized the distinction. Map tables are used to
319# give the property value for every code point (actually every code point
320# that doesn't map to a default value). Match tables are used for regular
321# expression matches, and are essentially the inverse mapping. Separating
322# the two allows more specialized methods, and error checks so that one
323# can't just take the intersection of two map tables, for example, as that
324# is nonsensical.
99870f4d 325#
232ed87f
KW
326# What about 'fate' and 'status'. The concept of a table's fate was created
327# late when it became clear that something more was needed. The difference
328# between this and 'status' is unclean, and could be improved if someone
329# wanted to spend the effort.
330#
23e33b60
KW
331# DEBUGGING
332#
678f13d5
KW
333# This program is written so it will run under miniperl. Occasionally changes
334# will cause an error where the backtrace doesn't work well under miniperl.
335# To diagnose the problem, you can instead run it under regular perl, if you
336# have one compiled.
337#
338# There is a good trace facility. To enable it, first sub DEBUG must be set
339# to return true. Then a line like
340#
341# local $to_trace = 1 if main::DEBUG;
342#
232ed87f
KW
343# can be added to enable tracing in its lexical scope (plus dynamic) or until
344# you insert another line:
678f13d5
KW
345#
346# local $to_trace = 0 if main::DEBUG;
347#
232ed87f 348# To actually trace, use a line like "trace $a, @b, %c, ...;
678f13d5
KW
349#
350# Some of the more complex subroutines already have trace statements in them.
351# Permanent trace statements should be like:
352#
353# trace ... if main::DEBUG && $to_trace;
354#
355# If there is just one or a few files that you're debugging, you can easily
356# cause most everything else to be skipped. Change the line
357#
358# my $debug_skip = 0;
359#
360# to 1, and every file whose object is in @input_file_objects and doesn't have
232ed87f
KW
361# a, 'non_skip => 1,' in its constructor will be skipped. However, skipping
362# Jamo.txt or UnicodeData.txt will likely cause fatal errors.
678f13d5 363#
b4a0206c 364# To compare the output tables, it may be useful to specify the -annotate
c4019d52
KW
365# flag. This causes the tables to expand so there is one entry for each
366# non-algorithmically named code point giving, currently its name, and its
367# graphic representation if printable (and you have a font that knows about
368# it). This makes it easier to see what the particular code points are in
369# each output table. The tables are usable, but because they don't have
370# ranges (for the most part), a Perl using them will run slower. Non-named
371# code points are annotated with a description of their status, and contiguous
372# ones with the same description will be output as a range rather than
373# individually. Algorithmically named characters are also output as ranges,
374# except when there are just a few contiguous ones.
375#
99870f4d
KW
376# FUTURE ISSUES
377#
378# The program would break if Unicode were to change its names so that
379# interior white space, underscores, or dashes differences were significant
380# within property and property value names.
381#
382# It might be easier to use the xml versions of the UCD if this program ever
383# would need heavy revision, and the ability to handle old versions was not
384# required.
385#
386# There is the potential for name collisions, in that Perl has chosen names
387# that Unicode could decide it also likes. There have been such collisions in
388# the past, with mostly Perl deciding to adopt the Unicode definition of the
389# name. However in the 5.2 Unicode beta testing, there were a number of such
390# collisions, which were withdrawn before the final release, because of Perl's
391# and other's protests. These all involved new properties which began with
392# 'Is'. Based on the protests, Unicode is unlikely to try that again. Also,
393# many of the Perl-defined synonyms, like Any, Word, etc, are listed in a
394# Unicode document, so they are unlikely to be used by Unicode for another
395# purpose. However, they might try something beginning with 'In', or use any
396# of the other Perl-defined properties. This program will warn you of name
397# collisions, and refuse to generate tables with them, but manual intervention
398# will be required in this event. One scheme that could be implemented, if
399# necessary, would be to have this program generate another file, or add a
400# field to mktables.lst that gives the date of first definition of a property.
401# Each new release of Unicode would use that file as a basis for the next
402# iteration. And the Perl synonym addition code could sort based on the age
403# of the property, so older properties get priority, and newer ones that clash
404# would be refused; hence existing code would not be impacted, and some other
405# synonym would have to be used for the new property. This is ugly, and
406# manual intervention would certainly be easier to do in the short run; lets
407# hope it never comes to this.
678f13d5 408#
99870f4d
KW
409# A NOTE ON UNIHAN
410#
411# This program can generate tables from the Unihan database. But it doesn't
412# by default, letting the CPAN module Unicode::Unihan handle them. Prior to
413# version 5.2, this database was in a single file, Unihan.txt. In 5.2 the
414# database was split into 8 different files, all beginning with the letters
415# 'Unihan'. This program will read those file(s) if present, but it needs to
416# know which of the many properties in the file(s) should have tables created
417# for them. It will create tables for any properties listed in
418# PropertyAliases.txt and PropValueAliases.txt, plus any listed in the
419# @cjk_properties array and the @cjk_property_values array. Thus, if a
420# property you want is not in those files of the release you are building
421# against, you must add it to those two arrays. Starting in 4.0, the
422# Unicode_Radical_Stroke was listed in those files, so if the Unihan database
423# is present in the directory, a table will be generated for that property.
424# In 5.2, several more properties were added. For your convenience, the two
5f7264c7 425# arrays are initialized with all the 6.0 listed properties that are also in
99870f4d
KW
426# earlier releases. But these are commented out. You can just uncomment the
427# ones you want, or use them as a template for adding entries for other
428# properties.
429#
430# You may need to adjust the entries to suit your purposes. setup_unihan(),
431# and filter_unihan_line() are the functions where this is done. This program
432# already does some adjusting to make the lines look more like the rest of the
433# Unicode DB; You can see what that is in filter_unihan_line()
434#
435# There is a bug in the 3.2 data file in which some values for the
436# kPrimaryNumeric property have commas and an unexpected comment. A filter
437# could be added for these; or for a particular installation, the Unihan.txt
438# file could be edited to fix them.
99870f4d 439#
678f13d5
KW
440# HOW TO ADD A FILE TO BE PROCESSED
441#
442# A new file from Unicode needs to have an object constructed for it in
443# @input_file_objects, probably at the end or at the end of the extracted
444# ones. The program should warn you if its name will clash with others on
445# restrictive file systems, like DOS. If so, figure out a better name, and
446# add lines to the README.perl file giving that. If the file is a character
232ed87f 447# property, it should be in the format that Unicode has implicitly
678f13d5
KW
448# standardized for such files for the more recently introduced ones.
449# If so, the Input_file constructor for @input_file_objects can just be the
450# file name and release it first appeared in. If not, then it should be
451# possible to construct an each_line_handler() to massage the line into the
452# standardized form.
453#
454# For non-character properties, more code will be needed. You can look at
455# the existing entries for clues.
456#
457# UNICODE VERSIONS NOTES
458#
459# The Unicode UCD has had a number of errors in it over the versions. And
460# these remain, by policy, in the standard for that version. Therefore it is
461# risky to correct them, because code may be expecting the error. So this
462# program doesn't generally make changes, unless the error breaks the Perl
463# core. As an example, some versions of 2.1.x Jamo.txt have the wrong value
464# for U+1105, which causes real problems for the algorithms for Jamo
465# calculations, so it is changed here.
466#
467# But it isn't so clear cut as to what to do about concepts that are
468# introduced in a later release; should they extend back to earlier releases
469# where the concept just didn't exist? It was easier to do this than to not,
470# so that's what was done. For example, the default value for code points not
471# in the files for various properties was probably undefined until changed by
472# some version. No_Block for blocks is such an example. This program will
473# assign No_Block even in Unicode versions that didn't have it. This has the
474# benefit that code being written doesn't have to special case earlier
475# versions; and the detriment that it doesn't match the Standard precisely for
476# the affected versions.
477#
478# Here are some observations about some of the issues in early versions:
479#
232ed87f
KW
480# Prior to version 3.0, there were 3 character decompositions. These are not
481# handled by Unicode::Normalize, nor will it compile when presented a version
482# that has them. However, you can trivially get it to compile by simply
483# ignoring those decompositions, by changing the croak to a carp. At the time
484# of this writing, the line (in cpan/Unicode-Normalize/mkheader) reads
485#
486# croak("Weird Canonical Decomposition of U+$h");
487#
488# Simply change to a carp. It will compile, but will not know about any three
489# character decomposition.
490
491# The number of code points in \p{alpha=True} halved in 2.1.9. It turns out
492# that the reason is that the CJK block starting at 4E00 was removed from
493# PropList, and was not put back in until 3.1.0. The Perl extension (the
494# single property name \p{alpha}) has the correct values. But the compound
495# form is simply not generated until 3.1, as it can be argued that prior to
496# this release, this was not an official property. The comments for
497# filter_old_style_proplist() give more details.
678f13d5
KW
498#
499# Unicode introduced the synonym Space for White_Space in 4.1. Perl has
500# always had a \p{Space}. In release 3.2 only, they are not synonymous. The
501# reason is that 3.2 introduced U+205F=medium math space, which was not
502# classed as white space, but Perl figured out that it should have been. 4.0
503# reclassified it correctly.
504#
505# Another change between 3.2 and 4.0 is the CCC property value ATBL. In 3.2
232ed87f
KW
506# this was erroneously a synonym for 202 (it should be 200). In 4.0, ATB
507# became 202, and ATBL was left with no code points, as all the ones that
508# mapped to 202 stayed mapped to 202. Thus if your program used the numeric
509# name for the class, it would not have been affected, but if it used the
510# mnemonic, it would have been.
678f13d5
KW
511#
512# \p{Script=Hrkt} (Katakana_Or_Hiragana) came in 4.0.1. Before that code
513# points which eventually came to have this script property value, instead
514# mapped to "Unknown". But in the next release all these code points were
515# moved to \p{sc=common} instead.
99870f4d
KW
516#
517# The default for missing code points for BidiClass is complicated. Starting
518# in 3.1.1, the derived file DBidiClass.txt handles this, but this program
519# tries to do the best it can for earlier releases. It is done in
520# process_PropertyAliases()
521#
232ed87f
KW
522# In version 2.1.2, the entry in UnicodeData.txt:
523# 0275;LATIN SMALL LETTER BARRED O;Ll;0;L;;;;;N;;;;019F;
524# should instead be
525# 0275;LATIN SMALL LETTER BARRED O;Ll;0;L;;;;;N;;;019F;;019F
526# Without this change, there are casing problems for this character.
527#
99870f4d
KW
528##############################################################################
529
530my $UNDEF = ':UNDEF:'; # String to print out for undefined values in tracing
531 # and errors
532my $MAX_LINE_WIDTH = 78;
533
534# Debugging aid to skip most files so as to not be distracted by them when
535# concentrating on the ones being debugged. Add
536# non_skip => 1,
537# to the constructor for those files you want processed when you set this.
538# Files with a first version number of 0 are special: they are always
c12f2655
KW
539# processed regardless of the state of this flag. Generally, Jamo.txt and
540# UnicodeData.txt must not be skipped if you want this program to not die
541# before normal completion.
99870f4d
KW
542my $debug_skip = 0;
543
e9c4b4f8
KW
544
545# Normally these are suppressed.
546my $write_Unicode_deprecated_tables = 0;
547
99870f4d
KW
548# Set to 1 to enable tracing.
549our $to_trace = 0;
550
551{ # Closure for trace: debugging aid
552 my $print_caller = 1; # ? Include calling subroutine name
553 my $main_with_colon = 'main::';
554 my $main_colon_length = length($main_with_colon);
555
556 sub trace {
557 return unless $to_trace; # Do nothing if global flag not set
558
559 my @input = @_;
560
561 local $DB::trace = 0;
562 $DB::trace = 0; # Quiet 'used only once' message
563
564 my $line_number;
565
566 # Loop looking up the stack to get the first non-trace caller
567 my $caller_line;
568 my $caller_name;
569 my $i = 0;
570 do {
571 $line_number = $caller_line;
572 (my $pkg, my $file, $caller_line, my $caller) = caller $i++;
573 $caller = $main_with_colon unless defined $caller;
574
575 $caller_name = $caller;
576
577 # get rid of pkg
578 $caller_name =~ s/.*:://;
579 if (substr($caller_name, 0, $main_colon_length)
580 eq $main_with_colon)
581 {
582 $caller_name = substr($caller_name, $main_colon_length);
583 }
584
585 } until ($caller_name ne 'trace');
586
587 # If the stack was empty, we were called from the top level
588 $caller_name = 'main' if ($caller_name eq ""
589 || $caller_name eq 'trace');
590
591 my $output = "";
592 foreach my $string (@input) {
593 #print STDERR __LINE__, ": ", join ", ", @input, "\n";
594 if (ref $string eq 'ARRAY' || ref $string eq 'HASH') {
595 $output .= simple_dumper($string);
596 }
597 else {
598 $string = "$string" if ref $string;
599 $string = $UNDEF unless defined $string;
600 chomp $string;
601 $string = '""' if $string eq "";
602 $output .= " " if $output ne ""
603 && $string ne ""
604 && substr($output, -1, 1) ne " "
605 && substr($string, 0, 1) ne " ";
606 $output .= $string;
607 }
608 }
609
99f78760
KW
610 print STDERR sprintf "%4d: ", $line_number if defined $line_number;
611 print STDERR "$caller_name: " if $print_caller;
99870f4d
KW
612 print STDERR $output, "\n";
613 return;
614 }
615}
616
617# This is for a rarely used development feature that allows you to compare two
618# versions of the Unicode standard without having to deal with changes caused
c12f2655
KW
619# by the code points introduced in the later version. Change the 0 to a
620# string containing a SINGLE dotted Unicode release number (e.g. "2.1"). Only
621# code points introduced in that release and earlier will be used; later ones
622# are thrown away. You use the version number of the earliest one you want to
623# compare; then run this program on directory structures containing each
624# release, and compare the outputs. These outputs will therefore include only
625# the code points common to both releases, and you can see the changes caused
626# just by the underlying release semantic changes. For versions earlier than
627# 3.2, you must copy a version of DAge.txt into the directory.
628my $string_compare_versions = DEBUG && 0; # e.g., "2.1";
99870f4d
KW
629my $compare_versions = DEBUG
630 && $string_compare_versions
631 && pack "C*", split /\./, $string_compare_versions;
632
633sub uniques {
634 # Returns non-duplicated input values. From "Perl Best Practices:
635 # Encapsulated Cleverness". p. 455 in first edition.
636
637 my %seen;
0e407844
NC
638 # Arguably this breaks encapsulation, if the goal is to permit multiple
639 # distinct objects to stringify to the same value, and be interchangeable.
640 # However, for this program, no two objects stringify identically, and all
641 # lists passed to this function are either objects or strings. So this
642 # doesn't affect correctness, but it does give a couple of percent speedup.
643 no overloading;
99870f4d
KW
644 return grep { ! $seen{$_}++ } @_;
645}
646
647$0 = File::Spec->canonpath($0);
648
649my $make_test_script = 0; # ? Should we output a test script
6b5ab373 650my $make_norm_test_script = 0; # ? Should we output a normalization test script
99870f4d
KW
651my $write_unchanged_files = 0; # ? Should we update the output files even if
652 # we don't think they have changed
653my $use_directory = ""; # ? Should we chdir somewhere.
654my $pod_directory; # input directory to store the pod file.
655my $pod_file = 'perluniprops';
656my $t_path; # Path to the .t test file
657my $file_list = 'mktables.lst'; # File to store input and output file names.
658 # This is used to speed up the build, by not
659 # executing the main body of the program if
660 # nothing on the list has changed since the
661 # previous build
662my $make_list = 1; # ? Should we write $file_list. Set to always
663 # make a list so that when the pumpking is
664 # preparing a release, s/he won't have to do
665 # special things
666my $glob_list = 0; # ? Should we try to include unknown .txt files
667 # in the input.
bd9ebcfd
KW
668my $output_range_counts = $debugging_build; # ? Should we include the number
669 # of code points in ranges in
670 # the output
558712cf 671my $annotate = 0; # ? Should character names be in the output
9ef2b94f 672
99870f4d
KW
673# Verbosity levels; 0 is quiet
674my $NORMAL_VERBOSITY = 1;
675my $PROGRESS = 2;
676my $VERBOSE = 3;
677
678my $verbosity = $NORMAL_VERBOSITY;
679
680# Process arguments
681while (@ARGV) {
cf25bb62
JH
682 my $arg = shift @ARGV;
683 if ($arg eq '-v') {
99870f4d
KW
684 $verbosity = $VERBOSE;
685 }
686 elsif ($arg eq '-p') {
687 $verbosity = $PROGRESS;
688 $| = 1; # Flush buffers as we go.
689 }
690 elsif ($arg eq '-q') {
691 $verbosity = 0;
692 }
693 elsif ($arg eq '-w') {
694 $write_unchanged_files = 1; # update the files even if havent changed
695 }
696 elsif ($arg eq '-check') {
6ae7e459
YO
697 my $this = shift @ARGV;
698 my $ok = shift @ARGV;
699 if ($this ne $ok) {
700 print "Skipping as check params are not the same.\n";
701 exit(0);
702 }
00a8df5c 703 }
99870f4d
KW
704 elsif ($arg eq '-P' && defined ($pod_directory = shift)) {
705 -d $pod_directory or croak "Directory '$pod_directory' doesn't exist";
706 }
3df51b85
KW
707 elsif ($arg eq '-maketest' || ($arg eq '-T' && defined ($t_path = shift)))
708 {
99870f4d 709 $make_test_script = 1;
99870f4d 710 }
6b5ab373
KW
711 elsif ($arg eq '-makenormtest')
712 {
713 $make_norm_test_script = 1;
714 }
99870f4d
KW
715 elsif ($arg eq '-makelist') {
716 $make_list = 1;
717 }
718 elsif ($arg eq '-C' && defined ($use_directory = shift)) {
719 -d $use_directory or croak "Unknown directory '$use_directory'";
720 }
721 elsif ($arg eq '-L') {
722
723 # Existence not tested until have chdir'd
724 $file_list = shift;
725 }
726 elsif ($arg eq '-globlist') {
727 $glob_list = 1;
728 }
729 elsif ($arg eq '-c') {
730 $output_range_counts = ! $output_range_counts
731 }
b4a0206c 732 elsif ($arg eq '-annotate') {
558712cf 733 $annotate = 1;
bd9ebcfd
KW
734 $debugging_build = 1;
735 $output_range_counts = 1;
9ef2b94f 736 }
99870f4d
KW
737 else {
738 my $with_c = 'with';
739 $with_c .= 'out' if $output_range_counts; # Complements the state
740 croak <<END;
741usage: $0 [-c|-p|-q|-v|-w] [-C dir] [-L filelist] [ -P pod_dir ]
742 [ -T test_file_path ] [-globlist] [-makelist] [-maketest]
743 [-check A B ]
744 -c : Output comments $with_c number of code points in ranges
745 -q : Quiet Mode: Only output serious warnings.
