This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
mktables: Add %prop_value_aliases to UCD.pl
[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
KW
33use Text::Tabs;
34
35sub DEBUG () { 0 } # Set to 0 for production; 1 for development
bd9ebcfd 36my $debugging_build = $Config{"ccflags"} =~ /-DDEBUGGING/;
99870f4d
KW
37
38##########################################################################
39#
40# mktables -- create the runtime Perl Unicode files (lib/unicore/.../*.pl),
41# from the Unicode database files (lib/unicore/.../*.txt), It also generates
42# a pod file and a .t file
43#
44# The structure of this file is:
45# First these introductory comments; then
46# code needed for everywhere, such as debugging stuff; then
47# code to handle input parameters; then
48# data structures likely to be of external interest (some of which depend on
49# the input parameters, so follows them; then
50# more data structures and subroutine and package (class) definitions; then
51# the small actual loop to process the input files and finish up; then
52# a __DATA__ section, for the .t tests
53#
5f7264c7 54# This program works on all releases of Unicode through at least 6.0. The
99870f4d
KW
55# outputs have been scrutinized most intently for release 5.1. The others
56# have been checked for somewhat more than just sanity. It can handle all
57# existing Unicode character properties in those releases.
58#
99870f4d
KW
59# This program is mostly about Unicode character (or code point) properties.
60# A property describes some attribute or quality of a code point, like if it
61# is lowercase or not, its name, what version of Unicode it was first defined
62# in, or what its uppercase equivalent is. Unicode deals with these disparate
63# possibilities by making all properties into mappings from each code point
64# into some corresponding value. In the case of it being lowercase or not,
65# the mapping is either to 'Y' or 'N' (or various synonyms thereof). Each
66# property maps each Unicode code point to a single value, called a "property
67# value". (Hence each Unicode property is a true mathematical function with
68# exactly one value per code point.)
69#
70# When using a property in a regular expression, what is desired isn't the
71# mapping of the code point to its property's value, but the reverse (or the
72# mathematical "inverse relation"): starting with the property value, "Does a
73# code point map to it?" These are written in a "compound" form:
74# \p{property=value}, e.g., \p{category=punctuation}. This program generates
75# files containing the lists of code points that map to each such regular
76# expression property value, one file per list
77#
78# There is also a single form shortcut that Perl adds for many of the commonly
79# used properties. This happens for all binary properties, plus script,
80# general_category, and block properties.
81#
82# Thus the outputs of this program are files. There are map files, mostly in
83# the 'To' directory; and there are list files for use in regular expression
84# matching, all in subdirectories of the 'lib' directory, with each
85# subdirectory being named for the property that the lists in it are for.
86# Bookkeeping, test, and documentation files are also generated.
87
88my $matches_directory = 'lib'; # Where match (\p{}) files go.
89my $map_directory = 'To'; # Where map files go.
90
91# DATA STRUCTURES
92#
93# The major data structures of this program are Property, of course, but also
94# Table. There are two kinds of tables, very similar to each other.
95# "Match_Table" is the data structure giving the list of code points that have
96# a particular property value, mentioned above. There is also a "Map_Table"
97# data structure which gives the property's mapping from code point to value.
98# There are two structures because the match tables need to be combined in
99# various ways, such as constructing unions, intersections, complements, etc.,
100# and the map ones don't. And there would be problems, perhaps subtle, if
101# a map table were inadvertently operated on in some of those ways.
102# The use of separate classes with operations defined on one but not the other
103# prevents accidentally confusing the two.
104#
105# At the heart of each table's data structure is a "Range_List", which is just
106# an ordered list of "Ranges", plus ancillary information, and methods to
107# operate on them. A Range is a compact way to store property information.
108# Each range has a starting code point, an ending code point, and a value that
109# is meant to apply to all the code points between the two end points,
110# inclusive. For a map table, this value is the property value for those
111# code points. Two such ranges could be written like this:
112# 0x41 .. 0x5A, 'Upper',
113# 0x61 .. 0x7A, 'Lower'
114#
115# Each range also has a type used as a convenience to classify the values.
116# Most ranges in this program will be Type 0, or normal, but there are some
117# ranges that have a non-zero type. These are used only in map tables, and
118# are for mappings that don't fit into the normal scheme of things. Mappings
119# that require a hash entry to communicate with utf8.c are one example;
120# another example is mappings for charnames.pm to use which indicate a name
121# that is algorithmically determinable from its code point (and vice-versa).
122# These are used to significantly compact these tables, instead of listing
123# each one of the tens of thousands individually.
124#
125# In a match table, the value of a range is irrelevant (and hence the type as
126# well, which will always be 0), and arbitrarily set to the null string.
127# Using the example above, there would be two match tables for those two
128# entries, one named Upper would contain the 0x41..0x5A range, and the other
129# named Lower would contain 0x61..0x7A.
130#
131# Actually, there are two types of range lists, "Range_Map" is the one
132# associated with map tables, and "Range_List" with match tables.
133# Again, this is so that methods can be defined on one and not the other so as
134# to prevent operating on them in incorrect ways.
135#
136# Eventually, most tables are written out to files to be read by utf8_heavy.pl
137# in the perl core. All tables could in theory be written, but some are
138# suppressed because there is no current practical use for them. It is easy
139# to change which get written by changing various lists that are near the top
140# of the actual code in this file. The table data structures contain enough
141# ancillary information to allow them to be treated as separate entities for
142# writing, such as the path to each one's file. There is a heading in each
143# map table that gives the format of its entries, and what the map is for all
144# the code points missing from it. (This allows tables to be more compact.)
678f13d5 145#
99870f4d
KW
146# The Property data structure contains one or more tables. All properties
147# contain a map table (except the $perl property which is a
148# pseudo-property containing only match tables), and any properties that
149# are usable in regular expression matches also contain various matching
150# tables, one for each value the property can have. A binary property can
151# have two values, True and False (or Y and N, which are preferred by Unicode
152# terminology). Thus each of these properties will have a map table that
153# takes every code point and maps it to Y or N (but having ranges cuts the
154# number of entries in that table way down), and two match tables, one
155# which has a list of all the code points that map to Y, and one for all the
156# code points that map to N. (For each of these, a third table is also
157# generated for the pseudo Perl property. It contains the identical code
158# points as the Y table, but can be written, not in the compound form, but in
159# a "single" form like \p{IsUppercase}.) Many properties are binary, but some
160# properties have several possible values, some have many, and properties like
161# Name have a different value for every named code point. Those will not,
162# unless the controlling lists are changed, have their match tables written
163# out. But all the ones which can be used in regular expression \p{} and \P{}
c12f2655
KW
164# constructs will. Prior to 5.14, generally a property would have either its
165# map table or its match tables written but not both. Again, what gets
166# written is controlled by lists which can easily be changed. Starting in
167# 5.14, advantage was taken of this, and all the map tables needed to
168# reconstruct the Unicode db are now written out, while suppressing the
169# Unicode .txt files that contain the data. Our tables are much more compact
170# than the .txt files, so a significant space savings was achieved.
171
172# Properties have a 'Type', like binary, or string, or enum depending on how
173# many match tables there are and the content of the maps. This 'Type' is
174# different than a range 'Type', so don't get confused by the two concepts
175# having the same name.
678f13d5 176#
99870f4d
KW
177# For information about the Unicode properties, see Unicode's UAX44 document:
178
179my $unicode_reference_url = 'http://www.unicode.org/reports/tr44/';
180
181# As stated earlier, this program will work on any release of Unicode so far.
182# Most obvious problems in earlier data have NOT been corrected except when
183# necessary to make Perl or this program work reasonably. For example, no
c12f2655
KW
184# folding information was given in early releases, so this program substitutes
185# lower case instead, just so that a regular expression with the /i option
186# will do something that actually gives the right results in many cases.
187# There are also a couple other corrections for version 1.1.5, commented at
188# the point they are made. As an example of corrections that weren't made
189# (but could be) is this statement from DerivedAge.txt: "The supplementary
190# private use code points and the non-character code points were assigned in
191# version 2.0, but not specifically listed in the UCD until versions 3.0 and
192# 3.1 respectively." (To be precise it was 3.0.1 not 3.0.0) More information
193# on Unicode version glitches is further down in these introductory comments.
99870f4d 194#
5f7264c7
KW
195# This program works on all non-provisional properties as of 6.0, though the
196# files for some are suppressed from apparent lack of demand for them. You
197# can change which are output by changing lists in this program.
678f13d5 198#
dc85bd38 199# The old version of mktables emphasized the term "Fuzzy" to mean Unicode's
99870f4d
KW
200# loose matchings rules (from Unicode TR18):
201#
202# The recommended names for UCD properties and property values are in
203# PropertyAliases.txt [Prop] and PropertyValueAliases.txt
204# [PropValue]. There are both abbreviated names and longer, more
205# descriptive names. It is strongly recommended that both names be
206# recognized, and that loose matching of property names be used,
207# whereby the case distinctions, whitespace, hyphens, and underbar
208# are ignored.
209# The program still allows Fuzzy to override its determination of if loose
210# matching should be used, but it isn't currently used, as it is no longer
211# needed; the calculations it makes are good enough.
678f13d5 212#
99870f4d
KW
213# SUMMARY OF HOW IT WORKS:
214#
215# Process arguments
216#
217# A list is constructed containing each input file that is to be processed
218#
219# Each file on the list is processed in a loop, using the associated handler
220# code for each:
221# The PropertyAliases.txt and PropValueAliases.txt files are processed
222# first. These files name the properties and property values.
223# Objects are created of all the property and property value names
224# that the rest of the input should expect, including all synonyms.
225# The other input files give mappings from properties to property
226# values. That is, they list code points and say what the mapping
227# is under the given property. Some files give the mappings for
228# just one property; and some for many. This program goes through
229# each file and populates the properties from them. Some properties
230# are listed in more than one file, and Unicode has set up a
231# precedence as to which has priority if there is a conflict. Thus
232# the order of processing matters, and this program handles the
233# conflict possibility by processing the overriding input files
234# last, so that if necessary they replace earlier values.
235# After this is all done, the program creates the property mappings not
236# furnished by Unicode, but derivable from what it does give.
237# The tables of code points that match each property value in each
238# property that is accessible by regular expressions are created.
239# The Perl-defined properties are created and populated. Many of these
240# require data determined from the earlier steps
241# Any Perl-defined synonyms are created, and name clashes between Perl
678f13d5 242# and Unicode are reconciled and warned about.
99870f4d
KW
243# All the properties are written to files
244# Any other files are written, and final warnings issued.
678f13d5 245#
99870f4d
KW
246# For clarity, a number of operators have been overloaded to work on tables:
247# ~ means invert (take all characters not in the set). The more
248# conventional '!' is not used because of the possibility of confusing
249# it with the actual boolean operation.
250# + means union
251# - means subtraction
252# & means intersection
253# The precedence of these is the order listed. Parentheses should be
254# copiously used. These are not a general scheme. The operations aren't
255# defined for a number of things, deliberately, to avoid getting into trouble.
256# Operations are done on references and affect the underlying structures, so
257# that the copy constructors for them have been overloaded to not return a new
258# clone, but the input object itself.
678f13d5 259#
99870f4d
KW
260# The bool operator is deliberately not overloaded to avoid confusion with
261# "should it mean if the object merely exists, or also is non-empty?".
99870f4d
KW
262#
263# WHY CERTAIN DESIGN DECISIONS WERE MADE
678f13d5
KW
264#
265# This program needs to be able to run under miniperl. Therefore, it uses a
266# minimum of other modules, and hence implements some things itself that could
267# be gotten from CPAN
268#
269# This program uses inputs published by the Unicode Consortium. These can
270# change incompatibly between releases without the Perl maintainers realizing
271# it. Therefore this program is now designed to try to flag these. It looks
272# at the directories where the inputs are, and flags any unrecognized files.
273# It keeps track of all the properties in the files it handles, and flags any
274# that it doesn't know how to handle. It also flags any input lines that
275# don't match the expected syntax, among other checks.
276#
277# It is also designed so if a new input file matches one of the known
278# templates, one hopefully just needs to add it to a list to have it
279# processed.
280#
281# As mentioned earlier, some properties are given in more than one file. In
282# particular, the files in the extracted directory are supposedly just
283# reformattings of the others. But they contain information not easily
284# derivable from the other files, including results for Unihan, which this
285# program doesn't ordinarily look at, and for unassigned code points. They
286# also have historically had errors or been incomplete. In an attempt to
287# create the best possible data, this program thus processes them first to
288# glean information missing from the other files; then processes those other
289# files to override any errors in the extracted ones. Much of the design was
290# driven by this need to store things and then possibly override them.
291#
292# It tries to keep fatal errors to a minimum, to generate something usable for
293# testing purposes. It always looks for files that could be inputs, and will
294# warn about any that it doesn't know how to handle (the -q option suppresses
295# the warning).
99870f4d 296#
678f13d5
KW
297# Why is there more than one type of range?
298# This simplified things. There are some very specialized code points that
299# have to be handled specially for output, such as Hangul syllable names.
300# By creating a range type (done late in the development process), it
301# allowed this to be stored with the range, and overridden by other input.
302# Originally these were stored in another data structure, and it became a
303# mess trying to decide if a second file that was for the same property was
304# overriding the earlier one or not.
305#
306# Why are there two kinds of tables, match and map?
307# (And there is a base class shared by the two as well.) As stated above,
308# they actually are for different things. Development proceeded much more
309# smoothly when I (khw) realized the distinction. Map tables are used to
310# give the property value for every code point (actually every code point
311# that doesn't map to a default value). Match tables are used for regular
312# expression matches, and are essentially the inverse mapping. Separating
313# the two allows more specialized methods, and error checks so that one
314# can't just take the intersection of two map tables, for example, as that
315# is nonsensical.
99870f4d 316#
23e33b60
KW
317# DEBUGGING
318#
678f13d5
KW
319# This program is written so it will run under miniperl. Occasionally changes
320# will cause an error where the backtrace doesn't work well under miniperl.
321# To diagnose the problem, you can instead run it under regular perl, if you
322# have one compiled.
323#
324# There is a good trace facility. To enable it, first sub DEBUG must be set
325# to return true. Then a line like
326#
327# local $to_trace = 1 if main::DEBUG;
328#
329# can be added to enable tracing in its lexical scope or until you insert
330# another line:
331#
332# local $to_trace = 0 if main::DEBUG;
333#
334# then use a line like "trace $a, @b, %c, ...;
335#
336# Some of the more complex subroutines already have trace statements in them.
337# Permanent trace statements should be like:
338#
339# trace ... if main::DEBUG && $to_trace;
340#
341# If there is just one or a few files that you're debugging, you can easily
342# cause most everything else to be skipped. Change the line
343#
344# my $debug_skip = 0;
345#
346# to 1, and every file whose object is in @input_file_objects and doesn't have
347# a, 'non_skip => 1,' in its constructor will be skipped.
348#
b4a0206c 349# To compare the output tables, it may be useful to specify the -annotate
c4019d52
KW
350# flag. This causes the tables to expand so there is one entry for each
351# non-algorithmically named code point giving, currently its name, and its
352# graphic representation if printable (and you have a font that knows about
353# it). This makes it easier to see what the particular code points are in
354# each output table. The tables are usable, but because they don't have
355# ranges (for the most part), a Perl using them will run slower. Non-named
356# code points are annotated with a description of their status, and contiguous
357# ones with the same description will be output as a range rather than
358# individually. Algorithmically named characters are also output as ranges,
359# except when there are just a few contiguous ones.
360#
99870f4d
KW
361# FUTURE ISSUES
362#
363# The program would break if Unicode were to change its names so that
364# interior white space, underscores, or dashes differences were significant
365# within property and property value names.
366#
367# It might be easier to use the xml versions of the UCD if this program ever
368# would need heavy revision, and the ability to handle old versions was not
369# required.
370#
371# There is the potential for name collisions, in that Perl has chosen names
372# that Unicode could decide it also likes. There have been such collisions in
373# the past, with mostly Perl deciding to adopt the Unicode definition of the
374# name. However in the 5.2 Unicode beta testing, there were a number of such
375# collisions, which were withdrawn before the final release, because of Perl's
376# and other's protests. These all involved new properties which began with
377# 'Is'. Based on the protests, Unicode is unlikely to try that again. Also,
378# many of the Perl-defined synonyms, like Any, Word, etc, are listed in a
379# Unicode document, so they are unlikely to be used by Unicode for another
380# purpose. However, they might try something beginning with 'In', or use any
381# of the other Perl-defined properties. This program will warn you of name
382# collisions, and refuse to generate tables with them, but manual intervention
383# will be required in this event. One scheme that could be implemented, if
384# necessary, would be to have this program generate another file, or add a
385# field to mktables.lst that gives the date of first definition of a property.
386# Each new release of Unicode would use that file as a basis for the next
387# iteration. And the Perl synonym addition code could sort based on the age
388# of the property, so older properties get priority, and newer ones that clash
389# would be refused; hence existing code would not be impacted, and some other
390# synonym would have to be used for the new property. This is ugly, and
391# manual intervention would certainly be easier to do in the short run; lets
392# hope it never comes to this.
678f13d5 393#
99870f4d
KW
394# A NOTE ON UNIHAN
395#
396# This program can generate tables from the Unihan database. But it doesn't
397# by default, letting the CPAN module Unicode::Unihan handle them. Prior to
398# version 5.2, this database was in a single file, Unihan.txt. In 5.2 the
399# database was split into 8 different files, all beginning with the letters
400# 'Unihan'. This program will read those file(s) if present, but it needs to
401# know which of the many properties in the file(s) should have tables created
402# for them. It will create tables for any properties listed in
403# PropertyAliases.txt and PropValueAliases.txt, plus any listed in the
404# @cjk_properties array and the @cjk_property_values array. Thus, if a
405# property you want is not in those files of the release you are building
406# against, you must add it to those two arrays. Starting in 4.0, the
407# Unicode_Radical_Stroke was listed in those files, so if the Unihan database
408# is present in the directory, a table will be generated for that property.
409# In 5.2, several more properties were added. For your convenience, the two
5f7264c7 410# arrays are initialized with all the 6.0 listed properties that are also in
99870f4d
KW
411# earlier releases. But these are commented out. You can just uncomment the
412# ones you want, or use them as a template for adding entries for other
413# properties.
414#
415# You may need to adjust the entries to suit your purposes. setup_unihan(),
416# and filter_unihan_line() are the functions where this is done. This program
417# already does some adjusting to make the lines look more like the rest of the
418# Unicode DB; You can see what that is in filter_unihan_line()
419#
420# There is a bug in the 3.2 data file in which some values for the
421# kPrimaryNumeric property have commas and an unexpected comment. A filter
422# could be added for these; or for a particular installation, the Unihan.txt
423# file could be edited to fix them.
99870f4d 424#
678f13d5
KW
425# HOW TO ADD A FILE TO BE PROCESSED
426#
427# A new file from Unicode needs to have an object constructed for it in
428# @input_file_objects, probably at the end or at the end of the extracted
429# ones. The program should warn you if its name will clash with others on
430# restrictive file systems, like DOS. If so, figure out a better name, and
431# add lines to the README.perl file giving that. If the file is a character
432# property, it should be in the format that Unicode has by default
433# standardized for such files for the more recently introduced ones.
434# If so, the Input_file constructor for @input_file_objects can just be the
435# file name and release it first appeared in. If not, then it should be
436# possible to construct an each_line_handler() to massage the line into the
437# standardized form.
438#
439# For non-character properties, more code will be needed. You can look at
440# the existing entries for clues.
441#
442# UNICODE VERSIONS NOTES
443#
444# The Unicode UCD has had a number of errors in it over the versions. And
445# these remain, by policy, in the standard for that version. Therefore it is
446# risky to correct them, because code may be expecting the error. So this
447# program doesn't generally make changes, unless the error breaks the Perl
448# core. As an example, some versions of 2.1.x Jamo.txt have the wrong value
449# for U+1105, which causes real problems for the algorithms for Jamo
450# calculations, so it is changed here.
