This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
mktables: Generate empty Name.pm
[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
KW
853# Properties that this program ignores.
854my @unimplemented_properties = (
855'Unicode_Radical_Stroke' # Remove if changing to handle this one.
856);
d73e5302 857
99870f4d
KW
858# There are several types of obsolete properties defined by Unicode. These
859# must be hand-edited for every new Unicode release.
860my %why_deprecated; # Generates a deprecated warning message if used.
861my %why_stabilized; # Documentation only
862my %why_obsolete; # Documentation only
863
864{ # Closure
865 my $simple = 'Perl uses the more complete version of this property';
866 my $unihan = 'Unihan properties are by default not enabled in the Perl core. Instead use CPAN: Unicode::Unihan';
867
868 my $other_properties = 'other properties';
869 my $contributory = "Used by Unicode internally for generating $other_properties and not intended to be used stand-alone";
5d294d41 870 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
871
872 %why_deprecated = (
5f7264c7 873 'Grapheme_Link' => 'Deprecated by Unicode: Duplicates ccc=vr (Canonical_Combining_Class=Virama)',
99870f4d
KW
874 'Jamo_Short_Name' => $contributory,
875 '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',
876 'Other_Alphabetic' => $contributory,
877 'Other_Default_Ignorable_Code_Point' => $contributory,
878 'Other_Grapheme_Extend' => $contributory,
879 'Other_ID_Continue' => $contributory,
880 'Other_ID_Start' => $contributory,
881 'Other_Lowercase' => $contributory,
882 'Other_Math' => $contributory,
883 'Other_Uppercase' => $contributory,
e22aaf5c
KW
884 'Expands_On_NFC' => $why_no_expand,
885 'Expands_On_NFD' => $why_no_expand,
886 'Expands_On_NFKC' => $why_no_expand,
887 'Expands_On_NFKD' => $why_no_expand,
99870f4d
KW
888 );
889
890 %why_suppressed = (
5f7264c7 891 # There is a lib/unicore/Decomposition.pl (used by Normalize.pm) which
99870f4d
KW
892 # contains the same information, but without the algorithmically
893 # determinable Hangul syllables'. This file is not published, so it's
894 # existence is not noted in the comment.
895 'Decomposition_Mapping' => 'Accessible via Unicode::Normalize',
896
897 'ISO_Comment' => 'Apparently no demand for it, but can access it through Unicode::UCD::charinfo. Obsoleted, and code points for it removed in Unicode 5.2',
99870f4d
KW
898
899 'Simple_Case_Folding' => "$simple. Can access this through Unicode::UCD::casefold",
900 'Simple_Lowercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo",
901 'Simple_Titlecase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo",
902 'Simple_Uppercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo",
903
904 'Name' => "Accessible via 'use charnames;'",
905 'Name_Alias' => "Accessible via 'use charnames;'",
906
5f7264c7 907 FC_NFKC_Closure => 'Supplanted in usage by NFKC_Casefold; otherwise not useful',
99870f4d
KW
908 );
909
910 # The following are suppressed because they were made contributory or
911 # deprecated by Unicode before Perl ever thought about supporting them.
e22aaf5c
KW
912 foreach my $property ('Jamo_Short_Name',
913 'Grapheme_Link',
914 'Expands_On_NFC',
915 'Expands_On_NFD',
916 'Expands_On_NFKC',
917 'Expands_On_NFKD'
918 ) {
99870f4d
KW
919 $why_suppressed{$property} = $why_deprecated{$property};
920 }
cf25bb62 921
99870f4d
KW
922 # Customize the message for all the 'Other_' properties
923 foreach my $property (keys %why_deprecated) {
924 next if (my $main_property = $property) !~ s/^Other_//;
925 $why_deprecated{$property} =~ s/$other_properties/the $main_property property (which should be used instead)/;
926 }
927}
928
929if ($v_version ge 4.0.0) {
930 $why_stabilized{'Hyphen'} = 'Use the Line_Break property instead; see www.unicode.org/reports/tr14';
5f7264c7
KW
931 if ($v_version ge 6.0.0) {
932 $why_deprecated{'Hyphen'} = 'Supplanted by Line_Break property values; see www.unicode.org/reports/tr14';
933 }
99870f4d 934}
5f7264c7 935if ($v_version ge 5.2.0 && $v_version lt 6.0.0) {
99870f4d 936 $why_obsolete{'ISO_Comment'} = 'Code points for it have been removed';
5f7264c7 937 if ($v_version ge 6.0.0) {
63f74647 938 $why_deprecated{'ISO_Comment'} = 'No longer needed for Unicode\'s internal chart generation; otherwise not useful, and code points for it have been removed';
5f7264c7 939 }
99870f4d
KW
940}
941
942# Probably obsolete forever
943if ($v_version ge v4.1.0) {
82aed44a
KW
944 $why_suppressed{'Script=Katakana_Or_Hiragana'} = 'Obsolete. All code points previously matched by this have been moved to "Script=Common".';
945}
946if ($v_version ge v6.0.0) {
2b352efd
KW
947 $why_suppressed{'Script=Katakana_Or_Hiragana'} .= ' Consider instead using "Script_Extensions=Katakana" or "Script_Extensions=Hiragana (or both)"';
948 $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
949}
950
951# This program can create files for enumerated-like properties, such as
952# 'Numeric_Type'. This file would be the same format as for a string
953# property, with a mapping from code point to its value, so you could look up,
954# for example, the script a code point is in. But no one so far wants this
955# mapping, or they have found another way to get it since this is a new
956# feature. So no file is generated except if it is in this list.
957my @output_mapped_properties = split "\n", <<END;
958END
959
c12f2655
KW
960# If you are using the Unihan database in a Unicode version before 5.2, you
961# need to add the properties that you want to extract from it to this table.
962# For your convenience, the properties in the 6.0 PropertyAliases.txt file are
963# listed, commented out
99870f4d
KW
964my @cjk_properties = split "\n", <<'END';
965#cjkAccountingNumeric; kAccountingNumeric
966#cjkOtherNumeric; kOtherNumeric
967#cjkPrimaryNumeric; kPrimaryNumeric
968#cjkCompatibilityVariant; kCompatibilityVariant
969#cjkIICore ; kIICore
970#cjkIRG_GSource; kIRG_GSource
971#cjkIRG_HSource; kIRG_HSource
972#cjkIRG_JSource; kIRG_JSource
973#cjkIRG_KPSource; kIRG_KPSource
974#cjkIRG_KSource; kIRG_KSource
975#cjkIRG_TSource; kIRG_TSource
976#cjkIRG_USource; kIRG_USource
977#cjkIRG_VSource; kIRG_VSource
978#cjkRSUnicode; kRSUnicode ; Unicode_Radical_Stroke; URS
979END
980
981# Similarly for the property values. For your convenience, the lines in the
5f7264c7 982# 6.0 PropertyAliases.txt file are listed. Just remove the first BUT NOT both
c12f2655 983# '#' marks (for Unicode versions before 5.2)
99870f4d
KW
984my @cjk_property_values = split "\n", <<'END';
985## @missing: 0000..10FFFF; cjkAccountingNumeric; NaN
986## @missing: 0000..10FFFF; cjkCompatibilityVariant; <code point>
987## @missing: 0000..10FFFF; cjkIICore; <none>
988## @missing: 0000..10FFFF; cjkIRG_GSource; <none>
989## @missing: 0000..10FFFF; cjkIRG_HSource; <none>
990## @missing: 0000..10FFFF; cjkIRG_JSource; <none>
991## @missing: 0000..10FFFF; cjkIRG_KPSource; <none>
992## @missing: 0000..10FFFF; cjkIRG_KSource; <none>
993## @missing: 0000..10FFFF; cjkIRG_TSource; <none>
994## @missing: 0000..10FFFF; cjkIRG_USource; <none>
995## @missing: 0000..10FFFF; cjkIRG_VSource; <none>
996## @missing: 0000..10FFFF; cjkOtherNumeric; NaN
997## @missing: 0000..10FFFF; cjkPrimaryNumeric; NaN
998## @missing: 0000..10FFFF; cjkRSUnicode; <none>
999END
1000
1001# The input files don't list every code point. Those not listed are to be
1002# defaulted to some value. Below are hard-coded what those values are for
1003# non-binary properties as of 5.1. Starting in 5.0, there are
1004# machine-parsable comment lines in the files the give the defaults; so this
1005# list shouldn't have to be extended. The claim is that all missing entries
1006# for binary properties will default to 'N'. Unicode tried to change that in
1007# 5.2, but the beta period produced enough protest that they backed off.
1008#
1009# The defaults for the fields that appear in UnicodeData.txt in this hash must
1010# be in the form that it expects. The others may be synonyms.
1011my $CODE_POINT = '<code point>';
1012my %default_mapping = (
1013 Age => "Unassigned",
1014 # Bidi_Class => Complicated; set in code
1015 Bidi_Mirroring_Glyph => "",
1016 Block => 'No_Block',
1017 Canonical_Combining_Class => 0,
1018 Case_Folding => $CODE_POINT,
1019 Decomposition_Mapping => $CODE_POINT,
1020 Decomposition_Type => 'None',
1021 East_Asian_Width => "Neutral",
1022 FC_NFKC_Closure => $CODE_POINT,
1023 General_Category => 'Cn',
1024 Grapheme_Cluster_Break => 'Other',
1025 Hangul_Syllable_Type => 'NA',
1026 ISO_Comment => "",
1027 Jamo_Short_Name => "",
1028 Joining_Group => "No_Joining_Group",
1029 # Joining_Type => Complicated; set in code
1030 kIICore => 'N', # Is converted to binary
1031 #Line_Break => Complicated; set in code
1032 Lowercase_Mapping => $CODE_POINT,
1033 Name => "",
1034 Name_Alias => "",
1035 NFC_QC => 'Yes',
1036 NFD_QC => 'Yes',
1037 NFKC_QC => 'Yes',
1038 NFKD_QC => 'Yes',
1039 Numeric_Type => 'None',
1040 Numeric_Value => 'NaN',
1041 Script => ($v_version le 4.1.0) ? 'Common' : 'Unknown',
1042 Sentence_Break => 'Other',
1043 Simple_Case_Folding => $CODE_POINT,
1044 Simple_Lowercase_Mapping => $CODE_POINT,
1045 Simple_Titlecase_Mapping => $CODE_POINT,
1046 Simple_Uppercase_Mapping => $CODE_POINT,
1047 Titlecase_Mapping => $CODE_POINT,
1048 Unicode_1_Name => "",
1049 Unicode_Radical_Stroke => "",
1050 Uppercase_Mapping => $CODE_POINT,
1051 Word_Break => 'Other',
1052);
1053
1054# Below are files that Unicode furnishes, but this program ignores, and why
1055my %ignored_files = (
1056 'CJKRadicals.txt' => 'Unihan data',
1057 'Index.txt' => 'An index, not actual data',
1058 'NamedSqProv.txt' => 'Not officially part of the Unicode standard; Append it to NamedSequences.txt if you want to process the contents.',
1059 'NamesList.txt' => 'Just adds commentary',
1060 'NormalizationCorrections.txt' => 'Data is already in other files.',
1061 'Props.txt' => 'Adds nothing to PropList.txt; only in very early releases',
1062 'ReadMe.txt' => 'Just comments',
1063 'README.TXT' => 'Just comments',
1064 'StandardizedVariants.txt' => 'Only for glyph changes, not a Unicode character property. Does not fit into current scheme where one code point is mapped',
5f7264c7
KW
1065 'EmojiSources.txt' => 'Not of general utility: for Japanese legacy cell-phone applications',
1066 'IndicMatraCategory.txt' => 'Provisional',
1067 'IndicSyllabicCategory.txt' => 'Provisional',
99870f4d
KW
1068);
1069
678f13d5 1070### End of externally interesting definitions, except for @input_file_objects
99870f4d
KW
1071
1072my $HEADER=<<"EOF";
1073# !!!!!!! DO NOT EDIT THIS FILE !!!!!!!
3df51b85
KW
1074# This file is machine-generated by $0 from the Unicode
1075# database, Version $string_version. Any changes made here will be lost!
cf25bb62
JH
1076EOF
1077
b6922eda 1078my $INTERNAL_ONLY=<<"EOF";
99870f4d
KW
1079
1080# !!!!!!! INTERNAL PERL USE ONLY !!!!!!!
fac53429
KW
1081# This file is for internal use by core Perl only. The format and even the
1082# name or existence of this file are subject to change without notice. Don't
1083# use it directly.
99870f4d
KW
1084EOF
1085
1086my $DEVELOPMENT_ONLY=<<"EOF";
1087# !!!!!!! DEVELOPMENT USE ONLY !!!!!!!
1088# This file contains information artificially constrained to code points
1089# present in Unicode release $string_compare_versions.
1090# IT CANNOT BE RELIED ON. It is for use during development only and should
23e33b60 1091# not be used for production.
b6922eda
KW
1092
1093EOF
1094
6189eadc
KW
1095my $MAX_UNICODE_CODEPOINT_STRING = "10FFFF";
1096my $MAX_UNICODE_CODEPOINT = hex $MAX_UNICODE_CODEPOINT_STRING;
1097my $MAX_UNICODE_CODEPOINTS = $MAX_UNICODE_CODEPOINT + 1;
99870f4d
KW
1098
1099# Matches legal code point. 4-6 hex numbers, If there are 6, the first
1100# two must be 10; if there are 5, the first must not be a 0. Written this way
8c32d378
KW
1101# to decrease backtracking. The first one allows the code point to be at the
1102# end of a word, but to work properly, the word shouldn't end with a valid hex
1103# character. The second one won't match a code point at the end of a word,
1104# and doesn't have the run-on issue
1105my $run_on_code_point_re =
1106 qr/ (?: 10[0-9A-F]{4} | [1-9A-F][0-9A-F]{4} | [0-9A-F]{4} ) \b/x;
1107my $code_point_re = qr/\b$run_on_code_point_re/;
99870f4d
KW
1108
1109# This matches the beginning of the line in the Unicode db files that give the
1110# defaults for code points not listed (i.e., missing) in the file. The code
1111# depends on this ending with a semi-colon, so it can assume it is a valid
1112# field when the line is split() by semi-colons
1113my $missing_defaults_prefix =
6189eadc 1114 qr/^#\s+\@missing:\s+0000\.\.$MAX_UNICODE_CODEPOINT_STRING\s*;/;
99870f4d
KW
1115
1116# Property types. Unicode has more types, but these are sufficient for our
1117# purposes.
1118my $UNKNOWN = -1; # initialized to illegal value
1119my $NON_STRING = 1; # Either binary or enum
1120my $BINARY = 2;
06f26c45
KW
1121my $FORCED_BINARY = 3; # Not a binary property, but, besides its normal
1122 # tables, additional true and false tables are
1123 # generated so that false is anything matching the
1124 # default value, and true is everything else.
1125my $ENUM = 4; # Include catalog
1126my $STRING = 5; # Anything else: string or misc
99870f4d
KW
1127
1128# Some input files have lines that give default values for code points not
1129# contained in the file. Sometimes these should be ignored.
1130my $NO_DEFAULTS = 0; # Must evaluate to false
1131my $NOT_IGNORED = 1;
1132my $IGNORED = 2;
1133
1134# Range types. Each range has a type. Most ranges are type 0, for normal,
1135# and will appear in the main body of the tables in the output files, but
1136# there are other types of ranges as well, listed below, that are specially
1137# handled. There are pseudo-types as well that will never be stored as a
1138# type, but will affect the calculation of the type.
