This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
mktables: Complement variable meaning for clarity
[perl5.git] / lib / unicore / mktables
CommitLineData
d73e5302 1#!/usr/bin/perl -w
99870f4d
KW
2
3# !!!!!!!!!!!!!! IF YOU MODIFY THIS FILE !!!!!!!!!!!!!!!!!!!!!!!!!
4# Any files created or read by this program should be listed in 'mktables.lst'
5# Use -makelist to regenerate it.
6
23e33b60
KW
7# Needs 'no overloading' to run faster on miniperl. Code commented out at the
8# subroutine objaddr can be used instead to work as far back (untested) as
f998e60c
KW
9# 5.8: needs pack "U". But almost all occurrences of objaddr have been
10# removed in favor of using 'no overloading'. You also would have to go
11# through and replace occurrences like:
ffe43484 12# my $addr = do { no overloading; pack 'J', $self; }
f998e60c
KW
13# with
14# my $addr = main::objaddr $self;
6c68572b 15# (or reverse commit 9b01bafde4b022706c3d6f947a0963f821b2e50b
051df77b
NC
16# that instituted the change to main::objaddr, and subsequent commits that
17# changed 0+$self to pack 'J', $self.)
6c68572b 18
cdcef19a
KW
19my $start_time;
20BEGIN { # Get the time the script started running; do it at compiliation to
21 # get it as close as possible
22 $start_time= time;
23}
24
25
23e33b60 26require 5.010_001;
d73e5302 27use strict;
99870f4d 28use warnings;
cf25bb62 29use Carp;
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
36
37##########################################################################
38#
39# mktables -- create the runtime Perl Unicode files (lib/unicore/.../*.pl),
40# from the Unicode database files (lib/unicore/.../*.txt), It also generates
41# a pod file and a .t file
42#
43# The structure of this file is:
44# First these introductory comments; then
45# code needed for everywhere, such as debugging stuff; then
46# code to handle input parameters; then
47# data structures likely to be of external interest (some of which depend on
48# the input parameters, so follows them; then
49# more data structures and subroutine and package (class) definitions; then
50# the small actual loop to process the input files and finish up; then
51# a __DATA__ section, for the .t tests
52#
53# This program works on all releases of Unicode through at least 5.2. The
54# outputs have been scrutinized most intently for release 5.1. The others
55# have been checked for somewhat more than just sanity. It can handle all
56# existing Unicode character properties in those releases.
57#
99870f4d
KW
58# This program is mostly about Unicode character (or code point) properties.
59# A property describes some attribute or quality of a code point, like if it
60# is lowercase or not, its name, what version of Unicode it was first defined
61# in, or what its uppercase equivalent is. Unicode deals with these disparate
62# possibilities by making all properties into mappings from each code point
63# into some corresponding value. In the case of it being lowercase or not,
64# the mapping is either to 'Y' or 'N' (or various synonyms thereof). Each
65# property maps each Unicode code point to a single value, called a "property
66# value". (Hence each Unicode property is a true mathematical function with
67# exactly one value per code point.)
68#
69# When using a property in a regular expression, what is desired isn't the
70# mapping of the code point to its property's value, but the reverse (or the
71# mathematical "inverse relation"): starting with the property value, "Does a
72# code point map to it?" These are written in a "compound" form:
73# \p{property=value}, e.g., \p{category=punctuation}. This program generates
74# files containing the lists of code points that map to each such regular
75# expression property value, one file per list
76#
77# There is also a single form shortcut that Perl adds for many of the commonly
78# used properties. This happens for all binary properties, plus script,
79# general_category, and block properties.
80#
81# Thus the outputs of this program are files. There are map files, mostly in
82# the 'To' directory; and there are list files for use in regular expression
83# matching, all in subdirectories of the 'lib' directory, with each
84# subdirectory being named for the property that the lists in it are for.
85# Bookkeeping, test, and documentation files are also generated.
86
87my $matches_directory = 'lib'; # Where match (\p{}) files go.
88my $map_directory = 'To'; # Where map files go.
89
90# DATA STRUCTURES
91#
92# The major data structures of this program are Property, of course, but also
93# Table. There are two kinds of tables, very similar to each other.
94# "Match_Table" is the data structure giving the list of code points that have
95# a particular property value, mentioned above. There is also a "Map_Table"
96# data structure which gives the property's mapping from code point to value.
97# There are two structures because the match tables need to be combined in
98# various ways, such as constructing unions, intersections, complements, etc.,
99# and the map ones don't. And there would be problems, perhaps subtle, if
100# a map table were inadvertently operated on in some of those ways.
101# The use of separate classes with operations defined on one but not the other
102# prevents accidentally confusing the two.
103#
104# At the heart of each table's data structure is a "Range_List", which is just
105# an ordered list of "Ranges", plus ancillary information, and methods to
106# operate on them. A Range is a compact way to store property information.
107# Each range has a starting code point, an ending code point, and a value that
108# is meant to apply to all the code points between the two end points,
109# inclusive. For a map table, this value is the property value for those
110# code points. Two such ranges could be written like this:
111# 0x41 .. 0x5A, 'Upper',
112# 0x61 .. 0x7A, 'Lower'
113#
114# Each range also has a type used as a convenience to classify the values.
115# Most ranges in this program will be Type 0, or normal, but there are some
116# ranges that have a non-zero type. These are used only in map tables, and
117# are for mappings that don't fit into the normal scheme of things. Mappings
118# that require a hash entry to communicate with utf8.c are one example;
119# another example is mappings for charnames.pm to use which indicate a name
120# that is algorithmically determinable from its code point (and vice-versa).
121# These are used to significantly compact these tables, instead of listing
122# each one of the tens of thousands individually.
123#
124# In a match table, the value of a range is irrelevant (and hence the type as
125# well, which will always be 0), and arbitrarily set to the null string.
126# Using the example above, there would be two match tables for those two
127# entries, one named Upper would contain the 0x41..0x5A range, and the other
128# named Lower would contain 0x61..0x7A.
129#
130# Actually, there are two types of range lists, "Range_Map" is the one
131# associated with map tables, and "Range_List" with match tables.
132# Again, this is so that methods can be defined on one and not the other so as
133# to prevent operating on them in incorrect ways.
134#
135# Eventually, most tables are written out to files to be read by utf8_heavy.pl
136# in the perl core. All tables could in theory be written, but some are
137# suppressed because there is no current practical use for them. It is easy
138# to change which get written by changing various lists that are near the top
139# of the actual code in this file. The table data structures contain enough
140# ancillary information to allow them to be treated as separate entities for
141# writing, such as the path to each one's file. There is a heading in each
142# map table that gives the format of its entries, and what the map is for all
143# the code points missing from it. (This allows tables to be more compact.)
678f13d5 144#
99870f4d
KW
145# The Property data structure contains one or more tables. All properties
146# contain a map table (except the $perl property which is a
147# pseudo-property containing only match tables), and any properties that
148# are usable in regular expression matches also contain various matching
149# tables, one for each value the property can have. A binary property can
150# have two values, True and False (or Y and N, which are preferred by Unicode
151# terminology). Thus each of these properties will have a map table that
152# takes every code point and maps it to Y or N (but having ranges cuts the
153# number of entries in that table way down), and two match tables, one
154# which has a list of all the code points that map to Y, and one for all the
155# code points that map to N. (For each of these, a third table is also
156# generated for the pseudo Perl property. It contains the identical code
157# points as the Y table, but can be written, not in the compound form, but in
158# a "single" form like \p{IsUppercase}.) Many properties are binary, but some
159# properties have several possible values, some have many, and properties like
160# Name have a different value for every named code point. Those will not,
161# unless the controlling lists are changed, have their match tables written
162# out. But all the ones which can be used in regular expression \p{} and \P{}
163# constructs will. Generally a property will have either its map table or its
164# match tables written but not both. Again, what gets written is controlled
165# by lists which can easily be changed.
678f13d5 166#
99870f4d
KW
167# For information about the Unicode properties, see Unicode's UAX44 document:
168
169my $unicode_reference_url = 'http://www.unicode.org/reports/tr44/';
170
171# As stated earlier, this program will work on any release of Unicode so far.
172# Most obvious problems in earlier data have NOT been corrected except when
173# necessary to make Perl or this program work reasonably. For example, no
174# folding information was given in early releases, so this program uses the
175# substitute of lower case, just so that a regular expression with the /i
176# option will do something that actually gives the right results in many
177# cases. There are also a couple other corrections for version 1.1.5,
178# commented at the point they are made. As an example of corrections that
179# weren't made (but could be) is this statement from DerivedAge.txt: "The
180# supplementary private use code points and the non-character code points were
181# assigned in version 2.0, but not specifically listed in the UCD until
182# versions 3.0 and 3.1 respectively." (To be precise it was 3.0.1 not 3.0.0)
183# More information on Unicode version glitches is further down in these
184# introductory comments.
185#
186# This program works on all properties as of 5.2, though the files for some
678f13d5
KW
187# are suppressed from apparent lack of demand for them. You can change which
188# are output by changing lists in this program.
189#
99870f4d
KW
190# The old version of mktables emphasized the term "Fuzzy" to mean Unocde's
191# loose matchings rules (from Unicode TR18):
192#
193# The recommended names for UCD properties and property values are in
194# PropertyAliases.txt [Prop] and PropertyValueAliases.txt
195# [PropValue]. There are both abbreviated names and longer, more
196# descriptive names. It is strongly recommended that both names be
197# recognized, and that loose matching of property names be used,
198# whereby the case distinctions, whitespace, hyphens, and underbar
199# are ignored.
200# The program still allows Fuzzy to override its determination of if loose
201# matching should be used, but it isn't currently used, as it is no longer
202# needed; the calculations it makes are good enough.
678f13d5 203#
99870f4d
KW
204# SUMMARY OF HOW IT WORKS:
205#
206# Process arguments
207#
208# A list is constructed containing each input file that is to be processed
209#
210# Each file on the list is processed in a loop, using the associated handler
211# code for each:
212# The PropertyAliases.txt and PropValueAliases.txt files are processed
213# first. These files name the properties and property values.
214# Objects are created of all the property and property value names
215# that the rest of the input should expect, including all synonyms.
216# The other input files give mappings from properties to property
217# values. That is, they list code points and say what the mapping
218# is under the given property. Some files give the mappings for
219# just one property; and some for many. This program goes through
220# each file and populates the properties from them. Some properties
221# are listed in more than one file, and Unicode has set up a
222# precedence as to which has priority if there is a conflict. Thus
223# the order of processing matters, and this program handles the
224# conflict possibility by processing the overriding input files
225# last, so that if necessary they replace earlier values.
226# After this is all done, the program creates the property mappings not
227# furnished by Unicode, but derivable from what it does give.
228# The tables of code points that match each property value in each
229# property that is accessible by regular expressions are created.
230# The Perl-defined properties are created and populated. Many of these
231# require data determined from the earlier steps
232# Any Perl-defined synonyms are created, and name clashes between Perl
678f13d5 233# and Unicode are reconciled and warned about.
99870f4d
KW
234# All the properties are written to files
235# Any other files are written, and final warnings issued.
678f13d5 236#
99870f4d
KW
237# For clarity, a number of operators have been overloaded to work on tables:
238# ~ means invert (take all characters not in the set). The more
239# conventional '!' is not used because of the possibility of confusing
240# it with the actual boolean operation.
241# + means union
242# - means subtraction
243# & means intersection
244# The precedence of these is the order listed. Parentheses should be
245# copiously used. These are not a general scheme. The operations aren't
246# defined for a number of things, deliberately, to avoid getting into trouble.
247# Operations are done on references and affect the underlying structures, so
248# that the copy constructors for them have been overloaded to not return a new
249# clone, but the input object itself.
678f13d5 250#
99870f4d
KW
251# The bool operator is deliberately not overloaded to avoid confusion with
252# "should it mean if the object merely exists, or also is non-empty?".
99870f4d
KW
253#
254# WHY CERTAIN DESIGN DECISIONS WERE MADE
678f13d5
KW
255#
256# This program needs to be able to run under miniperl. Therefore, it uses a
257# minimum of other modules, and hence implements some things itself that could
258# be gotten from CPAN
259#
260# This program uses inputs published by the Unicode Consortium. These can
261# change incompatibly between releases without the Perl maintainers realizing
262# it. Therefore this program is now designed to try to flag these. It looks
263# at the directories where the inputs are, and flags any unrecognized files.
264# It keeps track of all the properties in the files it handles, and flags any
265# that it doesn't know how to handle. It also flags any input lines that
266# don't match the expected syntax, among other checks.
267#
268# It is also designed so if a new input file matches one of the known
269# templates, one hopefully just needs to add it to a list to have it
270# processed.
271#
272# As mentioned earlier, some properties are given in more than one file. In
273# particular, the files in the extracted directory are supposedly just
274# reformattings of the others. But they contain information not easily
275# derivable from the other files, including results for Unihan, which this
276# program doesn't ordinarily look at, and for unassigned code points. They
277# also have historically had errors or been incomplete. In an attempt to
278# create the best possible data, this program thus processes them first to
279# glean information missing from the other files; then processes those other
280# files to override any errors in the extracted ones. Much of the design was
281# driven by this need to store things and then possibly override them.
282#
283# It tries to keep fatal errors to a minimum, to generate something usable for
284# testing purposes. It always looks for files that could be inputs, and will
285# warn about any that it doesn't know how to handle (the -q option suppresses
286# the warning).
99870f4d
KW
287#
288# Why have files written out for binary 'N' matches?
289# For binary properties, if you know the mapping for either Y or N; the
678f13d5
KW
290# other is trivial to construct, so could be done at Perl run-time by just
291# complementing the result, instead of having a file for it. That is, if
292# someone types in \p{foo: N}, Perl could translate that to \P{foo: Y} and
293# not need a file. The problem is communicating to Perl that a given
294# property is binary. Perl can't figure it out from looking at the N (or
295# No), as some non-binary properties have these as property values. So
296# rather than inventing a way to communicate this info back to the core,
297# which would have required changes there as well, it was simpler just to
298# add the extra tables.
299#
300# Why is there more than one type of range?
301# This simplified things. There are some very specialized code points that
302# have to be handled specially for output, such as Hangul syllable names.
303# By creating a range type (done late in the development process), it
304# allowed this to be stored with the range, and overridden by other input.
305# Originally these were stored in another data structure, and it became a
306# mess trying to decide if a second file that was for the same property was
307# overriding the earlier one or not.
308#
309# Why are there two kinds of tables, match and map?
310# (And there is a base class shared by the two as well.) As stated above,
311# they actually are for different things. Development proceeded much more
312# smoothly when I (khw) realized the distinction. Map tables are used to
313# give the property value for every code point (actually every code point
314# that doesn't map to a default value). Match tables are used for regular
315# expression matches, and are essentially the inverse mapping. Separating
316# the two allows more specialized methods, and error checks so that one
317# can't just take the intersection of two map tables, for example, as that
318# is nonsensical.
99870f4d
KW
319#
320# There are no match tables generated for matches of the null string. These
c1739a4a 321# would look like qr/\p{JSN=}/ currently without modifying the regex code.
678f13d5
KW
322# Perhaps something like them could be added if necessary. The JSN does have
323# a real code point U+110B that maps to the null string, but it is a
324# contributory property, and therefore not output by default. And it's easily
325# handled so far by making the null string the default where it is a
326# possibility.
99870f4d 327#
23e33b60
KW
328# DEBUGGING
329#
678f13d5
KW
330# This program is written so it will run under miniperl. Occasionally changes
331# will cause an error where the backtrace doesn't work well under miniperl.
332# To diagnose the problem, you can instead run it under regular perl, if you
333# have one compiled.
334#
335# There is a good trace facility. To enable it, first sub DEBUG must be set
336# to return true. Then a line like
337#
338# local $to_trace = 1 if main::DEBUG;
339#
340# can be added to enable tracing in its lexical scope or until you insert
341# another line:
342#
343# local $to_trace = 0 if main::DEBUG;
344#
345# then use a line like "trace $a, @b, %c, ...;
346#
347# Some of the more complex subroutines already have trace statements in them.
348# Permanent trace statements should be like:
349#
350# trace ... if main::DEBUG && $to_trace;
351#
352# If there is just one or a few files that you're debugging, you can easily
353# cause most everything else to be skipped. Change the line
354#
355# my $debug_skip = 0;
356#
357# to 1, and every file whose object is in @input_file_objects and doesn't have
358# a, 'non_skip => 1,' in its constructor will be skipped.
359#
b4a0206c 360# To compare the output tables, it may be useful to specify the -annotate
c4019d52
KW
361# flag. This causes the tables to expand so there is one entry for each
362# non-algorithmically named code point giving, currently its name, and its
363# graphic representation if printable (and you have a font that knows about
364# it). This makes it easier to see what the particular code points are in
365# each output table. The tables are usable, but because they don't have
366# ranges (for the most part), a Perl using them will run slower. Non-named
367# code points are annotated with a description of their status, and contiguous
368# ones with the same description will be output as a range rather than
369# individually. Algorithmically named characters are also output as ranges,
370# except when there are just a few contiguous ones.
371#
99870f4d
KW
372# FUTURE ISSUES
373#
374# The program would break if Unicode were to change its names so that
375# interior white space, underscores, or dashes differences were significant
376# within property and property value names.
377#
378# It might be easier to use the xml versions of the UCD if this program ever
379# would need heavy revision, and the ability to handle old versions was not
380# required.
381#
382# There is the potential for name collisions, in that Perl has chosen names
383# that Unicode could decide it also likes. There have been such collisions in
384# the past, with mostly Perl deciding to adopt the Unicode definition of the
385# name. However in the 5.2 Unicode beta testing, there were a number of such
386# collisions, which were withdrawn before the final release, because of Perl's
387# and other's protests. These all involved new properties which began with
388# 'Is'. Based on the protests, Unicode is unlikely to try that again. Also,
389# many of the Perl-defined synonyms, like Any, Word, etc, are listed in a
390# Unicode document, so they are unlikely to be used by Unicode for another
391# purpose. However, they might try something beginning with 'In', or use any
392# of the other Perl-defined properties. This program will warn you of name
393# collisions, and refuse to generate tables with them, but manual intervention
394# will be required in this event. One scheme that could be implemented, if
395# necessary, would be to have this program generate another file, or add a
396# field to mktables.lst that gives the date of first definition of a property.
397# Each new release of Unicode would use that file as a basis for the next
398# iteration. And the Perl synonym addition code could sort based on the age
399# of the property, so older properties get priority, and newer ones that clash
400# would be refused; hence existing code would not be impacted, and some other
401# synonym would have to be used for the new property. This is ugly, and
402# manual intervention would certainly be easier to do in the short run; lets
403# hope it never comes to this.
678f13d5 404#
99870f4d
KW
405# A NOTE ON UNIHAN
406#
407# This program can generate tables from the Unihan database. But it doesn't
408# by default, letting the CPAN module Unicode::Unihan handle them. Prior to
409# version 5.2, this database was in a single file, Unihan.txt. In 5.2 the
410# database was split into 8 different files, all beginning with the letters
411# 'Unihan'. This program will read those file(s) if present, but it needs to
412# know which of the many properties in the file(s) should have tables created
413# for them. It will create tables for any properties listed in
414# PropertyAliases.txt and PropValueAliases.txt, plus any listed in the
415# @cjk_properties array and the @cjk_property_values array. Thus, if a
416# property you want is not in those files of the release you are building
417# against, you must add it to those two arrays. Starting in 4.0, the
418# Unicode_Radical_Stroke was listed in those files, so if the Unihan database
419# is present in the directory, a table will be generated for that property.
