This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
mktables: Move table creation code
[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;
1e958ea9
KW
817push @tables_that_may_be_empty, 'Canonical_Combining_Class=CCC133'
818 if $v_version ge v6.2.0;
99870f4d
KW
819
820# The lists below are hashes, so the key is the item in the list, and the
821# value is the reason why it is in the list. This makes generation of
822# documentation easier.
823
824my %why_suppressed; # No file generated for these.
825
826# Files aren't generated for empty extraneous properties. This is arguable.
827# Extraneous properties generally come about because a property is no longer
828# used in a newer version of Unicode. If we generated a file without code
829# points, programs that used to work on that property will still execute
830# without errors. It just won't ever match (or will always match, with \P{}).
831# This means that the logic is now likely wrong. I (khw) think its better to
832# find this out by getting an error message. Just move them to the table
833# above to change this behavior
834my %why_suppress_if_empty_warn_if_not = (
835
836 # It is the only property that has ever officially been removed from the
837 # Standard. The database never contained any code points for it.
838 'Special_Case_Condition' => 'Obsolete',
839
840 # Apparently never official, but there were code points in some versions of
841 # old-style PropList.txt
842 'Non_Break' => 'Obsolete',
843);
844
845# These would normally go in the warn table just above, but they were changed
846# a long time before this program was written, so warnings about them are
847# moot.
848if ($v_version gt v3.2.0) {
849 push @tables_that_may_be_empty,
850 'Canonical_Combining_Class=Attached_Below_Left'
851}
852
5f7264c7 853# These are listed in the Property aliases file in 6.0, but Unihan is ignored
99870f4d
KW
854# unless explicitly added.
855if ($v_version ge v5.2.0) {
856 my $unihan = 'Unihan; remove from list if using Unihan';
ea25a9b2 857 foreach my $table (qw (
99870f4d
KW
858 kAccountingNumeric
859 kOtherNumeric
860 kPrimaryNumeric
861 kCompatibilityVariant
862 kIICore
863 kIRG_GSource
864 kIRG_HSource
865 kIRG_JSource
866 kIRG_KPSource
867 kIRG_MSource
868 kIRG_KSource
869 kIRG_TSource
870 kIRG_USource
871 kIRG_VSource
872 kRSUnicode
ea25a9b2 873 ))
99870f4d
KW
874 {
875 $why_suppress_if_empty_warn_if_not{$table} = $unihan;
876 }
ca12659b
NC
877}
878
272501f6
KW
879# Enum values for to_output_map() method in the Map_Table package.
880my $EXTERNAL_MAP = 1;
881my $INTERNAL_MAP = 2;
ce712c88 882my $OUTPUT_ADJUSTED = 3;
272501f6 883
fcf1973c
KW
884# To override computed values for writing the map tables for these properties.
885# The default for enum map tables is to write them out, so that the Unicode
886# .txt files can be removed, but all the data to compute any property value
887# for any code point is available in a more compact form.
888my %global_to_output_map = (
889 # Needed by UCD.pm, but don't want to publicize that it exists, so won't
c12f2655
KW
890 # get stuck supporting it if things change. Since it is a STRING
891 # property, it normally would be listed in the pod, but INTERNAL_MAP
892 # suppresses that.
fcf1973c
KW
893 Unicode_1_Name => $INTERNAL_MAP,
894
895 Present_In => 0, # Suppress, as easily computed from Age
fcf1973c 896 Block => 0, # Suppress, as Blocks.txt is retained.
53d34b6c
KW
897
898 # Suppress, as mapping can be found instead from the
899 # Perl_Decomposition_Mapping file
900 Decomposition_Type => 0,
fcf1973c
KW
901);
902
99870f4d 903# Properties that this program ignores.
230e0c16
KW
904my @unimplemented_properties;
905
906# With this release, it is automatically handled if the Unihan db is
907# downloaded
908push @unimplemented_properties, 'Unicode_Radical_Stroke' if $v_version le v5.2.0;
d73e5302 909
99870f4d
KW
910# There are several types of obsolete properties defined by Unicode. These
911# must be hand-edited for every new Unicode release.
912my %why_deprecated; # Generates a deprecated warning message if used.
913my %why_stabilized; # Documentation only
914my %why_obsolete; # Documentation only
915
916{ # Closure
917 my $simple = 'Perl uses the more complete version of this property';
918 my $unihan = 'Unihan properties are by default not enabled in the Perl core. Instead use CPAN: Unicode::Unihan';
919
920 my $other_properties = 'other properties';
921 my $contributory = "Used by Unicode internally for generating $other_properties and not intended to be used stand-alone";
5d294d41 922 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
923
924 %why_deprecated = (
5f7264c7 925 'Grapheme_Link' => 'Deprecated by Unicode: Duplicates ccc=vr (Canonical_Combining_Class=Virama)',
99870f4d
KW
926 'Jamo_Short_Name' => $contributory,
927 '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',
928 'Other_Alphabetic' => $contributory,
929 'Other_Default_Ignorable_Code_Point' => $contributory,
930 'Other_Grapheme_Extend' => $contributory,
931 'Other_ID_Continue' => $contributory,
932 'Other_ID_Start' => $contributory,
933 'Other_Lowercase' => $contributory,
934 'Other_Math' => $contributory,
935 'Other_Uppercase' => $contributory,
e22aaf5c
KW
936 'Expands_On_NFC' => $why_no_expand,
937 'Expands_On_NFD' => $why_no_expand,
938 'Expands_On_NFKC' => $why_no_expand,
939 'Expands_On_NFKD' => $why_no_expand,
99870f4d
KW
940 );
941
942 %why_suppressed = (
5f7264c7 943 # There is a lib/unicore/Decomposition.pl (used by Normalize.pm) which
99870f4d
KW
944 # contains the same information, but without the algorithmically
945 # determinable Hangul syllables'. This file is not published, so it's
946 # existence is not noted in the comment.
e0b29447 947 'Decomposition_Mapping' => 'Accessible via Unicode::Normalize or Unicode::UCD::prop_invmap()',
99870f4d 948
3111abc0
KW
949 'Indic_Matra_Category' => "Provisional",
950 'Indic_Syllabic_Category' => "Provisional",
951
5f8d1a89
KW
952 # Don't suppress ISO_Comment, as otherwise special handling is needed
953 # to differentiate between it and gc=c, which can be written as 'isc',
954 # which is the same characters as ISO_Comment's short name.
99870f4d 955
fbb93542 956 'Name' => "Accessible via \\N{...} or 'use charnames;' or Unicode::UCD::prop_invmap()",
e0b29447
KW
957
958 'Simple_Case_Folding' => "$simple. Can access this through Unicode::UCD::casefold or Unicode::UCD::prop_invmap()",
959 'Simple_Lowercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
960 'Simple_Titlecase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
961 'Simple_Uppercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo or Unicode::UCD::prop_invmap()",
99870f4d 962
5f7264c7 963 FC_NFKC_Closure => 'Supplanted in usage by NFKC_Casefold; otherwise not useful',
99870f4d
KW
964 );
965
1704a0ea
KW
966 foreach my $property (
967
968 # The following are suppressed because they were made contributory
969 # or deprecated by Unicode before Perl ever thought about
970 # supporting them.
971 'Jamo_Short_Name',
972 'Grapheme_Link',
973 'Expands_On_NFC',
974 'Expands_On_NFD',
975 'Expands_On_NFKC',
976 'Expands_On_NFKD',
977
978 # The following are suppressed because they have been marked
979 # as deprecated for a sufficient amount of time
980 'Other_Alphabetic',
981 'Other_Default_Ignorable_Code_Point',
982 'Other_Grapheme_Extend',
983 'Other_ID_Continue',
984 'Other_ID_Start',
985 'Other_Lowercase',
986 'Other_Math',
987 'Other_Uppercase',
e22aaf5c 988 ) {
99870f4d
KW
989 $why_suppressed{$property} = $why_deprecated{$property};
990 }
cf25bb62 991
99870f4d
KW
992 # Customize the message for all the 'Other_' properties
993 foreach my $property (keys %why_deprecated) {
994 next if (my $main_property = $property) !~ s/^Other_//;
995 $why_deprecated{$property} =~ s/$other_properties/the $main_property property (which should be used instead)/;
996 }
997}
998
e9c4b4f8
KW
999if ($write_Unicode_deprecated_tables) {
1000 foreach my $property (keys %why_suppressed) {
1001 delete $why_suppressed{$property} if $property =~
1002 / ^ Other | Grapheme /x;
1003 }
1004}
1005
99870f4d
KW
1006if ($v_version ge 4.0.0) {
1007 $why_stabilized{'Hyphen'} = 'Use the Line_Break property instead; see www.unicode.org/reports/tr14';
5f7264c7
KW
1008 if ($v_version ge 6.0.0) {
1009 $why_deprecated{'Hyphen'} = 'Supplanted by Line_Break property values; see www.unicode.org/reports/tr14';
1010 }
99870f4d 1011}
5f7264c7 1012if ($v_version ge 5.2.0 && $v_version lt 6.0.0) {
99870f4d 1013 $why_obsolete{'ISO_Comment'} = 'Code points for it have been removed';
5f7264c7 1014 if ($v_version ge 6.0.0) {
63f74647 1015 $why_deprecated{'ISO_Comment'} = 'No longer needed for Unicode\'s internal chart generation; otherwise not useful, and code points for it have been removed';
5f7264c7 1016 }
99870f4d
KW
1017}
1018
1019# Probably obsolete forever
1020if ($v_version ge v4.1.0) {
82aed44a
KW
1021 $why_suppressed{'Script=Katakana_Or_Hiragana'} = 'Obsolete. All code points previously matched by this have been moved to "Script=Common".';
1022}
1023if ($v_version ge v6.0.0) {
caa75395 1024 $why_suppressed{'Script=Katakana_Or_Hiragana'} .= ' Consider instead using "Script_Extensions=Katakana" or "Script_Extensions=Hiragana" (or both)';
2b352efd 1025 $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
1026}
1027
1028# This program can create files for enumerated-like properties, such as
1029# 'Numeric_Type'. This file would be the same format as for a string
1030# property, with a mapping from code point to its value, so you could look up,
1031# for example, the script a code point is in. But no one so far wants this
1032# mapping, or they have found another way to get it since this is a new
1033# feature. So no file is generated except if it is in this list.
1034my @output_mapped_properties = split "\n", <<END;
1035END
1036
c12f2655
KW
1037# If you are using the Unihan database in a Unicode version before 5.2, you
1038# need to add the properties that you want to extract from it to this table.
1039# For your convenience, the properties in the 6.0 PropertyAliases.txt file are
1040# listed, commented out
99870f4d
KW
1041my @cjk_properties = split "\n", <<'END';
1042#cjkAccountingNumeric; kAccountingNumeric
1043#cjkOtherNumeric; kOtherNumeric
1044#cjkPrimaryNumeric; kPrimaryNumeric
1045#cjkCompatibilityVariant; kCompatibilityVariant
1046#cjkIICore ; kIICore
1047#cjkIRG_GSource; kIRG_GSource
1048#cjkIRG_HSource; kIRG_HSource
1049#cjkIRG_JSource; kIRG_JSource
1050#cjkIRG_KPSource; kIRG_KPSource
1051#cjkIRG_KSource; kIRG_KSource
1052#cjkIRG_TSource; kIRG_TSource
1053#cjkIRG_USource; kIRG_USource
1054#cjkIRG_VSource; kIRG_VSource
1055#cjkRSUnicode; kRSUnicode ; Unicode_Radical_Stroke; URS
1056END
1057
1058# Similarly for the property values. For your convenience, the lines in the
5f7264c7 1059# 6.0 PropertyAliases.txt file are listed. Just remove the first BUT NOT both
c12f2655 1060# '#' marks (for Unicode versions before 5.2)
99870f4d
KW
1061my @cjk_property_values = split "\n", <<'END';
1062## @missing: 0000..10FFFF; cjkAccountingNumeric; NaN
1063## @missing: 0000..10FFFF; cjkCompatibilityVariant; <code point>
1064## @missing: 0000..10FFFF; cjkIICore; <none>
1065## @missing: 0000..10FFFF; cjkIRG_GSource; <none>
1066## @missing: 0000..10FFFF; cjkIRG_HSource; <none>
1067## @missing: 0000..10FFFF; cjkIRG_JSource; <none>
1068## @missing: 0000..10FFFF; cjkIRG_KPSource; <none>
1069## @missing: 0000..10FFFF; cjkIRG_KSource; <none>
1070## @missing: 0000..10FFFF; cjkIRG_TSource; <none>
1071## @missing: 0000..10FFFF; cjkIRG_USource; <none>
1072## @missing: 0000..10FFFF; cjkIRG_VSource; <none>
1073## @missing: 0000..10FFFF; cjkOtherNumeric; NaN
1074## @missing: 0000..10FFFF; cjkPrimaryNumeric; NaN
1075## @missing: 0000..10FFFF; cjkRSUnicode; <none>
1076END
1077
1078# The input files don't list every code point. Those not listed are to be
1079# defaulted to some value. Below are hard-coded what those values are for
1080# non-binary properties as of 5.1. Starting in 5.0, there are
caa75395 1081# machine-parsable comment lines in the files that give the defaults; so this
99870f4d
KW
1082# list shouldn't have to be extended. The claim is that all missing entries
1083# for binary properties will default to 'N'. Unicode tried to change that in
1084# 5.2, but the beta period produced enough protest that they backed off.
1085#
1086# The defaults for the fields that appear in UnicodeData.txt in this hash must
1087# be in the form that it expects. The others may be synonyms.
1088my $CODE_POINT = '<code point>';
1089my %default_mapping = (
1090 Age => "Unassigned",
1091 # Bidi_Class => Complicated; set in code
1092 Bidi_Mirroring_Glyph => "",
1093 Block => 'No_Block',
1094 Canonical_Combining_Class => 0,
1095 Case_Folding => $CODE_POINT,
1096 Decomposition_Mapping => $CODE_POINT,
1097 Decomposition_Type => 'None',
1098 East_Asian_Width => "Neutral",
1099 FC_NFKC_Closure => $CODE_POINT,
1100 General_Category => 'Cn',
1101 Grapheme_Cluster_Break => 'Other',
1102 Hangul_Syllable_Type => 'NA',
1103 ISO_Comment => "",
1104 Jamo_Short_Name => "",
1105 Joining_Group => "No_Joining_Group",
1106 # Joining_Type => Complicated; set in code
1107 kIICore => 'N', # Is converted to binary
1108 #Line_Break => Complicated; set in code
1109 Lowercase_Mapping => $CODE_POINT,
1110 Name => "",
1111 Name_Alias => "",
1112 NFC_QC => 'Yes',
1113 NFD_QC => 'Yes',
1114 NFKC_QC => 'Yes',
1115 NFKD_QC => 'Yes',
1116 Numeric_Type => 'None',
1117 Numeric_Value => 'NaN',
1118 Script => ($v_version le 4.1.0) ? 'Common' : 'Unknown',
1119 Sentence_Break => 'Other',
1120 Simple_Case_Folding => $CODE_POINT,
1121 Simple_Lowercase_Mapping => $CODE_POINT,
1122 Simple_Titlecase_Mapping => $CODE_POINT,
1123 Simple_Uppercase_Mapping => $CODE_POINT,
1124 Titlecase_Mapping => $CODE_POINT,
1125 Unicode_1_Name => "",
1126 Unicode_Radical_Stroke => "",
1127 Uppercase_Mapping => $CODE_POINT,
1128 Word_Break => 'Other',
1129);
1130
232ed87f
KW
1131# Below are files that Unicode furnishes, but this program ignores, and why.