451#
452# But it isn't so clear cut as to what to do about concepts that are
453# introduced in a later release; should they extend back to earlier releases
454# where the concept just didn't exist? It was easier to do this than to not,
455# so that's what was done. For example, the default value for code points not
456# in the files for various properties was probably undefined until changed by
457# some version. No_Block for blocks is such an example. This program will
458# assign No_Block even in Unicode versions that didn't have it. This has the
459# benefit that code being written doesn't have to special case earlier
460# versions; and the detriment that it doesn't match the Standard precisely for
461# the affected versions.
462#
463# Here are some observations about some of the issues in early versions:
464#
6426c51b 465# The number of code points in \p{alpha} halved in 2.1.9. It turns out that
678f13d5
KW
466# the reason is that the CJK block starting at 4E00 was removed from PropList,
467# and was not put back in until 3.1.0
468#
469# Unicode introduced the synonym Space for White_Space in 4.1. Perl has
470# always had a \p{Space}. In release 3.2 only, they are not synonymous. The
471# reason is that 3.2 introduced U+205F=medium math space, which was not
472# classed as white space, but Perl figured out that it should have been. 4.0
473# reclassified it correctly.
474#
475# Another change between 3.2 and 4.0 is the CCC property value ATBL. In 3.2
476# this was erroneously a synonym for 202. In 4.0, ATB became 202, and ATBL
477# was left with no code points, as all the ones that mapped to 202 stayed
478# mapped to 202. Thus if your program used the numeric name for the class,
479# it would not have been affected, but if it used the mnemonic, it would have
480# been.
481#
482# \p{Script=Hrkt} (Katakana_Or_Hiragana) came in 4.0.1. Before that code
483# points which eventually came to have this script property value, instead
484# mapped to "Unknown". But in the next release all these code points were
485# moved to \p{sc=common} instead.
99870f4d
KW
486#
487# The default for missing code points for BidiClass is complicated. Starting
488# in 3.1.1, the derived file DBidiClass.txt handles this, but this program
489# tries to do the best it can for earlier releases. It is done in
490# process_PropertyAliases()
491#
492##############################################################################
493
494my $UNDEF = ':UNDEF:'; # String to print out for undefined values in tracing
495 # and errors
496my $MAX_LINE_WIDTH = 78;
497
498# Debugging aid to skip most files so as to not be distracted by them when
499# concentrating on the ones being debugged. Add
500# non_skip => 1,
501# to the constructor for those files you want processed when you set this.
502# Files with a first version number of 0 are special: they are always
c12f2655
KW
503# processed regardless of the state of this flag. Generally, Jamo.txt and
504# UnicodeData.txt must not be skipped if you want this program to not die
505# before normal completion.
99870f4d
KW
506my $debug_skip = 0;
507
508# Set to 1 to enable tracing.
509our $to_trace = 0;
510
511{ # Closure for trace: debugging aid
512 my $print_caller = 1; # ? Include calling subroutine name
513 my $main_with_colon = 'main::';
514 my $main_colon_length = length($main_with_colon);
515
516 sub trace {
517 return unless $to_trace; # Do nothing if global flag not set
518
519 my @input = @_;
520
521 local $DB::trace = 0;
522 $DB::trace = 0; # Quiet 'used only once' message
523
524 my $line_number;
525
526 # Loop looking up the stack to get the first non-trace caller
527 my $caller_line;
528 my $caller_name;
529 my $i = 0;
530 do {
531 $line_number = $caller_line;
532 (my $pkg, my $file, $caller_line, my $caller) = caller $i++;
533 $caller = $main_with_colon unless defined $caller;
534
535 $caller_name = $caller;
536
537 # get rid of pkg
538 $caller_name =~ s/.*:://;
539 if (substr($caller_name, 0, $main_colon_length)
540 eq $main_with_colon)
541 {
542 $caller_name = substr($caller_name, $main_colon_length);
543 }
544
545 } until ($caller_name ne 'trace');
546
547 # If the stack was empty, we were called from the top level
548 $caller_name = 'main' if ($caller_name eq ""
549 || $caller_name eq 'trace');
550
551 my $output = "";
552 foreach my $string (@input) {
553 #print STDERR __LINE__, ": ", join ", ", @input, "\n";
554 if (ref $string eq 'ARRAY' || ref $string eq 'HASH') {
555 $output .= simple_dumper($string);
556 }
557 else {
558 $string = "$string" if ref $string;
559 $string = $UNDEF unless defined $string;
560 chomp $string;
561 $string = '""' if $string eq "";
562 $output .= " " if $output ne ""
563 && $string ne ""
564 && substr($output, -1, 1) ne " "
565 && substr($string, 0, 1) ne " ";
566 $output .= $string;
567 }
568 }
569
99f78760
KW
570 print STDERR sprintf "%4d: ", $line_number if defined $line_number;
571 print STDERR "$caller_name: " if $print_caller;
99870f4d
KW
572 print STDERR $output, "\n";
573 return;
574 }
575}
576
577# This is for a rarely used development feature that allows you to compare two
578# versions of the Unicode standard without having to deal with changes caused
c12f2655
KW
579# by the code points introduced in the later version. Change the 0 to a
580# string containing a SINGLE dotted Unicode release number (e.g. "2.1"). Only
581# code points introduced in that release and earlier will be used; later ones
582# are thrown away. You use the version number of the earliest one you want to
583# compare; then run this program on directory structures containing each
584# release, and compare the outputs. These outputs will therefore include only
585# the code points common to both releases, and you can see the changes caused
586# just by the underlying release semantic changes. For versions earlier than
587# 3.2, you must copy a version of DAge.txt into the directory.
588my $string_compare_versions = DEBUG && 0; # e.g., "2.1";
99870f4d
KW
589my $compare_versions = DEBUG
590 && $string_compare_versions
591 && pack "C*", split /\./, $string_compare_versions;
592
593sub uniques {
594 # Returns non-duplicated input values. From "Perl Best Practices:
595 # Encapsulated Cleverness". p. 455 in first edition.
596
597 my %seen;
0e407844
NC
598 # Arguably this breaks encapsulation, if the goal is to permit multiple
599 # distinct objects to stringify to the same value, and be interchangeable.
600 # However, for this program, no two objects stringify identically, and all
601 # lists passed to this function are either objects or strings. So this
602 # doesn't affect correctness, but it does give a couple of percent speedup.
603 no overloading;
99870f4d
KW
604 return grep { ! $seen{$_}++ } @_;
605}
606
607$0 = File::Spec->canonpath($0);
608
609my $make_test_script = 0; # ? Should we output a test script
610my $write_unchanged_files = 0; # ? Should we update the output files even if
611 # we don't think they have changed
612my $use_directory = ""; # ? Should we chdir somewhere.
613my $pod_directory; # input directory to store the pod file.
614my $pod_file = 'perluniprops';
615my $t_path; # Path to the .t test file
616my $file_list = 'mktables.lst'; # File to store input and output file names.
617 # This is used to speed up the build, by not
618 # executing the main body of the program if
619 # nothing on the list has changed since the
620 # previous build
621my $make_list = 1; # ? Should we write $file_list. Set to always
622 # make a list so that when the pumpking is
623 # preparing a release, s/he won't have to do
624 # special things
625my $glob_list = 0; # ? Should we try to include unknown .txt files
626 # in the input.
bd9ebcfd
KW
627my $output_range_counts = $debugging_build; # ? Should we include the number
628 # of code points in ranges in
629 # the output
558712cf 630my $annotate = 0; # ? Should character names be in the output
9ef2b94f 631
99870f4d
KW
632# Verbosity levels; 0 is quiet
633my $NORMAL_VERBOSITY = 1;
634my $PROGRESS = 2;
635my $VERBOSE = 3;
636
637my $verbosity = $NORMAL_VERBOSITY;
638
639# Process arguments
640while (@ARGV) {
cf25bb62
JH
641 my $arg = shift @ARGV;
642 if ($arg eq '-v') {
99870f4d
KW
643 $verbosity = $VERBOSE;
644 }
645 elsif ($arg eq '-p') {
646 $verbosity = $PROGRESS;
647 $| = 1; # Flush buffers as we go.
648 }
649 elsif ($arg eq '-q') {
650 $verbosity = 0;
651 }
652 elsif ($arg eq '-w') {
653 $write_unchanged_files = 1; # update the files even if havent changed
654 }
655 elsif ($arg eq '-check') {
6ae7e459
YO
656 my $this = shift @ARGV;
657 my $ok = shift @ARGV;
658 if ($this ne $ok) {
659 print "Skipping as check params are not the same.\n";
660 exit(0);
661 }
00a8df5c 662 }
99870f4d
KW
663 elsif ($arg eq '-P' && defined ($pod_directory = shift)) {
664 -d $pod_directory or croak "Directory '$pod_directory' doesn't exist";
665 }
3df51b85
KW
666 elsif ($arg eq '-maketest' || ($arg eq '-T' && defined ($t_path = shift)))
667 {
99870f4d 668 $make_test_script = 1;
99870f4d
KW
669 }
670 elsif ($arg eq '-makelist') {
671 $make_list = 1;
672 }
673 elsif ($arg eq '-C' && defined ($use_directory = shift)) {
674 -d $use_directory or croak "Unknown directory '$use_directory'";
675 }
676 elsif ($arg eq '-L') {
677
678 # Existence not tested until have chdir'd
679 $file_list = shift;
680 }
681 elsif ($arg eq '-globlist') {
682 $glob_list = 1;
683 }
684 elsif ($arg eq '-c') {
685 $output_range_counts = ! $output_range_counts
686 }
b4a0206c 687 elsif ($arg eq '-annotate') {
558712cf 688 $annotate = 1;
bd9ebcfd
KW
689 $debugging_build = 1;
690 $output_range_counts = 1;
9ef2b94f 691 }
99870f4d
KW
692 else {
693 my $with_c = 'with';
694 $with_c .= 'out' if $output_range_counts; # Complements the state
695 croak <<END;
696usage: $0 [-c|-p|-q|-v|-w] [-C dir] [-L filelist] [ -P pod_dir ]
697 [ -T test_file_path ] [-globlist] [-makelist] [-maketest]
698 [-check A B ]
699 -c : Output comments $with_c number of code points in ranges
700 -q : Quiet Mode: Only output serious warnings.
701 -p : Set verbosity level to normal plus show progress.
702 -v : Set Verbosity level high: Show progress and non-serious
703 warnings
704 -w : Write files regardless
705 -C dir : Change to this directory before proceeding. All relative paths
706 except those specified by the -P and -T options will be done
707 with respect to this directory.
708 -P dir : Output $pod_file file to directory 'dir'.
3df51b85 709 -T path : Create a test script as 'path'; overrides -maketest
99870f4d
KW
710 -L filelist : Use alternate 'filelist' instead of standard one
711 -globlist : Take as input all non-Test *.txt files in current and sub
712 directories
3df51b85
KW
713 -maketest : Make test script 'TestProp.pl' in current (or -C directory),
714 overrides -T
99870f4d 715 -makelist : Rewrite the file list $file_list based on current setup
b4a0206c 716 -annotate : Output an annotation for each character in the table files;
c4019d52
KW
717 useful for debugging mktables, looking at diffs; but is slow,
718 memory intensive; resulting tables are usable but slow and
719 very large.
99870f4d
KW
720 -check A B : Executes $0 only if A and B are the same
721END
722 }
723}
724
725# Stores the most-recently changed file. If none have changed, can skip the
726# build
aeab6150 727my $most_recent = (stat $0)[9]; # Do this before the chdir!
99870f4d
KW
728
729# Change directories now, because need to read 'version' early.
730if ($use_directory) {
3df51b85 731 if ($pod_directory && ! File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d
KW
732 $pod_directory = File::Spec->rel2abs($pod_directory);
733 }
3df51b85 734 if ($t_path && ! File::Spec->file_name_is_absolute($t_path)) {
99870f4d 735 $t_path = File::Spec->rel2abs($t_path);
00a8df5c 736 }
99870f4d 737 chdir $use_directory or croak "Failed to chdir to '$use_directory':$!";
3df51b85 738 if ($pod_directory && File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d 739 $pod_directory = File::Spec->abs2rel($pod_directory);
02b1aeec 740 }
3df51b85 741 if ($t_path && File::Spec->file_name_is_absolute($t_path)) {
99870f4d 742 $t_path = File::Spec->abs2rel($t_path);
02b1aeec 743 }
00a8df5c
YO
744}
745
99870f4d
KW
746# Get Unicode version into regular and v-string. This is done now because
747# various tables below get populated based on it. These tables are populated
748# here to be near the top of the file, and so easily seeable by those needing
749# to modify things.
750open my $VERSION, "<", "version"
751 or croak "$0: can't open required file 'version': $!\n";
752my $string_version = <$VERSION>;
753close $VERSION;
754chomp $string_version;
755my $v_version = pack "C*", split /\./, $string_version; # v string
756
757# The following are the complete names of properties with property values that
758# are known to not match any code points in some versions of Unicode, but that
759# may change in the future so they should be matchable, hence an empty file is
760# generated for them.
761my @tables_that_may_be_empty = (
762 'Joining_Type=Left_Joining',
763 );
764push @tables_that_may_be_empty, 'Script=Common' if $v_version le v4.0.1;
765push @tables_that_may_be_empty, 'Title' if $v_version lt v2.0.0;
766push @tables_that_may_be_empty, 'Script=Katakana_Or_Hiragana'
767 if $v_version ge v4.1.0;
82aed44a
KW
768push @tables_that_may_be_empty, 'Script_Extensions=Katakana_Or_Hiragana'
769 if $v_version ge v6.0.0;
99870f4d
KW
770
771# The lists below are hashes, so the key is the item in the list, and the
772# value is the reason why it is in the list. This makes generation of
773# documentation easier.
774
775my %why_suppressed; # No file generated for these.
776
777# Files aren't generated for empty extraneous properties. This is arguable.
778# Extraneous properties generally come about because a property is no longer
779# used in a newer version of Unicode. If we generated a file without code
780# points, programs that used to work on that property will still execute
781# without errors. It just won't ever match (or will always match, with \P{}).
782# This means that the logic is now likely wrong. I (khw) think its better to
783# find this out by getting an error message. Just move them to the table
784# above to change this behavior
785my %why_suppress_if_empty_warn_if_not = (
786
787 # It is the only property that has ever officially been removed from the
788 # Standard. The database never contained any code points for it.
789 'Special_Case_Condition' => 'Obsolete',
790
791 # Apparently never official, but there were code points in some versions of
792 # old-style PropList.txt
793 'Non_Break' => 'Obsolete',
794);
795
796# These would normally go in the warn table just above, but they were changed
797# a long time before this program was written, so warnings about them are
798# moot.
799if ($v_version gt v3.2.0) {
800 push @tables_that_may_be_empty,
801 'Canonical_Combining_Class=Attached_Below_Left'
802}
803
5f7264c7 804# These are listed in the Property aliases file in 6.0, but Unihan is ignored
99870f4d
KW
805# unless explicitly added.
806if ($v_version ge v5.2.0) {
807 my $unihan = 'Unihan; remove from list if using Unihan';
ea25a9b2 808 foreach my $table (qw (
99870f4d
KW
809 kAccountingNumeric
810 kOtherNumeric
811 kPrimaryNumeric
812 kCompatibilityVariant
813 kIICore
814 kIRG_GSource
815 kIRG_HSource
816 kIRG_JSource
817 kIRG_KPSource
818 kIRG_MSource
819 kIRG_KSource
820 kIRG_TSource
821 kIRG_USource
822 kIRG_VSource
823 kRSUnicode
ea25a9b2 824 ))
99870f4d
KW
825 {
826 $why_suppress_if_empty_warn_if_not{$table} = $unihan;
827 }
ca12659b
NC
828}
829
272501f6
KW
830# Enum values for to_output_map() method in the Map_Table package.
831my $EXTERNAL_MAP = 1;
832my $INTERNAL_MAP = 2;
833
fcf1973c
KW
834# To override computed values for writing the map tables for these properties.
835# The default for enum map tables is to write them out, so that the Unicode
836# .txt files can be removed, but all the data to compute any property value
837# for any code point is available in a more compact form.
838my %global_to_output_map = (
839 # Needed by UCD.pm, but don't want to publicize that it exists, so won't
c12f2655
KW
840 # get stuck supporting it if things change. Since it is a STRING
841 # property, it normally would be listed in the pod, but INTERNAL_MAP
842 # suppresses that.
fcf1973c
KW
843 Unicode_1_Name => $INTERNAL_MAP,
844
845 Present_In => 0, # Suppress, as easily computed from Age
fcf1973c 846 Block => 0, # Suppress, as Blocks.txt is retained.
53d34b6c
KW
847
848 # Suppress, as mapping can be found instead from the
849 # Perl_Decomposition_Mapping file
850 Decomposition_Type => 0,
fcf1973c
KW
851);
852
99870f4d 853# Properties that this program ignores.
230e0c16
KW
854my @unimplemented_properties;
855
856# With this release, it is automatically handled if the Unihan db is
857# downloaded
858push @unimplemented_properties, 'Unicode_Radical_Stroke' if $v_version le v5.2.0;
d73e5302 859
99870f4d
KW
860# There are several types of obsolete properties defined by Unicode. These
861# must be hand-edited for every new Unicode release.
862my %why_deprecated; # Generates a deprecated warning message if used.
863my %why_stabilized; # Documentation only
864my %why_obsolete; # Documentation only
865
866{ # Closure
867 my $simple = 'Perl uses the more complete version of this property';
868 my $unihan = 'Unihan properties are by default not enabled in the Perl core. Instead use CPAN: Unicode::Unihan';
869
870 my $other_properties = 'other properties';
871 my $contributory = "Used by Unicode internally for generating $other_properties and not intended to be used stand-alone";
5d294d41 872 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
873
874 %why_deprecated = (
5f7264c7 875 'Grapheme_Link' => 'Deprecated by Unicode: Duplicates ccc=vr (Canonical_Combining_Class=Virama)',
99870f4d
KW
876 'Jamo_Short_Name' => $contributory,
877 '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',
878 'Other_Alphabetic' => $contributory,
879 'Other_Default_Ignorable_Code_Point' => $contributory,
880 'Other_Grapheme_Extend' => $contributory,
881 'Other_ID_Continue' => $contributory,
882 'Other_ID_Start' => $contributory,
883 'Other_Lowercase' => $contributory,
884 'Other_Math' => $contributory,
885 'Other_Uppercase' => $contributory,
e22aaf5c
KW
886 'Expands_On_NFC' => $why_no_expand,
887 'Expands_On_NFD' => $why_no_expand,
888 'Expands_On_NFKC' => $why_no_expand,
889 'Expands_On_NFKD' => $why_no_expand,
99870f4d
KW
890 );
891
892 %why_suppressed = (
5f7264c7 893 # There is a lib/unicore/Decomposition.pl (used by Normalize.pm) which
99870f4d
KW
894 # contains the same information, but without the algorithmically
895 # determinable Hangul syllables'. This file is not published, so it's
896 # existence is not noted in the comment.