420# In 5.2, several more properties were added. For your convenience, the two
421# arrays are initialized with all the 5.2 listed properties that are also in
422# earlier releases. But these are commented out. You can just uncomment the
423# ones you want, or use them as a template for adding entries for other
424# properties.
425#
426# You may need to adjust the entries to suit your purposes. setup_unihan(),
427# and filter_unihan_line() are the functions where this is done. This program
428# already does some adjusting to make the lines look more like the rest of the
429# Unicode DB; You can see what that is in filter_unihan_line()
430#
431# There is a bug in the 3.2 data file in which some values for the
432# kPrimaryNumeric property have commas and an unexpected comment. A filter
433# could be added for these; or for a particular installation, the Unihan.txt
434# file could be edited to fix them.
99870f4d 435#
678f13d5
KW
436# HOW TO ADD A FILE TO BE PROCESSED
437#
438# A new file from Unicode needs to have an object constructed for it in
439# @input_file_objects, probably at the end or at the end of the extracted
440# ones. The program should warn you if its name will clash with others on
441# restrictive file systems, like DOS. If so, figure out a better name, and
442# add lines to the README.perl file giving that. If the file is a character
443# property, it should be in the format that Unicode has by default
444# standardized for such files for the more recently introduced ones.
445# If so, the Input_file constructor for @input_file_objects can just be the
446# file name and release it first appeared in. If not, then it should be
447# possible to construct an each_line_handler() to massage the line into the
448# standardized form.
449#
450# For non-character properties, more code will be needed. You can look at
451# the existing entries for clues.
452#
453# UNICODE VERSIONS NOTES
454#
455# The Unicode UCD has had a number of errors in it over the versions. And
456# these remain, by policy, in the standard for that version. Therefore it is
457# risky to correct them, because code may be expecting the error. So this
458# program doesn't generally make changes, unless the error breaks the Perl
459# core. As an example, some versions of 2.1.x Jamo.txt have the wrong value
460# for U+1105, which causes real problems for the algorithms for Jamo
461# calculations, so it is changed here.
462#
463# But it isn't so clear cut as to what to do about concepts that are
464# introduced in a later release; should they extend back to earlier releases
465# where the concept just didn't exist? It was easier to do this than to not,
466# so that's what was done. For example, the default value for code points not
467# in the files for various properties was probably undefined until changed by
468# some version. No_Block for blocks is such an example. This program will
469# assign No_Block even in Unicode versions that didn't have it. This has the
470# benefit that code being written doesn't have to special case earlier
471# versions; and the detriment that it doesn't match the Standard precisely for
472# the affected versions.
473#
474# Here are some observations about some of the issues in early versions:
475#
476# The number of code points in \p{alpha} halve in 2.1.9. It turns out that
477# the reason is that the CJK block starting at 4E00 was removed from PropList,
478# and was not put back in until 3.1.0
479#
480# Unicode introduced the synonym Space for White_Space in 4.1. Perl has
481# always had a \p{Space}. In release 3.2 only, they are not synonymous. The
482# reason is that 3.2 introduced U+205F=medium math space, which was not
483# classed as white space, but Perl figured out that it should have been. 4.0
484# reclassified it correctly.
485#
486# Another change between 3.2 and 4.0 is the CCC property value ATBL. In 3.2
487# this was erroneously a synonym for 202. In 4.0, ATB became 202, and ATBL
488# was left with no code points, as all the ones that mapped to 202 stayed
489# mapped to 202. Thus if your program used the numeric name for the class,
490# it would not have been affected, but if it used the mnemonic, it would have
491# been.
492#
493# \p{Script=Hrkt} (Katakana_Or_Hiragana) came in 4.0.1. Before that code
494# points which eventually came to have this script property value, instead
495# mapped to "Unknown". But in the next release all these code points were
496# moved to \p{sc=common} instead.
99870f4d
KW
497#
498# The default for missing code points for BidiClass is complicated. Starting
499# in 3.1.1, the derived file DBidiClass.txt handles this, but this program
500# tries to do the best it can for earlier releases. It is done in
501# process_PropertyAliases()
502#
503##############################################################################
504
505my $UNDEF = ':UNDEF:'; # String to print out for undefined values in tracing
506 # and errors
507my $MAX_LINE_WIDTH = 78;
508
509# Debugging aid to skip most files so as to not be distracted by them when
510# concentrating on the ones being debugged. Add
511# non_skip => 1,
512# to the constructor for those files you want processed when you set this.
513# Files with a first version number of 0 are special: they are always
514# processed regardless of the state of this flag.
515my $debug_skip = 0;
516
517# Set to 1 to enable tracing.
518our $to_trace = 0;
519
520{ # Closure for trace: debugging aid
521 my $print_caller = 1; # ? Include calling subroutine name
522 my $main_with_colon = 'main::';
523 my $main_colon_length = length($main_with_colon);
524
525 sub trace {
526 return unless $to_trace; # Do nothing if global flag not set
527
528 my @input = @_;
529
530 local $DB::trace = 0;
531 $DB::trace = 0; # Quiet 'used only once' message
532
533 my $line_number;
534
535 # Loop looking up the stack to get the first non-trace caller
536 my $caller_line;
537 my $caller_name;
538 my $i = 0;
539 do {
540 $line_number = $caller_line;
541 (my $pkg, my $file, $caller_line, my $caller) = caller $i++;
542 $caller = $main_with_colon unless defined $caller;
543
544 $caller_name = $caller;
545
546 # get rid of pkg
547 $caller_name =~ s/.*:://;
548 if (substr($caller_name, 0, $main_colon_length)
549 eq $main_with_colon)
550 {
551 $caller_name = substr($caller_name, $main_colon_length);
552 }
553
554 } until ($caller_name ne 'trace');
555
556 # If the stack was empty, we were called from the top level
557 $caller_name = 'main' if ($caller_name eq ""
558 || $caller_name eq 'trace');
559
560 my $output = "";
561 foreach my $string (@input) {
562 #print STDERR __LINE__, ": ", join ", ", @input, "\n";
563 if (ref $string eq 'ARRAY' || ref $string eq 'HASH') {
564 $output .= simple_dumper($string);
565 }
566 else {
567 $string = "$string" if ref $string;
568 $string = $UNDEF unless defined $string;
569 chomp $string;
570 $string = '""' if $string eq "";
571 $output .= " " if $output ne ""
572 && $string ne ""
573 && substr($output, -1, 1) ne " "
574 && substr($string, 0, 1) ne " ";
575 $output .= $string;
576 }
577 }
578
99f78760
KW
579 print STDERR sprintf "%4d: ", $line_number if defined $line_number;
580 print STDERR "$caller_name: " if $print_caller;
99870f4d
KW
581 print STDERR $output, "\n";
582 return;
583 }
584}
585
586# This is for a rarely used development feature that allows you to compare two
587# versions of the Unicode standard without having to deal with changes caused
588# by the code points introduced in the later verson. Change the 0 to a SINGLE
589# dotted Unicode release number (e.g. 2.1). Only code points introduced in
590# that release and earlier will be used; later ones are thrown away. You use
591# the version number of the earliest one you want to compare; then run this
592# program on directory structures containing each release, and compare the
593# outputs. These outputs will therefore include only the code points common
594# to both releases, and you can see the changes caused just by the underlying
595# release semantic changes. For versions earlier than 3.2, you must copy a
596# version of DAge.txt into the directory.
597my $string_compare_versions = DEBUG && 0; # e.g., v2.1;
598my $compare_versions = DEBUG
599 && $string_compare_versions
600 && pack "C*", split /\./, $string_compare_versions;
601
602sub uniques {
603 # Returns non-duplicated input values. From "Perl Best Practices:
604 # Encapsulated Cleverness". p. 455 in first edition.
605
606 my %seen;
0e407844
NC
607 # Arguably this breaks encapsulation, if the goal is to permit multiple
608 # distinct objects to stringify to the same value, and be interchangeable.
609 # However, for this program, no two objects stringify identically, and all
610 # lists passed to this function are either objects or strings. So this
611 # doesn't affect correctness, but it does give a couple of percent speedup.
612 no overloading;
99870f4d
KW
613 return grep { ! $seen{$_}++ } @_;
614}
615
616$0 = File::Spec->canonpath($0);
617
618my $make_test_script = 0; # ? Should we output a test script
619my $write_unchanged_files = 0; # ? Should we update the output files even if
620 # we don't think they have changed
621my $use_directory = ""; # ? Should we chdir somewhere.
622my $pod_directory; # input directory to store the pod file.
623my $pod_file = 'perluniprops';
624my $t_path; # Path to the .t test file
625my $file_list = 'mktables.lst'; # File to store input and output file names.
626 # This is used to speed up the build, by not
627 # executing the main body of the program if
628 # nothing on the list has changed since the
629 # previous build
630my $make_list = 1; # ? Should we write $file_list. Set to always
631 # make a list so that when the pumpking is
632 # preparing a release, s/he won't have to do
633 # special things
634my $glob_list = 0; # ? Should we try to include unknown .txt files
635 # in the input.
636my $output_range_counts = 1; # ? Should we include the number of code points
637 # in ranges in the output
558712cf 638my $annotate = 0; # ? Should character names be in the output
9ef2b94f 639
99870f4d
KW
640# Verbosity levels; 0 is quiet
641my $NORMAL_VERBOSITY = 1;
642my $PROGRESS = 2;
643my $VERBOSE = 3;
644
645my $verbosity = $NORMAL_VERBOSITY;
646
647# Process arguments
648while (@ARGV) {
cf25bb62
JH
649 my $arg = shift @ARGV;
650 if ($arg eq '-v') {
99870f4d
KW
651 $verbosity = $VERBOSE;
652 }
653 elsif ($arg eq '-p') {
654 $verbosity = $PROGRESS;
655 $| = 1; # Flush buffers as we go.
656 }
657 elsif ($arg eq '-q') {
658 $verbosity = 0;
659 }
660 elsif ($arg eq '-w') {
661 $write_unchanged_files = 1; # update the files even if havent changed
662 }
663 elsif ($arg eq '-check') {
6ae7e459
YO
664 my $this = shift @ARGV;
665 my $ok = shift @ARGV;
666 if ($this ne $ok) {
667 print "Skipping as check params are not the same.\n";
668 exit(0);
669 }
00a8df5c 670 }
99870f4d
KW
671 elsif ($arg eq '-P' && defined ($pod_directory = shift)) {
672 -d $pod_directory or croak "Directory '$pod_directory' doesn't exist";
673 }
3df51b85
KW
674 elsif ($arg eq '-maketest' || ($arg eq '-T' && defined ($t_path = shift)))
675 {
99870f4d 676 $make_test_script = 1;
99870f4d
KW
677 }
678 elsif ($arg eq '-makelist') {
679 $make_list = 1;
680 }
681 elsif ($arg eq '-C' && defined ($use_directory = shift)) {
682 -d $use_directory or croak "Unknown directory '$use_directory'";
683 }
684 elsif ($arg eq '-L') {
685
686 # Existence not tested until have chdir'd
687 $file_list = shift;
688 }
689 elsif ($arg eq '-globlist') {
690 $glob_list = 1;
691 }
692 elsif ($arg eq '-c') {
693 $output_range_counts = ! $output_range_counts
694 }
b4a0206c 695 elsif ($arg eq '-annotate') {
558712cf 696 $annotate = 1;
9ef2b94f 697 }
99870f4d
KW
698 else {
699 my $with_c = 'with';
700 $with_c .= 'out' if $output_range_counts; # Complements the state
701 croak <<END;
702usage: $0 [-c|-p|-q|-v|-w] [-C dir] [-L filelist] [ -P pod_dir ]
703 [ -T test_file_path ] [-globlist] [-makelist] [-maketest]
704 [-check A B ]
705 -c : Output comments $with_c number of code points in ranges
706 -q : Quiet Mode: Only output serious warnings.
707 -p : Set verbosity level to normal plus show progress.
708 -v : Set Verbosity level high: Show progress and non-serious
709 warnings
710 -w : Write files regardless
711 -C dir : Change to this directory before proceeding. All relative paths
712 except those specified by the -P and -T options will be done
713 with respect to this directory.
714 -P dir : Output $pod_file file to directory 'dir'.
3df51b85 715 -T path : Create a test script as 'path'; overrides -maketest
99870f4d
KW
716 -L filelist : Use alternate 'filelist' instead of standard one
717 -globlist : Take as input all non-Test *.txt files in current and sub
718 directories
3df51b85
KW
719 -maketest : Make test script 'TestProp.pl' in current (or -C directory),
720 overrides -T
99870f4d 721 -makelist : Rewrite the file list $file_list based on current setup
b4a0206c 722 -annotate : Output an annotation for each character in the table files;
c4019d52
KW
723 useful for debugging mktables, looking at diffs; but is slow,
724 memory intensive; resulting tables are usable but slow and
725 very large.
99870f4d
KW
726 -check A B : Executes $0 only if A and B are the same
727END
728 }
729}
730
731# Stores the most-recently changed file. If none have changed, can skip the
732# build
aeab6150 733my $most_recent = (stat $0)[9]; # Do this before the chdir!
99870f4d
KW
734
735# Change directories now, because need to read 'version' early.
736if ($use_directory) {
3df51b85 737 if ($pod_directory && ! File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d
KW
738 $pod_directory = File::Spec->rel2abs($pod_directory);
739 }
3df51b85 740 if ($t_path && ! File::Spec->file_name_is_absolute($t_path)) {
99870f4d 741 $t_path = File::Spec->rel2abs($t_path);
00a8df5c 742 }
99870f4d 743 chdir $use_directory or croak "Failed to chdir to '$use_directory':$!";
3df51b85 744 if ($pod_directory && File::Spec->file_name_is_absolute($pod_directory)) {
99870f4d 745 $pod_directory = File::Spec->abs2rel($pod_directory);
02b1aeec 746 }
3df51b85 747 if ($t_path && File::Spec->file_name_is_absolute($t_path)) {
99870f4d 748 $t_path = File::Spec->abs2rel($t_path);
02b1aeec 749 }
00a8df5c
YO
750}
751
99870f4d
KW
752# Get Unicode version into regular and v-string. This is done now because
753# various tables below get populated based on it. These tables are populated
754# here to be near the top of the file, and so easily seeable by those needing
755# to modify things.
756open my $VERSION, "<", "version"
757 or croak "$0: can't open required file 'version': $!\n";
758my $string_version = <$VERSION>;
759close $VERSION;
760chomp $string_version;
761my $v_version = pack "C*", split /\./, $string_version; # v string
762
763# The following are the complete names of properties with property values that
764# are known to not match any code points in some versions of Unicode, but that
765# may change in the future so they should be matchable, hence an empty file is
766# generated for them.
767my @tables_that_may_be_empty = (
768 'Joining_Type=Left_Joining',
769 );
770push @tables_that_may_be_empty, 'Script=Common' if $v_version le v4.0.1;
771push @tables_that_may_be_empty, 'Title' if $v_version lt v2.0.0;
772push @tables_that_may_be_empty, 'Script=Katakana_Or_Hiragana'
773 if $v_version ge v4.1.0;
774
775# The lists below are hashes, so the key is the item in the list, and the
776# value is the reason why it is in the list. This makes generation of
777# documentation easier.
778
779my %why_suppressed; # No file generated for these.
780
781# Files aren't generated for empty extraneous properties. This is arguable.
782# Extraneous properties generally come about because a property is no longer
783# used in a newer version of Unicode. If we generated a file without code
784# points, programs that used to work on that property will still execute
785# without errors. It just won't ever match (or will always match, with \P{}).
786# This means that the logic is now likely wrong. I (khw) think its better to
787# find this out by getting an error message. Just move them to the table
788# above to change this behavior
789my %why_suppress_if_empty_warn_if_not = (
790
791 # It is the only property that has ever officially been removed from the
792 # Standard. The database never contained any code points for it.
793 'Special_Case_Condition' => 'Obsolete',
794
795 # Apparently never official, but there were code points in some versions of
796 # old-style PropList.txt
797 'Non_Break' => 'Obsolete',
798);
799
800# These would normally go in the warn table just above, but they were changed
801# a long time before this program was written, so warnings about them are
802# moot.
803if ($v_version gt v3.2.0) {
804 push @tables_that_may_be_empty,
805 'Canonical_Combining_Class=Attached_Below_Left'
806}
807
808# These are listed in the Property aliases file in 5.2, but Unihan is ignored
809# unless explicitly added.
810if ($v_version ge v5.2.0) {
811 my $unihan = 'Unihan; remove from list if using Unihan';
ea25a9b2 812 foreach my $table (qw (
99870f4d
KW
813 kAccountingNumeric
814 kOtherNumeric
815 kPrimaryNumeric
816 kCompatibilityVariant
817 kIICore
818 kIRG_GSource
819 kIRG_HSource
820 kIRG_JSource
821 kIRG_KPSource
822 kIRG_MSource
823 kIRG_KSource
824 kIRG_TSource
825 kIRG_USource
826 kIRG_VSource
827 kRSUnicode
ea25a9b2 828 ))
99870f4d
KW
829 {
830 $why_suppress_if_empty_warn_if_not{$table} = $unihan;
831 }
ca12659b
NC
832}
833
99870f4d
KW
834# Properties that this program ignores.
835my @unimplemented_properties = (
836'Unicode_Radical_Stroke' # Remove if changing to handle this one.
837);
d73e5302 838
99870f4d
KW
839# There are several types of obsolete properties defined by Unicode. These
840# must be hand-edited for every new Unicode release.
841my %why_deprecated; # Generates a deprecated warning message if used.
842my %why_stabilized; # Documentation only
843my %why_obsolete; # Documentation only
844
845{ # Closure
846 my $simple = 'Perl uses the more complete version of this property';
847 my $unihan = 'Unihan properties are by default not enabled in the Perl core. Instead use CPAN: Unicode::Unihan';
848
849 my $other_properties = 'other properties';
850 my $contributory = "Used by Unicode internally for generating $other_properties and not intended to be used stand-alone";
851 my $why_no_expand = "Easily computed, and yet doesn't cover the common encoding forms (UTF-16/8)",
852
853 %why_deprecated = (
854 'Grapheme_Link' => 'Deprecated by Unicode. Use ccc=vr (Canonical_Combining_Class=Virama) instead',
855 'Jamo_Short_Name' => $contributory,
856 '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',
857 'Other_Alphabetic' => $contributory,
858 'Other_Default_Ignorable_Code_Point' => $contributory,
859 'Other_Grapheme_Extend' => $contributory,
860 'Other_ID_Continue' => $contributory,
861 'Other_ID_Start' => $contributory,
862 'Other_Lowercase' => $contributory,
863 'Other_Math' => $contributory,
864 'Other_Uppercase' => $contributory,
865 );
866
867 %why_suppressed = (
868 # There is a lib/unicore/Decomposition.pl (used by normalize.pm) which
869 # contains the same information, but without the algorithmically
870 # determinable Hangul syllables'. This file is not published, so it's
871 # existence is not noted in the comment.