1132# NormalizationCorrections.txt requires some more explanation. It documents
1133# the cumulative fixes to erroneous normalizations in earlier Unicode
1134# versions. Its main purpose is so that someone running on an earlier version
1135# can use this file to override what got published in that earlier release.
1136# It would be easy for mktables to read and handle this file. But all the
1137# corrections in it should already be in the other files for the release it
1138# is. To get it to actually mean something useful, someone would have to be
1139# using an earlier Unicode release, and copy it to the files for that release
1140# and recomplile. So far there has been no demand to do that, so this hasn't
1141# been implemented.
99870f4d 1142my %ignored_files = (
73ba1144
KW
1143 'CJKRadicals.txt' => 'Maps the kRSUnicode property values to corresponding code points',
1144 'Index.txt' => 'Alphabetical index of Unicode characters',
1145 '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',
1146 'NamesList.txt' => 'Annotated list of characters',
1147 'NormalizationCorrections.txt' => 'Documentation of corrections already incorporated into the Unicode data base',
1148 'Props.txt' => 'Only in very early releases; is a subset of F<PropList.txt> (which is used instead)',
1149 'ReadMe.txt' => 'Documentation',
1150 '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>',
1151 'EmojiSources.txt' => 'Maps certain Unicode code points to their legacy Japanese cell-phone values',
8ee2793f
KW
1152 'USourceData.txt' => 'Documentation of status and cross reference of proposals for encoding by Unicode of Unihan characters',
1153 'USourceData.pdf' => 'Documentation of status and cross reference of proposals for encoding by Unicode of Unihan characters',
73ba1144
KW
1154 'auxiliary/WordBreakTest.html' => 'Documentation of validation tests',
1155 'auxiliary/SentenceBreakTest.html' => 'Documentation of validation tests',
1156 'auxiliary/GraphemeBreakTest.html' => 'Documentation of validation tests',
1157 'auxiliary/LineBreakTest.html' => 'Documentation of validation tests',
99870f4d
KW
1158);
1159
1fec9f60
KW
1160my %skipped_files; # List of files that we skip
1161
678f13d5 1162### End of externally interesting definitions, except for @input_file_objects
99870f4d
KW
1163
1164my $HEADER=<<"EOF";
1165# !!!!!!! DO NOT EDIT THIS FILE !!!!!!!
3df51b85
KW
1166# This file is machine-generated by $0 from the Unicode
1167# database, Version $string_version. Any changes made here will be lost!
cf25bb62
JH
1168EOF
1169
126c3d4e 1170my $INTERNAL_ONLY_HEADER = <<"EOF";
99870f4d
KW
1171
1172# !!!!!!! INTERNAL PERL USE ONLY !!!!!!!
fac53429
KW
1173# This file is for internal use by core Perl only. The format and even the
1174# name or existence of this file are subject to change without notice. Don't
1175# use it directly.
99870f4d
KW
1176EOF
1177
1178my $DEVELOPMENT_ONLY=<<"EOF";
1179# !!!!!!! DEVELOPMENT USE ONLY !!!!!!!
1180# This file contains information artificially constrained to code points
1181# present in Unicode release $string_compare_versions.
1182# IT CANNOT BE RELIED ON. It is for use during development only and should
23e33b60 1183# not be used for production.
b6922eda
KW
1184
1185EOF
1186
6189eadc
KW
1187my $MAX_UNICODE_CODEPOINT_STRING = "10FFFF";
1188my $MAX_UNICODE_CODEPOINT = hex $MAX_UNICODE_CODEPOINT_STRING;
1189my $MAX_UNICODE_CODEPOINTS = $MAX_UNICODE_CODEPOINT + 1;
99870f4d
KW
1190
1191# Matches legal code point. 4-6 hex numbers, If there are 6, the first
1192# two must be 10; if there are 5, the first must not be a 0. Written this way
92199589
KW
1193# to decrease backtracking. The first regex allows the code point to be at
1194# the end of a word, but to work properly, the word shouldn't end with a valid
1195# hex character. The second one won't match a code point at the end of a
1196# word, and doesn't have the run-on issue
8c32d378
KW
1197my $run_on_code_point_re =
1198 qr/ (?: 10[0-9A-F]{4} | [1-9A-F][0-9A-F]{4} | [0-9A-F]{4} ) \b/x;
1199my $code_point_re = qr/\b$run_on_code_point_re/;
99870f4d
KW
1200
1201# This matches the beginning of the line in the Unicode db files that give the
1202# defaults for code points not listed (i.e., missing) in the file. The code
1203# depends on this ending with a semi-colon, so it can assume it is a valid
1204# field when the line is split() by semi-colons
1205my $missing_defaults_prefix =
6189eadc 1206 qr/^#\s+\@missing:\s+0000\.\.$MAX_UNICODE_CODEPOINT_STRING\s*;/;
99870f4d
KW
1207
1208# Property types. Unicode has more types, but these are sufficient for our
1209# purposes.
1210my $UNKNOWN = -1; # initialized to illegal value
1211my $NON_STRING = 1; # Either binary or enum
1212my $BINARY = 2;
06f26c45
KW
1213my $FORCED_BINARY = 3; # Not a binary property, but, besides its normal
1214 # tables, additional true and false tables are
1215 # generated so that false is anything matching the
1216 # default value, and true is everything else.
1217my $ENUM = 4; # Include catalog
1218my $STRING = 5; # Anything else: string or misc
99870f4d
KW
1219
1220# Some input files have lines that give default values for code points not
1221# contained in the file. Sometimes these should be ignored.
1222my $NO_DEFAULTS = 0; # Must evaluate to false
1223my $NOT_IGNORED = 1;
1224my $IGNORED = 2;
1225
1226# Range types. Each range has a type. Most ranges are type 0, for normal,
1227# and will appear in the main body of the tables in the output files, but
1228# there are other types of ranges as well, listed below, that are specially
1229# handled. There are pseudo-types as well that will never be stored as a
1230# type, but will affect the calculation of the type.
1231
1232# 0 is for normal, non-specials
1233my $MULTI_CP = 1; # Sequence of more than code point
1234my $HANGUL_SYLLABLE = 2;
1235my $CP_IN_NAME = 3; # The NAME contains the code point appended to it.
1236my $NULL = 4; # The map is to the null string; utf8.c can't
1237 # handle these, nor is there an accepted syntax
1238 # for them in \p{} constructs
f86864ac 1239my $COMPUTE_NO_MULTI_CP = 5; # Pseudo-type; means that ranges that would
99870f4d
KW
1240 # otherwise be $MULTI_CP type are instead type 0
1241
1242# process_generic_property_file() can accept certain overrides in its input.
1243# Each of these must begin AND end with $CMD_DELIM.
1244my $CMD_DELIM = "\a";
1245my $REPLACE_CMD = 'replace'; # Override the Replace
1246my $MAP_TYPE_CMD = 'map_type'; # Override the Type
1247
1248my $NO = 0;
1249my $YES = 1;
1250
1251# Values for the Replace argument to add_range.
1252# $NO # Don't replace; add only the code points not
1253 # already present.
1254my $IF_NOT_EQUIVALENT = 1; # Replace only under certain conditions; details in
1255 # the comments at the subroutine definition.
1256my $UNCONDITIONALLY = 2; # Replace without conditions.
9470941f 1257my $MULTIPLE_BEFORE = 4; # Don't replace, but add a duplicate record if
99870f4d 1258 # already there
7f4b1e25
KW
1259my $MULTIPLE_AFTER = 5; # Don't replace, but add a duplicate record if
1260 # already there
1261my $CROAK = 6; # Die with an error if is already there
99870f4d
KW
1262
1263# Flags to give property statuses. The phrases are to remind maintainers that
1264# if the flag is changed, the indefinite article referring to it in the
1265# documentation may need to be as well.
1266my $NORMAL = "";
99870f4d
KW
1267my $DEPRECATED = 'D';
1268my $a_bold_deprecated = "a 'B<$DEPRECATED>'";
1269my $A_bold_deprecated = "A 'B<$DEPRECATED>'";
1270my $DISCOURAGED = 'X';
1271my $a_bold_discouraged = "an 'B<$DISCOURAGED>'";
1272my $A_bold_discouraged = "An 'B<$DISCOURAGED>'";
1273my $STRICTER = 'T';
1274my $a_bold_stricter = "a 'B<$STRICTER>'";
1275my $A_bold_stricter = "A 'B<$STRICTER>'";
1276my $STABILIZED = 'S';
1277my $a_bold_stabilized = "an 'B<$STABILIZED>'";
1278my $A_bold_stabilized = "An 'B<$STABILIZED>'";
1279my $OBSOLETE = 'O';
1280my $a_bold_obsolete = "an 'B<$OBSOLETE>'";
1281my $A_bold_obsolete = "An 'B<$OBSOLETE>'";
1282
1283my %status_past_participles = (
1284 $DISCOURAGED => 'discouraged',
99870f4d
KW
1285 $STABILIZED => 'stabilized',
1286 $OBSOLETE => 'obsolete',
37e2e78e 1287 $DEPRECATED => 'deprecated',
99870f4d
KW
1288);
1289
395dfc19
KW
1290# Table fates. These are somewhat ordered, so that fates < $MAP_PROXIED should be
1291# externally documented.
301ba948 1292my $ORDINARY = 0; # The normal fate.
395dfc19
KW
1293my $MAP_PROXIED = 1; # The map table for the property isn't written out,
1294 # but there is a file written that can be used to
1295 # reconstruct this table
3cdaf629 1296my $INTERNAL_ONLY = 2; # The file for this table is written out, but it is
301ba948 1297 # for Perl's internal use only
3cdaf629
KW
1298my $SUPPRESSED = 3; # The file for this table is not written out, and as a
1299 # result, we don't bother to do many computations on
1300 # it.
1301my $PLACEHOLDER = 4; # Like $SUPPRESSED, but we go through all the
1302 # computations anyway, as the values are needed for
1303 # things to work. This happens when we have Perl
1304 # extensions that depend on Unicode tables that
1305 # wouldn't normally be in a given Unicode version.
301ba948 1306
f5817e0a
KW
1307# The format of the values of the tables:
1308my $EMPTY_FORMAT = "";
99870f4d
KW
1309my $BINARY_FORMAT = 'b';
1310my $DECIMAL_FORMAT = 'd';
1311my $FLOAT_FORMAT = 'f';
1312my $INTEGER_FORMAT = 'i';
1313my $HEX_FORMAT = 'x';
1314my $RATIONAL_FORMAT = 'r';
1315my $STRING_FORMAT = 's';
d11155ec 1316my $ADJUST_FORMAT = 'a';
a14f3cb1 1317my $DECOMP_STRING_FORMAT = 'c';
c3ff2976 1318my $STRING_WHITE_SPACE_LIST = 'sw';
99870f4d
KW
1319
1320my %map_table_formats = (
1321 $BINARY_FORMAT => 'binary',
1322 $DECIMAL_FORMAT => 'single decimal digit',
1323 $FLOAT_FORMAT => 'floating point number',
1324 $INTEGER_FORMAT => 'integer',
add63c13 1325 $HEX_FORMAT => 'non-negative hex whole number; a code point',
99870f4d 1326 $RATIONAL_FORMAT => 'rational: an integer or a fraction',
1a9d544b 1327 $STRING_FORMAT => 'string',
d11155ec 1328 $ADJUST_FORMAT => 'some entries need adjustment',
92f9d56c 1329 $DECOMP_STRING_FORMAT => 'Perl\'s internal (Normalize.pm) decomposition mapping',
c3ff2976 1330 $STRING_WHITE_SPACE_LIST => 'string, but some elements are interpreted as a list; white space occurs only as list item separators'
99870f4d
KW
1331);
1332
1333# Unicode didn't put such derived files in a separate directory at first.
1334my $EXTRACTED_DIR = (-d 'extracted') ? 'extracted' : "";
1335my $EXTRACTED = ($EXTRACTED_DIR) ? "$EXTRACTED_DIR/" : "";
1336my $AUXILIARY = 'auxiliary';
1337
1338# Hashes that will eventually go into Heavy.pl for the use of utf8_heavy.pl
9e4a1e86 1339# and into UCD.pl for the use of UCD.pm
99870f4d
KW
1340my %loose_to_file_of; # loosely maps table names to their respective
1341 # files
1342my %stricter_to_file_of; # same; but for stricter mapping.
315bfd4e 1343my %loose_property_to_file_of; # Maps a loose property name to its map file
89cf10cc
KW
1344my %file_to_swash_name; # Maps the file name to its corresponding key name
1345 # in the hash %utf8::SwashInfo
99870f4d
KW
1346my %nv_floating_to_rational; # maps numeric values floating point numbers to
1347 # their rational equivalent
c12f2655
KW
1348my %loose_property_name_of; # Loosely maps (non_string) property names to
1349 # standard form
86a52d1e 1350my %string_property_loose_to_name; # Same, for string properties.
c15fda25
KW
1351my %loose_defaults; # keys are of form "prop=value", where 'prop' is
1352 # the property name in standard loose form, and
1353 # 'value' is the default value for that property,
1354 # also in standard loose form.
9e4a1e86
KW
1355my %loose_to_standard_value; # loosely maps table names to the canonical
1356 # alias for them
2df7880f
KW
1357my %ambiguous_names; # keys are alias names (in standard form) that
1358 # have more than one possible meaning.
5d1df013
KW
1359my %prop_aliases; # Keys are standard property name; values are each
1360 # one's aliases
1e863613
KW
1361my %prop_value_aliases; # Keys of top level are standard property name;
1362 # values are keys to another hash, Each one is
1363 # one of the property's values, in standard form.
1364 # The values are that prop-val's aliases.
2df7880f 1365my %ucd_pod; # Holds entries that will go into the UCD section of the pod
99870f4d 1366
d867ccfb
KW
1367# Most properties are immune to caseless matching, otherwise you would get
1368# nonsensical results, as properties are a function of a code point, not
1369# everything that is caselessly equivalent to that code point. For example,
1370# Changes_When_Case_Folded('s') should be false, whereas caselessly it would
1371# be true because 's' and 'S' are equivalent caselessly. However,
1372# traditionally, [:upper:] and [:lower:] are equivalent caselessly, so we
1373# extend that concept to those very few properties that are like this. Each
1374# such property will match the full range caselessly. They are hard-coded in
1375# the program; it's not worth trying to make it general as it's extremely
1376# unlikely that they will ever change.
1377my %caseless_equivalent_to;
1378
99870f4d
KW
1379# These constants names and values were taken from the Unicode standard,
1380# version 5.1, section 3.12. They are used in conjunction with Hangul
6e5a209b
KW
1381# syllables. The '_string' versions are so generated tables can retain the
1382# hex format, which is the more familiar value
1383my $SBase_string = "0xAC00";
1384my $SBase = CORE::hex $SBase_string;
1385my $LBase_string = "0x1100";
1386my $LBase = CORE::hex $LBase_string;
1387my $VBase_string = "0x1161";
1388my $VBase = CORE::hex $VBase_string;
1389my $TBase_string = "0x11A7";
1390my $TBase = CORE::hex $TBase_string;
99870f4d
KW
1391my $SCount = 11172;
1392my $LCount = 19;
1393my $VCount = 21;
1394my $TCount = 28;
1395my $NCount = $VCount * $TCount;
1396
1397# For Hangul syllables; These store the numbers from Jamo.txt in conjunction
1398# with the above published constants.
1399my %Jamo;
1400my %Jamo_L; # Leading consonants
1401my %Jamo_V; # Vowels
1402my %Jamo_T; # Trailing consonants
1403
bb1dd3da
KW
1404# For code points whose name contains its ordinal as a '-ABCD' suffix.