897 'Decomposition_Mapping' => 'Accessible via Unicode::Normalize',
898
899 'ISO_Comment' => 'Apparently no demand for it, but can access it through Unicode::UCD::charinfo. Obsoleted, and code points for it removed in Unicode 5.2',
99870f4d
KW
900
901 'Simple_Case_Folding' => "$simple. Can access this through Unicode::UCD::casefold",
902 'Simple_Lowercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo",
903 'Simple_Titlecase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo",
904 'Simple_Uppercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo",
905
906 'Name' => "Accessible via 'use charnames;'",
907 'Name_Alias' => "Accessible via 'use charnames;'",
908
5f7264c7 909 FC_NFKC_Closure => 'Supplanted in usage by NFKC_Casefold; otherwise not useful',
99870f4d
KW
910 );
911
912 # The following are suppressed because they were made contributory or
913 # deprecated by Unicode before Perl ever thought about supporting them.
e22aaf5c
KW
914 foreach my $property ('Jamo_Short_Name',
915 'Grapheme_Link',
916 'Expands_On_NFC',
917 'Expands_On_NFD',
918 'Expands_On_NFKC',
919 'Expands_On_NFKD'
920 ) {
99870f4d
KW
921 $why_suppressed{$property} = $why_deprecated{$property};
922 }
cf25bb62 923
99870f4d
KW
924 # Customize the message for all the 'Other_' properties
925 foreach my $property (keys %why_deprecated) {
926 next if (my $main_property = $property) !~ s/^Other_//;
927 $why_deprecated{$property} =~ s/$other_properties/the $main_property property (which should be used instead)/;
928 }
929}
930
931if ($v_version ge 4.0.0) {
932 $why_stabilized{'Hyphen'} = 'Use the Line_Break property instead; see www.unicode.org/reports/tr14';
5f7264c7
KW
933 if ($v_version ge 6.0.0) {
934 $why_deprecated{'Hyphen'} = 'Supplanted by Line_Break property values; see www.unicode.org/reports/tr14';
935 }
99870f4d 936}
5f7264c7 937if ($v_version ge 5.2.0 && $v_version lt 6.0.0) {
99870f4d 938 $why_obsolete{'ISO_Comment'} = 'Code points for it have been removed';
5f7264c7 939 if ($v_version ge 6.0.0) {
63f74647 940 $why_deprecated{'ISO_Comment'} = 'No longer needed for Unicode\'s internal chart generation; otherwise not useful, and code points for it have been removed';
5f7264c7 941 }
99870f4d
KW
942}
943
944# Probably obsolete forever
945if ($v_version ge v4.1.0) {
82aed44a
KW
946 $why_suppressed{'Script=Katakana_Or_Hiragana'} = 'Obsolete. All code points previously matched by this have been moved to "Script=Common".';
947}
948if ($v_version ge v6.0.0) {
2b352efd
KW
949 $why_suppressed{'Script=Katakana_Or_Hiragana'} .= ' Consider instead using "Script_Extensions=Katakana" or "Script_Extensions=Hiragana (or both)"';
950 $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
951}
952
953# This program can create files for enumerated-like properties, such as
954# 'Numeric_Type'. This file would be the same format as for a string
955# property, with a mapping from code point to its value, so you could look up,
956# for example, the script a code point is in. But no one so far wants this
957# mapping, or they have found another way to get it since this is a new
958# feature. So no file is generated except if it is in this list.
959my @output_mapped_properties = split "\n", <<END;
960END
961
c12f2655
KW
962# If you are using the Unihan database in a Unicode version before 5.2, you
963# need to add the properties that you want to extract from it to this table.
964# For your convenience, the properties in the 6.0 PropertyAliases.txt file are
965# listed, commented out
99870f4d
KW
966my @cjk_properties = split "\n", <<'END';
967#cjkAccountingNumeric; kAccountingNumeric
968#cjkOtherNumeric; kOtherNumeric
969#cjkPrimaryNumeric; kPrimaryNumeric
970#cjkCompatibilityVariant; kCompatibilityVariant
971#cjkIICore ; kIICore
972#cjkIRG_GSource; kIRG_GSource
973#cjkIRG_HSource; kIRG_HSource
974#cjkIRG_JSource; kIRG_JSource
975#cjkIRG_KPSource; kIRG_KPSource
976#cjkIRG_KSource; kIRG_KSource
977#cjkIRG_TSource; kIRG_TSource
978#cjkIRG_USource; kIRG_USource
979#cjkIRG_VSource; kIRG_VSource
980#cjkRSUnicode; kRSUnicode ; Unicode_Radical_Stroke; URS
981END
982
983# Similarly for the property values. For your convenience, the lines in the
5f7264c7 984# 6.0 PropertyAliases.txt file are listed. Just remove the first BUT NOT both
c12f2655 985# '#' marks (for Unicode versions before 5.2)
99870f4d
KW
986my @cjk_property_values = split "\n", <<'END';
987## @missing: 0000..10FFFF; cjkAccountingNumeric; NaN
988## @missing: 0000..10FFFF; cjkCompatibilityVariant; <code point>
989## @missing: 0000..10FFFF; cjkIICore; <none>
990## @missing: 0000..10FFFF; cjkIRG_GSource; <none>
991## @missing: 0000..10FFFF; cjkIRG_HSource; <none>
992## @missing: 0000..10FFFF; cjkIRG_JSource; <none>
993## @missing: 0000..10FFFF; cjkIRG_KPSource; <none>
994## @missing: 0000..10FFFF; cjkIRG_KSource; <none>
995## @missing: 0000..10FFFF; cjkIRG_TSource; <none>
996## @missing: 0000..10FFFF; cjkIRG_USource; <none>
997## @missing: 0000..10FFFF; cjkIRG_VSource; <none>
998## @missing: 0000..10FFFF; cjkOtherNumeric; NaN
999## @missing: 0000..10FFFF; cjkPrimaryNumeric; NaN
1000## @missing: 0000..10FFFF; cjkRSUnicode; <none>
1001END
1002
1003# The input files don't list every code point. Those not listed are to be
1004# defaulted to some value. Below are hard-coded what those values are for
1005# non-binary properties as of 5.1. Starting in 5.0, there are
1006# machine-parsable comment lines in the files the give the defaults; so this
1007# list shouldn't have to be extended. The claim is that all missing entries
1008# for binary properties will default to 'N'. Unicode tried to change that in
1009# 5.2, but the beta period produced enough protest that they backed off.
1010#
1011# The defaults for the fields that appear in UnicodeData.txt in this hash must
1012# be in the form that it expects. The others may be synonyms.
1013my $CODE_POINT = '<code point>';
1014my %default_mapping = (
1015 Age => "Unassigned",
1016 # Bidi_Class => Complicated; set in code
1017 Bidi_Mirroring_Glyph => "",
1018 Block => 'No_Block',
1019 Canonical_Combining_Class => 0,
1020 Case_Folding => $CODE_POINT,
1021 Decomposition_Mapping => $CODE_POINT,
1022 Decomposition_Type => 'None',
1023 East_Asian_Width => "Neutral",
1024 FC_NFKC_Closure => $CODE_POINT,
1025 General_Category => 'Cn',
1026 Grapheme_Cluster_Break => 'Other',
1027 Hangul_Syllable_Type => 'NA',
1028 ISO_Comment => "",
1029 Jamo_Short_Name => "",
1030 Joining_Group => "No_Joining_Group",
1031 # Joining_Type => Complicated; set in code
1032 kIICore => 'N', # Is converted to binary
1033 #Line_Break => Complicated; set in code
1034 Lowercase_Mapping => $CODE_POINT,
1035 Name => "",
1036 Name_Alias => "",
1037 NFC_QC => 'Yes',
1038 NFD_QC => 'Yes',
1039 NFKC_QC => 'Yes',
1040 NFKD_QC => 'Yes',
1041 Numeric_Type => 'None',
1042 Numeric_Value => 'NaN',
1043 Script => ($v_version le 4.1.0) ? 'Common' : 'Unknown',
1044 Sentence_Break => 'Other',
1045 Simple_Case_Folding => $CODE_POINT,
1046 Simple_Lowercase_Mapping => $CODE_POINT,
1047 Simple_Titlecase_Mapping => $CODE_POINT,
1048 Simple_Uppercase_Mapping => $CODE_POINT,
1049 Titlecase_Mapping => $CODE_POINT,
1050 Unicode_1_Name => "",
1051 Unicode_Radical_Stroke => "",
1052 Uppercase_Mapping => $CODE_POINT,
1053 Word_Break => 'Other',
1054);
1055
1056# Below are files that Unicode furnishes, but this program ignores, and why
1057my %ignored_files = (
73ba1144
KW
1058 'CJKRadicals.txt' => 'Maps the kRSUnicode property values to corresponding code points',
1059 'Index.txt' => 'Alphabetical index of Unicode characters',
1060 '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',
1061 'NamesList.txt' => 'Annotated list of characters',
1062 'NormalizationCorrections.txt' => 'Documentation of corrections already incorporated into the Unicode data base',
1063 'Props.txt' => 'Only in very early releases; is a subset of F<PropList.txt> (which is used instead)',
1064 'ReadMe.txt' => 'Documentation',
1065 '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>',
1066 'EmojiSources.txt' => 'Maps certain Unicode code points to their legacy Japanese cell-phone values',
1067 'IndicMatraCategory.txt' => 'Provisional; for the analysis and processing of Indic scripts',
1068 'IndicSyllabicCategory.txt' => 'Provisional; for the analysis and processing of Indic scripts',
1069 'auxiliary/WordBreakTest.html' => 'Documentation of validation tests',
1070 'auxiliary/SentenceBreakTest.html' => 'Documentation of validation tests',
1071 'auxiliary/GraphemeBreakTest.html' => 'Documentation of validation tests',
1072 'auxiliary/LineBreakTest.html' => 'Documentation of validation tests',
99870f4d
KW
1073);
1074
678f13d5 1075### End of externally interesting definitions, except for @input_file_objects
99870f4d
KW
1076
1077my $HEADER=<<"EOF";
1078# !!!!!!! DO NOT EDIT THIS FILE !!!!!!!
3df51b85
KW
1079# This file is machine-generated by $0 from the Unicode
1080# database, Version $string_version. Any changes made here will be lost!
cf25bb62
JH
1081EOF
1082
126c3d4e 1083my $INTERNAL_ONLY_HEADER = <<"EOF";
99870f4d
KW
1084
1085# !!!!!!! INTERNAL PERL USE ONLY !!!!!!!
fac53429
KW
1086# This file is for internal use by core Perl only. The format and even the
1087# name or existence of this file are subject to change without notice. Don't
1088# use it directly.
99870f4d
KW
1089EOF
1090
1091my $DEVELOPMENT_ONLY=<<"EOF";
1092# !!!!!!! DEVELOPMENT USE ONLY !!!!!!!
1093# This file contains information artificially constrained to code points
1094# present in Unicode release $string_compare_versions.
1095# IT CANNOT BE RELIED ON. It is for use during development only and should
23e33b60 1096# not be used for production.
b6922eda
KW
1097
1098EOF
1099
6189eadc
KW
1100my $MAX_UNICODE_CODEPOINT_STRING = "10FFFF";
1101my $MAX_UNICODE_CODEPOINT = hex $MAX_UNICODE_CODEPOINT_STRING;
1102my $MAX_UNICODE_CODEPOINTS = $MAX_UNICODE_CODEPOINT + 1;
99870f4d
KW
1103
1104# Matches legal code point. 4-6 hex numbers, If there are 6, the first
1105# two must be 10; if there are 5, the first must not be a 0. Written this way
92199589
KW
1106# to decrease backtracking. The first regex allows the code point to be at
1107# the end of a word, but to work properly, the word shouldn't end with a valid
1108# hex character. The second one won't match a code point at the end of a
1109# word, and doesn't have the run-on issue
8c32d378
KW
1110my $run_on_code_point_re =
1111 qr/ (?: 10[0-9A-F]{4} | [1-9A-F][0-9A-F]{4} | [0-9A-F]{4} ) \b/x;
1112my $code_point_re = qr/\b$run_on_code_point_re/;
99870f4d
KW
1113
1114# This matches the beginning of the line in the Unicode db files that give the
1115# defaults for code points not listed (i.e., missing) in the file. The code
1116# depends on this ending with a semi-colon, so it can assume it is a valid
1117# field when the line is split() by semi-colons
1118my $missing_defaults_prefix =
6189eadc 1119 qr/^#\s+\@missing:\s+0000\.\.$MAX_UNICODE_CODEPOINT_STRING\s*;/;
99870f4d
KW
1120
1121# Property types. Unicode has more types, but these are sufficient for our
1122# purposes.
1123my $UNKNOWN = -1; # initialized to illegal value
1124my $NON_STRING = 1; # Either binary or enum
1125my $BINARY = 2;
06f26c45
KW
1126my $FORCED_BINARY = 3; # Not a binary property, but, besides its normal
1127 # tables, additional true and false tables are
1128 # generated so that false is anything matching the
1129 # default value, and true is everything else.
1130my $ENUM = 4; # Include catalog
1131my $STRING = 5; # Anything else: string or misc
99870f4d
KW
1132
1133# Some input files have lines that give default values for code points not
1134# contained in the file. Sometimes these should be ignored.
1135my $NO_DEFAULTS = 0; # Must evaluate to false
1136my $NOT_IGNORED = 1;
1137my $IGNORED = 2;
1138
1139# Range types. Each range has a type. Most ranges are type 0, for normal,
1140# and will appear in the main body of the tables in the output files, but
1141# there are other types of ranges as well, listed below, that are specially
1142# handled. There are pseudo-types as well that will never be stored as a
1143# type, but will affect the calculation of the type.
1144
1145# 0 is for normal, non-specials
1146my $MULTI_CP = 1; # Sequence of more than code point
1147my $HANGUL_SYLLABLE = 2;
1148my $CP_IN_NAME = 3; # The NAME contains the code point appended to it.
1149my $NULL = 4; # The map is to the null string; utf8.c can't
1150 # handle these, nor is there an accepted syntax
1151 # for them in \p{} constructs
f86864ac 1152my $COMPUTE_NO_MULTI_CP = 5; # Pseudo-type; means that ranges that would
99870f4d
KW
1153 # otherwise be $MULTI_CP type are instead type 0
1154
1155# process_generic_property_file() can accept certain overrides in its input.
1156# Each of these must begin AND end with $CMD_DELIM.
1157my $CMD_DELIM = "\a";
1158my $REPLACE_CMD = 'replace'; # Override the Replace
1159my $MAP_TYPE_CMD = 'map_type'; # Override the Type
1160
1161my $NO = 0;
1162my $YES = 1;
1163
1164# Values for the Replace argument to add_range.
1165# $NO # Don't replace; add only the code points not
1166 # already present.
1167my $IF_NOT_EQUIVALENT = 1; # Replace only under certain conditions; details in
1168 # the comments at the subroutine definition.
1169my $UNCONDITIONALLY = 2; # Replace without conditions.
1170my $MULTIPLE = 4; # Don't replace, but add a duplicate record if
1171 # already there
56343c78 1172my $CROAK = 5; # Die with an error if is already there
99870f4d
KW
1173
1174# Flags to give property statuses. The phrases are to remind maintainers that
1175# if the flag is changed, the indefinite article referring to it in the
1176# documentation may need to be as well.
1177my $NORMAL = "";
99870f4d
KW
1178my $DEPRECATED = 'D';
1179my $a_bold_deprecated = "a 'B<$DEPRECATED>'";
1180my $A_bold_deprecated = "A 'B<$DEPRECATED>'";
1181my $DISCOURAGED = 'X';
1182my $a_bold_discouraged = "an 'B<$DISCOURAGED>'";
1183my $A_bold_discouraged = "An 'B<$DISCOURAGED>'";
1184my $STRICTER = 'T';
1185my $a_bold_stricter = "a 'B<$STRICTER>'";
1186my $A_bold_stricter = "A 'B<$STRICTER>'";
1187my $STABILIZED = 'S';
1188my $a_bold_stabilized = "an 'B<$STABILIZED>'";
1189my $A_bold_stabilized = "An 'B<$STABILIZED>'";
1190my $OBSOLETE = 'O';
1191my $a_bold_obsolete = "an 'B<$OBSOLETE>'";
1192my $A_bold_obsolete = "An 'B<$OBSOLETE>'";
1193
1194my %status_past_participles = (
1195 $DISCOURAGED => 'discouraged',
99870f4d
KW
1196 $STABILIZED => 'stabilized',
1197 $OBSOLETE => 'obsolete',
37e2e78e 1198 $DEPRECATED => 'deprecated',
99870f4d
KW
1199);
1200
395dfc19
KW
1201# Table fates. These are somewhat ordered, so that fates < $MAP_PROXIED should be
1202# externally documented.
301ba948 1203my $ORDINARY = 0; # The normal fate.
395dfc19
KW
1204my $MAP_PROXIED = 1; # The map table for the property isn't written out,
1205 # but there is a file written that can be used to
1206 # reconstruct this table
301ba948
KW
1207my $SUPPRESSED = 3; # The file for this table is not written out.
1208my $INTERNAL_ONLY = 4; # The file for this table is written out, but it is
1209 # for Perl's internal use only
1210my $PLACEHOLDER = 5; # A property that is defined as a placeholder in a
1211 # Unicode version that doesn't have it, but we need it
1212 # to be defined, if empty, to have things work.
1213 # Implies no pod entry generated
1214
f5817e0a
KW
1215# The format of the values of the tables:
1216my $EMPTY_FORMAT = "";
99870f4d
KW
1217my $BINARY_FORMAT = 'b';
1218my $DECIMAL_FORMAT = 'd';
1219my $FLOAT_FORMAT = 'f';
1220my $INTEGER_FORMAT = 'i';
1221my $HEX_FORMAT = 'x';
1222my $RATIONAL_FORMAT = 'r';
1223my $STRING_FORMAT = 's';
a14f3cb1 1224my $DECOMP_STRING_FORMAT = 'c';
c3ff2976 1225my $STRING_WHITE_SPACE_LIST = 'sw';
99870f4d
KW
1226
1227my %map_table_formats = (
1228 $BINARY_FORMAT => 'binary',
1229 $DECIMAL_FORMAT => 'single decimal digit',
1230 $FLOAT_FORMAT => 'floating point number',
1231 $INTEGER_FORMAT => 'integer',
add63c13 1232 $HEX_FORMAT => 'non-negative hex whole number; a code point',
99870f4d 1233 $RATIONAL_FORMAT => 'rational: an integer or a fraction',
1a9d544b 1234 $STRING_FORMAT => 'string',
92f9d56c 1235 $DECOMP_STRING_FORMAT => 'Perl\'s internal (Normalize.pm) decomposition mapping',
c3ff2976 1236 $STRING_WHITE_SPACE_LIST => 'string, but some elements are interpreted as a list; white space occurs only as list item separators'
99870f4d
KW
1237);
1238
1239# Unicode didn't put such derived files in a separate directory at first.
1240my $EXTRACTED_DIR = (-d 'extracted') ? 'extracted' : "";
1241my $EXTRACTED = ($EXTRACTED_DIR) ? "$EXTRACTED_DIR/" : "";
1242my $AUXILIARY = 'auxiliary';
1243
1244# Hashes that will eventually go into Heavy.pl for the use of utf8_heavy.pl
9e4a1e86 1245# and into UCD.pl for the use of UCD.pm
99870f4d
KW
1246my %loose_to_file_of; # loosely maps table names to their respective
1247 # files
1248my %stricter_to_file_of; # same; but for stricter mapping.
1249my %nv_floating_to_rational; # maps numeric values floating point numbers to
1250 # their rational equivalent
c12f2655
KW
1251my %loose_property_name_of; # Loosely maps (non_string) property names to
1252 # standard form
86a52d1e 1253my %string_property_loose_to_name; # Same, for string properties.
9e4a1e86
KW
1254my %loose_to_standard_value; # loosely maps table names to the canonical
1255 # alias for them
5d1df013
KW
1256my %prop_aliases; # Keys are standard property name; values are each
1257 # one's aliases
1e863613
KW
1258my %prop_value_aliases; # Keys of top level are standard property name;
1259 # values are keys to another hash, Each one is
1260 # one of the property's values, in standard form.
1261 # The values are that prop-val's aliases.
99870f4d 1262
d867ccfb
KW
1263# Most properties are immune to caseless matching, otherwise you would get
1264# nonsensical results, as properties are a function of a code point, not
1265# everything that is caselessly equivalent to that code point. For example,
1266# Changes_When_Case_Folded('s') should be false, whereas caselessly it would
1267# be true because 's' and 'S' are equivalent caselessly. However,
1268# traditionally, [:upper:] and [:lower:] are equivalent caselessly, so we
1269# extend that concept to those very few properties that are like this. Each
1270# such property will match the full range caselessly. They are hard-coded in
1271# the program; it's not worth trying to make it general as it's extremely
1272# unlikely that they will ever change.