746 -p : Set verbosity level to normal plus show progress.
747 -v : Set Verbosity level high: Show progress and non-serious
748 warnings
749 -w : Write files regardless
750 -C dir : Change to this directory before proceeding. All relative paths
751 except those specified by the -P and -T options will be done
752 with respect to this directory.
753 -P dir : Output $pod_file file to directory 'dir'.
3df51b85 754 -T path : Create a test script as 'path'; overrides -maketest
99870f4d
KW
755 -L filelist : Use alternate 'filelist' instead of standard one
756 -globlist : Take as input all non-Test *.txt files in current and sub
757 directories
3df51b85
KW
758 -maketest : Make test script 'TestProp.pl' in current (or -C directory),
759 overrides -T
99870f4d 760 -makelist : Rewrite the file list $file_list based on current setup
b4a0206c 761 -annotate : Output an annotation for each character in the table files;
c4019d52 762 useful for debugging mktables, looking at diffs; but is slow,
b318e5e5
KW
763 memory intensive; resulting tables are usable but are slow and
764 very large (and currently fail the Unicode::UCD.t tests).
99870f4d
KW
765 -check A B : Executes $0 only if A and B are the same
766END
767 }
768}
769
770# Stores the most-recently changed file. If none have changed, can skip the
771# build
aeab6150 772my $most_recent = (stat $0)[9]; # Do this before the chdir!
99870f4d
KW
773
774# Change directories now, because need to read 'version' early.
775if ($use_directory) {
3df51b85 776 if ($pod_directory && ! File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d
KW
777 $pod_directory = File::Spec->rel2abs($pod_directory);
778 }
3df51b85 779 if ($t_path && ! File::Spec->file_name_is_absolute($t_path)) {
99870f4d 780 $t_path = File::Spec->rel2abs($t_path);
00a8df5c 781 }
99870f4d 782 chdir $use_directory or croak "Failed to chdir to '$use_directory':$!";
3df51b85 783 if ($pod_directory && File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d 784 $pod_directory = File::Spec->abs2rel($pod_directory);
02b1aeec 785 }
3df51b85 786 if ($t_path && File::Spec->file_name_is_absolute($t_path)) {
99870f4d 787 $t_path = File::Spec->abs2rel($t_path);
02b1aeec 788 }
00a8df5c
YO
789}
790
99870f4d
KW
791# Get Unicode version into regular and v-string. This is done now because
792# various tables below get populated based on it. These tables are populated
793# here to be near the top of the file, and so easily seeable by those needing
794# to modify things.
795open my $VERSION, "<", "version"
796 or croak "$0: can't open required file 'version': $!\n";
797my $string_version = <$VERSION>;
798close $VERSION;
799chomp $string_version;
800my $v_version = pack "C*", split /\./, $string_version; # v string
801
802# The following are the complete names of properties with property values that
803# are known to not match any code points in some versions of Unicode, but that
804# may change in the future so they should be matchable, hence an empty file is
805# generated for them.
806my @tables_that_may_be_empty = (
807 'Joining_Type=Left_Joining',
808 );
809push @tables_that_may_be_empty, 'Script=Common' if $v_version le v4.0.1;
810push @tables_that_may_be_empty, 'Title' if $v_version lt v2.0.0;
811push @tables_that_may_be_empty, 'Script=Katakana_Or_Hiragana'
812 if $v_version ge v4.1.0;
82aed44a
KW
813push @tables_that_may_be_empty, 'Script_Extensions=Katakana_Or_Hiragana'
814 if $v_version ge v6.0.0;
f583b44c
KW
815push @tables_that_may_be_empty, 'Grapheme_Cluster_Break=Prepend'
816 if $v_version ge v6.1.0;
99870f4d
KW
817
818# The lists below are hashes, so the key is the item in the list, and the
819# value is the reason why it is in the list. This makes generation of
820# documentation easier.
821
822my %why_suppressed; # No file generated for these.
823
824# Files aren't generated for empty extraneous properties. This is arguable.
825# Extraneous properties generally come about because a property is no longer
826# used in a newer version of Unicode. If we generated a file without code
827# points, programs that used to work on that property will still execute
828# without errors. It just won't ever match (or will always match, with \P{}).
829# This means that the logic is now likely wrong. I (khw) think its better to
830# find this out by getting an error message. Just move them to the table
831# above to change this behavior
832my %why_suppress_if_empty_warn_if_not = (
833
834 # It is the only property that has ever officially been removed from the
835 # Standard. The database never contained any code points for it.
836 'Special_Case_Condition' => 'Obsolete',
837
838 # Apparently never official, but there were code points in some versions of
839 # old-style PropList.txt
840 'Non_Break' => 'Obsolete',
841);
842
843# These would normally go in the warn table just above, but they were changed
844# a long time before this program was written, so warnings about them are
845# moot.
846if ($v_version gt v3.2.0) {
847 push @tables_that_may_be_empty,
848 'Canonical_Combining_Class=Attached_Below_Left'
849}
850
5f7264c7 851# These are listed in the Property aliases file in 6.0, but Unihan is ignored
99870f4d
KW
852# unless explicitly added.
853if ($v_version ge v5.2.0) {
854 my $unihan = 'Unihan; remove from list if using Unihan';
ea25a9b2 855 foreach my $table (qw (
99870f4d
KW
856 kAccountingNumeric
857 kOtherNumeric
858 kPrimaryNumeric
859 kCompatibilityVariant
860 kIICore
861 kIRG_GSource
862 kIRG_HSource
863 kIRG_JSource
864 kIRG_KPSource
865 kIRG_MSource
866 kIRG_KSource
867 kIRG_TSource
868 kIRG_USource
869 kIRG_VSource
870 kRSUnicode
ea25a9b2 871 ))
99870f4d
KW
872 {
873 $why_suppress_if_empty_warn_if_not{$table} = $unihan;
874 }
ca12659b
NC
875}
876
272501f6
KW
877# Enum values for to_output_map() method in the Map_Table package.
878my $EXTERNAL_MAP = 1;
879my $INTERNAL_MAP = 2;
ce712c88 880my $OUTPUT_ADJUSTED = 3;
272501f6 881
fcf1973c
KW
882# To override computed values for writing the map tables for these properties.
883# The default for enum map tables is to write them out, so that the Unicode
884# .txt files can be removed, but all the data to compute any property value
885# for any code point is available in a more compact form.
886my %global_to_output_map = (
887 # Needed by UCD.pm, but don't want to publicize that it exists, so won't
c12f2655
KW
888 # get stuck supporting it if things change. Since it is a STRING
889 # property, it normally would be listed in the pod, but INTERNAL_MAP
890 # suppresses that.
fcf1973c
KW
891 Unicode_1_Name => $INTERNAL_MAP,
892
893 Present_In => 0, # Suppress, as easily computed from Age
fcf1973c 894 Block => 0, # Suppress, as Blocks.txt is retained.
53d34b6c
KW
895
896 # Suppress, as mapping can be found instead from the
897 # Perl_Decomposition_Mapping file
898 Decomposition_Type => 0,
fcf1973c
KW
899);
900
99870f4d 901# Properties that this program ignores.
230e0c16
KW
902my @unimplemented_properties;
903
904# With this release, it is automatically handled if the Unihan db is
905# downloaded
906push @unimplemented_properties, 'Unicode_Radical_Stroke' if $v_version le v5.2.0;
d73e5302 907
99870f4d
KW
908# There are several types of obsolete properties defined by Unicode. These
909# must be hand-edited for every new Unicode release.
910my %why_deprecated; # Generates a deprecated warning message if used.
911my %why_stabilized; # Documentation only
912my %why_obsolete; # Documentation only
913
914{ # Closure
915 my $simple = 'Perl uses the more complete version of this property';
916 my $unihan = 'Unihan properties are by default not enabled in the Perl core. Instead use CPAN: Unicode::Unihan';
917
918 my $other_properties = 'other properties';
919 my $contributory = "Used by Unicode internally for generating $other_properties and not intended to be used stand-alone";
5d294d41 920 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
921
922 %why_deprecated = (
5f7264c7 923 'Grapheme_Link' => 'Deprecated by Unicode: Duplicates ccc=vr (Canonical_Combining_Class=Virama)',
99870f4d
KW
924 'Jamo_Short_Name' => $contributory,
925 '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',
926 'Other_Alphabetic' => $contributory,
927 'Other_Default_Ignorable_Code_Point' => $contributory,
928 'Other_Grapheme_Extend' => $contributory,
929 'Other_ID_Continue' => $contributory,
930 'Other_ID_Start' => $contributory,
931 'Other_Lowercase' => $contributory,
932 'Other_Math' => $contributory,
933 'Other_Uppercase' => $contributory,
e22aaf5c
KW
934 'Expands_On_NFC' => $why_no_expand,
935 'Expands_On_NFD' => $why_no_expand,
936 'Expands_On_NFKC' => $why_no_expand,
937 'Expands_On_NFKD' => $why_no_expand,
99870f4d
KW
938 );
939
940 %why_suppressed = (
5f7264c7 941 # There is a lib/unicore/Decomposition.pl (used by Normalize.pm) which
99870f4d
KW
942 # contains the same information, but without the algorithmically
943 # determinable Hangul syllables'. This file is not published, so it's
944 # existence is not noted in the comment.
e0b29447 945 'Decomposition_Mapping' => 'Accessible via Unicode::Normalize or Unicode::UCD::prop_invmap()',
99870f4d 946
3111abc0
KW
947 'Indic_Matra_Category' => "Provisional",
948 'Indic_Syllabic_Category' => "Provisional",
949
5f8d1a89
KW
950 # Don't suppress ISO_Comment, as otherwise special handling is needed
951 # to differentiate between it and gc=c, which can be written as 'isc',
952 # which is the same characters as ISO_Comment's short name.
99870f4d 953
fbb93542 954 'Name' => "Accessible via \\N{...} or 'use charnames;' or Unicode::UCD::prop_invmap()",
e0b29447
KW
955
956 'Simple_Case_Folding' => "$simple. Can access this through Unicode::UCD::casefold or Unicode::UCD::prop_invmap()",
957 'Simple_Lowercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
958 'Simple_Titlecase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
959 'Simple_Uppercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
99870f4d 960
5f7264c7 961 FC_NFKC_Closure => 'Supplanted in usage by NFKC_Casefold; otherwise not useful',
99870f4d
KW
962 );
963
1704a0ea
KW
964 foreach my $property (
965
966 # The following are suppressed because they were made contributory
967 # or deprecated by Unicode before Perl ever thought about
968 # supporting them.
969 'Jamo_Short_Name',
970 'Grapheme_Link',
971 'Expands_On_NFC',
972 'Expands_On_NFD',
973 'Expands_On_NFKC',
974 'Expands_On_NFKD',
975
976 # The following are suppressed because they have been marked
977 # as deprecated for a sufficient amount of time
978 'Other_Alphabetic',
979 'Other_Default_Ignorable_Code_Point',
980 'Other_Grapheme_Extend',
981 'Other_ID_Continue',
982 'Other_ID_Start',
983 'Other_Lowercase',
984 'Other_Math',
985 'Other_Uppercase',
e22aaf5c 986 ) {
99870f4d
KW
987 $why_suppressed{$property} = $why_deprecated{$property};
988 }
cf25bb62 989
99870f4d
KW
990 # Customize the message for all the 'Other_' properties
991 foreach my $property (keys %why_deprecated) {
992 next if (my $main_property = $property) !~ s/^Other_//;
993 $why_deprecated{$property} =~ s/$other_properties/the $main_property property (which should be used instead)/;
994 }
995}
996
e9c4b4f8
KW
997if ($write_Unicode_deprecated_tables) {
998 foreach my $property (keys %why_suppressed) {
999 delete $why_suppressed{$property} if $property =~
1000 / ^ Other | Grapheme /x;
1001 }
1002}
1003
99870f4d
KW
1004if ($v_version ge 4.0.0) {
1005 $why_stabilized{'Hyphen'} = 'Use the Line_Break property instead; see www.unicode.org/reports/tr14';
5f7264c7
KW
1006 if ($v_version ge 6.0.0) {
1007 $why_deprecated{'Hyphen'} = 'Supplanted by Line_Break property values; see www.unicode.org/reports/tr14';
1008 }
99870f4d 1009}
5f7264c7 1010if ($v_version ge 5.2.0 && $v_version lt 6.0.0) {
99870f4d 1011 $why_obsolete{'ISO_Comment'} = 'Code points for it have been removed';
5f7264c7 1012 if ($v_version ge 6.0.0) {
63f74647 1013 $why_deprecated{'ISO_Comment'} = 'No longer needed for Unicode\'s internal chart generation; otherwise not useful, and code points for it have been removed';
5f7264c7 1014 }
99870f4d
KW
1015}
1016
1017# Probably obsolete forever
1018if ($v_version ge v4.1.0) {
82aed44a
KW
1019 $why_suppressed{'Script=Katakana_Or_Hiragana'} = 'Obsolete. All code points previously matched by this have been moved to "Script=Common".';
1020}
1021if ($v_version ge v6.0.0) {
2b352efd
KW
1022 $why_suppressed{'Script=Katakana_Or_Hiragana'} .= ' Consider instead using "Script_Extensions=Katakana" or "Script_Extensions=Hiragana (or both)"';
1023 $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
1024}
1025
1026# This program can create files for enumerated-like properties, such as
1027# 'Numeric_Type'. This file would be the same format as for a string
1028# property, with a mapping from code point to its value, so you could look up,
1029# for example, the script a code point is in. But no one so far wants this
1030# mapping, or they have found another way to get it since this is a new
1031# feature. So no file is generated except if it is in this list.
1032my @output_mapped_properties = split "\n", <<END;
1033END
1034
c12f2655
KW
1035# If you are using the Unihan database in a Unicode version before 5.2, you
1036# need to add the properties that you want to extract from it to this table.
1037# For your convenience, the properties in the 6.0 PropertyAliases.txt file are
1038# listed, commented out
99870f4d
KW
1039my @cjk_properties = split "\n", <<'END';
1040#cjkAccountingNumeric; kAccountingNumeric
1041#cjkOtherNumeric; kOtherNumeric
1042#cjkPrimaryNumeric; kPrimaryNumeric
1043#cjkCompatibilityVariant; kCompatibilityVariant
1044#cjkIICore ; kIICore
1045#cjkIRG_GSource; kIRG_GSource
1046#cjkIRG_HSource; kIRG_HSource
1047#cjkIRG_JSource; kIRG_JSource
1048#cjkIRG_KPSource; kIRG_KPSource
1049#cjkIRG_KSource; kIRG_KSource
1050#cjkIRG_TSource; kIRG_TSource
1051#cjkIRG_USource; kIRG_USource
1052#cjkIRG_VSource; kIRG_VSource
1053#cjkRSUnicode; kRSUnicode ; Unicode_Radical_Stroke; URS
1054END
1055
1056# Similarly for the property values. For your convenience, the lines in the
5f7264c7 1057# 6.0 PropertyAliases.txt file are listed. Just remove the first BUT NOT both
c12f2655 1058# '#' marks (for Unicode versions before 5.2)
99870f4d
KW
1059my @cjk_property_values = split "\n", <<'END';
1060## @missing: 0000..10FFFF; cjkAccountingNumeric; NaN
1061## @missing: 0000..10FFFF; cjkCompatibilityVariant; <code point>
1062## @missing: 0000..10FFFF; cjkIICore; <none>
1063## @missing: 0000..10FFFF; cjkIRG_GSource; <none>
1064## @missing: 0000..10FFFF; cjkIRG_HSource; <none>
1065## @missing: 0000..10FFFF; cjkIRG_JSource; <none>
1066## @missing: 0000..10FFFF; cjkIRG_KPSource; <none>
1067## @missing: 0000..10FFFF; cjkIRG_KSource; <none>
1068## @missing: 0000..10FFFF; cjkIRG_TSource; <none>
1069## @missing: 0000..10FFFF; cjkIRG_USource; <none>
1070## @missing: 0000..10FFFF; cjkIRG_VSource; <none>
1071## @missing: 0000..10FFFF; cjkOtherNumeric; NaN
1072## @missing: 0000..10FFFF; cjkPrimaryNumeric; NaN
1073## @missing: 0000..10FFFF; cjkRSUnicode; <none>
1074END
1075
1076# The input files don't list every code point. Those not listed are to be
1077# defaulted to some value. Below are hard-coded what those values are for
1078# non-binary properties as of 5.1. Starting in 5.0, there are
1079# machine-parsable comment lines in the files the give the defaults; so this
1080# list shouldn't have to be extended. The claim is that all missing entries
1081# for binary properties will default to 'N'. Unicode tried to change that in
1082# 5.2, but the beta period produced enough protest that they backed off.
1083#
1084# The defaults for the fields that appear in UnicodeData.txt in this hash must
1085# be in the form that it expects. The others may be synonyms.
1086my $CODE_POINT = '<code point>';
1087my %default_mapping = (
1088 Age => "Unassigned",
1089 # Bidi_Class => Complicated; set in code
1090 Bidi_Mirroring_Glyph => "",
1091 Block => 'No_Block',
1092 Canonical_Combining_Class => 0,
1093 Case_Folding => $CODE_POINT,
1094 Decomposition_Mapping => $CODE_POINT,
1095 Decomposition_Type => 'None',
1096 East_Asian_Width => "Neutral",
1097 FC_NFKC_Closure => $CODE_POINT,
1098 General_Category => 'Cn',
1099 Grapheme_Cluster_Break => 'Other',
1100 Hangul_Syllable_Type => 'NA',
1101 ISO_Comment => "",
1102 Jamo_Short_Name => "",
1103 Joining_Group => "No_Joining_Group",
1104 # Joining_Type => Complicated; set in code
1105 kIICore => 'N', # Is converted to binary
1106 #Line_Break => Complicated; set in code
1107 Lowercase_Mapping => $CODE_POINT,
1108 Name => "",
1109 Name_Alias => "",
1110 NFC_QC => 'Yes',
1111 NFD_QC => 'Yes',
1112 NFKC_QC => 'Yes',
1113 NFKD_QC => 'Yes',
1114 Numeric_Type => 'None',
1115 Numeric_Value => 'NaN',
1116 Script => ($v_version le 4.1.0) ? 'Common' : 'Unknown',
1117 Sentence_Break => 'Other',
1118 Simple_Case_Folding => $CODE_POINT,
1119 Simple_Lowercase_Mapping => $CODE_POINT,
1120 Simple_Titlecase_Mapping => $CODE_POINT,
1121 Simple_Uppercase_Mapping => $CODE_POINT,
1122 Titlecase_Mapping => $CODE_POINT,
1123 Unicode_1_Name => "",
1124 Unicode_Radical_Stroke => "",
1125 Uppercase_Mapping => $CODE_POINT,
1126 Word_Break => 'Other',
1127);
1128
232ed87f
KW
1129# Below are files that Unicode furnishes, but this program ignores, and why.