1405# The key is the base name of the code point, and the value is an
1406# array giving all the ranges that use this base name. Each range
1407# is actually a hash giving the 'low' and 'high' values of it.
1408my %names_ending_in_code_point;
1409my %loose_names_ending_in_code_point; # Same as above, but has blanks, dashes
1410 # removed from the names
1411# Inverse mapping. The list of ranges that have these kinds of
1412# names. Each element contains the low, high, and base names in an
1413# anonymous hash.
1414my @code_points_ending_in_code_point;
1415
6b5ab373
KW
1416# To hold Unicode's normalization test suite
1417my @normalization_tests;
1418
bb1dd3da
KW
1419# Boolean: does this Unicode version have the hangul syllables, and are we
1420# writing out a table for them?
1421my $has_hangul_syllables = 0;
1422
1423# Does this Unicode version have code points whose names end in their
1424# respective code points, and are we writing out a table for them? 0 for no;
1425# otherwise points to first property that a table is needed for them, so that
1426# if multiple tables are needed, we don't create duplicates
1427my $needing_code_points_ending_in_code_point = 0;
1428
37e2e78e 1429my @backslash_X_tests; # List of tests read in for testing \X
99870f4d
KW
1430my @unhandled_properties; # Will contain a list of properties found in
1431 # the input that we didn't process.
f86864ac 1432my @match_properties; # Properties that have match tables, to be
99870f4d
KW
1433 # listed in the pod
1434my @map_properties; # Properties that get map files written
1435my @named_sequences; # NamedSequences.txt contents.
1436my %potential_files; # Generated list of all .txt files in the directory
1437 # structure so we can warn if something is being
1438 # ignored.
1439my @files_actually_output; # List of files we generated.
1440my @more_Names; # Some code point names are compound; this is used
1441 # to store the extra components of them.
1442my $MIN_FRACTION_LENGTH = 3; # How many digits of a floating point number at
1443 # the minimum before we consider it equivalent to a
1444 # candidate rational
1445my $MAX_FLOATING_SLOP = 10 ** - $MIN_FRACTION_LENGTH; # And in floating terms
1446
1447# These store references to certain commonly used property objects
1448my $gc;
1449my $perl;
1450my $block;
3e20195b
KW
1451my $perl_charname;
1452my $print;
7fc6cb55 1453my $Any;
359523e2 1454my $script;
99870f4d
KW
1455
1456# Are there conflicting names because of beginning with 'In_', or 'Is_'
1457my $has_In_conflicts = 0;
1458my $has_Is_conflicts = 0;
1459
1460sub internal_file_to_platform ($) {
1461 # Convert our file paths which have '/' separators to those of the
1462 # platform.
1463
1464 my $file = shift;
1465 return undef unless defined $file;
1466
1467 return File::Spec->join(split '/', $file);
d07a55ed 1468}
5beb625e 1469
99870f4d
KW
1470sub file_exists ($) { # platform independent '-e'. This program internally
1471 # uses slash as a path separator.
1472 my $file = shift;
1473 return 0 if ! defined $file;
1474 return -e internal_file_to_platform($file);
1475}
5beb625e 1476
99870f4d 1477sub objaddr($) {
23e33b60
KW
1478 # Returns the address of the blessed input object.
1479 # It doesn't check for blessedness because that would do a string eval
1480 # every call, and the program is structured so that this is never called
1481 # for a non-blessed object.
99870f4d 1482
23e33b60 1483 no overloading; # If overloaded, numifying below won't work.
99870f4d
KW
1484
1485 # Numifying a ref gives its address.
051df77b 1486 return pack 'J', $_[0];
99870f4d
KW
1487}
1488
558712cf 1489# These are used only if $annotate is true.
c4019d52
KW
1490# The entire range of Unicode characters is examined to populate these
1491# after all the input has been processed. But most can be skipped, as they
1492# have the same descriptive phrases, such as being unassigned
1493my @viacode; # Contains the 1 million character names
1494my @printable; # boolean: And are those characters printable?
1495my @annotate_char_type; # Contains a type of those characters, specifically
1496 # for the purposes of annotation.
1497my $annotate_ranges; # A map of ranges of code points that have the same
98dc9551 1498 # name for the purposes of annotation. They map to the
c4019d52
KW
1499 # upper edge of the range, so that the end point can
1500 # be immediately found. This is used to skip ahead to
1501 # the end of a range, and avoid processing each
1502 # individual code point in it.
1503my $unassigned_sans_noncharacters; # A Range_List of the unassigned
1504 # characters, but excluding those which are
1505 # also noncharacter code points
1506
1507# The annotation types are an extension of the regular range types, though
1508# some of the latter are folded into one. Make the new types negative to
1509# avoid conflicting with the regular types
1510my $SURROGATE_TYPE = -1;
1511my $UNASSIGNED_TYPE = -2;
1512my $PRIVATE_USE_TYPE = -3;
1513my $NONCHARACTER_TYPE = -4;
1514my $CONTROL_TYPE = -5;
1515my $UNKNOWN_TYPE = -6; # Used only if there is a bug in this program
1516
1517sub populate_char_info ($) {
558712cf 1518 # Used only with the $annotate option. Populates the arrays with the
c4019d52
KW
1519 # input code point's info that are needed for outputting more detailed
1520 # comments. If calling context wants a return, it is the end point of
1521 # any contiguous range of characters that share essentially the same info
1522
1523 my $i = shift;
1524 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1525
1526 $viacode[$i] = $perl_charname->value_of($i) || "";
1527
1528 # A character is generally printable if Unicode says it is,
1529 # but below we make sure that most Unicode general category 'C' types
1530 # aren't.
1531 $printable[$i] = $print->contains($i);
1532
1533 $annotate_char_type[$i] = $perl_charname->type_of($i) || 0;
1534
1535 # Only these two regular types are treated specially for annotations
1536 # purposes
1537 $annotate_char_type[$i] = 0 if $annotate_char_type[$i] != $CP_IN_NAME
1538 && $annotate_char_type[$i] != $HANGUL_SYLLABLE;
1539
1540 # Give a generic name to all code points that don't have a real name.
1541 # We output ranges, if applicable, for these. Also calculate the end
1542 # point of the range.
1543 my $end;
1544 if (! $viacode[$i]) {
1d025d66
KW
1545 my $nonchar;
1546 if ($gc-> table('Private_use')->contains($i)) {
c4019d52
KW
1547 $viacode[$i] = 'Private Use';
1548 $annotate_char_type[$i] = $PRIVATE_USE_TYPE;
1549 $printable[$i] = 0;
1550 $end = $gc->table('Private_Use')->containing_range($i)->end;
1551 }
1d025d66
KW
1552 elsif ((defined ($nonchar =
1553 Property::property_ref('Noncharacter_Code_Point'))
1554 && $nonchar->table('Y')->contains($i)))
c4019d52
KW
1555 {
1556 $viacode[$i] = 'Noncharacter';
1557 $annotate_char_type[$i] = $NONCHARACTER_TYPE;
1558 $printable[$i] = 0;
1559 $end = property_ref('Noncharacter_Code_Point')->table('Y')->
1560 containing_range($i)->end;
1561 }
1562 elsif ($gc-> table('Control')->contains($i)) {
c71dea7f 1563 $viacode[$i] = property_ref('Name_Alias')->value_of($i) || 'Control';
c4019d52
KW
1564 $annotate_char_type[$i] = $CONTROL_TYPE;
1565 $printable[$i] = 0;
c4019d52
KW
1566 }
1567 elsif ($gc-> table('Unassigned')->contains($i)) {
c4019d52
KW
1568 $annotate_char_type[$i] = $UNASSIGNED_TYPE;
1569 $printable[$i] = 0;
1d025d66
KW
1570 if ($v_version lt v2.0.0) { # No blocks in earliest releases
1571 $viacode[$i] = 'Unassigned';
1572 $end = $gc-> table('Unassigned')->containing_range($i)->end;
1573 }
1574 else {
1575 $viacode[$i] = 'Unassigned, block=' . $block-> value_of($i);
c4019d52 1576
bf06c733
KW
1577 # Because we name the unassigned by the blocks they are in, it
1578 # can't go past the end of that block, and it also can't go
1579 # past the unassigned range it is in. The special table makes
1580 # sure that the non-characters, which are unassigned, are
1581 # separated out.
1582 $end = min($block->containing_range($i)->end,
1583 $unassigned_sans_noncharacters->
1584 containing_range($i)->end);
1d025d66
KW
1585 }
1586 }
1587 elsif ($v_version lt v2.0.0) { # No surrogates in earliest releases
1588 $viacode[$i] = $gc->value_of($i);
1589 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1590 $printable[$i] = 0;
1591 }
1592 elsif ($gc-> table('Surrogate')->contains($i)) {
1593 $viacode[$i] = 'Surrogate';
1594 $annotate_char_type[$i] = $SURROGATE_TYPE;
1595 $printable[$i] = 0;
1596 $end = $gc->table('Surrogate')->containing_range($i)->end;
13ca76ff
KW
1597 }
1598 else {
1599 Carp::my_carp_bug("Can't figure out how to annotate "
1600 . sprintf("U+%04X", $i)
1601 . ". Proceeding anyway.");
c4019d52
KW
1602 $viacode[$i] = 'UNKNOWN';
1603 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1604 $printable[$i] = 0;
1605 }
1606 }
1607
1608 # Here, has a name, but if it's one in which the code point number is
1609 # appended to the name, do that.
1610 elsif ($annotate_char_type[$i] == $CP_IN_NAME) {
1611 $viacode[$i] .= sprintf("-%04X", $i);
1612 $end = $perl_charname->containing_range($i)->end;
1613 }
1614
1615 # And here, has a name, but if it's a hangul syllable one, replace it with
1616 # the correct name from the Unicode algorithm
1617 elsif ($annotate_char_type[$i] == $HANGUL_SYLLABLE) {
1618 use integer;
1619 my $SIndex = $i - $SBase;
1620 my $L = $LBase + $SIndex / $NCount;
1621 my $V = $VBase + ($SIndex % $NCount) / $TCount;
1622 my $T = $TBase + $SIndex % $TCount;
1623 $viacode[$i] = "HANGUL SYLLABLE $Jamo{$L}$Jamo{$V}";
1624 $viacode[$i] .= $Jamo{$T} if $T != $TBase;
1625 $end = $perl_charname->containing_range($i)->end;
1626 }
1627
1628 return if ! defined wantarray;
1629 return $i if ! defined $end; # If not a range, return the input
1630
1631 # Save this whole range so can find the end point quickly
1632 $annotate_ranges->add_map($i, $end, $end);
1633
1634 return $end;
1635}
1636
23e33b60
KW
1637# Commented code below should work on Perl 5.8.
1638## This 'require' doesn't necessarily work in miniperl, and even if it does,
1639## the native perl version of it (which is what would operate under miniperl)
1640## is extremely slow, as it does a string eval every call.
7e017d6d 1641#my $has_fast_scalar_util = $^X !~ /miniperl/
23e33b60
KW
1642# && defined eval "require Scalar::Util";
1643#
1644#sub objaddr($) {
1645# # Returns the address of the blessed input object. Uses the XS version if
1646# # available. It doesn't check for blessedness because that would do a
1647# # string eval every call, and the program is structured so that this is
1648# # never called for a non-blessed object.
1649#
1650# return Scalar::Util::refaddr($_[0]) if $has_fast_scalar_util;
1651#
1652# # Check at least that is a ref.
1653# my $pkg = ref($_[0]) or return undef;
1654#
1655# # Change to a fake package to defeat any overloaded stringify
1656# bless $_[0], 'main::Fake';
1657#
1658# # Numifying a ref gives its address.
051df77b 1659# my $addr = pack 'J', $_[0];
23e33b60
KW
1660#
1661# # Return to original class
1662# bless $_[0], $pkg;
1663# return $addr;
1664#}
1665
99870f4d
KW
1666sub max ($$) {
1667 my $a = shift;
1668 my $b = shift;
1669 return $a if $a >= $b;
1670 return $b;
1671}
1672
1673sub min ($$) {
1674 my $a = shift;
1675 my $b = shift;
1676 return $a if $a <= $b;
1677 return $b;
1678}
1679
1680sub clarify_number ($) {
1681 # This returns the input number with underscores inserted every 3 digits
1682 # in large (5 digits or more) numbers. Input must be entirely digits, not
1683 # checked.
1684
1685 my $number = shift;
1686 my $pos = length($number) - 3;
1687 return $number if $pos <= 1;
1688 while ($pos > 0) {
1689 substr($number, $pos, 0) = '_';
1690 $pos -= 3;
5beb625e 1691 }
99870f4d 1692 return $number;
99598c8c
JH
1693}
1694
12ac2576 1695
99870f4d 1696package Carp;
7ebf06b3 1697
99870f4d
KW
1698# These routines give a uniform treatment of messages in this program. They
1699# are placed in the Carp package to cause the stack trace to not include them,
1700# although an alternative would be to use another package and set @CARP_NOT
1701# for it.
12ac2576 1702
99870f4d 1703our $Verbose = 1 if main::DEBUG; # Useful info when debugging
12ac2576 1704
99f78760
KW
1705# This is a work-around suggested by Nicholas Clark to fix a problem with Carp
1706# and overload trying to load Scalar:Util under miniperl. See
1707# http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2009-11/msg01057.html
1708undef $overload::VERSION;
1709
99870f4d
KW
1710sub my_carp {
1711 my $message = shift || "";
1712 my $nofold = shift || 0;
7ebf06b3 1713
99870f4d
KW
1714 if ($message) {
1715 $message = main::join_lines($message);
1716 $message =~ s/^$0: *//; # Remove initial program name
1717 $message =~ s/[.;,]+$//; # Remove certain ending punctuation
1718 $message = "\n$0: $message;";
12ac2576 1719
99870f4d
KW
1720 # Fold the message with program name, semi-colon end punctuation
1721 # (which looks good with the message that carp appends to it), and a
1722 # hanging indent for continuation lines.
1723 $message = main::simple_fold($message, "", 4) unless $nofold;
1724 $message =~ s/\n$//; # Remove the trailing nl so what carp
1725 # appends is to the same line
1726 }
12ac2576 1727
99870f4d 1728 return $message if defined wantarray; # If a caller just wants the msg
12ac2576 1729
99870f4d
KW
1730 carp $message;
1731 return;
1732}
7ebf06b3 1733
99870f4d
KW
1734sub my_carp_bug {
1735 # This is called when it is clear that the problem is caused by a bug in
1736 # this program.