1139
1140# 0 is for normal, non-specials
1141my $MULTI_CP = 1; # Sequence of more than code point
1142my $HANGUL_SYLLABLE = 2;
1143my $CP_IN_NAME = 3; # The NAME contains the code point appended to it.
1144my $NULL = 4; # The map is to the null string; utf8.c can't
1145 # handle these, nor is there an accepted syntax
1146 # for them in \p{} constructs
f86864ac 1147my $COMPUTE_NO_MULTI_CP = 5; # Pseudo-type; means that ranges that would
99870f4d
KW
1148 # otherwise be $MULTI_CP type are instead type 0
1149
1150# process_generic_property_file() can accept certain overrides in its input.
1151# Each of these must begin AND end with $CMD_DELIM.
1152my $CMD_DELIM = "\a";
1153my $REPLACE_CMD = 'replace'; # Override the Replace
1154my $MAP_TYPE_CMD = 'map_type'; # Override the Type
1155
1156my $NO = 0;
1157my $YES = 1;
1158
1159# Values for the Replace argument to add_range.
1160# $NO # Don't replace; add only the code points not
1161 # already present.
1162my $IF_NOT_EQUIVALENT = 1; # Replace only under certain conditions; details in
1163 # the comments at the subroutine definition.
1164my $UNCONDITIONALLY = 2; # Replace without conditions.
1165my $MULTIPLE = 4; # Don't replace, but add a duplicate record if
1166 # already there
56343c78 1167my $CROAK = 5; # Die with an error if is already there
99870f4d
KW
1168
1169# Flags to give property statuses. The phrases are to remind maintainers that
1170# if the flag is changed, the indefinite article referring to it in the
1171# documentation may need to be as well.
1172my $NORMAL = "";
1173my $SUPPRESSED = 'z'; # The character should never actually be seen, since
1174 # it is suppressed
c12f2655
KW
1175my $PLACEHOLDER = 'P'; # A property that is defined as a placeholder in a
1176 # Unicode version that doesn't have it, but we need it
1177 # to be defined, if empty, to have things work.
1178 # Implies no pod entry generated
99870f4d
KW
1179my $DEPRECATED = 'D';
1180my $a_bold_deprecated = "a 'B<$DEPRECATED>'";
1181my $A_bold_deprecated = "A 'B<$DEPRECATED>'";
1182my $DISCOURAGED = 'X';
1183my $a_bold_discouraged = "an 'B<$DISCOURAGED>'";
1184my $A_bold_discouraged = "An 'B<$DISCOURAGED>'";
1185my $STRICTER = 'T';
1186my $a_bold_stricter = "a 'B<$STRICTER>'";
1187my $A_bold_stricter = "A 'B<$STRICTER>'";
1188my $STABILIZED = 'S';
1189my $a_bold_stabilized = "an 'B<$STABILIZED>'";
1190my $A_bold_stabilized = "An 'B<$STABILIZED>'";
1191my $OBSOLETE = 'O';
1192my $a_bold_obsolete = "an 'B<$OBSOLETE>'";
1193my $A_bold_obsolete = "An 'B<$OBSOLETE>'";
1194
1195my %status_past_participles = (
1196 $DISCOURAGED => 'discouraged',
1197 $SUPPRESSED => 'should never be generated',
1198 $STABILIZED => 'stabilized',
1199 $OBSOLETE => 'obsolete',
37e2e78e 1200 $DEPRECATED => 'deprecated',
99870f4d
KW
1201);
1202
f5817e0a
KW
1203# The format of the values of the tables:
1204my $EMPTY_FORMAT = "";
99870f4d
KW
1205my $BINARY_FORMAT = 'b';
1206my $DECIMAL_FORMAT = 'd';
1207my $FLOAT_FORMAT = 'f';
1208my $INTEGER_FORMAT = 'i';
1209my $HEX_FORMAT = 'x';
1210my $RATIONAL_FORMAT = 'r';
1211my $STRING_FORMAT = 's';
a14f3cb1 1212my $DECOMP_STRING_FORMAT = 'c';
c3ff2976 1213my $STRING_WHITE_SPACE_LIST = 'sw';
99870f4d
KW
1214
1215my %map_table_formats = (
1216 $BINARY_FORMAT => 'binary',
1217 $DECIMAL_FORMAT => 'single decimal digit',
1218 $FLOAT_FORMAT => 'floating point number',
1219 $INTEGER_FORMAT => 'integer',
add63c13 1220 $HEX_FORMAT => 'non-negative hex whole number; a code point',
99870f4d 1221 $RATIONAL_FORMAT => 'rational: an integer or a fraction',
1a9d544b 1222 $STRING_FORMAT => 'string',
92f9d56c 1223 $DECOMP_STRING_FORMAT => 'Perl\'s internal (Normalize.pm) decomposition mapping',
c3ff2976 1224 $STRING_WHITE_SPACE_LIST => 'string, but some elements are interpreted as a list; white space occurs only as list item separators'
99870f4d
KW
1225);
1226
1227# Unicode didn't put such derived files in a separate directory at first.
1228my $EXTRACTED_DIR = (-d 'extracted') ? 'extracted' : "";
1229my $EXTRACTED = ($EXTRACTED_DIR) ? "$EXTRACTED_DIR/" : "";
1230my $AUXILIARY = 'auxiliary';
1231
1232# Hashes that will eventually go into Heavy.pl for the use of utf8_heavy.pl
1233my %loose_to_file_of; # loosely maps table names to their respective
1234 # files
1235my %stricter_to_file_of; # same; but for stricter mapping.
1236my %nv_floating_to_rational; # maps numeric values floating point numbers to
1237 # their rational equivalent
c12f2655
KW
1238my %loose_property_name_of; # Loosely maps (non_string) property names to
1239 # standard form
99870f4d 1240
d867ccfb
KW
1241# Most properties are immune to caseless matching, otherwise you would get
1242# nonsensical results, as properties are a function of a code point, not
1243# everything that is caselessly equivalent to that code point. For example,
1244# Changes_When_Case_Folded('s') should be false, whereas caselessly it would
1245# be true because 's' and 'S' are equivalent caselessly. However,
1246# traditionally, [:upper:] and [:lower:] are equivalent caselessly, so we
1247# extend that concept to those very few properties that are like this. Each
1248# such property will match the full range caselessly. They are hard-coded in
1249# the program; it's not worth trying to make it general as it's extremely
1250# unlikely that they will ever change.
1251my %caseless_equivalent_to;
1252
99870f4d
KW
1253# These constants names and values were taken from the Unicode standard,
1254# version 5.1, section 3.12. They are used in conjunction with Hangul
6e5a209b
KW
1255# syllables. The '_string' versions are so generated tables can retain the
1256# hex format, which is the more familiar value
1257my $SBase_string = "0xAC00";
1258my $SBase = CORE::hex $SBase_string;
1259my $LBase_string = "0x1100";
1260my $LBase = CORE::hex $LBase_string;
1261my $VBase_string = "0x1161";
1262my $VBase = CORE::hex $VBase_string;
1263my $TBase_string = "0x11A7";
1264my $TBase = CORE::hex $TBase_string;
99870f4d
KW
1265my $SCount = 11172;
1266my $LCount = 19;
1267my $VCount = 21;
1268my $TCount = 28;
1269my $NCount = $VCount * $TCount;
1270
1271# For Hangul syllables; These store the numbers from Jamo.txt in conjunction
1272# with the above published constants.
1273my %Jamo;
1274my %Jamo_L; # Leading consonants
1275my %Jamo_V; # Vowels
1276my %Jamo_T; # Trailing consonants
1277
37e2e78e 1278my @backslash_X_tests; # List of tests read in for testing \X
99870f4d
KW
1279my @unhandled_properties; # Will contain a list of properties found in
1280 # the input that we didn't process.
f86864ac 1281my @match_properties; # Properties that have match tables, to be
99870f4d
KW
1282 # listed in the pod
1283my @map_properties; # Properties that get map files written
1284my @named_sequences; # NamedSequences.txt contents.
1285my %potential_files; # Generated list of all .txt files in the directory
1286 # structure so we can warn if something is being
1287 # ignored.
1288my @files_actually_output; # List of files we generated.
1289my @more_Names; # Some code point names are compound; this is used
1290 # to store the extra components of them.
1291my $MIN_FRACTION_LENGTH = 3; # How many digits of a floating point number at
1292 # the minimum before we consider it equivalent to a
1293 # candidate rational
1294my $MAX_FLOATING_SLOP = 10 ** - $MIN_FRACTION_LENGTH; # And in floating terms
1295
1296# These store references to certain commonly used property objects
1297my $gc;
1298my $perl;
1299my $block;
3e20195b
KW
1300my $perl_charname;
1301my $print;
7fc6cb55 1302my $Any;
359523e2 1303my $script;
99870f4d
KW
1304
1305# Are there conflicting names because of beginning with 'In_', or 'Is_'
1306my $has_In_conflicts = 0;
1307my $has_Is_conflicts = 0;
1308
1309sub internal_file_to_platform ($) {
1310 # Convert our file paths which have '/' separators to those of the
1311 # platform.
1312
1313 my $file = shift;
1314 return undef unless defined $file;
1315
1316 return File::Spec->join(split '/', $file);
d07a55ed 1317}
5beb625e 1318
99870f4d
KW
1319sub file_exists ($) { # platform independent '-e'. This program internally
1320 # uses slash as a path separator.
1321 my $file = shift;
1322 return 0 if ! defined $file;
1323 return -e internal_file_to_platform($file);
1324}
5beb625e 1325
99870f4d 1326sub objaddr($) {
23e33b60
KW
1327 # Returns the address of the blessed input object.
1328 # It doesn't check for blessedness because that would do a string eval
1329 # every call, and the program is structured so that this is never called
1330 # for a non-blessed object.
99870f4d 1331
23e33b60 1332 no overloading; # If overloaded, numifying below won't work.
99870f4d
KW
1333
1334 # Numifying a ref gives its address.
051df77b 1335 return pack 'J', $_[0];
99870f4d
KW
1336}
1337
558712cf 1338# These are used only if $annotate is true.
c4019d52
KW
1339# The entire range of Unicode characters is examined to populate these
1340# after all the input has been processed. But most can be skipped, as they
1341# have the same descriptive phrases, such as being unassigned
1342my @viacode; # Contains the 1 million character names
1343my @printable; # boolean: And are those characters printable?
1344my @annotate_char_type; # Contains a type of those characters, specifically
1345 # for the purposes of annotation.
1346my $annotate_ranges; # A map of ranges of code points that have the same
98dc9551 1347 # name for the purposes of annotation. They map to the
c4019d52
KW
1348 # upper edge of the range, so that the end point can
1349 # be immediately found. This is used to skip ahead to
1350 # the end of a range, and avoid processing each
1351 # individual code point in it.
1352my $unassigned_sans_noncharacters; # A Range_List of the unassigned
1353 # characters, but excluding those which are
1354 # also noncharacter code points
1355
1356# The annotation types are an extension of the regular range types, though
1357# some of the latter are folded into one. Make the new types negative to
1358# avoid conflicting with the regular types
1359my $SURROGATE_TYPE = -1;
1360my $UNASSIGNED_TYPE = -2;
1361my $PRIVATE_USE_TYPE = -3;
1362my $NONCHARACTER_TYPE = -4;
1363my $CONTROL_TYPE = -5;
1364my $UNKNOWN_TYPE = -6; # Used only if there is a bug in this program
1365
1366sub populate_char_info ($) {
558712cf 1367 # Used only with the $annotate option. Populates the arrays with the
c4019d52
KW
1368 # input code point's info that are needed for outputting more detailed
1369 # comments. If calling context wants a return, it is the end point of
1370 # any contiguous range of characters that share essentially the same info
1371
1372 my $i = shift;
1373 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1374
1375 $viacode[$i] = $perl_charname->value_of($i) || "";
1376
1377 # A character is generally printable if Unicode says it is,
1378 # but below we make sure that most Unicode general category 'C' types
1379 # aren't.
1380 $printable[$i] = $print->contains($i);
1381
1382 $annotate_char_type[$i] = $perl_charname->type_of($i) || 0;
1383
1384 # Only these two regular types are treated specially for annotations
1385 # purposes
1386 $annotate_char_type[$i] = 0 if $annotate_char_type[$i] != $CP_IN_NAME
1387 && $annotate_char_type[$i] != $HANGUL_SYLLABLE;
1388
1389 # Give a generic name to all code points that don't have a real name.
1390 # We output ranges, if applicable, for these. Also calculate the end
1391 # point of the range.
1392 my $end;
1393 if (! $viacode[$i]) {
1394 if ($gc-> table('Surrogate')->contains($i)) {
1395 $viacode[$i] = 'Surrogate';
1396 $annotate_char_type[$i] = $SURROGATE_TYPE;
1397 $printable[$i] = 0;
1398 $end = $gc->table('Surrogate')->containing_range($i)->end;
1399 }
1400 elsif ($gc-> table('Private_use')->contains($i)) {
1401 $viacode[$i] = 'Private Use';
1402 $annotate_char_type[$i] = $PRIVATE_USE_TYPE;
1403 $printable[$i] = 0;
1404 $end = $gc->table('Private_Use')->containing_range($i)->end;
1405 }
1406 elsif (Property::property_ref('Noncharacter_Code_Point')-> table('Y')->
1407 contains($i))
1408 {
1409 $viacode[$i] = 'Noncharacter';
1410 $annotate_char_type[$i] = $NONCHARACTER_TYPE;
1411 $printable[$i] = 0;
1412 $end = property_ref('Noncharacter_Code_Point')->table('Y')->
1413 containing_range($i)->end;
1414 }
1415 elsif ($gc-> table('Control')->contains($i)) {
1416 $viacode[$i] = 'Control';
1417 $annotate_char_type[$i] = $CONTROL_TYPE;
1418 $printable[$i] = 0;
1419 $end = 0x81 if $i == 0x80; # Hard-code this one known case
1420 }
1421 elsif ($gc-> table('Unassigned')->contains($i)) {
1422 $viacode[$i] = 'Unassigned, block=' . $block-> value_of($i);
1423 $annotate_char_type[$i] = $UNASSIGNED_TYPE;
1424 $printable[$i] = 0;
1425
1426 # Because we name the unassigned by the blocks they are in, it
1427 # can't go past the end of that block, and it also can't go past
1428 # the unassigned range it is in. The special table makes sure
1429 # that the non-characters, which are unassigned, are separated
1430 # out.
1431 $end = min($block->containing_range($i)->end,
1432 $unassigned_sans_noncharacters-> containing_range($i)->
1433 end);
13ca76ff
KW
1434 }
1435 else {
1436 Carp::my_carp_bug("Can't figure out how to annotate "
1437 . sprintf("U+%04X", $i)
1438 . ". Proceeding anyway.");
c4019d52
KW
1439 $viacode[$i] = 'UNKNOWN';
1440 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1441 $printable[$i] = 0;
1442 }
1443 }
1444
1445 # Here, has a name, but if it's one in which the code point number is
1446 # appended to the name, do that.