872 'Decomposition_Mapping' => 'Accessible via Unicode::Normalize',
873
874 'ISO_Comment' => 'Apparently no demand for it, but can access it through Unicode::UCD::charinfo. Obsoleted, and code points for it removed in Unicode 5.2',
875 'Unicode_1_Name' => "$simple, and no apparent demand for it, but can access it through Unicode::UCD::charinfo. If there is no later name for a code point, then this one is used instead in charnames",
876
877 'Simple_Case_Folding' => "$simple. Can access this through Unicode::UCD::casefold",
878 'Simple_Lowercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo",
879 'Simple_Titlecase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo",
880 'Simple_Uppercase_Mapping' => "$simple. Can access this through Unicode::UCD::charinfo",
881
882 'Name' => "Accessible via 'use charnames;'",
883 'Name_Alias' => "Accessible via 'use charnames;'",
884
885 # These are sort of jumping the gun; deprecation is proposed for
886 # Unicode version 6.0, but they have never been exposed by Perl, and
887 # likely are soon to be deprecated, so best not to expose them.
888 FC_NFKC_Closure => 'Use NFKC_Casefold instead',
889 Expands_On_NFC => $why_no_expand,
890 Expands_On_NFD => $why_no_expand,
891 Expands_On_NFKC => $why_no_expand,
892 Expands_On_NFKD => $why_no_expand,
893 );
894
895 # The following are suppressed because they were made contributory or
896 # deprecated by Unicode before Perl ever thought about supporting them.
897 foreach my $property ('Jamo_Short_Name', 'Grapheme_Link') {
898 $why_suppressed{$property} = $why_deprecated{$property};
899 }
cf25bb62 900
99870f4d
KW
901 # Customize the message for all the 'Other_' properties
902 foreach my $property (keys %why_deprecated) {
903 next if (my $main_property = $property) !~ s/^Other_//;
904 $why_deprecated{$property} =~ s/$other_properties/the $main_property property (which should be used instead)/;
905 }
906}
907
908if ($v_version ge 4.0.0) {
909 $why_stabilized{'Hyphen'} = 'Use the Line_Break property instead; see www.unicode.org/reports/tr14';
910}
911if ($v_version ge 5.2.0) {
912 $why_obsolete{'ISO_Comment'} = 'Code points for it have been removed';
913}
914
915# Probably obsolete forever
916if ($v_version ge v4.1.0) {
917 $why_suppressed{'Script=Katakana_Or_Hiragana'} = 'Obsolete. All code points previously matched by this have been moved to "Script=Common"';
918}
919
920# This program can create files for enumerated-like properties, such as
921# 'Numeric_Type'. This file would be the same format as for a string
922# property, with a mapping from code point to its value, so you could look up,
923# for example, the script a code point is in. But no one so far wants this
924# mapping, or they have found another way to get it since this is a new
925# feature. So no file is generated except if it is in this list.
926my @output_mapped_properties = split "\n", <<END;
927END
928
929# If you are using the Unihan database, you need to add the properties that
930# you want to extract from it to this table. For your convenience, the
931# properties in the 5.2 PropertyAliases.txt file are listed, commented out
932my @cjk_properties = split "\n", <<'END';
933#cjkAccountingNumeric; kAccountingNumeric
934#cjkOtherNumeric; kOtherNumeric
935#cjkPrimaryNumeric; kPrimaryNumeric
936#cjkCompatibilityVariant; kCompatibilityVariant
937#cjkIICore ; kIICore
938#cjkIRG_GSource; kIRG_GSource
939#cjkIRG_HSource; kIRG_HSource
940#cjkIRG_JSource; kIRG_JSource
941#cjkIRG_KPSource; kIRG_KPSource
942#cjkIRG_KSource; kIRG_KSource
943#cjkIRG_TSource; kIRG_TSource
944#cjkIRG_USource; kIRG_USource
945#cjkIRG_VSource; kIRG_VSource
946#cjkRSUnicode; kRSUnicode ; Unicode_Radical_Stroke; URS
947END
948
949# Similarly for the property values. For your convenience, the lines in the
950# 5.2 PropertyAliases.txt file are listed. Just remove the first BUT NOT both
951# '#' marks
952my @cjk_property_values = split "\n", <<'END';
953## @missing: 0000..10FFFF; cjkAccountingNumeric; NaN
954## @missing: 0000..10FFFF; cjkCompatibilityVariant; <code point>
955## @missing: 0000..10FFFF; cjkIICore; <none>
956## @missing: 0000..10FFFF; cjkIRG_GSource; <none>
957## @missing: 0000..10FFFF; cjkIRG_HSource; <none>
958## @missing: 0000..10FFFF; cjkIRG_JSource; <none>
959## @missing: 0000..10FFFF; cjkIRG_KPSource; <none>
960## @missing: 0000..10FFFF; cjkIRG_KSource; <none>
961## @missing: 0000..10FFFF; cjkIRG_TSource; <none>
962## @missing: 0000..10FFFF; cjkIRG_USource; <none>
963## @missing: 0000..10FFFF; cjkIRG_VSource; <none>
964## @missing: 0000..10FFFF; cjkOtherNumeric; NaN
965## @missing: 0000..10FFFF; cjkPrimaryNumeric; NaN
966## @missing: 0000..10FFFF; cjkRSUnicode; <none>
967END
968
969# The input files don't list every code point. Those not listed are to be
970# defaulted to some value. Below are hard-coded what those values are for
971# non-binary properties as of 5.1. Starting in 5.0, there are
972# machine-parsable comment lines in the files the give the defaults; so this
973# list shouldn't have to be extended. The claim is that all missing entries
974# for binary properties will default to 'N'. Unicode tried to change that in
975# 5.2, but the beta period produced enough protest that they backed off.
976#
977# The defaults for the fields that appear in UnicodeData.txt in this hash must
978# be in the form that it expects. The others may be synonyms.
979my $CODE_POINT = '<code point>';
980my %default_mapping = (
981 Age => "Unassigned",
982 # Bidi_Class => Complicated; set in code
983 Bidi_Mirroring_Glyph => "",
984 Block => 'No_Block',
985 Canonical_Combining_Class => 0,
986 Case_Folding => $CODE_POINT,
987 Decomposition_Mapping => $CODE_POINT,
988 Decomposition_Type => 'None',
989 East_Asian_Width => "Neutral",
990 FC_NFKC_Closure => $CODE_POINT,
991 General_Category => 'Cn',
992 Grapheme_Cluster_Break => 'Other',
993 Hangul_Syllable_Type => 'NA',
994 ISO_Comment => "",
995 Jamo_Short_Name => "",
996 Joining_Group => "No_Joining_Group",
997 # Joining_Type => Complicated; set in code
998 kIICore => 'N', # Is converted to binary
999 #Line_Break => Complicated; set in code
1000 Lowercase_Mapping => $CODE_POINT,
1001 Name => "",
1002 Name_Alias => "",
1003 NFC_QC => 'Yes',
1004 NFD_QC => 'Yes',
1005 NFKC_QC => 'Yes',
1006 NFKD_QC => 'Yes',
1007 Numeric_Type => 'None',
1008 Numeric_Value => 'NaN',
1009 Script => ($v_version le 4.1.0) ? 'Common' : 'Unknown',
1010 Sentence_Break => 'Other',
1011 Simple_Case_Folding => $CODE_POINT,
1012 Simple_Lowercase_Mapping => $CODE_POINT,
1013 Simple_Titlecase_Mapping => $CODE_POINT,
1014 Simple_Uppercase_Mapping => $CODE_POINT,
1015 Titlecase_Mapping => $CODE_POINT,
1016 Unicode_1_Name => "",
1017 Unicode_Radical_Stroke => "",
1018 Uppercase_Mapping => $CODE_POINT,
1019 Word_Break => 'Other',
1020);
1021
1022# Below are files that Unicode furnishes, but this program ignores, and why
1023my %ignored_files = (
1024 'CJKRadicals.txt' => 'Unihan data',
1025 'Index.txt' => 'An index, not actual data',
1026 'NamedSqProv.txt' => 'Not officially part of the Unicode standard; Append it to NamedSequences.txt if you want to process the contents.',
1027 'NamesList.txt' => 'Just adds commentary',
1028 'NormalizationCorrections.txt' => 'Data is already in other files.',
1029 'Props.txt' => 'Adds nothing to PropList.txt; only in very early releases',
1030 'ReadMe.txt' => 'Just comments',
1031 'README.TXT' => 'Just comments',
1032 'StandardizedVariants.txt' => 'Only for glyph changes, not a Unicode character property. Does not fit into current scheme where one code point is mapped',
1033);
1034
678f13d5 1035### End of externally interesting definitions, except for @input_file_objects
99870f4d
KW
1036
1037my $HEADER=<<"EOF";
1038# !!!!!!! DO NOT EDIT THIS FILE !!!!!!!
3df51b85
KW
1039# This file is machine-generated by $0 from the Unicode
1040# database, Version $string_version. Any changes made here will be lost!
cf25bb62
JH
1041EOF
1042
b6922eda 1043my $INTERNAL_ONLY=<<"EOF";
99870f4d
KW
1044
1045# !!!!!!! INTERNAL PERL USE ONLY !!!!!!!
b6922eda 1046# This file is for internal use by the Perl program only. The format and even
99870f4d
KW
1047# the name or existence of this file are subject to change without notice.
1048# Don't use it directly.
1049EOF
1050
1051my $DEVELOPMENT_ONLY=<<"EOF";
1052# !!!!!!! DEVELOPMENT USE ONLY !!!!!!!
1053# This file contains information artificially constrained to code points
1054# present in Unicode release $string_compare_versions.
1055# IT CANNOT BE RELIED ON. It is for use during development only and should
23e33b60 1056# not be used for production.
b6922eda
KW
1057
1058EOF
1059
99870f4d
KW
1060my $LAST_UNICODE_CODEPOINT_STRING = "10FFFF";
1061my $LAST_UNICODE_CODEPOINT = hex $LAST_UNICODE_CODEPOINT_STRING;
1062my $MAX_UNICODE_CODEPOINTS = $LAST_UNICODE_CODEPOINT + 1;
1063
1064# Matches legal code point. 4-6 hex numbers, If there are 6, the first
1065# two must be 10; if there are 5, the first must not be a 0. Written this way
1066# to decrease backtracking
1067my $code_point_re =
1068 qr/ \b (?: 10[0-9A-F]{4} | [1-9A-F][0-9A-F]{4} | [0-9A-F]{4} ) \b/x;
1069
1070# This matches the beginning of the line in the Unicode db files that give the
1071# defaults for code points not listed (i.e., missing) in the file. The code
1072# depends on this ending with a semi-colon, so it can assume it is a valid
1073# field when the line is split() by semi-colons
1074my $missing_defaults_prefix =
1075 qr/^#\s+\@missing:\s+0000\.\.$LAST_UNICODE_CODEPOINT_STRING\s*;/;
1076
1077# Property types. Unicode has more types, but these are sufficient for our
1078# purposes.
1079my $UNKNOWN = -1; # initialized to illegal value
1080my $NON_STRING = 1; # Either binary or enum
1081my $BINARY = 2;
1082my $ENUM = 3; # Include catalog
1083my $STRING = 4; # Anything else: string or misc
1084
1085# Some input files have lines that give default values for code points not
1086# contained in the file. Sometimes these should be ignored.
1087my $NO_DEFAULTS = 0; # Must evaluate to false
1088my $NOT_IGNORED = 1;
1089my $IGNORED = 2;
1090
1091# Range types. Each range has a type. Most ranges are type 0, for normal,
1092# and will appear in the main body of the tables in the output files, but
1093# there are other types of ranges as well, listed below, that are specially
1094# handled. There are pseudo-types as well that will never be stored as a
1095# type, but will affect the calculation of the type.
1096
1097# 0 is for normal, non-specials
1098my $MULTI_CP = 1; # Sequence of more than code point
1099my $HANGUL_SYLLABLE = 2;
1100my $CP_IN_NAME = 3; # The NAME contains the code point appended to it.
1101my $NULL = 4; # The map is to the null string; utf8.c can't
1102 # handle these, nor is there an accepted syntax
1103 # for them in \p{} constructs
f86864ac 1104my $COMPUTE_NO_MULTI_CP = 5; # Pseudo-type; means that ranges that would
99870f4d
KW
1105 # otherwise be $MULTI_CP type are instead type 0
1106
1107# process_generic_property_file() can accept certain overrides in its input.
1108# Each of these must begin AND end with $CMD_DELIM.
1109my $CMD_DELIM = "\a";
1110my $REPLACE_CMD = 'replace'; # Override the Replace
1111my $MAP_TYPE_CMD = 'map_type'; # Override the Type
1112
1113my $NO = 0;
1114my $YES = 1;
1115
1116# Values for the Replace argument to add_range.
1117# $NO # Don't replace; add only the code points not
1118 # already present.
1119my $IF_NOT_EQUIVALENT = 1; # Replace only under certain conditions; details in
1120 # the comments at the subroutine definition.
1121my $UNCONDITIONALLY = 2; # Replace without conditions.
1122my $MULTIPLE = 4; # Don't replace, but add a duplicate record if
1123 # already there
1124
1125# Flags to give property statuses. The phrases are to remind maintainers that
1126# if the flag is changed, the indefinite article referring to it in the
1127# documentation may need to be as well.
1128my $NORMAL = "";
1129my $SUPPRESSED = 'z'; # The character should never actually be seen, since
1130 # it is suppressed
37e2e78e 1131my $PLACEHOLDER = 'P'; # Implies no pod entry generated
99870f4d
KW
1132my $DEPRECATED = 'D';
1133my $a_bold_deprecated = "a 'B<$DEPRECATED>'";
1134my $A_bold_deprecated = "A 'B<$DEPRECATED>'";
1135my $DISCOURAGED = 'X';
1136my $a_bold_discouraged = "an 'B<$DISCOURAGED>'";
1137my $A_bold_discouraged = "An 'B<$DISCOURAGED>'";
1138my $STRICTER = 'T';
1139my $a_bold_stricter = "a 'B<$STRICTER>'";
1140my $A_bold_stricter = "A 'B<$STRICTER>'";
1141my $STABILIZED = 'S';
1142my $a_bold_stabilized = "an 'B<$STABILIZED>'";
1143my $A_bold_stabilized = "An 'B<$STABILIZED>'";
1144my $OBSOLETE = 'O';
1145my $a_bold_obsolete = "an 'B<$OBSOLETE>'";
1146my $A_bold_obsolete = "An 'B<$OBSOLETE>'";
1147
1148my %status_past_participles = (
1149 $DISCOURAGED => 'discouraged',
1150 $SUPPRESSED => 'should never be generated',
1151 $STABILIZED => 'stabilized',
1152 $OBSOLETE => 'obsolete',
37e2e78e 1153 $DEPRECATED => 'deprecated',
99870f4d
KW
1154);
1155
f5817e0a
KW
1156# The format of the values of the tables:
1157my $EMPTY_FORMAT = "";
99870f4d
KW
1158my $BINARY_FORMAT = 'b';
1159my $DECIMAL_FORMAT = 'd';
1160my $FLOAT_FORMAT = 'f';
1161my $INTEGER_FORMAT = 'i';
1162my $HEX_FORMAT = 'x';
1163my $RATIONAL_FORMAT = 'r';
1164my $STRING_FORMAT = 's';
a14f3cb1 1165my $DECOMP_STRING_FORMAT = 'c';
99870f4d
KW
1166
1167my %map_table_formats = (
1168 $BINARY_FORMAT => 'binary',
1169 $DECIMAL_FORMAT => 'single decimal digit',
1170 $FLOAT_FORMAT => 'floating point number',
1171 $INTEGER_FORMAT => 'integer',
1172 $HEX_FORMAT => 'positive hex whole number; a code point',
1173 $RATIONAL_FORMAT => 'rational: an integer or a fraction',
1a9d544b 1174 $STRING_FORMAT => 'string',
92f9d56c 1175 $DECOMP_STRING_FORMAT => 'Perl\'s internal (Normalize.pm) decomposition mapping',
99870f4d
KW
1176);
1177
1178# Unicode didn't put such derived files in a separate directory at first.
1179my $EXTRACTED_DIR = (-d 'extracted') ? 'extracted' : "";
1180my $EXTRACTED = ($EXTRACTED_DIR) ? "$EXTRACTED_DIR/" : "";
1181my $AUXILIARY = 'auxiliary';
1182
1183# Hashes that will eventually go into Heavy.pl for the use of utf8_heavy.pl
1184my %loose_to_file_of; # loosely maps table names to their respective
1185 # files
1186my %stricter_to_file_of; # same; but for stricter mapping.
1187my %nv_floating_to_rational; # maps numeric values floating point numbers to
1188 # their rational equivalent
1189my %loose_property_name_of; # Loosely maps property names to standard form
1190
1191# These constants names and values were taken from the Unicode standard,
1192# version 5.1, section 3.12. They are used in conjunction with Hangul
6e5a209b
KW
1193# syllables. The '_string' versions are so generated tables can retain the
1194# hex format, which is the more familiar value
1195my $SBase_string = "0xAC00";
1196my $SBase = CORE::hex $SBase_string;
1197my $LBase_string = "0x1100";
1198my $LBase = CORE::hex $LBase_string;
1199my $VBase_string = "0x1161";
1200my $VBase = CORE::hex $VBase_string;
1201my $TBase_string = "0x11A7";
1202my $TBase = CORE::hex $TBase_string;
99870f4d
KW
1203my $SCount = 11172;
1204my $LCount = 19;
1205my $VCount = 21;
1206my $TCount = 28;
1207my $NCount = $VCount * $TCount;
1208
1209# For Hangul syllables; These store the numbers from Jamo.txt in conjunction
1210# with the above published constants.
1211my %Jamo;
1212my %Jamo_L; # Leading consonants
1213my %Jamo_V; # Vowels
1214my %Jamo_T; # Trailing consonants
1215
37e2e78e 1216my @backslash_X_tests; # List of tests read in for testing \X
99870f4d
KW
1217my @unhandled_properties; # Will contain a list of properties found in
1218 # the input that we didn't process.
f86864ac 1219my @match_properties; # Properties that have match tables, to be
99870f4d
KW
1220 # listed in the pod
1221my @map_properties; # Properties that get map files written
1222my @named_sequences; # NamedSequences.txt contents.
1223my %potential_files; # Generated list of all .txt files in the directory
1224 # structure so we can warn if something is being
1225 # ignored.
1226my @files_actually_output; # List of files we generated.
1227my @more_Names; # Some code point names are compound; this is used
1228 # to store the extra components of them.
1229my $MIN_FRACTION_LENGTH = 3; # How many digits of a floating point number at
1230 # the minimum before we consider it equivalent to a
1231 # candidate rational
1232my $MAX_FLOATING_SLOP = 10 ** - $MIN_FRACTION_LENGTH; # And in floating terms
1233
1234# These store references to certain commonly used property objects
1235my $gc;
1236my $perl;
1237my $block;
3e20195b
KW
1238my $perl_charname;
1239my $print;
99870f4d
KW
1240
1241# Are there conflicting names because of beginning with 'In_', or 'Is_'
1242my $has_In_conflicts = 0;
1243my $has_Is_conflicts = 0;
1244
1245sub internal_file_to_platform ($) {
1246 # Convert our file paths which have '/' separators to those of the
1247 # platform.