7ebf06b3 1737
99870f4d
KW
1738 my $message = shift;
1739 $message =~ s/^$0: *//;
1740 $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");
1741 carp $message;
1742 return;
1743}
7ebf06b3 1744
99870f4d
KW
1745sub carp_too_few_args {
1746 if (@_ != 2) {
1747 my_carp_bug("Wrong number of arguments: to 'carp_too_few_arguments'. No action taken.");
1748 return;
12ac2576 1749 }
7ebf06b3 1750
99870f4d
KW
1751 my $args_ref = shift;
1752 my $count = shift;
7ebf06b3 1753
99870f4d
KW
1754 my_carp_bug("Need at least $count arguments to "
1755 . (caller 1)[3]
1756 . ". Instead got: '"
1757 . join ', ', @$args_ref
1758 . "'. No action taken.");
1759 return;
12ac2576
JP
1760}
1761
99870f4d
KW
1762sub carp_extra_args {
1763 my $args_ref = shift;
1764 my_carp_bug("Too many arguments to 'carp_extra_args': (" . join(', ', @_) . "); Extras ignored.") if @_;
12ac2576 1765
99870f4d
KW
1766 unless (ref $args_ref) {
1767 my_carp_bug("Argument to 'carp_extra_args' ($args_ref) must be a ref. Not checking arguments.");
1768 return;
1769 }
1770 my ($package, $file, $line) = caller;
1771 my $subroutine = (caller 1)[3];
cf25bb62 1772
99870f4d
KW
1773 my $list;
1774 if (ref $args_ref eq 'HASH') {
1775 foreach my $key (keys %$args_ref) {
1776 $args_ref->{$key} = $UNDEF unless defined $args_ref->{$key};
cf25bb62 1777 }
99870f4d 1778 $list = join ', ', each %{$args_ref};
cf25bb62 1779 }
99870f4d
KW
1780 elsif (ref $args_ref eq 'ARRAY') {
1781 foreach my $arg (@$args_ref) {
1782 $arg = $UNDEF unless defined $arg;
1783 }
1784 $list = join ', ', @$args_ref;
1785 }
1786 else {
1787 my_carp_bug("Can't cope with ref "
1788 . ref($args_ref)
1789 . " . argument to 'carp_extra_args'. Not checking arguments.");
1790 return;
1791 }
1792
1793 my_carp_bug("Unrecognized parameters in options: '$list' to $subroutine. Skipped.");
1794 return;
d73e5302
JH
1795}
1796
99870f4d
KW
1797package main;
1798
1799{ # Closure
1800
1801 # This program uses the inside-out method for objects, as recommended in
1802 # "Perl Best Practices". This closure aids in generating those. There
1803 # are two routines. setup_package() is called once per package to set
1804 # things up, and then set_access() is called for each hash representing a
1805 # field in the object. These routines arrange for the object to be
1806 # properly destroyed when no longer used, and for standard accessor
1807 # functions to be generated. If you need more complex accessors, just
1808 # write your own and leave those accesses out of the call to set_access().
1809 # More details below.
1810
1811 my %constructor_fields; # fields that are to be used in constructors; see
1812 # below
1813
1814 # The values of this hash will be the package names as keys to other
1815 # hashes containing the name of each field in the package as keys, and
1816 # references to their respective hashes as values.
1817 my %package_fields;
1818
1819 sub setup_package {
1820 # Sets up the package, creating standard DESTROY and dump methods
1821 # (unless already defined). The dump method is used in debugging by
1822 # simple_dumper().
1823 # The optional parameters are:
1824 # a) a reference to a hash, that gets populated by later
1825 # set_access() calls with one of the accesses being
1826 # 'constructor'. The caller can then refer to this, but it is
1827 # not otherwise used by these two routines.
1828 # b) a reference to a callback routine to call during destruction
1829 # of the object, before any fields are actually destroyed
1830
1831 my %args = @_;
1832 my $constructor_ref = delete $args{'Constructor_Fields'};
1833 my $destroy_callback = delete $args{'Destroy_Callback'};
1834 Carp::carp_extra_args(\@_) if main::DEBUG && %args;
1835
1836 my %fields;
1837 my $package = (caller)[0];
1838
1839 $package_fields{$package} = \%fields;
1840 $constructor_fields{$package} = $constructor_ref;
1841
1842 unless ($package->can('DESTROY')) {
1843 my $destroy_name = "${package}::DESTROY";
1844 no strict "refs";
1845
1846 # Use typeglob to give the anonymous subroutine the name we want
1847 *$destroy_name = sub {
1848 my $self = shift;
ffe43484 1849 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1850
1851 $self->$destroy_callback if $destroy_callback;
1852 foreach my $field (keys %{$package_fields{$package}}) {
1853 #print STDERR __LINE__, ": Destroying ", ref $self, " ", sprintf("%04X", $addr), ": ", $field, "\n";
1854 delete $package_fields{$package}{$field}{$addr};
1855 }
1856 return;
1857 }
1858 }
1859
1860 unless ($package->can('dump')) {
1861 my $dump_name = "${package}::dump";
1862 no strict "refs";
1863 *$dump_name = sub {
1864 my $self = shift;
1865 return dump_inside_out($self, $package_fields{$package}, @_);
1866 }
1867 }
1868 return;
1869 }
1870
1871 sub set_access {
1872 # Arrange for the input field to be garbage collected when no longer
1873 # needed. Also, creates standard accessor functions for the field
1874 # based on the optional parameters-- none if none of these parameters:
1875 # 'addable' creates an 'add_NAME()' accessor function.
1876 # 'readable' or 'readable_array' creates a 'NAME()' accessor
1877 # function.
1878 # 'settable' creates a 'set_NAME()' accessor function.
1879 # 'constructor' doesn't create an accessor function, but adds the
1880 # field to the hash that was previously passed to
1881 # setup_package();
1882 # Any of the accesses can be abbreviated down, so that 'a', 'ad',
1883 # 'add' etc. all mean 'addable'.
1884 # The read accessor function will work on both array and scalar
1885 # values. If another accessor in the parameter list is 'a', the read
1886 # access assumes an array. You can also force it to be array access
1887 # by specifying 'readable_array' instead of 'readable'
1888 #
1889 # A sort-of 'protected' access can be set-up by preceding the addable,
1890 # readable or settable with some initial portion of 'protected_' (but,
1891 # the underscore is required), like 'p_a', 'pro_set', etc. The
1892 # "protection" is only by convention. All that happens is that the
1893 # accessor functions' names begin with an underscore. So instead of
1894 # calling set_foo, the call is _set_foo. (Real protection could be
c1739a4a 1895 # accomplished by having a new subroutine, end_package, called at the
99870f4d
KW
1896 # end of each package, and then storing the __LINE__ ranges and
1897 # checking them on every accessor. But that is way overkill.)
1898
1899 # We create anonymous subroutines as the accessors and then use
1900 # typeglobs to assign them to the proper package and name
1901
1902 my $name = shift; # Name of the field
1903 my $field = shift; # Reference to the inside-out hash containing the
1904 # field
1905
1906 my $package = (caller)[0];
1907
1908 if (! exists $package_fields{$package}) {
1909 croak "$0: Must call 'setup_package' before 'set_access'";
1910 }
d73e5302 1911
99870f4d
KW
1912 # Stash the field so DESTROY can get it.
1913 $package_fields{$package}{$name} = $field;
cf25bb62 1914
99870f4d
KW
1915 # Remaining arguments are the accessors. For each...
1916 foreach my $access (@_) {
1917 my $access = lc $access;
cf25bb62 1918
99870f4d 1919 my $protected = "";
cf25bb62 1920
99870f4d
KW
1921 # Match the input as far as it goes.
1922 if ($access =~ /^(p[^_]*)_/) {
1923 $protected = $1;
1924 if (substr('protected_', 0, length $protected)
1925 eq $protected)
1926 {
1927
1928 # Add 1 for the underscore not included in $protected
1929 $access = substr($access, length($protected) + 1);
1930 $protected = '_';
1931 }
1932 else {
1933 $protected = "";
1934 }
1935 }
1936
1937 if (substr('addable', 0, length $access) eq $access) {
1938 my $subname = "${package}::${protected}add_$name";
1939 no strict "refs";
1940
1941 # add_ accessor. Don't add if already there, which we
1942 # determine using 'eq' for scalars and '==' otherwise.
1943 *$subname = sub {
1944 use strict "refs";
1945 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
1946 my $self = shift;
1947 my $value = shift;
ffe43484 1948 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1949 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1950 if (ref $value) {
f998e60c 1951 return if grep { $value == $_ } @{$field->{$addr}};
99870f4d
KW
1952 }
1953 else {
f998e60c 1954 return if grep { $value eq $_ } @{$field->{$addr}};
99870f4d 1955 }
f998e60c 1956 push @{$field->{$addr}}, $value;
99870f4d
KW
1957 return;
1958 }
1959 }
1960 elsif (substr('constructor', 0, length $access) eq $access) {
1961 if ($protected) {
1962 Carp::my_carp_bug("Can't set-up 'protected' constructors")
1963 }
1964 else {
1965 $constructor_fields{$package}{$name} = $field;
1966 }
1967 }
1968 elsif (substr('readable_array', 0, length $access) eq $access) {
1969
1970 # Here has read access. If one of the other parameters for
1971 # access is array, or this one specifies array (by being more
1972 # than just 'readable_'), then create a subroutine that
1973 # assumes the data is an array. Otherwise just a scalar
1974 my $subname = "${package}::${protected}$name";
1975 if (grep { /^a/i } @_
1976 or length($access) > length('readable_'))
1977 {
1978 no strict "refs";
1979 *$subname = sub {
1980 use strict "refs";
23e33b60 1981 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
ffe43484 1982 my $addr = do { no overloading; pack 'J', $_[0]; };
99870f4d
KW
1983 if (ref $field->{$addr} ne 'ARRAY') {
1984 my $type = ref $field->{$addr};
1985 $type = 'scalar' unless $type;
1986 Carp::my_carp_bug("Trying to read $name as an array when it is a $type. Big problems.");
1987 return;
1988 }
1989 return scalar @{$field->{$addr}} unless wantarray;
1990
1991 # Make a copy; had problems with caller modifying the
1992 # original otherwise
1993 my @return = @{$field->{$addr}};
1994 return @return;
1995 }
1996 }
1997 else {
1998
1999 # Here not an array value, a simpler function.
2000 no strict "refs";
2001 *$subname = sub {
2002 use strict "refs";
23e33b60 2003 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
f998e60c 2004 no overloading;
051df77b 2005 return $field->{pack 'J', $_[0]};
99870f4d
KW
2006 }
2007 }
2008 }
2009 elsif (substr('settable', 0, length $access) eq $access) {
2010 my $subname = "${package}::${protected}set_$name";
2011 no strict "refs";
2012 *$subname = sub {
2013 use strict "refs";
23e33b60
KW
2014 if (main::DEBUG) {
2015 return Carp::carp_too_few_args(\@_, 2) if @_ < 2;
2016 Carp::carp_extra_args(\@_) if @_ > 2;
2017 }
2018 # $self is $_[0]; $value is $_[1]
f998e60c 2019 no overloading;
051df77b 2020 $field->{pack 'J', $_[0]} = $_[1];
99870f4d
KW
2021 return;
2022 }
2023 }
2024 else {
2025 Carp::my_carp_bug("Unknown accessor type $access. No accessor set.");
2026 }
cf25bb62 2027 }
99870f4d 2028 return;
cf25bb62 2029 }
99870f4d
KW
2030}
2031
2032package Input_file;
2033
2034# All input files use this object, which stores various attributes about them,
2035# and provides for convenient, uniform handling. The run method wraps the
2036# processing. It handles all the bookkeeping of opening, reading, and closing
2037# the file, returning only significant input lines.
2038#
2039# Each object gets a handler which processes the body of the file, and is
2040# called by run(). Most should use the generic, default handler, which has
2041# code scrubbed to handle things you might not expect. A handler should
2042# basically be a while(next_line()) {...} loop.
2043#
2044# You can also set up handlers to
537124e4 2045# 1) call before the first line is read, for pre processing
99870f4d
KW
2046# 2) call to adjust each line of the input before the main handler gets them
2047# 3) call upon EOF before the main handler exits its loop
537124e4 2048# 4) call at the end, for post processing
99870f4d
KW
2049#
2050# $_ is used to store the input line, and is to be filtered by the
2051# each_line_handler()s. So, if the format of the line is not in the desired
2052# format for the main handler, these are used to do that adjusting. They can
2053# be stacked (by enclosing them in an [ anonymous array ] in the constructor,
2054# so the $_ output of one is used as the input to the next. None of the other
2055# handlers are stackable, but could easily be changed to be so.
2056#
2057# Most of the handlers can call insert_lines() or insert_adjusted_lines()
2058# which insert the parameters as lines to be processed before the next input
2059# file line is read. This allows the EOF handler to flush buffers, for
2060# example. The difference between the two routines is that the lines inserted
2061# by insert_lines() are subjected to the each_line_handler()s. (So if you
2062# called it from such a handler, you would get infinite recursion.) Lines
2063# inserted by insert_adjusted_lines() go directly to the main handler without
2064# any adjustments. If the post-processing handler calls any of these, there
2065# will be no effect. Some error checking for these conditions could be added,
2066# but it hasn't been done.
2067#
2068# carp_bad_line() should be called to warn of bad input lines, which clears $_
2069# to prevent further processing of the line. This routine will output the
2070# message as a warning once, and then keep a count of the lines that have the
2071# same message, and output that count at the end of the file's processing.
2072# This keeps the number of messages down to a manageable amount.
2073#
2074# get_missings() should be called to retrieve any @missing input lines.
2075# Messages will be raised if this isn't done if the options aren't to ignore
2076# missings.
2077
2078sub trace { return main::trace(@_); }
2079
99870f4d
KW
2080{ # Closure
2081 # Keep track of fields that are to be put into the constructor.
2082 my %constructor_fields;
2083
2084 main::setup_package(Constructor_Fields => \%constructor_fields);
2085
2086 my %file; # Input file name, required
2087 main::set_access('file', \%file, qw{ c r });
2088
2089 my %first_released; # Unicode version file was first released in, required
2090 main::set_access('first_released', \%first_released, qw{ c r });
2091
2092 my %handler; # Subroutine to process the input file, defaults to
2093 # 'process_generic_property_file'
2094 main::set_access('handler', \%handler, qw{ c });
2095
2096 my %property;
2097 # name of property this file is for. defaults to none, meaning not
2098 # applicable, or is otherwise determinable, for example, from each line.
696609bf 2099 main::set_access('property', \%property, qw{ c r });
99870f4d
KW
2100
2101 my %optional;
2102 # If this is true, the file is optional. If not present, no warning is
2103 # output. If it is present, the string given by this parameter is
2104 # evaluated, and if false the file is not processed.
2105 main::set_access('optional', \%optional, 'c', 'r');
2106
2107 my %non_skip;
2108 # This is used for debugging, to skip processing of all but a few input
2109 # files. Add 'non_skip => 1' to the constructor for those files you want
2110 # processed when you set the $debug_skip global.
2111 main::set_access('non_skip', \%non_skip, 'c');
2112
37e2e78e 2113 my %skip;
09ca89ce
KW
2114 # This is used to skip processing of this input file semi-permanently,
2115 # when it evaluates to true. The value should be the reason the file is
2116 # being skipped. It is used for files that we aren't planning to process
2117 # anytime soon, but want to allow to be in the directory and not raise a
2118 # message that we are not handling. Mostly for test files. This is in
2119 # contrast to the non_skip element, which is supposed to be used very
2120 # temporarily for debugging. Sets 'optional' to 1. Also, files that we
2121 # pretty much will never look at can be placed in the global
1fec9f60 2122 # %ignored_files instead. Ones used here will be added to %skipped files
37e2e78e
KW
2123 main::set_access('skip', \%skip, 'c');
2124
99870f4d
KW
2125 my %each_line_handler;
2126 # list of subroutines to look at and filter each non-comment line in the
2127 # file. defaults to none. The subroutines are called in order, each is
2128 # to adjust $_ for the next one, and the final one adjusts it for
2129 # 'handler'
2130 main::set_access('each_line_handler', \%each_line_handler, 'c');
2131
2132 my %has_missings_defaults;
2133 # ? Are there lines in the file giving default values for code points
2134 # missing from it?. Defaults to NO_DEFAULTS. Otherwise NOT_IGNORED is
2135 # the norm, but IGNORED means it has such lines, but the handler doesn't
2136 # use them. Having these three states allows us to catch changes to the
2137 # UCD that this program should track
2138 main::set_access('has_missings_defaults',
2139 \%has_missings_defaults, qw{ c r });
2140
2141 my %pre_handler;
2142 # Subroutine to call before doing anything else in the file. If undef, no
2143 # such handler is called.