1447 elsif ($annotate_char_type[$i] == $CP_IN_NAME) {
1448 $viacode[$i] .= sprintf("-%04X", $i);
1449 $end = $perl_charname->containing_range($i)->end;
1450 }
1451
1452 # And here, has a name, but if it's a hangul syllable one, replace it with
1453 # the correct name from the Unicode algorithm
1454 elsif ($annotate_char_type[$i] == $HANGUL_SYLLABLE) {
1455 use integer;
1456 my $SIndex = $i - $SBase;
1457 my $L = $LBase + $SIndex / $NCount;
1458 my $V = $VBase + ($SIndex % $NCount) / $TCount;
1459 my $T = $TBase + $SIndex % $TCount;
1460 $viacode[$i] = "HANGUL SYLLABLE $Jamo{$L}$Jamo{$V}";
1461 $viacode[$i] .= $Jamo{$T} if $T != $TBase;
1462 $end = $perl_charname->containing_range($i)->end;
1463 }
1464
1465 return if ! defined wantarray;
1466 return $i if ! defined $end; # If not a range, return the input
1467
1468 # Save this whole range so can find the end point quickly
1469 $annotate_ranges->add_map($i, $end, $end);
1470
1471 return $end;
1472}
1473
23e33b60
KW
1474# Commented code below should work on Perl 5.8.
1475## This 'require' doesn't necessarily work in miniperl, and even if it does,
1476## the native perl version of it (which is what would operate under miniperl)
1477## is extremely slow, as it does a string eval every call.
1478#my $has_fast_scalar_util = $\18 !~ /miniperl/
1479# && defined eval "require Scalar::Util";
1480#
1481#sub objaddr($) {
1482# # Returns the address of the blessed input object. Uses the XS version if
1483# # available. It doesn't check for blessedness because that would do a
1484# # string eval every call, and the program is structured so that this is
1485# # never called for a non-blessed object.
1486#
1487# return Scalar::Util::refaddr($_[0]) if $has_fast_scalar_util;
1488#
1489# # Check at least that is a ref.
1490# my $pkg = ref($_[0]) or return undef;
1491#
1492# # Change to a fake package to defeat any overloaded stringify
1493# bless $_[0], 'main::Fake';
1494#
1495# # Numifying a ref gives its address.
051df77b 1496# my $addr = pack 'J', $_[0];
23e33b60
KW
1497#
1498# # Return to original class
1499# bless $_[0], $pkg;
1500# return $addr;
1501#}
1502
99870f4d
KW
1503sub max ($$) {
1504 my $a = shift;
1505 my $b = shift;
1506 return $a if $a >= $b;
1507 return $b;
1508}
1509
1510sub min ($$) {
1511 my $a = shift;
1512 my $b = shift;
1513 return $a if $a <= $b;
1514 return $b;
1515}
1516
1517sub clarify_number ($) {
1518 # This returns the input number with underscores inserted every 3 digits
1519 # in large (5 digits or more) numbers. Input must be entirely digits, not
1520 # checked.
1521
1522 my $number = shift;
1523 my $pos = length($number) - 3;
1524 return $number if $pos <= 1;
1525 while ($pos > 0) {
1526 substr($number, $pos, 0) = '_';
1527 $pos -= 3;
5beb625e 1528 }
99870f4d 1529 return $number;
99598c8c
JH
1530}
1531
12ac2576 1532
99870f4d 1533package Carp;
7ebf06b3 1534
99870f4d
KW
1535# These routines give a uniform treatment of messages in this program. They
1536# are placed in the Carp package to cause the stack trace to not include them,
1537# although an alternative would be to use another package and set @CARP_NOT
1538# for it.
12ac2576 1539
99870f4d 1540our $Verbose = 1 if main::DEBUG; # Useful info when debugging
12ac2576 1541
99f78760
KW
1542# This is a work-around suggested by Nicholas Clark to fix a problem with Carp
1543# and overload trying to load Scalar:Util under miniperl. See
1544# http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2009-11/msg01057.html
1545undef $overload::VERSION;
1546
99870f4d
KW
1547sub my_carp {
1548 my $message = shift || "";
1549 my $nofold = shift || 0;
7ebf06b3 1550
99870f4d
KW
1551 if ($message) {
1552 $message = main::join_lines($message);
1553 $message =~ s/^$0: *//; # Remove initial program name
1554 $message =~ s/[.;,]+$//; # Remove certain ending punctuation
1555 $message = "\n$0: $message;";
12ac2576 1556
99870f4d
KW
1557 # Fold the message with program name, semi-colon end punctuation
1558 # (which looks good with the message that carp appends to it), and a
1559 # hanging indent for continuation lines.
1560 $message = main::simple_fold($message, "", 4) unless $nofold;
1561 $message =~ s/\n$//; # Remove the trailing nl so what carp
1562 # appends is to the same line
1563 }
12ac2576 1564
99870f4d 1565 return $message if defined wantarray; # If a caller just wants the msg
12ac2576 1566
99870f4d
KW
1567 carp $message;
1568 return;
1569}
7ebf06b3 1570
99870f4d
KW
1571sub my_carp_bug {
1572 # This is called when it is clear that the problem is caused by a bug in
1573 # this program.
7ebf06b3 1574
99870f4d
KW
1575 my $message = shift;
1576 $message =~ s/^$0: *//;
1577 $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");
1578 carp $message;
1579 return;
1580}
7ebf06b3 1581
99870f4d
KW
1582sub carp_too_few_args {
1583 if (@_ != 2) {
1584 my_carp_bug("Wrong number of arguments: to 'carp_too_few_arguments'. No action taken.");
1585 return;
12ac2576 1586 }
7ebf06b3 1587
99870f4d
KW
1588 my $args_ref = shift;
1589 my $count = shift;
7ebf06b3 1590
99870f4d
KW
1591 my_carp_bug("Need at least $count arguments to "
1592 . (caller 1)[3]
1593 . ". Instead got: '"
1594 . join ', ', @$args_ref
1595 . "'. No action taken.");
1596 return;
12ac2576
JP
1597}
1598
99870f4d
KW
1599sub carp_extra_args {
1600 my $args_ref = shift;
1601 my_carp_bug("Too many arguments to 'carp_extra_args': (" . join(', ', @_) . "); Extras ignored.") if @_;
12ac2576 1602
99870f4d
KW
1603 unless (ref $args_ref) {
1604 my_carp_bug("Argument to 'carp_extra_args' ($args_ref) must be a ref. Not checking arguments.");
1605 return;
1606 }
1607 my ($package, $file, $line) = caller;
1608 my $subroutine = (caller 1)[3];
cf25bb62 1609
99870f4d
KW
1610 my $list;
1611 if (ref $args_ref eq 'HASH') {
1612 foreach my $key (keys %$args_ref) {
1613 $args_ref->{$key} = $UNDEF unless defined $args_ref->{$key};
cf25bb62 1614 }
99870f4d 1615 $list = join ', ', each %{$args_ref};
cf25bb62 1616 }
99870f4d
KW
1617 elsif (ref $args_ref eq 'ARRAY') {
1618 foreach my $arg (@$args_ref) {
1619 $arg = $UNDEF unless defined $arg;
1620 }
1621 $list = join ', ', @$args_ref;
1622 }
1623 else {
1624 my_carp_bug("Can't cope with ref "
1625 . ref($args_ref)
1626 . " . argument to 'carp_extra_args'. Not checking arguments.");
1627 return;
1628 }
1629
1630 my_carp_bug("Unrecognized parameters in options: '$list' to $subroutine. Skipped.");
1631 return;
d73e5302
JH
1632}
1633
99870f4d
KW
1634package main;
1635
1636{ # Closure
1637
1638 # This program uses the inside-out method for objects, as recommended in
1639 # "Perl Best Practices". This closure aids in generating those. There
1640 # are two routines. setup_package() is called once per package to set
1641 # things up, and then set_access() is called for each hash representing a
1642 # field in the object. These routines arrange for the object to be
1643 # properly destroyed when no longer used, and for standard accessor
1644 # functions to be generated. If you need more complex accessors, just
1645 # write your own and leave those accesses out of the call to set_access().
1646 # More details below.
1647
1648 my %constructor_fields; # fields that are to be used in constructors; see
1649 # below
1650
1651 # The values of this hash will be the package names as keys to other
1652 # hashes containing the name of each field in the package as keys, and
1653 # references to their respective hashes as values.
1654 my %package_fields;
1655
1656 sub setup_package {
1657 # Sets up the package, creating standard DESTROY and dump methods
1658 # (unless already defined). The dump method is used in debugging by
1659 # simple_dumper().
1660 # The optional parameters are:
1661 # a) a reference to a hash, that gets populated by later
1662 # set_access() calls with one of the accesses being
1663 # 'constructor'. The caller can then refer to this, but it is
1664 # not otherwise used by these two routines.
1665 # b) a reference to a callback routine to call during destruction
1666 # of the object, before any fields are actually destroyed
1667
1668 my %args = @_;
1669 my $constructor_ref = delete $args{'Constructor_Fields'};
1670 my $destroy_callback = delete $args{'Destroy_Callback'};
1671 Carp::carp_extra_args(\@_) if main::DEBUG && %args;
1672
1673 my %fields;
1674 my $package = (caller)[0];
1675
1676 $package_fields{$package} = \%fields;
1677 $constructor_fields{$package} = $constructor_ref;
1678
1679 unless ($package->can('DESTROY')) {
1680 my $destroy_name = "${package}::DESTROY";
1681 no strict "refs";
1682
1683 # Use typeglob to give the anonymous subroutine the name we want
1684 *$destroy_name = sub {
1685 my $self = shift;
ffe43484 1686 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1687
1688 $self->$destroy_callback if $destroy_callback;
1689 foreach my $field (keys %{$package_fields{$package}}) {
1690 #print STDERR __LINE__, ": Destroying ", ref $self, " ", sprintf("%04X", $addr), ": ", $field, "\n";
1691 delete $package_fields{$package}{$field}{$addr};
1692 }
1693 return;
1694 }
1695 }
1696
1697 unless ($package->can('dump')) {
1698 my $dump_name = "${package}::dump";
1699 no strict "refs";
1700 *$dump_name = sub {
1701 my $self = shift;
1702 return dump_inside_out($self, $package_fields{$package}, @_);
1703 }
1704 }
1705 return;
1706 }
1707
1708 sub set_access {
1709 # Arrange for the input field to be garbage collected when no longer
1710 # needed. Also, creates standard accessor functions for the field
1711 # based on the optional parameters-- none if none of these parameters:
1712 # 'addable' creates an 'add_NAME()' accessor function.
1713 # 'readable' or 'readable_array' creates a 'NAME()' accessor
1714 # function.
1715 # 'settable' creates a 'set_NAME()' accessor function.
1716 # 'constructor' doesn't create an accessor function, but adds the
1717 # field to the hash that was previously passed to
1718 # setup_package();
1719 # Any of the accesses can be abbreviated down, so that 'a', 'ad',
1720 # 'add' etc. all mean 'addable'.
1721 # The read accessor function will work on both array and scalar
1722 # values. If another accessor in the parameter list is 'a', the read
1723 # access assumes an array. You can also force it to be array access
1724 # by specifying 'readable_array' instead of 'readable'
1725 #
1726 # A sort-of 'protected' access can be set-up by preceding the addable,
1727 # readable or settable with some initial portion of 'protected_' (but,
1728 # the underscore is required), like 'p_a', 'pro_set', etc. The
1729 # "protection" is only by convention. All that happens is that the
1730 # accessor functions' names begin with an underscore. So instead of
1731 # calling set_foo, the call is _set_foo. (Real protection could be
c1739a4a 1732 # accomplished by having a new subroutine, end_package, called at the
99870f4d
KW
1733 # end of each package, and then storing the __LINE__ ranges and
1734 # checking them on every accessor. But that is way overkill.)
1735
1736 # We create anonymous subroutines as the accessors and then use
1737 # typeglobs to assign them to the proper package and name
1738
1739 my $name = shift; # Name of the field
1740 my $field = shift; # Reference to the inside-out hash containing the
1741 # field
1742
1743 my $package = (caller)[0];
1744
1745 if (! exists $package_fields{$package}) {
1746 croak "$0: Must call 'setup_package' before 'set_access'";
1747 }
d73e5302 1748
99870f4d
KW
1749 # Stash the field so DESTROY can get it.
1750 $package_fields{$package}{$name} = $field;
cf25bb62 1751
99870f4d
KW
1752 # Remaining arguments are the accessors. For each...
1753 foreach my $access (@_) {
1754 my $access = lc $access;
cf25bb62 1755
99870f4d 1756 my $protected = "";
cf25bb62 1757
99870f4d
KW
1758 # Match the input as far as it goes.
1759 if ($access =~ /^(p[^_]*)_/) {
1760 $protected = $1;
1761 if (substr('protected_', 0, length $protected)
1762 eq $protected)
1763 {
1764
1765 # Add 1 for the underscore not included in $protected
1766 $access = substr($access, length($protected) + 1);
1767 $protected = '_';
1768 }
1769 else {
1770 $protected = "";
1771 }
1772 }
1773
1774 if (substr('addable', 0, length $access) eq $access) {
1775 my $subname = "${package}::${protected}add_$name";
1776 no strict "refs";
1777
1778 # add_ accessor. Don't add if already there, which we
1779 # determine using 'eq' for scalars and '==' otherwise.
1780 *$subname = sub {
1781 use strict "refs";
1782 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
1783 my $self = shift;
1784 my $value = shift;
ffe43484 1785 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1786 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1787 if (ref $value) {
f998e60c 1788 return if grep { $value == $_ } @{$field->{$addr}};
99870f4d
KW
1789 }
1790 else {
f998e60c 1791 return if grep { $value eq $_ } @{$field->{$addr}};
99870f4d 1792 }
f998e60c 1793 push @{$field->{$addr}}, $value;
99870f4d
KW
1794 return;
1795 }
1796 }
1797 elsif (substr('constructor', 0, length $access) eq $access) {
1798 if ($protected) {
1799 Carp::my_carp_bug("Can't set-up 'protected' constructors")
1800 }
1801 else {
1802 $constructor_fields{$package}{$name} = $field;
1803 }
1804 }
1805 elsif (substr('readable_array', 0, length $access) eq $access) {
1806
1807 # Here has read access. If one of the other parameters for
1808 # access is array, or this one specifies array (by being more
1809 # than just 'readable_'), then create a subroutine that
1810 # assumes the data is an array. Otherwise just a scalar
1811 my $subname = "${package}::${protected}$name";
1812 if (grep { /^a/i } @_
1813 or length($access) > length('readable_'))
1814 {
1815 no strict "refs";
1816 *$subname = sub {
1817 use strict "refs";
23e33b60 1818 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
ffe43484 1819 my $addr = do { no overloading; pack 'J', $_[0]; };
99870f4d
KW
1820 if (ref $field->{$addr} ne 'ARRAY') {
1821 my $type = ref $field->{$addr};
1822 $type = 'scalar' unless $type;
1823 Carp::my_carp_bug("Trying to read $name as an array when it is a $type. Big problems.");
1824 return;
1825 }
1826 return scalar @{$field->{$addr}} unless wantarray;
1827
1828 # Make a copy; had problems with caller modifying the
1829 # original otherwise
1830 my @return = @{$field->{$addr}};
1831 return @return;
1832 }
1833 }
1834 else {
1835
1836 # Here not an array value, a simpler function.
1837 no strict "refs";
1838 *$subname = sub {
1839 use strict "refs";
23e33b60 1840 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
f998e60c 1841 no overloading;
051df77b 1842 return $field->{pack 'J', $_[0]};
99870f4d
KW
1843 }
1844 }
1845 }
1846 elsif (substr('settable', 0, length $access) eq $access) {
1847 my $subname = "${package}::${protected}set_$name";
1848 no strict "refs";
1849 *$subname = sub {
1850 use strict "refs";
23e33b60
KW
1851 if (main::DEBUG) {
1852 return Carp::carp_too_few_args(\@_, 2) if @_ < 2;
1853 Carp::carp_extra_args(\@_) if @_ > 2;
1854 }
1855 # $self is $_[0]; $value is $_[1]
f998e60c 1856 no overloading;
051df77b 1857 $field->{pack 'J', $_[0]} = $_[1];
99870f4d
KW
1858 return;
1859 }
1860 }
1861 else {
1862 Carp::my_carp_bug("Unknown accessor type $access. No accessor set.");
1863 }
cf25bb62 1864 }
99870f4d 1865 return;
cf25bb62 1866 }
99870f4d
KW
1867}
1868
1869package Input_file;
1870
1871# All input files use this object, which stores various attributes about them,
1872# and provides for convenient, uniform handling. The run method wraps the
1873# processing. It handles all the bookkeeping of opening, reading, and closing
1874# the file, returning only significant input lines.