1248
1249 my $file = shift;
1250 return undef unless defined $file;
1251
1252 return File::Spec->join(split '/', $file);
d07a55ed 1253}
5beb625e 1254
99870f4d
KW
1255sub file_exists ($) { # platform independent '-e'. This program internally
1256 # uses slash as a path separator.
1257 my $file = shift;
1258 return 0 if ! defined $file;
1259 return -e internal_file_to_platform($file);
1260}
5beb625e 1261
99870f4d 1262sub objaddr($) {
23e33b60
KW
1263 # Returns the address of the blessed input object.
1264 # It doesn't check for blessedness because that would do a string eval
1265 # every call, and the program is structured so that this is never called
1266 # for a non-blessed object.
99870f4d 1267
23e33b60 1268 no overloading; # If overloaded, numifying below won't work.
99870f4d
KW
1269
1270 # Numifying a ref gives its address.
051df77b 1271 return pack 'J', $_[0];
99870f4d
KW
1272}
1273
558712cf 1274# These are used only if $annotate is true.
c4019d52
KW
1275# The entire range of Unicode characters is examined to populate these
1276# after all the input has been processed. But most can be skipped, as they
1277# have the same descriptive phrases, such as being unassigned
1278my @viacode; # Contains the 1 million character names
1279my @printable; # boolean: And are those characters printable?
1280my @annotate_char_type; # Contains a type of those characters, specifically
1281 # for the purposes of annotation.
1282my $annotate_ranges; # A map of ranges of code points that have the same
1283 # name for the purposes of annoation. They map to the
1284 # upper edge of the range, so that the end point can
1285 # be immediately found. This is used to skip ahead to
1286 # the end of a range, and avoid processing each
1287 # individual code point in it.
1288my $unassigned_sans_noncharacters; # A Range_List of the unassigned
1289 # characters, but excluding those which are
1290 # also noncharacter code points
1291
1292# The annotation types are an extension of the regular range types, though
1293# some of the latter are folded into one. Make the new types negative to
1294# avoid conflicting with the regular types
1295my $SURROGATE_TYPE = -1;
1296my $UNASSIGNED_TYPE = -2;
1297my $PRIVATE_USE_TYPE = -3;
1298my $NONCHARACTER_TYPE = -4;
1299my $CONTROL_TYPE = -5;
1300my $UNKNOWN_TYPE = -6; # Used only if there is a bug in this program
1301
1302sub populate_char_info ($) {
558712cf 1303 # Used only with the $annotate option. Populates the arrays with the
c4019d52
KW
1304 # input code point's info that are needed for outputting more detailed
1305 # comments. If calling context wants a return, it is the end point of
1306 # any contiguous range of characters that share essentially the same info
1307
1308 my $i = shift;
1309 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1310
1311 $viacode[$i] = $perl_charname->value_of($i) || "";
1312
1313 # A character is generally printable if Unicode says it is,
1314 # but below we make sure that most Unicode general category 'C' types
1315 # aren't.
1316 $printable[$i] = $print->contains($i);
1317
1318 $annotate_char_type[$i] = $perl_charname->type_of($i) || 0;
1319
1320 # Only these two regular types are treated specially for annotations
1321 # purposes
1322 $annotate_char_type[$i] = 0 if $annotate_char_type[$i] != $CP_IN_NAME
1323 && $annotate_char_type[$i] != $HANGUL_SYLLABLE;
1324
1325 # Give a generic name to all code points that don't have a real name.
1326 # We output ranges, if applicable, for these. Also calculate the end
1327 # point of the range.
1328 my $end;
1329 if (! $viacode[$i]) {
1330 if ($gc-> table('Surrogate')->contains($i)) {
1331 $viacode[$i] = 'Surrogate';
1332 $annotate_char_type[$i] = $SURROGATE_TYPE;
1333 $printable[$i] = 0;
1334 $end = $gc->table('Surrogate')->containing_range($i)->end;
1335 }
1336 elsif ($gc-> table('Private_use')->contains($i)) {
1337 $viacode[$i] = 'Private Use';
1338 $annotate_char_type[$i] = $PRIVATE_USE_TYPE;
1339 $printable[$i] = 0;
1340 $end = $gc->table('Private_Use')->containing_range($i)->end;
1341 }
1342 elsif (Property::property_ref('Noncharacter_Code_Point')-> table('Y')->
1343 contains($i))
1344 {
1345 $viacode[$i] = 'Noncharacter';
1346 $annotate_char_type[$i] = $NONCHARACTER_TYPE;
1347 $printable[$i] = 0;
1348 $end = property_ref('Noncharacter_Code_Point')->table('Y')->
1349 containing_range($i)->end;
1350 }
1351 elsif ($gc-> table('Control')->contains($i)) {
1352 $viacode[$i] = 'Control';
1353 $annotate_char_type[$i] = $CONTROL_TYPE;
1354 $printable[$i] = 0;
1355 $end = 0x81 if $i == 0x80; # Hard-code this one known case
1356 }
1357 elsif ($gc-> table('Unassigned')->contains($i)) {
1358 $viacode[$i] = 'Unassigned, block=' . $block-> value_of($i);
1359 $annotate_char_type[$i] = $UNASSIGNED_TYPE;
1360 $printable[$i] = 0;
1361
1362 # Because we name the unassigned by the blocks they are in, it
1363 # can't go past the end of that block, and it also can't go past
1364 # the unassigned range it is in. The special table makes sure
1365 # that the non-characters, which are unassigned, are separated
1366 # out.
1367 $end = min($block->containing_range($i)->end,
1368 $unassigned_sans_noncharacters-> containing_range($i)->
1369 end);
1370 } else {
1371 my_carp_bug("Can't figure out how to annotate"
1372 . sprintf("U+%04X", $i)
1373 . "Proceeding anyway.");
1374 $viacode[$i] = 'UNKNOWN';
1375 $annotate_char_type[$i] = $UNKNOWN_TYPE;
1376 $printable[$i] = 0;
1377 }
1378 }
1379
1380 # Here, has a name, but if it's one in which the code point number is
1381 # appended to the name, do that.
1382 elsif ($annotate_char_type[$i] == $CP_IN_NAME) {
1383 $viacode[$i] .= sprintf("-%04X", $i);
1384 $end = $perl_charname->containing_range($i)->end;
1385 }
1386
1387 # And here, has a name, but if it's a hangul syllable one, replace it with
1388 # the correct name from the Unicode algorithm
1389 elsif ($annotate_char_type[$i] == $HANGUL_SYLLABLE) {
1390 use integer;
1391 my $SIndex = $i - $SBase;
1392 my $L = $LBase + $SIndex / $NCount;
1393 my $V = $VBase + ($SIndex % $NCount) / $TCount;
1394 my $T = $TBase + $SIndex % $TCount;
1395 $viacode[$i] = "HANGUL SYLLABLE $Jamo{$L}$Jamo{$V}";
1396 $viacode[$i] .= $Jamo{$T} if $T != $TBase;
1397 $end = $perl_charname->containing_range($i)->end;
1398 }
1399
1400 return if ! defined wantarray;
1401 return $i if ! defined $end; # If not a range, return the input
1402
1403 # Save this whole range so can find the end point quickly
1404 $annotate_ranges->add_map($i, $end, $end);
1405
1406 return $end;
1407}
1408
23e33b60
KW
1409# Commented code below should work on Perl 5.8.
1410## This 'require' doesn't necessarily work in miniperl, and even if it does,
1411## the native perl version of it (which is what would operate under miniperl)
1412## is extremely slow, as it does a string eval every call.
1413#my $has_fast_scalar_util = $\18 !~ /miniperl/
1414# && defined eval "require Scalar::Util";
1415#
1416#sub objaddr($) {
1417# # Returns the address of the blessed input object. Uses the XS version if
1418# # available. It doesn't check for blessedness because that would do a
1419# # string eval every call, and the program is structured so that this is
1420# # never called for a non-blessed object.
1421#
1422# return Scalar::Util::refaddr($_[0]) if $has_fast_scalar_util;
1423#
1424# # Check at least that is a ref.
1425# my $pkg = ref($_[0]) or return undef;
1426#
1427# # Change to a fake package to defeat any overloaded stringify
1428# bless $_[0], 'main::Fake';
1429#
1430# # Numifying a ref gives its address.
051df77b 1431# my $addr = pack 'J', $_[0];
23e33b60
KW
1432#
1433# # Return to original class
1434# bless $_[0], $pkg;
1435# return $addr;
1436#}
1437
99870f4d
KW
1438sub max ($$) {
1439 my $a = shift;
1440 my $b = shift;
1441 return $a if $a >= $b;
1442 return $b;
1443}
1444
1445sub min ($$) {
1446 my $a = shift;
1447 my $b = shift;
1448 return $a if $a <= $b;
1449 return $b;
1450}
1451
1452sub clarify_number ($) {
1453 # This returns the input number with underscores inserted every 3 digits
1454 # in large (5 digits or more) numbers. Input must be entirely digits, not
1455 # checked.
1456
1457 my $number = shift;
1458 my $pos = length($number) - 3;
1459 return $number if $pos <= 1;
1460 while ($pos > 0) {
1461 substr($number, $pos, 0) = '_';
1462 $pos -= 3;
5beb625e 1463 }
99870f4d 1464 return $number;
99598c8c
JH
1465}
1466
12ac2576 1467
99870f4d 1468package Carp;
7ebf06b3 1469
99870f4d
KW
1470# These routines give a uniform treatment of messages in this program. They
1471# are placed in the Carp package to cause the stack trace to not include them,
1472# although an alternative would be to use another package and set @CARP_NOT
1473# for it.
12ac2576 1474
99870f4d 1475our $Verbose = 1 if main::DEBUG; # Useful info when debugging
12ac2576 1476
99f78760
KW
1477# This is a work-around suggested by Nicholas Clark to fix a problem with Carp
1478# and overload trying to load Scalar:Util under miniperl. See
1479# http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2009-11/msg01057.html
1480undef $overload::VERSION;
1481
99870f4d
KW
1482sub my_carp {
1483 my $message = shift || "";
1484 my $nofold = shift || 0;
7ebf06b3 1485
99870f4d
KW
1486 if ($message) {
1487 $message = main::join_lines($message);
1488 $message =~ s/^$0: *//; # Remove initial program name
1489 $message =~ s/[.;,]+$//; # Remove certain ending punctuation
1490 $message = "\n$0: $message;";
12ac2576 1491
99870f4d
KW
1492 # Fold the message with program name, semi-colon end punctuation
1493 # (which looks good with the message that carp appends to it), and a
1494 # hanging indent for continuation lines.
1495 $message = main::simple_fold($message, "", 4) unless $nofold;
1496 $message =~ s/\n$//; # Remove the trailing nl so what carp
1497 # appends is to the same line
1498 }
12ac2576 1499
99870f4d 1500 return $message if defined wantarray; # If a caller just wants the msg
12ac2576 1501
99870f4d
KW
1502 carp $message;
1503 return;
1504}
7ebf06b3 1505
99870f4d
KW
1506sub my_carp_bug {
1507 # This is called when it is clear that the problem is caused by a bug in
1508 # this program.
7ebf06b3 1509
99870f4d
KW
1510 my $message = shift;
1511 $message =~ s/^$0: *//;
1512 $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");
1513 carp $message;
1514 return;
1515}
7ebf06b3 1516
99870f4d
KW
1517sub carp_too_few_args {
1518 if (@_ != 2) {
1519 my_carp_bug("Wrong number of arguments: to 'carp_too_few_arguments'. No action taken.");
1520 return;
12ac2576 1521 }
7ebf06b3 1522
99870f4d
KW
1523 my $args_ref = shift;
1524 my $count = shift;
7ebf06b3 1525
99870f4d
KW
1526 my_carp_bug("Need at least $count arguments to "
1527 . (caller 1)[3]
1528 . ". Instead got: '"
1529 . join ', ', @$args_ref
1530 . "'. No action taken.");
1531 return;
12ac2576
JP
1532}
1533
99870f4d
KW
1534sub carp_extra_args {
1535 my $args_ref = shift;
1536 my_carp_bug("Too many arguments to 'carp_extra_args': (" . join(', ', @_) . "); Extras ignored.") if @_;
12ac2576 1537
99870f4d
KW
1538 unless (ref $args_ref) {
1539 my_carp_bug("Argument to 'carp_extra_args' ($args_ref) must be a ref. Not checking arguments.");
1540 return;
1541 }
1542 my ($package, $file, $line) = caller;
1543 my $subroutine = (caller 1)[3];
cf25bb62 1544
99870f4d
KW
1545 my $list;
1546 if (ref $args_ref eq 'HASH') {
1547 foreach my $key (keys %$args_ref) {
1548 $args_ref->{$key} = $UNDEF unless defined $args_ref->{$key};
cf25bb62 1549 }
99870f4d 1550 $list = join ', ', each %{$args_ref};
cf25bb62 1551 }
99870f4d
KW
1552 elsif (ref $args_ref eq 'ARRAY') {
1553 foreach my $arg (@$args_ref) {
1554 $arg = $UNDEF unless defined $arg;
1555 }
1556 $list = join ', ', @$args_ref;
1557 }
1558 else {
1559 my_carp_bug("Can't cope with ref "
1560 . ref($args_ref)
1561 . " . argument to 'carp_extra_args'. Not checking arguments.");
1562 return;
1563 }
1564
1565 my_carp_bug("Unrecognized parameters in options: '$list' to $subroutine. Skipped.");
1566 return;
d73e5302
JH
1567}
1568
99870f4d
KW
1569package main;
1570
1571{ # Closure
1572
1573 # This program uses the inside-out method for objects, as recommended in
1574 # "Perl Best Practices". This closure aids in generating those. There
1575 # are two routines. setup_package() is called once per package to set
1576 # things up, and then set_access() is called for each hash representing a
1577 # field in the object. These routines arrange for the object to be
1578 # properly destroyed when no longer used, and for standard accessor
1579 # functions to be generated. If you need more complex accessors, just
1580 # write your own and leave those accesses out of the call to set_access().
1581 # More details below.
1582
1583 my %constructor_fields; # fields that are to be used in constructors; see
1584 # below
1585
1586 # The values of this hash will be the package names as keys to other
1587 # hashes containing the name of each field in the package as keys, and
1588 # references to their respective hashes as values.
1589 my %package_fields;
1590
1591 sub setup_package {
1592 # Sets up the package, creating standard DESTROY and dump methods
1593 # (unless already defined). The dump method is used in debugging by
1594 # simple_dumper().
1595 # The optional parameters are:
1596 # a) a reference to a hash, that gets populated by later
1597 # set_access() calls with one of the accesses being
1598 # 'constructor'. The caller can then refer to this, but it is
1599 # not otherwise used by these two routines.
1600 # b) a reference to a callback routine to call during destruction
1601 # of the object, before any fields are actually destroyed
1602
1603 my %args = @_;
1604 my $constructor_ref = delete $args{'Constructor_Fields'};
1605 my $destroy_callback = delete $args{'Destroy_Callback'};
1606 Carp::carp_extra_args(\@_) if main::DEBUG && %args;
1607
1608 my %fields;
1609 my $package = (caller)[0];
1610
1611 $package_fields{$package} = \%fields;
1612 $constructor_fields{$package} = $constructor_ref;
1613
1614 unless ($package->can('DESTROY')) {
1615 my $destroy_name = "${package}::DESTROY";
1616 no strict "refs";
1617
1618 # Use typeglob to give the anonymous subroutine the name we want
1619 *$destroy_name = sub {
1620 my $self = shift;
ffe43484 1621 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1622
1623 $self->$destroy_callback if $destroy_callback;
1624 foreach my $field (keys %{$package_fields{$package}}) {
1625 #print STDERR __LINE__, ": Destroying ", ref $self, " ", sprintf("%04X", $addr), ": ", $field, "\n";
1626 delete $package_fields{$package}{$field}{$addr};
1627 }
1628 return;
1629 }
1630 }
1631
1632 unless ($package->can('dump')) {
1633 my $dump_name = "${package}::dump";
1634 no strict "refs";
1635 *$dump_name = sub {
1636 my $self = shift;
1637 return dump_inside_out($self, $package_fields{$package}, @_);
1638 }
1639 }
1640 return;
1641 }
1642
1643 sub set_access {
1644 # Arrange for the input field to be garbage collected when no longer
1645 # needed. Also, creates standard accessor functions for the field
1646 # based on the optional parameters-- none if none of these parameters:
1647 # 'addable' creates an 'add_NAME()' accessor function.
1648 # 'readable' or 'readable_array' creates a 'NAME()' accessor
1649 # function.
1650 # 'settable' creates a 'set_NAME()' accessor function.
1651 # 'constructor' doesn't create an accessor function, but adds the
1652 # field to the hash that was previously passed to
1653 # setup_package();
1654 # Any of the accesses can be abbreviated down, so that 'a', 'ad',
1655 # 'add' etc. all mean 'addable'.
1656 # The read accessor function will work on both array and scalar
1657 # values. If another accessor in the parameter list is 'a', the read
1658 # access assumes an array. You can also force it to be array access
1659 # by specifying 'readable_array' instead of 'readable'
1660 #
1661 # A sort-of 'protected' access can be set-up by preceding the addable,
1662 # readable or settable with some initial portion of 'protected_' (but,
1663 # the underscore is required), like 'p_a', 'pro_set', etc. The
1664 # "protection" is only by convention. All that happens is that the
1665 # accessor functions' names begin with an underscore. So instead of
1666 # calling set_foo, the call is _set_foo. (Real protection could be
c1739a4a 1667 # accomplished by having a new subroutine, end_package, called at the
99870f4d
KW
1668 # end of each package, and then storing the __LINE__ ranges and
1669 # checking them on every accessor. But that is way overkill.)
1670
1671 # We create anonymous subroutines as the accessors and then use
1672 # typeglobs to assign them to the proper package and name
1673
1674 my $name = shift; # Name of the field
1675 my $field = shift; # Reference to the inside-out hash containing the
1676 # field
1677
1678 my $package = (caller)[0];
1679
1680 if (! exists $package_fields{$package}) {
1681 croak "$0: Must call 'setup_package' before 'set_access'";
1682 }
d73e5302 1683
99870f4d
KW
1684 # Stash the field so DESTROY can get it.
1685 $package_fields{$package}{$name} = $field;
cf25bb62 1686
99870f4d
KW
1687 # Remaining arguments are the accessors. For each...
1688 foreach my $access (@_) {
1689 my $access = lc $access;
cf25bb62 1690
99870f4d 1691 my $protected = "";
cf25bb62 1692
99870f4d
KW
1693 # Match the input as far as it goes.
1694 if ($access =~ /^(p[^_]*)_/) {
1695 $protected = $1;
1696 if (substr('protected_', 0, length $protected)
1697 eq $protected)
1698 {
1699
1700 # Add 1 for the underscore not included in $protected
1701 $access = substr($access, length($protected) + 1);
1702 $protected = '_';
1703 }
1704 else {
1705 $protected = "";
1706 }
1707 }
1708
1709 if (substr('addable', 0, length $access) eq $access) {
1710 my $subname = "${package}::${protected}add_$name";
1711 no strict "refs";
1712
1713 # add_ accessor. Don't add if already there, which we
1714 # determine using 'eq' for scalars and '==' otherwise.