1273my %caseless_equivalent_to;
1274
99870f4d
KW
1275# These constants names and values were taken from the Unicode standard,
1276# version 5.1, section 3.12. They are used in conjunction with Hangul
6e5a209b
KW
1277# syllables. The '_string' versions are so generated tables can retain the
1278# hex format, which is the more familiar value
1279my $SBase_string = "0xAC00";
1280my $SBase = CORE::hex $SBase_string;
1281my $LBase_string = "0x1100";
1282my $LBase = CORE::hex $LBase_string;
1283my $VBase_string = "0x1161";
1284my $VBase = CORE::hex $VBase_string;
1285my $TBase_string = "0x11A7";
1286my $TBase = CORE::hex $TBase_string;
99870f4d
KW
1287my $SCount = 11172;
1288my $LCount = 19;
1289my $VCount = 21;
1290my $TCount = 28;
1291my $NCount = $VCount * $TCount;
1292
1293# For Hangul syllables; These store the numbers from Jamo.txt in conjunction
1294# with the above published constants.
1295my %Jamo;
1296my %Jamo_L; # Leading consonants
1297my %Jamo_V; # Vowels
1298my %Jamo_T; # Trailing consonants
1299
bb1dd3da
KW
1300# For code points whose name contains its ordinal as a '-ABCD' suffix.
1301# The key is the base name of the code point, and the value is an
1302# array giving all the ranges that use this base name. Each range
1303# is actually a hash giving the 'low' and 'high' values of it.
1304my %names_ending_in_code_point;
1305my %loose_names_ending_in_code_point; # Same as above, but has blanks, dashes
1306 # removed from the names
1307# Inverse mapping. The list of ranges that have these kinds of
1308# names. Each element contains the low, high, and base names in an
1309# anonymous hash.
1310my @code_points_ending_in_code_point;
1311
1312# Boolean: does this Unicode version have the hangul syllables, and are we
1313# writing out a table for them?
1314my $has_hangul_syllables = 0;
1315
1316# Does this Unicode version have code points whose names end in their
1317# respective code points, and are we writing out a table for them? 0 for no;
1318# otherwise points to first property that a table is needed for them, so that
1319# if multiple tables are needed, we don't create duplicates
1320my $needing_code_points_ending_in_code_point = 0;
1321
37e2e78e 1322my @backslash_X_tests; # List of tests read in for testing \X
99870f4d
KW
1323my @unhandled_properties; # Will contain a list of properties found in
1324 # the input that we didn't process.
f86864ac 1325my @match_properties; # Properties that have match tables, to be
99870f4d
KW
1326 # listed in the pod
1327my @map_properties; # Properties that get map files written
1328my @named_sequences; # NamedSequences.txt contents.
1329my %potential_files; # Generated list of all .txt files in the directory
1330 # structure so we can warn if something is being
1331 # ignored.
1332my @files_actually_output; # List of files we generated.
1333my @more_Names; # Some code point names are compound; this is used
1334 # to store the extra components of them.
1335my $MIN_FRACTION_LENGTH = 3; # How many digits of a floating point number at
1336 # the minimum before we consider it equivalent to a
1337 # candidate rational
1338my $MAX_FLOATING_SLOP = 10 ** - $MIN_FRACTION_LENGTH; # And in floating terms
1339
1340# These store references to certain commonly used property objects
1341my $gc;
1342my $perl;
1343my $block;
3e20195b
KW
1344my $perl_charname;
1345my $print;
7fc6cb55 1346my $Any;
359523e2 1347my $script;
99870f4d
KW
1348
1349# Are there conflicting names because of beginning with 'In_', or 'Is_'
1350my $has_In_conflicts = 0;
1351my $has_Is_conflicts = 0;
1352
1353sub internal_file_to_platform ($) {
1354 # Convert our file paths which have '/' separators to those of the
1355 # platform.
1356
1357 my $file = shift;
1358 return undef unless defined $file;
1359
1360 return File::Spec->join(split '/', $file);
d07a55ed 1361}
5beb625e 1362
99870f4d
KW
1363sub file_exists ($) { # platform independent '-e'. This program internally
1364 # uses slash as a path separator.
1365 my $file = shift;
1366 return 0 if ! defined $file;
1367 return -e internal_file_to_platform($file);
1368}
5beb625e 1369
99870f4d 1370sub objaddr($) {
23e33b60
KW
1371 # Returns the address of the blessed input object.
1372 # It doesn't check for blessedness because that would do a string eval
1373 # every call, and the program is structured so that this is never called
1374 # for a non-blessed object.
99870f4d 1375
23e33b60 1376 no overloading; # If overloaded, numifying below won't work.
99870f4d
KW
1377
1378 # Numifying a ref gives its address.
051df77b 1379 return pack 'J', $_[0];
99870f4d
KW
1380}
1381
558712cf 1382# These are used only if $annotate is true.
c4019d52
KW
1383# The entire range of Unicode characters is examined to populate these
1384# after all the input has been processed. But most can be skipped, as they
1385# have the same descriptive phrases, such as being unassigned
1386my @viacode; # Contains the 1 million character names
1387my @printable; # boolean: And are those characters printable?
1388my @annotate_char_type; # Contains a type of those characters, specifically
1389 # for the purposes of annotation.
1390my $annotate_ranges; # A map of ranges of code points that have the same
98dc9551 1391 # name for the purposes of annotation. They map to the
c4019d52
KW
1392 # upper edge of the range, so that the end point can
1393 # be immediately found. This is used to skip ahead to
1394 # the end of a range, and avoid processing each
1395 # individual code point in it.
1396my $unassigned_sans_noncharacters; # A Range_List of the unassigned
1397 # characters, but excluding those which are
1398 # also noncharacter code points
1399
1400# The annotation types are an extension of the regular range types, though
1401# some of the latter are folded into one. Make the new types negative to
1402# avoid conflicting with the regular types
1403my $SURROGATE_TYPE = -1;
1404my $UNASSIGNED_TYPE = -2;
1405my $PRIVATE_USE_TYPE = -3;
1406my $NONCHARACTER_TYPE = -4;
1407my $CONTROL_TYPE = -5;
1408my $UNKNOWN_TYPE = -6; # Used only if there is a bug in this program
1409
1410sub populate_char_info ($) {
558712cf 1411 # Used only with the $annotate option. Populates the arrays with the
c4019d52
KW
1412 # input code point's info that are needed for outputting more detailed
1413 # comments. If calling context wants a return, it is the end point of
1414 # any contiguous range of characters that share essentially the same info
1415
1416 my $i = shift;
1417 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1418
1419 $viacode[$i] = $perl_charname->value_of($i) || "";
1420
1421 # A character is generally printable if Unicode says it is,
1422 # but below we make sure that most Unicode general category 'C' types
1423 # aren't.
1424 $printable[$i] = $print->contains($i);
1425
1426 $annotate_char_type[$i] = $perl_charname->type_of($i) || 0;
1427
1428 # Only these two regular types are treated specially for annotations
1429 # purposes
1430 $annotate_char_type[$i] = 0 if $annotate_char_type[$i] != $CP_IN_NAME
1431 && $annotate_char_type[$i] != $HANGUL_SYLLABLE;
1432
1433 # Give a generic name to all code points that don't have a real name.
1434 # We output ranges, if applicable, for these. Also calculate the end
1435 # point of the range.
1436 my $end;
1437 if (! $viacode[$i]) {
1438 if ($gc-> table('Surrogate')->contains($i)) {
1439 $viacode[$i] = 'Surrogate';
1440 $annotate_char_type[$i] = $SURROGATE_TYPE;
1441 $printable[$i] = 0;
1442 $end = $gc->table('Surrogate')->containing_range($i)->end;
1443 }
1444 elsif ($gc-> table('Private_use')->contains($i)) {
1445 $viacode[$i] = 'Private Use';
1446 $annotate_char_type[$i] = $PRIVATE_USE_TYPE;
1447 $printable[$i] = 0;
1448 $end = $gc->table('Private_Use')->containing_range($i)->end;
1449 }
1450 elsif (Property::property_ref('Noncharacter_Code_Point')-> table('Y')->
1451 contains($i))
1452 {
1453 $viacode[$i] = 'Noncharacter';
1454 $annotate_char_type[$i] = $NONCHARACTER_TYPE;
1455 $printable[$i] = 0;
1456 $end = property_ref('Noncharacter_Code_Point')->table('Y')->
1457 containing_range($i)->end;
1458 }
1459 elsif ($gc-> table('Control')->contains($i)) {
1460 $viacode[$i] = 'Control';
1461 $annotate_char_type[$i] = $CONTROL_TYPE;
1462 $printable[$i] = 0;
1463 $end = 0x81 if $i == 0x80; # Hard-code this one known case
1464 }
1465 elsif ($gc-> table('Unassigned')->contains($i)) {
1466 $viacode[$i] = 'Unassigned, block=' . $block-> value_of($i);
1467 $annotate_char_type[$i] = $UNASSIGNED_TYPE;
1468 $printable[$i] = 0;
1469
1470 # Because we name the unassigned by the blocks they are in, it
1471 # can't go past the end of that block, and it also can't go past
1472 # the unassigned range it is in. The special table makes sure
1473 # that the non-characters, which are unassigned, are separated
1474 # out.
1475 $end = min($block->containing_range($i)->end,
1476 $unassigned_sans_noncharacters-> containing_range($i)->
1477 end);
13ca76ff
KW
1478 }
1479 else {
1480 Carp::my_carp_bug("Can't figure out how to annotate "
1481 . sprintf("U+%04X", $i)
1482 . ". Proceeding anyway.");
c4019d52
KW
1483 $viacode[$i] = 'UNKNOWN';
1484 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1485 $printable[$i] = 0;
1486 }
1487 }
1488
1489 # Here, has a name, but if it's one in which the code point number is
1490 # appended to the name, do that.
1491 elsif ($annotate_char_type[$i] == $CP_IN_NAME) {
1492 $viacode[$i] .= sprintf("-%04X", $i);
1493 $end = $perl_charname->containing_range($i)->end;
1494 }
1495
1496 # And here, has a name, but if it's a hangul syllable one, replace it with
1497 # the correct name from the Unicode algorithm
1498 elsif ($annotate_char_type[$i] == $HANGUL_SYLLABLE) {
1499 use integer;
1500 my $SIndex = $i - $SBase;
1501 my $L = $LBase + $SIndex / $NCount;
1502 my $V = $VBase + ($SIndex % $NCount) / $TCount;
1503 my $T = $TBase + $SIndex % $TCount;
1504 $viacode[$i] = "HANGUL SYLLABLE $Jamo{$L}$Jamo{$V}";
1505 $viacode[$i] .= $Jamo{$T} if $T != $TBase;
1506 $end = $perl_charname->containing_range($i)->end;
1507 }
1508
1509 return if ! defined wantarray;
1510 return $i if ! defined $end; # If not a range, return the input
1511
1512 # Save this whole range so can find the end point quickly
1513 $annotate_ranges->add_map($i, $end, $end);
1514
1515 return $end;
1516}
1517
23e33b60
KW
1518# Commented code below should work on Perl 5.8.
1519## This 'require' doesn't necessarily work in miniperl, and even if it does,
1520## the native perl version of it (which is what would operate under miniperl)
1521## is extremely slow, as it does a string eval every call.
1522#my $has_fast_scalar_util = $\18 !~ /miniperl/
1523# && defined eval "require Scalar::Util";
1524#
1525#sub objaddr($) {
1526# # Returns the address of the blessed input object. Uses the XS version if
1527# # available. It doesn't check for blessedness because that would do a
1528# # string eval every call, and the program is structured so that this is
1529# # never called for a non-blessed object.
1530#
1531# return Scalar::Util::refaddr($_[0]) if $has_fast_scalar_util;
1532#
1533# # Check at least that is a ref.
1534# my $pkg = ref($_[0]) or return undef;
1535#
1536# # Change to a fake package to defeat any overloaded stringify
1537# bless $_[0], 'main::Fake';
1538#
1539# # Numifying a ref gives its address.
051df77b 1540# my $addr = pack 'J', $_[0];
23e33b60
KW
1541#
1542# # Return to original class
1543# bless $_[0], $pkg;
1544# return $addr;
1545#}
1546
99870f4d
KW
1547sub max ($$) {
1548 my $a = shift;
1549 my $b = shift;
1550 return $a if $a >= $b;
1551 return $b;
1552}
1553
1554sub min ($$) {
1555 my $a = shift;
1556 my $b = shift;
1557 return $a if $a <= $b;
1558 return $b;
1559}
1560
1561sub clarify_number ($) {
1562 # This returns the input number with underscores inserted every 3 digits
1563 # in large (5 digits or more) numbers. Input must be entirely digits, not
1564 # checked.
1565
1566 my $number = shift;
1567 my $pos = length($number) - 3;
1568 return $number if $pos <= 1;
1569 while ($pos > 0) {
1570 substr($number, $pos, 0) = '_';
1571 $pos -= 3;
5beb625e 1572 }
99870f4d 1573 return $number;
99598c8c
JH
1574}
1575
12ac2576 1576
99870f4d 1577package Carp;
7ebf06b3 1578
99870f4d
KW
1579# These routines give a uniform treatment of messages in this program. They
1580# are placed in the Carp package to cause the stack trace to not include them,
1581# although an alternative would be to use another package and set @CARP_NOT
1582# for it.
12ac2576 1583
99870f4d 1584our $Verbose = 1 if main::DEBUG; # Useful info when debugging
12ac2576 1585
99f78760
KW
1586# This is a work-around suggested by Nicholas Clark to fix a problem with Carp
1587# and overload trying to load Scalar:Util under miniperl. See
1588# http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2009-11/msg01057.html
1589undef $overload::VERSION;
1590
99870f4d
KW
1591sub my_carp {
1592 my $message = shift || "";
1593 my $nofold = shift || 0;
7ebf06b3 1594
99870f4d
KW
1595 if ($message) {
1596 $message = main::join_lines($message);
1597 $message =~ s/^$0: *//; # Remove initial program name
1598 $message =~ s/[.;,]+$//; # Remove certain ending punctuation
1599 $message = "\n$0: $message;";
12ac2576 1600
99870f4d
KW
1601 # Fold the message with program name, semi-colon end punctuation
1602 # (which looks good with the message that carp appends to it), and a
1603 # hanging indent for continuation lines.
1604 $message = main::simple_fold($message, "", 4) unless $nofold;
1605 $message =~ s/\n$//; # Remove the trailing nl so what carp
1606 # appends is to the same line
1607 }
12ac2576 1608
99870f4d 1609 return $message if defined wantarray; # If a caller just wants the msg
12ac2576 1610
99870f4d
KW
1611 carp $message;
1612 return;
1613}
7ebf06b3 1614
99870f4d
KW
1615sub my_carp_bug {
1616 # This is called when it is clear that the problem is caused by a bug in
1617 # this program.
7ebf06b3 1618
99870f4d
KW
1619 my $message = shift;
1620 $message =~ s/^$0: *//;
1621 $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");
1622 carp $message;
1623 return;
1624}
7ebf06b3 1625
99870f4d
KW
1626sub carp_too_few_args {
1627 if (@_ != 2) {
1628 my_carp_bug("Wrong number of arguments: to 'carp_too_few_arguments'. No action taken.");
1629 return;
12ac2576 1630 }
7ebf06b3 1631
99870f4d
KW
1632 my $args_ref = shift;
1633 my $count = shift;
7ebf06b3 1634
99870f4d
KW
1635 my_carp_bug("Need at least $count arguments to "
1636 . (caller 1)[3]
1637 . ". Instead got: '"
1638 . join ', ', @$args_ref
1639 . "'. No action taken.");
1640 return;
12ac2576
JP
1641}
1642
99870f4d
KW
1643sub carp_extra_args {
1644 my $args_ref = shift;
1645 my_carp_bug("Too many arguments to 'carp_extra_args': (" . join(', ', @_) . "); Extras ignored.") if @_;
12ac2576 1646
99870f4d
KW
1647 unless (ref $args_ref) {
1648 my_carp_bug("Argument to 'carp_extra_args' ($args_ref) must be a ref. Not checking arguments.");
1649 return;
1650 }
1651 my ($package, $file, $line) = caller;
1652 my $subroutine = (caller 1)[3];
cf25bb62 1653
99870f4d
KW
1654 my $list;
1655 if (ref $args_ref eq 'HASH') {
1656 foreach my $key (keys %$args_ref) {
1657 $args_ref->{$key} = $UNDEF unless defined $args_ref->{$key};
cf25bb62 1658 }
99870f4d 1659 $list = join ', ', each %{$args_ref};
cf25bb62 1660 }
99870f4d
KW
1661 elsif (ref $args_ref eq 'ARRAY') {
1662 foreach my $arg (@$args_ref) {
1663 $arg = $UNDEF unless defined $arg;
1664 }
1665 $list = join ', ', @$args_ref;
1666 }
1667 else {
1668 my_carp_bug("Can't cope with ref "
1669 . ref($args_ref)
1670 . " . argument to 'carp_extra_args'. Not checking arguments.");
1671 return;
1672 }
1673
1674 my_carp_bug("Unrecognized parameters in options: '$list' to $subroutine. Skipped.");
1675 return;
d73e5302
JH
1676}
1677
99870f4d
KW
1678package main;
1679
1680{ # Closure
1681
1682 # This program uses the inside-out method for objects, as recommended in
1683 # "Perl Best Practices". This closure aids in generating those. There
1684 # are two routines. setup_package() is called once per package to set
1685 # things up, and then set_access() is called for each hash representing a
1686 # field in the object. These routines arrange for the object to be
1687 # properly destroyed when no longer used, and for standard accessor
1688 # functions to be generated. If you need more complex accessors, just
1689 # write your own and leave those accesses out of the call to set_access().
1690 # More details below.
1691
1692 my %constructor_fields; # fields that are to be used in constructors; see
1693 # below
1694
1695 # The values of this hash will be the package names as keys to other
1696 # hashes containing the name of each field in the package as keys, and
1697 # references to their respective hashes as values.
1698 my %package_fields;
1699
1700 sub setup_package {
1701 # Sets up the package, creating standard DESTROY and dump methods
1702 # (unless already defined). The dump method is used in debugging by
1703 # simple_dumper().
1704 # The optional parameters are:
1705 # a) a reference to a hash, that gets populated by later
1706 # set_access() calls with one of the accesses being
1707 # 'constructor'. The caller can then refer to this, but it is
1708 # not otherwise used by these two routines.
1709 # b) a reference to a callback routine to call during destruction
1710 # of the object, before any fields are actually destroyed
1711
1712 my %args = @_;
1713 my $constructor_ref = delete $args{'Constructor_Fields'};
1714 my $destroy_callback = delete $args{'Destroy_Callback'};
1715 Carp::carp_extra_args(\@_) if main::DEBUG && %args;
1716
1717 my %fields;
1718 my $package = (caller)[0];
1719
1720 $package_fields{$package} = \%fields;
1721 $constructor_fields{$package} = $constructor_ref;
1722
1723 unless ($package->can('DESTROY')) {
1724 my $destroy_name = "${package}::DESTROY";
1725 no strict "refs";
1726
1727 # Use typeglob to give the anonymous subroutine the name we want
1728 *$destroy_name = sub {
1729 my $self = shift;
ffe43484 1730 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1731
1732 $self->$destroy_callback if $destroy_callback;
1733 foreach my $field (keys %{$package_fields{$package}}) {
1734 #print STDERR __LINE__, ": Destroying ", ref $self, " ", sprintf("%04X", $addr), ": ", $field, "\n";
1735 delete $package_fields{$package}{$field}{$addr};
1736 }
1737 return;
1738 }
1739 }
1740
1741 unless ($package->can('dump')) {
1742 my $dump_name = "${package}::dump";
1743 no strict "refs";
1744 *$dump_name = sub {
1745 my $self = shift;
1746 return dump_inside_out($self, $package_fields{$package}, @_);
1747 }
1748 }
1749 return;
1750 }
1751
1752 sub set_access {
1753 # Arrange for the input field to be garbage collected when no longer
1754 # needed. Also, creates standard accessor functions for the field
1755 # based on the optional parameters-- none if none of these parameters:
1756 # 'addable' creates an 'add_NAME()' accessor function.
1757 # 'readable' or 'readable_array' creates a 'NAME()' accessor
1758 # function.
1759 # 'settable' creates a 'set_NAME()' accessor function.
1760 # 'constructor' doesn't create an accessor function, but adds the
1761 # field to the hash that was previously passed to
1762 # setup_package();
1763 # Any of the accesses can be abbreviated down, so that 'a', 'ad',
1764 # 'add' etc. all mean 'addable'.