2144 main::set_access('pre_handler', \%pre_handler, qw{ c });
2145
2146 my %eof_handler;
2147 # Subroutine to call upon getting an EOF on the input file, but before
2148 # that is returned to the main handler. This is to allow buffers to be
2149 # flushed. The handler is expected to call insert_lines() or
2150 # insert_adjusted() with the buffered material
2151 main::set_access('eof_handler', \%eof_handler, qw{ c r });
2152
2153 my %post_handler;
2154 # Subroutine to call after all the lines of the file are read in and
2155 # processed. If undef, no such handler is called.
2156 main::set_access('post_handler', \%post_handler, qw{ c });
2157
2158 my %progress_message;
2159 # Message to print to display progress in lieu of the standard one
2160 main::set_access('progress_message', \%progress_message, qw{ c });
2161
2162 my %handle;
2163 # cache open file handle, internal. Is undef if file hasn't been
2164 # processed at all, empty if has;
2165 main::set_access('handle', \%handle);
2166
2167 my %added_lines;
2168 # cache of lines added virtually to the file, internal
2169 main::set_access('added_lines', \%added_lines);
2170
2171 my %errors;
2172 # cache of errors found, internal
2173 main::set_access('errors', \%errors);
2174
2175 my %missings;
2176 # storage of '@missing' defaults lines
2177 main::set_access('missings', \%missings);
2178
2179 sub new {
2180 my $class = shift;
2181
2182 my $self = bless \do{ my $anonymous_scalar }, $class;
ffe43484 2183 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2184
2185 # Set defaults
2186 $handler{$addr} = \&main::process_generic_property_file;
2187 $non_skip{$addr} = 0;
37e2e78e 2188 $skip{$addr} = 0;
99870f4d
KW
2189 $has_missings_defaults{$addr} = $NO_DEFAULTS;
2190 $handle{$addr} = undef;
2191 $added_lines{$addr} = [ ];
2192 $each_line_handler{$addr} = [ ];
2193 $errors{$addr} = { };
2194 $missings{$addr} = [ ];
2195
2196 # Two positional parameters.
99f78760 2197 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
99870f4d
KW
2198 $file{$addr} = main::internal_file_to_platform(shift);
2199 $first_released{$addr} = shift;
2200
2201 # The rest of the arguments are key => value pairs
2202 # %constructor_fields has been set up earlier to list all possible
2203 # ones. Either set or push, depending on how the default has been set
2204 # up just above.
2205 my %args = @_;
2206 foreach my $key (keys %args) {
2207 my $argument = $args{$key};
2208
2209 # Note that the fields are the lower case of the constructor keys
2210 my $hash = $constructor_fields{lc $key};
2211 if (! defined $hash) {
2212 Carp::my_carp_bug("Unrecognized parameters '$key => $argument' to new() for $self. Skipped");
2213 next;
2214 }
2215 if (ref $hash->{$addr} eq 'ARRAY') {
2216 if (ref $argument eq 'ARRAY') {
2217 foreach my $argument (@{$argument}) {
2218 next if ! defined $argument;
2219 push @{$hash->{$addr}}, $argument;
2220 }
2221 }
2222 else {
2223 push @{$hash->{$addr}}, $argument if defined $argument;
2224 }
2225 }
2226 else {
2227 $hash->{$addr} = $argument;
2228 }
2229 delete $args{$key};
2230 };
2231
2232 # If the file has a property for it, it means that the property is not
2233 # listed in the file's entries. So add a handler to the list of line
2234 # handlers to insert the property name into the lines, to provide a
2235 # uniform interface to the final processing subroutine.
2236 # the final code doesn't have to worry about that.
2237 if ($property{$addr}) {
2238 push @{$each_line_handler{$addr}}, \&_insert_property_into_line;
2239 }
2240
2241 if ($non_skip{$addr} && ! $debug_skip && $verbosity) {
2242 print "Warning: " . __PACKAGE__ . " constructor for $file{$addr} has useless 'non_skip' in it\n";
a3a8c5f0 2243 }
99870f4d 2244
09ca89ce
KW
2245 # If skipping, set to optional, and add to list of ignored files,
2246 # including its reason
2247 if ($skip{$addr}) {
2248 $optional{$addr} = 1;
1fec9f60 2249 $skipped_files{$file{$addr}} = $skip{$addr}
09ca89ce 2250 }
37e2e78e 2251
99870f4d 2252 return $self;
d73e5302
JH
2253 }
2254
cf25bb62 2255
99870f4d
KW
2256 use overload
2257 fallback => 0,
2258 qw("") => "_operator_stringify",
2259 "." => \&main::_operator_dot,
1285127e 2260 ".=" => \&main::_operator_dot_equal,
99870f4d 2261 ;
cf25bb62 2262
99870f4d
KW
2263 sub _operator_stringify {
2264 my $self = shift;
cf25bb62 2265
99870f4d 2266 return __PACKAGE__ . " object for " . $self->file;
d73e5302 2267 }
d73e5302 2268
99870f4d
KW
2269 # flag to make sure extracted files are processed early
2270 my $seen_non_extracted_non_age = 0;
d73e5302 2271
99870f4d
KW
2272 sub run {
2273 # Process the input object $self. This opens and closes the file and
2274 # calls all the handlers for it. Currently, this can only be called
2275 # once per file, as it destroy's the EOF handler
d73e5302 2276
99870f4d
KW
2277 my $self = shift;
2278 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
b6922eda 2279
ffe43484 2280 my $addr = do { no overloading; pack 'J', $self; };
b6922eda 2281
99870f4d 2282 my $file = $file{$addr};
d73e5302 2283
99870f4d
KW
2284 # Don't process if not expecting this file (because released later
2285 # than this Unicode version), and isn't there. This means if someone
2286 # copies it into an earlier version's directory, we will go ahead and
2287 # process it.
2288 return if $first_released{$addr} gt $v_version && ! -e $file;
2289
2290 # If in debugging mode and this file doesn't have the non-skip
2291 # flag set, and isn't one of the critical files, skip it.
2292 if ($debug_skip
2293 && $first_released{$addr} ne v0
2294 && ! $non_skip{$addr})
2295 {
2296 print "Skipping $file in debugging\n" if $verbosity;
2297 return;
2298 }
2299
2300 # File could be optional
37e2e78e 2301 if ($optional{$addr}) {
99870f4d
KW
2302 return unless -e $file;
2303 my $result = eval $optional{$addr};
2304 if (! defined $result) {
2305 Carp::my_carp_bug("Got '$@' when tried to eval $optional{$addr}. $file Skipped.");
2306 return;
2307 }
2308 if (! $result) {
2309 if ($verbosity) {
2310 print STDERR "Skipping processing input file '$file' because '$optional{$addr}' is not true\n";
2311 }
2312 return;
2313 }
2314 }
2315
2316 if (! defined $file || ! -e $file) {
2317
2318 # If the file doesn't exist, see if have internal data for it
2319 # (based on first_released being 0).
2320 if ($first_released{$addr} eq v0) {
2321 $handle{$addr} = 'pretend_is_open';
2322 }
2323 else {
2324 if (! $optional{$addr} # File could be optional
2325 && $v_version ge $first_released{$addr})
2326 {
2327 print STDERR "Skipping processing input file '$file' because not found\n" if $v_version ge $first_released{$addr};
2328 }
2329 return;
2330 }
2331 }
2332 else {
2333
37e2e78e
KW
2334 # Here, the file exists. Some platforms may change the case of
2335 # its name
99870f4d 2336 if ($seen_non_extracted_non_age) {
517956bf 2337 if ($file =~ /$EXTRACTED/i) {
1675ea0d 2338 Carp::my_carp_bug(main::join_lines(<<END
99f78760 2339$file should be processed just after the 'Prop...Alias' files, and before
99870f4d
KW
2340anything not in the $EXTRACTED_DIR directory. Proceeding, but the results may
2341have subtle problems
2342END
2343 ));
2344 }
2345 }
2346 elsif ($EXTRACTED_DIR
2347 && $first_released{$addr} ne v0
517956bf
CB
2348 && $file !~ /$EXTRACTED/i
2349 && lc($file) ne 'dage.txt')
99870f4d
KW
2350 {
2351 # We don't set this (by the 'if' above) if we have no
2352 # extracted directory, so if running on an early version,
2353 # this test won't work. Not worth worrying about.
2354 $seen_non_extracted_non_age = 1;
2355 }
2356
2357 # And mark the file as having being processed, and warn if it
2358 # isn't a file we are expecting. As we process the files,
2359 # they are deleted from the hash, so any that remain at the
2360 # end of the program are files that we didn't process.
517956bf 2361 my $fkey = File::Spec->rel2abs($file);
faf3cf6b
KW
2362 my $expecting = delete $potential_files{lc($fkey)};
2363
678f13d5
KW
2364 Carp::my_carp("Was not expecting '$file'.") if
2365 ! $expecting
99870f4d
KW
2366 && ! defined $handle{$addr};
2367
37e2e78e
KW
2368 # Having deleted from expected files, we can quit if not to do
2369 # anything. Don't print progress unless really want verbosity
2370 if ($skip{$addr}) {
2371 print "Skipping $file.\n" if $verbosity >= $VERBOSE;
2372 return;
2373 }
2374
99870f4d
KW
2375 # Open the file, converting the slashes used in this program
2376 # into the proper form for the OS
2377 my $file_handle;
2378 if (not open $file_handle, "<", $file) {
2379 Carp::my_carp("Can't open $file. Skipping: $!");
2380 return 0;
2381 }
2382 $handle{$addr} = $file_handle; # Cache the open file handle
2383 }
2384
2385 if ($verbosity >= $PROGRESS) {
2386 if ($progress_message{$addr}) {
2387 print "$progress_message{$addr}\n";
2388 }
2389 else {
2390 # If using a virtual file, say so.
2391 print "Processing ", (-e $file)
2392 ? $file
2393 : "substitute $file",
2394 "\n";
2395 }
2396 }
2397
2398
2399 # Call any special handler for before the file.
2400 &{$pre_handler{$addr}}($self) if $pre_handler{$addr};
2401
2402 # Then the main handler
2403 &{$handler{$addr}}($self);
2404
2405 # Then any special post-file handler.
2406 &{$post_handler{$addr}}($self) if $post_handler{$addr};
2407
2408 # If any errors have been accumulated, output the counts (as the first
2409 # error message in each class was output when it was encountered).
2410 if ($errors{$addr}) {
2411 my $total = 0;
2412 my $types = 0;
2413 foreach my $error (keys %{$errors{$addr}}) {
2414 $total += $errors{$addr}->{$error};
2415 delete $errors{$addr}->{$error};
2416 $types++;
2417 }
2418 if ($total > 1) {
2419 my $message
2420 = "A total of $total lines had errors in $file. ";
2421
2422 $message .= ($types == 1)
2423 ? '(Only the first one was displayed.)'
2424 : '(Only the first of each type was displayed.)';
2425 Carp::my_carp($message);
2426 }
2427 }
2428
2429 if (@{$missings{$addr}}) {
2430 Carp::my_carp_bug("Handler for $file didn't look at all the \@missing lines. Generated tables likely are wrong");
2431 }
2432
2433 # If a real file handle, close it.
2434 close $handle{$addr} or Carp::my_carp("Can't close $file: $!") if
2435 ref $handle{$addr};
2436 $handle{$addr} = ""; # Uses empty to indicate that has already seen
2437 # the file, as opposed to undef
2438 return;
2439 }
2440
2441 sub next_line {
2442 # Sets $_ to be the next logical input line, if any. Returns non-zero
2443 # if such a line exists. 'logical' means that any lines that have
2444 # been added via insert_lines() will be returned in $_ before the file
2445 # is read again.
2446
2447 my $self = shift;
2448 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2449
ffe43484 2450 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2451
2452 # Here the file is open (or if the handle is not a ref, is an open
2453 # 'virtual' file). Get the next line; any inserted lines get priority
2454 # over the file itself.
2455 my $adjusted;
2456
2457 LINE:
2458 while (1) { # Loop until find non-comment, non-empty line
2459 #local $to_trace = 1 if main::DEBUG;
2460 my $inserted_ref = shift @{$added_lines{$addr}};
2461 if (defined $inserted_ref) {
2462 ($adjusted, $_) = @{$inserted_ref};
2463 trace $adjusted, $_ if main::DEBUG && $to_trace;
2464 return 1 if $adjusted;
2465 }
2466 else {
2467 last if ! ref $handle{$addr}; # Don't read unless is real file
2468 last if ! defined ($_ = readline $handle{$addr});
2469 }
2470 chomp;
2471 trace $_ if main::DEBUG && $to_trace;
2472
2473 # See if this line is the comment line that defines what property
2474 # value that code points that are not listed in the file should
2475 # have. The format or existence of these lines is not guaranteed
2476 # by Unicode since they are comments, but the documentation says
2477 # that this was added for machine-readability, so probably won't
2478 # change. This works starting in Unicode Version 5.0. They look
2479 # like:
2480 #
2481 # @missing: 0000..10FFFF; Not_Reordered
2482 # @missing: 0000..10FFFF; Decomposition_Mapping; <code point>
2483 # @missing: 0000..10FFFF; ; NaN
2484 #
2485 # Save the line for a later get_missings() call.
2486 if (/$missing_defaults_prefix/) {
2487 if ($has_missings_defaults{$addr} == $NO_DEFAULTS) {
2488 $self->carp_bad_line("Unexpected \@missing line. Assuming no missing entries");
2489 }
2490 elsif ($has_missings_defaults{$addr} == $NOT_IGNORED) {
2491 my @defaults = split /\s* ; \s*/x, $_;
2492
2493 # The first field is the @missing, which ends in a
2494 # semi-colon, so can safely shift.
2495 shift @defaults;
2496
2497 # Some of these lines may have empty field placeholders
2498 # which get in the way. An example is:
2499 # @missing: 0000..10FFFF; ; NaN
2500 # Remove them. Process starting from the top so the
2501 # splice doesn't affect things still to be looked at.
2502 for (my $i = @defaults - 1; $i >= 0; $i--) {
2503 next if $defaults[$i] ne "";
2504 splice @defaults, $i, 1;
2505 }
2506
2507 # What's left should be just the property (maybe) and the
2508 # default. Having only one element means it doesn't have
2509 # the property.
2510 my $default;
2511 my $property;
2512 if (@defaults >= 1) {
2513 if (@defaults == 1) {
2514 $default = $defaults[0];
2515 }
2516 else {
2517 $property = $defaults[0];
2518 $default = $defaults[1];
2519 }
2520 }
2521
2522 if (@defaults < 1
2523 || @defaults > 2
2524 || ($default =~ /^</
2525 && $default !~ /^<code *point>$/i
09f8d0ac
KW
2526 && $default !~ /^<none>$/i
2527 && $default !~ /^<script>$/i))
99870f4d
KW
2528 {
2529 $self->carp_bad_line("Unrecognized \@missing line: $_. Assuming no missing entries");
2530 }
2531 else {
2532
2533 # If the property is missing from the line, it should
2534 # be the one for the whole file
2535 $property = $property{$addr} if ! defined $property;
2536
2537 # Change <none> to the null string, which is what it
2538 # really means. If the default is the code point
2539 # itself, set it to <code point>, which is what
2540 # Unicode uses (but sometimes they've forgotten the
2541 # space)
2542 if ($default =~ /^<none>$/i) {
2543 $default = "";
2544 }
2545 elsif ($default =~ /^<code *point>$/i) {
2546 $default = $CODE_POINT;
2547 }
09f8d0ac
KW
2548 elsif ($default =~ /^<script>$/i) {
2549
2550 # Special case this one. Currently is from
2551 # ScriptExtensions.txt, and means for all unlisted
2552 # code points, use their Script property values.