1715 *$subname = sub {
1716 use strict "refs";
1717 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
1718 my $self = shift;
1719 my $value = shift;
ffe43484 1720 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1721 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
1722 if (ref $value) {
f998e60c 1723 return if grep { $value == $_ } @{$field->{$addr}};
99870f4d
KW
1724 }
1725 else {
f998e60c 1726 return if grep { $value eq $_ } @{$field->{$addr}};
99870f4d 1727 }
f998e60c 1728 push @{$field->{$addr}}, $value;
99870f4d
KW
1729 return;
1730 }
1731 }
1732 elsif (substr('constructor', 0, length $access) eq $access) {
1733 if ($protected) {
1734 Carp::my_carp_bug("Can't set-up 'protected' constructors")
1735 }
1736 else {
1737 $constructor_fields{$package}{$name} = $field;
1738 }
1739 }
1740 elsif (substr('readable_array', 0, length $access) eq $access) {
1741
1742 # Here has read access. If one of the other parameters for
1743 # access is array, or this one specifies array (by being more
1744 # than just 'readable_'), then create a subroutine that
1745 # assumes the data is an array. Otherwise just a scalar
1746 my $subname = "${package}::${protected}$name";
1747 if (grep { /^a/i } @_
1748 or length($access) > length('readable_'))
1749 {
1750 no strict "refs";
1751 *$subname = sub {
1752 use strict "refs";
23e33b60 1753 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
ffe43484 1754 my $addr = do { no overloading; pack 'J', $_[0]; };
99870f4d
KW
1755 if (ref $field->{$addr} ne 'ARRAY') {
1756 my $type = ref $field->{$addr};
1757 $type = 'scalar' unless $type;
1758 Carp::my_carp_bug("Trying to read $name as an array when it is a $type. Big problems.");
1759 return;
1760 }
1761 return scalar @{$field->{$addr}} unless wantarray;
1762
1763 # Make a copy; had problems with caller modifying the
1764 # original otherwise
1765 my @return = @{$field->{$addr}};
1766 return @return;
1767 }
1768 }
1769 else {
1770
1771 # Here not an array value, a simpler function.
1772 no strict "refs";
1773 *$subname = sub {
1774 use strict "refs";
23e33b60 1775 Carp::carp_extra_args(\@_) if main::DEBUG && @_ > 1;
f998e60c 1776 no overloading;
051df77b 1777 return $field->{pack 'J', $_[0]};
99870f4d
KW
1778 }
1779 }
1780 }
1781 elsif (substr('settable', 0, length $access) eq $access) {
1782 my $subname = "${package}::${protected}set_$name";
1783 no strict "refs";
1784 *$subname = sub {
1785 use strict "refs";
23e33b60
KW
1786 if (main::DEBUG) {
1787 return Carp::carp_too_few_args(\@_, 2) if @_ < 2;
1788 Carp::carp_extra_args(\@_) if @_ > 2;
1789 }
1790 # $self is $_[0]; $value is $_[1]
f998e60c 1791 no overloading;
051df77b 1792 $field->{pack 'J', $_[0]} = $_[1];
99870f4d
KW
1793 return;
1794 }
1795 }
1796 else {
1797 Carp::my_carp_bug("Unknown accessor type $access. No accessor set.");
1798 }
cf25bb62 1799 }
99870f4d 1800 return;
cf25bb62 1801 }
99870f4d
KW
1802}
1803
1804package Input_file;
1805
1806# All input files use this object, which stores various attributes about them,
1807# and provides for convenient, uniform handling. The run method wraps the
1808# processing. It handles all the bookkeeping of opening, reading, and closing
1809# the file, returning only significant input lines.
1810#
1811# Each object gets a handler which processes the body of the file, and is
1812# called by run(). Most should use the generic, default handler, which has
1813# code scrubbed to handle things you might not expect. A handler should
1814# basically be a while(next_line()) {...} loop.
1815#
1816# You can also set up handlers to
1817# 1) call before the first line is read for pre processing
1818# 2) call to adjust each line of the input before the main handler gets them
1819# 3) call upon EOF before the main handler exits its loop
1820# 4) call at the end for post processing
1821#
1822# $_ is used to store the input line, and is to be filtered by the
1823# each_line_handler()s. So, if the format of the line is not in the desired
1824# format for the main handler, these are used to do that adjusting. They can
1825# be stacked (by enclosing them in an [ anonymous array ] in the constructor,
1826# so the $_ output of one is used as the input to the next. None of the other
1827# handlers are stackable, but could easily be changed to be so.
1828#
1829# Most of the handlers can call insert_lines() or insert_adjusted_lines()
1830# which insert the parameters as lines to be processed before the next input
1831# file line is read. This allows the EOF handler to flush buffers, for
1832# example. The difference between the two routines is that the lines inserted
1833# by insert_lines() are subjected to the each_line_handler()s. (So if you
1834# called it from such a handler, you would get infinite recursion.) Lines
1835# inserted by insert_adjusted_lines() go directly to the main handler without
1836# any adjustments. If the post-processing handler calls any of these, there
1837# will be no effect. Some error checking for these conditions could be added,
1838# but it hasn't been done.
1839#
1840# carp_bad_line() should be called to warn of bad input lines, which clears $_
1841# to prevent further processing of the line. This routine will output the
1842# message as a warning once, and then keep a count of the lines that have the
1843# same message, and output that count at the end of the file's processing.
1844# This keeps the number of messages down to a manageable amount.
1845#
1846# get_missings() should be called to retrieve any @missing input lines.
1847# Messages will be raised if this isn't done if the options aren't to ignore
1848# missings.
1849
1850sub trace { return main::trace(@_); }
1851
99870f4d
KW
1852{ # Closure
1853 # Keep track of fields that are to be put into the constructor.
1854 my %constructor_fields;
1855
1856 main::setup_package(Constructor_Fields => \%constructor_fields);
1857
1858 my %file; # Input file name, required
1859 main::set_access('file', \%file, qw{ c r });
1860
1861 my %first_released; # Unicode version file was first released in, required
1862 main::set_access('first_released', \%first_released, qw{ c r });
1863
1864 my %handler; # Subroutine to process the input file, defaults to
1865 # 'process_generic_property_file'
1866 main::set_access('handler', \%handler, qw{ c });
1867
1868 my %property;
1869 # name of property this file is for. defaults to none, meaning not
1870 # applicable, or is otherwise determinable, for example, from each line.
1871 main::set_access('property', \%property, qw{ c });
1872
1873 my %optional;
1874 # If this is true, the file is optional. If not present, no warning is
1875 # output. If it is present, the string given by this parameter is
1876 # evaluated, and if false the file is not processed.
1877 main::set_access('optional', \%optional, 'c', 'r');
1878
1879 my %non_skip;
1880 # This is used for debugging, to skip processing of all but a few input
1881 # files. Add 'non_skip => 1' to the constructor for those files you want
1882 # processed when you set the $debug_skip global.
1883 main::set_access('non_skip', \%non_skip, 'c');
1884
37e2e78e
KW
1885 my %skip;
1886 # This is used to skip processing of this input file semi-permanently.
1887 # It is used for files that we aren't planning to process anytime soon,
1888 # but want to allow to be in the directory and not raise a message that we
1889 # are not handling. Mostly for test files. This is in contrast to the
1890 # non_skip element, which is supposed to be used very temporarily for
1891 # debugging. Sets 'optional' to 1
1892 main::set_access('skip', \%skip, 'c');
1893
99870f4d
KW
1894 my %each_line_handler;
1895 # list of subroutines to look at and filter each non-comment line in the
1896 # file. defaults to none. The subroutines are called in order, each is
1897 # to adjust $_ for the next one, and the final one adjusts it for
1898 # 'handler'
1899 main::set_access('each_line_handler', \%each_line_handler, 'c');
1900
1901 my %has_missings_defaults;
1902 # ? Are there lines in the file giving default values for code points
1903 # missing from it?. Defaults to NO_DEFAULTS. Otherwise NOT_IGNORED is
1904 # the norm, but IGNORED means it has such lines, but the handler doesn't
1905 # use them. Having these three states allows us to catch changes to the
1906 # UCD that this program should track
1907 main::set_access('has_missings_defaults',
1908 \%has_missings_defaults, qw{ c r });
1909
1910 my %pre_handler;
1911 # Subroutine to call before doing anything else in the file. If undef, no
1912 # such handler is called.
1913 main::set_access('pre_handler', \%pre_handler, qw{ c });
1914
1915 my %eof_handler;
1916 # Subroutine to call upon getting an EOF on the input file, but before
1917 # that is returned to the main handler. This is to allow buffers to be
1918 # flushed. The handler is expected to call insert_lines() or
1919 # insert_adjusted() with the buffered material
1920 main::set_access('eof_handler', \%eof_handler, qw{ c r });
1921
1922 my %post_handler;
1923 # Subroutine to call after all the lines of the file are read in and
1924 # processed. If undef, no such handler is called.
1925 main::set_access('post_handler', \%post_handler, qw{ c });
1926
1927 my %progress_message;
1928 # Message to print to display progress in lieu of the standard one
1929 main::set_access('progress_message', \%progress_message, qw{ c });
1930
1931 my %handle;
1932 # cache open file handle, internal. Is undef if file hasn't been
1933 # processed at all, empty if has;
1934 main::set_access('handle', \%handle);
1935
1936 my %added_lines;
1937 # cache of lines added virtually to the file, internal
1938 main::set_access('added_lines', \%added_lines);
1939
1940 my %errors;
1941 # cache of errors found, internal
1942 main::set_access('errors', \%errors);
1943
1944 my %missings;
1945 # storage of '@missing' defaults lines
1946 main::set_access('missings', \%missings);
1947
1948 sub new {
1949 my $class = shift;
1950
1951 my $self = bless \do{ my $anonymous_scalar }, $class;
ffe43484 1952 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
1953
1954 # Set defaults
1955 $handler{$addr} = \&main::process_generic_property_file;
1956 $non_skip{$addr} = 0;
37e2e78e 1957 $skip{$addr} = 0;
99870f4d
KW
1958 $has_missings_defaults{$addr} = $NO_DEFAULTS;
1959 $handle{$addr} = undef;
1960 $added_lines{$addr} = [ ];
1961 $each_line_handler{$addr} = [ ];
1962 $errors{$addr} = { };
1963 $missings{$addr} = [ ];
1964
1965 # Two positional parameters.
99f78760 1966 return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
99870f4d
KW
1967 $file{$addr} = main::internal_file_to_platform(shift);
1968 $first_released{$addr} = shift;
1969
1970 # The rest of the arguments are key => value pairs
1971 # %constructor_fields has been set up earlier to list all possible
1972 # ones. Either set or push, depending on how the default has been set
1973 # up just above.
1974 my %args = @_;
1975 foreach my $key (keys %args) {
1976 my $argument = $args{$key};
1977
1978 # Note that the fields are the lower case of the constructor keys
1979 my $hash = $constructor_fields{lc $key};
1980 if (! defined $hash) {
1981 Carp::my_carp_bug("Unrecognized parameters '$key => $argument' to new() for $self. Skipped");
1982 next;
1983 }
1984 if (ref $hash->{$addr} eq 'ARRAY') {
1985 if (ref $argument eq 'ARRAY') {
1986 foreach my $argument (@{$argument}) {
1987 next if ! defined $argument;
1988 push @{$hash->{$addr}}, $argument;
1989 }
1990 }
1991 else {
1992 push @{$hash->{$addr}}, $argument if defined $argument;
1993 }
1994 }
1995 else {
1996 $hash->{$addr} = $argument;
1997 }
1998 delete $args{$key};
1999 };
2000
2001 # If the file has a property for it, it means that the property is not
2002 # listed in the file's entries. So add a handler to the list of line
2003 # handlers to insert the property name into the lines, to provide a
2004 # uniform interface to the final processing subroutine.
2005 # the final code doesn't have to worry about that.
2006 if ($property{$addr}) {
2007 push @{$each_line_handler{$addr}}, \&_insert_property_into_line;
2008 }
2009
2010 if ($non_skip{$addr} && ! $debug_skip && $verbosity) {
2011 print "Warning: " . __PACKAGE__ . " constructor for $file{$addr} has useless 'non_skip' in it\n";
a3a8c5f0 2012 }
99870f4d 2013
37e2e78e
KW
2014 $optional{$addr} = 1 if $skip{$addr};
2015
99870f4d 2016 return $self;
d73e5302
JH
2017 }
2018
cf25bb62 2019
99870f4d
KW
2020 use overload
2021 fallback => 0,
2022 qw("") => "_operator_stringify",
2023 "." => \&main::_operator_dot,
2024 ;
cf25bb62 2025
99870f4d
KW
2026 sub _operator_stringify {
2027 my $self = shift;
cf25bb62 2028
99870f4d 2029 return __PACKAGE__ . " object for " . $self->file;
d73e5302 2030 }
d73e5302 2031
99870f4d
KW
2032 # flag to make sure extracted files are processed early
2033 my $seen_non_extracted_non_age = 0;
d73e5302 2034
99870f4d
KW
2035 sub run {
2036 # Process the input object $self. This opens and closes the file and
2037 # calls all the handlers for it. Currently, this can only be called
2038 # once per file, as it destroy's the EOF handler
d73e5302 2039
99870f4d
KW
2040 my $self = shift;
2041 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
b6922eda 2042
ffe43484 2043 my $addr = do { no overloading; pack 'J', $self; };
b6922eda 2044
99870f4d 2045 my $file = $file{$addr};
d73e5302 2046
99870f4d
KW
2047 # Don't process if not expecting this file (because released later
2048 # than this Unicode version), and isn't there. This means if someone
2049 # copies it into an earlier version's directory, we will go ahead and
2050 # process it.
2051 return if $first_released{$addr} gt $v_version && ! -e $file;
2052
2053 # If in debugging mode and this file doesn't have the non-skip
2054 # flag set, and isn't one of the critical files, skip it.
2055 if ($debug_skip
2056 && $first_released{$addr} ne v0
2057 && ! $non_skip{$addr})
2058 {
2059 print "Skipping $file in debugging\n" if $verbosity;
2060 return;
2061 }
2062
2063 # File could be optional
37e2e78e 2064 if ($optional{$addr}) {
99870f4d
KW
2065 return unless -e $file;
2066 my $result = eval $optional{$addr};
2067 if (! defined $result) {
2068 Carp::my_carp_bug("Got '$@' when tried to eval $optional{$addr}. $file Skipped.");
2069 return;
2070 }
2071 if (! $result) {
2072 if ($verbosity) {
2073 print STDERR "Skipping processing input file '$file' because '$optional{$addr}' is not true\n";
2074 }
2075 return;
2076 }
2077 }
2078
2079 if (! defined $file || ! -e $file) {
2080
2081 # If the file doesn't exist, see if have internal data for it
2082 # (based on first_released being 0).
2083 if ($first_released{$addr} eq v0) {
2084 $handle{$addr} = 'pretend_is_open';
2085 }
2086 else {
2087 if (! $optional{$addr} # File could be optional
2088 && $v_version ge $first_released{$addr})
2089 {
2090 print STDERR "Skipping processing input file '$file' because not found\n" if $v_version ge $first_released{$addr};
2091 }
2092 return;
2093 }
2094 }
2095 else {
2096
37e2e78e
KW
2097 # Here, the file exists. Some platforms may change the case of
2098 # its name
99870f4d 2099 if ($seen_non_extracted_non_age) {
517956bf 2100 if ($file =~ /$EXTRACTED/i) {
99870f4d 2101 Carp::my_carp_bug(join_lines(<<END
99f78760 2102$file should be processed just after the 'Prop...Alias' files, and before
99870f4d
KW
2103anything not in the $EXTRACTED_DIR directory. Proceeding, but the results may
2104have subtle problems
2105END
2106 ));
2107 }
2108 }
2109 elsif ($EXTRACTED_DIR
2110 && $first_released{$addr} ne v0
517956bf
CB
2111 && $file !~ /$EXTRACTED/i
2112 && lc($file) ne 'dage.txt')
99870f4d
KW
2113 {
2114 # We don't set this (by the 'if' above) if we have no
2115 # extracted directory, so if running on an early version,
2116 # this test won't work. Not worth worrying about.
2117 $seen_non_extracted_non_age = 1;
2118 }
2119
2120 # And mark the file as having being processed, and warn if it
2121 # isn't a file we are expecting. As we process the files,
2122 # they are deleted from the hash, so any that remain at the
2123 # end of the program are files that we didn't process.
517956bf
CB
2124 my $fkey = File::Spec->rel2abs($file);
2125 my $expecting = delete $potential_files{$fkey};
2126 $expecting = delete $potential_files{lc($fkey)} unless defined $expecting;
678f13d5
KW
2127 Carp::my_carp("Was not expecting '$file'.") if
2128 ! $expecting
99870f4d
KW
2129 && ! defined $handle{$addr};
2130
37e2e78e
KW
2131 # Having deleted from expected files, we can quit if not to do
2132 # anything. Don't print progress unless really want verbosity
2133 if ($skip{$addr}) {
2134 print "Skipping $file.\n" if $verbosity >= $VERBOSE;
2135 return;
2136 }
2137
99870f4d
KW
2138 # Open the file, converting the slashes used in this program
2139 # into the proper form for the OS
2140 my $file_handle;
2141 if (not open $file_handle, "<", $file) {
2142 Carp::my_carp("Can't open $file. Skipping: $!");
2143 return 0;
2144 }
2145 $handle{$addr} = $file_handle; # Cache the open file handle
2146 }
2147
2148 if ($verbosity >= $PROGRESS) {
2149 if ($progress_message{$addr}) {
2150 print "$progress_message{$addr}\n";
2151 }
2152 else {
2153 # If using a virtual file, say so.
2154 print "Processing ", (-e $file)
2155 ? $file
2156 : "substitute $file",
2157 "\n";
2158 }
2159 }
2160
2161
2162 # Call any special handler for before the file.
2163 &{$pre_handler{$addr}}($self) if $pre_handler{$addr};
2164
2165 # Then the main handler
2166 &{$handler{$addr}}($self);
2167
2168 # Then any special post-file handler.
2169 &{$post_handler{$addr}}($self) if $post_handler{$addr};
2170
2171 # If any errors have been accumulated, output the counts (as the first
2172 # error message in each class was output when it was encountered).
2173 if ($errors{$addr}) {
2174 my $total = 0;
2175 my $types = 0;
2176 foreach my $error (keys %{$errors{$addr}}) {
2177 $total += $errors{$addr}->{$error};
2178 delete $errors{$addr}->{$error};
2179 $types++;
2180 }
2181 if ($total > 1) {
2182 my $message
2183 = "A total of $total lines had errors in $file. ";
2184
2185 $message .= ($types == 1)
2186 ? '(Only the first one was displayed.)'
2187 : '(Only the first of each type was displayed.)';
2188 Carp::my_carp($message);
2189 }
2190 }
2191
2192 if (@{$missings{$addr}}) {
2193 Carp::my_carp_bug("Handler for $file didn't look at all the \@missing lines. Generated tables likely are wrong");
2194 }
2195
2196 # If a real file handle, close it.