1130# NormalizationCorrections.txt requires some more explanation. It documents
1131# the cumulative fixes to erroneous normalizations in earlier Unicode
1132# versions. Its main purpose is so that someone running on an earlier version
1133# can use this file to override what got published in that earlier release.
1134# It would be easy for mktables to read and handle this file. But all the
1135# corrections in it should already be in the other files for the release it
1136# is. To get it to actually mean something useful, someone would have to be
1137# using an earlier Unicode release, and copy it to the files for that release
1138# and recomplile. So far there has been no demand to do that, so this hasn't
1139# been implemented.
99870f4d 1140my %ignored_files = (
73ba1144
KW
1141 'CJKRadicals.txt' => 'Maps the kRSUnicode property values to corresponding code points',
1142 'Index.txt' => 'Alphabetical index of Unicode characters',
1143 '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',
1144 'NamesList.txt' => 'Annotated list of characters',
1145 'NormalizationCorrections.txt' => 'Documentation of corrections already incorporated into the Unicode data base',
1146 'Props.txt' => 'Only in very early releases; is a subset of F<PropList.txt> (which is used instead)',
1147 'ReadMe.txt' => 'Documentation',
1148 '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>',
1149 'EmojiSources.txt' => 'Maps certain Unicode code points to their legacy Japanese cell-phone values',
73ba1144
KW
1150 'auxiliary/WordBreakTest.html' => 'Documentation of validation tests',
1151 'auxiliary/SentenceBreakTest.html' => 'Documentation of validation tests',
1152 'auxiliary/GraphemeBreakTest.html' => 'Documentation of validation tests',
1153 'auxiliary/LineBreakTest.html' => 'Documentation of validation tests',
99870f4d
KW
1154);
1155
1fec9f60
KW
1156my %skipped_files; # List of files that we skip
1157
678f13d5 1158### End of externally interesting definitions, except for @input_file_objects
99870f4d
KW
1159
1160my $HEADER=<<"EOF";
1161# !!!!!!! DO NOT EDIT THIS FILE !!!!!!!
3df51b85
KW
1162# This file is machine-generated by $0 from the Unicode
1163# database, Version $string_version. Any changes made here will be lost!
cf25bb62
JH
1164EOF
1165
126c3d4e 1166my $INTERNAL_ONLY_HEADER = <<"EOF";
99870f4d
KW
1167
1168# !!!!!!! INTERNAL PERL USE ONLY !!!!!!!
fac53429
KW
1169# This file is for internal use by core Perl only. The format and even the
1170# name or existence of this file are subject to change without notice. Don't
1171# use it directly.
99870f4d
KW
1172EOF
1173
1174my $DEVELOPMENT_ONLY=<<"EOF";
1175# !!!!!!! DEVELOPMENT USE ONLY !!!!!!!
1176# This file contains information artificially constrained to code points
1177# present in Unicode release $string_compare_versions.
1178# IT CANNOT BE RELIED ON. It is for use during development only and should
23e33b60 1179# not be used for production.
b6922eda
KW
1180
1181EOF
1182
6189eadc
KW
1183my $MAX_UNICODE_CODEPOINT_STRING = "10FFFF";
1184my $MAX_UNICODE_CODEPOINT = hex $MAX_UNICODE_CODEPOINT_STRING;
1185my $MAX_UNICODE_CODEPOINTS = $MAX_UNICODE_CODEPOINT + 1;
99870f4d
KW
1186
1187# Matches legal code point. 4-6 hex numbers, If there are 6, the first
1188# two must be 10; if there are 5, the first must not be a 0. Written this way
92199589
KW
1189# to decrease backtracking. The first regex allows the code point to be at
1190# the end of a word, but to work properly, the word shouldn't end with a valid
1191# hex character. The second one won't match a code point at the end of a
1192# word, and doesn't have the run-on issue
8c32d378
KW
1193my $run_on_code_point_re =
1194 qr/ (?: 10[0-9A-F]{4} | [1-9A-F][0-9A-F]{4} | [0-9A-F]{4} ) \b/x;
1195my $code_point_re = qr/\b$run_on_code_point_re/;
99870f4d
KW
1196
1197# This matches the beginning of the line in the Unicode db files that give the
1198# defaults for code points not listed (i.e., missing) in the file. The code
1199# depends on this ending with a semi-colon, so it can assume it is a valid
1200# field when the line is split() by semi-colons
1201my $missing_defaults_prefix =
6189eadc 1202 qr/^#\s+\@missing:\s+0000\.\.$MAX_UNICODE_CODEPOINT_STRING\s*;/;
99870f4d
KW
1203
1204# Property types. Unicode has more types, but these are sufficient for our
1205# purposes.
1206my $UNKNOWN = -1; # initialized to illegal value
1207my $NON_STRING = 1; # Either binary or enum
1208my $BINARY = 2;
06f26c45
KW
1209my $FORCED_BINARY = 3; # Not a binary property, but, besides its normal
1210 # tables, additional true and false tables are
1211 # generated so that false is anything matching the
1212 # default value, and true is everything else.
1213my $ENUM = 4; # Include catalog
1214my $STRING = 5; # Anything else: string or misc
99870f4d
KW
1215
1216# Some input files have lines that give default values for code points not
1217# contained in the file. Sometimes these should be ignored.
1218my $NO_DEFAULTS = 0; # Must evaluate to false
1219my $NOT_IGNORED = 1;
1220my $IGNORED = 2;
1221
1222# Range types. Each range has a type. Most ranges are type 0, for normal,
1223# and will appear in the main body of the tables in the output files, but
1224# there are other types of ranges as well, listed below, that are specially
1225# handled. There are pseudo-types as well that will never be stored as a
1226# type, but will affect the calculation of the type.
1227
1228# 0 is for normal, non-specials
1229my $MULTI_CP = 1; # Sequence of more than code point
1230my $HANGUL_SYLLABLE = 2;
1231my $CP_IN_NAME = 3; # The NAME contains the code point appended to it.
1232my $NULL = 4; # The map is to the null string; utf8.c can't
1233 # handle these, nor is there an accepted syntax
1234 # for them in \p{} constructs
f86864ac 1235my $COMPUTE_NO_MULTI_CP = 5; # Pseudo-type; means that ranges that would
99870f4d
KW
1236 # otherwise be $MULTI_CP type are instead type 0
1237
1238# process_generic_property_file() can accept certain overrides in its input.
1239# Each of these must begin AND end with $CMD_DELIM.
1240my $CMD_DELIM = "\a";
1241my $REPLACE_CMD = 'replace'; # Override the Replace
1242my $MAP_TYPE_CMD = 'map_type'; # Override the Type
1243
1244my $NO = 0;
1245my $YES = 1;
1246
1247# Values for the Replace argument to add_range.
1248# $NO # Don't replace; add only the code points not
1249 # already present.
1250my $IF_NOT_EQUIVALENT = 1; # Replace only under certain conditions; details in
1251 # the comments at the subroutine definition.
1252my $UNCONDITIONALLY = 2; # Replace without conditions.
9470941f 1253my $MULTIPLE_BEFORE = 4; # Don't replace, but add a duplicate record if
99870f4d 1254 # already there
7f4b1e25
KW
1255my $MULTIPLE_AFTER = 5; # Don't replace, but add a duplicate record if
1256 # already there
1257my $CROAK = 6; # Die with an error if is already there
99870f4d
KW
1258
1259# Flags to give property statuses. The phrases are to remind maintainers that
1260# if the flag is changed, the indefinite article referring to it in the
1261# documentation may need to be as well.
1262my $NORMAL = "";
99870f4d
KW
1263my $DEPRECATED = 'D';
1264my $a_bold_deprecated = "a 'B<$DEPRECATED>'";
1265my $A_bold_deprecated = "A 'B<$DEPRECATED>'";
1266my $DISCOURAGED = 'X';
1267my $a_bold_discouraged = "an 'B<$DISCOURAGED>'";
1268my $A_bold_discouraged = "An 'B<$DISCOURAGED>'";
1269my $STRICTER = 'T';
1270my $a_bold_stricter = "a 'B<$STRICTER>'";
1271my $A_bold_stricter = "A 'B<$STRICTER>'";
1272my $STABILIZED = 'S';
1273my $a_bold_stabilized = "an 'B<$STABILIZED>'";
1274my $A_bold_stabilized = "An 'B<$STABILIZED>'";
1275my $OBSOLETE = 'O';
1276my $a_bold_obsolete = "an 'B<$OBSOLETE>'";
1277my $A_bold_obsolete = "An 'B<$OBSOLETE>'";
1278
1279my %status_past_participles = (
1280 $DISCOURAGED => 'discouraged',
99870f4d
KW
1281 $STABILIZED => 'stabilized',
1282 $OBSOLETE => 'obsolete',
37e2e78e 1283 $DEPRECATED => 'deprecated',
99870f4d
KW
1284);
1285
395dfc19
KW
1286# Table fates. These are somewhat ordered, so that fates < $MAP_PROXIED should be
1287# externally documented.
301ba948 1288my $ORDINARY = 0; # The normal fate.
395dfc19
KW
1289my $MAP_PROXIED = 1; # The map table for the property isn't written out,
1290 # but there is a file written that can be used to
1291 # reconstruct this table
3cdaf629 1292my $INTERNAL_ONLY = 2; # The file for this table is written out, but it is
301ba948 1293 # for Perl's internal use only
3cdaf629
KW
1294my $SUPPRESSED = 3; # The file for this table is not written out, and as a
1295 # result, we don't bother to do many computations on
1296 # it.
1297my $PLACEHOLDER = 4; # Like $SUPPRESSED, but we go through all the
1298 # computations anyway, as the values are needed for
1299 # things to work. This happens when we have Perl
1300 # extensions that depend on Unicode tables that
1301 # wouldn't normally be in a given Unicode version.
301ba948 1302
f5817e0a
KW
1303# The format of the values of the tables:
1304my $EMPTY_FORMAT = "";
99870f4d
KW
1305my $BINARY_FORMAT = 'b';
1306my $DECIMAL_FORMAT = 'd';
1307my $FLOAT_FORMAT = 'f';
1308my $INTEGER_FORMAT = 'i';
1309my $HEX_FORMAT = 'x';
1310my $RATIONAL_FORMAT = 'r';
1311my $STRING_FORMAT = 's';
d11155ec 1312my $ADJUST_FORMAT = 'a';
a14f3cb1 1313my $DECOMP_STRING_FORMAT = 'c';
c3ff2976 1314my $STRING_WHITE_SPACE_LIST = 'sw';
99870f4d
KW
1315
1316my %map_table_formats = (
1317 $BINARY_FORMAT => 'binary',
1318 $DECIMAL_FORMAT => 'single decimal digit',
1319 $FLOAT_FORMAT => 'floating point number',
1320 $INTEGER_FORMAT => 'integer',
add63c13 1321 $HEX_FORMAT => 'non-negative hex whole number; a code point',
99870f4d 1322 $RATIONAL_FORMAT => 'rational: an integer or a fraction',
1a9d544b 1323 $STRING_FORMAT => 'string',
d11155ec 1324 $ADJUST_FORMAT => 'some entries need adjustment',
92f9d56c 1325 $DECOMP_STRING_FORMAT => 'Perl\'s internal (Normalize.pm) decomposition mapping',
c3ff2976 1326 $STRING_WHITE_SPACE_LIST => 'string, but some elements are interpreted as a list; white space occurs only as list item separators'
99870f4d
KW
1327);
1328
1329# Unicode didn't put such derived files in a separate directory at first.
1330my $EXTRACTED_DIR = (-d 'extracted') ? 'extracted' : "";
1331my $EXTRACTED = ($EXTRACTED_DIR) ? "$EXTRACTED_DIR/" : "";
1332my $AUXILIARY = 'auxiliary';
1333
1334# Hashes that will eventually go into Heavy.pl for the use of utf8_heavy.pl
9e4a1e86 1335# and into UCD.pl for the use of UCD.pm
99870f4d
KW
1336my %loose_to_file_of; # loosely maps table names to their respective
1337 # files
1338my %stricter_to_file_of; # same; but for stricter mapping.
315bfd4e 1339my %loose_property_to_file_of; # Maps a loose property name to its map file
89cf10cc
KW
1340my %file_to_swash_name; # Maps the file name to its corresponding key name
1341 # in the hash %utf8::SwashInfo
99870f4d
KW
1342my %nv_floating_to_rational; # maps numeric values floating point numbers to
1343 # their rational equivalent
c12f2655
KW
1344my %loose_property_name_of; # Loosely maps (non_string) property names to
1345 # standard form
86a52d1e 1346my %string_property_loose_to_name; # Same, for string properties.
c15fda25
KW
1347my %loose_defaults; # keys are of form "prop=value", where 'prop' is
1348 # the property name in standard loose form, and
1349 # 'value' is the default value for that property,
1350 # also in standard loose form.
9e4a1e86
KW
1351my %loose_to_standard_value; # loosely maps table names to the canonical
1352 # alias for them
2df7880f
KW
1353my %ambiguous_names; # keys are alias names (in standard form) that
1354 # have more than one possible meaning.
5d1df013
KW
1355my %prop_aliases; # Keys are standard property name; values are each
1356 # one's aliases
1e863613
KW
1357my %prop_value_aliases; # Keys of top level are standard property name;
1358 # values are keys to another hash, Each one is
1359 # one of the property's values, in standard form.
1360 # The values are that prop-val's aliases.
2df7880f 1361my %ucd_pod; # Holds entries that will go into the UCD section of the pod
99870f4d 1362
d867ccfb
KW
1363# Most properties are immune to caseless matching, otherwise you would get
1364# nonsensical results, as properties are a function of a code point, not
1365# everything that is caselessly equivalent to that code point. For example,
1366# Changes_When_Case_Folded('s') should be false, whereas caselessly it would
1367# be true because 's' and 'S' are equivalent caselessly. However,
1368# traditionally, [:upper:] and [:lower:] are equivalent caselessly, so we
1369# extend that concept to those very few properties that are like this. Each
1370# such property will match the full range caselessly. They are hard-coded in
1371# the program; it's not worth trying to make it general as it's extremely
1372# unlikely that they will ever change.
1373my %caseless_equivalent_to;
1374
99870f4d
KW
1375# These constants names and values were taken from the Unicode standard,
1376# version 5.1, section 3.12. They are used in conjunction with Hangul
6e5a209b
KW
1377# syllables. The '_string' versions are so generated tables can retain the
1378# hex format, which is the more familiar value
1379my $SBase_string = "0xAC00";
1380my $SBase = CORE::hex $SBase_string;
1381my $LBase_string = "0x1100";
1382my $LBase = CORE::hex $LBase_string;
1383my $VBase_string = "0x1161";
1384my $VBase = CORE::hex $VBase_string;
1385my $TBase_string = "0x11A7";
1386my $TBase = CORE::hex $TBase_string;
99870f4d
KW
1387my $SCount = 11172;
1388my $LCount = 19;
1389my $VCount = 21;
1390my $TCount = 28;
1391my $NCount = $VCount * $TCount;
1392
1393# For Hangul syllables; These store the numbers from Jamo.txt in conjunction
1394# with the above published constants.
1395my %Jamo;
1396my %Jamo_L; # Leading consonants
1397my %Jamo_V; # Vowels
1398my %Jamo_T; # Trailing consonants
1399
bb1dd3da
KW
1400# For code points whose name contains its ordinal as a '-ABCD' suffix.
1401# The key is the base name of the code point, and the value is an
1402# array giving all the ranges that use this base name. Each range
1403# is actually a hash giving the 'low' and 'high' values of it.
1404my %names_ending_in_code_point;
1405my %loose_names_ending_in_code_point; # Same as above, but has blanks, dashes
1406 # removed from the names
1407# Inverse mapping. The list of ranges that have these kinds of
1408# names. Each element contains the low, high, and base names in an
1409# anonymous hash.
1410my @code_points_ending_in_code_point;
1411
6b5ab373
KW
1412# To hold Unicode's normalization test suite
1413my @normalization_tests;
1414
bb1dd3da
KW
1415# Boolean: does this Unicode version have the hangul syllables, and are we
1416# writing out a table for them?
1417my $has_hangul_syllables = 0;
1418
1419# Does this Unicode version have code points whose names end in their
1420# respective code points, and are we writing out a table for them? 0 for no;
1421# otherwise points to first property that a table is needed for them, so that
1422# if multiple tables are needed, we don't create duplicates
1423my $needing_code_points_ending_in_code_point = 0;
1424
37e2e78e 1425my @backslash_X_tests; # List of tests read in for testing \X
99870f4d
KW
1426my @unhandled_properties; # Will contain a list of properties found in
1427 # the input that we didn't process.
f86864ac 1428my @match_properties; # Properties that have match tables, to be
99870f4d
KW
1429 # listed in the pod
1430my @map_properties; # Properties that get map files written
1431my @named_sequences; # NamedSequences.txt contents.
1432my %potential_files; # Generated list of all .txt files in the directory
1433 # structure so we can warn if something is being
1434 # ignored.
1435my @files_actually_output; # List of files we generated.
1436my @more_Names; # Some code point names are compound; this is used
1437 # to store the extra components of them.