2553 # For the code points not listed in that file, the
2554 # default value is 'Unknown'.
2555 $default = "Unknown";
2556 }
99870f4d
KW
2557
2558 # Store them as a sub-arrays with both components.
2559 push @{$missings{$addr}}, [ $default, $property ];
2560 }
2561 }
2562
2563 # There is nothing for the caller to process on this comment
2564 # line.
2565 next;
2566 }
2567
2568 # Remove comments and trailing space, and skip this line if the
2569 # result is empty
2570 s/#.*//;
2571 s/\s+$//;
2572 next if /^$/;
2573
2574 # Call any handlers for this line, and skip further processing of
2575 # the line if the handler sets the line to null.
2576 foreach my $sub_ref (@{$each_line_handler{$addr}}) {
2577 &{$sub_ref}($self);
2578 next LINE if /^$/;
2579 }
2580
2581 # Here the line is ok. return success.
2582 return 1;
2583 } # End of looping through lines.
2584
2585 # If there is an EOF handler, call it (only once) and if it generates
2586 # more lines to process go back in the loop to handle them.
2587 if ($eof_handler{$addr}) {
2588 &{$eof_handler{$addr}}($self);
2589 $eof_handler{$addr} = ""; # Currently only get one shot at it.
2590 goto LINE if $added_lines{$addr};
2591 }
2592
2593 # Return failure -- no more lines.
2594 return 0;
2595
2596 }
2597
2598# Not currently used, not fully tested.
2599# sub peek {
2600# # Non-destructive look-ahead one non-adjusted, non-comment, non-blank
2601# # record. Not callable from an each_line_handler(), nor does it call
2602# # an each_line_handler() on the line.
2603#
2604# my $self = shift;
ffe43484 2605# my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2606#
2607# foreach my $inserted_ref (@{$added_lines{$addr}}) {
2608# my ($adjusted, $line) = @{$inserted_ref};
2609# next if $adjusted;
2610#
2611# # Remove comments and trailing space, and return a non-empty
2612# # resulting line
2613# $line =~ s/#.*//;
2614# $line =~ s/\s+$//;
2615# return $line if $line ne "";
2616# }
2617#
2618# return if ! ref $handle{$addr}; # Don't read unless is real file
2619# while (1) { # Loop until find non-comment, non-empty line
2620# local $to_trace = 1 if main::DEBUG;
2621# trace $_ if main::DEBUG && $to_trace;
2622# return if ! defined (my $line = readline $handle{$addr});
2623# chomp $line;
2624# push @{$added_lines{$addr}}, [ 0, $line ];
2625#
2626# $line =~ s/#.*//;
2627# $line =~ s/\s+$//;
2628# return $line if $line ne "";
2629# }
2630#
2631# return;
2632# }
2633
2634
2635 sub insert_lines {
2636 # Lines can be inserted so that it looks like they were in the input
2637 # file at the place it was when this routine is called. See also
2638 # insert_adjusted_lines(). Lines inserted via this routine go through
2639 # any each_line_handler()
2640
2641 my $self = shift;
2642
2643 # Each inserted line is an array, with the first element being 0 to
2644 # indicate that this line hasn't been adjusted, and needs to be
2645 # processed.
f998e60c 2646 no overloading;
051df77b 2647 push @{$added_lines{pack 'J', $self}}, map { [ 0, $_ ] } @_;
99870f4d
KW
2648 return;
2649 }
2650
2651 sub insert_adjusted_lines {
2652 # Lines can be inserted so that it looks like they were in the input
2653 # file at the place it was when this routine is called. See also
2654 # insert_lines(). Lines inserted via this routine are already fully
2655 # adjusted, ready to be processed; each_line_handler()s handlers will
2656 # not be called. This means this is not a completely general
2657 # facility, as only the last each_line_handler on the stack should
2658 # call this. It could be made more general, by passing to each of the
2659 # line_handlers their position on the stack, which they would pass on
2660 # to this routine, and that would replace the boolean first element in
2661 # the anonymous array pushed here, so that the next_line routine could
2662 # use that to call only those handlers whose index is after it on the
2663 # stack. But this is overkill for what is needed now.
2664
2665 my $self = shift;
2666 trace $_[0] if main::DEBUG && $to_trace;
2667
2668 # Each inserted line is an array, with the first element being 1 to
2669 # indicate that this line has been adjusted
f998e60c 2670 no overloading;
051df77b 2671 push @{$added_lines{pack 'J', $self}}, map { [ 1, $_ ] } @_;
99870f4d
KW
2672 return;
2673 }
2674
2675 sub get_missings {
2676 # Returns the stored up @missings lines' values, and clears the list.
2677 # The values are in an array, consisting of the default in the first
2678 # element, and the property in the 2nd. However, since these lines
2679 # can be stacked up, the return is an array of all these arrays.
2680
2681 my $self = shift;
2682 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2683
ffe43484 2684 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2685
2686 # If not accepting a list return, just return the first one.
2687 return shift @{$missings{$addr}} unless wantarray;
2688
2689 my @return = @{$missings{$addr}};
2690 undef @{$missings{$addr}};
2691 return @return;
2692 }
2693
2694 sub _insert_property_into_line {
2695 # Add a property field to $_, if this file requires it.
2696
f998e60c 2697 my $self = shift;
ffe43484 2698 my $addr = do { no overloading; pack 'J', $self; };
f998e60c 2699 my $property = $property{$addr};
99870f4d
KW
2700 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2701
2702 $_ =~ s/(;|$)/; $property$1/;
2703 return;
2704 }
2705
2706 sub carp_bad_line {
2707 # Output consistent error messages, using either a generic one, or the
2708 # one given by the optional parameter. To avoid gazillions of the
2709 # same message in case the syntax of a file is way off, this routine
2710 # only outputs the first instance of each message, incrementing a
2711 # count so the totals can be output at the end of the file.
2712
2713 my $self = shift;
2714 my $message = shift;
2715 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2716
ffe43484 2717 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2718
2719 $message = 'Unexpected line' unless $message;
2720
2721 # No trailing punctuation so as to fit with our addenda.
2722 $message =~ s/[.:;,]$//;
2723
2724 # If haven't seen this exact message before, output it now. Otherwise
2725 # increment the count of how many times it has occurred
2726 unless ($errors{$addr}->{$message}) {
2727 Carp::my_carp("$message in '$_' in "
f998e60c 2728 . $file{$addr}
99870f4d
KW
2729 . " at line $.. Skipping this line;");
2730 $errors{$addr}->{$message} = 1;
2731 }
2732 else {
2733 $errors{$addr}->{$message}++;
2734 }
2735
2736 # Clear the line to prevent any further (meaningful) processing of it.
2737 $_ = "";
2738
2739 return;
2740 }
2741} # End closure
2742
2743package Multi_Default;
2744
2745# Certain properties in early versions of Unicode had more than one possible
2746# default for code points missing from the files. In these cases, one
2747# default applies to everything left over after all the others are applied,
2748# and for each of the others, there is a description of which class of code
2749# points applies to it. This object helps implement this by storing the
2750# defaults, and for all but that final default, an eval string that generates
2751# the class that it applies to.
2752
2753
2754{ # Closure
2755
2756 main::setup_package();
2757
2758 my %class_defaults;
2759 # The defaults structure for the classes
2760 main::set_access('class_defaults', \%class_defaults);
2761
2762 my %other_default;
2763 # The default that applies to everything left over.
2764 main::set_access('other_default', \%other_default, 'r');
2765
2766
2767 sub new {
2768 # The constructor is called with default => eval pairs, terminated by
2769 # the left-over default. e.g.
2770 # Multi_Default->new(
2771 # 'T' => '$gc->table("Mn") + $gc->table("Cf") - 0x200C
2772 # - 0x200D',
2773 # 'R' => 'some other expression that evaluates to code points',
2774 # .
2775 # .
2776 # .
2777 # 'U'));
2778
2779 my $class = shift;
2780
2781 my $self = bless \do{my $anonymous_scalar}, $class;
ffe43484 2782 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2783
2784 while (@_ > 1) {
2785 my $default = shift;
2786 my $eval = shift;
2787 $class_defaults{$addr}->{$default} = $eval;
2788 }
2789
2790 $other_default{$addr} = shift;
2791
2792 return $self;
2793 }
2794
2795 sub get_next_defaults {
2796 # Iterates and returns the next class of defaults.
2797 my $self = shift;
2798 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2799
ffe43484 2800 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2801
2802 return each %{$class_defaults{$addr}};
2803 }
2804}
2805
2806package Alias;
2807
2808# An alias is one of the names that a table goes by. This class defines them
2809# including some attributes. Everything is currently setup in the
2810# constructor.
2811
2812
2813{ # Closure
2814
2815 main::setup_package();
2816
2817 my %name;
2818 main::set_access('name', \%name, 'r');
2819
2820 my %loose_match;
c12f2655 2821 # Should this name match loosely or not.
99870f4d
KW
2822 main::set_access('loose_match', \%loose_match, 'r');
2823
33e96e72
KW
2824 my %make_re_pod_entry;
2825 # Some aliases should not get their own entries in the re section of the
2826 # pod, because they are covered by a wild-card, and some we want to
2827 # discourage use of. Binary
f82fe4ba 2828 main::set_access('make_re_pod_entry', \%make_re_pod_entry, 'r', 's');
99870f4d 2829
fd1e3e84
KW
2830 my %ucd;
2831 # Is this documented to be accessible via Unicode::UCD
2832 main::set_access('ucd', \%ucd, 'r', 's');
2833
99870f4d
KW
2834 my %status;
2835 # Aliases have a status, like deprecated, or even suppressed (which means
2836 # they don't appear in documentation). Enum
2837 main::set_access('status', \%status, 'r');
2838
0eac1e20 2839 my %ok_as_filename;
99870f4d
KW
2840 # Similarly, some aliases should not be considered as usable ones for
2841 # external use, such as file names, or we don't want documentation to
2842 # recommend them. Boolean
0eac1e20 2843 main::set_access('ok_as_filename', \%ok_as_filename, 'r');
99870f4d
KW
2844
2845 sub new {
2846 my $class = shift;
2847
2848 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2849 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2850
2851 $name{$addr} = shift;
2852 $loose_match{$addr} = shift;
33e96e72 2853 $make_re_pod_entry{$addr} = shift;
0eac1e20 2854 $ok_as_filename{$addr} = shift;
99870f4d 2855 $status{$addr} = shift;
fd1e3e84 2856 $ucd{$addr} = shift;
99870f4d
KW
2857
2858 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2859
2860 # Null names are never ok externally
0eac1e20 2861 $ok_as_filename{$addr} = 0 if $name{$addr} eq "";
99870f4d
KW
2862
2863 return $self;
2864 }
2865}
2866
2867package Range;
2868
2869# A range is the basic unit for storing code points, and is described in the
2870# comments at the beginning of the program. Each range has a starting code
2871# point; an ending code point (not less than the starting one); a value
2872# that applies to every code point in between the two end-points, inclusive;
2873# and an enum type that applies to the value. The type is for the user's
2874# convenience, and has no meaning here, except that a non-zero type is
2875# considered to not obey the normal Unicode rules for having standard forms.
2876#
2877# The same structure is used for both map and match tables, even though in the
2878# latter, the value (and hence type) is irrelevant and could be used as a
2879# comment. In map tables, the value is what all the code points in the range
2880# map to. Type 0 values have the standardized version of the value stored as
2881# well, so as to not have to recalculate it a lot.
2882
2883sub trace { return main::trace(@_); }
2884
2885{ # Closure
2886
2887 main::setup_package();
2888
2889 my %start;
2890 main::set_access('start', \%start, 'r', 's');
2891
2892 my %end;
2893 main::set_access('end', \%end, 'r', 's');
2894
2895 my %value;
2896 main::set_access('value', \%value, 'r');
2897
2898 my %type;
2899 main::set_access('type', \%type, 'r');
2900
2901 my %standard_form;
2902 # The value in internal standard form. Defined only if the type is 0.
2903 main::set_access('standard_form', \%standard_form);
2904
2905 # Note that if these fields change, the dump() method should as well
2906
2907 sub new {
2908 return Carp::carp_too_few_args(\@_, 3) if main::DEBUG && @_ < 3;
2909 my $class = shift;
2910
2911 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2912 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2913
2914 $start{$addr} = shift;
2915 $end{$addr} = shift;
2916
2917 my %args = @_;
2918
2919 my $value = delete $args{'Value'}; # Can be 0
2920 $value = "" unless defined $value;
2921 $value{$addr} = $value;
2922
2923 $type{$addr} = delete $args{'Type'} || 0;
2924
2925 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2926
99870f4d
KW
2927 return $self;
2928 }
2929
2930 use overload
2931 fallback => 0,
2932 qw("") => "_operator_stringify",
2933 "." => \&main::_operator_dot,
1285127e 2934 ".=" => \&main::_operator_dot_equal,
99870f4d
KW
2935 ;
2936
2937 sub _operator_stringify {
2938 my $self = shift;
ffe43484 2939 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2940
2941 # Output it like '0041..0065 (value)'
2942 my $return = sprintf("%04X", $start{$addr})
2943 . '..'
2944 . sprintf("%04X", $end{$addr});
2945 my $value = $value{$addr};
2946 my $type = $type{$addr};
2947 $return .= ' (';
2948 $return .= "$value";
2949 $return .= ", Type=$type" if $type != 0;
2950 $return .= ')';
2951
2952 return $return;
2953 }
2954
2955 sub standard_form {
c292d35a
NC
2956 # Calculate the standard form only if needed, and cache the result.
2957 # The standard form is the value itself if the type is special.
2958 # This represents a considerable CPU and memory saving - at the time
2959 # of writing there are 368676 non-special objects, but the standard
2960 # form is only requested for 22047 of them - ie about 6%.
99870f4d
KW
2961
2962 my $self = shift;
2963 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2964
ffe43484 2965 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2966
2967 return $standard_form{$addr} if defined $standard_form{$addr};
c292d35a
NC
2968
2969 my $value = $value{$addr};
2970 return $value if $type{$addr};
2971 return $standard_form{$addr} = main::standardize($value);
99870f4d
KW
2972 }
2973
2974 sub dump {
2975 # Human, not machine readable. For machine readable, comment out this
2976 # entire routine and let the standard one take effect.
2977 my $self = shift;
2978 my $indent = shift;
2979 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2980
ffe43484 2981 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2982
2983 my $return = $indent
2984 . sprintf("%04X", $start{$addr})
2985 . '..'
2986 . sprintf("%04X", $end{$addr})
2987 . " '$value{$addr}';";
2988 if (! defined $standard_form{$addr}) {
2989 $return .= "(type=$type{$addr})";
2990 }
2991 elsif ($standard_form{$addr} ne $value{$addr}) {
2992 $return .= "(standard '$standard_form{$addr}')";
2993 }
2994 return $return;
2995 }
2996} # End closure
2997
2998package _Range_List_Base;
2999
3000# Base class for range lists. A range list is simply an ordered list of
3001# ranges, so that the ranges with the lowest starting numbers are first in it.
3002#
3003# When a new range is added that is adjacent to an existing range that has the
3004# same value and type, it merges with it to form a larger range.
3005#
3006# Ranges generally do not overlap, except that there can be multiple entries
3007# of single code point ranges. This is because of NameAliases.txt.