2197 close $handle{$addr} or Carp::my_carp("Can't close $file: $!") if
2198 ref $handle{$addr};
2199 $handle{$addr} = ""; # Uses empty to indicate that has already seen
2200 # the file, as opposed to undef
2201 return;
2202 }
2203
2204 sub next_line {
2205 # Sets $_ to be the next logical input line, if any. Returns non-zero
2206 # if such a line exists. 'logical' means that any lines that have
2207 # been added via insert_lines() will be returned in $_ before the file
2208 # is read again.
2209
2210 my $self = shift;
2211 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2212
ffe43484 2213 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2214
2215 # Here the file is open (or if the handle is not a ref, is an open
2216 # 'virtual' file). Get the next line; any inserted lines get priority
2217 # over the file itself.
2218 my $adjusted;
2219
2220 LINE:
2221 while (1) { # Loop until find non-comment, non-empty line
2222 #local $to_trace = 1 if main::DEBUG;
2223 my $inserted_ref = shift @{$added_lines{$addr}};
2224 if (defined $inserted_ref) {
2225 ($adjusted, $_) = @{$inserted_ref};
2226 trace $adjusted, $_ if main::DEBUG && $to_trace;
2227 return 1 if $adjusted;
2228 }
2229 else {
2230 last if ! ref $handle{$addr}; # Don't read unless is real file
2231 last if ! defined ($_ = readline $handle{$addr});
2232 }
2233 chomp;
2234 trace $_ if main::DEBUG && $to_trace;
2235
2236 # See if this line is the comment line that defines what property
2237 # value that code points that are not listed in the file should
2238 # have. The format or existence of these lines is not guaranteed
2239 # by Unicode since they are comments, but the documentation says
2240 # that this was added for machine-readability, so probably won't
2241 # change. This works starting in Unicode Version 5.0. They look
2242 # like:
2243 #
2244 # @missing: 0000..10FFFF; Not_Reordered
2245 # @missing: 0000..10FFFF; Decomposition_Mapping; <code point>
2246 # @missing: 0000..10FFFF; ; NaN
2247 #
2248 # Save the line for a later get_missings() call.
2249 if (/$missing_defaults_prefix/) {
2250 if ($has_missings_defaults{$addr} == $NO_DEFAULTS) {
2251 $self->carp_bad_line("Unexpected \@missing line. Assuming no missing entries");
2252 }
2253 elsif ($has_missings_defaults{$addr} == $NOT_IGNORED) {
2254 my @defaults = split /\s* ; \s*/x, $_;
2255
2256 # The first field is the @missing, which ends in a
2257 # semi-colon, so can safely shift.
2258 shift @defaults;
2259
2260 # Some of these lines may have empty field placeholders
2261 # which get in the way. An example is:
2262 # @missing: 0000..10FFFF; ; NaN
2263 # Remove them. Process starting from the top so the
2264 # splice doesn't affect things still to be looked at.
2265 for (my $i = @defaults - 1; $i >= 0; $i--) {
2266 next if $defaults[$i] ne "";
2267 splice @defaults, $i, 1;
2268 }
2269
2270 # What's left should be just the property (maybe) and the
2271 # default. Having only one element means it doesn't have
2272 # the property.
2273 my $default;
2274 my $property;
2275 if (@defaults >= 1) {
2276 if (@defaults == 1) {
2277 $default = $defaults[0];
2278 }
2279 else {
2280 $property = $defaults[0];
2281 $default = $defaults[1];
2282 }
2283 }
2284
2285 if (@defaults < 1
2286 || @defaults > 2
2287 || ($default =~ /^</
2288 && $default !~ /^<code *point>$/i
2289 && $default !~ /^<none>$/i))
2290 {
2291 $self->carp_bad_line("Unrecognized \@missing line: $_. Assuming no missing entries");
2292 }
2293 else {
2294
2295 # If the property is missing from the line, it should
2296 # be the one for the whole file
2297 $property = $property{$addr} if ! defined $property;
2298
2299 # Change <none> to the null string, which is what it
2300 # really means. If the default is the code point
2301 # itself, set it to <code point>, which is what
2302 # Unicode uses (but sometimes they've forgotten the
2303 # space)
2304 if ($default =~ /^<none>$/i) {
2305 $default = "";
2306 }
2307 elsif ($default =~ /^<code *point>$/i) {
2308 $default = $CODE_POINT;
2309 }
2310
2311 # Store them as a sub-arrays with both components.
2312 push @{$missings{$addr}}, [ $default, $property ];
2313 }
2314 }
2315
2316 # There is nothing for the caller to process on this comment
2317 # line.
2318 next;
2319 }
2320
2321 # Remove comments and trailing space, and skip this line if the
2322 # result is empty
2323 s/#.*//;
2324 s/\s+$//;
2325 next if /^$/;
2326
2327 # Call any handlers for this line, and skip further processing of
2328 # the line if the handler sets the line to null.
2329 foreach my $sub_ref (@{$each_line_handler{$addr}}) {
2330 &{$sub_ref}($self);
2331 next LINE if /^$/;
2332 }
2333
2334 # Here the line is ok. return success.
2335 return 1;
2336 } # End of looping through lines.
2337
2338 # If there is an EOF handler, call it (only once) and if it generates
2339 # more lines to process go back in the loop to handle them.
2340 if ($eof_handler{$addr}) {
2341 &{$eof_handler{$addr}}($self);
2342 $eof_handler{$addr} = ""; # Currently only get one shot at it.
2343 goto LINE if $added_lines{$addr};
2344 }
2345
2346 # Return failure -- no more lines.
2347 return 0;
2348
2349 }
2350
2351# Not currently used, not fully tested.
2352# sub peek {
2353# # Non-destructive look-ahead one non-adjusted, non-comment, non-blank
2354# # record. Not callable from an each_line_handler(), nor does it call
2355# # an each_line_handler() on the line.
2356#
2357# my $self = shift;
ffe43484 2358# my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2359#
2360# foreach my $inserted_ref (@{$added_lines{$addr}}) {
2361# my ($adjusted, $line) = @{$inserted_ref};
2362# next if $adjusted;
2363#
2364# # Remove comments and trailing space, and return a non-empty
2365# # resulting line
2366# $line =~ s/#.*//;
2367# $line =~ s/\s+$//;
2368# return $line if $line ne "";
2369# }
2370#
2371# return if ! ref $handle{$addr}; # Don't read unless is real file
2372# while (1) { # Loop until find non-comment, non-empty line
2373# local $to_trace = 1 if main::DEBUG;
2374# trace $_ if main::DEBUG && $to_trace;
2375# return if ! defined (my $line = readline $handle{$addr});
2376# chomp $line;
2377# push @{$added_lines{$addr}}, [ 0, $line ];
2378#
2379# $line =~ s/#.*//;
2380# $line =~ s/\s+$//;
2381# return $line if $line ne "";
2382# }
2383#
2384# return;
2385# }
2386
2387
2388 sub insert_lines {
2389 # Lines can be inserted so that it looks like they were in the input
2390 # file at the place it was when this routine is called. See also
2391 # insert_adjusted_lines(). Lines inserted via this routine go through
2392 # any each_line_handler()
2393
2394 my $self = shift;
2395
2396 # Each inserted line is an array, with the first element being 0 to
2397 # indicate that this line hasn't been adjusted, and needs to be
2398 # processed.
f998e60c 2399 no overloading;
051df77b 2400 push @{$added_lines{pack 'J', $self}}, map { [ 0, $_ ] } @_;
99870f4d
KW
2401 return;
2402 }
2403
2404 sub insert_adjusted_lines {
2405 # Lines can be inserted so that it looks like they were in the input
2406 # file at the place it was when this routine is called. See also
2407 # insert_lines(). Lines inserted via this routine are already fully
2408 # adjusted, ready to be processed; each_line_handler()s handlers will
2409 # not be called. This means this is not a completely general
2410 # facility, as only the last each_line_handler on the stack should
2411 # call this. It could be made more general, by passing to each of the
2412 # line_handlers their position on the stack, which they would pass on
2413 # to this routine, and that would replace the boolean first element in
2414 # the anonymous array pushed here, so that the next_line routine could
2415 # use that to call only those handlers whose index is after it on the
2416 # stack. But this is overkill for what is needed now.
2417
2418 my $self = shift;
2419 trace $_[0] if main::DEBUG && $to_trace;
2420
2421 # Each inserted line is an array, with the first element being 1 to
2422 # indicate that this line has been adjusted
f998e60c 2423 no overloading;
051df77b 2424 push @{$added_lines{pack 'J', $self}}, map { [ 1, $_ ] } @_;
99870f4d
KW
2425 return;
2426 }
2427
2428 sub get_missings {
2429 # Returns the stored up @missings lines' values, and clears the list.
2430 # The values are in an array, consisting of the default in the first
2431 # element, and the property in the 2nd. However, since these lines
2432 # can be stacked up, the return is an array of all these arrays.
2433
2434 my $self = shift;
2435 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2436
ffe43484 2437 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2438
2439 # If not accepting a list return, just return the first one.
2440 return shift @{$missings{$addr}} unless wantarray;
2441
2442 my @return = @{$missings{$addr}};
2443 undef @{$missings{$addr}};
2444 return @return;
2445 }
2446
2447 sub _insert_property_into_line {
2448 # Add a property field to $_, if this file requires it.
2449
f998e60c 2450 my $self = shift;
ffe43484 2451 my $addr = do { no overloading; pack 'J', $self; };
f998e60c 2452 my $property = $property{$addr};
99870f4d
KW
2453 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2454
2455 $_ =~ s/(;|$)/; $property$1/;
2456 return;
2457 }
2458
2459 sub carp_bad_line {
2460 # Output consistent error messages, using either a generic one, or the
2461 # one given by the optional parameter. To avoid gazillions of the
2462 # same message in case the syntax of a file is way off, this routine
2463 # only outputs the first instance of each message, incrementing a
2464 # count so the totals can be output at the end of the file.
2465
2466 my $self = shift;
2467 my $message = shift;
2468 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2469
ffe43484 2470 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2471
2472 $message = 'Unexpected line' unless $message;
2473
2474 # No trailing punctuation so as to fit with our addenda.
2475 $message =~ s/[.:;,]$//;
2476
2477 # If haven't seen this exact message before, output it now. Otherwise
2478 # increment the count of how many times it has occurred
2479 unless ($errors{$addr}->{$message}) {
2480 Carp::my_carp("$message in '$_' in "
f998e60c 2481 . $file{$addr}
99870f4d
KW
2482 . " at line $.. Skipping this line;");
2483 $errors{$addr}->{$message} = 1;
2484 }
2485 else {
2486 $errors{$addr}->{$message}++;
2487 }
2488
2489 # Clear the line to prevent any further (meaningful) processing of it.
2490 $_ = "";
2491
2492 return;
2493 }
2494} # End closure
2495
2496package Multi_Default;
2497
2498# Certain properties in early versions of Unicode had more than one possible
2499# default for code points missing from the files. In these cases, one
2500# default applies to everything left over after all the others are applied,
2501# and for each of the others, there is a description of which class of code
2502# points applies to it. This object helps implement this by storing the
2503# defaults, and for all but that final default, an eval string that generates
2504# the class that it applies to.
2505
2506
2507{ # Closure
2508
2509 main::setup_package();
2510
2511 my %class_defaults;
2512 # The defaults structure for the classes
2513 main::set_access('class_defaults', \%class_defaults);
2514
2515 my %other_default;
2516 # The default that applies to everything left over.
2517 main::set_access('other_default', \%other_default, 'r');
2518
2519
2520 sub new {
2521 # The constructor is called with default => eval pairs, terminated by
2522 # the left-over default. e.g.
2523 # Multi_Default->new(
2524 # 'T' => '$gc->table("Mn") + $gc->table("Cf") - 0x200C
2525 # - 0x200D',
2526 # 'R' => 'some other expression that evaluates to code points',
2527 # .
2528 # .
2529 # .
2530 # 'U'));
2531
2532 my $class = shift;
2533
2534 my $self = bless \do{my $anonymous_scalar}, $class;
ffe43484 2535 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2536
2537 while (@_ > 1) {
2538 my $default = shift;
2539 my $eval = shift;
2540 $class_defaults{$addr}->{$default} = $eval;
2541 }
2542
2543 $other_default{$addr} = shift;
2544
2545 return $self;
2546 }
2547
2548 sub get_next_defaults {
2549 # Iterates and returns the next class of defaults.
2550 my $self = shift;
2551 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2552
ffe43484 2553 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2554
2555 return each %{$class_defaults{$addr}};
2556 }
2557}
2558
2559package Alias;
2560
2561# An alias is one of the names that a table goes by. This class defines them
2562# including some attributes. Everything is currently setup in the
2563# constructor.
2564
2565
2566{ # Closure
2567
2568 main::setup_package();
2569
2570 my %name;
2571 main::set_access('name', \%name, 'r');
2572
2573 my %loose_match;
2574 # Determined by the constructor code if this name should match loosely or
2575 # not. The constructor parameters can override this, but it isn't fully
2576 # implemented, as should have ability to override Unicode one's via
2577 # something like a set_loose_match()
2578 main::set_access('loose_match', \%loose_match, 'r');
2579
2580 my %make_pod_entry;
2581 # Some aliases should not get their own entries because they are covered
2582 # by a wild-card, and some we want to discourage use of. Binary
2583 main::set_access('make_pod_entry', \%make_pod_entry, 'r');
2584
2585 my %status;
2586 # Aliases have a status, like deprecated, or even suppressed (which means
2587 # they don't appear in documentation). Enum
2588 main::set_access('status', \%status, 'r');
2589
2590 my %externally_ok;
2591 # Similarly, some aliases should not be considered as usable ones for
2592 # external use, such as file names, or we don't want documentation to
2593 # recommend them. Boolean
2594 main::set_access('externally_ok', \%externally_ok, 'r');
2595
2596 sub new {
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 $name{$addr} = shift;
2603 $loose_match{$addr} = shift;
2604 $make_pod_entry{$addr} = shift;
2605 $externally_ok{$addr} = shift;
2606 $status{$addr} = shift;
2607
2608 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2609
2610 # Null names are never ok externally
2611 $externally_ok{$addr} = 0 if $name{$addr} eq "";
2612
2613 return $self;
2614 }
2615}
2616
2617package Range;
2618
2619# A range is the basic unit for storing code points, and is described in the
2620# comments at the beginning of the program. Each range has a starting code
2621# point; an ending code point (not less than the starting one); a value
2622# that applies to every code point in between the two end-points, inclusive;
2623# and an enum type that applies to the value. The type is for the user's
2624# convenience, and has no meaning here, except that a non-zero type is
2625# considered to not obey the normal Unicode rules for having standard forms.
2626#
2627# The same structure is used for both map and match tables, even though in the
2628# latter, the value (and hence type) is irrelevant and could be used as a
2629# comment. In map tables, the value is what all the code points in the range
2630# map to. Type 0 values have the standardized version of the value stored as
2631# well, so as to not have to recalculate it a lot.
2632
2633sub trace { return main::trace(@_); }
2634
2635{ # Closure
2636
2637 main::setup_package();
2638
2639 my %start;
2640 main::set_access('start', \%start, 'r', 's');
2641
2642 my %end;
2643 main::set_access('end', \%end, 'r', 's');
2644
2645 my %value;
2646 main::set_access('value', \%value, 'r');
2647
2648 my %type;
2649 main::set_access('type', \%type, 'r');
2650
2651 my %standard_form;
2652 # The value in internal standard form. Defined only if the type is 0.
2653 main::set_access('standard_form', \%standard_form);
2654
2655 # Note that if these fields change, the dump() method should as well
2656
2657 sub new {
2658 return Carp::carp_too_few_args(\@_, 3) if main::DEBUG && @_ < 3;
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 $start{$addr} = shift;
2665 $end{$addr} = shift;
2666
2667 my %args = @_;
2668
2669 my $value = delete $args{'Value'}; # Can be 0
2670 $value = "" unless defined $value;
2671 $value{$addr} = $value;
2672
2673 $type{$addr} = delete $args{'Type'} || 0;
2674
2675 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2676
2677 if (! $type{$addr}) {
2678 $standard_form{$addr} = main::standardize($value);
2679 }
2680
2681 return $self;
2682 }
2683
2684 use overload
2685 fallback => 0,
2686 qw("") => "_operator_stringify",
2687 "." => \&main::_operator_dot,
2688 ;
2689
2690 sub _operator_stringify {
2691 my $self = shift;
ffe43484 2692 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2693
2694 # Output it like '0041..0065 (value)'
2695 my $return = sprintf("%04X", $start{$addr})
2696 . '..'
2697 . sprintf("%04X", $end{$addr});
2698 my $value = $value{$addr};
2699 my $type = $type{$addr};
2700 $return .= ' (';
2701 $return .= "$value";
2702 $return .= ", Type=$type" if $type != 0;
2703 $return .= ')';
2704
2705 return $return;
2706 }
2707
2708 sub standard_form {
2709 # The standard form is the value itself if the standard form is
2710 # undefined (that is if the value is special)
2711
2712 my $self = shift;
2713 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2714
ffe43484 2715 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2716
2717 return $standard_form{$addr} if defined $standard_form{$addr};
2718 return $value{$addr};
2719 }
2720
2721 sub dump {
2722 # Human, not machine readable. For machine readable, comment out this
2723 # entire routine and let the standard one take effect.
2724 my $self = shift;
2725 my $indent = shift;
2726 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2727
ffe43484 2728 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2729
2730 my $return = $indent
2731 . sprintf("%04X", $start{$addr})
2732 . '..'
2733 . sprintf("%04X", $end{$addr})
2734 . " '$value{$addr}';";
2735 if (! defined $standard_form{$addr}) {
2736 $return .= "(type=$type{$addr})";
2737 }
2738 elsif ($standard_form{$addr} ne $value{$addr}) {
2739 $return .= "(standard '$standard_form{$addr}')";
2740 }
2741 return $return;
2742 }
2743} # End closure
2744
2745package _Range_List_Base;
2746
2747# Base class for range lists. A range list is simply an ordered list of
2748# ranges, so that the ranges with the lowest starting numbers are first in it.
2749#
2750# When a new range is added that is adjacent to an existing range that has the
2751# same value and type, it merges with it to form a larger range.
2752#
2753# Ranges generally do not overlap, except that there can be multiple entries
2754# of single code point ranges. This is because of NameAliases.txt.
2755#
2756# In this program, there is a standard value such that if two different
2757# values, have the same standard value, they are considered equivalent. This
2758# value was chosen so that it gives correct results on Unicode data
2759
2760# There are a number of methods to manipulate range lists, and some operators
2761# are overloaded to handle them.
2762
99870f4d
KW
2763sub trace { return main::trace(@_); }
2764
2765{ # Closure
2766
2767 our $addr;
2768
2769 main::setup_package();
2770
2771 my %ranges;
2772 # The list of ranges
2773 main::set_access('ranges', \%ranges, 'readable_array');
2774
2775 my %max;
2776 # The highest code point in the list. This was originally a method, but
2777 # actual measurements said it was used a lot.
2778 main::set_access('max', \%max, 'r');
2779
2780 my %each_range_iterator;
2781 # Iterator position for each_range()
2782 main::set_access('each_range_iterator', \%each_range_iterator);
2783
2784 my %owner_name_of;
2785 # Name of parent this is attached to, if any. Solely for better error
2786 # messages.