1875#
1876# Each object gets a handler which processes the body of the file, and is
1877# called by run(). Most should use the generic, default handler, which has
1878# code scrubbed to handle things you might not expect. A handler should
1879# basically be a while(next_line()) {...} loop.
1880#
1881# You can also set up handlers to
1882# 1) call before the first line is read for pre processing
1883# 2) call to adjust each line of the input before the main handler gets them
1884# 3) call upon EOF before the main handler exits its loop
1885# 4) call at the end for post processing
1886#
1887# $_ is used to store the input line, and is to be filtered by the
1888# each_line_handler()s. So, if the format of the line is not in the desired
1889# format for the main handler, these are used to do that adjusting. They can
1890# be stacked (by enclosing them in an [ anonymous array ] in the constructor,
1891# so the $_ output of one is used as the input to the next. None of the other
1892# handlers are stackable, but could easily be changed to be so.
1893#
1894# Most of the handlers can call insert_lines() or insert_adjusted_lines()
1895# which insert the parameters as lines to be processed before the next input
1896# file line is read. This allows the EOF handler to flush buffers, for
1897# example. The difference between the two routines is that the lines inserted
1898# by insert_lines() are subjected to the each_line_handler()s. (So if you
1899# called it from such a handler, you would get infinite recursion.) Lines
1900# inserted by insert_adjusted_lines() go directly to the main handler without
1901# any adjustments. If the post-processing handler calls any of these, there
1902# will be no effect. Some error checking for these conditions could be added,
1903# but it hasn't been done.
1904#
1905# carp_bad_line() should be called to warn of bad input lines, which clears $_
1906# to prevent further processing of the line. This routine will output the
1907# message as a warning once, and then keep a count of the lines that have the
1908# same message, and output that count at the end of the file's processing.
1909# This keeps the number of messages down to a manageable amount.
1910#
1911# get_missings() should be called to retrieve any @missing input lines.
1912# Messages will be raised if this isn't done if the options aren't to ignore
1913# missings.
1914
1915sub trace { return main::trace(@_); }
1916
99870f4d
KW
1917{ # Closure
1918 # Keep track of fields that are to be put into the constructor.
1919 my %constructor_fields;
1920
1921 main::setup_package(Constructor_Fields => \%constructor_fields);
1922
1923 my %file; # Input file name, required
1924 main::set_access('file', \%file, qw{ c r });
1925
1926 my %first_released; # Unicode version file was first released in, required
1927 main::set_access('first_released', \%first_released, qw{ c r });
1928
1929 my %handler; # Subroutine to process the input file, defaults to
1930 # 'process_generic_property_file'
1931 main::set_access('handler', \%handler, qw{ c });
1932
1933 my %property;
1934 # name of property this file is for. defaults to none, meaning not
1935 # applicable, or is otherwise determinable, for example, from each line.
1936 main::set_access('property', \%property, qw{ c });
1937
1938 my %optional;
1939 # If this is true, the file is optional. If not present, no warning is
1940 # output. If it is present, the string given by this parameter is
1941 # evaluated, and if false the file is not processed.
1942 main::set_access('optional', \%optional, 'c', 'r');
1943
1944 my %non_skip;
1945 # This is used for debugging, to skip processing of all but a few input
1946 # files. Add 'non_skip => 1' to the constructor for those files you want
1947 # processed when you set the $debug_skip global.
1948 main::set_access('non_skip', \%non_skip, 'c');
1949
37e2e78e
KW
1950 my %skip;
1951 # This is used to skip processing of this input file semi-permanently.
1952 # It is used for files that we aren't planning to process anytime soon,
1953 # but want to allow to be in the directory and not raise a message that we
1954 # are not handling. Mostly for test files. This is in contrast to the
1955 # non_skip element, which is supposed to be used very temporarily for
1956 # debugging. Sets 'optional' to 1
1957 main::set_access('skip', \%skip, 'c');
1958
99870f4d
KW
1959 my %each_line_handler;
1960 # list of subroutines to look at and filter each non-comment line in the
1961 # file. defaults to none. The subroutines are called in order, each is
1962 # to adjust $_ for the next one, and the final one adjusts it for
1963 # 'handler'
1964 main::set_access('each_line_handler', \%each_line_handler, 'c');
1965
1966 my %has_missings_defaults;
1967 # ? Are there lines in the file giving default values for code points
1968 # missing from it?. Defaults to NO_DEFAULTS. Otherwise NOT_IGNORED is
1969 # the norm, but IGNORED means it has such lines, but the handler doesn't
1970 # use them. Having these three states allows us to catch changes to the
1971 # UCD that this program should track
1972 main::set_access('has_missings_defaults',
1973 \%has_missings_defaults, qw{ c r });
1974
1975 my %pre_handler;
1976 # Subroutine to call before doing anything else in the file. If undef, no
1977 # such handler is called.
1978 main::set_access('pre_handler', \%pre_handler, qw{ c });
1979
1980 my %eof_handler;
1981 # Subroutine to call upon getting an EOF on the input file, but before
1982 # that is returned to the main handler. This is to allow buffers to be
1983 # flushed. The handler is expected to call insert_lines() or
1984 # insert_adjusted() with the buffered material
1985 main::set_access('eof_handler', \%eof_handler, qw{ c r });
1986
1987 my %post_handler;
1988 # Subroutine to call after all the lines of the file are read in and
1989 # processed. If undef, no such handler is called.
1990 main::set_access('post_handler', \%post_handler, qw{ c });
1991
1992 my %progress_message;
1993 # Message to print to display progress in lieu of the standard one
1994 main::set_access('progress_message', \%progress_message, qw{ c });
1995
1996 my %handle;
1997 # cache open file handle, internal. Is undef if file hasn't been
1998 # processed at all, empty if has;
1999 main::set_access('handle', \%handle);
2000
2001 my %added_lines;
2002 # cache of lines added virtually to the file, internal
2003 main::set_access('added_lines', \%added_lines);
2004
2005 my %errors;
2006 # cache of errors found, internal
2007 main::set_access('errors', \%errors);
2008
2009 my %missings;
2010 # storage of '@missing' defaults lines
2011 main::set_access('missings', \%missings);
2012
2013 sub new {
2014 my $class = shift;
2015
2016 my $self = bless \do{ my $anonymous_scalar }, $class;
ffe43484 2017 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2018
2019 # Set defaults
2020 $handler{$addr} = \&main::process_generic_property_file;
2021 $non_skip{$addr} = 0;
37e2e78e 2022 $skip{$addr} = 0;
99870f4d
KW
2023 $has_missings_defaults{$addr} = $NO_DEFAULTS;
2024 $handle{$addr} = undef;
2025 $added_lines{$addr} = [ ];
2026 $each_line_handler{$addr} = [ ];
2027 $errors{$addr} = { };
2028 $missings{$addr} = [ ];
2029
2030 # Two positional parameters.
99f78760 2031 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
99870f4d
KW
2032 $file{$addr} = main::internal_file_to_platform(shift);
2033 $first_released{$addr} = shift;
2034
2035 # The rest of the arguments are key => value pairs
2036 # %constructor_fields has been set up earlier to list all possible
2037 # ones. Either set or push, depending on how the default has been set
2038 # up just above.
2039 my %args = @_;
2040 foreach my $key (keys %args) {
2041 my $argument = $args{$key};
2042
2043 # Note that the fields are the lower case of the constructor keys
2044 my $hash = $constructor_fields{lc $key};
2045 if (! defined $hash) {
2046 Carp::my_carp_bug("Unrecognized parameters '$key => $argument' to new() for $self. Skipped");
2047 next;
2048 }
2049 if (ref $hash->{$addr} eq 'ARRAY') {
2050 if (ref $argument eq 'ARRAY') {
2051 foreach my $argument (@{$argument}) {
2052 next if ! defined $argument;
2053 push @{$hash->{$addr}}, $argument;
2054 }
2055 }
2056 else {
2057 push @{$hash->{$addr}}, $argument if defined $argument;
2058 }
2059 }
2060 else {
2061 $hash->{$addr} = $argument;
2062 }
2063 delete $args{$key};
2064 };
2065
2066 # If the file has a property for it, it means that the property is not
2067 # listed in the file's entries. So add a handler to the list of line
2068 # handlers to insert the property name into the lines, to provide a
2069 # uniform interface to the final processing subroutine.
2070 # the final code doesn't have to worry about that.
2071 if ($property{$addr}) {
2072 push @{$each_line_handler{$addr}}, \&_insert_property_into_line;
2073 }
2074
2075 if ($non_skip{$addr} && ! $debug_skip && $verbosity) {
2076 print "Warning: " . __PACKAGE__ . " constructor for $file{$addr} has useless 'non_skip' in it\n";
a3a8c5f0 2077 }
99870f4d 2078
37e2e78e
KW
2079 $optional{$addr} = 1 if $skip{$addr};
2080
99870f4d 2081 return $self;
d73e5302
JH
2082 }
2083
cf25bb62 2084
99870f4d
KW
2085 use overload
2086 fallback => 0,
2087 qw("") => "_operator_stringify",
2088 "." => \&main::_operator_dot,
2089 ;
cf25bb62 2090
99870f4d
KW
2091 sub _operator_stringify {
2092 my $self = shift;
cf25bb62 2093
99870f4d 2094 return __PACKAGE__ . " object for " . $self->file;
d73e5302 2095 }
d73e5302 2096
99870f4d
KW
2097 # flag to make sure extracted files are processed early
2098 my $seen_non_extracted_non_age = 0;
d73e5302 2099
99870f4d
KW
2100 sub run {
2101 # Process the input object $self. This opens and closes the file and
2102 # calls all the handlers for it. Currently, this can only be called
2103 # once per file, as it destroy's the EOF handler
d73e5302 2104
99870f4d
KW
2105 my $self = shift;
2106 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
b6922eda 2107
ffe43484 2108 my $addr = do { no overloading; pack 'J', $self; };
b6922eda 2109
99870f4d 2110 my $file = $file{$addr};
d73e5302 2111
99870f4d
KW
2112 # Don't process if not expecting this file (because released later
2113 # than this Unicode version), and isn't there. This means if someone
2114 # copies it into an earlier version's directory, we will go ahead and
2115 # process it.
2116 return if $first_released{$addr} gt $v_version && ! -e $file;
2117
2118 # If in debugging mode and this file doesn't have the non-skip
2119 # flag set, and isn't one of the critical files, skip it.
2120 if ($debug_skip
2121 && $first_released{$addr} ne v0
2122 && ! $non_skip{$addr})
2123 {
2124 print "Skipping $file in debugging\n" if $verbosity;
2125 return;
2126 }
2127
2128 # File could be optional
37e2e78e 2129 if ($optional{$addr}) {
99870f4d
KW
2130 return unless -e $file;
2131 my $result = eval $optional{$addr};
2132 if (! defined $result) {
2133 Carp::my_carp_bug("Got '$@' when tried to eval $optional{$addr}. $file Skipped.");
2134 return;
2135 }
2136 if (! $result) {
2137 if ($verbosity) {
2138 print STDERR "Skipping processing input file '$file' because '$optional{$addr}' is not true\n";
2139 }
2140 return;
2141 }
2142 }
2143
2144 if (! defined $file || ! -e $file) {
2145
2146 # If the file doesn't exist, see if have internal data for it
2147 # (based on first_released being 0).
2148 if ($first_released{$addr} eq v0) {
2149 $handle{$addr} = 'pretend_is_open';
2150 }
2151 else {
2152 if (! $optional{$addr} # File could be optional
2153 && $v_version ge $first_released{$addr})
2154 {
2155 print STDERR "Skipping processing input file '$file' because not found\n" if $v_version ge $first_released{$addr};
2156 }
2157 return;
2158 }
2159 }
2160 else {
2161
37e2e78e
KW
2162 # Here, the file exists. Some platforms may change the case of
2163 # its name
99870f4d 2164 if ($seen_non_extracted_non_age) {
517956bf 2165 if ($file =~ /$EXTRACTED/i) {
99870f4d 2166 Carp::my_carp_bug(join_lines(<<END
99f78760 2167$file should be processed just after the 'Prop...Alias' files, and before
99870f4d
KW
2168anything not in the $EXTRACTED_DIR directory. Proceeding, but the results may
2169have subtle problems
2170END
2171 ));
2172 }
2173 }
2174 elsif ($EXTRACTED_DIR
2175 && $first_released{$addr} ne v0
517956bf
CB
2176 && $file !~ /$EXTRACTED/i
2177 && lc($file) ne 'dage.txt')
99870f4d
KW
2178 {
2179 # We don't set this (by the 'if' above) if we have no
2180 # extracted directory, so if running on an early version,
2181 # this test won't work. Not worth worrying about.
2182 $seen_non_extracted_non_age = 1;
2183 }
2184
2185 # And mark the file as having being processed, and warn if it
2186 # isn't a file we are expecting. As we process the files,
2187 # they are deleted from the hash, so any that remain at the
2188 # end of the program are files that we didn't process.
517956bf
CB
2189 my $fkey = File::Spec->rel2abs($file);
2190 my $expecting = delete $potential_files{$fkey};
2191 $expecting = delete $potential_files{lc($fkey)} unless defined $expecting;
678f13d5
KW
2192 Carp::my_carp("Was not expecting '$file'.") if
2193 ! $expecting
99870f4d
KW
2194 && ! defined $handle{$addr};
2195
37e2e78e
KW
2196 # Having deleted from expected files, we can quit if not to do
2197 # anything. Don't print progress unless really want verbosity
2198 if ($skip{$addr}) {
2199 print "Skipping $file.\n" if $verbosity >= $VERBOSE;
2200 return;
2201 }
2202
99870f4d
KW
2203 # Open the file, converting the slashes used in this program
2204 # into the proper form for the OS
2205 my $file_handle;
2206 if (not open $file_handle, "<", $file) {
2207 Carp::my_carp("Can't open $file. Skipping: $!");
2208 return 0;
2209 }
2210 $handle{$addr} = $file_handle; # Cache the open file handle
2211 }
2212
2213 if ($verbosity >= $PROGRESS) {
2214 if ($progress_message{$addr}) {
2215 print "$progress_message{$addr}\n";
2216 }
2217 else {
2218 # If using a virtual file, say so.
2219 print "Processing ", (-e $file)
2220 ? $file
2221 : "substitute $file",
2222 "\n";
2223 }
2224 }
2225
2226
2227 # Call any special handler for before the file.
2228 &{$pre_handler{$addr}}($self) if $pre_handler{$addr};
2229
2230 # Then the main handler
2231 &{$handler{$addr}}($self);
2232
2233 # Then any special post-file handler.
2234 &{$post_handler{$addr}}($self) if $post_handler{$addr};
2235
2236 # If any errors have been accumulated, output the counts (as the first
2237 # error message in each class was output when it was encountered).