1438my $MIN_FRACTION_LENGTH = 3; # How many digits of a floating point number at
1439 # the minimum before we consider it equivalent to a
1440 # candidate rational
1441my $MAX_FLOATING_SLOP = 10 ** - $MIN_FRACTION_LENGTH; # And in floating terms
1442
1443# These store references to certain commonly used property objects
1444my $gc;
1445my $perl;
1446my $block;
3e20195b
KW
1447my $perl_charname;
1448my $print;
7fc6cb55 1449my $Any;
359523e2 1450my $script;
99870f4d
KW
1451
1452# Are there conflicting names because of beginning with 'In_', or 'Is_'
1453my $has_In_conflicts = 0;
1454my $has_Is_conflicts = 0;
1455
1456sub internal_file_to_platform ($) {
1457 # Convert our file paths which have '/' separators to those of the
1458 # platform.
1459
1460 my $file = shift;
1461 return undef unless defined $file;
1462
1463 return File::Spec->join(split '/', $file);
d07a55ed 1464}
5beb625e 1465
99870f4d
KW
1466sub file_exists ($) { # platform independent '-e'. This program internally
1467 # uses slash as a path separator.
1468 my $file = shift;
1469 return 0 if ! defined $file;
1470 return -e internal_file_to_platform($file);
1471}
5beb625e 1472
99870f4d 1473sub objaddr($) {
23e33b60
KW
1474 # Returns the address of the blessed input object.
1475 # It doesn't check for blessedness because that would do a string eval
1476 # every call, and the program is structured so that this is never called
1477 # for a non-blessed object.
99870f4d 1478
23e33b60 1479 no overloading; # If overloaded, numifying below won't work.
99870f4d
KW
1480
1481 # Numifying a ref gives its address.
051df77b 1482 return pack 'J', $_[0];
99870f4d
KW
1483}
1484
558712cf 1485# These are used only if $annotate is true.
c4019d52
KW
1486# The entire range of Unicode characters is examined to populate these
1487# after all the input has been processed. But most can be skipped, as they
1488# have the same descriptive phrases, such as being unassigned
1489my @viacode; # Contains the 1 million character names
1490my @printable; # boolean: And are those characters printable?
1491my @annotate_char_type; # Contains a type of those characters, specifically
1492 # for the purposes of annotation.
1493my $annotate_ranges; # A map of ranges of code points that have the same
98dc9551 1494 # name for the purposes of annotation. They map to the
c4019d52
KW
1495 # upper edge of the range, so that the end point can
1496 # be immediately found. This is used to skip ahead to
1497 # the end of a range, and avoid processing each
1498 # individual code point in it.
1499my $unassigned_sans_noncharacters; # A Range_List of the unassigned
1500 # characters, but excluding those which are
1501 # also noncharacter code points
1502
1503# The annotation types are an extension of the regular range types, though
1504# some of the latter are folded into one. Make the new types negative to
1505# avoid conflicting with the regular types
1506my $SURROGATE_TYPE = -1;
1507my $UNASSIGNED_TYPE = -2;
1508my $PRIVATE_USE_TYPE = -3;
1509my $NONCHARACTER_TYPE = -4;
1510my $CONTROL_TYPE = -5;
1511my $UNKNOWN_TYPE = -6; # Used only if there is a bug in this program
1512
1513sub populate_char_info ($) {
558712cf 1514 # Used only with the $annotate option. Populates the arrays with the
c4019d52
KW
1515 # input code point's info that are needed for outputting more detailed
1516 # comments. If calling context wants a return, it is the end point of
1517 # any contiguous range of characters that share essentially the same info
1518
1519 my $i = shift;
1520 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1521
1522 $viacode[$i] = $perl_charname->value_of($i) || "";
1523
1524 # A character is generally printable if Unicode says it is,
1525 # but below we make sure that most Unicode general category 'C' types
1526 # aren't.
1527 $printable[$i] = $print->contains($i);
1528
1529 $annotate_char_type[$i] = $perl_charname->type_of($i) || 0;
1530
1531 # Only these two regular types are treated specially for annotations
1532 # purposes
1533 $annotate_char_type[$i] = 0 if $annotate_char_type[$i] != $CP_IN_NAME
1534 && $annotate_char_type[$i] != $HANGUL_SYLLABLE;
1535
1536 # Give a generic name to all code points that don't have a real name.
1537 # We output ranges, if applicable, for these. Also calculate the end
1538 # point of the range.
1539 my $end;
1540 if (! $viacode[$i]) {
1d025d66
KW
1541 my $nonchar;
1542 if ($gc-> table('Private_use')->contains($i)) {
c4019d52
KW
1543 $viacode[$i] = 'Private Use';
1544 $annotate_char_type[$i] = $PRIVATE_USE_TYPE;
1545 $printable[$i] = 0;
1546 $end = $gc->table('Private_Use')->containing_range($i)->end;
1547 }
1d025d66
KW
1548 elsif ((defined ($nonchar =
1549 Property::property_ref('Noncharacter_Code_Point'))
1550 && $nonchar->table('Y')->contains($i)))
c4019d52
KW
1551 {
1552 $viacode[$i] = 'Noncharacter';
1553 $annotate_char_type[$i] = $NONCHARACTER_TYPE;
1554 $printable[$i] = 0;
1555 $end = property_ref('Noncharacter_Code_Point')->table('Y')->
1556 containing_range($i)->end;
1557 }
1558 elsif ($gc-> table('Control')->contains($i)) {
1559 $viacode[$i] = 'Control';
1560 $annotate_char_type[$i] = $CONTROL_TYPE;
1561 $printable[$i] = 0;
1562 $end = 0x81 if $i == 0x80; # Hard-code this one known case
1563 }
1564 elsif ($gc-> table('Unassigned')->contains($i)) {
c4019d52
KW
1565 $annotate_char_type[$i] = $UNASSIGNED_TYPE;
1566 $printable[$i] = 0;
1d025d66
KW
1567 if ($v_version lt v2.0.0) { # No blocks in earliest releases
1568 $viacode[$i] = 'Unassigned';
1569 $end = $gc-> table('Unassigned')->containing_range($i)->end;
1570 }
1571 else {
1572 $viacode[$i] = 'Unassigned, block=' . $block-> value_of($i);
c4019d52 1573
bf06c733
KW
1574 # Because we name the unassigned by the blocks they are in, it
1575 # can't go past the end of that block, and it also can't go
1576 # past the unassigned range it is in. The special table makes
1577 # sure that the non-characters, which are unassigned, are
1578 # separated out.
1579 $end = min($block->containing_range($i)->end,
1580 $unassigned_sans_noncharacters->
1581 containing_range($i)->end);
1d025d66
KW
1582 }
1583 }
1584 elsif ($v_version lt v2.0.0) { # No surrogates in earliest releases
1585 $viacode[$i] = $gc->value_of($i);
1586 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1587 $printable[$i] = 0;
1588 }
1589 elsif ($gc-> table('Surrogate')->contains($i)) {
1590 $viacode[$i] = 'Surrogate';
1591 $annotate_char_type[$i] = $SURROGATE_TYPE;
1592 $printable[$i] = 0;
1593 $end = $gc->table('Surrogate')->containing_range($i)->end;
13ca76ff
KW
1594 }
1595 else {
1596 Carp::my_carp_bug("Can't figure out how to annotate "
1597 . sprintf("U+%04X", $i)
1598 . ". Proceeding anyway.");
c4019d52
KW
1599 $viacode[$i] = 'UNKNOWN';
1600 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1601 $printable[$i] = 0;
1602 }
1603 }
1604
1605 # Here, has a name, but if it's one in which the code point number is
1606 # appended to the name, do that.
1607 elsif ($annotate_char_type[$i] == $CP_IN_NAME) {
1608 $viacode[$i] .= sprintf("-%04X", $i);
1609 $end = $perl_charname->containing_range($i)->end;
1610 }
1611
1612 # And here, has a name, but if it's a hangul syllable one, replace it with
1613 # the correct name from the Unicode algorithm
1614 elsif ($annotate_char_type[$i] == $HANGUL_SYLLABLE) {
1615 use integer;
1616 my $SIndex = $i - $SBase;
1617 my $L = $LBase + $SIndex / $NCount;
1618 my $V = $VBase + ($SIndex % $NCount) / $TCount;
1619 my $T = $TBase + $SIndex % $TCount;
1620 $viacode[$i] = "HANGUL SYLLABLE $Jamo{$L}$Jamo{$V}";
1621 $viacode[$i] .= $Jamo{$T} if $T != $TBase;
1622 $end = $perl_charname->containing_range($i)->end;
1623 }
1624
1625 return if ! defined wantarray;
1626 return $i if ! defined $end; # If not a range, return the input
1627
1628 # Save this whole range so can find the end point quickly
1629 $annotate_ranges->add_map($i, $end, $end);
1630
1631 return $end;
1632}
1633
23e33b60
KW
1634# Commented code below should work on Perl 5.8.
1635## This 'require' doesn't necessarily work in miniperl, and even if it does,
1636## the native perl version of it (which is what would operate under miniperl)
1637## is extremely slow, as it does a string eval every call.
1638#my $has_fast_scalar_util = $\18 !~ /miniperl/
1639# && defined eval "require Scalar::Util";
1640#
1641#sub objaddr($) {
1642# # Returns the address of the blessed input object. Uses the XS version if
1643# # available. It doesn't check for blessedness because that would do a
1644# # string eval every call, and the program is structured so that this is
1645# # never called for a non-blessed object.
1646#
1647# return Scalar::Util::refaddr($_[0]) if $has_fast_scalar_util;
1648#
1649# # Check at least that is a ref.
1650# my $pkg = ref($_[0]) or return undef;
1651#
1652# # Change to a fake package to defeat any overloaded stringify
1653# bless $_[0], 'main::Fake';
1654#
1655# # Numifying a ref gives its address.
051df77b 1656# my $addr = pack 'J', $_[0];
23e33b60
KW
1657#
1658# # Return to original class
1659# bless $_[0], $pkg;
1660# return $addr;
1661#}
1662
99870f4d
KW
1663sub max ($$) {
1664 my $a = shift;
1665 my $b = shift;
1666 return $a if $a >= $b;
1667 return $b;
1668}
1669
1670sub min ($$) {
1671 my $a = shift;
1672 my $b = shift;
1673 return $a if $a <= $b;
1674 return $b;
1675}
1676
1677sub clarify_number ($) {
1678 # This returns the input number with underscores inserted every 3 digits
1679 # in large (5 digits or more) numbers. Input must be entirely digits, not
1680 # checked.
1681
1682 my $number = shift;
1683 my $pos = length($number) - 3;
1684 return $number if $pos <= 1;
1685 while ($pos > 0) {
1686 substr($number, $pos, 0) = '_';
1687 $pos -= 3;
5beb625e 1688 }
99870f4d 1689 return $number;
99598c8c
JH
1690}
1691
12ac2576 1692
99870f4d 1693package Carp;
7ebf06b3 1694
99870f4d
KW
1695# These routines give a uniform treatment of messages in this program. They
1696# are placed in the Carp package to cause the stack trace to not include them,
1697# although an alternative would be to use another package and set @CARP_NOT
1698# for it.
12ac2576 1699
99870f4d 1700our $Verbose = 1 if main::DEBUG; # Useful info when debugging
12ac2576 1701
99f78760
KW
1702# This is a work-around suggested by Nicholas Clark to fix a problem with Carp
1703# and overload trying to load Scalar:Util under miniperl. See
1704# http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2009-11/msg01057.html
1705undef $overload::VERSION;
1706
99870f4d
KW
1707sub my_carp {
1708 my $message = shift || "";
1709 my $nofold = shift || 0;
7ebf06b3 1710
99870f4d
KW
1711 if ($message) {
1712 $message = main::join_lines($message);
1713 $message =~ s/^$0: *//; # Remove initial program name
1714 $message =~ s/[.;,]+$//; # Remove certain ending punctuation
1715 $message = "\n$0: $message;";
12ac2576 1716
99870f4d
KW
1717 # Fold the message with program name, semi-colon end punctuation
1718 # (which looks good with the message that carp appends to it), and a
1719 # hanging indent for continuation lines.
1720 $message = main::simple_fold($message, "", 4) unless $nofold;
1721 $message =~ s/\n$//; # Remove the trailing nl so what carp
1722 # appends is to the same line
1723 }
12ac2576 1724
99870f4d 1725 return $message if defined wantarray; # If a caller just wants the msg
12ac2576 1726
99870f4d
KW
1727 carp $message;
1728 return;
1729}
7ebf06b3 1730
99870f4d
KW
1731sub my_carp_bug {
1732 # This is called when it is clear that the problem is caused by a bug in
1733 # this program.
7ebf06b3 1734
99870f4d
KW
1735 my $message = shift;
1736 $message =~ s/^$0: *//;
1737 $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");
1738 carp $message;
1739 return;
1740}
7ebf06b3 1741
99870f4d
KW
1742sub carp_too_few_args {
1743 if (@_ != 2) {
1744 my_carp_bug("Wrong number of arguments: to 'carp_too_few_arguments'. No action taken.");
1745 return;
12ac2576 1746 }
7ebf06b3 1747
99870f4d
KW
1748 my $args_ref = shift;
1749 my $count = shift;
7ebf06b3 1750
99870f4d
KW
1751 my_carp_bug("Need at least $count arguments to "
1752 . (caller 1)[3]
1753 . ". Instead got: '"
1754 . join ', ', @$args_ref
1755 . "'. No action taken.");
1756 return;
12ac2576
JP
1757}
1758
99870f4d
KW
1759sub carp_extra_args {
1760 my $args_ref = shift;
1761 my_carp_bug("Too many arguments to 'carp_extra_args': (" . join(', ', @_) . "); Extras ignored.") if @_;
12ac2576 1762
99870f4d
KW
1763 unless (ref $args_ref) {
1764 my_carp_bug("Argument to 'carp_extra_args' ($args_ref) must be a ref. Not checking arguments.");
1765 return;
1766 }
1767 my ($package, $file, $line) = caller;
1768 my $subroutine = (caller 1)[3];
cf25bb62 1769
99870f4d
KW
1770 my $list;
1771 if (ref $args_ref eq 'HASH') {
1772 foreach my $key (keys %$args_ref) {
1773 $args_ref->{$key} = $UNDEF unless defined $args_ref->{$key};
cf25bb62 1774 }
99870f4d 1775 $list = join ', ', each %{$args_ref};
cf25bb62 1776 }
99870f4d
KW
1777 elsif (ref $args_ref eq 'ARRAY') {
1778 foreach my $arg (@$args_ref) {
1779 $arg = $UNDEF unless defined $arg;
1780 }
1781 $list = join ', ', @$args_ref;
1782 }
1783 else {
1784 my_carp_bug("Can't cope with ref "
1785 . ref($args_ref)
1786 . " . argument to 'carp_extra_args'. Not checking arguments.");
1787 return;
1788 }
1789
1790 my_carp_bug("Unrecognized parameters in options: '$list' to $subroutine. Skipped.");
1791 return;
d73e5302
JH
1792}
1793
99870f4d
KW
1794package main;
1795
1796{ # Closure
1797
1798 # This program uses the inside-out method for objects, as recommended in
1799 # "Perl Best Practices". This closure aids in generating those. There
1800 # are two routines. setup_package() is called once per package to set
1801 # things up, and then set_access() is called for each hash representing a
1802 # field in the object. These routines arrange for the object to be
1803 # properly destroyed when no longer used, and for standard accessor
1804 # functions to be generated. If you need more complex accessors, just
1805 # write your own and leave those accesses out of the call to set_access().
1806 # More details below.
1807
1808 my %constructor_fields; # fields that are to be used in constructors; see
1809 # below
1810
1811 # The values of this hash will be the package names as keys to other
1812 # hashes containing the name of each field in the package as keys, and
1813 # references to their respective hashes as values.
1814 my %package_fields;
1815
1816 sub setup_package {
1817 # Sets up the package, creating standard DESTROY and dump methods
1818 # (unless already defined). The dump method is used in debugging by
1819 # simple_dumper().
1820 # The optional parameters are:
1821 # a) a reference to a hash, that gets populated by later
1822 # set_access() calls with one of the accesses being
1823 # 'constructor'. The caller can then refer to this, but it is
1824 # not otherwise used by these two routines.
1825 # b) a reference to a callback routine to call during destruction
1826 # of the object, before any fields are actually destroyed
1827
1828 my %args = @_;
1829 my $constructor_ref = delete $args{'Constructor_Fields'};
1830 my $destroy_callback = delete $args{'Destroy_Callback'};
1831 Carp::carp_extra_args(\@_) if main::DEBUG && %args;
1832
1833 my %fields;
1834 my $package = (caller)[0];
1835
1836 $package_fields{$package} = \%fields;
1837 $constructor_fields{$package} = $constructor_ref;
1838
1839 unless ($package->can('DESTROY')) {
1840 my $destroy_name = "${package}::DESTROY";
1841 no strict "refs";
1842
1843 # Use typeglob to give the anonymous subroutine the name we want
1844 *$destroy_name = sub {
1845 my $self = shift;
ffe43484 1846 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1847
1848 $self->$destroy_callback if $destroy_callback;
1849 foreach my $field (keys %{$package_fields{$package}}) {
1850 #print STDERR __LINE__, ": Destroying ", ref $self, " ", sprintf("%04X", $addr), ": ", $field, "\n";
1851 delete $package_fields{$package}{$field}{$addr};
1852 }
1853 return;
1854 }
1855 }
1856
1857 unless ($package->can('dump')) {
1858 my $dump_name = "${package}::dump";
1859 no strict "refs";
1860 *$dump_name = sub {
1861 my $self = shift;
1862 return dump_inside_out($self, $package_fields{$package}, @_);
1863 }
1864 }
1865 return;
1866 }
1867
1868 sub set_access {
1869 # Arrange for the input field to be garbage collected when no longer
1870 # needed. Also, creates standard accessor functions for the field
1871 # based on the optional parameters-- none if none of these parameters:
1872 # 'addable' creates an 'add_NAME()' accessor function.
1873 # 'readable' or 'readable_array' creates a 'NAME()' accessor
1874 # function.
1875 # 'settable' creates a 'set_NAME()' accessor function.
1876 # 'constructor' doesn't create an accessor function, but adds the
1877 # field to the hash that was previously passed to
1878 # setup_package();
1879 # Any of the accesses can be abbreviated down, so that 'a', 'ad',
1880 # 'add' etc. all mean 'addable'.