2787 main::set_access('owner_name_of', \%owner_name_of, 'p_r');
2788
2789 my %_search_ranges_cache;
2790 # A cache of the previous result from _search_ranges(), for better
2791 # performance
2792 main::set_access('_search_ranges_cache', \%_search_ranges_cache);
2793
2794 sub new {
2795 my $class = shift;
2796 my %args = @_;
2797
2798 # Optional initialization data for the range list.
2799 my $initialize = delete $args{'Initialize'};
2800
2801 my $self;
2802
2803 # Use _union() to initialize. _union() returns an object of this
2804 # class, which means that it will call this constructor recursively.
2805 # But it won't have this $initialize parameter so that it won't
2806 # infinitely loop on this.
2807 return _union($class, $initialize, %args) if defined $initialize;
2808
2809 $self = bless \do { my $anonymous_scalar }, $class;
ffe43484 2810 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2811
2812 # Optional parent object, only for debug info.
2813 $owner_name_of{$addr} = delete $args{'Owner'};
2814 $owner_name_of{$addr} = "" if ! defined $owner_name_of{$addr};
2815
2816 # Stringify, in case it is an object.
2817 $owner_name_of{$addr} = "$owner_name_of{$addr}";
2818
2819 # This is used only for error messages, and so a colon is added
2820 $owner_name_of{$addr} .= ": " if $owner_name_of{$addr} ne "";
2821
2822 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
2823
2824 # Max is initialized to a negative value that isn't adjacent to 0,
2825 # for simpler tests
2826 $max{$addr} = -2;
2827
2828 $_search_ranges_cache{$addr} = 0;
2829 $ranges{$addr} = [];
2830
2831 return $self;
2832 }
2833
2834 use overload
2835 fallback => 0,
2836 qw("") => "_operator_stringify",
2837 "." => \&main::_operator_dot,
2838 ;
2839
2840 sub _operator_stringify {
2841 my $self = shift;
ffe43484 2842 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2843
2844 return "Range_List attached to '$owner_name_of{$addr}'"
2845 if $owner_name_of{$addr};
2846 return "anonymous Range_List " . \$self;
2847 }
2848
2849 sub _union {
2850 # Returns the union of the input code points. It can be called as
2851 # either a constructor or a method. If called as a method, the result
2852 # will be a new() instance of the calling object, containing the union
2853 # of that object with the other parameter's code points; if called as
2854 # a constructor, the first parameter gives the class the new object
2855 # should be, and the second parameter gives the code points to go into
2856 # it.
2857 # In either case, there are two parameters looked at by this routine;
2858 # any additional parameters are passed to the new() constructor.
2859 #
2860 # The code points can come in the form of some object that contains
2861 # ranges, and has a conventionally named method to access them; or
2862 # they can be an array of individual code points (as integers); or
2863 # just a single code point.
2864 #
2865 # If they are ranges, this routine doesn't make any effort to preserve
2866 # the range values of one input over the other. Therefore this base
2867 # class should not allow _union to be called from other than
2868 # initialization code, so as to prevent two tables from being added
2869 # together where the range values matter. The general form of this
2870 # routine therefore belongs in a derived class, but it was moved here
2871 # to avoid duplication of code. The failure to overload this in this
2872 # class keeps it safe.
2873 #
2874
2875 my $self;
2876 my @args; # Arguments to pass to the constructor
2877
2878 my $class = shift;
2879
2880 # If a method call, will start the union with the object itself, and
2881 # the class of the new object will be the same as self.
2882 if (ref $class) {
2883 $self = $class;
2884 $class = ref $self;
2885 push @args, $self;
2886 }
2887
2888 # Add the other required parameter.
2889 push @args, shift;
2890 # Rest of parameters are passed on to the constructor
2891
2892 # Accumulate all records from both lists.
2893 my @records;
2894 for my $arg (@args) {
2895 #local $to_trace = 0 if main::DEBUG;
2896 trace "argument = $arg" if main::DEBUG && $to_trace;
2897 if (! defined $arg) {
2898 my $message = "";
2899 if (defined $self) {
f998e60c 2900 no overloading;
051df77b 2901 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
2902 }
2903 Carp::my_carp_bug($message .= "Undefined argument to _union. No union done.");
2904 return;
2905 }
2906 $arg = [ $arg ] if ! ref $arg;
2907 my $type = ref $arg;
2908 if ($type eq 'ARRAY') {
2909 foreach my $element (@$arg) {
2910 push @records, Range->new($element, $element);
2911 }
2912 }
2913 elsif ($arg->isa('Range')) {
2914 push @records, $arg;
2915 }
2916 elsif ($arg->can('ranges')) {
2917 push @records, $arg->ranges;
2918 }
2919 else {
2920 my $message = "";
2921 if (defined $self) {
f998e60c 2922 no overloading;
051df77b 2923 $message .= $owner_name_of{pack 'J', $self};
99870f4d
KW
2924 }
2925 Carp::my_carp_bug($message . "Cannot take the union of a $type. No union done.");
2926 return;
2927 }
2928 }
2929
2930 # Sort with the range containing the lowest ordinal first, but if
2931 # two ranges start at the same code point, sort with the bigger range
2932 # of the two first, because it takes fewer cycles.
2933 @records = sort { ($a->start <=> $b->start)
2934 or
2935 # if b is shorter than a, b->end will be
2936 # less than a->end, and we want to select
2937 # a, so want to return -1
2938 ($b->end <=> $a->end)
2939 } @records;
2940
2941 my $new = $class->new(@_);
2942
2943 # Fold in records so long as they add new information.
2944 for my $set (@records) {
2945 my $start = $set->start;
2946 my $end = $set->end;
2947 my $value = $set->value;
2948 if ($start > $new->max) {
2949 $new->_add_delete('+', $start, $end, $value);
2950 }
2951 elsif ($end > $new->max) {
2952 $new->_add_delete('+', $new->max +1, $end, $value);
2953 }
2954 }
2955
2956 return $new;
2957 }
2958
2959 sub range_count { # Return the number of ranges in the range list
2960 my $self = shift;
2961 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2962
f998e60c 2963 no overloading;
051df77b 2964 return scalar @{$ranges{pack 'J', $self}};
99870f4d
KW
2965 }
2966
2967 sub min {
2968 # Returns the minimum code point currently in the range list, or if
2969 # the range list is empty, 2 beyond the max possible. This is a
2970 # method because used so rarely, that not worth saving between calls,
2971 # and having to worry about changing it as ranges are added and
2972 # deleted.
2973
2974 my $self = shift;
2975 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2976
ffe43484 2977 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
2978
2979 # If the range list is empty, return a large value that isn't adjacent
2980 # to any that could be in the range list, for simpler tests
2981 return $LAST_UNICODE_CODEPOINT + 2 unless scalar @{$ranges{$addr}};
2982 return $ranges{$addr}->[0]->start;
2983 }
2984
2985 sub contains {
2986 # Boolean: Is argument in the range list? If so returns $i such that:
2987 # range[$i]->end < $codepoint <= range[$i+1]->end
2988 # which is one beyond what you want; this is so that the 0th range
2989 # doesn't return false
2990 my $self = shift;
2991 my $codepoint = shift;
2992 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
2993
99870f4d
KW
2994 my $i = $self->_search_ranges($codepoint);
2995 return 0 unless defined $i;
2996
2997 # The search returns $i, such that
2998 # range[$i-1]->end < $codepoint <= range[$i]->end
2999 # So is in the table if and only iff it is at least the start position
3000 # of range $i.
f998e60c 3001 no overloading;
051df77b 3002 return 0 if $ranges{pack 'J', $self}->[$i]->start > $codepoint;
99870f4d
KW
3003 return $i + 1;
3004 }
3005
2f7a8815
KW
3006 sub containing_range {
3007 # Returns the range object that contains the code point, undef if none
3008
3009 my $self = shift;
3010 my $codepoint = shift;
3011 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3012
3013 my $i = $self->contains($codepoint);
3014 return unless $i;
3015
3016 # contains() returns 1 beyond where we should look
3017 no overloading;
3018 return $ranges{pack 'J', $self}->[$i-1];
3019 }
3020
99870f4d
KW
3021 sub value_of {
3022 # Returns the value associated with the code point, undef if none
3023
3024 my $self = shift;
3025 my $codepoint = shift;
3026 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3027
d69c231b
KW
3028 my $range = $self->containing_range($codepoint);
3029 return unless defined $range;
99870f4d 3030
d69c231b 3031 return $range->value;
99870f4d
KW
3032 }
3033
0a9dbafc
KW
3034 sub type_of {
3035 # Returns the type of the range containing the code point, undef if
3036 # the code point is not in the table
3037
3038 my $self = shift;
3039 my $codepoint = shift;
3040 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3041
3042 my $range = $self->containing_range($codepoint);
3043 return unless defined $range;
3044
3045 return $range->type;
3046 }
3047
99870f4d
KW
3048 sub _search_ranges {
3049 # Find the range in the list which contains a code point, or where it
3050 # should go if were to add it. That is, it returns $i, such that:
3051 # range[$i-1]->end < $codepoint <= range[$i]->end
3052 # Returns undef if no such $i is possible (e.g. at end of table), or
3053 # if there is an error.
3054
3055 my $self = shift;
3056 my $code_point = shift;
3057 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3058
ffe43484 3059 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3060
3061 return if $code_point > $max{$addr};
3062 my $r = $ranges{$addr}; # The current list of ranges
3063 my $range_list_size = scalar @$r;
3064 my $i;
3065
3066 use integer; # want integer division
3067
3068 # Use the cached result as the starting guess for this one, because,
3069 # an experiment on 5.1 showed that 90% of the time the cache was the
3070 # same as the result on the next call (and 7% it was one less).
3071 $i = $_search_ranges_cache{$addr};
3072 $i = 0 if $i >= $range_list_size; # Reset if no longer valid (prob.
3073 # from an intervening deletion
3074 #local $to_trace = 1 if main::DEBUG;
3075 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);
3076 return $i if $code_point <= $r->[$i]->end
3077 && ($i == 0 || $r->[$i-1]->end < $code_point);
3078
3079 # Here the cache doesn't yield the correct $i. Try adding 1.
3080 if ($i < $range_list_size - 1
3081 && $r->[$i]->end < $code_point &&
3082 $code_point <= $r->[$i+1]->end)
3083 {
3084 $i++;
3085 trace "next \$i is correct: $i" if main::DEBUG && $to_trace;
3086 $_search_ranges_cache{$addr} = $i;
3087 return $i;
3088 }
3089
3090 # Here, adding 1 also didn't work. We do a binary search to
3091 # find the correct position, starting with current $i
3092 my $lower = 0;
3093 my $upper = $range_list_size - 1;
3094 while (1) {
3095 trace "top of loop i=$i:", sprintf("%04X", $r->[$lower]->start), "[$lower] .. ", sprintf("%04X", $r->[$i]->start), "[$i] .. ", sprintf("%04X", $r->[$upper]->start), "[$upper]" if main::DEBUG && $to_trace;
3096
3097 if ($code_point <= $r->[$i]->end) {
3098
3099 # Here we have met the upper constraint. We can quit if we
3100 # also meet the lower one.
3101 last if $i == 0 || $r->[$i-1]->end < $code_point;
3102
3103 $upper = $i; # Still too high.
3104
3105 }
3106 else {
3107
3108 # Here, $r[$i]->end < $code_point, so look higher up.
3109 $lower = $i;
3110 }
3111
3112 # Split search domain in half to try again.
3113 my $temp = ($upper + $lower) / 2;
3114
3115 # No point in continuing unless $i changes for next time
3116 # in the loop.
3117 if ($temp == $i) {
3118
3119 # We can't reach the highest element because of the averaging.
3120 # So if one below the upper edge, force it there and try one
3121 # more time.
3122 if ($i == $range_list_size - 2) {
3123
3124 trace "Forcing to upper edge" if main::DEBUG && $to_trace;
3125 $i = $range_list_size - 1;
3126
3127 # Change $lower as well so if fails next time through,
3128 # taking the average will yield the same $i, and we will
3129 # quit with the error message just below.
3130 $lower = $i;
3131 next;
3132 }
3133 Carp::my_carp_bug("$owner_name_of{$addr}Can't find where the range ought to go. No action taken.");
3134 return;
3135 }
3136 $i = $temp;
3137 } # End of while loop
3138
3139 if (main::DEBUG && $to_trace) {
3140 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i;
3141 trace "i= [ $i ]", $r->[$i];
3142 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < $range_list_size - 1;
3143 }
3144
3145 # Here we have found the offset. Cache it as a starting point for the
3146 # next call.
3147 $_search_ranges_cache{$addr} = $i;
3148 return $i;
3149 }
3150
3151 sub _add_delete {
3152 # Add, replace or delete ranges to or from a list. The $type
3153 # parameter gives which:
3154 # '+' => insert or replace a range, returning a list of any changed
3155 # ranges.
3156 # '-' => delete a range, returning a list of any deleted ranges.
3157 #
3158 # The next three parameters give respectively the start, end, and
3159 # value associated with the range. 'value' should be null unless the
3160 # operation is '+';
3161 #
3162 # The range list is kept sorted so that the range with the lowest
3163 # starting position is first in the list, and generally, adjacent
c1739a4a 3164 # ranges with the same values are merged into a single larger one (see
99870f4d
KW
3165 # exceptions below).
3166 #
c1739a4a 3167 # There are more parameters; all are key => value pairs:
99870f4d
KW
3168 # Type gives the type of the value. It is only valid for '+'.
3169 # All ranges have types; if this parameter is omitted, 0 is
3170 # assumed. Ranges with type 0 are assumed to obey the
3171 # Unicode rules for casing, etc; ranges with other types are
3172 # not. Otherwise, the type is arbitrary, for the caller's
3173 # convenience, and looked at only by this routine to keep
3174 # adjacent ranges of different types from being merged into
3175 # a single larger range, and when Replace =>
3176 # $IF_NOT_EQUIVALENT is specified (see just below).
3177 # Replace determines what to do if the range list already contains
3178 # ranges which coincide with all or portions of the input
3179 # range. It is only valid for '+':
3180 # => $NO means that the new value is not to replace
3181 # any existing ones, but any empty gaps of the
3182 # range list coinciding with the input range
3183 # will be filled in with the new value.
3184 # => $UNCONDITIONALLY means to replace the existing values with
3185 # this one unconditionally. However, if the
3186 # new and old values are identical, the
3187 # replacement is skipped to save cycles
3188 # => $IF_NOT_EQUIVALENT means to replace the existing values
3189 # with this one if they are not equivalent.
3190 # Ranges are equivalent if their types are the
c1739a4a 3191 # same, and they are the same string; or if
99870f4d
KW
3192 # both are type 0 ranges, if their Unicode
3193 # standard forms are identical. In this last
3194 # case, the routine chooses the more "modern"
3195 # one to use. This is because some of the
3196 # older files are formatted with values that
3197 # are, for example, ALL CAPs, whereas the
3198 # derived files have a more modern style,
3199 # which looks better. By looking for this
3200 # style when the pre-existing and replacement
3201 # standard forms are the same, we can move to
3202 # the modern style
3203 # => $MULTIPLE means that if this range duplicates an
3204 # existing one, but has a different value,
3205 # don't replace the existing one, but insert
3206 # this, one so that the same range can occur
3207 # multiple times.
3208 # => anything else is the same as => $IF_NOT_EQUIVALENT
3209 #
c1739a4a
KW
3210 # "same value" means identical for non-type-0 ranges, and it means
3211 # having the same standard forms for type-0 ranges.
99870f4d
KW
3212
3213 return Carp::carp_too_few_args(\@_, 5) if main::DEBUG && @_ < 5;
3214
3215 my $self = shift;
3216 my $operation = shift; # '+' for add/replace; '-' for delete;
3217 my $start = shift;
3218 my $end = shift;
3219 my $value = shift;
3220
3221 my %args = @_;
3222
3223 $value = "" if not defined $value; # warning: $value can be "0"
3224
3225 my $replace = delete $args{'Replace'};
3226 $replace = $IF_NOT_EQUIVALENT unless defined $replace;
3227
3228 my $type = delete $args{'Type'};
3229 $type = 0 unless defined $type;
3230
3231 Carp::carp_extra_args(\%args) if main::DEBUG && %args;
3232
ffe43484 3233 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3234
3235 if ($operation ne '+' && $operation ne '-') {
3236 Carp::my_carp_bug("$owner_name_of{$addr}First parameter to _add_delete must be '+' or '-'. No action taken.");
3237 return;
3238 }
3239 unless (defined $start && defined $end) {
3240 Carp::my_carp_bug("$owner_name_of{$addr}Undefined start and/or end to _add_delete. No action taken.");
3241 return;
3242 }
3243 unless ($end >= $start) {
3244 Carp::my_carp_bug("$owner_name_of{$addr}End of range (" . sprintf("%04X", $end) . ") must not be before start (" . sprintf("%04X", $start) . "). No action taken.");
3245 return;
3246 }
3247 #local $to_trace = 1 if main::DEBUG;
3248
3249 if ($operation eq '-') {
3250 if ($replace != $IF_NOT_EQUIVALENT) {
3251 Carp::my_carp_bug("$owner_name_of{$addr}Replace => \$IF_NOT_EQUIVALENT is required when deleting a range from a range list. Assuming Replace => \$IF_NOT_EQUIVALENT.");
3252 $replace = $IF_NOT_EQUIVALENT;
3253 }
3254 if ($type) {
3255 Carp::my_carp_bug("$owner_name_of{$addr}Type => 0 is required when deleting a range from a range list. Assuming Type => 0.");
3256 $type = 0;
3257 }
3258 if ($value ne "") {
3259 Carp::my_carp_bug("$owner_name_of{$addr}Value => \"\" is required when deleting a range from a range list. Assuming Value => \"\".");
3260 $value = "";
3261 }
3262 }
3263
3264 my $r = $ranges{$addr}; # The current list of ranges
3265 my $range_list_size = scalar @$r; # And its size
3266 my $max = $max{$addr}; # The current high code point in
3267 # the list of ranges
3268
3269 # Do a special case requiring fewer machine cycles when the new range
3270 # starts after the current highest point. The Unicode input data is
3271 # structured so this is common.
3272 if ($start > $max) {
3273
3274 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) type=$type" if main::DEBUG && $to_trace;
3275 return if $operation eq '-'; # Deleting a non-existing range is a
3276 # no-op
3277
3278 # If the new range doesn't logically extend the current final one
3279 # in the range list, create a new range at the end of the range
3280 # list. (max cleverly is initialized to a negative number not
3281 # adjacent to 0 if the range list is empty, so even adding a range
3282 # to an empty range list starting at 0 will have this 'if'
3283 # succeed.)
3284 if ($start > $max + 1 # non-adjacent means can't extend.
3285 || @{$r}[-1]->value ne $value # values differ, can't extend.
3286 || @{$r}[-1]->type != $type # types differ, can't extend.
3287 ) {
3288 push @$r, Range->new($start, $end,
3289 Value => $value,
3290 Type => $type);
3291 }
3292 else {
3293
3294 # Here, the new range starts just after the current highest in
3295 # the range list, and they have the same type and value.
3296 # Extend the current range to incorporate the new one.
3297 @{$r}[-1]->set_end($end);
3298 }
3299
3300 # This becomes the new maximum.