1765 # The read accessor function will work on both array and scalar
1766 # values. If another accessor in the parameter list is 'a', the read
1767 # access assumes an array. You can also force it to be array access
1768 # by specifying 'readable_array' instead of 'readable'
1769 #
1770 # A sort-of 'protected' access can be set-up by preceding the addable,
1771 # readable or settable with some initial portion of 'protected_' (but,
1772 # the underscore is required), like 'p_a', 'pro_set', etc. The
1773 # "protection" is only by convention. All that happens is that the
1774 # accessor functions' names begin with an underscore. So instead of
1775 # calling set_foo, the call is _set_foo. (Real protection could be
c1739a4a 1776 # accomplished by having a new subroutine, end_package, called at the
99870f4d
KW
1777 # end of each package, and then storing the __LINE__ ranges and
1778 # checking them on every accessor. But that is way overkill.)
1779
1780 # We create anonymous subroutines as the accessors and then use
1781 # typeglobs to assign them to the proper package and name
1782
1783 my $name = shift; # Name of the field
1784 my $field = shift; # Reference to the inside-out hash containing the
1785 # field
1786
1787 my $package = (caller)[0];
1788
1789 if (! exists $package_fields{$package}) {
1790 croak "$0: Must call 'setup_package' before 'set_access'";
1791 }
d73e5302 1792
99870f4d
KW
1793 # Stash the field so DESTROY can get it.
1794 $package_fields{$package}{$name} = $field;
cf25bb62 1795
99870f4d
KW
1796 # Remaining arguments are the accessors. For each...
1797 foreach my $access (@_) {
1798 my $access = lc $access;
cf25bb62 1799
99870f4d 1800 my $protected = "";
cf25bb62 1801
99870f4d
KW
1802 # Match the input as far as it goes.
1803 if ($access =~ /^(p[^_]*)_/) {
1804 $protected = $1;
1805 if (substr('protected_', 0, length $protected)
1806 eq $protected)
1807 {
1808
1809 # Add 1 for the underscore not included in $protected
1810 $access = substr($access, length($protected) + 1);
1811 $protected = '_';
1812 }
1813 else {
1814 $protected = "";
1815 }
1816 }
1817
1818 if (substr('addable', 0, length $access) eq $access) {
1819 my $subname = "${package}::${protected}add_$name";
1820 no strict "refs";
1821
1822 # add_ accessor. Don't add if already there, which we
1823 # determine using 'eq' for scalars and '==' otherwise.
1824 *$subname = sub {
1825 use strict "refs";
1826 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
1827 my $self = shift;
1828 my $value = shift;
ffe43484 1829 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1830 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1831 if (ref $value) {
f998e60c 1832 return if grep { $value == $_ } @{$field->{$addr}};
99870f4d
KW
1833 }
1834 else {
f998e60c 1835 return if grep { $value eq $_ } @{$field->{$addr}};
99870f4d 1836 }
f998e60c 1837 push @{$field->{$addr}}, $value;
99870f4d
KW
1838 return;
1839 }
1840 }
1841 elsif (substr('constructor', 0, length $access) eq $access) {
1842 if ($protected) {
1843 Carp::my_carp_bug("Can't set-up 'protected' constructors")
1844 }
1845 else {
1846 $constructor_fields{$package}{$name} = $field;
1847 }
1848 }
1849 elsif (substr('readable_array', 0, length $access) eq $access) {
1850
1851 # Here has read access. If one of the other parameters for
1852 # access is array, or this one specifies array (by being more
1853 # than just 'readable_'), then create a subroutine that
1854 # assumes the data is an array. Otherwise just a scalar
1855 my $subname = "${package}::${protected}$name";
1856 if (grep { /^a/i } @_
1857 or length($access) > length('readable_'))
1858 {
1859 no strict "refs";
1860 *$subname = sub {
1861 use strict "refs";
23e33b60 1862 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
ffe43484 1863 my $addr = do { no overloading; pack 'J', $_[0]; };
99870f4d
KW
1864 if (ref $field->{$addr} ne 'ARRAY') {
1865 my $type = ref $field->{$addr};
1866 $type = 'scalar' unless $type;
1867 Carp::my_carp_bug("Trying to read $name as an array when it is a $type. Big problems.");
1868 return;
1869 }
1870 return scalar @{$field->{$addr}} unless wantarray;
1871
1872 # Make a copy; had problems with caller modifying the
1873 # original otherwise
1874 my @return = @{$field->{$addr}};
1875 return @return;
1876 }
1877 }
1878 else {
1879
1880 # Here not an array value, a simpler function.
1881 no strict "refs";
1882 *$subname = sub {
1883 use strict "refs";
23e33b60 1884 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
f998e60c 1885 no overloading;
051df77b 1886 return $field->{pack 'J', $_[0]};
99870f4d
KW
1887 }
1888 }
1889 }
1890 elsif (substr('settable', 0, length $access) eq $access) {
1891 my $subname = "${package}::${protected}set_$name";
1892 no strict "refs";
1893 *$subname = sub {
1894 use strict "refs";
23e33b60
KW
1895 if (main::DEBUG) {
1896 return Carp::carp_too_few_args(\@_, 2) if @_ < 2;
1897 Carp::carp_extra_args(\@_) if @_ > 2;
1898 }
1899 # $self is $_[0]; $value is $_[1]
f998e60c 1900 no overloading;
051df77b 1901 $field->{pack 'J', $_[0]} = $_[1];
99870f4d
KW
1902 return;
1903 }
1904 }
1905 else {
1906 Carp::my_carp_bug("Unknown accessor type $access. No accessor set.");
1907 }
cf25bb62 1908 }
99870f4d 1909 return;
cf25bb62 1910 }
99870f4d
KW
1911}
1912
1913package Input_file;
1914
1915# All input files use this object, which stores various attributes about them,
1916# and provides for convenient, uniform handling. The run method wraps the
1917# processing. It handles all the bookkeeping of opening, reading, and closing
1918# the file, returning only significant input lines.
1919#
1920# Each object gets a handler which processes the body of the file, and is
1921# called by run(). Most should use the generic, default handler, which has
1922# code scrubbed to handle things you might not expect. A handler should
1923# basically be a while(next_line()) {...} loop.
1924#
1925# You can also set up handlers to
1926# 1) call before the first line is read for pre processing
1927# 2) call to adjust each line of the input before the main handler gets them
1928# 3) call upon EOF before the main handler exits its loop
1929# 4) call at the end for post processing
1930#
1931# $_ is used to store the input line, and is to be filtered by the
1932# each_line_handler()s. So, if the format of the line is not in the desired
1933# format for the main handler, these are used to do that adjusting. They can
1934# be stacked (by enclosing them in an [ anonymous array ] in the constructor,
1935# so the $_ output of one is used as the input to the next. None of the other
1936# handlers are stackable, but could easily be changed to be so.
1937#
1938# Most of the handlers can call insert_lines() or insert_adjusted_lines()
1939# which insert the parameters as lines to be processed before the next input
1940# file line is read. This allows the EOF handler to flush buffers, for
1941# example. The difference between the two routines is that the lines inserted
1942# by insert_lines() are subjected to the each_line_handler()s. (So if you
1943# called it from such a handler, you would get infinite recursion.) Lines
1944# inserted by insert_adjusted_lines() go directly to the main handler without
1945# any adjustments. If the post-processing handler calls any of these, there
1946# will be no effect. Some error checking for these conditions could be added,
1947# but it hasn't been done.
1948#
1949# carp_bad_line() should be called to warn of bad input lines, which clears $_
1950# to prevent further processing of the line. This routine will output the
1951# message as a warning once, and then keep a count of the lines that have the
1952# same message, and output that count at the end of the file's processing.
1953# This keeps the number of messages down to a manageable amount.
1954#
1955# get_missings() should be called to retrieve any @missing input lines.
1956# Messages will be raised if this isn't done if the options aren't to ignore
1957# missings.
1958
1959sub trace { return main::trace(@_); }
1960
99870f4d
KW
1961{ # Closure
1962 # Keep track of fields that are to be put into the constructor.
1963 my %constructor_fields;
1964
1965 main::setup_package(Constructor_Fields => \%constructor_fields);
1966
1967 my %file; # Input file name, required
1968 main::set_access('file', \%file, qw{ c r });
1969
1970 my %first_released; # Unicode version file was first released in, required
1971 main::set_access('first_released', \%first_released, qw{ c r });
1972
1973 my %handler; # Subroutine to process the input file, defaults to
1974 # 'process_generic_property_file'
1975 main::set_access('handler', \%handler, qw{ c });
1976
1977 my %property;
1978 # name of property this file is for. defaults to none, meaning not
1979 # applicable, or is otherwise determinable, for example, from each line.
1980 main::set_access('property', \%property, qw{ c });
1981
1982 my %optional;
1983 # If this is true, the file is optional. If not present, no warning is
1984 # output. If it is present, the string given by this parameter is
1985 # evaluated, and if false the file is not processed.
1986 main::set_access('optional', \%optional, 'c', 'r');
1987
1988 my %non_skip;
1989 # This is used for debugging, to skip processing of all but a few input
1990 # files. Add 'non_skip => 1' to the constructor for those files you want
1991 # processed when you set the $debug_skip global.
1992 main::set_access('non_skip', \%non_skip, 'c');
1993
37e2e78e 1994 my %skip;
09ca89ce
KW
1995 # This is used to skip processing of this input file semi-permanently,
1996 # when it evaluates to true. The value should be the reason the file is
1997 # being skipped. It is used for files that we aren't planning to process
1998 # anytime soon, but want to allow to be in the directory and not raise a
1999 # message that we are not handling. Mostly for test files. This is in
2000 # contrast to the non_skip element, which is supposed to be used very
2001 # temporarily for debugging. Sets 'optional' to 1. Also, files that we
2002 # pretty much will never look at can be placed in the global
2003 # %ignored_files instead. Ones used here will be added to that list.
37e2e78e
KW
2004 main::set_access('skip', \%skip, 'c');
2005
99870f4d
KW
2006 my %each_line_handler;
2007 # list of subroutines to look at and filter each non-comment line in the
2008 # file. defaults to none. The subroutines are called in order, each is
2009 # to adjust $_ for the next one, and the final one adjusts it for
2010 # 'handler'
2011 main::set_access('each_line_handler', \%each_line_handler, 'c');
2012
2013 my %has_missings_defaults;
2014 # ? Are there lines in the file giving default values for code points
2015 # missing from it?. Defaults to NO_DEFAULTS. Otherwise NOT_IGNORED is
2016 # the norm, but IGNORED means it has such lines, but the handler doesn't
2017 # use them. Having these three states allows us to catch changes to the
2018 # UCD that this program should track
2019 main::set_access('has_missings_defaults',
2020 \%has_missings_defaults, qw{ c r });
2021
2022 my %pre_handler;
2023 # Subroutine to call before doing anything else in the file. If undef, no
2024 # such handler is called.
2025 main::set_access('pre_handler', \%pre_handler, qw{ c });
2026
2027 my %eof_handler;
2028 # Subroutine to call upon getting an EOF on the input file, but before
2029 # that is returned to the main handler. This is to allow buffers to be
2030 # flushed. The handler is expected to call insert_lines() or
2031 # insert_adjusted() with the buffered material
2032 main::set_access('eof_handler', \%eof_handler, qw{ c r });
2033
2034 my %post_handler;
2035 # Subroutine to call after all the lines of the file are read in and
2036 # processed. If undef, no such handler is called.
2037 main::set_access('post_handler', \%post_handler, qw{ c });
2038
2039 my %progress_message;
2040 # Message to print to display progress in lieu of the standard one
2041 main::set_access('progress_message', \%progress_message, qw{ c });
2042
2043 my %handle;
2044 # cache open file handle, internal. Is undef if file hasn't been
2045 # processed at all, empty if has;
2046 main::set_access('handle', \%handle);
2047
2048 my %added_lines;
2049 # cache of lines added virtually to the file, internal
2050 main::set_access('added_lines', \%added_lines);
2051
2052 my %errors;
2053 # cache of errors found, internal
2054 main::set_access('errors', \%errors);
2055
2056 my %missings;
2057 # storage of '@missing' defaults lines
2058 main::set_access('missings', \%missings);
2059
2060 sub new {
2061 my $class = shift;
2062
2063 my $self = bless \do{ my $anonymous_scalar }, $class;
ffe43484 2064 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2065
2066 # Set defaults
2067 $handler{$addr} = \&main::process_generic_property_file;
2068 $non_skip{$addr} = 0;
37e2e78e 2069 $skip{$addr} = 0;
99870f4d
KW
2070 $has_missings_defaults{$addr} = $NO_DEFAULTS;
2071 $handle{$addr} = undef;
2072 $added_lines{$addr} = [ ];
2073 $each_line_handler{$addr} = [ ];
2074 $errors{$addr} = { };
2075 $missings{$addr} = [ ];
2076
2077 # Two positional parameters.
99f78760 2078 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
99870f4d
KW
2079 $file{$addr} = main::internal_file_to_platform(shift);
2080 $first_released{$addr} = shift;
2081
2082 # The rest of the arguments are key => value pairs
2083 # %constructor_fields has been set up earlier to list all possible
2084 # ones. Either set or push, depending on how the default has been set
2085 # up just above.
2086 my %args = @_;
2087 foreach my $key (keys %args) {
2088 my $argument = $args{$key};
2089
2090 # Note that the fields are the lower case of the constructor keys
2091 my $hash = $constructor_fields{lc $key};
2092 if (! defined $hash) {
2093 Carp::my_carp_bug("Unrecognized parameters '$key => $argument' to new() for $self. Skipped");
2094 next;
2095 }
2096 if (ref $hash->{$addr} eq 'ARRAY') {
2097 if (ref $argument eq 'ARRAY') {
2098 foreach my $argument (@{$argument}) {
2099 next if ! defined $argument;
2100 push @{$hash->{$addr}}, $argument;
2101 }
2102 }
2103 else {
2104 push @{$hash->{$addr}}, $argument if defined $argument;
2105 }
2106 }
2107 else {
2108 $hash->{$addr} = $argument;
2109 }
2110 delete $args{$key};
2111 };
2112
2113 # If the file has a property for it, it means that the property is not
2114 # listed in the file's entries. So add a handler to the list of line
2115 # handlers to insert the property name into the lines, to provide a
2116 # uniform interface to the final processing subroutine.
2117 # the final code doesn't have to worry about that.
2118 if ($property{$addr}) {
2119 push @{$each_line_handler{$addr}}, \&_insert_property_into_line;
2120 }
2121
2122 if ($non_skip{$addr} && ! $debug_skip && $verbosity) {
2123 print "Warning: " . __PACKAGE__ . " constructor for $file{$addr} has useless 'non_skip' in it\n";
a3a8c5f0 2124 }
99870f4d 2125
09ca89ce
KW
2126 # If skipping, set to optional, and add to list of ignored files,
2127 # including its reason
2128 if ($skip{$addr}) {
2129 $optional{$addr} = 1;
2130 $ignored_files{$file{$addr}} = $skip{$addr}
2131 }
37e2e78e 2132
99870f4d 2133 return $self;
d73e5302
JH
2134 }
2135
cf25bb62 2136
99870f4d
KW
2137 use overload
2138 fallback => 0,
2139 qw("") => "_operator_stringify",
2140 "." => \&main::_operator_dot,
2141 ;
cf25bb62 2142
99870f4d
KW
2143 sub _operator_stringify {
2144 my $self = shift;
cf25bb62 2145
99870f4d 2146 return __PACKAGE__ . " object for " . $self->file;
d73e5302 2147 }
d73e5302 2148
99870f4d
KW
2149 # flag to make sure extracted files are processed early
2150 my $seen_non_extracted_non_age = 0;
d73e5302 2151
99870f4d
KW
2152 sub run {
2153 # Process the input object $self. This opens and closes the file and
2154 # calls all the handlers for it. Currently, this can only be called
2155 # once per file, as it destroy's the EOF handler
d73e5302 2156
99870f4d
KW
2157 my $self = shift;
2158 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
b6922eda 2159
ffe43484 2160 my $addr = do { no overloading; pack 'J', $self; };
b6922eda 2161
99870f4d 2162 my $file = $file{$addr};
d73e5302 2163
99870f4d
KW
2164 # Don't process if not expecting this file (because released later
2165 # than this Unicode version), and isn't there. This means if someone
2166 # copies it into an earlier version's directory, we will go ahead and
2167 # process it.
2168 return if $first_released{$addr} gt $v_version && ! -e $file;
2169
2170 # If in debugging mode and this file doesn't have the non-skip
2171 # flag set, and isn't one of the critical files, skip it.
2172 if ($debug_skip
2173 && $first_released{$addr} ne v0
2174 && ! $non_skip{$addr})
2175 {
2176 print "Skipping $file in debugging\n" if $verbosity;
2177 return;
2178 }
2179
2180 # File could be optional
37e2e78e 2181 if ($optional{$addr}) {
99870f4d
KW
2182 return unless -e $file;
2183 my $result = eval $optional{$addr};
2184 if (! defined $result) {
2185 Carp::my_carp_bug("Got '$@' when tried to eval $optional{$addr}. $file Skipped.");
2186 return;
2187 }
2188 if (! $result) {
2189 if ($verbosity) {
2190 print STDERR "Skipping processing input file '$file' because '$optional{$addr}' is not true\n";
2191 }
2192 return;
2193 }
2194 }
2195
2196 if (! defined $file || ! -e $file) {
2197
2198 # If the file doesn't exist, see if have internal data for it
2199 # (based on first_released being 0).
2200 if ($first_released{$addr} eq v0) {
2201 $handle{$addr} = 'pretend_is_open';
2202 }
2203 else {
2204 if (! $optional{$addr} # File could be optional
2205 && $v_version ge $first_released{$addr})
2206 {
2207 print STDERR "Skipping processing input file '$file' because not found\n" if $v_version ge $first_released{$addr};
2208 }
2209 return;
2210 }
2211 }
2212 else {
2213
37e2e78e
KW
2214 # Here, the file exists. Some platforms may change the case of
2215 # its name
99870f4d 2216 if ($seen_non_extracted_non_age) {
517956bf 2217 if ($file =~ /$EXTRACTED/i) {
99870f4d 2218 Carp::my_carp_bug(join_lines(<<END
99f78760 2219$file should be processed just after the 'Prop...Alias' files, and before
99870f4d
KW
2220anything not in the $EXTRACTED_DIR directory. Proceeding, but the results may
2221have subtle problems
2222END
2223 ));
2224 }
2225 }
2226 elsif ($EXTRACTED_DIR
2227 && $first_released{$addr} ne v0
517956bf
CB
2228 && $file !~ /$EXTRACTED/i
2229 && lc($file) ne 'dage.txt')
99870f4d
KW
2230 {
2231 # We don't set this (by the 'if' above) if we have no
2232 # extracted directory, so if running on an early version,
2233 # this test won't work. Not worth worrying about.
2234 $seen_non_extracted_non_age = 1;
2235 }
2236
2237 # And mark the file as having being processed, and warn if it
2238 # isn't a file we are expecting. As we process the files,
2239 # they are deleted from the hash, so any that remain at the
2240 # end of the program are files that we didn't process.
517956bf
CB
2241 my $fkey = File::Spec->rel2abs($file);
2242 my $expecting = delete $potential_files{$fkey};
2243 $expecting = delete $potential_files{lc($fkey)} unless defined $expecting;
678f13d5
KW
2244 Carp::my_carp("Was not expecting '$file'.") if
2245 ! $expecting
99870f4d
KW
2246 && ! defined $handle{$addr};
2247
37e2e78e
KW
2248 # Having deleted from expected files, we can quit if not to do
2249 # anything. Don't print progress unless really want verbosity
2250 if ($skip{$addr}) {
2251 print "Skipping $file.\n" if $verbosity >= $VERBOSE;
2252 return;
2253 }
2254
99870f4d
KW
2255 # Open the file, converting the slashes used in this program
2256 # into the proper form for the OS
2257 my $file_handle;
2258 if (not open $file_handle, "<", $file) {
2259 Carp::my_carp("Can't open $file. Skipping: $!");
2260 return 0;
2261 }
2262 $handle{$addr} = $file_handle; # Cache the open file handle
2263 }
2264
2265 if ($verbosity >= $PROGRESS) {
2266 if ($progress_message{$addr}) {
2267 print "$progress_message{$addr}\n";
2268 }
2269 else {
2270 # If using a virtual file, say so.