1881 # The read accessor function will work on both array and scalar
1882 # values. If another accessor in the parameter list is 'a', the read
1883 # access assumes an array. You can also force it to be array access
1884 # by specifying 'readable_array' instead of 'readable'
1885 #
1886 # A sort-of 'protected' access can be set-up by preceding the addable,
1887 # readable or settable with some initial portion of 'protected_' (but,
1888 # the underscore is required), like 'p_a', 'pro_set', etc. The
1889 # "protection" is only by convention. All that happens is that the
1890 # accessor functions' names begin with an underscore. So instead of
1891 # calling set_foo, the call is _set_foo. (Real protection could be
c1739a4a 1892 # accomplished by having a new subroutine, end_package, called at the
99870f4d
KW
1893 # end of each package, and then storing the __LINE__ ranges and
1894 # checking them on every accessor. But that is way overkill.)
1895
1896 # We create anonymous subroutines as the accessors and then use
1897 # typeglobs to assign them to the proper package and name
1898
1899 my $name = shift; # Name of the field
1900 my $field = shift; # Reference to the inside-out hash containing the
1901 # field
1902
1903 my $package = (caller)[0];
1904
1905 if (! exists $package_fields{$package}) {
1906 croak "$0: Must call 'setup_package' before 'set_access'";
1907 }
d73e5302 1908
99870f4d
KW
1909 # Stash the field so DESTROY can get it.
1910 $package_fields{$package}{$name} = $field;
cf25bb62 1911
99870f4d
KW
1912 # Remaining arguments are the accessors. For each...
1913 foreach my $access (@_) {
1914 my $access = lc $access;
cf25bb62 1915
99870f4d 1916 my $protected = "";
cf25bb62 1917
99870f4d
KW
1918 # Match the input as far as it goes.
1919 if ($access =~ /^(p[^_]*)_/) {
1920 $protected = $1;
1921 if (substr('protected_', 0, length $protected)
1922 eq $protected)
1923 {
1924
1925 # Add 1 for the underscore not included in $protected
1926 $access = substr($access, length($protected) + 1);
1927 $protected = '_';
1928 }
1929 else {
1930 $protected = "";
1931 }
1932 }
1933
1934 if (substr('addable', 0, length $access) eq $access) {
1935 my $subname = "${package}::${protected}add_$name";
1936 no strict "refs";
1937
1938 # add_ accessor. Don't add if already there, which we
1939 # determine using 'eq' for scalars and '==' otherwise.
1940 *$subname = sub {
1941 use strict "refs";
1942 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
1943 my $self = shift;
1944 my $value = shift;
ffe43484 1945 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1946 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1947 if (ref $value) {
f998e60c 1948 return if grep { $value == $_ } @{$field->{$addr}};
99870f4d
KW
1949 }
1950 else {
f998e60c 1951 return if grep { $value eq $_ } @{$field->{$addr}};
99870f4d 1952 }
f998e60c 1953 push @{$field->{$addr}}, $value;
99870f4d
KW
1954 return;
1955 }
1956 }
1957 elsif (substr('constructor', 0, length $access) eq $access) {
1958 if ($protected) {
1959 Carp::my_carp_bug("Can't set-up 'protected' constructors")
1960 }
1961 else {
1962 $constructor_fields{$package}{$name} = $field;
1963 }
1964 }
1965 elsif (substr('readable_array', 0, length $access) eq $access) {
1966
1967 # Here has read access. If one of the other parameters for
1968 # access is array, or this one specifies array (by being more
1969 # than just 'readable_'), then create a subroutine that
1970 # assumes the data is an array. Otherwise just a scalar
1971 my $subname = "${package}::${protected}$name";
1972 if (grep { /^a/i } @_
1973 or length($access) > length('readable_'))
1974 {
1975 no strict "refs";
1976 *$subname = sub {
1977 use strict "refs";
23e33b60 1978 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
ffe43484 1979 my $addr = do { no overloading; pack 'J', $_[0]; };
99870f4d
KW
1980 if (ref $field->{$addr} ne 'ARRAY') {
1981 my $type = ref $field->{$addr};
1982 $type = 'scalar' unless $type;
1983 Carp::my_carp_bug("Trying to read $name as an array when it is a $type. Big problems.");
1984 return;
1985 }
1986 return scalar @{$field->{$addr}} unless wantarray;
1987
1988 # Make a copy; had problems with caller modifying the
1989 # original otherwise
1990 my @return = @{$field->{$addr}};
1991 return @return;
1992 }
1993 }
1994 else {
1995
1996 # Here not an array value, a simpler function.
1997 no strict "refs";
1998 *$subname = sub {
1999 use strict "refs";
23e33b60 2000 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
f998e60c 2001 no overloading;
051df77b 2002 return $field->{pack 'J', $_[0]};
99870f4d
KW
2003 }
2004 }
2005 }
2006 elsif (substr('settable', 0, length $access) eq $access) {
2007 my $subname = "${package}::${protected}set_$name";
2008 no strict "refs";
2009 *$subname = sub {
2010 use strict "refs";
23e33b60
KW
2011 if (main::DEBUG) {
2012 return Carp::carp_too_few_args(\@_, 2) if @_ < 2;
2013 Carp::carp_extra_args(\@_) if @_ > 2;
2014 }
2015 # $self is $_[0]; $value is $_[1]
f998e60c 2016 no overloading;
051df77b 2017 $field->{pack 'J', $_[0]} = $_[1];
99870f4d
KW
2018 return;
2019 }
2020 }
2021 else {
2022 Carp::my_carp_bug("Unknown accessor type $access. No accessor set.");
2023 }
cf25bb62 2024 }
99870f4d 2025 return;
cf25bb62 2026 }
99870f4d
KW
2027}
2028
2029package Input_file;
2030
2031# All input files use this object, which stores various attributes about them,
2032# and provides for convenient, uniform handling. The run method wraps the
2033# processing. It handles all the bookkeeping of opening, reading, and closing
2034# the file, returning only significant input lines.
2035#
2036# Each object gets a handler which processes the body of the file, and is
2037# called by run(). Most should use the generic, default handler, which has
2038# code scrubbed to handle things you might not expect. A handler should
2039# basically be a while(next_line()) {...} loop.
2040#
2041# You can also set up handlers to
2042# 1) call before the first line is read for pre processing
2043# 2) call to adjust each line of the input before the main handler gets them
2044# 3) call upon EOF before the main handler exits its loop
2045# 4) call at the end for post processing
2046#
2047# $_ is used to store the input line, and is to be filtered by the
2048# each_line_handler()s. So, if the format of the line is not in the desired
2049# format for the main handler, these are used to do that adjusting. They can
2050# be stacked (by enclosing them in an [ anonymous array ] in the constructor,
2051# so the $_ output of one is used as the input to the next. None of the other
2052# handlers are stackable, but could easily be changed to be so.
2053#
2054# Most of the handlers can call insert_lines() or insert_adjusted_lines()
2055# which insert the parameters as lines to be processed before the next input
2056# file line is read. This allows the EOF handler to flush buffers, for
2057# example. The difference between the two routines is that the lines inserted
2058# by insert_lines() are subjected to the each_line_handler()s. (So if you
2059# called it from such a handler, you would get infinite recursion.) Lines
2060# inserted by insert_adjusted_lines() go directly to the main handler without
2061# any adjustments. If the post-processing handler calls any of these, there
2062# will be no effect. Some error checking for these conditions could be added,
2063# but it hasn't been done.
2064#
2065# carp_bad_line() should be called to warn of bad input lines, which clears $_
2066# to prevent further processing of the line. This routine will output the
2067# message as a warning once, and then keep a count of the lines that have the
2068# same message, and output that count at the end of the file's processing.
2069# This keeps the number of messages down to a manageable amount.
2070#
2071# get_missings() should be called to retrieve any @missing input lines.
2072# Messages will be raised if this isn't done if the options aren't to ignore
2073# missings.
2074
2075sub trace { return main::trace(@_); }
2076
99870f4d
KW
2077{ # Closure
2078 # Keep track of fields that are to be put into the constructor.
2079 my %constructor_fields;
2080
2081 main::setup_package(Constructor_Fields => \%constructor_fields);
2082
2083 my %file; # Input file name, required
2084 main::set_access('file', \%file, qw{ c r });
2085
2086 my %first_released; # Unicode version file was first released in, required
2087 main::set_access('first_released', \%first_released, qw{ c r });
2088
2089 my %handler; # Subroutine to process the input file, defaults to
2090 # 'process_generic_property_file'
2091 main::set_access('handler', \%handler, qw{ c });
2092
2093 my %property;
2094 # name of property this file is for. defaults to none, meaning not
2095 # applicable, or is otherwise determinable, for example, from each line.
696609bf 2096 main::set_access('property', \%property, qw{ c r });
99870f4d
KW
2097
2098 my %optional;
2099 # If this is true, the file is optional. If not present, no warning is
2100 # output. If it is present, the string given by this parameter is
2101 # evaluated, and if false the file is not processed.
2102 main::set_access('optional', \%optional, 'c', 'r');
2103
2104 my %non_skip;
2105 # This is used for debugging, to skip processing of all but a few input
2106 # files. Add 'non_skip => 1' to the constructor for those files you want
2107 # processed when you set the $debug_skip global.
2108 main::set_access('non_skip', \%non_skip, 'c');
2109
37e2e78e 2110 my %skip;
09ca89ce
KW
2111 # This is used to skip processing of this input file semi-permanently,
2112 # when it evaluates to true. The value should be the reason the file is
2113 # being skipped. It is used for files that we aren't planning to process
2114 # anytime soon, but want to allow to be in the directory and not raise a
2115 # message that we are not handling. Mostly for test files. This is in
2116 # contrast to the non_skip element, which is supposed to be used very
2117 # temporarily for debugging. Sets 'optional' to 1. Also, files that we
2118 # pretty much will never look at can be placed in the global
1fec9f60 2119 # %ignored_files instead. Ones used here will be added to %skipped files
37e2e78e
KW
2120 main::set_access('skip', \%skip, 'c');
2121
99870f4d
KW
2122 my %each_line_handler;
2123 # list of subroutines to look at and filter each non-comment line in the
2124 # file. defaults to none. The subroutines are called in order, each is
2125 # to adjust $_ for the next one, and the final one adjusts it for
2126 # 'handler'
2127 main::set_access('each_line_handler', \%each_line_handler, 'c');
2128
2129 my %has_missings_defaults;
2130 # ? Are there lines in the file giving default values for code points
2131 # missing from it?. Defaults to NO_DEFAULTS. Otherwise NOT_IGNORED is
2132 # the norm, but IGNORED means it has such lines, but the handler doesn't
2133 # use them. Having these three states allows us to catch changes to the
2134 # UCD that this program should track
2135 main::set_access('has_missings_defaults',
2136 \%has_missings_defaults, qw{ c r });
2137
2138 my %pre_handler;
2139 # Subroutine to call before doing anything else in the file. If undef, no
2140 # such handler is called.
2141 main::set_access('pre_handler', \%pre_handler, qw{ c });
2142
2143 my %eof_handler;
2144 # Subroutine to call upon getting an EOF on the input file, but before
2145 # that is returned to the main handler. This is to allow buffers to be
2146 # flushed. The handler is expected to call insert_lines() or
2147 # insert_adjusted() with the buffered material
2148 main::set_access('eof_handler', \%eof_handler, qw{ c r });
2149
2150 my %post_handler;
2151 # Subroutine to call after all the lines of the file are read in and
2152 # processed. If undef, no such handler is called.
2153 main::set_access('post_handler', \%post_handler, qw{ c });
2154
2155 my %progress_message;
2156 # Message to print to display progress in lieu of the standard one
2157 main::set_access('progress_message', \%progress_message, qw{ c });
2158
2159 my %handle;
2160 # cache open file handle, internal. Is undef if file hasn't been
2161 # processed at all, empty if has;
2162 main::set_access('handle', \%handle);
2163
2164 my %added_lines;
2165 # cache of lines added virtually to the file, internal
2166 main::set_access('added_lines', \%added_lines);
2167
2168 my %errors;
2169 # cache of errors found, internal
2170 main::set_access('errors', \%errors);
2171
2172 my %missings;
2173 # storage of '@missing' defaults lines
2174 main::set_access('missings', \%missings);
2175
2176 sub new {
2177 my $class = shift;
2178
2179 my $self = bless \do{ my $anonymous_scalar }, $class;
ffe43484 2180 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2181
2182 # Set defaults
2183 $handler{$addr} = \&main::process_generic_property_file;
2184 $non_skip{$addr} = 0;
37e2e78e 2185 $skip{$addr} = 0;
99870f4d
KW
2186 $has_missings_defaults{$addr} = $NO_DEFAULTS;
2187 $handle{$addr} = undef;
2188 $added_lines{$addr} = [ ];
2189 $each_line_handler{$addr} = [ ];
2190 $errors{$addr} = { };
2191 $missings{$addr} = [ ];
2192
2193 # Two positional parameters.
99f78760 2194 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
99870f4d
KW
2195 $file{$addr} = main::internal_file_to_platform(shift);
2196 $first_released{$addr} = shift;
2197
2198 # The rest of the arguments are key => value pairs
2199 # %constructor_fields has been set up earlier to list all possible
2200 # ones. Either set or push, depending on how the default has been set
2201 # up just above.
2202 my %args = @_;
2203 foreach my $key (keys %args) {
2204 my $argument = $args{$key};
2205
2206 # Note that the fields are the lower case of the constructor keys
2207 my $hash = $constructor_fields{lc $key};
2208 if (! defined $hash) {
2209 Carp::my_carp_bug("Unrecognized parameters '$key => $argument' to new() for $self. Skipped");
2210 next;
2211 }
2212 if (ref $hash->{$addr} eq 'ARRAY') {
2213 if (ref $argument eq 'ARRAY') {
2214 foreach my $argument (@{$argument}) {
2215 next if ! defined $argument;
2216 push @{$hash->{$addr}}, $argument;
2217 }
2218 }
2219 else {
2220 push @{$hash->{$addr}}, $argument if defined $argument;
2221 }
2222 }
2223 else {
2224 $hash->{$addr} = $argument;
2225 }
2226 delete $args{$key};
2227 };
2228
2229 # If the file has a property for it, it means that the property is not
2230 # listed in the file's entries. So add a handler to the list of line
2231 # handlers to insert the property name into the lines, to provide a
2232 # uniform interface to the final processing subroutine.
2233 # the final code doesn't have to worry about that.
2234 if ($property{$addr}) {
2235 push @{$each_line_handler{$addr}}, \&_insert_property_into_line;
2236 }
2237
2238 if ($non_skip{$addr} && ! $debug_skip && $verbosity) {
2239 print "Warning: " . __PACKAGE__ . " constructor for $file{$addr} has useless 'non_skip' in it\n";
a3a8c5f0 2240 }
99870f4d 2241
09ca89ce
KW
2242 # If skipping, set to optional, and add to list of ignored files,
2243 # including its reason
2244 if ($skip{$addr}) {
2245 $optional{$addr} = 1;
1fec9f60 2246 $skipped_files{$file{$addr}} = $skip{$addr}
09ca89ce 2247 }
37e2e78e 2248
99870f4d 2249 return $self;
d73e5302
JH
2250 }
2251
cf25bb62 2252
99870f4d
KW
2253 use overload
2254 fallback => 0,
2255 qw("") => "_operator_stringify",
2256 "." => \&main::_operator_dot,
1285127e 2257 ".=" => \&main::_operator_dot_equal,
99870f4d 2258 ;
cf25bb62 2259
99870f4d
KW
2260 sub _operator_stringify {
2261 my $self = shift;
cf25bb62 2262
99870f4d 2263 return __PACKAGE__ . " object for " . $self->file;
d73e5302 2264 }
d73e5302 2265
99870f4d
KW
2266 # flag to make sure extracted files are processed early
2267 my $seen_non_extracted_non_age = 0;
d73e5302 2268
99870f4d
KW
2269 sub run {
2270 # Process the input object $self. This opens and closes the file and
2271 # calls all the handlers for it. Currently, this can only be called
2272 # once per file, as it destroy's the EOF handler
d73e5302 2273
99870f4d
KW
2274 my $self = shift;
2275 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
b6922eda 2276
ffe43484 2277 my $addr = do { no overloading; pack 'J', $self; };
b6922eda 2278
99870f4d 2279 my $file = $file{$addr};
d73e5302 2280
99870f4d
KW
2281 # Don't process if not expecting this file (because released later
2282 # than this Unicode version), and isn't there. This means if someone
2283 # copies it into an earlier version's directory, we will go ahead and
2284 # process it.
2285 return if $first_released{$addr} gt $v_version && ! -e $file;
2286
2287 # If in debugging mode and this file doesn't have the non-skip
2288 # flag set, and isn't one of the critical files, skip it.
2289 if ($debug_skip
2290 && $first_released{$addr} ne v0
2291 && ! $non_skip{$addr})
2292 {
2293 print "Skipping $file in debugging\n" if $verbosity;
2294 return;
2295 }
2296
2297 # File could be optional
37e2e78e 2298 if ($optional{$addr}) {
99870f4d
KW
2299 return unless -e $file;
2300 my $result = eval $optional{$addr};
2301 if (! defined $result) {
2302 Carp::my_carp_bug("Got '$@' when tried to eval $optional{$addr}. $file Skipped.");
2303 return;
2304 }
2305 if (! $result) {
2306 if ($verbosity) {
2307 print STDERR "Skipping processing input file '$file' because '$optional{$addr}' is not true\n";
2308 }
2309 return;
2310 }
2311 }
2312
2313 if (! defined $file || ! -e $file) {
2314
2315 # If the file doesn't exist, see if have internal data for it
2316 # (based on first_released being 0).
2317 if ($first_released{$addr} eq v0) {
2318 $handle{$addr} = 'pretend_is_open';
2319 }
2320 else {
2321 if (! $optional{$addr} # File could be optional
2322 && $v_version ge $first_released{$addr})
2323 {
2324 print STDERR "Skipping processing input file '$file' because not found\n" if $v_version ge $first_released{$addr};
2325 }
2326 return;
2327 }
2328 }
2329 else {
2330
37e2e78e
KW
2331 # Here, the file exists. Some platforms may change the case of
2332 # its name
99870f4d 2333 if ($seen_non_extracted_non_age) {
517956bf 2334 if ($file =~ /$EXTRACTED/i) {
1675ea0d 2335 Carp::my_carp_bug(main::join_lines(<<END
99f78760 2336$file should be processed just after the 'Prop...Alias' files, and before
99870f4d
KW
2337anything not in the $EXTRACTED_DIR directory. Proceeding, but the results may
2338have subtle problems
2339END
2340 ));
2341 }
2342 }
2343 elsif ($EXTRACTED_DIR
2344 && $first_released{$addr} ne v0
517956bf
CB
2345 && $file !~ /$EXTRACTED/i
2346 && lc($file) ne 'dage.txt')
99870f4d
KW
2347 {
2348 # We don't set this (by the 'if' above) if we have no
2349 # extracted directory, so if running on an early version,
2350 # this test won't work. Not worth worrying about.