2238 if ($errors{$addr}) {
2239 my $total = 0;
2240 my $types = 0;
2241 foreach my $error (keys %{$errors{$addr}}) {
2242 $total += $errors{$addr}->{$error};
2243 delete $errors{$addr}->{$error};
2244 $types++;
2245 }
2246 if ($total > 1) {
2247 my $message
2248 = "A total of $total lines had errors in $file. ";
2249
2250 $message .= ($types == 1)
2251 ? '(Only the first one was displayed.)'
2252 : '(Only the first of each type was displayed.)';
2253 Carp::my_carp($message);
2254 }
2255 }
2256
2257 if (@{$missings{$addr}}) {
2258 Carp::my_carp_bug("Handler for $file didn't look at all the \@missing lines. Generated tables likely are wrong");
2259 }
2260
2261 # If a real file handle, close it.
2262 close $handle{$addr} or Carp::my_carp("Can't close $file: $!") if
2263 ref $handle{$addr};
2264 $handle{$addr} = ""; # Uses empty to indicate that has already seen
2265 # the file, as opposed to undef
2266 return;
2267 }
2268
2269 sub next_line {
2270 # Sets $_ to be the next logical input line, if any. Returns non-zero
2271 # if such a line exists. 'logical' means that any lines that have
2272 # been added via insert_lines() will be returned in $_ before the file
2273 # is read again.
2274
2275 my $self = shift;
2276 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2277
ffe43484 2278 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2279
2280 # Here the file is open (or if the handle is not a ref, is an open
2281 # 'virtual' file). Get the next line; any inserted lines get priority
2282 # over the file itself.
2283 my $adjusted;
2284
2285 LINE:
2286 while (1) { # Loop until find non-comment, non-empty line
2287 #local $to_trace = 1 if main::DEBUG;
2288 my $inserted_ref = shift @{$added_lines{$addr}};
2289 if (defined $inserted_ref) {
2290 ($adjusted, $_) = @{$inserted_ref};
2291 trace $adjusted, $_ if main::DEBUG && $to_trace;
2292 return 1 if $adjusted;
2293 }
2294 else {
2295 last if ! ref $handle{$addr}; # Don't read unless is real file
2296 last if ! defined ($_ = readline $handle{$addr});
2297 }
2298 chomp;
2299 trace $_ if main::DEBUG && $to_trace;
2300
2301 # See if this line is the comment line that defines what property
2302 # value that code points that are not listed in the file should
2303 # have. The format or existence of these lines is not guaranteed
2304 # by Unicode since they are comments, but the documentation says
2305 # that this was added for machine-readability, so probably won't
2306 # change. This works starting in Unicode Version 5.0. They look
2307 # like:
2308 #
2309 # @missing: 0000..10FFFF; Not_Reordered
2310 # @missing: 0000..10FFFF; Decomposition_Mapping; <code point>
2311 # @missing: 0000..10FFFF; ; NaN
2312 #
2313 # Save the line for a later get_missings() call.
2314 if (/$missing_defaults_prefix/) {
2315 if ($has_missings_defaults{$addr} == $NO_DEFAULTS) {
2316 $self->carp_bad_line("Unexpected \@missing line. Assuming no missing entries");
2317 }
2318 elsif ($has_missings_defaults{$addr} == $NOT_IGNORED) {
2319 my @defaults = split /\s* ; \s*/x, $_;
2320
2321 # The first field is the @missing, which ends in a
2322 # semi-colon, so can safely shift.
2323 shift @defaults;
2324
2325 # Some of these lines may have empty field placeholders
2326 # which get in the way. An example is:
2327 # @missing: 0000..10FFFF; ; NaN
2328 # Remove them. Process starting from the top so the
2329 # splice doesn't affect things still to be looked at.
2330 for (my $i = @defaults - 1; $i >= 0; $i--) {
2331 next if $defaults[$i] ne "";
2332 splice @defaults, $i, 1;
2333 }
2334
2335 # What's left should be just the property (maybe) and the
2336 # default. Having only one element means it doesn't have
2337 # the property.
2338 my $default;
2339 my $property;
2340 if (@defaults >= 1) {
2341 if (@defaults == 1) {
2342 $default = $defaults[0];
2343 }
2344 else {
2345 $property = $defaults[0];
2346 $default = $defaults[1];
2347 }
2348 }
2349
2350 if (@defaults < 1
2351 || @defaults > 2
2352 || ($default =~ /^</
2353 && $default !~ /^<code *point>$/i
2354 && $default !~ /^<none>$/i))
2355 {
2356 $self->carp_bad_line("Unrecognized \@missing line: $_. Assuming no missing entries");
2357 }
2358 else {
2359
2360 # If the property is missing from the line, it should
2361 # be the one for the whole file
2362 $property = $property{$addr} if ! defined $property;
2363
2364 # Change <none> to the null string, which is what it
2365 # really means. If the default is the code point
2366 # itself, set it to <code point>, which is what
2367 # Unicode uses (but sometimes they've forgotten the
2368 # space)
2369 if ($default =~ /^<none>$/i) {
2370 $default = "";
2371 }
2372 elsif ($default =~ /^<code *point>$/i) {
2373 $default = $CODE_POINT;
2374 }
2375
2376 # Store them as a sub-arrays with both components.
2377 push @{$missings{$addr}}, [ $default, $property ];
2378 }
2379 }
2380
2381 # There is nothing for the caller to process on this comment
2382 # line.
2383 next;
2384 }
2385
2386 # Remove comments and trailing space, and skip this line if the
2387 # result is empty
2388 s/#.*//;
2389 s/\s+$//;
2390 next if /^$/;
2391
2392 # Call any handlers for this line, and skip further processing of
2393 # the line if the handler sets the line to null.
2394 foreach my $sub_ref (@{$each_line_handler{$addr}}) {
2395 &{$sub_ref}($self);
2396 next LINE if /^$/;
2397 }
2398
2399 # Here the line is ok. return success.
2400 return 1;
2401 } # End of looping through lines.
2402
2403 # If there is an EOF handler, call it (only once) and if it generates
2404 # more lines to process go back in the loop to handle them.
2405 if ($eof_handler{$addr}) {
2406 &{$eof_handler{$addr}}($self);
2407 $eof_handler{$addr} = ""; # Currently only get one shot at it.
2408 goto LINE if $added_lines{$addr};
2409 }
2410
2411 # Return failure -- no more lines.
2412 return 0;
2413
2414 }
2415
2416# Not currently used, not fully tested.
2417# sub peek {
2418# # Non-destructive look-ahead one non-adjusted, non-comment, non-blank
2419# # record. Not callable from an each_line_handler(), nor does it call
2420# # an each_line_handler() on the line.
2421#
2422# my $self = shift;
ffe43484 2423# my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2424#
2425# foreach my $inserted_ref (@{$added_lines{$addr}}) {
2426# my ($adjusted, $line) = @{$inserted_ref};
2427# next if $adjusted;
2428#
2429# # Remove comments and trailing space, and return a non-empty
2430# # resulting line
2431# $line =~ s/#.*//;
2432# $line =~ s/\s+$//;
2433# return $line if $line ne "";
2434# }
2435#
2436# return if ! ref $handle{$addr}; # Don't read unless is real file
2437# while (1) { # Loop until find non-comment, non-empty line
2438# local $to_trace = 1 if main::DEBUG;
2439# trace $_ if main::DEBUG && $to_trace;
2440# return if ! defined (my $line = readline $handle{$addr});
2441# chomp $line;
2442# push @{$added_lines{$addr}}, [ 0, $line ];
2443#
2444# $line =~ s/#.*//;
2445# $line =~ s/\s+$//;
2446# return $line if $line ne "";
2447# }
2448#
2449# return;
2450# }
2451
2452
2453 sub insert_lines {
2454 # Lines can be inserted so that it looks like they were in the input
2455 # file at the place it was when this routine is called. See also
2456 # insert_adjusted_lines(). Lines inserted via this routine go through
2457 # any each_line_handler()
2458
2459 my $self = shift;
2460
2461 # Each inserted line is an array, with the first element being 0 to
2462 # indicate that this line hasn't been adjusted, and needs to be
2463 # processed.
f998e60c 2464 no overloading;
051df77b 2465 push @{$added_lines{pack 'J', $self}}, map { [ 0, $_ ] } @_;
99870f4d
KW
2466 return;
2467 }
2468
2469 sub insert_adjusted_lines {
2470 # Lines can be inserted so that it looks like they were in the input
2471 # file at the place it was when this routine is called. See also
2472 # insert_lines(). Lines inserted via this routine are already fully
2473 # adjusted, ready to be processed; each_line_handler()s handlers will
2474 # not be called. This means this is not a completely general
2475 # facility, as only the last each_line_handler on the stack should
2476 # call this. It could be made more general, by passing to each of the
2477 # line_handlers their position on the stack, which they would pass on
2478 # to this routine, and that would replace the boolean first element in
2479 # the anonymous array pushed here, so that the next_line routine could
2480 # use that to call only those handlers whose index is after it on the
2481 # stack. But this is overkill for what is needed now.
2482
2483 my $self = shift;
2484 trace $_[0] if main::DEBUG && $to_trace;
2485
2486 # Each inserted line is an array, with the first element being 1 to
2487 # indicate that this line has been adjusted
f998e60c 2488 no overloading;
051df77b 2489 push @{$added_lines{pack 'J', $self}}, map { [ 1, $_ ] } @_;
99870f4d
KW
2490 return;
2491 }
2492
2493 sub get_missings {
2494 # Returns the stored up @missings lines' values, and clears the list.
2495 # The values are in an array, consisting of the default in the first
2496 # element, and the property in the 2nd. However, since these lines
2497 # can be stacked up, the return is an array of all these arrays.
2498
2499 my $self = shift;
2500 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2501
ffe43484 2502 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2503
2504 # If not accepting a list return, just return the first one.
2505 return shift @{$missings{$addr}} unless wantarray;
2506
2507 my @return = @{$missings{$addr}};
2508 undef @{$missings{$addr}};
2509 return @return;
2510 }
2511
2512 sub _insert_property_into_line {
2513 # Add a property field to $_, if this file requires it.
2514
f998e60c 2515 my $self = shift;
ffe43484 2516 my $addr = do { no overloading; pack 'J', $self; };
f998e60c 2517 my $property = $property{$addr};
99870f4d
KW
2518 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2519
2520 $_ =~ s/(;|$)/; $property$1/;
2521 return;
2522 }
2523
2524 sub carp_bad_line {
2525 # Output consistent error messages, using either a generic one, or the
2526 # one given by the optional parameter. To avoid gazillions of the
2527 # same message in case the syntax of a file is way off, this routine
2528 # only outputs the first instance of each message, incrementing a
2529 # count so the totals can be output at the end of the file.
2530
2531 my $self = shift;
2532 my $message = shift;
2533 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2534
ffe43484 2535 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2536
2537 $message = 'Unexpected line' unless $message;
2538
2539 # No trailing punctuation so as to fit with our addenda.
2540 $message =~ s/[.:;,]$//;
2541
2542 # If haven't seen this exact message before, output it now. Otherwise
2543 # increment the count of how many times it has occurred
2544 unless ($errors{$addr}->{$message}) {
2545 Carp::my_carp("$message in '$_' in "
f998e60c 2546 . $file{$addr}
99870f4d
KW
2547 . " at line $.. Skipping this line;");
2548 $errors{$addr}->{$message} = 1;
2549 }
2550 else {
2551 $errors{$addr}->{$message}++;
2552 }
2553
2554 # Clear the line to prevent any further (meaningful) processing of it.
2555 $_ = "";
2556
2557 return;
2558 }
2559} # End closure
2560
2561package Multi_Default;
2562
2563# Certain properties in early versions of Unicode had more than one possible
2564# default for code points missing from the files. In these cases, one
2565# default applies to everything left over after all the others are applied,
2566# and for each of the others, there is a description of which class of code
2567# points applies to it. This object helps implement this by storing the
2568# defaults, and for all but that final default, an eval string that generates
2569# the class that it applies to.
2570
2571
2572{ # Closure
2573
2574 main::setup_package();
2575
2576 my %class_defaults;
2577 # The defaults structure for the classes
2578 main::set_access('class_defaults', \%class_defaults);
2579
2580 my %other_default;
2581 # The default that applies to everything left over.
2582 main::set_access('other_default', \%other_default, 'r');
2583
2584
2585 sub new {
2586 # The constructor is called with default => eval pairs, terminated by
2587 # the left-over default. e.g.
2588 # Multi_Default->new(
2589 # 'T' => '$gc->table("Mn") + $gc->table("Cf") - 0x200C
2590 # - 0x200D',
2591 # 'R' => 'some other expression that evaluates to code points',
2592 # .
2593 # .
2594 # .
2595 # 'U'));
2596
2597 my $class = shift;
2598
2599 my $self = bless \do{my $anonymous_scalar}, $class;
ffe43484 2600 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2601
2602 while (@_ > 1) {
2603 my $default = shift;
2604 my $eval = shift;
2605 $class_defaults{$addr}->{$default} = $eval;
2606 }
2607
2608 $other_default{$addr} = shift;
2609
2610 return $self;
2611 }
2612
2613 sub get_next_defaults {
2614 # Iterates and returns the next class of defaults.
2615 my $self = shift;
2616 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2617
ffe43484 2618 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2619
2620 return each %{$class_defaults{$addr}};
2621 }
2622}
2623
2624package Alias;
2625
2626# An alias is one of the names that a table goes by. This class defines them
2627# including some attributes. Everything is currently setup in the
2628# constructor.
2629
2630
2631{ # Closure
2632
2633 main::setup_package();
2634
2635 my %name;
2636 main::set_access('name', \%name, 'r');
2637
2638 my %loose_match;
c12f2655 2639 # Should this name match loosely or not.
99870f4d
KW
2640 main::set_access('loose_match', \%loose_match, 'r');
2641
2642 my %make_pod_entry;
2643 # Some aliases should not get their own entries because they are covered
2644 # by a wild-card, and some we want to discourage use of. Binary
2645 main::set_access('make_pod_entry', \%make_pod_entry, 'r');
2646
2647 my %status;
2648 # Aliases have a status, like deprecated, or even suppressed (which means
2649 # they don't appear in documentation). Enum
2650 main::set_access('status', \%status, 'r');
2651
2652 my %externally_ok;
2653 # Similarly, some aliases should not be considered as usable ones for
2654 # external use, such as file names, or we don't want documentation to
2655 # recommend them. Boolean
2656 main::set_access('externally_ok', \%externally_ok, 'r');
2657
2658 sub new {
2659 my $class = shift;
2660
2661 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2662 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2663
2664 $name{$addr} = shift;
2665 $loose_match{$addr} = shift;
2666 $make_pod_entry{$addr} = shift;
2667 $externally_ok{$addr} = shift;
2668 $status{$addr} = shift;
2669
2670 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2671
2672 # Null names are never ok externally
2673 $externally_ok{$addr} = 0 if $name{$addr} eq "";
2674
2675 return $self;
2676 }
2677}
2678
2679package Range;
2680
2681# A range is the basic unit for storing code points, and is described in the
2682# comments at the beginning of the program. Each range has a starting code
2683# point; an ending code point (not less than the starting one); a value
2684# that applies to every code point in between the two end-points, inclusive;
2685# and an enum type that applies to the value. The type is for the user's
2686# convenience, and has no meaning here, except that a non-zero type is
2687# considered to not obey the normal Unicode rules for having standard forms.
2688#
2689# The same structure is used for both map and match tables, even though in the
2690# latter, the value (and hence type) is irrelevant and could be used as a
2691# comment. In map tables, the value is what all the code points in the range
2692# map to. Type 0 values have the standardized version of the value stored as
2693# well, so as to not have to recalculate it a lot.