2271 print "Processing ", (-e $file)
2272 ? $file
2273 : "substitute $file",
2274 "\n";
2275 }
2276 }
2277
2278
2279 # Call any special handler for before the file.
2280 &{$pre_handler{$addr}}($self) if $pre_handler{$addr};
2281
2282 # Then the main handler
2283 &{$handler{$addr}}($self);
2284
2285 # Then any special post-file handler.
2286 &{$post_handler{$addr}}($self) if $post_handler{$addr};
2287
2288 # If any errors have been accumulated, output the counts (as the first
2289 # error message in each class was output when it was encountered).
2290 if ($errors{$addr}) {
2291 my $total = 0;
2292 my $types = 0;
2293 foreach my $error (keys %{$errors{$addr}}) {
2294 $total += $errors{$addr}->{$error};
2295 delete $errors{$addr}->{$error};
2296 $types++;
2297 }
2298 if ($total > 1) {
2299 my $message
2300 = "A total of $total lines had errors in $file. ";
2301
2302 $message .= ($types == 1)
2303 ? '(Only the first one was displayed.)'
2304 : '(Only the first of each type was displayed.)';
2305 Carp::my_carp($message);
2306 }
2307 }
2308
2309 if (@{$missings{$addr}}) {
2310 Carp::my_carp_bug("Handler for $file didn't look at all the \@missing lines. Generated tables likely are wrong");
2311 }
2312
2313 # If a real file handle, close it.
2314 close $handle{$addr} or Carp::my_carp("Can't close $file: $!") if
2315 ref $handle{$addr};
2316 $handle{$addr} = ""; # Uses empty to indicate that has already seen
2317 # the file, as opposed to undef
2318 return;
2319 }
2320
2321 sub next_line {
2322 # Sets $_ to be the next logical input line, if any. Returns non-zero
2323 # if such a line exists. 'logical' means that any lines that have
2324 # been added via insert_lines() will be returned in $_ before the file
2325 # is read again.
2326
2327 my $self = shift;
2328 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2329
ffe43484 2330 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2331
2332 # Here the file is open (or if the handle is not a ref, is an open
2333 # 'virtual' file). Get the next line; any inserted lines get priority
2334 # over the file itself.
2335 my $adjusted;
2336
2337 LINE:
2338 while (1) { # Loop until find non-comment, non-empty line
2339 #local $to_trace = 1 if main::DEBUG;
2340 my $inserted_ref = shift @{$added_lines{$addr}};
2341 if (defined $inserted_ref) {
2342 ($adjusted, $_) = @{$inserted_ref};
2343 trace $adjusted, $_ if main::DEBUG && $to_trace;
2344 return 1 if $adjusted;
2345 }
2346 else {
2347 last if ! ref $handle{$addr}; # Don't read unless is real file
2348 last if ! defined ($_ = readline $handle{$addr});
2349 }
2350 chomp;
2351 trace $_ if main::DEBUG && $to_trace;
2352
2353 # See if this line is the comment line that defines what property
2354 # value that code points that are not listed in the file should
2355 # have. The format or existence of these lines is not guaranteed
2356 # by Unicode since they are comments, but the documentation says
2357 # that this was added for machine-readability, so probably won't
2358 # change. This works starting in Unicode Version 5.0. They look
2359 # like:
2360 #
2361 # @missing: 0000..10FFFF; Not_Reordered
2362 # @missing: 0000..10FFFF; Decomposition_Mapping; <code point>
2363 # @missing: 0000..10FFFF; ; NaN
2364 #
2365 # Save the line for a later get_missings() call.
2366 if (/$missing_defaults_prefix/) {
2367 if ($has_missings_defaults{$addr} == $NO_DEFAULTS) {
2368 $self->carp_bad_line("Unexpected \@missing line. Assuming no missing entries");
2369 }
2370 elsif ($has_missings_defaults{$addr} == $NOT_IGNORED) {
2371 my @defaults = split /\s* ; \s*/x, $_;
2372
2373 # The first field is the @missing, which ends in a
2374 # semi-colon, so can safely shift.
2375 shift @defaults;
2376
2377 # Some of these lines may have empty field placeholders
2378 # which get in the way. An example is:
2379 # @missing: 0000..10FFFF; ; NaN
2380 # Remove them. Process starting from the top so the
2381 # splice doesn't affect things still to be looked at.
2382 for (my $i = @defaults - 1; $i >= 0; $i--) {
2383 next if $defaults[$i] ne "";
2384 splice @defaults, $i, 1;
2385 }
2386
2387 # What's left should be just the property (maybe) and the
2388 # default. Having only one element means it doesn't have
2389 # the property.
2390 my $default;
2391 my $property;
2392 if (@defaults >= 1) {
2393 if (@defaults == 1) {
2394 $default = $defaults[0];
2395 }
2396 else {
2397 $property = $defaults[0];
2398 $default = $defaults[1];
2399 }
2400 }
2401
2402 if (@defaults < 1
2403 || @defaults > 2
2404 || ($default =~ /^</
2405 && $default !~ /^<code *point>$/i
2406 && $default !~ /^<none>$/i))
2407 {
2408 $self->carp_bad_line("Unrecognized \@missing line: $_. Assuming no missing entries");
2409 }
2410 else {
2411
2412 # If the property is missing from the line, it should
2413 # be the one for the whole file
2414 $property = $property{$addr} if ! defined $property;
2415
2416 # Change <none> to the null string, which is what it
2417 # really means. If the default is the code point
2418 # itself, set it to <code point>, which is what
2419 # Unicode uses (but sometimes they've forgotten the
2420 # space)
2421 if ($default =~ /^<none>$/i) {
2422 $default = "";
2423 }
2424 elsif ($default =~ /^<code *point>$/i) {
2425 $default = $CODE_POINT;
2426 }
2427
2428 # Store them as a sub-arrays with both components.
2429 push @{$missings{$addr}}, [ $default, $property ];
2430 }
2431 }
2432
2433 # There is nothing for the caller to process on this comment
2434 # line.
2435 next;
2436 }
2437
2438 # Remove comments and trailing space, and skip this line if the
2439 # result is empty
2440 s/#.*//;
2441 s/\s+$//;
2442 next if /^$/;
2443
2444 # Call any handlers for this line, and skip further processing of
2445 # the line if the handler sets the line to null.
2446 foreach my $sub_ref (@{$each_line_handler{$addr}}) {
2447 &{$sub_ref}($self);
2448 next LINE if /^$/;
2449 }
2450
2451 # Here the line is ok. return success.
2452 return 1;
2453 } # End of looping through lines.
2454
2455 # If there is an EOF handler, call it (only once) and if it generates
2456 # more lines to process go back in the loop to handle them.
2457 if ($eof_handler{$addr}) {
2458 &{$eof_handler{$addr}}($self);
2459 $eof_handler{$addr} = ""; # Currently only get one shot at it.
2460 goto LINE if $added_lines{$addr};
2461 }
2462
2463 # Return failure -- no more lines.
2464 return 0;
2465
2466 }
2467
2468# Not currently used, not fully tested.
2469# sub peek {
2470# # Non-destructive look-ahead one non-adjusted, non-comment, non-blank
2471# # record. Not callable from an each_line_handler(), nor does it call
2472# # an each_line_handler() on the line.
2473#
2474# my $self = shift;
ffe43484 2475# my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2476#
2477# foreach my $inserted_ref (@{$added_lines{$addr}}) {
2478# my ($adjusted, $line) = @{$inserted_ref};
2479# next if $adjusted;
2480#
2481# # Remove comments and trailing space, and return a non-empty
2482# # resulting line
2483# $line =~ s/#.*//;
2484# $line =~ s/\s+$//;
2485# return $line if $line ne "";
2486# }
2487#
2488# return if ! ref $handle{$addr}; # Don't read unless is real file
2489# while (1) { # Loop until find non-comment, non-empty line
2490# local $to_trace = 1 if main::DEBUG;
2491# trace $_ if main::DEBUG && $to_trace;
2492# return if ! defined (my $line = readline $handle{$addr});
2493# chomp $line;
2494# push @{$added_lines{$addr}}, [ 0, $line ];
2495#
2496# $line =~ s/#.*//;
2497# $line =~ s/\s+$//;
2498# return $line if $line ne "";
2499# }
2500#
2501# return;
2502# }
2503
2504
2505 sub insert_lines {
2506 # Lines can be inserted so that it looks like they were in the input
2507 # file at the place it was when this routine is called. See also
2508 # insert_adjusted_lines(). Lines inserted via this routine go through
2509 # any each_line_handler()
2510
2511 my $self = shift;
2512
2513 # Each inserted line is an array, with the first element being 0 to
2514 # indicate that this line hasn't been adjusted, and needs to be
2515 # processed.
f998e60c 2516 no overloading;
051df77b 2517 push @{$added_lines{pack 'J', $self}}, map { [ 0, $_ ] } @_;
99870f4d
KW
2518 return;
2519 }
2520
2521 sub insert_adjusted_lines {
2522 # Lines can be inserted so that it looks like they were in the input
2523 # file at the place it was when this routine is called. See also
2524 # insert_lines(). Lines inserted via this routine are already fully
2525 # adjusted, ready to be processed; each_line_handler()s handlers will
2526 # not be called. This means this is not a completely general
2527 # facility, as only the last each_line_handler on the stack should
2528 # call this. It could be made more general, by passing to each of the
2529 # line_handlers their position on the stack, which they would pass on
2530 # to this routine, and that would replace the boolean first element in
2531 # the anonymous array pushed here, so that the next_line routine could
2532 # use that to call only those handlers whose index is after it on the
2533 # stack. But this is overkill for what is needed now.
2534
2535 my $self = shift;
2536 trace $_[0] if main::DEBUG && $to_trace;
2537
2538 # Each inserted line is an array, with the first element being 1 to
2539 # indicate that this line has been adjusted
f998e60c 2540 no overloading;
051df77b 2541 push @{$added_lines{pack 'J', $self}}, map { [ 1, $_ ] } @_;
99870f4d
KW
2542 return;
2543 }
2544
2545 sub get_missings {
2546 # Returns the stored up @missings lines' values, and clears the list.
2547 # The values are in an array, consisting of the default in the first
2548 # element, and the property in the 2nd. However, since these lines
2549 # can be stacked up, the return is an array of all these arrays.
2550
2551 my $self = shift;
2552 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2553
ffe43484 2554 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2555
2556 # If not accepting a list return, just return the first one.
2557 return shift @{$missings{$addr}} unless wantarray;
2558
2559 my @return = @{$missings{$addr}};
2560 undef @{$missings{$addr}};
2561 return @return;
2562 }
2563
2564 sub _insert_property_into_line {
2565 # Add a property field to $_, if this file requires it.
2566
f998e60c 2567 my $self = shift;
ffe43484 2568 my $addr = do { no overloading; pack 'J', $self; };
f998e60c 2569 my $property = $property{$addr};
99870f4d
KW
2570 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2571
2572 $_ =~ s/(;|$)/; $property$1/;
2573 return;
2574 }
2575
2576 sub carp_bad_line {
2577 # Output consistent error messages, using either a generic one, or the
2578 # one given by the optional parameter. To avoid gazillions of the
2579 # same message in case the syntax of a file is way off, this routine
2580 # only outputs the first instance of each message, incrementing a
2581 # count so the totals can be output at the end of the file.
2582
2583 my $self = shift;
2584 my $message = shift;
2585 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2586
ffe43484 2587 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2588
2589 $message = 'Unexpected line' unless $message;
2590
2591 # No trailing punctuation so as to fit with our addenda.
2592 $message =~ s/[.:;,]$//;
2593
2594 # If haven't seen this exact message before, output it now. Otherwise
2595 # increment the count of how many times it has occurred
2596 unless ($errors{$addr}->{$message}) {
2597 Carp::my_carp("$message in '$_' in "
f998e60c 2598 . $file{$addr}
99870f4d
KW
2599 . " at line $.. Skipping this line;");
2600 $errors{$addr}->{$message} = 1;
2601 }
2602 else {
2603 $errors{$addr}->{$message}++;
2604 }
2605
2606 # Clear the line to prevent any further (meaningful) processing of it.
2607 $_ = "";
2608
2609 return;
2610 }
2611} # End closure
2612
2613package Multi_Default;
2614
2615# Certain properties in early versions of Unicode had more than one possible
2616# default for code points missing from the files. In these cases, one
2617# default applies to everything left over after all the others are applied,
2618# and for each of the others, there is a description of which class of code
2619# points applies to it. This object helps implement this by storing the
2620# defaults, and for all but that final default, an eval string that generates
2621# the class that it applies to.
2622
2623
2624{ # Closure
2625
2626 main::setup_package();
2627
2628 my %class_defaults;
2629 # The defaults structure for the classes
2630 main::set_access('class_defaults', \%class_defaults);
2631
2632 my %other_default;
2633 # The default that applies to everything left over.
2634 main::set_access('other_default', \%other_default, 'r');
2635
2636
2637 sub new {
2638 # The constructor is called with default => eval pairs, terminated by
2639 # the left-over default. e.g.
2640 # Multi_Default->new(
2641 # 'T' => '$gc->table("Mn") + $gc->table("Cf") - 0x200C
2642 # - 0x200D',
2643 # 'R' => 'some other expression that evaluates to code points',
2644 # .
2645 # .
2646 # .
2647 # 'U'));
2648
2649 my $class = shift;
2650
2651 my $self = bless \do{my $anonymous_scalar}, $class;
ffe43484 2652 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2653
2654 while (@_ > 1) {
2655 my $default = shift;
2656 my $eval = shift;
2657 $class_defaults{$addr}->{$default} = $eval;
2658 }
2659
2660 $other_default{$addr} = shift;
2661
2662 return $self;
2663 }
2664
2665 sub get_next_defaults {
2666 # Iterates and returns the next class of defaults.
2667 my $self = shift;
2668 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2669
ffe43484 2670 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2671
2672 return each %{$class_defaults{$addr}};
2673 }
2674}
2675
2676package Alias;
2677
2678# An alias is one of the names that a table goes by. This class defines them
2679# including some attributes. Everything is currently setup in the
2680# constructor.
2681
2682
2683{ # Closure
2684
2685 main::setup_package();
2686
2687 my %name;
2688 main::set_access('name', \%name, 'r');
2689
2690 my %loose_match;
c12f2655 2691 # Should this name match loosely or not.
99870f4d
KW
2692 main::set_access('loose_match', \%loose_match, 'r');
2693
33e96e72
KW
2694 my %make_re_pod_entry;
2695 # Some aliases should not get their own entries in the re section of the
2696 # pod, because they are covered by a wild-card, and some we want to
2697 # discourage use of. Binary
f82fe4ba 2698 main::set_access('make_re_pod_entry', \%make_re_pod_entry, 'r', 's');
99870f4d 2699
fd1e3e84
KW
2700 my %ucd;
2701 # Is this documented to be accessible via Unicode::UCD
2702 main::set_access('ucd', \%ucd, 'r', 's');
2703
99870f4d
KW
2704 my %status;
2705 # Aliases have a status, like deprecated, or even suppressed (which means
2706 # they don't appear in documentation). Enum
2707 main::set_access('status', \%status, 'r');
2708
2709 my %externally_ok;
2710 # Similarly, some aliases should not be considered as usable ones for
2711 # external use, such as file names, or we don't want documentation to
2712 # recommend them. Boolean
2713 main::set_access('externally_ok', \%externally_ok, 'r');
2714
2715 sub new {
2716 my $class = shift;
2717
2718 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2719 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2720
2721 $name{$addr} = shift;
2722 $loose_match{$addr} = shift;
33e96e72 2723 $make_re_pod_entry{$addr} = shift;
99870f4d
KW
2724 $externally_ok{$addr} = shift;
2725 $status{$addr} = shift;
fd1e3e84 2726 $ucd{$addr} = shift;
99870f4d
KW
2727
2728 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2729
2730 # Null names are never ok externally
2731 $externally_ok{$addr} = 0 if $name{$addr} eq "";
2732
2733 return $self;
2734 }
2735}
2736
2737package Range;
2738
2739# A range is the basic unit for storing code points, and is described in the
2740# comments at the beginning of the program. Each range has a starting code
2741# point; an ending code point (not less than the starting one); a value
2742# that applies to every code point in between the two end-points, inclusive;
2743# and an enum type that applies to the value. The type is for the user's
2744# convenience, and has no meaning here, except that a non-zero type is
2745# considered to not obey the normal Unicode rules for having standard forms.
2746#
2747# The same structure is used for both map and match tables, even though in the
2748# latter, the value (and hence type) is irrelevant and could be used as a
2749# comment. In map tables, the value is what all the code points in the range
2750# map to. Type 0 values have the standardized version of the value stored as
2751# well, so as to not have to recalculate it a lot.
2752
2753sub trace { return main::trace(@_); }
2754
2755{ # Closure
2756
2757 main::setup_package();
2758
2759 my %start;
2760 main::set_access('start', \%start, 'r', 's');
2761
2762 my %end;
2763 main::set_access('end', \%end, 'r', 's');
2764
2765 my %value;
2766 main::set_access('value', \%value, 'r');
2767
2768 my %type;
2769 main::set_access('type', \%type, 'r');
2770
2771 my %standard_form;
2772 # The value in internal standard form. Defined only if the type is 0.
2773 main::set_access('standard_form', \%standard_form);
2774
2775 # Note that if these fields change, the dump() method should as well
2776
2777 sub new {
2778 return Carp::carp_too_few_args(\@_, 3) if main::DEBUG && @_ < 3;
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 $start{$addr} = shift;
2785 $end{$addr} = shift;
2786
2787 my %args = @_;
2788
2789 my $value = delete $args{'Value'}; # Can be 0
2790 $value = "" unless defined $value;
2791 $value{$addr} = $value;
2792
2793 $type{$addr} = delete $args{'Type'} || 0;
2794
2795 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2796
2797 if (! $type{$addr}) {
2798 $standard_form{$addr} = main::standardize($value);
2799 }
2800
2801 return $self;
2802 }
2803
2804 use overload
2805 fallback => 0,
2806 qw("") => "_operator_stringify",
2807 "." => \&main::_operator_dot,
2808 ;
2809
2810 sub _operator_stringify {
2811 my $self = shift;
ffe43484 2812 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2813
2814 # Output it like '0041..0065 (value)'
2815 my $return = sprintf("%04X", $start{$addr})
2816 . '..'
2817 . sprintf("%04X", $end{$addr});
2818 my $value = $value{$addr};
2819 my $type = $type{$addr};
2820 $return .= ' (';
2821 $return .= "$value";
2822 $return .= ", Type=$type" if $type != 0;
2823 $return .= ')';
2824
2825 return $return;
2826 }
2827
2828 sub standard_form {
2829 # The standard form is the value itself if the standard form is
2830 # undefined (that is if the value is special)
2831
2832 my $self = shift;
2833 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2834
ffe43484 2835 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2836
2837 return $standard_form{$addr} if defined $standard_form{$addr};
2838 return $value{$addr};
2839 }
2840
2841 sub dump {
2842 # Human, not machine readable. For machine readable, comment out this
2843 # entire routine and let the standard one take effect.
2844 my $self = shift;
2845 my $indent = shift;
2846 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2847
ffe43484 2848 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2849
2850 my $return = $indent
2851 . sprintf("%04X", $start{$addr})
2852 . '..'
2853 . sprintf("%04X", $end{$addr})
2854 . " '$value{$addr}';";
2855 if (! defined $standard_form{$addr}) {
2856 $return .= "(type=$type{$addr})";
2857 }
2858 elsif ($standard_form{$addr} ne $value{$addr}) {
2859 $return .= "(standard '$standard_form{$addr}')";
2860 }
2861 return $return;
2862 }
2863} # End closure
2864
2865package _Range_List_Base;
2866
2867# Base class for range lists. A range list is simply an ordered list of
2868# ranges, so that the ranges with the lowest starting numbers are first in it.