2351 $seen_non_extracted_non_age = 1;
2352 }
2353
2354 # And mark the file as having being processed, and warn if it
2355 # isn't a file we are expecting. As we process the files,
2356 # they are deleted from the hash, so any that remain at the
2357 # end of the program are files that we didn't process.
517956bf 2358 my $fkey = File::Spec->rel2abs($file);
faf3cf6b
KW
2359 my $expecting = delete $potential_files{lc($fkey)};
2360
678f13d5
KW
2361 Carp::my_carp("Was not expecting '$file'.") if
2362 ! $expecting
99870f4d
KW
2363 && ! defined $handle{$addr};
2364
37e2e78e
KW
2365 # Having deleted from expected files, we can quit if not to do
2366 # anything. Don't print progress unless really want verbosity
2367 if ($skip{$addr}) {
2368 print "Skipping $file.\n" if $verbosity >= $VERBOSE;
2369 return;
2370 }
2371
99870f4d
KW
2372 # Open the file, converting the slashes used in this program
2373 # into the proper form for the OS
2374 my $file_handle;
2375 if (not open $file_handle, "<", $file) {
2376 Carp::my_carp("Can't open $file. Skipping: $!");
2377 return 0;
2378 }
2379 $handle{$addr} = $file_handle; # Cache the open file handle
2380 }
2381
2382 if ($verbosity >= $PROGRESS) {
2383 if ($progress_message{$addr}) {
2384 print "$progress_message{$addr}\n";
2385 }
2386 else {
2387 # If using a virtual file, say so.
2388 print "Processing ", (-e $file)
2389 ? $file
2390 : "substitute $file",
2391 "\n";
2392 }
2393 }
2394
2395
2396 # Call any special handler for before the file.
2397 &{$pre_handler{$addr}}($self) if $pre_handler{$addr};
2398
2399 # Then the main handler
2400 &{$handler{$addr}}($self);
2401
2402 # Then any special post-file handler.
2403 &{$post_handler{$addr}}($self) if $post_handler{$addr};
2404
2405 # If any errors have been accumulated, output the counts (as the first
2406 # error message in each class was output when it was encountered).
2407 if ($errors{$addr}) {
2408 my $total = 0;
2409 my $types = 0;
2410 foreach my $error (keys %{$errors{$addr}}) {
2411 $total += $errors{$addr}->{$error};
2412 delete $errors{$addr}->{$error};
2413 $types++;
2414 }
2415 if ($total > 1) {
2416 my $message
2417 = "A total of $total lines had errors in $file. ";
2418
2419 $message .= ($types == 1)
2420 ? '(Only the first one was displayed.)'
2421 : '(Only the first of each type was displayed.)';
2422 Carp::my_carp($message);
2423 }
2424 }
2425
2426 if (@{$missings{$addr}}) {
2427 Carp::my_carp_bug("Handler for $file didn't look at all the \@missing lines. Generated tables likely are wrong");
2428 }
2429
2430 # If a real file handle, close it.
2431 close $handle{$addr} or Carp::my_carp("Can't close $file: $!") if
2432 ref $handle{$addr};
2433 $handle{$addr} = ""; # Uses empty to indicate that has already seen
2434 # the file, as opposed to undef
2435 return;
2436 }
2437
2438 sub next_line {
2439 # Sets $_ to be the next logical input line, if any. Returns non-zero
2440 # if such a line exists. 'logical' means that any lines that have
2441 # been added via insert_lines() will be returned in $_ before the file
2442 # is read again.
2443
2444 my $self = shift;
2445 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2446
ffe43484 2447 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2448
2449 # Here the file is open (or if the handle is not a ref, is an open
2450 # 'virtual' file). Get the next line; any inserted lines get priority
2451 # over the file itself.
2452 my $adjusted;
2453
2454 LINE:
2455 while (1) { # Loop until find non-comment, non-empty line
2456 #local $to_trace = 1 if main::DEBUG;
2457 my $inserted_ref = shift @{$added_lines{$addr}};
2458 if (defined $inserted_ref) {
2459 ($adjusted, $_) = @{$inserted_ref};
2460 trace $adjusted, $_ if main::DEBUG && $to_trace;
2461 return 1 if $adjusted;
2462 }
2463 else {
2464 last if ! ref $handle{$addr}; # Don't read unless is real file
2465 last if ! defined ($_ = readline $handle{$addr});
2466 }
2467 chomp;
2468 trace $_ if main::DEBUG && $to_trace;
2469
2470 # See if this line is the comment line that defines what property
2471 # value that code points that are not listed in the file should
2472 # have. The format or existence of these lines is not guaranteed
2473 # by Unicode since they are comments, but the documentation says
2474 # that this was added for machine-readability, so probably won't
2475 # change. This works starting in Unicode Version 5.0. They look
2476 # like:
2477 #
2478 # @missing: 0000..10FFFF; Not_Reordered
2479 # @missing: 0000..10FFFF; Decomposition_Mapping; <code point>
2480 # @missing: 0000..10FFFF; ; NaN
2481 #
2482 # Save the line for a later get_missings() call.
2483 if (/$missing_defaults_prefix/) {
2484 if ($has_missings_defaults{$addr} == $NO_DEFAULTS) {
2485 $self->carp_bad_line("Unexpected \@missing line. Assuming no missing entries");
2486 }
2487 elsif ($has_missings_defaults{$addr} == $NOT_IGNORED) {
2488 my @defaults = split /\s* ; \s*/x, $_;
2489
2490 # The first field is the @missing, which ends in a
2491 # semi-colon, so can safely shift.
2492 shift @defaults;
2493
2494 # Some of these lines may have empty field placeholders
2495 # which get in the way. An example is:
2496 # @missing: 0000..10FFFF; ; NaN
2497 # Remove them. Process starting from the top so the
2498 # splice doesn't affect things still to be looked at.
2499 for (my $i = @defaults - 1; $i >= 0; $i--) {
2500 next if $defaults[$i] ne "";
2501 splice @defaults, $i, 1;
2502 }
2503
2504 # What's left should be just the property (maybe) and the
2505 # default. Having only one element means it doesn't have
2506 # the property.
2507 my $default;
2508 my $property;
2509 if (@defaults >= 1) {
2510 if (@defaults == 1) {
2511 $default = $defaults[0];
2512 }
2513 else {
2514 $property = $defaults[0];
2515 $default = $defaults[1];
2516 }
2517 }
2518
2519 if (@defaults < 1
2520 || @defaults > 2
2521 || ($default =~ /^</
2522 && $default !~ /^<code *point>$/i
09f8d0ac
KW
2523 && $default !~ /^<none>$/i
2524 && $default !~ /^<script>$/i))
99870f4d
KW
2525 {
2526 $self->carp_bad_line("Unrecognized \@missing line: $_. Assuming no missing entries");
2527 }
2528 else {
2529
2530 # If the property is missing from the line, it should
2531 # be the one for the whole file
2532 $property = $property{$addr} if ! defined $property;
2533
2534 # Change <none> to the null string, which is what it
2535 # really means. If the default is the code point
2536 # itself, set it to <code point>, which is what
2537 # Unicode uses (but sometimes they've forgotten the
2538 # space)
2539 if ($default =~ /^<none>$/i) {
2540 $default = "";
2541 }
2542 elsif ($default =~ /^<code *point>$/i) {
2543 $default = $CODE_POINT;
2544 }
09f8d0ac
KW
2545 elsif ($default =~ /^<script>$/i) {
2546
2547 # Special case this one. Currently is from
2548 # ScriptExtensions.txt, and means for all unlisted
2549 # code points, use their Script property values.
2550 # For the code points not listed in that file, the
2551 # default value is 'Unknown'.
2552 $default = "Unknown";
2553 }
99870f4d
KW
2554
2555 # Store them as a sub-arrays with both components.
2556 push @{$missings{$addr}}, [ $default, $property ];
2557 }
2558 }
2559
2560 # There is nothing for the caller to process on this comment
2561 # line.
2562 next;
2563 }
2564
2565 # Remove comments and trailing space, and skip this line if the
2566 # result is empty
2567 s/#.*//;
2568 s/\s+$//;
2569 next if /^$/;
2570
2571 # Call any handlers for this line, and skip further processing of
2572 # the line if the handler sets the line to null.
2573 foreach my $sub_ref (@{$each_line_handler{$addr}}) {
2574 &{$sub_ref}($self);
2575 next LINE if /^$/;
2576 }
2577
2578 # Here the line is ok. return success.
2579 return 1;
2580 } # End of looping through lines.
2581
2582 # If there is an EOF handler, call it (only once) and if it generates
2583 # more lines to process go back in the loop to handle them.
2584 if ($eof_handler{$addr}) {
2585 &{$eof_handler{$addr}}($self);
2586 $eof_handler{$addr} = ""; # Currently only get one shot at it.
2587 goto LINE if $added_lines{$addr};
2588 }
2589
2590 # Return failure -- no more lines.
2591 return 0;
2592
2593 }
2594
2595# Not currently used, not fully tested.
2596# sub peek {
2597# # Non-destructive look-ahead one non-adjusted, non-comment, non-blank
2598# # record. Not callable from an each_line_handler(), nor does it call
2599# # an each_line_handler() on the line.
2600#
2601# my $self = shift;
ffe43484 2602# my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2603#
2604# foreach my $inserted_ref (@{$added_lines{$addr}}) {
2605# my ($adjusted, $line) = @{$inserted_ref};
2606# next if $adjusted;
2607#
2608# # Remove comments and trailing space, and return a non-empty
2609# # resulting line
2610# $line =~ s/#.*//;
2611# $line =~ s/\s+$//;
2612# return $line if $line ne "";
2613# }
2614#
2615# return if ! ref $handle{$addr}; # Don't read unless is real file
2616# while (1) { # Loop until find non-comment, non-empty line
2617# local $to_trace = 1 if main::DEBUG;
2618# trace $_ if main::DEBUG && $to_trace;
2619# return if ! defined (my $line = readline $handle{$addr});
2620# chomp $line;
2621# push @{$added_lines{$addr}}, [ 0, $line ];
2622#
2623# $line =~ s/#.*//;
2624# $line =~ s/\s+$//;
2625# return $line if $line ne "";
2626# }
2627#
2628# return;
2629# }
2630
2631
2632 sub insert_lines {
2633 # Lines can be inserted so that it looks like they were in the input
2634 # file at the place it was when this routine is called. See also
2635 # insert_adjusted_lines(). Lines inserted via this routine go through
2636 # any each_line_handler()
2637
2638 my $self = shift;
2639
2640 # Each inserted line is an array, with the first element being 0 to
2641 # indicate that this line hasn't been adjusted, and needs to be
2642 # processed.
f998e60c 2643 no overloading;
051df77b 2644 push @{$added_lines{pack 'J', $self}}, map { [ 0, $_ ] } @_;
99870f4d
KW
2645 return;
2646 }
2647
2648 sub insert_adjusted_lines {
2649 # Lines can be inserted so that it looks like they were in the input
2650 # file at the place it was when this routine is called. See also
2651 # insert_lines(). Lines inserted via this routine are already fully
2652 # adjusted, ready to be processed; each_line_handler()s handlers will
2653 # not be called. This means this is not a completely general
2654 # facility, as only the last each_line_handler on the stack should
2655 # call this. It could be made more general, by passing to each of the
2656 # line_handlers their position on the stack, which they would pass on
2657 # to this routine, and that would replace the boolean first element in
2658 # the anonymous array pushed here, so that the next_line routine could
2659 # use that to call only those handlers whose index is after it on the
2660 # stack. But this is overkill for what is needed now.
2661
2662 my $self = shift;
2663 trace $_[0] if main::DEBUG && $to_trace;
2664
2665 # Each inserted line is an array, with the first element being 1 to
2666 # indicate that this line has been adjusted
f998e60c 2667 no overloading;
051df77b 2668 push @{$added_lines{pack 'J', $self}}, map { [ 1, $_ ] } @_;
99870f4d
KW
2669 return;
2670 }
2671
2672 sub get_missings {
2673 # Returns the stored up @missings lines' values, and clears the list.
2674 # The values are in an array, consisting of the default in the first
2675 # element, and the property in the 2nd. However, since these lines
2676 # can be stacked up, the return is an array of all these arrays.
2677
2678 my $self = shift;
2679 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2680
ffe43484 2681 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2682
2683 # If not accepting a list return, just return the first one.
2684 return shift @{$missings{$addr}} unless wantarray;
2685
2686 my @return = @{$missings{$addr}};
2687 undef @{$missings{$addr}};
2688 return @return;
2689 }
2690
2691 sub _insert_property_into_line {
2692 # Add a property field to $_, if this file requires it.
2693
f998e60c 2694 my $self = shift;
ffe43484 2695 my $addr = do { no overloading; pack 'J', $self; };
f998e60c 2696 my $property = $property{$addr};
99870f4d
KW
2697 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2698
2699 $_ =~ s/(;|$)/; $property$1/;
2700 return;
2701 }
2702
2703 sub carp_bad_line {
2704 # Output consistent error messages, using either a generic one, or the
2705 # one given by the optional parameter. To avoid gazillions of the
2706 # same message in case the syntax of a file is way off, this routine
2707 # only outputs the first instance of each message, incrementing a
2708 # count so the totals can be output at the end of the file.
2709
2710 my $self = shift;
2711 my $message = shift;
2712 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2713
ffe43484 2714 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2715
2716 $message = 'Unexpected line' unless $message;
2717
2718 # No trailing punctuation so as to fit with our addenda.
2719 $message =~ s/[.:;,]$//;
2720
2721 # If haven't seen this exact message before, output it now. Otherwise
2722 # increment the count of how many times it has occurred
2723 unless ($errors{$addr}->{$message}) {
2724 Carp::my_carp("$message in '$_' in "
f998e60c 2725 . $file{$addr}
99870f4d
KW
2726 . " at line $.. Skipping this line;");
2727 $errors{$addr}->{$message} = 1;
2728 }
2729 else {
2730 $errors{$addr}->{$message}++;
2731 }
2732
2733 # Clear the line to prevent any further (meaningful) processing of it.
2734 $_ = "";
2735
2736 return;
2737 }
2738} # End closure
2739
2740package Multi_Default;
2741
2742# Certain properties in early versions of Unicode had more than one possible
2743# default for code points missing from the files. In these cases, one
2744# default applies to everything left over after all the others are applied,
2745# and for each of the others, there is a description of which class of code
2746# points applies to it. This object helps implement this by storing the
2747# defaults, and for all but that final default, an eval string that generates
2748# the class that it applies to.
2749
2750
2751{ # Closure
2752
2753 main::setup_package();
2754
2755 my %class_defaults;
2756 # The defaults structure for the classes
2757 main::set_access('class_defaults', \%class_defaults);
2758
2759 my %other_default;
2760 # The default that applies to everything left over.
2761 main::set_access('other_default', \%other_default, 'r');
2762
2763
2764 sub new {
2765 # The constructor is called with default => eval pairs, terminated by
2766 # the left-over default. e.g.
2767 # Multi_Default->new(
2768 # 'T' => '$gc->table("Mn") + $gc->table("Cf") - 0x200C
2769 # - 0x200D',
2770 # 'R' => 'some other expression that evaluates to code points',
2771 # .
2772 # .
2773 # .
2774 # 'U'));
2775
2776 my $class = shift;
2777
2778 my $self = bless \do{my $anonymous_scalar}, $class;
ffe43484 2779 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2780
2781 while (@_ > 1) {
2782 my $default = shift;
2783 my $eval = shift;
2784 $class_defaults{$addr}->{$default} = $eval;
2785 }
2786
2787 $other_default{$addr} = shift;
2788
2789 return $self;
2790 }
2791
2792 sub get_next_defaults {
2793 # Iterates and returns the next class of defaults.
2794 my $self = shift;
2795 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2796
ffe43484 2797 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2798
2799 return each %{$class_defaults{$addr}};
2800 }
2801}
2802
2803package Alias;
2804
2805# An alias is one of the names that a table goes by. This class defines them
2806# including some attributes. Everything is currently setup in the
2807# constructor.
2808
2809
2810{ # Closure
2811
2812 main::setup_package();
2813
2814 my %name;
2815 main::set_access('name', \%name, 'r');
2816
2817 my %loose_match;
c12f2655 2818 # Should this name match loosely or not.
99870f4d
KW
2819 main::set_access('loose_match', \%loose_match, 'r');
2820
33e96e72
KW
2821 my %make_re_pod_entry;
2822 # Some aliases should not get their own entries in the re section of the
2823 # pod, because they are covered by a wild-card, and some we want to
2824 # discourage use of. Binary
f82fe4ba 2825 main::set_access('make_re_pod_entry', \%make_re_pod_entry, 'r', 's');
99870f4d 2826
fd1e3e84
KW
2827 my %ucd;
2828 # Is this documented to be accessible via Unicode::UCD
2829 main::set_access('ucd', \%ucd, 'r', 's');
2830
99870f4d
KW
2831 my %status;
2832 # Aliases have a status, like deprecated, or even suppressed (which means
2833 # they don't appear in documentation). Enum
2834 main::set_access('status', \%status, 'r');
2835
0eac1e20 2836 my %ok_as_filename;
99870f4d
KW
2837 # Similarly, some aliases should not be considered as usable ones for
2838 # external use, such as file names, or we don't want documentation to
2839 # recommend them. Boolean
0eac1e20 2840 main::set_access('ok_as_filename', \%ok_as_filename, 'r');
99870f4d
KW
2841
2842 sub new {
2843 my $class = shift;
2844
2845 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2846 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2847
2848 $name{$addr} = shift;
2849 $loose_match{$addr} = shift;
33e96e72 2850 $make_re_pod_entry{$addr} = shift;
0eac1e20 2851 $ok_as_filename{$addr} = shift;
99870f4d 2852 $status{$addr} = shift;
fd1e3e84 2853 $ucd{$addr} = shift;
99870f4d
KW
2854
2855 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2856
2857 # Null names are never ok externally
0eac1e20 2858 $ok_as_filename{$addr} = 0 if $name{$addr} eq "";
99870f4d
KW
2859
2860 return $self;
2861 }
2862}
2863
2864package Range;
2865
2866# A range is the basic unit for storing code points, and is described in the
2867# comments at the beginning of the program. Each range has a starting code
2868# point; an ending code point (not less than the starting one); a value
2869# that applies to every code point in between the two end-points, inclusive;
2870# and an enum type that applies to the value. The type is for the user's
2871# convenience, and has no meaning here, except that a non-zero type is
2872# considered to not obey the normal Unicode rules for having standard forms.