2694
2695sub trace { return main::trace(@_); }
2696
2697{ # Closure
2698
2699 main::setup_package();
2700
2701 my %start;
2702 main::set_access('start', \%start, 'r', 's');
2703
2704 my %end;
2705 main::set_access('end', \%end, 'r', 's');
2706
2707 my %value;
2708 main::set_access('value', \%value, 'r');
2709
2710 my %type;
2711 main::set_access('type', \%type, 'r');
2712
2713 my %standard_form;
2714 # The value in internal standard form. Defined only if the type is 0.
2715 main::set_access('standard_form', \%standard_form);
2716
2717 # Note that if these fields change, the dump() method should as well
2718
2719 sub new {
2720 return Carp::carp_too_few_args(\@_, 3) if main::DEBUG && @_ < 3;
2721 my $class = shift;
2722
2723 my $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2724 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2725
2726 $start{$addr} = shift;
2727 $end{$addr} = shift;
2728
2729 my %args = @_;
2730
2731 my $value = delete $args{'Value'}; # Can be 0
2732 $value = "" unless defined $value;
2733 $value{$addr} = $value;
2734
2735 $type{$addr} = delete $args{'Type'} || 0;
2736
2737 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2738
2739 if (! $type{$addr}) {
2740 $standard_form{$addr} = main::standardize($value);
2741 }
2742
2743 return $self;
2744 }
2745
2746 use overload
2747 fallback => 0,
2748 qw("") => "_operator_stringify",
2749 "." => \&main::_operator_dot,
2750 ;
2751
2752 sub _operator_stringify {
2753 my $self = shift;
ffe43484 2754 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2755
2756 # Output it like '0041..0065 (value)'
2757 my $return = sprintf("%04X", $start{$addr})
2758 . '..'
2759 . sprintf("%04X", $end{$addr});
2760 my $value = $value{$addr};
2761 my $type = $type{$addr};
2762 $return .= ' (';
2763 $return .= "$value";
2764 $return .= ", Type=$type" if $type != 0;
2765 $return .= ')';
2766
2767 return $return;
2768 }
2769
2770 sub standard_form {
2771 # The standard form is the value itself if the standard form is
2772 # undefined (that is if the value is special)
2773
2774 my $self = shift;
2775 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2776
ffe43484 2777 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2778
2779 return $standard_form{$addr} if defined $standard_form{$addr};
2780 return $value{$addr};
2781 }
2782
2783 sub dump {
2784 # Human, not machine readable. For machine readable, comment out this
2785 # entire routine and let the standard one take effect.
2786 my $self = shift;
2787 my $indent = shift;
2788 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2789
ffe43484 2790 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2791
2792 my $return = $indent
2793 . sprintf("%04X", $start{$addr})
2794 . '..'
2795 . sprintf("%04X", $end{$addr})
2796 . " '$value{$addr}';";
2797 if (! defined $standard_form{$addr}) {
2798 $return .= "(type=$type{$addr})";
2799 }
2800 elsif ($standard_form{$addr} ne $value{$addr}) {
2801 $return .= "(standard '$standard_form{$addr}')";
2802 }
2803 return $return;
2804 }
2805} # End closure
2806
2807package _Range_List_Base;
2808
2809# Base class for range lists. A range list is simply an ordered list of
2810# ranges, so that the ranges with the lowest starting numbers are first in it.
2811#
2812# When a new range is added that is adjacent to an existing range that has the
2813# same value and type, it merges with it to form a larger range.
2814#
2815# Ranges generally do not overlap, except that there can be multiple entries
2816# of single code point ranges. This is because of NameAliases.txt.
2817#
2818# In this program, there is a standard value such that if two different
2819# values, have the same standard value, they are considered equivalent. This
2820# value was chosen so that it gives correct results on Unicode data
2821
2822# There are a number of methods to manipulate range lists, and some operators
2823# are overloaded to handle them.
2824
99870f4d
KW
2825sub trace { return main::trace(@_); }
2826
2827{ # Closure
2828
2829 our $addr;
2830
2831 main::setup_package();
2832
2833 my %ranges;
2834 # The list of ranges
2835 main::set_access('ranges', \%ranges, 'readable_array');
2836
2837 my %max;
2838 # The highest code point in the list. This was originally a method, but
2839 # actual measurements said it was used a lot.
2840 main::set_access('max', \%max, 'r');
2841
2842 my %each_range_iterator;
2843 # Iterator position for each_range()
2844 main::set_access('each_range_iterator', \%each_range_iterator);
2845
2846 my %owner_name_of;
2847 # Name of parent this is attached to, if any. Solely for better error
2848 # messages.
2849 main::set_access('owner_name_of', \%owner_name_of, 'p_r');
2850
2851 my %_search_ranges_cache;
2852 # A cache of the previous result from _search_ranges(), for better
2853 # performance
2854 main::set_access('_search_ranges_cache', \%_search_ranges_cache);
2855
2856 sub new {
2857 my $class = shift;
2858 my %args = @_;
2859
2860 # Optional initialization data for the range list.
2861 my $initialize = delete $args{'Initialize'};
2862
2863 my $self;
2864
2865 # Use _union() to initialize. _union() returns an object of this
2866 # class, which means that it will call this constructor recursively.
2867 # But it won't have this $initialize parameter so that it won't
2868 # infinitely loop on this.
2869 return _union($class, $initialize, %args) if defined $initialize;
2870
2871 $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2872 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2873
2874 # Optional parent object, only for debug info.
2875 $owner_name_of{$addr} = delete $args{'Owner'};
2876 $owner_name_of{$addr} = "" if ! defined $owner_name_of{$addr};
2877
2878 # Stringify, in case it is an object.
2879 $owner_name_of{$addr} = "$owner_name_of{$addr}";
2880
2881 # This is used only for error messages, and so a colon is added
2882 $owner_name_of{$addr} .= ": " if $owner_name_of{$addr} ne "";
2883
2884 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2885
2886 # Max is initialized to a negative value that isn't adjacent to 0,
2887 # for simpler tests
2888 $max{$addr} = -2;
2889
2890 $_search_ranges_cache{$addr} = 0;
2891 $ranges{$addr} = [];
2892
2893 return $self;
2894 }
2895
2896 use overload
2897 fallback => 0,
2898 qw("") => "_operator_stringify",
2899 "." => \&main::_operator_dot,
2900 ;
2901
2902 sub _operator_stringify {
2903 my $self = shift;
ffe43484 2904 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2905
2906 return "Range_List attached to '$owner_name_of{$addr}'"
2907 if $owner_name_of{$addr};
2908 return "anonymous Range_List " . \$self;
2909 }
2910
2911 sub _union {
2912 # Returns the union of the input code points. It can be called as
2913 # either a constructor or a method. If called as a method, the result
2914 # will be a new() instance of the calling object, containing the union
2915 # of that object with the other parameter's code points; if called as
2916 # a constructor, the first parameter gives the class the new object
2917 # should be, and the second parameter gives the code points to go into
2918 # it.
2919 # In either case, there are two parameters looked at by this routine;
2920 # any additional parameters are passed to the new() constructor.
2921 #
2922 # The code points can come in the form of some object that contains
2923 # ranges, and has a conventionally named method to access them; or
2924 # they can be an array of individual code points (as integers); or
2925 # just a single code point.
2926 #
2927 # If they are ranges, this routine doesn't make any effort to preserve
2928 # the range values of one input over the other. Therefore this base
2929 # class should not allow _union to be called from other than
2930 # initialization code, so as to prevent two tables from being added
2931 # together where the range values matter. The general form of this
2932 # routine therefore belongs in a derived class, but it was moved here
2933 # to avoid duplication of code. The failure to overload this in this
2934 # class keeps it safe.
2935 #
2936
2937 my $self;
2938 my @args; # Arguments to pass to the constructor
2939
2940 my $class = shift;
2941
2942 # If a method call, will start the union with the object itself, and
2943 # the class of the new object will be the same as self.
2944 if (ref $class) {
2945 $self = $class;
2946 $class = ref $self;
2947 push @args, $self;
2948 }
2949
2950 # Add the other required parameter.
2951 push @args, shift;
2952 # Rest of parameters are passed on to the constructor
2953
2954 # Accumulate all records from both lists.
2955 my @records;
2956 for my $arg (@args) {
2957 #local $to_trace = 0 if main::DEBUG;
2958 trace "argument = $arg" if main::DEBUG && $to_trace;
2959 if (! defined $arg) {
2960 my $message = "";
2961 if (defined $self) {
f998e60c 2962 no overloading;
051df77b 2963 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
2964 }
2965 Carp::my_carp_bug($message .= "Undefined argument to _union. No union done.");
2966 return;
2967 }
2968 $arg = [ $arg ] if ! ref $arg;
2969 my $type = ref $arg;
2970 if ($type eq 'ARRAY') {
2971 foreach my $element (@$arg) {
2972 push @records, Range->new($element, $element);
2973 }
2974 }
2975 elsif ($arg->isa('Range')) {
2976 push @records, $arg;
2977 }
2978 elsif ($arg->can('ranges')) {
2979 push @records, $arg->ranges;
2980 }
2981 else {
2982 my $message = "";
2983 if (defined $self) {
f998e60c 2984 no overloading;
051df77b 2985 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
2986 }
2987 Carp::my_carp_bug($message . "Cannot take the union of a $type. No union done.");
2988 return;
2989 }
2990 }
2991
2992 # Sort with the range containing the lowest ordinal first, but if
2993 # two ranges start at the same code point, sort with the bigger range
2994 # of the two first, because it takes fewer cycles.
2995 @records = sort { ($a->start <=> $b->start)
2996 or
2997 # if b is shorter than a, b->end will be
2998 # less than a->end, and we want to select
2999 # a, so want to return -1
3000 ($b->end <=> $a->end)
3001 } @records;
3002
3003 my $new = $class->new(@_);
3004
3005 # Fold in records so long as they add new information.
3006 for my $set (@records) {
3007 my $start = $set->start;
3008 my $end = $set->end;
3009 my $value = $set->value;
3010 if ($start > $new->max) {
3011 $new->_add_delete('+', $start, $end, $value);
3012 }
3013 elsif ($end > $new->max) {
3014 $new->_add_delete('+', $new->max +1, $end, $value);
3015 }
3016 }
3017
3018 return $new;
3019 }
3020
3021 sub range_count { # Return the number of ranges in the range list
3022 my $self = shift;
3023 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3024
f998e60c 3025 no overloading;
051df77b 3026 return scalar @{$ranges{pack 'J', $self}};
99870f4d
KW
3027 }
3028
3029 sub min {
3030 # Returns the minimum code point currently in the range list, or if
3031 # the range list is empty, 2 beyond the max possible. This is a
3032 # method because used so rarely, that not worth saving between calls,
3033 # and having to worry about changing it as ranges are added and
3034 # deleted.
3035
3036 my $self = shift;
3037 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3038
ffe43484 3039 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3040
3041 # If the range list is empty, return a large value that isn't adjacent
3042 # to any that could be in the range list, for simpler tests
6189eadc 3043 return $MAX_UNICODE_CODEPOINT + 2 unless scalar @{$ranges{$addr}};
99870f4d
KW
3044 return $ranges{$addr}->[0]->start;
3045 }
3046
3047 sub contains {
3048 # Boolean: Is argument in the range list? If so returns $i such that:
3049 # range[$i]->end < $codepoint <= range[$i+1]->end
3050 # which is one beyond what you want; this is so that the 0th range
3051 # doesn't return false
3052 my $self = shift;
3053 my $codepoint = shift;
3054 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3055
99870f4d
KW
3056 my $i = $self->_search_ranges($codepoint);
3057 return 0 unless defined $i;
3058
3059 # The search returns $i, such that
3060 # range[$i-1]->end < $codepoint <= range[$i]->end
3061 # So is in the table if and only iff it is at least the start position
3062 # of range $i.
f998e60c 3063 no overloading;
051df77b 3064 return 0 if $ranges{pack 'J', $self}->[$i]->start > $codepoint;
99870f4d
KW
3065 return $i + 1;
3066 }
3067
2f7a8815
KW
3068 sub containing_range {
3069 # Returns the range object that contains the code point, undef if none
3070
3071 my $self = shift;
3072 my $codepoint = shift;
3073 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3074
3075 my $i = $self->contains($codepoint);
3076 return unless $i;
3077
3078 # contains() returns 1 beyond where we should look
3079 no overloading;
3080 return $ranges{pack 'J', $self}->[$i-1];
3081 }
3082
99870f4d
KW
3083 sub value_of {
3084 # Returns the value associated with the code point, undef if none
3085
3086 my $self = shift;
3087 my $codepoint = shift;
3088 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3089
d69c231b
KW
3090 my $range = $self->containing_range($codepoint);
3091 return unless defined $range;
99870f4d 3092
d69c231b 3093 return $range->value;
99870f4d
KW
3094 }
3095
0a9dbafc
KW
3096 sub type_of {
3097 # Returns the type of the range containing the code point, undef if
3098 # the code point is not in the table
3099
3100 my $self = shift;
3101 my $codepoint = shift;
3102 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3103
3104 my $range = $self->containing_range($codepoint);
3105 return unless defined $range;
3106
3107 return $range->type;
3108 }
3109
99870f4d
KW
3110 sub _search_ranges {
3111 # Find the range in the list which contains a code point, or where it
3112 # should go if were to add it. That is, it returns $i, such that:
3113 # range[$i-1]->end < $codepoint <= range[$i]->end
3114 # Returns undef if no such $i is possible (e.g. at end of table), or
3115 # if there is an error.
3116
3117 my $self = shift;
3118 my $code_point = shift;
3119 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3120
ffe43484 3121 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3122
3123 return if $code_point > $max{$addr};
3124 my $r = $ranges{$addr}; # The current list of ranges
3125 my $range_list_size = scalar @$r;
3126 my $i;
3127
3128 use integer; # want integer division
3129
3130 # Use the cached result as the starting guess for this one, because,
3131 # an experiment on 5.1 showed that 90% of the time the cache was the
3132 # same as the result on the next call (and 7% it was one less).
3133 $i = $_search_ranges_cache{$addr};
3134 $i = 0 if $i >= $range_list_size; # Reset if no longer valid (prob.
3135 # from an intervening deletion
3136 #local $to_trace = 1 if main::DEBUG;
3137 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);
3138 return $i if $code_point <= $r->[$i]->end
3139 && ($i == 0 || $r->[$i-1]->end < $code_point);
3140
3141 # Here the cache doesn't yield the correct $i. Try adding 1.
3142 if ($i < $range_list_size - 1
3143 && $r->[$i]->end < $code_point &&
3144 $code_point <= $r->[$i+1]->end)
3145 {
3146 $i++;
3147 trace "next \$i is correct: $i" if main::DEBUG && $to_trace;
3148 $_search_ranges_cache{$addr} = $i;
3149 return $i;
3150 }
3151
3152 # Here, adding 1 also didn't work. We do a binary search to
3153 # find the correct position, starting with current $i
3154 my $lower = 0;
3155 my $upper = $range_list_size - 1;
3156 while (1) {
3157 trace "top of loop i=$i:", sprintf("%04X", $r->[$lower]->start), "[$lower] .. ", sprintf("%04X", $r->[$i]->start), "[$i] .. ", sprintf("%04X", $r->[$upper]->start), "[$upper]" if main::DEBUG && $to_trace;
3158
3159 if ($code_point <= $r->[$i]->end) {
3160
3161 # Here we have met the upper constraint. We can quit if we
3162 # also meet the lower one.