3008#
3009# In this program, there is a standard value such that if two different
3010# values, have the same standard value, they are considered equivalent. This
3011# value was chosen so that it gives correct results on Unicode data
3012
3013# There are a number of methods to manipulate range lists, and some operators
3014# are overloaded to handle them.
3015
99870f4d
KW
3016sub trace { return main::trace(@_); }
3017
3018{ # Closure
3019
3020 our $addr;
3021
5b348b71
KW
3022 # Max is initialized to a negative value that isn't adjacent to 0, for
3023 # simpler tests
3024 my $max_init = -2;
3025
99870f4d
KW
3026 main::setup_package();
3027
3028 my %ranges;
3029 # The list of ranges
3030 main::set_access('ranges', \%ranges, 'readable_array');
3031
3032 my %max;
3033 # The highest code point in the list. This was originally a method, but
3034 # actual measurements said it was used a lot.
3035 main::set_access('max', \%max, 'r');
3036
3037 my %each_range_iterator;
3038 # Iterator position for each_range()
3039 main::set_access('each_range_iterator', \%each_range_iterator);
3040
3041 my %owner_name_of;
3042 # Name of parent this is attached to, if any. Solely for better error
3043 # messages.
3044 main::set_access('owner_name_of', \%owner_name_of, 'p_r');
3045
3046 my %_search_ranges_cache;
3047 # A cache of the previous result from _search_ranges(), for better
3048 # performance
3049 main::set_access('_search_ranges_cache', \%_search_ranges_cache);
3050
3051 sub new {
3052 my $class = shift;
3053 my %args = @_;
3054
3055 # Optional initialization data for the range list.
3056 my $initialize = delete $args{'Initialize'};
3057
3058 my $self;
3059
3060 # Use _union() to initialize. _union() returns an object of this
3061 # class, which means that it will call this constructor recursively.
3062 # But it won't have this $initialize parameter so that it won't
3063 # infinitely loop on this.
3064 return _union($class, $initialize, %args) if defined $initialize;
3065
3066 $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 3067 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3068
3069 # Optional parent object, only for debug info.
3070 $owner_name_of{$addr} = delete $args{'Owner'};
3071 $owner_name_of{$addr} = "" if ! defined $owner_name_of{$addr};
3072
3073 # Stringify, in case it is an object.
3074 $owner_name_of{$addr} = "$owner_name_of{$addr}";
3075
3076 # This is used only for error messages, and so a colon is added
3077 $owner_name_of{$addr} .= ": " if $owner_name_of{$addr} ne "";
3078
3079 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3080
5b348b71 3081 $max{$addr} = $max_init;
99870f4d
KW
3082
3083 $_search_ranges_cache{$addr} = 0;
3084 $ranges{$addr} = [];
3085
3086 return $self;
3087 }
3088
3089 use overload
3090 fallback => 0,
3091 qw("") => "_operator_stringify",
3092 "." => \&main::_operator_dot,
1285127e 3093 ".=" => \&main::_operator_dot_equal,
99870f4d
KW
3094 ;
3095
3096 sub _operator_stringify {
3097 my $self = shift;
ffe43484 3098 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3099
3100 return "Range_List attached to '$owner_name_of{$addr}'"
3101 if $owner_name_of{$addr};
3102 return "anonymous Range_List " . \$self;
3103 }
3104
3105 sub _union {
3106 # Returns the union of the input code points. It can be called as
3107 # either a constructor or a method. If called as a method, the result
3108 # will be a new() instance of the calling object, containing the union
3109 # of that object with the other parameter's code points; if called as
d59563d0 3110 # a constructor, the first parameter gives the class that the new object
99870f4d
KW
3111 # should be, and the second parameter gives the code points to go into
3112 # it.
3113 # In either case, there are two parameters looked at by this routine;
3114 # any additional parameters are passed to the new() constructor.
3115 #
3116 # The code points can come in the form of some object that contains
3117 # ranges, and has a conventionally named method to access them; or
3118 # they can be an array of individual code points (as integers); or
3119 # just a single code point.
3120 #
3121 # If they are ranges, this routine doesn't make any effort to preserve
3198cc57
KW
3122 # the range values and types of one input over the other. Therefore
3123 # this base class should not allow _union to be called from other than
99870f4d
KW
3124 # initialization code, so as to prevent two tables from being added
3125 # together where the range values matter. The general form of this
3126 # routine therefore belongs in a derived class, but it was moved here
3127 # to avoid duplication of code. The failure to overload this in this
3128 # class keeps it safe.
3198cc57
KW
3129 #
3130 # It does make the effort during initialization to accept tables with
3131 # multiple values for the same code point, and to preserve the order
3132 # of these. If there is only one input range or range set, it doesn't
3133 # sort (as it should already be sorted to the desired order), and will
3134 # accept multiple values per code point. Otherwise it will merge
3135 # multiple values into a single one.
99870f4d
KW
3136
3137 my $self;
3138 my @args; # Arguments to pass to the constructor
3139
3140 my $class = shift;
3141
3142 # If a method call, will start the union with the object itself, and
3143 # the class of the new object will be the same as self.
3144 if (ref $class) {
3145 $self = $class;
3146 $class = ref $self;
3147 push @args, $self;
3148 }
3149
3150 # Add the other required parameter.
3151 push @args, shift;
3152 # Rest of parameters are passed on to the constructor
3153
3154 # Accumulate all records from both lists.
3155 my @records;
3198cc57 3156 my $input_count = 0;
99870f4d
KW
3157 for my $arg (@args) {
3158 #local $to_trace = 0 if main::DEBUG;
3159 trace "argument = $arg" if main::DEBUG && $to_trace;
3160 if (! defined $arg) {
3161 my $message = "";
3162 if (defined $self) {
f998e60c 3163 no overloading;
051df77b 3164 $message .= $owner_name_of{pack 'J', $self};
99870f4d 3165 }
ada6088e 3166 Carp::my_carp_bug($message . "Undefined argument to _union. No union done.");
99870f4d
KW
3167 return;
3168 }
3198cc57 3169
99870f4d
KW
3170 $arg = [ $arg ] if ! ref $arg;
3171 my $type = ref $arg;
3172 if ($type eq 'ARRAY') {
3173 foreach my $element (@$arg) {
3174 push @records, Range->new($element, $element);
3198cc57 3175 $input_count++;
99870f4d
KW
3176 }
3177 }
3178 elsif ($arg->isa('Range')) {
3179 push @records, $arg;
3198cc57 3180 $input_count++;
99870f4d
KW
3181 }
3182 elsif ($arg->can('ranges')) {
3183 push @records, $arg->ranges;
3198cc57 3184 $input_count++;
99870f4d
KW
3185 }
3186 else {
3187 my $message = "";
3188 if (defined $self) {
f998e60c 3189 no overloading;
051df77b 3190 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
3191 }
3192 Carp::my_carp_bug($message . "Cannot take the union of a $type. No union done.");
3193 return;
3194 }
3195 }
3196
3197 # Sort with the range containing the lowest ordinal first, but if
3198 # two ranges start at the same code point, sort with the bigger range
3199 # of the two first, because it takes fewer cycles.
3198cc57
KW
3200 if ($input_count > 1) {
3201 @records = sort { ($a->start <=> $b->start)
99870f4d
KW
3202 or
3203 # if b is shorter than a, b->end will be
3204 # less than a->end, and we want to select
3205 # a, so want to return -1
3206 ($b->end <=> $a->end)
3207 } @records;
3198cc57 3208 }
99870f4d
KW
3209
3210 my $new = $class->new(@_);
3211
3212 # Fold in records so long as they add new information.
3213 for my $set (@records) {
3214 my $start = $set->start;
3215 my $end = $set->end;
d59563d0 3216 my $value = $set->value;
3198cc57 3217 my $type = $set->type;
99870f4d 3218 if ($start > $new->max) {
3198cc57 3219 $new->_add_delete('+', $start, $end, $value, Type => $type);
99870f4d
KW
3220 }
3221 elsif ($end > $new->max) {
3198cc57
KW
3222 $new->_add_delete('+', $new->max +1, $end, $value,
3223 Type => $type);
3224 }
3225 elsif ($input_count == 1) {
3226 # Here, overlaps existing range, but is from a single input,
3227 # so preserve the multiple values from that input.
3228 $new->_add_delete('+', $start, $end, $value, Type => $type,
3229 Replace => $MULTIPLE_AFTER);
99870f4d
KW
3230 }
3231 }
3232
3233 return $new;
3234 }
3235
3236 sub range_count { # Return the number of ranges in the range list
3237 my $self = shift;
3238 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3239
f998e60c 3240 no overloading;
051df77b 3241 return scalar @{$ranges{pack 'J', $self}};
99870f4d
KW
3242 }
3243
3244 sub min {
3245 # Returns the minimum code point currently in the range list, or if
3246 # the range list is empty, 2 beyond the max possible. This is a
3247 # method because used so rarely, that not worth saving between calls,
3248 # and having to worry about changing it as ranges are added and
3249 # deleted.
3250
3251 my $self = shift;
3252 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3253
ffe43484 3254 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3255
3256 # If the range list is empty, return a large value that isn't adjacent
3257 # to any that could be in the range list, for simpler tests
6189eadc 3258 return $MAX_UNICODE_CODEPOINT + 2 unless scalar @{$ranges{$addr}};
99870f4d
KW
3259 return $ranges{$addr}->[0]->start;
3260 }
3261
3262 sub contains {
3263 # Boolean: Is argument in the range list? If so returns $i such that:
3264 # range[$i]->end < $codepoint <= range[$i+1]->end
3265 # which is one beyond what you want; this is so that the 0th range
3266 # doesn't return false
3267 my $self = shift;
3268 my $codepoint = shift;
3269 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3270
99870f4d
KW
3271 my $i = $self->_search_ranges($codepoint);
3272 return 0 unless defined $i;
3273
3274 # The search returns $i, such that
3275 # range[$i-1]->end < $codepoint <= range[$i]->end
3276 # So is in the table if and only iff it is at least the start position
3277 # of range $i.
f998e60c 3278 no overloading;
051df77b 3279 return 0 if $ranges{pack 'J', $self}->[$i]->start > $codepoint;
99870f4d
KW
3280 return $i + 1;
3281 }
3282
2f7a8815
KW
3283 sub containing_range {
3284 # Returns the range object that contains the code point, undef if none
3285
3286 my $self = shift;
3287 my $codepoint = shift;
3288 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3289
3290 my $i = $self->contains($codepoint);
3291 return unless $i;
3292
3293 # contains() returns 1 beyond where we should look
3294 no overloading;
3295 return $ranges{pack 'J', $self}->[$i-1];
3296 }
3297
99870f4d
KW
3298 sub value_of {
3299 # Returns the value associated with the code point, undef if none
3300
3301 my $self = shift;
3302 my $codepoint = shift;
3303 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3304
d69c231b
KW
3305 my $range = $self->containing_range($codepoint);
3306 return unless defined $range;
99870f4d 3307
d69c231b 3308 return $range->value;
99870f4d
KW
3309 }
3310
0a9dbafc
KW
3311 sub type_of {
3312 # Returns the type of the range containing the code point, undef if
3313 # the code point is not in the table
3314
3315 my $self = shift;
3316 my $codepoint = shift;
3317 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3318
3319 my $range = $self->containing_range($codepoint);
3320 return unless defined $range;
3321
3322 return $range->type;
3323 }
3324
99870f4d
KW
3325 sub _search_ranges {
3326 # Find the range in the list which contains a code point, or where it
3327 # should go if were to add it. That is, it returns $i, such that:
3328 # range[$i-1]->end < $codepoint <= range[$i]->end
3329 # Returns undef if no such $i is possible (e.g. at end of table), or
3330 # if there is an error.
3331
3332 my $self = shift;
3333 my $code_point = shift;
3334 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3335
ffe43484 3336 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3337
3338 return if $code_point > $max{$addr};
3339 my $r = $ranges{$addr}; # The current list of ranges
3340 my $range_list_size = scalar @$r;
3341 my $i;
3342
3343 use integer; # want integer division
3344
3345 # Use the cached result as the starting guess for this one, because,
3346 # an experiment on 5.1 showed that 90% of the time the cache was the
3347 # same as the result on the next call (and 7% it was one less).
3348 $i = $_search_ranges_cache{$addr};
3349 $i = 0 if $i >= $range_list_size; # Reset if no longer valid (prob.
3350 # from an intervening deletion
3351 #local $to_trace = 1 if main::DEBUG;
3352 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);
3353 return $i if $code_point <= $r->[$i]->end
3354 && ($i == 0 || $r->[$i-1]->end < $code_point);
3355
3356 # Here the cache doesn't yield the correct $i. Try adding 1.
3357 if ($i < $range_list_size - 1
3358 && $r->[$i]->end < $code_point &&
3359 $code_point <= $r->[$i+1]->end)
3360 {
3361 $i++;
3362 trace "next \$i is correct: $i" if main::DEBUG && $to_trace;
3363 $_search_ranges_cache{$addr} = $i;
3364 return $i;
3365 }
3366
3367 # Here, adding 1 also didn't work. We do a binary search to
3368 # find the correct position, starting with current $i
3369 my $lower = 0;
3370 my $upper = $range_list_size - 1;
3371 while (1) {
3372 trace "top of loop i=$i:", sprintf("%04X", $r->[$lower]->start), "[$lower] .. ", sprintf("%04X", $r->[$i]->start), "[$i] .. ", sprintf("%04X", $r->[$upper]->start), "[$upper]" if main::DEBUG && $to_trace;
3373
3374 if ($code_point <= $r->[$i]->end) {
3375
3376 # Here we have met the upper constraint. We can quit if we
3377 # also meet the lower one.
3378 last if $i == 0 || $r->[$i-1]->end < $code_point;
3379
3380 $upper = $i; # Still too high.
3381
3382 }
3383 else {
3384
3385 # Here, $r[$i]->end < $code_point, so look higher up.
3386 $lower = $i;
3387 }
3388
3389 # Split search domain in half to try again.
3390 my $temp = ($upper + $lower) / 2;
3391
3392 # No point in continuing unless $i changes for next time
3393 # in the loop.
3394 if ($temp == $i) {
3395
3396 # We can't reach the highest element because of the averaging.
3397 # So if one below the upper edge, force it there and try one
3398 # more time.
3399 if ($i == $range_list_size - 2) {
3400
3401 trace "Forcing to upper edge" if main::DEBUG && $to_trace;
3402 $i = $range_list_size - 1;
3403
3404 # Change $lower as well so if fails next time through,
3405 # taking the average will yield the same $i, and we will
3406 # quit with the error message just below.
3407 $lower = $i;
3408 next;
3409 }
3410 Carp::my_carp_bug("$owner_name_of{$addr}Can't find where the range ought to go. No action taken.");
3411 return;
3412 }
3413 $i = $temp;
3414 } # End of while loop
3415
3416 if (main::DEBUG && $to_trace) {
3417 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i;
3418 trace "i= [ $i ]", $r->[$i];
3419 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < $range_list_size - 1;
3420 }
3421
3422 # Here we have found the offset. Cache it as a starting point for the
3423 # next call.
3424 $_search_ranges_cache{$addr} = $i;
3425 return $i;
3426 }
3427
3428 sub _add_delete {
3429 # Add, replace or delete ranges to or from a list. The $type
3430 # parameter gives which:
3431 # '+' => insert or replace a range, returning a list of any changed
3432 # ranges.
3433 # '-' => delete a range, returning a list of any deleted ranges.