2869#
2870# When a new range is added that is adjacent to an existing range that has the
2871# same value and type, it merges with it to form a larger range.
2872#
2873# Ranges generally do not overlap, except that there can be multiple entries
2874# of single code point ranges. This is because of NameAliases.txt.
2875#
2876# In this program, there is a standard value such that if two different
2877# values, have the same standard value, they are considered equivalent. This
2878# value was chosen so that it gives correct results on Unicode data
2879
2880# There are a number of methods to manipulate range lists, and some operators
2881# are overloaded to handle them.
2882
99870f4d
KW
2883sub trace { return main::trace(@_); }
2884
2885{ # Closure
2886
2887 our $addr;
2888
2889 main::setup_package();
2890
2891 my %ranges;
2892 # The list of ranges
2893 main::set_access('ranges', \%ranges, 'readable_array');
2894
2895 my %max;
2896 # The highest code point in the list. This was originally a method, but
2897 # actual measurements said it was used a lot.
2898 main::set_access('max', \%max, 'r');
2899
2900 my %each_range_iterator;
2901 # Iterator position for each_range()
2902 main::set_access('each_range_iterator', \%each_range_iterator);
2903
2904 my %owner_name_of;
2905 # Name of parent this is attached to, if any. Solely for better error
2906 # messages.
2907 main::set_access('owner_name_of', \%owner_name_of, 'p_r');
2908
2909 my %_search_ranges_cache;
2910 # A cache of the previous result from _search_ranges(), for better
2911 # performance
2912 main::set_access('_search_ranges_cache', \%_search_ranges_cache);
2913
2914 sub new {
2915 my $class = shift;
2916 my %args = @_;
2917
2918 # Optional initialization data for the range list.
2919 my $initialize = delete $args{'Initialize'};
2920
2921 my $self;
2922
2923 # Use _union() to initialize. _union() returns an object of this
2924 # class, which means that it will call this constructor recursively.
2925 # But it won't have this $initialize parameter so that it won't
2926 # infinitely loop on this.
2927 return _union($class, $initialize, %args) if defined $initialize;
2928
2929 $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2930 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2931
2932 # Optional parent object, only for debug info.
2933 $owner_name_of{$addr} = delete $args{'Owner'};
2934 $owner_name_of{$addr} = "" if ! defined $owner_name_of{$addr};
2935
2936 # Stringify, in case it is an object.
2937 $owner_name_of{$addr} = "$owner_name_of{$addr}";
2938
2939 # This is used only for error messages, and so a colon is added
2940 $owner_name_of{$addr} .= ": " if $owner_name_of{$addr} ne "";
2941
2942 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2943
2944 # Max is initialized to a negative value that isn't adjacent to 0,
2945 # for simpler tests
2946 $max{$addr} = -2;
2947
2948 $_search_ranges_cache{$addr} = 0;
2949 $ranges{$addr} = [];
2950
2951 return $self;
2952 }
2953
2954 use overload
2955 fallback => 0,
2956 qw("") => "_operator_stringify",
2957 "." => \&main::_operator_dot,
2958 ;
2959
2960 sub _operator_stringify {
2961 my $self = shift;
ffe43484 2962 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2963
2964 return "Range_List attached to '$owner_name_of{$addr}'"
2965 if $owner_name_of{$addr};
2966 return "anonymous Range_List " . \$self;
2967 }
2968
2969 sub _union {
2970 # Returns the union of the input code points. It can be called as
2971 # either a constructor or a method. If called as a method, the result
2972 # will be a new() instance of the calling object, containing the union
2973 # of that object with the other parameter's code points; if called as
2974 # a constructor, the first parameter gives the class the new object
2975 # should be, and the second parameter gives the code points to go into
2976 # it.
2977 # In either case, there are two parameters looked at by this routine;
2978 # any additional parameters are passed to the new() constructor.
2979 #
2980 # The code points can come in the form of some object that contains
2981 # ranges, and has a conventionally named method to access them; or
2982 # they can be an array of individual code points (as integers); or
2983 # just a single code point.
2984 #
2985 # If they are ranges, this routine doesn't make any effort to preserve
2986 # the range values of one input over the other. Therefore this base
2987 # class should not allow _union to be called from other than
2988 # initialization code, so as to prevent two tables from being added
2989 # together where the range values matter. The general form of this
2990 # routine therefore belongs in a derived class, but it was moved here
2991 # to avoid duplication of code. The failure to overload this in this
2992 # class keeps it safe.
2993 #
2994
2995 my $self;
2996 my @args; # Arguments to pass to the constructor
2997
2998 my $class = shift;
2999
3000 # If a method call, will start the union with the object itself, and
3001 # the class of the new object will be the same as self.
3002 if (ref $class) {
3003 $self = $class;
3004 $class = ref $self;
3005 push @args, $self;
3006 }
3007
3008 # Add the other required parameter.
3009 push @args, shift;
3010 # Rest of parameters are passed on to the constructor
3011
3012 # Accumulate all records from both lists.
3013 my @records;
3014 for my $arg (@args) {
3015 #local $to_trace = 0 if main::DEBUG;
3016 trace "argument = $arg" if main::DEBUG && $to_trace;
3017 if (! defined $arg) {
3018 my $message = "";
3019 if (defined $self) {
f998e60c 3020 no overloading;
051df77b 3021 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
3022 }
3023 Carp::my_carp_bug($message .= "Undefined argument to _union. No union done.");
3024 return;
3025 }
3026 $arg = [ $arg ] if ! ref $arg;
3027 my $type = ref $arg;
3028 if ($type eq 'ARRAY') {
3029 foreach my $element (@$arg) {
3030 push @records, Range->new($element, $element);
3031 }
3032 }
3033 elsif ($arg->isa('Range')) {
3034 push @records, $arg;
3035 }
3036 elsif ($arg->can('ranges')) {
3037 push @records, $arg->ranges;
3038 }
3039 else {
3040 my $message = "";
3041 if (defined $self) {
f998e60c 3042 no overloading;
051df77b 3043 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
3044 }
3045 Carp::my_carp_bug($message . "Cannot take the union of a $type. No union done.");
3046 return;
3047 }
3048 }
3049
3050 # Sort with the range containing the lowest ordinal first, but if
3051 # two ranges start at the same code point, sort with the bigger range
3052 # of the two first, because it takes fewer cycles.
3053 @records = sort { ($a->start <=> $b->start)
3054 or
3055 # if b is shorter than a, b->end will be
3056 # less than a->end, and we want to select
3057 # a, so want to return -1
3058 ($b->end <=> $a->end)
3059 } @records;
3060
3061 my $new = $class->new(@_);
3062
3063 # Fold in records so long as they add new information.
3064 for my $set (@records) {
3065 my $start = $set->start;
3066 my $end = $set->end;
3067 my $value = $set->value;
3068 if ($start > $new->max) {
3069 $new->_add_delete('+', $start, $end, $value);
3070 }
3071 elsif ($end > $new->max) {
3072 $new->_add_delete('+', $new->max +1, $end, $value);
3073 }
3074 }
3075
3076 return $new;
3077 }
3078
3079 sub range_count { # Return the number of ranges in the range list
3080 my $self = shift;
3081 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3082
f998e60c 3083 no overloading;
051df77b 3084 return scalar @{$ranges{pack 'J', $self}};
99870f4d
KW
3085 }
3086
3087 sub min {
3088 # Returns the minimum code point currently in the range list, or if
3089 # the range list is empty, 2 beyond the max possible. This is a
3090 # method because used so rarely, that not worth saving between calls,
3091 # and having to worry about changing it as ranges are added and
3092 # deleted.
3093
3094 my $self = shift;
3095 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3096
ffe43484 3097 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3098
3099 # If the range list is empty, return a large value that isn't adjacent
3100 # to any that could be in the range list, for simpler tests
6189eadc 3101 return $MAX_UNICODE_CODEPOINT + 2 unless scalar @{$ranges{$addr}};
99870f4d
KW
3102 return $ranges{$addr}->[0]->start;
3103 }
3104
3105 sub contains {
3106 # Boolean: Is argument in the range list? If so returns $i such that:
3107 # range[$i]->end < $codepoint <= range[$i+1]->end
3108 # which is one beyond what you want; this is so that the 0th range
3109 # doesn't return false
3110 my $self = shift;
3111 my $codepoint = shift;
3112 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3113
99870f4d
KW
3114 my $i = $self->_search_ranges($codepoint);
3115 return 0 unless defined $i;
3116
3117 # The search returns $i, such that
3118 # range[$i-1]->end < $codepoint <= range[$i]->end
3119 # So is in the table if and only iff it is at least the start position
3120 # of range $i.
f998e60c 3121 no overloading;
051df77b 3122 return 0 if $ranges{pack 'J', $self}->[$i]->start > $codepoint;
99870f4d
KW
3123 return $i + 1;
3124 }
3125
2f7a8815
KW
3126 sub containing_range {
3127 # Returns the range object that contains the code point, undef if none
3128
3129 my $self = shift;
3130 my $codepoint = shift;
3131 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3132
3133 my $i = $self->contains($codepoint);
3134 return unless $i;
3135
3136 # contains() returns 1 beyond where we should look
3137 no overloading;
3138 return $ranges{pack 'J', $self}->[$i-1];
3139 }
3140
99870f4d
KW
3141 sub value_of {
3142 # Returns the value associated with the code point, undef if none
3143
3144 my $self = shift;
3145 my $codepoint = shift;
3146 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3147
d69c231b
KW
3148 my $range = $self->containing_range($codepoint);
3149 return unless defined $range;
99870f4d 3150
d69c231b 3151 return $range->value;
99870f4d
KW
3152 }
3153
0a9dbafc
KW
3154 sub type_of {
3155 # Returns the type of the range containing the code point, undef if
3156 # the code point is not in the table
3157
3158 my $self = shift;
3159 my $codepoint = shift;
3160 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3161
3162 my $range = $self->containing_range($codepoint);
3163 return unless defined $range;
3164
3165 return $range->type;
3166 }
3167
99870f4d
KW
3168 sub _search_ranges {
3169 # Find the range in the list which contains a code point, or where it
3170 # should go if were to add it. That is, it returns $i, such that:
3171 # range[$i-1]->end < $codepoint <= range[$i]->end
3172 # Returns undef if no such $i is possible (e.g. at end of table), or
3173 # if there is an error.
3174
3175 my $self = shift;
3176 my $code_point = shift;
3177 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3178
ffe43484 3179 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3180
3181 return if $code_point > $max{$addr};
3182 my $r = $ranges{$addr}; # The current list of ranges
3183 my $range_list_size = scalar @$r;
3184 my $i;
3185
3186 use integer; # want integer division
3187
3188 # Use the cached result as the starting guess for this one, because,
3189 # an experiment on 5.1 showed that 90% of the time the cache was the
3190 # same as the result on the next call (and 7% it was one less).
3191 $i = $_search_ranges_cache{$addr};
3192 $i = 0 if $i >= $range_list_size; # Reset if no longer valid (prob.
3193 # from an intervening deletion
3194 #local $to_trace = 1 if main::DEBUG;
3195 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);
3196 return $i if $code_point <= $r->[$i]->end
3197 && ($i == 0 || $r->[$i-1]->end < $code_point);
3198
3199 # Here the cache doesn't yield the correct $i. Try adding 1.
3200 if ($i < $range_list_size - 1
3201 && $r->[$i]->end < $code_point &&
3202 $code_point <= $r->[$i+1]->end)
3203 {
3204 $i++;
3205 trace "next \$i is correct: $i" if main::DEBUG && $to_trace;
3206 $_search_ranges_cache{$addr} = $i;
3207 return $i;
3208 }
3209
3210 # Here, adding 1 also didn't work. We do a binary search to
3211 # find the correct position, starting with current $i
3212 my $lower = 0;
3213 my $upper = $range_list_size - 1;
3214 while (1) {
3215 trace "top of loop i=$i:", sprintf("%04X", $r->[$lower]->start), "[$lower] .. ", sprintf("%04X", $r->[$i]->start), "[$i] .. ", sprintf("%04X", $r->[$upper]->start), "[$upper]" if main::DEBUG && $to_trace;
3216
3217 if ($code_point <= $r->[$i]->end) {
3218
3219 # Here we have met the upper constraint. We can quit if we
3220 # also meet the lower one.
3221 last if $i == 0 || $r->[$i-1]->end < $code_point;
3222
3223 $upper = $i; # Still too high.
3224
3225 }
3226 else {
3227
3228 # Here, $r[$i]->end < $code_point, so look higher up.
3229 $lower = $i;
3230 }
3231
3232 # Split search domain in half to try again.
3233 my $temp = ($upper + $lower) / 2;
3234
3235 # No point in continuing unless $i changes for next time
3236 # in the loop.
3237 if ($temp == $i) {
3238
3239 # We can't reach the highest element because of the averaging.
3240 # So if one below the upper edge, force it there and try one
3241 # more time.
3242 if ($i == $range_list_size - 2) {
3243
3244 trace "Forcing to upper edge" if main::DEBUG && $to_trace;
3245 $i = $range_list_size - 1;
3246
3247 # Change $lower as well so if fails next time through,
3248 # taking the average will yield the same $i, and we will
3249 # quit with the error message just below.
3250 $lower = $i;
3251 next;
3252 }
3253 Carp::my_carp_bug("$owner_name_of{$addr}Can't find where the range ought to go. No action taken.");
3254 return;
3255 }
3256 $i = $temp;
3257 } # End of while loop
3258
3259 if (main::DEBUG && $to_trace) {
3260 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i;
3261 trace "i= [ $i ]", $r->[$i];
3262 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < $range_list_size - 1;
3263 }
3264
3265 # Here we have found the offset. Cache it as a starting point for the
3266 # next call.
3267 $_search_ranges_cache{$addr} = $i;
3268 return $i;
3269 }
3270
3271 sub _add_delete {
3272 # Add, replace or delete ranges to or from a list. The $type
3273 # parameter gives which:
3274 # '+' => insert or replace a range, returning a list of any changed
3275 # ranges.
3276 # '-' => delete a range, returning a list of any deleted ranges.
3277 #
3278 # The next three parameters give respectively the start, end, and
3279 # value associated with the range. 'value' should be null unless the
3280 # operation is '+';
3281 #
3282 # The range list is kept sorted so that the range with the lowest
3283 # starting position is first in the list, and generally, adjacent
c1739a4a 3284 # ranges with the same values are merged into a single larger one (see
99870f4d
KW
3285 # exceptions below).
3286 #
c1739a4a 3287 # There are more parameters; all are key => value pairs:
99870f4d
KW
3288 # Type gives the type of the value. It is only valid for '+'.
3289 # All ranges have types; if this parameter is omitted, 0 is
3290 # assumed. Ranges with type 0 are assumed to obey the
3291 # Unicode rules for casing, etc; ranges with other types are
3292 # not. Otherwise, the type is arbitrary, for the caller's
3293 # convenience, and looked at only by this routine to keep
3294 # adjacent ranges of different types from being merged into
3295 # a single larger range, and when Replace =>
3296 # $IF_NOT_EQUIVALENT is specified (see just below).
3297 # Replace determines what to do if the range list already contains
3298 # ranges which coincide with all or portions of the input
3299 # range. It is only valid for '+':
3300 # => $NO means that the new value is not to replace
3301 # any existing ones, but any empty gaps of the
3302 # range list coinciding with the input range
3303 # will be filled in with the new value.
3304 # => $UNCONDITIONALLY means to replace the existing values with
3305 # this one unconditionally. However, if the
3306 # new and old values are identical, the
3307 # replacement is skipped to save cycles
3308 # => $IF_NOT_EQUIVALENT means to replace the existing values
3309 # with this one if they are not equivalent.
3310 # Ranges are equivalent if their types are the
c1739a4a 3311 # same, and they are the same string; or if
99870f4d
KW
3312 # both are type 0 ranges, if their Unicode
3313 # standard forms are identical. In this last
3314 # case, the routine chooses the more "modern"
3315 # one to use. This is because some of the
3316 # older files are formatted with values that
3317 # are, for example, ALL CAPs, whereas the
3318 # derived files have a more modern style,
3319 # which looks better. By looking for this
3320 # style when the pre-existing and replacement
3321 # standard forms are the same, we can move to
3322 # the modern style
3323 # => $MULTIPLE means that if this range duplicates an
3324 # existing one, but has a different value,
3325 # don't replace the existing one, but insert
3326 # this, one so that the same range can occur
53d84487
KW
3327 # multiple times. They are stored LIFO, so
3328 # that the final one inserted is the first one
3329 # returned in an ordered search of the table.
99870f4d
KW
3330 # => anything else is the same as => $IF_NOT_EQUIVALENT
3331 #
c1739a4a
KW
3332 # "same value" means identical for non-type-0 ranges, and it means
3333 # having the same standard forms for type-0 ranges.
99870f4d
KW
3334
3335 return Carp::carp_too_few_args(\@_, 5) if main::DEBUG && @_ < 5;
3336
3337 my $self = shift;
3338 my $operation = shift; # '+' for add/replace; '-' for delete;
3339 my $start = shift;
3340 my $end = shift;
3341 my $value = shift;
3342
3343 my %args = @_;
3344
3345 $value = "" if not defined $value; # warning: $value can be "0"
3346
3347 my $replace = delete $args{'Replace'};
3348 $replace = $IF_NOT_EQUIVALENT unless defined $replace;
3349
3350 my $type = delete $args{'Type'};
3351 $type = 0 unless defined $type;
3352
3353 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3354
ffe43484 3355 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3356
3357 if ($operation ne '+' && $operation ne '-') {
3358 Carp::my_carp_bug("$owner_name_of{$addr}First parameter to _add_delete must be '+' or '-'. No action taken.");
3359 return;
3360 }
3361 unless (defined $start && defined $end) {
3362 Carp::my_carp_bug("$owner_name_of{$addr}Undefined start and/or end to _add_delete. No action taken.");
3363 return;
3364 }
3365 unless ($end >= $start) {
3366 Carp::my_carp_bug("$owner_name_of{$addr}End of range (" . sprintf("%04X", $end) . ") must not be before start (" . sprintf("%04X", $start) . "). No action taken.");
3367 return;
3368 }
3369 #local $to_trace = 1 if main::DEBUG;
3370
3371 if ($operation eq '-') {
3372 if ($replace != $IF_NOT_EQUIVALENT) {
3373 Carp::my_carp_bug("$owner_name_of{$addr}Replace => \$IF_NOT_EQUIVALENT is required when deleting a range from a range list. Assuming Replace => \$IF_NOT_EQUIVALENT.");
3374 $replace = $IF_NOT_EQUIVALENT;
3375 }
3376 if ($type) {
3377 Carp::my_carp_bug("$owner_name_of{$addr}Type => 0 is required when deleting a range from a range list. Assuming Type => 0.");
3378 $type = 0;
3379 }
3380 if ($value ne "") {
3381 Carp::my_carp_bug("$owner_name_of{$addr}Value => \"\" is required when deleting a range from a range list. Assuming Value => \"\".");
3382 $value = "";
3383 }
3384 }
3385
3386 my $r = $ranges{$addr}; # The current list of ranges
3387 my $range_list_size = scalar @$r; # And its size
3388 my $max = $max{$addr}; # The current high code point in
3389 # the list of ranges
3390
3391 # Do a special case requiring fewer machine cycles when the new range
3392 # starts after the current highest point. The Unicode input data is
3393 # structured so this is common.
3394 if ($start > $max) {
3395
3396 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) type=$type" if main::DEBUG && $to_trace;
3397 return if $operation eq '-'; # Deleting a non-existing range is a
3398 # no-op
3399
3400 # If the new range doesn't logically extend the current final one
3401 # in the range list, create a new range at the end of the range
3402 # list. (max cleverly is initialized to a negative number not
3403 # adjacent to 0 if the range list is empty, so even adding a range
3404 # to an empty range list starting at 0 will have this 'if'
3405 # succeed.)
3406 if ($start > $max + 1 # non-adjacent means can't extend.