2873#
2874# The same structure is used for both map and match tables, even though in the
2875# latter, the value (and hence type) is irrelevant and could be used as a
2876# comment. In map tables, the value is what all the code points in the range
2877# map to. Type 0 values have the standardized version of the value stored as
2878# well, so as to not have to recalculate it a lot.
2879
2880sub trace { return main::trace(@_); }
2881
2882{ # Closure
2883
2884 main::setup_package();
2885
2886 my %start;
2887 main::set_access('start', \%start, 'r', 's');
2888
2889 my %end;
2890 main::set_access('end', \%end, 'r', 's');
2891
2892 my %value;
2893 main::set_access('value', \%value, 'r');
2894
2895 my %type;
2896 main::set_access('type', \%type, 'r');
2897
2898 my %standard_form;
2899 # The value in internal standard form. Defined only if the type is 0.
2900 main::set_access('standard_form', \%standard_form);
2901
2902 # Note that if these fields change, the dump() method should as well
2903
2904 sub new {
2905 return Carp::carp_too_few_args(\@_, 3) if main::DEBUG && @_ < 3;
2906 my $class = shift;
2907
2908 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2909 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2910
2911 $start{$addr} = shift;
2912 $end{$addr} = shift;
2913
2914 my %args = @_;
2915
2916 my $value = delete $args{'Value'}; # Can be 0
2917 $value = "" unless defined $value;
2918 $value{$addr} = $value;
2919
2920 $type{$addr} = delete $args{'Type'} || 0;
2921
2922 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2923
2924 if (! $type{$addr}) {
2925 $standard_form{$addr} = main::standardize($value);
2926 }
2927
2928 return $self;
2929 }
2930
2931 use overload
2932 fallback => 0,
2933 qw("") => "_operator_stringify",
2934 "." => \&main::_operator_dot,
1285127e 2935 ".=" => \&main::_operator_dot_equal,
99870f4d
KW
2936 ;
2937
2938 sub _operator_stringify {
2939 my $self = shift;
ffe43484 2940 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2941
2942 # Output it like '0041..0065 (value)'
2943 my $return = sprintf("%04X", $start{$addr})
2944 . '..'
2945 . sprintf("%04X", $end{$addr});
2946 my $value = $value{$addr};
2947 my $type = $type{$addr};
2948 $return .= ' (';
2949 $return .= "$value";
2950 $return .= ", Type=$type" if $type != 0;
2951 $return .= ')';
2952
2953 return $return;
2954 }
2955
2956 sub standard_form {
2957 # The standard form is the value itself if the standard form is
2958 # undefined (that is if the value is special)
2959
2960 my $self = shift;
2961 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2962
ffe43484 2963 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2964
2965 return $standard_form{$addr} if defined $standard_form{$addr};
2966 return $value{$addr};
2967 }
2968
2969 sub dump {
2970 # Human, not machine readable. For machine readable, comment out this
2971 # entire routine and let the standard one take effect.
2972 my $self = shift;
2973 my $indent = shift;
2974 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2975
ffe43484 2976 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2977
2978 my $return = $indent
2979 . sprintf("%04X", $start{$addr})
2980 . '..'
2981 . sprintf("%04X", $end{$addr})
2982 . " '$value{$addr}';";
2983 if (! defined $standard_form{$addr}) {
2984 $return .= "(type=$type{$addr})";
2985 }
2986 elsif ($standard_form{$addr} ne $value{$addr}) {
2987 $return .= "(standard '$standard_form{$addr}')";
2988 }
2989 return $return;
2990 }
2991} # End closure
2992
2993package _Range_List_Base;
2994
2995# Base class for range lists. A range list is simply an ordered list of
2996# ranges, so that the ranges with the lowest starting numbers are first in it.
2997#
2998# When a new range is added that is adjacent to an existing range that has the
2999# same value and type, it merges with it to form a larger range.
3000#
3001# Ranges generally do not overlap, except that there can be multiple entries
3002# of single code point ranges. This is because of NameAliases.txt.
3003#
3004# In this program, there is a standard value such that if two different
3005# values, have the same standard value, they are considered equivalent. This
3006# value was chosen so that it gives correct results on Unicode data
3007
3008# There are a number of methods to manipulate range lists, and some operators
3009# are overloaded to handle them.
3010
99870f4d
KW
3011sub trace { return main::trace(@_); }
3012
3013{ # Closure
3014
3015 our $addr;
3016
3017 main::setup_package();
3018
3019 my %ranges;
3020 # The list of ranges
3021 main::set_access('ranges', \%ranges, 'readable_array');
3022
3023 my %max;
3024 # The highest code point in the list. This was originally a method, but
3025 # actual measurements said it was used a lot.
3026 main::set_access('max', \%max, 'r');
3027
3028 my %each_range_iterator;
3029 # Iterator position for each_range()
3030 main::set_access('each_range_iterator', \%each_range_iterator);
3031
3032 my %owner_name_of;
3033 # Name of parent this is attached to, if any. Solely for better error
3034 # messages.
3035 main::set_access('owner_name_of', \%owner_name_of, 'p_r');
3036
3037 my %_search_ranges_cache;
3038 # A cache of the previous result from _search_ranges(), for better
3039 # performance
3040 main::set_access('_search_ranges_cache', \%_search_ranges_cache);
3041
3042 sub new {
3043 my $class = shift;
3044 my %args = @_;
3045
3046 # Optional initialization data for the range list.
3047 my $initialize = delete $args{'Initialize'};
3048
3049 my $self;
3050
3051 # Use _union() to initialize. _union() returns an object of this
3052 # class, which means that it will call this constructor recursively.
3053 # But it won't have this $initialize parameter so that it won't
3054 # infinitely loop on this.
3055 return _union($class, $initialize, %args) if defined $initialize;
3056
3057 $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 3058 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3059
3060 # Optional parent object, only for debug info.
3061 $owner_name_of{$addr} = delete $args{'Owner'};
3062 $owner_name_of{$addr} = "" if ! defined $owner_name_of{$addr};
3063
3064 # Stringify, in case it is an object.
3065 $owner_name_of{$addr} = "$owner_name_of{$addr}";
3066
3067 # This is used only for error messages, and so a colon is added
3068 $owner_name_of{$addr} .= ": " if $owner_name_of{$addr} ne "";
3069
3070 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3071
3072 # Max is initialized to a negative value that isn't adjacent to 0,
3073 # for simpler tests
3074 $max{$addr} = -2;
3075
3076 $_search_ranges_cache{$addr} = 0;
3077 $ranges{$addr} = [];
3078
3079 return $self;
3080 }
3081
3082 use overload
3083 fallback => 0,
3084 qw("") => "_operator_stringify",
3085 "." => \&main::_operator_dot,
1285127e 3086 ".=" => \&main::_operator_dot_equal,
99870f4d
KW
3087 ;
3088
3089 sub _operator_stringify {
3090 my $self = shift;
ffe43484 3091 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3092
3093 return "Range_List attached to '$owner_name_of{$addr}'"
3094 if $owner_name_of{$addr};
3095 return "anonymous Range_List " . \$self;
3096 }
3097
3098 sub _union {
3099 # Returns the union of the input code points. It can be called as
3100 # either a constructor or a method. If called as a method, the result
3101 # will be a new() instance of the calling object, containing the union
3102 # of that object with the other parameter's code points; if called as
d59563d0 3103 # a constructor, the first parameter gives the class that the new object
99870f4d
KW
3104 # should be, and the second parameter gives the code points to go into
3105 # it.
3106 # In either case, there are two parameters looked at by this routine;
3107 # any additional parameters are passed to the new() constructor.
3108 #
3109 # The code points can come in the form of some object that contains
3110 # ranges, and has a conventionally named method to access them; or
3111 # they can be an array of individual code points (as integers); or
3112 # just a single code point.
3113 #
3114 # If they are ranges, this routine doesn't make any effort to preserve
3198cc57
KW
3115 # the range values and types of one input over the other. Therefore
3116 # this base class should not allow _union to be called from other than
99870f4d
KW
3117 # initialization code, so as to prevent two tables from being added
3118 # together where the range values matter. The general form of this
3119 # routine therefore belongs in a derived class, but it was moved here
3120 # to avoid duplication of code. The failure to overload this in this
3121 # class keeps it safe.
3198cc57
KW
3122 #
3123 # It does make the effort during initialization to accept tables with
3124 # multiple values for the same code point, and to preserve the order
3125 # of these. If there is only one input range or range set, it doesn't
3126 # sort (as it should already be sorted to the desired order), and will
3127 # accept multiple values per code point. Otherwise it will merge
3128 # multiple values into a single one.
99870f4d
KW
3129
3130 my $self;
3131 my @args; # Arguments to pass to the constructor
3132
3133 my $class = shift;
3134
3135 # If a method call, will start the union with the object itself, and
3136 # the class of the new object will be the same as self.
3137 if (ref $class) {
3138 $self = $class;
3139 $class = ref $self;
3140 push @args, $self;
3141 }
3142
3143 # Add the other required parameter.
3144 push @args, shift;
3145 # Rest of parameters are passed on to the constructor
3146
3147 # Accumulate all records from both lists.
3148 my @records;
3198cc57 3149 my $input_count = 0;
99870f4d
KW
3150 for my $arg (@args) {
3151 #local $to_trace = 0 if main::DEBUG;
3152 trace "argument = $arg" if main::DEBUG && $to_trace;
3153 if (! defined $arg) {
3154 my $message = "";
3155 if (defined $self) {
f998e60c 3156 no overloading;
051df77b 3157 $message .= $owner_name_of{pack 'J', $self};
99870f4d 3158 }
ada6088e 3159 Carp::my_carp_bug($message . "Undefined argument to _union. No union done.");
99870f4d
KW
3160 return;
3161 }
3198cc57 3162
99870f4d
KW
3163 $arg = [ $arg ] if ! ref $arg;
3164 my $type = ref $arg;
3165 if ($type eq 'ARRAY') {
3166 foreach my $element (@$arg) {
3167 push @records, Range->new($element, $element);
3198cc57 3168 $input_count++;
99870f4d
KW
3169 }
3170 }
3171 elsif ($arg->isa('Range')) {
3172 push @records, $arg;
3198cc57 3173 $input_count++;
99870f4d
KW
3174 }
3175 elsif ($arg->can('ranges')) {
3176 push @records, $arg->ranges;
3198cc57 3177 $input_count++;
99870f4d
KW
3178 }
3179 else {
3180 my $message = "";
3181 if (defined $self) {
f998e60c 3182 no overloading;
051df77b 3183 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
3184 }
3185 Carp::my_carp_bug($message . "Cannot take the union of a $type. No union done.");
3186 return;
3187 }
3188 }
3189
3190 # Sort with the range containing the lowest ordinal first, but if
3191 # two ranges start at the same code point, sort with the bigger range
3192 # of the two first, because it takes fewer cycles.
3198cc57
KW
3193 if ($input_count > 1) {
3194 @records = sort { ($a->start <=> $b->start)
99870f4d
KW
3195 or
3196 # if b is shorter than a, b->end will be
3197 # less than a->end, and we want to select
3198 # a, so want to return -1
3199 ($b->end <=> $a->end)
3200 } @records;
3198cc57 3201 }
99870f4d
KW
3202
3203 my $new = $class->new(@_);
3204
3205 # Fold in records so long as they add new information.
3206 for my $set (@records) {
3207 my $start = $set->start;
3208 my $end = $set->end;
d59563d0 3209 my $value = $set->value;
3198cc57 3210 my $type = $set->type;
99870f4d 3211 if ($start > $new->max) {
3198cc57 3212 $new->_add_delete('+', $start, $end, $value, Type => $type);
99870f4d
KW
3213 }
3214 elsif ($end > $new->max) {
3198cc57
KW
3215 $new->_add_delete('+', $new->max +1, $end, $value,
3216 Type => $type);
3217 }
3218 elsif ($input_count == 1) {
3219 # Here, overlaps existing range, but is from a single input,
3220 # so preserve the multiple values from that input.
3221 $new->_add_delete('+', $start, $end, $value, Type => $type,
3222 Replace => $MULTIPLE_AFTER);
99870f4d
KW
3223 }
3224 }
3225
3226 return $new;
3227 }
3228
3229 sub range_count { # Return the number of ranges in the range list
3230 my $self = shift;
3231 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3232
f998e60c 3233 no overloading;
051df77b 3234 return scalar @{$ranges{pack 'J', $self}};
99870f4d
KW
3235 }
3236
3237 sub min {
3238 # Returns the minimum code point currently in the range list, or if
3239 # the range list is empty, 2 beyond the max possible. This is a
3240 # method because used so rarely, that not worth saving between calls,
3241 # and having to worry about changing it as ranges are added and
3242 # deleted.
3243
3244 my $self = shift;
3245 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3246
ffe43484 3247 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3248
3249 # If the range list is empty, return a large value that isn't adjacent
3250 # to any that could be in the range list, for simpler tests
6189eadc 3251 return $MAX_UNICODE_CODEPOINT + 2 unless scalar @{$ranges{$addr}};
99870f4d
KW
3252 return $ranges{$addr}->[0]->start;
3253 }
3254
3255 sub contains {
3256 # Boolean: Is argument in the range list? If so returns $i such that:
3257 # range[$i]->end < $codepoint <= range[$i+1]->end
3258 # which is one beyond what you want; this is so that the 0th range
3259 # doesn't return false
3260 my $self = shift;
3261 my $codepoint = shift;
3262 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3263
99870f4d
KW
3264 my $i = $self->_search_ranges($codepoint);
3265 return 0 unless defined $i;
3266
3267 # The search returns $i, such that
3268 # range[$i-1]->end < $codepoint <= range[$i]->end
3269 # So is in the table if and only iff it is at least the start position
3270 # of range $i.
f998e60c 3271 no overloading;
051df77b 3272 return 0 if $ranges{pack 'J', $self}->[$i]->start > $codepoint;
99870f4d
KW
3273 return $i + 1;
3274 }
3275
2f7a8815
KW
3276 sub containing_range {
3277 # Returns the range object that contains the code point, undef if none
3278
3279 my $self = shift;
3280 my $codepoint = shift;
3281 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3282
3283 my $i = $self->contains($codepoint);
3284 return unless $i;
3285
3286 # contains() returns 1 beyond where we should look
3287 no overloading;
3288 return $ranges{pack 'J', $self}->[$i-1];
3289 }
3290
99870f4d
KW
3291 sub value_of {
3292 # Returns the value associated with the code point, undef if none
3293
3294 my $self = shift;
3295 my $codepoint = shift;
3296 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3297
d69c231b
KW
3298 my $range = $self->containing_range($codepoint);
3299 return unless defined $range;
99870f4d 3300
d69c231b 3301 return $range->value;
99870f4d
KW
3302 }
3303
0a9dbafc
KW
3304 sub type_of {
3305 # Returns the type of the range containing the code point, undef if
3306 # the code point is not in the table
3307
3308 my $self = shift;
3309 my $codepoint = shift;
3310 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3311
3312 my $range = $self->containing_range($codepoint);
3313 return unless defined $range;
3314
3315 return $range->type;
3316 }
3317
99870f4d
KW
3318 sub _search_ranges {
3319 # Find the range in the list which contains a code point, or where it
3320 # should go if were to add it. That is, it returns $i, such that:
3321 # range[$i-1]->end < $codepoint <= range[$i]->end
3322 # Returns undef if no such $i is possible (e.g. at end of table), or
3323 # if there is an error.
3324
3325 my $self = shift;
3326 my $code_point = shift;
3327 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3328
ffe43484 3329 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3330
3331 return if $code_point > $max{$addr};
3332 my $r = $ranges{$addr}; # The current list of ranges
3333 my $range_list_size = scalar @$r;
3334 my $i;
3335
3336 use integer; # want integer division
3337
3338 # Use the cached result as the starting guess for this one, because,
3339 # an experiment on 5.1 showed that 90% of the time the cache was the
3340 # same as the result on the next call (and 7% it was one less).
3341 $i = $_search_ranges_cache{$addr};
3342 $i = 0 if $i >= $range_list_size; # Reset if no longer valid (prob.
3343 # from an intervening deletion
3344 #local $to_trace = 1 if main::DEBUG;
3345 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);
3346 return $i if $code_point <= $r->[$i]->end
3347 && ($i == 0 || $r->[$i-1]->end < $code_point);
3348
3349 # Here the cache doesn't yield the correct $i. Try adding 1.
3350 if ($i < $range_list_size - 1
3351 && $r->[$i]->end < $code_point &&
3352 $code_point <= $r->[$i+1]->end)
3353 {
3354 $i++;
3355 trace "next \$i is correct: $i" if main::DEBUG && $to_trace;
3356 $_search_ranges_cache{$addr} = $i;
3357 return $i;
3358 }
3359
3360 # Here, adding 1 also didn't work. We do a binary search to
3361 # find the correct position, starting with current $i
3362 my $lower = 0;
3363 my $upper = $range_list_size - 1;
3364 while (1) {
3365 trace "top of loop i=$i:", sprintf("%04X", $r->[$lower]->start), "[$lower] .. ", sprintf("%04X", $r->[$i]->start), "[$i] .. ", sprintf("%04X", $r->[$upper]->start), "[$upper]" if main::DEBUG && $to_trace;
3366
3367 if ($code_point <= $r->[$i]->end) {
3368
3369 # Here we have met the upper constraint. We can quit if we
3370 # also meet the lower one.
3371 last if $i == 0 || $r->[$i-1]->end < $code_point;
3372
3373 $upper = $i; # Still too high.
3374
3375 }
3376 else {
3377
3378 # Here, $r[$i]->end < $code_point, so look higher up.
3379 $lower = $i;
3380 }
3381
3382 # Split search domain in half to try again.
3383 my $temp = ($upper + $lower) / 2;
3384
3385 # No point in continuing unless $i changes for next time
3386 # in the loop.