3163 last if $i == 0 || $r->[$i-1]->end < $code_point;
3164
3165 $upper = $i; # Still too high.
3166
3167 }
3168 else {
3169
3170 # Here, $r[$i]->end < $code_point, so look higher up.
3171 $lower = $i;
3172 }
3173
3174 # Split search domain in half to try again.
3175 my $temp = ($upper + $lower) / 2;
3176
3177 # No point in continuing unless $i changes for next time
3178 # in the loop.
3179 if ($temp == $i) {
3180
3181 # We can't reach the highest element because of the averaging.
3182 # So if one below the upper edge, force it there and try one
3183 # more time.
3184 if ($i == $range_list_size - 2) {
3185
3186 trace "Forcing to upper edge" if main::DEBUG && $to_trace;
3187 $i = $range_list_size - 1;
3188
3189 # Change $lower as well so if fails next time through,
3190 # taking the average will yield the same $i, and we will
3191 # quit with the error message just below.
3192 $lower = $i;
3193 next;
3194 }
3195 Carp::my_carp_bug("$owner_name_of{$addr}Can't find where the range ought to go. No action taken.");
3196 return;
3197 }
3198 $i = $temp;
3199 } # End of while loop
3200
3201 if (main::DEBUG && $to_trace) {
3202 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i;
3203 trace "i= [ $i ]", $r->[$i];
3204 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < $range_list_size - 1;
3205 }
3206
3207 # Here we have found the offset. Cache it as a starting point for the
3208 # next call.
3209 $_search_ranges_cache{$addr} = $i;
3210 return $i;
3211 }
3212
3213 sub _add_delete {
3214 # Add, replace or delete ranges to or from a list. The $type
3215 # parameter gives which:
3216 # '+' => insert or replace a range, returning a list of any changed
3217 # ranges.
3218 # '-' => delete a range, returning a list of any deleted ranges.
3219 #
3220 # The next three parameters give respectively the start, end, and
3221 # value associated with the range. 'value' should be null unless the
3222 # operation is '+';
3223 #
3224 # The range list is kept sorted so that the range with the lowest
3225 # starting position is first in the list, and generally, adjacent
c1739a4a 3226 # ranges with the same values are merged into a single larger one (see
99870f4d
KW
3227 # exceptions below).
3228 #
c1739a4a 3229 # There are more parameters; all are key => value pairs:
99870f4d
KW
3230 # Type gives the type of the value. It is only valid for '+'.
3231 # All ranges have types; if this parameter is omitted, 0 is
3232 # assumed. Ranges with type 0 are assumed to obey the
3233 # Unicode rules for casing, etc; ranges with other types are
3234 # not. Otherwise, the type is arbitrary, for the caller's
3235 # convenience, and looked at only by this routine to keep
3236 # adjacent ranges of different types from being merged into
3237 # a single larger range, and when Replace =>
3238 # $IF_NOT_EQUIVALENT is specified (see just below).
3239 # Replace determines what to do if the range list already contains
3240 # ranges which coincide with all or portions of the input
3241 # range. It is only valid for '+':
3242 # => $NO means that the new value is not to replace
3243 # any existing ones, but any empty gaps of the
3244 # range list coinciding with the input range
3245 # will be filled in with the new value.
3246 # => $UNCONDITIONALLY means to replace the existing values with
3247 # this one unconditionally. However, if the
3248 # new and old values are identical, the
3249 # replacement is skipped to save cycles
3250 # => $IF_NOT_EQUIVALENT means to replace the existing values
3251 # with this one if they are not equivalent.
3252 # Ranges are equivalent if their types are the
c1739a4a 3253 # same, and they are the same string; or if
99870f4d
KW
3254 # both are type 0 ranges, if their Unicode
3255 # standard forms are identical. In this last
3256 # case, the routine chooses the more "modern"
3257 # one to use. This is because some of the
3258 # older files are formatted with values that
3259 # are, for example, ALL CAPs, whereas the
3260 # derived files have a more modern style,
3261 # which looks better. By looking for this
3262 # style when the pre-existing and replacement
3263 # standard forms are the same, we can move to
3264 # the modern style
3265 # => $MULTIPLE means that if this range duplicates an
3266 # existing one, but has a different value,
3267 # don't replace the existing one, but insert
3268 # this, one so that the same range can occur
53d84487
KW
3269 # multiple times. They are stored LIFO, so
3270 # that the final one inserted is the first one
3271 # returned in an ordered search of the table.
99870f4d
KW
3272 # => anything else is the same as => $IF_NOT_EQUIVALENT
3273 #
c1739a4a
KW
3274 # "same value" means identical for non-type-0 ranges, and it means
3275 # having the same standard forms for type-0 ranges.
99870f4d
KW
3276
3277 return Carp::carp_too_few_args(\@_, 5) if main::DEBUG && @_ < 5;
3278
3279 my $self = shift;
3280 my $operation = shift; # '+' for add/replace; '-' for delete;
3281 my $start = shift;
3282 my $end = shift;
3283 my $value = shift;
3284
3285 my %args = @_;
3286
3287 $value = "" if not defined $value; # warning: $value can be "0"
3288
3289 my $replace = delete $args{'Replace'};
3290 $replace = $IF_NOT_EQUIVALENT unless defined $replace;
3291
3292 my $type = delete $args{'Type'};
3293 $type = 0 unless defined $type;
3294
3295 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3296
ffe43484 3297 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3298
3299 if ($operation ne '+' && $operation ne '-') {
3300 Carp::my_carp_bug("$owner_name_of{$addr}First parameter to _add_delete must be '+' or '-'. No action taken.");
3301 return;
3302 }
3303 unless (defined $start && defined $end) {
3304 Carp::my_carp_bug("$owner_name_of{$addr}Undefined start and/or end to _add_delete. No action taken.");
3305 return;
3306 }
3307 unless ($end >= $start) {
3308 Carp::my_carp_bug("$owner_name_of{$addr}End of range (" . sprintf("%04X", $end) . ") must not be before start (" . sprintf("%04X", $start) . "). No action taken.");
3309 return;
3310 }
3311 #local $to_trace = 1 if main::DEBUG;
3312
3313 if ($operation eq '-') {
3314 if ($replace != $IF_NOT_EQUIVALENT) {
3315 Carp::my_carp_bug("$owner_name_of{$addr}Replace => \$IF_NOT_EQUIVALENT is required when deleting a range from a range list. Assuming Replace => \$IF_NOT_EQUIVALENT.");
3316 $replace = $IF_NOT_EQUIVALENT;
3317 }
3318 if ($type) {
3319 Carp::my_carp_bug("$owner_name_of{$addr}Type => 0 is required when deleting a range from a range list. Assuming Type => 0.");
3320 $type = 0;
3321 }
3322 if ($value ne "") {
3323 Carp::my_carp_bug("$owner_name_of{$addr}Value => \"\" is required when deleting a range from a range list. Assuming Value => \"\".");
3324 $value = "";
3325 }
3326 }
3327
3328 my $r = $ranges{$addr}; # The current list of ranges
3329 my $range_list_size = scalar @$r; # And its size
3330 my $max = $max{$addr}; # The current high code point in
3331 # the list of ranges
3332
3333 # Do a special case requiring fewer machine cycles when the new range
3334 # starts after the current highest point. The Unicode input data is
3335 # structured so this is common.
3336 if ($start > $max) {
3337
3338 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) type=$type" if main::DEBUG && $to_trace;
3339 return if $operation eq '-'; # Deleting a non-existing range is a
3340 # no-op
3341
3342 # If the new range doesn't logically extend the current final one
3343 # in the range list, create a new range at the end of the range
3344 # list. (max cleverly is initialized to a negative number not
3345 # adjacent to 0 if the range list is empty, so even adding a range
3346 # to an empty range list starting at 0 will have this 'if'
3347 # succeed.)
3348 if ($start > $max + 1 # non-adjacent means can't extend.
3349 || @{$r}[-1]->value ne $value # values differ, can't extend.
3350 || @{$r}[-1]->type != $type # types differ, can't extend.
3351 ) {
3352 push @$r, Range->new($start, $end,
3353 Value => $value,
3354 Type => $type);
3355 }
3356 else {
3357
3358 # Here, the new range starts just after the current highest in
3359 # the range list, and they have the same type and value.
3360 # Extend the current range to incorporate the new one.
3361 @{$r}[-1]->set_end($end);
3362 }
3363
3364 # This becomes the new maximum.
3365 $max{$addr} = $end;
3366
3367 return;
3368 }
3369 #local $to_trace = 0 if main::DEBUG;
3370
3371 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) replace=$replace" if main::DEBUG && $to_trace;
3372
3373 # Here, the input range isn't after the whole rest of the range list.
3374 # Most likely 'splice' will be needed. The rest of the routine finds
3375 # the needed splice parameters, and if necessary, does the splice.
3376 # First, find the offset parameter needed by the splice function for
3377 # the input range. Note that the input range may span multiple
3378 # existing ones, but we'll worry about that later. For now, just find
3379 # the beginning. If the input range is to be inserted starting in a
3380 # position not currently in the range list, it must (obviously) come
3381 # just after the range below it, and just before the range above it.
3382 # Slightly less obviously, it will occupy the position currently
3383 # occupied by the range that is to come after it. More formally, we
3384 # are looking for the position, $i, in the array of ranges, such that:
3385 #
3386 # r[$i-1]->start <= r[$i-1]->end < $start < r[$i]->start <= r[$i]->end
3387 #
3388 # (The ordered relationships within existing ranges are also shown in
3389 # the equation above). However, if the start of the input range is
3390 # within an existing range, the splice offset should point to that
3391 # existing range's position in the list; that is $i satisfies a
3392 # somewhat different equation, namely:
3393 #
3394 #r[$i-1]->start <= r[$i-1]->end < r[$i]->start <= $start <= r[$i]->end
3395 #
3396 # More briefly, $start can come before or after r[$i]->start, and at
3397 # this point, we don't know which it will be. However, these
3398 # two equations share these constraints:
3399 #
3400 # r[$i-1]->end < $start <= r[$i]->end
3401 #
3402 # And that is good enough to find $i.
3403
3404 my $i = $self->_search_ranges($start);
3405 if (! defined $i) {
3406 Carp::my_carp_bug("Searching $self for range beginning with $start unexpectedly returned undefined. Operation '$operation' not performed");
3407 return;
3408 }
3409
3410 # The search function returns $i such that:
3411 #
3412 # r[$i-1]->end < $start <= r[$i]->end
3413 #
3414 # That means that $i points to the first range in the range list
3415 # that could possibly be affected by this operation. We still don't
3416 # know if the start of the input range is within r[$i], or if it
3417 # points to empty space between r[$i-1] and r[$i].
3418 trace "[$i] is the beginning splice point. Existing range there is ", $r->[$i] if main::DEBUG && $to_trace;
3419
3420 # Special case the insertion of data that is not to replace any
3421 # existing data.
3422 if ($replace == $NO) { # If $NO, has to be operation '+'
3423 #local $to_trace = 1 if main::DEBUG;
3424 trace "Doesn't replace" if main::DEBUG && $to_trace;
3425
3426 # Here, the new range is to take effect only on those code points
3427 # that aren't already in an existing range. This can be done by
3428 # looking through the existing range list and finding the gaps in
3429 # the ranges that this new range affects, and then calling this
3430 # function recursively on each of those gaps, leaving untouched
3431 # anything already in the list. Gather up a list of the changed
3432 # gaps first so that changes to the internal state as new ranges
3433 # are added won't be a problem.
3434 my @gap_list;
3435
3436 # First, if the starting point of the input range is outside an
3437 # existing one, there is a gap from there to the beginning of the
3438 # existing range -- add a span to fill the part that this new
3439 # range occupies
3440 if ($start < $r->[$i]->start) {
3441 push @gap_list, Range->new($start,
3442 main::min($end,
3443 $r->[$i]->start - 1),
3444 Type => $type);
3445 trace "gap before $r->[$i] [$i], will add", $gap_list[-1] if main::DEBUG && $to_trace;
3446 }
3447
3448 # Then look through the range list for other gaps until we reach
3449 # the highest range affected by the input one.
3450 my $j;
3451 for ($j = $i+1; $j < $range_list_size; $j++) {
3452 trace "j=[$j]", $r->[$j] if main::DEBUG && $to_trace;
3453 last if $end < $r->[$j]->start;
3454
3455 # If there is a gap between when this range starts and the
3456 # previous one ends, add a span to fill it. Note that just
3457 # because there are two ranges doesn't mean there is a
3458 # non-zero gap between them. It could be that they have
3459 # different values or types
3460 if ($r->[$j-1]->end + 1 != $r->[$j]->start) {
3461 push @gap_list,
3462 Range->new($r->[$j-1]->end + 1,
3463 $r->[$j]->start - 1,
3464 Type => $type);
3465 trace "gap between $r->[$j-1] and $r->[$j] [$j], will add: $gap_list[-1]" if main::DEBUG && $to_trace;
3466 }
3467 }
3468
3469 # Here, we have either found an existing range in the range list,
3470 # beyond the area affected by the input one, or we fell off the
3471 # end of the loop because the input range affects the whole rest
3472 # of the range list. In either case, $j is 1 higher than the
3473 # highest affected range. If $j == $i, it means that there are no
3474 # affected ranges, that the entire insertion is in the gap between
3475 # r[$i-1], and r[$i], which we already have taken care of before
3476 # the loop.
3477 # On the other hand, if there are affected ranges, it might be
3478 # that there is a gap that needs filling after the final such
3479 # range to the end of the input range
3480 if ($r->[$j-1]->end < $end) {
3481 push @gap_list, Range->new(main::max($start,
3482 $r->[$j-1]->end + 1),
3483 $end,
3484 Type => $type);
3485 trace "gap after $r->[$j-1], will add $gap_list[-1]" if main::DEBUG && $to_trace;
3486 }
3487
3488 # Call recursively to fill in all the gaps.
3489 foreach my $gap (@gap_list) {
3490 $self->_add_delete($operation,
3491 $gap->start,
3492 $gap->end,
3493 $value,
3494 Type => $type);
3495 }
3496
3497 return;
3498 }
3499
53d84487
KW
3500 # Here, we have taken care of the case where $replace is $NO.
3501 # Remember that here, r[$i-1]->end < $start <= r[$i]->end
3502 # If inserting a multiple record, this is where it goes, before the
3503 # first (if any) existing one. This implies an insertion, and no
3504 # change to any existing ranges. Note that $i can be -1 if this new
3505 # range doesn't actually duplicate any existing, and comes at the
3506 # beginning of the list.
3507 if ($replace == $MULTIPLE) {
3508
3509 if ($start != $end) {
3510 Carp::my_carp_bug("$owner_name_of{$addr}Can't cope with adding a multiple record when the range ($start..$end) contains more than one code point. No action taken.");
3511 return;
3512 }
3513
3514 # Don't add an exact duplicate, as it isn't really a multiple
3515 if ($end >= $r->[$i]->start) {
1f6798c4
KW
3516 my $existing_value = $r->[$i]->value;
3517 my $existing_type = $r->[$i]->type;
3518 return if $value eq $existing_value && $type eq $existing_type;
3519
3520 # If the multiple value is part of an existing range, we want
3521 # to split up that range, so that only the single code point
3522 # is affected. To do this, we first call ourselves
3523 # recursively to delete that code point from the table, having
3524 # preserved its current data above. Then we call ourselves
3525 # recursively again to add the new multiple, which we know by
3526 # the test just above is different than the current code
3527 # point's value, so it will become a range containing a single
3528 # code point: just itself. Finally, we add back in the
3529 # pre-existing code point, which will again be a single code
3530 # point range. Because 'i' likely will have changed as a
3531 # result of these operations, we can't just continue on, but
3532 # do this operation recursively as well.