3434 #
3435 # The next three parameters give respectively the start, end, and
3436 # value associated with the range. 'value' should be null unless the
3437 # operation is '+';
3438 #
3439 # The range list is kept sorted so that the range with the lowest
3440 # starting position is first in the list, and generally, adjacent
c1739a4a 3441 # ranges with the same values are merged into a single larger one (see
99870f4d
KW
3442 # exceptions below).
3443 #
c1739a4a 3444 # There are more parameters; all are key => value pairs:
99870f4d
KW
3445 # Type gives the type of the value. It is only valid for '+'.
3446 # All ranges have types; if this parameter is omitted, 0 is
3447 # assumed. Ranges with type 0 are assumed to obey the
3448 # Unicode rules for casing, etc; ranges with other types are
3449 # not. Otherwise, the type is arbitrary, for the caller's
3450 # convenience, and looked at only by this routine to keep
3451 # adjacent ranges of different types from being merged into
3452 # a single larger range, and when Replace =>
3453 # $IF_NOT_EQUIVALENT is specified (see just below).
3454 # Replace determines what to do if the range list already contains
3455 # ranges which coincide with all or portions of the input
3456 # range. It is only valid for '+':
3457 # => $NO means that the new value is not to replace
3458 # any existing ones, but any empty gaps of the
3459 # range list coinciding with the input range
3460 # will be filled in with the new value.
3461 # => $UNCONDITIONALLY means to replace the existing values with
3462 # this one unconditionally. However, if the
3463 # new and old values are identical, the
3464 # replacement is skipped to save cycles
3465 # => $IF_NOT_EQUIVALENT means to replace the existing values
d59563d0 3466 # (the default) with this one if they are not equivalent.
99870f4d 3467 # Ranges are equivalent if their types are the
c1739a4a 3468 # same, and they are the same string; or if
99870f4d
KW
3469 # both are type 0 ranges, if their Unicode
3470 # standard forms are identical. In this last
3471 # case, the routine chooses the more "modern"
3472 # one to use. This is because some of the
3473 # older files are formatted with values that
3474 # are, for example, ALL CAPs, whereas the
3475 # derived files have a more modern style,
3476 # which looks better. By looking for this
3477 # style when the pre-existing and replacement
3478 # standard forms are the same, we can move to
3479 # the modern style
9470941f 3480 # => $MULTIPLE_BEFORE means that if this range duplicates an
99870f4d
KW
3481 # existing one, but has a different value,
3482 # don't replace the existing one, but insert
3483 # this, one so that the same range can occur
53d84487
KW
3484 # multiple times. They are stored LIFO, so
3485 # that the final one inserted is the first one
3486 # returned in an ordered search of the table.
6901521e
KW
3487 # If this is an exact duplicate, including the
3488 # value, the original will be moved to be
3489 # first, before any other duplicate ranges
3490 # with different values.
7f4b1e25
KW
3491 # => $MULTIPLE_AFTER is like $MULTIPLE_BEFORE, but is stored
3492 # FIFO, so that this one is inserted after all
6901521e
KW
3493 # others that currently exist. If this is an
3494 # exact duplicate, including value, of an
3495 # existing range, this one is discarded
3496 # (leaving the existing one in its original,
3497 # higher priority position
99870f4d
KW
3498 # => anything else is the same as => $IF_NOT_EQUIVALENT
3499 #
c1739a4a
KW
3500 # "same value" means identical for non-type-0 ranges, and it means
3501 # having the same standard forms for type-0 ranges.
99870f4d
KW
3502
3503 return Carp::carp_too_few_args(\@_, 5) if main::DEBUG && @_ < 5;
3504
3505 my $self = shift;
3506 my $operation = shift; # '+' for add/replace; '-' for delete;
3507 my $start = shift;
3508 my $end = shift;
3509 my $value = shift;
3510
3511 my %args = @_;
3512
3513 $value = "" if not defined $value; # warning: $value can be "0"
3514
3515 my $replace = delete $args{'Replace'};
3516 $replace = $IF_NOT_EQUIVALENT unless defined $replace;
3517
3518 my $type = delete $args{'Type'};
3519 $type = 0 unless defined $type;
3520
3521 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3522
ffe43484 3523 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3524
3525 if ($operation ne '+' && $operation ne '-') {
3526 Carp::my_carp_bug("$owner_name_of{$addr}First parameter to _add_delete must be '+' or '-'. No action taken.");
3527 return;
3528 }
3529 unless (defined $start && defined $end) {
3530 Carp::my_carp_bug("$owner_name_of{$addr}Undefined start and/or end to _add_delete. No action taken.");
3531 return;
3532 }
3533 unless ($end >= $start) {
3534 Carp::my_carp_bug("$owner_name_of{$addr}End of range (" . sprintf("%04X", $end) . ") must not be before start (" . sprintf("%04X", $start) . "). No action taken.");
3535 return;
3536 }
556ca434
KW
3537 if ($end > $MAX_UNICODE_CODEPOINT && $operation eq '+') {
3538 Carp::my_carp("$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");
3539 }
99870f4d
KW
3540 #local $to_trace = 1 if main::DEBUG;
3541
3542 if ($operation eq '-') {
3543 if ($replace != $IF_NOT_EQUIVALENT) {
3544 Carp::my_carp_bug("$owner_name_of{$addr}Replace => \$IF_NOT_EQUIVALENT is required when deleting a range from a range list. Assuming Replace => \$IF_NOT_EQUIVALENT.");
3545 $replace = $IF_NOT_EQUIVALENT;
3546 }
3547 if ($type) {
3548 Carp::my_carp_bug("$owner_name_of{$addr}Type => 0 is required when deleting a range from a range list. Assuming Type => 0.");
3549 $type = 0;
3550 }
3551 if ($value ne "") {
3552 Carp::my_carp_bug("$owner_name_of{$addr}Value => \"\" is required when deleting a range from a range list. Assuming Value => \"\".");
3553 $value = "";
3554 }
3555 }
3556
3557 my $r = $ranges{$addr}; # The current list of ranges
3558 my $range_list_size = scalar @$r; # And its size
3559 my $max = $max{$addr}; # The current high code point in
3560 # the list of ranges
3561
3562 # Do a special case requiring fewer machine cycles when the new range
3563 # starts after the current highest point. The Unicode input data is
3564 # structured so this is common.
3565 if ($start > $max) {
3566
52d4d76a 3567 trace "$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
3568 return if $operation eq '-'; # Deleting a non-existing range is a
3569 # no-op
3570
3571 # If the new range doesn't logically extend the current final one
3572 # in the range list, create a new range at the end of the range
3573 # list. (max cleverly is initialized to a negative number not
3574 # adjacent to 0 if the range list is empty, so even adding a range
3575 # to an empty range list starting at 0 will have this 'if'
3576 # succeed.)
3577 if ($start > $max + 1 # non-adjacent means can't extend.
3578 || @{$r}[-1]->value ne $value # values differ, can't extend.
3579 || @{$r}[-1]->type != $type # types differ, can't extend.
3580 ) {
3581 push @$r, Range->new($start, $end,
3582 Value => $value,
3583 Type => $type);
3584 }
3585 else {
3586
3587 # Here, the new range starts just after the current highest in
3588 # the range list, and they have the same type and value.
3589 # Extend the current range to incorporate the new one.
3590 @{$r}[-1]->set_end($end);
3591 }
3592
3593 # This becomes the new maximum.
3594 $max{$addr} = $end;
3595
3596 return;
3597 }
3598 #local $to_trace = 0 if main::DEBUG;
3599
3600 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) replace=$replace" if main::DEBUG && $to_trace;
3601
3602 # Here, the input range isn't after the whole rest of the range list.
3603 # Most likely 'splice' will be needed. The rest of the routine finds
3604 # the needed splice parameters, and if necessary, does the splice.
3605 # First, find the offset parameter needed by the splice function for
3606 # the input range. Note that the input range may span multiple
3607 # existing ones, but we'll worry about that later. For now, just find
3608 # the beginning. If the input range is to be inserted starting in a
3609 # position not currently in the range list, it must (obviously) come
3610 # just after the range below it, and just before the range above it.
3611 # Slightly less obviously, it will occupy the position currently
3612 # occupied by the range that is to come after it. More formally, we
3613 # are looking for the position, $i, in the array of ranges, such that:
3614 #
3615 # r[$i-1]->start <= r[$i-1]->end < $start < r[$i]->start <= r[$i]->end
3616 #
3617 # (The ordered relationships within existing ranges are also shown in
3618 # the equation above). However, if the start of the input range is
3619 # within an existing range, the splice offset should point to that
3620 # existing range's position in the list; that is $i satisfies a
3621 # somewhat different equation, namely:
3622 #
3623 #r[$i-1]->start <= r[$i-1]->end < r[$i]->start <= $start <= r[$i]->end
3624 #
3625 # More briefly, $start can come before or after r[$i]->start, and at
3626 # this point, we don't know which it will be. However, these
3627 # two equations share these constraints:
3628 #
3629 # r[$i-1]->end < $start <= r[$i]->end
3630 #
3631 # And that is good enough to find $i.
3632
3633 my $i = $self->_search_ranges($start);
3634 if (! defined $i) {
3635 Carp::my_carp_bug("Searching $self for range beginning with $start unexpectedly returned undefined. Operation '$operation' not performed");
3636 return;
3637 }
3638
3639 # The search function returns $i such that:
3640 #
3641 # r[$i-1]->end < $start <= r[$i]->end
3642 #
3643 # That means that $i points to the first range in the range list
3644 # that could possibly be affected by this operation. We still don't
3645 # know if the start of the input range is within r[$i], or if it
3646 # points to empty space between r[$i-1] and r[$i].
3647 trace "[$i] is the beginning splice point. Existing range there is ", $r->[$i] if main::DEBUG && $to_trace;
3648
3649 # Special case the insertion of data that is not to replace any
3650 # existing data.
3651 if ($replace == $NO) { # If $NO, has to be operation '+'
3652 #local $to_trace = 1 if main::DEBUG;
3653 trace "Doesn't replace" if main::DEBUG && $to_trace;
3654
3655 # Here, the new range is to take effect only on those code points
3656 # that aren't already in an existing range. This can be done by
3657 # looking through the existing range list and finding the gaps in
3658 # the ranges that this new range affects, and then calling this
3659 # function recursively on each of those gaps, leaving untouched
3660 # anything already in the list. Gather up a list of the changed
3661 # gaps first so that changes to the internal state as new ranges
3662 # are added won't be a problem.
3663 my @gap_list;
3664
3665 # First, if the starting point of the input range is outside an
3666 # existing one, there is a gap from there to the beginning of the
3667 # existing range -- add a span to fill the part that this new
3668 # range occupies
3669 if ($start < $r->[$i]->start) {
3670 push @gap_list, Range->new($start,
3671 main::min($end,
3672 $r->[$i]->start - 1),
3673 Type => $type);
3674 trace "gap before $r->[$i] [$i], will add", $gap_list[-1] if main::DEBUG && $to_trace;
3675 }
3676
3677 # Then look through the range list for other gaps until we reach
3678 # the highest range affected by the input one.
3679 my $j;
3680 for ($j = $i+1; $j < $range_list_size; $j++) {
3681 trace "j=[$j]", $r->[$j] if main::DEBUG && $to_trace;
3682 last if $end < $r->[$j]->start;
3683
3684 # If there is a gap between when this range starts and the
3685 # previous one ends, add a span to fill it. Note that just
3686 # because there are two ranges doesn't mean there is a
3687 # non-zero gap between them. It could be that they have
3688 # different values or types
3689 if ($r->[$j-1]->end + 1 != $r->[$j]->start) {
3690 push @gap_list,
3691 Range->new($r->[$j-1]->end + 1,
3692 $r->[$j]->start - 1,
3693 Type => $type);
3694 trace "gap between $r->[$j-1] and $r->[$j] [$j], will add: $gap_list[-1]" if main::DEBUG && $to_trace;
3695 }
3696 }
3697
3698 # Here, we have either found an existing range in the range list,
3699 # beyond the area affected by the input one, or we fell off the
3700 # end of the loop because the input range affects the whole rest
3701 # of the range list. In either case, $j is 1 higher than the
3702 # highest affected range. If $j == $i, it means that there are no
3703 # affected ranges, that the entire insertion is in the gap between
3704 # r[$i-1], and r[$i], which we already have taken care of before
3705 # the loop.
3706 # On the other hand, if there are affected ranges, it might be
3707 # that there is a gap that needs filling after the final such
3708 # range to the end of the input range
3709 if ($r->[$j-1]->end < $end) {
3710 push @gap_list, Range->new(main::max($start,
3711 $r->[$j-1]->end + 1),
3712 $end,
3713 Type => $type);
3714 trace "gap after $r->[$j-1], will add $gap_list[-1]" if main::DEBUG && $to_trace;
3715 }
3716
3717 # Call recursively to fill in all the gaps.
3718 foreach my $gap (@gap_list) {
3719 $self->_add_delete($operation,
3720 $gap->start,
3721 $gap->end,
3722 $value,
3723 Type => $type);
3724 }
3725
3726 return;
3727 }
3728
53d84487
KW
3729 # Here, we have taken care of the case where $replace is $NO.
3730 # Remember that here, r[$i-1]->end < $start <= r[$i]->end
3731 # If inserting a multiple record, this is where it goes, before the
7f4b1e25
KW
3732 # first (if any) existing one if inserting LIFO. (If this is to go
3733 # afterwards, FIFO, we below move the pointer to there.) These imply
3734 # an insertion, and no change to any existing ranges. Note that $i
3735 # can be -1 if this new range doesn't actually duplicate any existing,
3736 # and comes at the beginning of the list.
3737 if ($replace == $MULTIPLE_BEFORE || $replace == $MULTIPLE_AFTER) {
53d84487
KW
3738
3739 if ($start != $end) {
3740 Carp::my_carp_bug("$owner_name_of{$addr}Can't cope with adding a multiple record when the range ($start..$end) contains more than one code point. No action taken.");
3741 return;
3742 }
3743
19155fcc 3744 # If the new code point is within a current range ...
53d84487 3745 if ($end >= $r->[$i]->start) {
19155fcc
KW
3746
3747 # Don't add an exact duplicate, as it isn't really a multiple
1f6798c4
KW
3748 my $existing_value = $r->[$i]->value;
3749 my $existing_type = $r->[$i]->type;
3750 return if $value eq $existing_value && $type eq $existing_type;
3751
3752 # If the multiple value is part of an existing range, we want
3753 # to split up that range, so that only the single code point
3754 # is affected. To do this, we first call ourselves
3755 # recursively to delete that code point from the table, having
3756 # preserved its current data above. Then we call ourselves
3757 # recursively again to add the new multiple, which we know by
3758 # the test just above is different than the current code
3759 # point's value, so it will become a range containing a single
3760 # code point: just itself. Finally, we add back in the
3761 # pre-existing code point, which will again be a single code
3762 # point range. Because 'i' likely will have changed as a
3763 # result of these operations, we can't just continue on, but
7f4b1e25
KW
3764 # do this operation recursively as well. If we are inserting
3765 # LIFO, the pre-existing code point needs to go after the new
3766 # one, so use MULTIPLE_AFTER; and vice versa.
53d84487 3767 if ($r->[$i]->start != $r->[$i]->end) {
1f6798c4
KW
3768 $self->_add_delete('-', $start, $end, "");
3769 $self->_add_delete('+', $start, $end, $value, Type => $type);
7f4b1e25
KW
3770 return $self->_add_delete('+',
3771 $start, $end,
3772 $existing_value,
3773 Type => $existing_type,
3774 Replace => ($replace == $MULTIPLE_BEFORE)
3775 ? $MULTIPLE_AFTER
3776 : $MULTIPLE_BEFORE);