3387 if ($temp == $i) {
3388
3389 # We can't reach the highest element because of the averaging.
3390 # So if one below the upper edge, force it there and try one
3391 # more time.
3392 if ($i == $range_list_size - 2) {
3393
3394 trace "Forcing to upper edge" if main::DEBUG && $to_trace;
3395 $i = $range_list_size - 1;
3396
3397 # Change $lower as well so if fails next time through,
3398 # taking the average will yield the same $i, and we will
3399 # quit with the error message just below.
3400 $lower = $i;
3401 next;
3402 }
3403 Carp::my_carp_bug("$owner_name_of{$addr}Can't find where the range ought to go. No action taken.");
3404 return;
3405 }
3406 $i = $temp;
3407 } # End of while loop
3408
3409 if (main::DEBUG && $to_trace) {
3410 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i;
3411 trace "i= [ $i ]", $r->[$i];
3412 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < $range_list_size - 1;
3413 }
3414
3415 # Here we have found the offset. Cache it as a starting point for the
3416 # next call.
3417 $_search_ranges_cache{$addr} = $i;
3418 return $i;
3419 }
3420
3421 sub _add_delete {
3422 # Add, replace or delete ranges to or from a list. The $type
3423 # parameter gives which:
3424 # '+' => insert or replace a range, returning a list of any changed
3425 # ranges.
3426 # '-' => delete a range, returning a list of any deleted ranges.
3427 #
3428 # The next three parameters give respectively the start, end, and
3429 # value associated with the range. 'value' should be null unless the
3430 # operation is '+';
3431 #
3432 # The range list is kept sorted so that the range with the lowest
3433 # starting position is first in the list, and generally, adjacent
c1739a4a 3434 # ranges with the same values are merged into a single larger one (see
99870f4d
KW
3435 # exceptions below).
3436 #
c1739a4a 3437 # There are more parameters; all are key => value pairs:
99870f4d
KW
3438 # Type gives the type of the value. It is only valid for '+'.
3439 # All ranges have types; if this parameter is omitted, 0 is
3440 # assumed. Ranges with type 0 are assumed to obey the
3441 # Unicode rules for casing, etc; ranges with other types are
3442 # not. Otherwise, the type is arbitrary, for the caller's
3443 # convenience, and looked at only by this routine to keep
3444 # adjacent ranges of different types from being merged into
3445 # a single larger range, and when Replace =>
3446 # $IF_NOT_EQUIVALENT is specified (see just below).
3447 # Replace determines what to do if the range list already contains
3448 # ranges which coincide with all or portions of the input
3449 # range. It is only valid for '+':
3450 # => $NO means that the new value is not to replace
3451 # any existing ones, but any empty gaps of the
3452 # range list coinciding with the input range
3453 # will be filled in with the new value.
3454 # => $UNCONDITIONALLY means to replace the existing values with
3455 # this one unconditionally. However, if the
3456 # new and old values are identical, the
3457 # replacement is skipped to save cycles
3458 # => $IF_NOT_EQUIVALENT means to replace the existing values
d59563d0 3459 # (the default) with this one if they are not equivalent.
99870f4d 3460 # Ranges are equivalent if their types are the
c1739a4a 3461 # same, and they are the same string; or if
99870f4d
KW
3462 # both are type 0 ranges, if their Unicode
3463 # standard forms are identical. In this last
3464 # case, the routine chooses the more "modern"
3465 # one to use. This is because some of the
3466 # older files are formatted with values that
3467 # are, for example, ALL CAPs, whereas the
3468 # derived files have a more modern style,
3469 # which looks better. By looking for this
3470 # style when the pre-existing and replacement
3471 # standard forms are the same, we can move to
3472 # the modern style
9470941f 3473 # => $MULTIPLE_BEFORE means that if this range duplicates an
99870f4d
KW
3474 # existing one, but has a different value,
3475 # don't replace the existing one, but insert
3476 # this, one so that the same range can occur
53d84487
KW
3477 # multiple times. They are stored LIFO, so
3478 # that the final one inserted is the first one
3479 # returned in an ordered search of the table.
6901521e
KW
3480 # If this is an exact duplicate, including the
3481 # value, the original will be moved to be
3482 # first, before any other duplicate ranges
3483 # with different values.
7f4b1e25
KW
3484 # => $MULTIPLE_AFTER is like $MULTIPLE_BEFORE, but is stored
3485 # FIFO, so that this one is inserted after all
6901521e
KW
3486 # others that currently exist. If this is an
3487 # exact duplicate, including value, of an
3488 # existing range, this one is discarded
3489 # (leaving the existing one in its original,
3490 # higher priority position
99870f4d
KW
3491 # => anything else is the same as => $IF_NOT_EQUIVALENT
3492 #
c1739a4a
KW
3493 # "same value" means identical for non-type-0 ranges, and it means
3494 # having the same standard forms for type-0 ranges.
99870f4d
KW
3495
3496 return Carp::carp_too_few_args(\@_, 5) if main::DEBUG && @_ < 5;
3497
3498 my $self = shift;
3499 my $operation = shift; # '+' for add/replace; '-' for delete;
3500 my $start = shift;
3501 my $end = shift;
3502 my $value = shift;
3503
3504 my %args = @_;
3505
3506 $value = "" if not defined $value; # warning: $value can be "0"
3507
3508 my $replace = delete $args{'Replace'};
3509 $replace = $IF_NOT_EQUIVALENT unless defined $replace;
3510
3511 my $type = delete $args{'Type'};
3512 $type = 0 unless defined $type;
3513
3514 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3515
ffe43484 3516 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3517
3518 if ($operation ne '+' && $operation ne '-') {
3519 Carp::my_carp_bug("$owner_name_of{$addr}First parameter to _add_delete must be '+' or '-'. No action taken.");
3520 return;
3521 }
3522 unless (defined $start && defined $end) {
3523 Carp::my_carp_bug("$owner_name_of{$addr}Undefined start and/or end to _add_delete. No action taken.");
3524 return;
3525 }
3526 unless ($end >= $start) {
3527 Carp::my_carp_bug("$owner_name_of{$addr}End of range (" . sprintf("%04X", $end) . ") must not be before start (" . sprintf("%04X", $start) . "). No action taken.");
3528 return;
3529 }
556ca434
KW
3530 if ($end > $MAX_UNICODE_CODEPOINT && $operation eq '+') {
3531 Carp::my_carp("$owner_name_of{$addr}Warning: Range '" . sprintf("%04X..%04X", $start, $end) . ") is above the Unicode maximum of " . sprintf("%04X", $MAX_UNICODE_CODEPOINT) . ". Adding it anyway");
3532 }
99870f4d
KW
3533 #local $to_trace = 1 if main::DEBUG;
3534
3535 if ($operation eq '-') {
3536 if ($replace != $IF_NOT_EQUIVALENT) {
3537 Carp::my_carp_bug("$owner_name_of{$addr}Replace => \$IF_NOT_EQUIVALENT is required when deleting a range from a range list. Assuming Replace => \$IF_NOT_EQUIVALENT.");
3538 $replace = $IF_NOT_EQUIVALENT;
3539 }
3540 if ($type) {
3541 Carp::my_carp_bug("$owner_name_of{$addr}Type => 0 is required when deleting a range from a range list. Assuming Type => 0.");
3542 $type = 0;
3543 }
3544 if ($value ne "") {
3545 Carp::my_carp_bug("$owner_name_of{$addr}Value => \"\" is required when deleting a range from a range list. Assuming Value => \"\".");
3546 $value = "";
3547 }
3548 }
3549
3550 my $r = $ranges{$addr}; # The current list of ranges
3551 my $range_list_size = scalar @$r; # And its size
3552 my $max = $max{$addr}; # The current high code point in
3553 # the list of ranges
3554
3555 # Do a special case requiring fewer machine cycles when the new range
3556 # starts after the current highest point. The Unicode input data is
3557 # structured so this is common.
3558 if ($start > $max) {
3559
52d4d76a 3560 trace "$owner_name_of{$addr} $operation", sprintf("%04X..%04X (%s) type=%d; prev max=%04X", $start, $end, $value, $type, $max) if main::DEBUG && $to_trace;
99870f4d
KW
3561 return if $operation eq '-'; # Deleting a non-existing range is a
3562 # no-op
3563
3564 # If the new range doesn't logically extend the current final one
3565 # in the range list, create a new range at the end of the range
3566 # list. (max cleverly is initialized to a negative number not
3567 # adjacent to 0 if the range list is empty, so even adding a range
3568 # to an empty range list starting at 0 will have this 'if'
3569 # succeed.)
3570 if ($start > $max + 1 # non-adjacent means can't extend.
3571 || @{$r}[-1]->value ne $value # values differ, can't extend.
3572 || @{$r}[-1]->type != $type # types differ, can't extend.
3573 ) {
3574 push @$r, Range->new($start, $end,
3575 Value => $value,
3576 Type => $type);
3577 }
3578 else {
3579
3580 # Here, the new range starts just after the current highest in
3581 # the range list, and they have the same type and value.
3582 # Extend the current range to incorporate the new one.
3583 @{$r}[-1]->set_end($end);
3584 }
3585
3586 # This becomes the new maximum.
3587 $max{$addr} = $end;
3588
3589 return;
3590 }
3591 #local $to_trace = 0 if main::DEBUG;
3592
3593 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) replace=$replace" if main::DEBUG && $to_trace;
3594
3595 # Here, the input range isn't after the whole rest of the range list.
3596 # Most likely 'splice' will be needed. The rest of the routine finds
3597 # the needed splice parameters, and if necessary, does the splice.
3598 # First, find the offset parameter needed by the splice function for
3599 # the input range. Note that the input range may span multiple
3600 # existing ones, but we'll worry about that later. For now, just find
3601 # the beginning. If the input range is to be inserted starting in a
3602 # position not currently in the range list, it must (obviously) come
3603 # just after the range below it, and just before the range above it.
3604 # Slightly less obviously, it will occupy the position currently
3605 # occupied by the range that is to come after it. More formally, we
3606 # are looking for the position, $i, in the array of ranges, such that:
3607 #
3608 # r[$i-1]->start <= r[$i-1]->end < $start < r[$i]->start <= r[$i]->end
3609 #
3610 # (The ordered relationships within existing ranges are also shown in
3611 # the equation above). However, if the start of the input range is
3612 # within an existing range, the splice offset should point to that
3613 # existing range's position in the list; that is $i satisfies a
3614 # somewhat different equation, namely:
3615 #
3616 #r[$i-1]->start <= r[$i-1]->end < r[$i]->start <= $start <= r[$i]->end
3617 #
3618 # More briefly, $start can come before or after r[$i]->start, and at
3619 # this point, we don't know which it will be. However, these
3620 # two equations share these constraints:
3621 #
3622 # r[$i-1]->end < $start <= r[$i]->end
3623 #
3624 # And that is good enough to find $i.
3625
3626 my $i = $self->_search_ranges($start);
3627 if (! defined $i) {
3628 Carp::my_carp_bug("Searching $self for range beginning with $start unexpectedly returned undefined. Operation '$operation' not performed");
3629 return;
3630 }
3631
3632 # The search function returns $i such that:
3633 #
3634 # r[$i-1]->end < $start <= r[$i]->end
3635 #
3636 # That means that $i points to the first range in the range list
3637 # that could possibly be affected by this operation. We still don't
3638 # know if the start of the input range is within r[$i], or if it
3639 # points to empty space between r[$i-1] and r[$i].
3640 trace "[$i] is the beginning splice point. Existing range there is ", $r->[$i] if main::DEBUG && $to_trace;
3641
3642 # Special case the insertion of data that is not to replace any
3643 # existing data.
3644 if ($replace == $NO) { # If $NO, has to be operation '+'
3645 #local $to_trace = 1 if main::DEBUG;
3646 trace "Doesn't replace" if main::DEBUG && $to_trace;
3647
3648 # Here, the new range is to take effect only on those code points
3649 # that aren't already in an existing range. This can be done by
3650 # looking through the existing range list and finding the gaps in
3651 # the ranges that this new range affects, and then calling this
3652 # function recursively on each of those gaps, leaving untouched
3653 # anything already in the list. Gather up a list of the changed
3654 # gaps first so that changes to the internal state as new ranges
3655 # are added won't be a problem.
3656 my @gap_list;
3657
3658 # First, if the starting point of the input range is outside an
3659 # existing one, there is a gap from there to the beginning of the
3660 # existing range -- add a span to fill the part that this new
3661 # range occupies
3662 if ($start < $r->[$i]->start) {
3663 push @gap_list, Range->new($start,
3664 main::min($end,
3665 $r->[$i]->start - 1),
3666 Type => $type);
3667 trace "gap before $r->[$i] [$i], will add", $gap_list[-1] if main::DEBUG && $to_trace;
3668 }
3669
3670 # Then look through the range list for other gaps until we reach
3671 # the highest range affected by the input one.
3672 my $j;
3673 for ($j = $i+1; $j < $range_list_size; $j++) {
3674 trace "j=[$j]", $r->[$j] if main::DEBUG && $to_trace;
3675 last if $end < $r->[$j]->start;
3676
3677 # If there is a gap between when this range starts and the
3678 # previous one ends, add a span to fill it. Note that just
3679 # because there are two ranges doesn't mean there is a
3680 # non-zero gap between them. It could be that they have
3681 # different values or types
3682 if ($r->[$j-1]->end + 1 != $r->[$j]->start) {
3683 push @gap_list,
3684 Range->new($r->[$j-1]->end + 1,
3685 $r->[$j]->start - 1,
3686 Type => $type);
3687 trace "gap between $r->[$j-1] and $r->[$j] [$j], will add: $gap_list[-1]" if main::DEBUG && $to_trace;
3688 }
3689 }
3690
3691 # Here, we have either found an existing range in the range list,
3692 # beyond the area affected by the input one, or we fell off the
3693 # end of the loop because the input range affects the whole rest
3694 # of the range list. In either case, $j is 1 higher than the
3695 # highest affected range. If $j == $i, it means that there are no
3696 # affected ranges, that the entire insertion is in the gap between
3697 # r[$i-1], and r[$i], which we already have taken care of before
3698 # the loop.
3699 # On the other hand, if there are affected ranges, it might be
3700 # that there is a gap that needs filling after the final such
3701 # range to the end of the input range
3702 if ($r->[$j-1]->end < $end) {
3703 push @gap_list, Range->new(main::max($start,
3704 $r->[$j-1]->end + 1),
3705 $end,
3706 Type => $type);
3707 trace "gap after $r->[$j-1], will add $gap_list[-1]" if main::DEBUG && $to_trace;
3708 }
3709
3710 # Call recursively to fill in all the gaps.
3711 foreach my $gap (@gap_list) {
3712 $self->_add_delete($operation,
3713 $gap->start,
3714 $gap->end,
3715 $value,
3716 Type => $type);
3717 }
3718
3719 return;
3720 }
3721
53d84487
KW
3722 # Here, we have taken care of the case where $replace is $NO.
3723 # Remember that here, r[$i-1]->end < $start <= r[$i]->end
3724 # If inserting a multiple record, this is where it goes, before the
7f4b1e25
KW
3725 # first (if any) existing one if inserting LIFO. (If this is to go
3726 # afterwards, FIFO, we below move the pointer to there.) These imply
3727 # an insertion, and no change to any existing ranges. Note that $i
3728 # can be -1 if this new range doesn't actually duplicate any existing,
3729 # and comes at the beginning of the list.
3730 if ($replace == $MULTIPLE_BEFORE || $replace == $MULTIPLE_AFTER) {
53d84487
KW
3731
3732 if ($start != $end) {
3733 Carp::my_carp_bug("$owner_name_of{$addr}Can't cope with adding a multiple record when the range ($start..$end) contains more than one code point. No action taken.");
3734 return;
3735 }
3736
19155fcc 3737 # If the new code point is within a current range ...
53d84487 3738 if ($end >= $r->[$i]->start) {
19155fcc
KW
3739
3740 # Don't add an exact duplicate, as it isn't really a multiple
1f6798c4
KW
3741 my $existing_value = $r->[$i]->value;
3742 my $existing_type = $r->[$i]->type;
3743 return if $value eq $existing_value && $type eq $existing_type;
3744
3745 # If the multiple value is part of an existing range, we want
3746 # to split up that range, so that only the single code point
3747 # is affected. To do this, we first call ourselves
3748 # recursively to delete that code point from the table, having
3749 # preserved its current data above. Then we call ourselves
3750 # recursively again to add the new multiple, which we know by
3751 # the test just above is different than the current code
3752 # point's value, so it will become a range containing a single
3753 # code point: just itself. Finally, we add back in the
3754 # pre-existing code point, which will again be a single code
3755 # point range. Because 'i' likely will have changed as a
3756 # result of these operations, we can't just continue on, but
7f4b1e25
KW
3757 # do this operation recursively as well. If we are inserting
3758 # LIFO, the pre-existing code point needs to go after the new
3759 # one, so use MULTIPLE_AFTER; and vice versa.
53d84487 3760 if ($r->[$i]->start != $r->[$i]->end) {
1f6798c4
KW
3761 $self->_add_delete('-', $start, $end, "");
3762 $self->_add_delete('+', $start, $end, $value, Type => $type);
7f4b1e25
KW
3763 return $self->_add_delete('+',
3764 $start, $end,
3765 $existing_value,
3766 Type => $existing_type,
3767 Replace => ($replace == $MULTIPLE_BEFORE)
3768 ? $MULTIPLE_AFTER
3769 : $MULTIPLE_BEFORE);
3770 }
3771 }
3772
3773 # If to place this new record after, move to beyond all existing
1722e378 3774 # ones; but don't add this one if identical to any of them, as it
6901521e
KW
3775 # isn't really a multiple. This leaves the original order, so
3776 # that the current request is ignored. The reasoning is that the
3777 # previous request that wanted this record to have high priority
3778 # should have precedence.
7f4b1e25
KW
3779 if ($replace == $MULTIPLE_AFTER) {
3780 while ($i < @$r && $r->[$i]->start == $start) {
1722e378
KW
3781 return if $value eq $r->[$i]->value
3782 && $type eq $r->[$i]->type;
7f4b1e25 3783 $i++;
53d84487 3784 }
53d84487 3785 }
6901521e
KW
3786 else {
3787 # If instead we are to place this new record before any
3788 # existing ones, remove any identical ones that come after it.
3789 # This change