3407 || @{$r}[-1]->value ne $value # values differ, can't extend.
3408 || @{$r}[-1]->type != $type # types differ, can't extend.
3409 ) {
3410 push @$r, Range->new($start, $end,
3411 Value => $value,
3412 Type => $type);
3413 }
3414 else {
3415
3416 # Here, the new range starts just after the current highest in
3417 # the range list, and they have the same type and value.
3418 # Extend the current range to incorporate the new one.
3419 @{$r}[-1]->set_end($end);
3420 }
3421
3422 # This becomes the new maximum.
3423 $max{$addr} = $end;
3424
3425 return;
3426 }
3427 #local $to_trace = 0 if main::DEBUG;
3428
3429 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) replace=$replace" if main::DEBUG && $to_trace;
3430
3431 # Here, the input range isn't after the whole rest of the range list.
3432 # Most likely 'splice' will be needed. The rest of the routine finds
3433 # the needed splice parameters, and if necessary, does the splice.
3434 # First, find the offset parameter needed by the splice function for
3435 # the input range. Note that the input range may span multiple
3436 # existing ones, but we'll worry about that later. For now, just find
3437 # the beginning. If the input range is to be inserted starting in a
3438 # position not currently in the range list, it must (obviously) come
3439 # just after the range below it, and just before the range above it.
3440 # Slightly less obviously, it will occupy the position currently
3441 # occupied by the range that is to come after it. More formally, we
3442 # are looking for the position, $i, in the array of ranges, such that:
3443 #
3444 # r[$i-1]->start <= r[$i-1]->end < $start < r[$i]->start <= r[$i]->end
3445 #
3446 # (The ordered relationships within existing ranges are also shown in
3447 # the equation above). However, if the start of the input range is
3448 # within an existing range, the splice offset should point to that
3449 # existing range's position in the list; that is $i satisfies a
3450 # somewhat different equation, namely:
3451 #
3452 #r[$i-1]->start <= r[$i-1]->end < r[$i]->start <= $start <= r[$i]->end
3453 #
3454 # More briefly, $start can come before or after r[$i]->start, and at
3455 # this point, we don't know which it will be. However, these
3456 # two equations share these constraints:
3457 #
3458 # r[$i-1]->end < $start <= r[$i]->end
3459 #
3460 # And that is good enough to find $i.
3461
3462 my $i = $self->_search_ranges($start);
3463 if (! defined $i) {
3464 Carp::my_carp_bug("Searching $self for range beginning with $start unexpectedly returned undefined. Operation '$operation' not performed");
3465 return;
3466 }
3467
3468 # The search function returns $i such that:
3469 #
3470 # r[$i-1]->end < $start <= r[$i]->end
3471 #
3472 # That means that $i points to the first range in the range list
3473 # that could possibly be affected by this operation. We still don't
3474 # know if the start of the input range is within r[$i], or if it
3475 # points to empty space between r[$i-1] and r[$i].
3476 trace "[$i] is the beginning splice point. Existing range there is ", $r->[$i] if main::DEBUG && $to_trace;
3477
3478 # Special case the insertion of data that is not to replace any
3479 # existing data.
3480 if ($replace == $NO) { # If $NO, has to be operation '+'
3481 #local $to_trace = 1 if main::DEBUG;
3482 trace "Doesn't replace" if main::DEBUG && $to_trace;
3483
3484 # Here, the new range is to take effect only on those code points
3485 # that aren't already in an existing range. This can be done by
3486 # looking through the existing range list and finding the gaps in
3487 # the ranges that this new range affects, and then calling this
3488 # function recursively on each of those gaps, leaving untouched
3489 # anything already in the list. Gather up a list of the changed
3490 # gaps first so that changes to the internal state as new ranges
3491 # are added won't be a problem.
3492 my @gap_list;
3493
3494 # First, if the starting point of the input range is outside an
3495 # existing one, there is a gap from there to the beginning of the
3496 # existing range -- add a span to fill the part that this new
3497 # range occupies
3498 if ($start < $r->[$i]->start) {
3499 push @gap_list, Range->new($start,
3500 main::min($end,
3501 $r->[$i]->start - 1),
3502 Type => $type);
3503 trace "gap before $r->[$i] [$i], will add", $gap_list[-1] if main::DEBUG && $to_trace;
3504 }
3505
3506 # Then look through the range list for other gaps until we reach
3507 # the highest range affected by the input one.
3508 my $j;
3509 for ($j = $i+1; $j < $range_list_size; $j++) {
3510 trace "j=[$j]", $r->[$j] if main::DEBUG && $to_trace;
3511 last if $end < $r->[$j]->start;
3512
3513 # If there is a gap between when this range starts and the
3514 # previous one ends, add a span to fill it. Note that just
3515 # because there are two ranges doesn't mean there is a
3516 # non-zero gap between them. It could be that they have
3517 # different values or types
3518 if ($r->[$j-1]->end + 1 != $r->[$j]->start) {
3519 push @gap_list,
3520 Range->new($r->[$j-1]->end + 1,
3521 $r->[$j]->start - 1,
3522 Type => $type);
3523 trace "gap between $r->[$j-1] and $r->[$j] [$j], will add: $gap_list[-1]" if main::DEBUG && $to_trace;
3524 }
3525 }
3526
3527 # Here, we have either found an existing range in the range list,
3528 # beyond the area affected by the input one, or we fell off the
3529 # end of the loop because the input range affects the whole rest
3530 # of the range list. In either case, $j is 1 higher than the
3531 # highest affected range. If $j == $i, it means that there are no
3532 # affected ranges, that the entire insertion is in the gap between
3533 # r[$i-1], and r[$i], which we already have taken care of before
3534 # the loop.
3535 # On the other hand, if there are affected ranges, it might be
3536 # that there is a gap that needs filling after the final such
3537 # range to the end of the input range
3538 if ($r->[$j-1]->end < $end) {
3539 push @gap_list, Range->new(main::max($start,
3540 $r->[$j-1]->end + 1),
3541 $end,
3542 Type => $type);
3543 trace "gap after $r->[$j-1], will add $gap_list[-1]" if main::DEBUG && $to_trace;
3544 }
3545
3546 # Call recursively to fill in all the gaps.
3547 foreach my $gap (@gap_list) {
3548 $self->_add_delete($operation,
3549 $gap->start,
3550 $gap->end,
3551 $value,
3552 Type => $type);
3553 }
3554
3555 return;
3556 }
3557
53d84487
KW
3558 # Here, we have taken care of the case where $replace is $NO.
3559 # Remember that here, r[$i-1]->end < $start <= r[$i]->end
3560 # If inserting a multiple record, this is where it goes, before the
3561 # first (if any) existing one. This implies an insertion, and no
3562 # change to any existing ranges. Note that $i can be -1 if this new
3563 # range doesn't actually duplicate any existing, and comes at the
3564 # beginning of the list.
3565 if ($replace == $MULTIPLE) {
3566
3567 if ($start != $end) {
3568 Carp::my_carp_bug("$owner_name_of{$addr}Can't cope with adding a multiple record when the range ($start..$end) contains more than one code point. No action taken.");
3569 return;
3570 }
3571
3572 # Don't add an exact duplicate, as it isn't really a multiple
3573 if ($end >= $r->[$i]->start) {
1f6798c4
KW
3574 my $existing_value = $r->[$i]->value;
3575 my $existing_type = $r->[$i]->type;
3576 return if $value eq $existing_value && $type eq $existing_type;
3577
3578 # If the multiple value is part of an existing range, we want
3579 # to split up that range, so that only the single code point
3580 # is affected. To do this, we first call ourselves
3581 # recursively to delete that code point from the table, having
3582 # preserved its current data above. Then we call ourselves
3583 # recursively again to add the new multiple, which we know by
3584 # the test just above is different than the current code
3585 # point's value, so it will become a range containing a single
3586 # code point: just itself. Finally, we add back in the
3587 # pre-existing code point, which will again be a single code
3588 # point range. Because 'i' likely will have changed as a
3589 # result of these operations, we can't just continue on, but
3590 # do this operation recursively as well.
53d84487 3591 if ($r->[$i]->start != $r->[$i]->end) {
1f6798c4
KW
3592 $self->_add_delete('-', $start, $end, "");
3593 $self->_add_delete('+', $start, $end, $value, Type => $type);
3594 return $self->_add_delete('+', $start, $end, $existing_value, Type => $existing_type, Replace => $MULTIPLE);
53d84487 3595 }
53d84487
KW
3596 }
3597
3598 trace "Adding multiple record at $i with $start..$end, $value" if main::DEBUG && $to_trace;
3599 my @return = splice @$r,
3600 $i,
3601 0,
3602 Range->new($start,
3603 $end,
3604 Value => $value,
3605 Type => $type);
3606 if (main::DEBUG && $to_trace) {
3607 trace "After splice:";
3608 trace 'i-2=[', $i-2, ']', $r->[$i-2] if $i >= 2;
3609 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i >= 1;
3610 trace "i =[", $i, "]", $r->[$i] if $i >= 0;
3611 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < @$r - 1;
3612 trace 'i+2=[', $i+2, ']', $r->[$i+2] if $i < @$r - 2;
3613 trace 'i+3=[', $i+3, ']', $r->[$i+3] if $i < @$r - 3;
3614 }
3615 return @return;
3616 }
3617
3618 # Here, we have taken care of $NO and $MULTIPLE replaces. This leaves
3619 # delete, insert, and replace either unconditionally or if not
3620 # equivalent. $i still points to the first potential affected range.
3621 # Now find the highest range affected, which will determine the length
3622 # parameter to splice. (The input range can span multiple existing
3623 # ones.) If this isn't a deletion, while we are looking through the
3624 # range list, see also if this is a replacement rather than a clean
3625 # insertion; that is if it will change the values of at least one
3626 # existing range. Start off assuming it is an insert, until find it
3627 # isn't.
3628 my $clean_insert = $operation eq '+';
99870f4d
KW
3629 my $j; # This will point to the highest affected range
3630
3631 # For non-zero types, the standard form is the value itself;
3632 my $standard_form = ($type) ? $value : main::standardize($value);
3633
3634 for ($j = $i; $j < $range_list_size; $j++) {
3635 trace "Looking for highest affected range; the one at $j is ", $r->[$j] if main::DEBUG && $to_trace;
3636
3637 # If find a range that it doesn't overlap into, we can stop
3638 # searching
3639 last if $end < $r->[$j]->start;
3640
969a34cc
KW
3641 # Here, overlaps the range at $j. If the values don't match,
3642 # and so far we think this is a clean insertion, it becomes a
3643 # non-clean insertion, i.e., a 'change' or 'replace' instead.
3644 if ($clean_insert) {
99870f4d 3645 if ($r->[$j]->standard_form ne $standard_form) {
969a34cc 3646 $clean_insert = 0;
56343c78
KW
3647 if ($replace == $CROAK) {
3648 main::croak("The range to add "
3649 . sprintf("%04X", $start)
3650 . '-'
3651 . sprintf("%04X", $end)
3652 . " with value '$value' overlaps an existing range $r->[$j]");
3653 }
99870f4d
KW
3654 }
3655 else {
3656
3657 # Here, the two values are essentially the same. If the
3658 # two are actually identical, replacing wouldn't change
3659 # anything so skip it.
3660 my $pre_existing = $r->[$j]->value;
3661 if ($pre_existing ne $value) {
3662
3663 # Here the new and old standardized values are the
3664 # same, but the non-standardized values aren't. If
3665 # replacing unconditionally, then replace
3666 if( $replace == $UNCONDITIONALLY) {
969a34cc 3667 $clean_insert = 0;
99870f4d
KW
3668 }
3669 else {
3670
3671 # Here, are replacing conditionally. Decide to
3672 # replace or not based on which appears to look
3673 # the "nicest". If one is mixed case and the
3674 # other isn't, choose the mixed case one.
3675 my $new_mixed = $value =~ /[A-Z]/
3676 && $value =~ /[a-z]/;
3677 my $old_mixed = $pre_existing =~ /[A-Z]/
3678 && $pre_existing =~ /[a-z]/;
3679
3680 if ($old_mixed != $new_mixed) {
969a34cc 3681 $clean_insert = 0 if $new_mixed;
99870f4d 3682 if (main::DEBUG && $to_trace) {
969a34cc
KW
3683 if ($clean_insert) {
3684 trace "Retaining $pre_existing over $value";
99870f4d
KW
3685 }
3686 else {
969a34cc 3687 trace "Replacing $pre_existing with $value";
99870f4d
KW
3688 }
3689 }
3690 }
3691 else {
3692
3693 # Here casing wasn't different between the two.
3694 # If one has hyphens or underscores and the
3695 # other doesn't, choose the one with the
3696 # punctuation.
3697 my $new_punct = $value =~ /[-_]/;
3698 my $old_punct = $pre_existing =~ /[-_]/;
3699
3700 if ($old_punct != $new_punct) {
969a34cc 3701 $clean_insert = 0 if $new_punct;
99870f4d 3702 if (main::DEBUG && $to_trace) {
969a34cc
KW
3703 if ($clean_insert) {
3704 trace "Retaining $pre_existing over $value";
99870f4d
KW
3705 }
3706 else {
969a34cc 3707 trace "Replacing $pre_existing with $value";
99870f4d
KW
3708 }
3709 }
3710 } # else existing one is just as "good";
3711 # retain it to save cycles.
3712 }
3713 }
3714 }
3715 }
3716 }
3717 } # End of loop looking for highest affected range.
3718
3719 # Here, $j points to one beyond the highest range that this insertion
3720 # affects (hence to beyond the range list if that range is the final
3721 # one in the range list).
3722
3723 # The splice length is all the affected ranges. Get it before
3724 # subtracting, for efficiency, so we don't have to later add 1.
3725 my $length = $j - $i;
3726
3727 $j--; # $j now points to the highest affected range.
3728 trace "Final affected range is $j: $r->[$j]" if main::DEBUG && $to_trace;
3729
99870f4d
KW
3730 # Here, have taken care of $NO and $MULTIPLE replaces.
3731 # $j points to the highest affected range. But it can be < $i or even
3732 # -1. These happen only if the insertion is entirely in the gap
3733 # between r[$i-1] and r[$i]. Here's why: j < i means that the j loop
3734 # above exited first time through with $end < $r->[$i]->start. (And
3735 # then we subtracted one from j) This implies also that $start <
3736 # $r->[$i]->start, but we know from above that $r->[$i-1]->end <
3737 # $start, so the entire input range is in the gap.
3738 if ($j < $i) {
3739
3740 # Here the entire input range is in the gap before $i.
3741
3742 if (main::DEBUG && $to_trace) {
3743 if ($i) {
3744 trace "Entire range is between $r->[$i-1] and $r->[$i]";
3745 }
3746 else {
3747 trace "Entire range is before $r->[$i]";
3748 }
3749 }
3750 return if $operation ne '+'; # Deletion of a non-existent range is
3751 # a no-op
3752 }
3753 else {
3754
969a34cc
KW
3755 # Here part of the input range is not in the gap before $i. Thus,
3756 # there is at least one affected one, and $j points to the highest
3757 # such one.
99870f4d
KW
3758
3759 # At this point, here is the situation:
3760 # This is not an insertion of a multiple, nor of tentative ($NO)
3761 # data.
3762 # $i points to the first element in the current range list that
3763 # may be affected by this operation. In fact, we know
3764 # that the range at $i is affected because we are in
3765 # the else branch of this 'if'
3766 # $j points to the highest affected range.
3767 # In other words,
3768 # r[$i-1]->end < $start <= r[$i]->end
3769 # And:
3770 # r[$i-1]->end < $start <= $end <= r[$j]->end
3771 #
3772 # Also:
969a34cc
KW
3773 # $clean_insert is a boolean which is set true if and only if
3774 # this is a "clean insertion", i.e., not a change nor a
3775 # deletion (multiple was handled above).
99870f4d
KW
3776
3777 # We now have enough information to decide if this call is a no-op
969a34cc
KW
3778 # or not. It is a no-op if this is an insertion of already
3779 # existing data.
99870f4d 3780
969a34cc 3781 if (main::DEBUG && $to_trace && $clean_insert
99870f4d
KW
3782 && $i == $j
3783 && $start >= $r->[$i]->start)
3784 {
3785 trace "no-op";
3786 }
969a34cc 3787 return if $clean_insert
99870f4d
KW
3788 && $i == $j # more than one affected range => not no-op
3789
3790 # Here, r[$i-1]->end < $start <= $end <= r[$i]->end
3791 # Further, $start and/or $end is >= r[$i]->start
3792 # The test below hence guarantees that
3793 # r[$i]->start < $start <= $end <= r[$i]->end
3794 # This means the input range is contained entirely in
3795 # the one at $i, so is a no-op
3796 && $start >= $r->[$i]->start;
3797 }
3798
3799 # Here, we know that some action will have to be taken. We have
3800 # calculated the offset and length (though adjustments may be needed)
3801 # for the splice. Now start constructing the replacement list.
3802 my @replacement;
3803 my $splice_start = $i;
3804
3805 my $extends_below;
3806 my $extends_above;
3807
3808 # See if should extend any adjacent ranges.
3809 if ($operation eq '-') { # Don't extend deletions
3810 $extends_below = $extends_above = 0;
3811 }
3812 else { # Here, should extend any adjacent ranges. See if there are
3813 # any.
3814 $extends_below = ($i > 0
3815 # can't extend unless adjacent
3816 && $r->[$i-1]->end == $start -1
3817 # can't extend unless are same standard value
3818 && $r->[$i-1]->standard_form eq $standard_form
3819 # can't extend unless share type
3820 && $r->[$i-1]->type == $type);
3821 $extends_above = ($j+1 < $range_list_size
3822 && $r->[$j+1]->start == $end +1
3823 && $r->[$j+1]->standard_form eq $standard_form
23822bda 3824 && $r->[$j+1]->type == $type);
99870f4d
KW
3825 }
3826 if ($extends_below && $extends_above) { # Adds to both
3827 $splice_start--; # start replace at element below
3828 $length += 2; # will replace on both sides
3829 trace "Extends both below and above ranges" if main::DEBUG && $to_trace;
3830
3831 # The result will fill in any gap, replacing both sides, and
3832 # create one large range.
3833 @replacement = Range->new($r->[$i-1]->start,
3834 $r->[$j+1]->end,
3835 Value => $value,
3836 Type => $type);
3837 }
3838 else {
3839
3840 # Here we know that the result won't just be the conglomeration of
3841 # a new range with both its adjacent neighbors. But it could
3842 # extend one of them.
3843
3844 if ($extends_below) {
3845
3846 # Here the new element adds to the one below, but not to the
3847 # one above. If inserting, and only to that one range, can
3848 # just change its ending to include the new one.
969a34cc 3849 if ($length == 0 && $clean_insert) {
99870f4d
KW
3850 $r->[$i-1]->set_end($end);
3851 trace "inserted range extends range to below so it is now $r->[$i-1]" if main::DEBUG && $to_trace;
3852 return;
3853 }
3854 else {
3855 trace "Changing inserted range to start at ", sprintf("%04X", $r->[$i-1]->start), " instead of ", sprintf("%04X", $start) if main::DEBUG && $to_trace;
3856 $splice_start--; # start replace at element below
3857 $length++; # will replace the element below
3858 $start = $r->[$i-1]->start;
3859 }
3860 }
3861 elsif ($extends_above) {
3862
3863 # Here the new element adds to the one above, but not below.
3864 # Mirror the code above
969a34cc 3865 if ($length == 0 && $clean_insert) {
99870f4d
KW
3866 $r->[$j+1]->set_start($start);
3867 trace "inserted range extends range to above so it is now $r->[$j+1]" if main::DEBUG && $to_trace;
3868 return;
3869 }
3870 else {
3871 trace "Changing inserted range to end at ", sprintf("%04X", $r->[$j+1]->end), " instead of ", sprintf("%04X", $end) if main::DEBUG && $to_trace;
3872 $length++; # will replace the element above
3873 $end = $r->[$j+1]->end;
3874 }
3875 }
3876
3877 trace "Range at $i is $r->[$i]" if main::DEBUG && $to_trace;
3878
3879 # Finally, here we know there will have to be a splice.