53d84487 3533 if ($r->[$i]->start != $r->[$i]->end) {
1f6798c4
KW
3534 $self->_add_delete('-', $start, $end, "");
3535 $self->_add_delete('+', $start, $end, $value, Type => $type);
3536 return $self->_add_delete('+', $start, $end, $existing_value, Type => $existing_type, Replace => $MULTIPLE);
53d84487 3537 }
53d84487
KW
3538 }
3539
3540 trace "Adding multiple record at $i with $start..$end, $value" if main::DEBUG && $to_trace;
3541 my @return = splice @$r,
3542 $i,
3543 0,
3544 Range->new($start,
3545 $end,
3546 Value => $value,
3547 Type => $type);
3548 if (main::DEBUG && $to_trace) {
3549 trace "After splice:";
3550 trace 'i-2=[', $i-2, ']', $r->[$i-2] if $i >= 2;
3551 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i >= 1;
3552 trace "i =[", $i, "]", $r->[$i] if $i >= 0;
3553 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < @$r - 1;
3554 trace 'i+2=[', $i+2, ']', $r->[$i+2] if $i < @$r - 2;
3555 trace 'i+3=[', $i+3, ']', $r->[$i+3] if $i < @$r - 3;
3556 }
3557 return @return;
3558 }
3559
3560 # Here, we have taken care of $NO and $MULTIPLE replaces. This leaves
3561 # delete, insert, and replace either unconditionally or if not
3562 # equivalent. $i still points to the first potential affected range.
3563 # Now find the highest range affected, which will determine the length
3564 # parameter to splice. (The input range can span multiple existing
3565 # ones.) If this isn't a deletion, while we are looking through the
3566 # range list, see also if this is a replacement rather than a clean
3567 # insertion; that is if it will change the values of at least one
3568 # existing range. Start off assuming it is an insert, until find it
3569 # isn't.
3570 my $clean_insert = $operation eq '+';
99870f4d
KW
3571 my $j; # This will point to the highest affected range
3572
3573 # For non-zero types, the standard form is the value itself;
3574 my $standard_form = ($type) ? $value : main::standardize($value);
3575
3576 for ($j = $i; $j < $range_list_size; $j++) {
3577 trace "Looking for highest affected range; the one at $j is ", $r->[$j] if main::DEBUG && $to_trace;
3578
3579 # If find a range that it doesn't overlap into, we can stop
3580 # searching
3581 last if $end < $r->[$j]->start;
3582
969a34cc
KW
3583 # Here, overlaps the range at $j. If the values don't match,
3584 # and so far we think this is a clean insertion, it becomes a
3585 # non-clean insertion, i.e., a 'change' or 'replace' instead.
3586 if ($clean_insert) {
99870f4d 3587 if ($r->[$j]->standard_form ne $standard_form) {
969a34cc 3588 $clean_insert = 0;
56343c78
KW
3589 if ($replace == $CROAK) {
3590 main::croak("The range to add "
3591 . sprintf("%04X", $start)
3592 . '-'
3593 . sprintf("%04X", $end)
3594 . " with value '$value' overlaps an existing range $r->[$j]");
3595 }
99870f4d
KW
3596 }
3597 else {
3598
3599 # Here, the two values are essentially the same. If the
3600 # two are actually identical, replacing wouldn't change
3601 # anything so skip it.
3602 my $pre_existing = $r->[$j]->value;
3603 if ($pre_existing ne $value) {
3604
3605 # Here the new and old standardized values are the
3606 # same, but the non-standardized values aren't. If
3607 # replacing unconditionally, then replace
3608 if( $replace == $UNCONDITIONALLY) {
969a34cc 3609 $clean_insert = 0;
99870f4d
KW
3610 }
3611 else {
3612
3613 # Here, are replacing conditionally. Decide to
3614 # replace or not based on which appears to look
3615 # the "nicest". If one is mixed case and the
3616 # other isn't, choose the mixed case one.
3617 my $new_mixed = $value =~ /[A-Z]/
3618 && $value =~ /[a-z]/;
3619 my $old_mixed = $pre_existing =~ /[A-Z]/
3620 && $pre_existing =~ /[a-z]/;
3621
3622 if ($old_mixed != $new_mixed) {
969a34cc 3623 $clean_insert = 0 if $new_mixed;
99870f4d 3624 if (main::DEBUG && $to_trace) {
969a34cc
KW
3625 if ($clean_insert) {
3626 trace "Retaining $pre_existing over $value";
99870f4d
KW
3627 }
3628 else {
969a34cc 3629 trace "Replacing $pre_existing with $value";
99870f4d
KW
3630 }
3631 }
3632 }
3633 else {
3634
3635 # Here casing wasn't different between the two.
3636 # If one has hyphens or underscores and the
3637 # other doesn't, choose the one with the
3638 # punctuation.
3639 my $new_punct = $value =~ /[-_]/;
3640 my $old_punct = $pre_existing =~ /[-_]/;
3641
3642 if ($old_punct != $new_punct) {
969a34cc 3643 $clean_insert = 0 if $new_punct;
99870f4d 3644 if (main::DEBUG && $to_trace) {
969a34cc
KW
3645 if ($clean_insert) {
3646 trace "Retaining $pre_existing over $value";
99870f4d
KW
3647 }
3648 else {
969a34cc 3649 trace "Replacing $pre_existing with $value";
99870f4d
KW
3650 }
3651 }
3652 } # else existing one is just as "good";
3653 # retain it to save cycles.
3654 }
3655 }
3656 }
3657 }
3658 }
3659 } # End of loop looking for highest affected range.
3660
3661 # Here, $j points to one beyond the highest range that this insertion
3662 # affects (hence to beyond the range list if that range is the final
3663 # one in the range list).
3664
3665 # The splice length is all the affected ranges. Get it before
3666 # subtracting, for efficiency, so we don't have to later add 1.
3667 my $length = $j - $i;
3668
3669 $j--; # $j now points to the highest affected range.
3670 trace "Final affected range is $j: $r->[$j]" if main::DEBUG && $to_trace;
3671
99870f4d
KW
3672 # Here, have taken care of $NO and $MULTIPLE replaces.
3673 # $j points to the highest affected range. But it can be < $i or even
3674 # -1. These happen only if the insertion is entirely in the gap
3675 # between r[$i-1] and r[$i]. Here's why: j < i means that the j loop
3676 # above exited first time through with $end < $r->[$i]->start. (And
3677 # then we subtracted one from j) This implies also that $start <
3678 # $r->[$i]->start, but we know from above that $r->[$i-1]->end <
3679 # $start, so the entire input range is in the gap.
3680 if ($j < $i) {
3681
3682 # Here the entire input range is in the gap before $i.
3683
3684 if (main::DEBUG && $to_trace) {
3685 if ($i) {
3686 trace "Entire range is between $r->[$i-1] and $r->[$i]";
3687 }
3688 else {
3689 trace "Entire range is before $r->[$i]";
3690 }
3691 }
3692 return if $operation ne '+'; # Deletion of a non-existent range is
3693 # a no-op
3694 }
3695 else {
3696
969a34cc
KW
3697 # Here part of the input range is not in the gap before $i. Thus,
3698 # there is at least one affected one, and $j points to the highest
3699 # such one.
99870f4d
KW
3700
3701 # At this point, here is the situation:
3702 # This is not an insertion of a multiple, nor of tentative ($NO)
3703 # data.
3704 # $i points to the first element in the current range list that
3705 # may be affected by this operation. In fact, we know
3706 # that the range at $i is affected because we are in
3707 # the else branch of this 'if'
3708 # $j points to the highest affected range.
3709 # In other words,
3710 # r[$i-1]->end < $start <= r[$i]->end
3711 # And:
3712 # r[$i-1]->end < $start <= $end <= r[$j]->end
3713 #
3714 # Also:
969a34cc
KW
3715 # $clean_insert is a boolean which is set true if and only if
3716 # this is a "clean insertion", i.e., not a change nor a
3717 # deletion (multiple was handled above).
99870f4d
KW
3718
3719 # We now have enough information to decide if this call is a no-op
969a34cc
KW
3720 # or not. It is a no-op if this is an insertion of already
3721 # existing data.
99870f4d 3722
969a34cc 3723 if (main::DEBUG && $to_trace && $clean_insert
99870f4d
KW
3724 && $i == $j
3725 && $start >= $r->[$i]->start)
3726 {
3727 trace "no-op";
3728 }
969a34cc 3729 return if $clean_insert
99870f4d
KW
3730 && $i == $j # more than one affected range => not no-op
3731
3732 # Here, r[$i-1]->end < $start <= $end <= r[$i]->end
3733 # Further, $start and/or $end is >= r[$i]->start
3734 # The test below hence guarantees that
3735 # r[$i]->start < $start <= $end <= r[$i]->end
3736 # This means the input range is contained entirely in
3737 # the one at $i, so is a no-op
3738 && $start >= $r->[$i]->start;
3739 }
3740
3741 # Here, we know that some action will have to be taken. We have
3742 # calculated the offset and length (though adjustments may be needed)
3743 # for the splice. Now start constructing the replacement list.
3744 my @replacement;
3745 my $splice_start = $i;
3746
3747 my $extends_below;
3748 my $extends_above;
3749
3750 # See if should extend any adjacent ranges.
3751 if ($operation eq '-') { # Don't extend deletions
3752 $extends_below = $extends_above = 0;
3753 }
3754 else { # Here, should extend any adjacent ranges. See if there are
3755 # any.
3756 $extends_below = ($i > 0
3757 # can't extend unless adjacent
3758 && $r->[$i-1]->end == $start -1
3759 # can't extend unless are same standard value
3760 && $r->[$i-1]->standard_form eq $standard_form
3761 # can't extend unless share type
3762 && $r->[$i-1]->type == $type);
3763 $extends_above = ($j+1 < $range_list_size
3764 && $r->[$j+1]->start == $end +1
3765 && $r->[$j+1]->standard_form eq $standard_form
23822bda 3766 && $r->[$j+1]->type == $type);
99870f4d
KW
3767 }
3768 if ($extends_below && $extends_above) { # Adds to both
3769 $splice_start--; # start replace at element below
3770 $length += 2; # will replace on both sides
3771 trace "Extends both below and above ranges" if main::DEBUG && $to_trace;
3772
3773 # The result will fill in any gap, replacing both sides, and
3774 # create one large range.
3775 @replacement = Range->new($r->[$i-1]->start,
3776 $r->[$j+1]->end,
3777 Value => $value,
3778 Type => $type);
3779 }
3780 else {
3781
3782 # Here we know that the result won't just be the conglomeration of
3783 # a new range with both its adjacent neighbors. But it could
3784 # extend one of them.
3785
3786 if ($extends_below) {
3787
3788 # Here the new element adds to the one below, but not to the
3789 # one above. If inserting, and only to that one range, can
3790 # just change its ending to include the new one.
969a34cc 3791 if ($length == 0 && $clean_insert) {
99870f4d
KW
3792 $r->[$i-1]->set_end($end);
3793 trace "inserted range extends range to below so it is now $r->[$i-1]" if main::DEBUG && $to_trace;
3794 return;
3795 }
3796 else {
3797 trace "Changing inserted range to start at ", sprintf("%04X", $r->[$i-1]->start), " instead of ", sprintf("%04X", $start) if main::DEBUG && $to_trace;
3798 $splice_start--; # start replace at element below
3799 $length++; # will replace the element below
3800 $start = $r->[$i-1]->start;
3801 }
3802 }
3803 elsif ($extends_above) {
3804
3805 # Here the new element adds to the one above, but not below.
3806 # Mirror the code above
969a34cc 3807 if ($length == 0 && $clean_insert) {
99870f4d
KW
3808 $r->[$j+1]->set_start($start);
3809 trace "inserted range extends range to above so it is now $r->[$j+1]" if main::DEBUG && $to_trace;
3810 return;
3811 }
3812 else {
3813 trace "Changing inserted range to end at ", sprintf("%04X", $r->[$j+1]->end), " instead of ", sprintf("%04X", $end) if main::DEBUG && $to_trace;
3814 $length++; # will replace the element above
3815 $end = $r->[$j+1]->end;
3816 }
3817 }
3818
3819 trace "Range at $i is $r->[$i]" if main::DEBUG && $to_trace;
3820
3821 # Finally, here we know there will have to be a splice.
3822 # If the change or delete affects only the highest portion of the
3823 # first affected range, the range will have to be split. The
3824 # splice will remove the whole range, but will replace it by a new
3825 # range containing just the unaffected part. So, in this case,
3826 # add to the replacement list just this unaffected portion.
3827 if (! $extends_below
3828 && $start > $r->[$i]->start && $start <= $r->[$i]->end)
3829 {
3830 push @replacement,
3831 Range->new($r->[$i]->start,
3832 $start - 1,
3833 Value => $r->[$i]->value,
3834 Type => $r->[$i]->type);
3835 }
3836
3837 # In the case of an insert or change, but not a delete, we have to
3838 # put in the new stuff; this comes next.
3839 if ($operation eq '+') {
3840 push @replacement, Range->new($start,
3841 $end,
3842 Value => $value,
3843 Type => $type);
3844 }
3845
3846 trace "Range at $j is $r->[$j]" if main::DEBUG && $to_trace && $j != $i;
3847 #trace "$end >=", $r->[$j]->start, " && $end <", $r->[$j]->end if main::DEBUG && $to_trace;
3848
3849 # And finally, if we're changing or deleting only a portion of the
3850 # highest affected range, it must be split, as the lowest one was.
3851 if (! $extends_above
3852 && $j >= 0 # Remember that j can be -1 if before first
3853 # current element
3854 && $end >= $r->[$j]->start
3855 && $end < $r->[$j]->end)
3856 {
3857 push @replacement,
3858 Range->new($end + 1,
3859 $r->[$j]->end,
3860 Value => $r->[$j]->value,
3861 Type => $r->[$j]->type);
3862 }
3863 }
3864
3865 # And do the splice, as calculated above
3866 if (main::DEBUG && $to_trace) {
3867 trace "replacing $length element(s) at $i with ";
3868 foreach my $replacement (@replacement) {
3869 trace " $replacement";
3870 }
3871 trace "Before splice:";
3872 trace 'i-2=[', $i-2, ']', $r->[$i-2] if $i >= 2;
3873 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i >= 1;
3874 trace "i =[", $i, "]", $r->[$i];
3875 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < @$r - 1;
3876 trace 'i+2=[', $i+2, ']', $r->[$i+2] if $i < @$r - 2;
3877 }
3878
3879 my @return = splice @$r, $splice_start, $length, @replacement;
3880
3881 if (main::DEBUG && $to_trace) {
3882 trace "After splice:";
3883 trace 'i-2=[', $i-2, ']', $r->[$i-2] if $i >= 2;
3884 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i >= 1;
3885 trace "i =[", $i, "]", $r->[$i];
3886 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < @$r - 1;
3887 trace 'i+2=[', $i+2, ']', $r->[$i+2] if $i < @$r - 2;
e6451557 3888 trace "removed ", @return if @return;
99870f4d
KW
3889 }
3890
3891 # An actual deletion could have changed the maximum in the list.
3892 # There was no deletion if the splice didn't return something, but
3893 # otherwise recalculate it. This is done too rarely to worry about
3894 # performance.
3895 if ($operation eq '-' && @return) {
3896 $max{$addr} = $r->[-1]->end;
3897 }
3898 return @return;
3899 }
3900