3301 $max{$addr} = $end;
3302
3303 return;
3304 }
3305 #local $to_trace = 0 if main::DEBUG;
3306
3307 trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) replace=$replace" if main::DEBUG && $to_trace;
3308
3309 # Here, the input range isn't after the whole rest of the range list.
3310 # Most likely 'splice' will be needed. The rest of the routine finds
3311 # the needed splice parameters, and if necessary, does the splice.
3312 # First, find the offset parameter needed by the splice function for
3313 # the input range. Note that the input range may span multiple
3314 # existing ones, but we'll worry about that later. For now, just find
3315 # the beginning. If the input range is to be inserted starting in a
3316 # position not currently in the range list, it must (obviously) come
3317 # just after the range below it, and just before the range above it.
3318 # Slightly less obviously, it will occupy the position currently
3319 # occupied by the range that is to come after it. More formally, we
3320 # are looking for the position, $i, in the array of ranges, such that:
3321 #
3322 # r[$i-1]->start <= r[$i-1]->end < $start < r[$i]->start <= r[$i]->end
3323 #
3324 # (The ordered relationships within existing ranges are also shown in
3325 # the equation above). However, if the start of the input range is
3326 # within an existing range, the splice offset should point to that
3327 # existing range's position in the list; that is $i satisfies a
3328 # somewhat different equation, namely:
3329 #
3330 #r[$i-1]->start <= r[$i-1]->end < r[$i]->start <= $start <= r[$i]->end
3331 #
3332 # More briefly, $start can come before or after r[$i]->start, and at
3333 # this point, we don't know which it will be. However, these
3334 # two equations share these constraints:
3335 #
3336 # r[$i-1]->end < $start <= r[$i]->end
3337 #
3338 # And that is good enough to find $i.
3339
3340 my $i = $self->_search_ranges($start);
3341 if (! defined $i) {
3342 Carp::my_carp_bug("Searching $self for range beginning with $start unexpectedly returned undefined. Operation '$operation' not performed");
3343 return;
3344 }
3345
3346 # The search function returns $i such that:
3347 #
3348 # r[$i-1]->end < $start <= r[$i]->end
3349 #
3350 # That means that $i points to the first range in the range list
3351 # that could possibly be affected by this operation. We still don't
3352 # know if the start of the input range is within r[$i], or if it
3353 # points to empty space between r[$i-1] and r[$i].
3354 trace "[$i] is the beginning splice point. Existing range there is ", $r->[$i] if main::DEBUG && $to_trace;
3355
3356 # Special case the insertion of data that is not to replace any
3357 # existing data.
3358 if ($replace == $NO) { # If $NO, has to be operation '+'
3359 #local $to_trace = 1 if main::DEBUG;
3360 trace "Doesn't replace" if main::DEBUG && $to_trace;
3361
3362 # Here, the new range is to take effect only on those code points
3363 # that aren't already in an existing range. This can be done by
3364 # looking through the existing range list and finding the gaps in
3365 # the ranges that this new range affects, and then calling this
3366 # function recursively on each of those gaps, leaving untouched
3367 # anything already in the list. Gather up a list of the changed
3368 # gaps first so that changes to the internal state as new ranges
3369 # are added won't be a problem.
3370 my @gap_list;
3371
3372 # First, if the starting point of the input range is outside an
3373 # existing one, there is a gap from there to the beginning of the
3374 # existing range -- add a span to fill the part that this new
3375 # range occupies
3376 if ($start < $r->[$i]->start) {
3377 push @gap_list, Range->new($start,
3378 main::min($end,
3379 $r->[$i]->start - 1),
3380 Type => $type);
3381 trace "gap before $r->[$i] [$i], will add", $gap_list[-1] if main::DEBUG && $to_trace;
3382 }
3383
3384 # Then look through the range list for other gaps until we reach
3385 # the highest range affected by the input one.
3386 my $j;
3387 for ($j = $i+1; $j < $range_list_size; $j++) {
3388 trace "j=[$j]", $r->[$j] if main::DEBUG && $to_trace;
3389 last if $end < $r->[$j]->start;
3390
3391 # If there is a gap between when this range starts and the
3392 # previous one ends, add a span to fill it. Note that just
3393 # because there are two ranges doesn't mean there is a
3394 # non-zero gap between them. It could be that they have
3395 # different values or types
3396 if ($r->[$j-1]->end + 1 != $r->[$j]->start) {
3397 push @gap_list,
3398 Range->new($r->[$j-1]->end + 1,
3399 $r->[$j]->start - 1,
3400 Type => $type);
3401 trace "gap between $r->[$j-1] and $r->[$j] [$j], will add: $gap_list[-1]" if main::DEBUG && $to_trace;
3402 }
3403 }
3404
3405 # Here, we have either found an existing range in the range list,
3406 # beyond the area affected by the input one, or we fell off the
3407 # end of the loop because the input range affects the whole rest
3408 # of the range list. In either case, $j is 1 higher than the
3409 # highest affected range. If $j == $i, it means that there are no
3410 # affected ranges, that the entire insertion is in the gap between
3411 # r[$i-1], and r[$i], which we already have taken care of before
3412 # the loop.
3413 # On the other hand, if there are affected ranges, it might be
3414 # that there is a gap that needs filling after the final such
3415 # range to the end of the input range
3416 if ($r->[$j-1]->end < $end) {
3417 push @gap_list, Range->new(main::max($start,
3418 $r->[$j-1]->end + 1),
3419 $end,
3420 Type => $type);
3421 trace "gap after $r->[$j-1], will add $gap_list[-1]" if main::DEBUG && $to_trace;
3422 }
3423
3424 # Call recursively to fill in all the gaps.
3425 foreach my $gap (@gap_list) {
3426 $self->_add_delete($operation,
3427 $gap->start,
3428 $gap->end,
3429 $value,
3430 Type => $type);
3431 }
3432
3433 return;
3434 }
3435
3436 # Here, we have taken care of the case where $replace is $NO, which
3437 # means that whatever action we now take is done unconditionally. It
3438 # still could be that this call will result in a no-op, if duplicates
3439 # aren't allowed, and we are inserting a range that merely duplicates
3440 # data already in the range list; or also if deleting a non-existent
3441 # range.
3442 # $i still points to the first potential affected range. Now find the
3443 # highest range affected, which will determine the length parameter to
3444 # splice. (The input range can span multiple existing ones.) While
969a34cc
KW
3445 # we are looking through the range list, see also if this is a "clean
3446 # insertion" that doesn't change the values in any existing range. It
3447 # isn't a clean insertion if it is a deletion, and we know it is a
3448 # clean insertion if we are adding duplicates (because by definition
3449 # they don't affect existing ranges), so we don't need to do the check
3450 # for either of those cases. But otherwise, start off assuming it is
3451 # a clean insertion until proven otherwise.
3452 my $clean_insert = ($operation eq '+' && $replace != $MULTIPLE);
99870f4d
KW
3453 my $j; # This will point to the highest affected range
3454
3455 # For non-zero types, the standard form is the value itself;
3456 my $standard_form = ($type) ? $value : main::standardize($value);
3457
3458 for ($j = $i; $j < $range_list_size; $j++) {
3459 trace "Looking for highest affected range; the one at $j is ", $r->[$j] if main::DEBUG && $to_trace;
3460
3461 # If find a range that it doesn't overlap into, we can stop
3462 # searching
3463 last if $end < $r->[$j]->start;
3464
969a34cc
KW
3465 # Here, overlaps the range at $j. If the values don't match,
3466 # and so far we think this is a clean insertion, it becomes a
3467 # non-clean insertion, i.e., a 'change' or 'replace' instead.
3468 if ($clean_insert) {
99870f4d 3469 if ($r->[$j]->standard_form ne $standard_form) {
969a34cc 3470 $clean_insert = 0;
99870f4d
KW
3471 }
3472 else {
3473
3474 # Here, the two values are essentially the same. If the
3475 # two are actually identical, replacing wouldn't change
3476 # anything so skip it.
3477 my $pre_existing = $r->[$j]->value;
3478 if ($pre_existing ne $value) {
3479
3480 # Here the new and old standardized values are the
3481 # same, but the non-standardized values aren't. If
3482 # replacing unconditionally, then replace
3483 if( $replace == $UNCONDITIONALLY) {
969a34cc 3484 $clean_insert = 0;
99870f4d
KW
3485 }
3486 else {
3487
3488 # Here, are replacing conditionally. Decide to
3489 # replace or not based on which appears to look
3490 # the "nicest". If one is mixed case and the
3491 # other isn't, choose the mixed case one.
3492 my $new_mixed = $value =~ /[A-Z]/
3493 && $value =~ /[a-z]/;
3494 my $old_mixed = $pre_existing =~ /[A-Z]/
3495 && $pre_existing =~ /[a-z]/;
3496
3497 if ($old_mixed != $new_mixed) {
969a34cc 3498 $clean_insert = 0 if $new_mixed;
99870f4d 3499 if (main::DEBUG && $to_trace) {
969a34cc
KW
3500 if ($clean_insert) {
3501 trace "Retaining $pre_existing over $value";
99870f4d
KW
3502 }
3503 else {
969a34cc 3504 trace "Replacing $pre_existing with $value";
99870f4d
KW
3505 }
3506 }
3507 }
3508 else {
3509
3510 # Here casing wasn't different between the two.
3511 # If one has hyphens or underscores and the
3512 # other doesn't, choose the one with the
3513 # punctuation.
3514 my $new_punct = $value =~ /[-_]/;
3515 my $old_punct = $pre_existing =~ /[-_]/;
3516
3517 if ($old_punct != $new_punct) {
969a34cc 3518 $clean_insert = 0 if $new_punct;
99870f4d 3519 if (main::DEBUG && $to_trace) {
969a34cc
KW
3520 if ($clean_insert) {
3521 trace "Retaining $pre_existing over $value";
99870f4d
KW
3522 }
3523 else {
969a34cc 3524 trace "Replacing $pre_existing with $value";
99870f4d
KW
3525 }
3526 }
3527 } # else existing one is just as "good";
3528 # retain it to save cycles.
3529 }
3530 }
3531 }
3532 }
3533 }
3534 } # End of loop looking for highest affected range.
3535
3536 # Here, $j points to one beyond the highest range that this insertion
3537 # affects (hence to beyond the range list if that range is the final
3538 # one in the range list).
3539
3540 # The splice length is all the affected ranges. Get it before
3541 # subtracting, for efficiency, so we don't have to later add 1.
3542 my $length = $j - $i;
3543
3544 $j--; # $j now points to the highest affected range.
3545 trace "Final affected range is $j: $r->[$j]" if main::DEBUG && $to_trace;
3546
3547 # If inserting a multiple record, this is where it goes, after all the
3548 # existing ones for this range. This implies an insertion, and no
3549 # change to any existing ranges. Note that $j can be -1 if this new
3550 # range doesn't actually duplicate any existing, and comes at the
3551 # beginning of the list, in which case we can handle it like any other
3552 # insertion, and is easier to do so.
3553 if ($replace == $MULTIPLE && $j >= 0) {
3554
3555 # This restriction could be remedied with a little extra work, but
3556 # it won't hopefully ever be necessary
3557 if ($r->[$j]->start != $r->[$j]->end) {
3558 Carp::my_carp_bug("$owner_name_of{$addr}Can't cope with adding a multiple when the other range ($r->[$j]) contains more than one code point. No action taken.");
3559 return;
3560 }
3561
3562 # Don't add an exact duplicate, as it isn't really a multiple
3563 return if $value eq $r->[$j]->value && $type eq $r->[$j]->type;
3564
3565 trace "Adding multiple record at $j+1 with $start..$end, $value" if main::DEBUG && $to_trace;
3566 my @return = splice @$r,
3567 $j+1,
3568 0,
3569 Range->new($start,
3570 $end,
3571 Value => $value,
3572 Type => $type);
3573 if (main::DEBUG && $to_trace) {
3574 trace "After splice:";
3575 trace 'j-2=[', $j-2, ']', $r->[$j-2] if $j >= 2;
3576 trace 'j-1=[', $j-1, ']', $r->[$j-1] if $j >= 1;
3577 trace "j =[", $j, "]", $r->[$j] if $j >= 0;
3578 trace 'j+1=[', $j+1, ']', $r->[$j+1] if $j < @$r - 1;
3579 trace 'j+2=[', $j+2, ']', $r->[$j+2] if $j < @$r - 2;
3580 trace 'j+3=[', $j+3, ']', $r->[$j+3] if $j < @$r - 3;
3581 }
3582 return @return;
3583 }
3584
3585 # Here, have taken care of $NO and $MULTIPLE replaces.
3586 # $j points to the highest affected range. But it can be < $i or even
3587 # -1. These happen only if the insertion is entirely in the gap
3588 # between r[$i-1] and r[$i]. Here's why: j < i means that the j loop
3589 # above exited first time through with $end < $r->[$i]->start. (And
3590 # then we subtracted one from j) This implies also that $start <
3591 # $r->[$i]->start, but we know from above that $r->[$i-1]->end <
3592 # $start, so the entire input range is in the gap.
3593 if ($j < $i) {
3594
3595 # Here the entire input range is in the gap before $i.
3596
3597 if (main::DEBUG && $to_trace) {
3598 if ($i) {
3599 trace "Entire range is between $r->[$i-1] and $r->[$i]";
3600 }
3601 else {
3602 trace "Entire range is before $r->[$i]";
3603 }
3604 }
3605 return if $operation ne '+'; # Deletion of a non-existent range is
3606 # a no-op
3607 }
3608 else {
3609
969a34cc
KW
3610 # Here part of the input range is not in the gap before $i. Thus,
3611 # there is at least one affected one, and $j points to the highest
3612 # such one.
99870f4d
KW
3613
3614 # At this point, here is the situation:
3615 # This is not an insertion of a multiple, nor of tentative ($NO)
3616 # data.
3617 # $i points to the first element in the current range list that
3618 # may be affected by this operation. In fact, we know
3619 # that the range at $i is affected because we are in
3620 # the else branch of this 'if'
3621 # $j points to the highest affected range.
3622 # In other words,
3623 # r[$i-1]->end < $start <= r[$i]->end
3624 # And:
3625 # r[$i-1]->end < $start <= $end <= r[$j]->end
3626 #
3627 # Also:
969a34cc
KW
3628 # $clean_insert is a boolean which is set true if and only if
3629 # this is a "clean insertion", i.e., not a change nor a
3630 # deletion (multiple was handled above).
99870f4d
KW
3631
3632 # We now have enough information to decide if this call is a no-op
969a34cc
KW
3633 # or not. It is a no-op if this is an insertion of already
3634 # existing data.
99870f4d 3635
969a34cc 3636 if (main::DEBUG && $to_trace && $clean_insert
99870f4d
KW
3637 && $i == $j
3638 && $start >= $r->[$i]->start)
3639 {
3640 trace "no-op";
3641 }
969a34cc 3642 return if $clean_insert
99870f4d
KW
3643 && $i == $j # more than one affected range => not no-op
3644
3645 # Here, r[$i-1]->end < $start <= $end <= r[$i]->end
3646 # Further, $start and/or $end is >= r[$i]->start
3647 # The test below hence guarantees that
3648 # r[$i]->start < $start <= $end <= r[$i]->end
3649 # This means the input range is contained entirely in
3650 # the one at $i, so is a no-op
3651 && $start >= $r->[$i]->start;
3652 }
3653
3654 # Here, we know that some action will have to be taken. We have
3655 # calculated the offset and length (though adjustments may be needed)
3656 # for the splice. Now start constructing the replacement list.
3657 my @replacement;
3658 my $splice_start = $i;
3659
3660 my $extends_below;
3661 my $extends_above;
3662
3663 # See if should extend any adjacent ranges.
3664 if ($operation eq '-') { # Don't extend deletions
3665 $extends_below = $extends_above = 0;
3666 }
3667 else { # Here, should extend any adjacent ranges. See if there are
3668 # any.
3669 $extends_below = ($i > 0
3670 # can't extend unless adjacent
3671 && $r->[$i-1]->end == $start -1
3672 # can't extend unless are same standard value
3673 && $r->[$i-1]->standard_form eq $standard_form
3674 # can't extend unless share type
3675 && $r->[$i-1]->type == $type);
3676 $extends_above = ($j+1 < $range_list_size
3677 && $r->[$j+1]->start == $end +1
3678 && $r->[$j+1]->standard_form eq $standard_form
3679 && $r->[$j-1]->type == $type);
3680 }
3681 if ($extends_below && $extends_above) { # Adds to both
3682 $splice_start--; # start replace at element below
3683 $length += 2; # will replace on both sides
3684 trace "Extends both below and above ranges" if main::DEBUG && $to_trace;
3685
3686 # The result will fill in any gap, replacing both sides, and
3687 # create one large range.
3688 @replacement = Range->new($r->[$i-1]->start,
3689 $r->[$j+1]->end,
3690 Value => $value,
3691 Type => $type);
3692 }
3693 else {
3694
3695 # Here we know that the result won't just be the conglomeration of
3696 # a new range with both its adjacent neighbors. But it could
3697 # extend one of them.
3698
3699 if ($extends_below) {
3700
3701 # Here the new element adds to the one below, but not to the
3702 # one above. If inserting, and only to that one range, can
3703 # just change its ending to include the new one.
969a34cc 3704 if ($length == 0 && $clean_insert) {
99870f4d
KW
3705 $r->[$i-1]->set_end($end);
3706 trace "inserted range extends range to below so it is now $r->[$i-1]" if main::DEBUG && $to_trace;
3707 return;
3708 }
3709 else {
3710 trace "Changing inserted range to start at ", sprintf("%04X", $r->[$i-1]->start), " instead of ", sprintf("%04X", $start) if main::DEBUG && $to_trace;
3711 $splice_start--; # start replace at element below
3712 $length++; # will replace the element below
3713 $start = $r->[$i-1]->start;
3714 }
3715 }
3716 elsif ($extends_above) {
3717
3718 # Here the new element adds to the one above, but not below.
3719 # Mirror the code above
969a34cc 3720 if ($length == 0 && $clean_insert) {
99870f4d
KW
3721 $r->[$j+1]->set_start($start);
3722 trace "inserted range extends range to above so it is now $r->[$j+1]" if main::DEBUG && $to_trace;
3723 return;
3724 }
3725 else {
3726 trace "Changing inserted range to end at ", sprintf("%04X", $r->[$j+1]->end), " instead of ", sprintf("%04X", $end) if main::DEBUG && $to_trace;
3727 $length++; # will replace the element above
3728 $end = $r->[$j+1]->end;
3729 }
3730 }
3731
3732 trace "Range at $i is $r->[$i]" if main::DEBUG && $to_trace;
3733
3734 # Finally, here we know there will have to be a splice.
3735 # If the change or delete affects only the highest portion of the
3736 # first affected range, the range will have to be split. The
3737 # splice will remove the whole range, but will replace it by a new
3738 # range containing just the unaffected part. So, in this case,
3739 # add to the replacement list just this unaffected portion.
3740 if (! $extends_below
3741 && $start > $r->[$i]->start && $start <= $r->[$i]->end)
3742 {
3743 push @replacement,
3744 Range->new($r->[$i]->start,
3745 $start - 1,
3746 Value => $r->[$i]->value,
3747 Type => $r->[$i]->type);
3748 }
3749
3750 # In the case of an insert or change, but not a delete, we have to
3751 # put in the new stuff; this comes next.
3752 if ($operation eq '+') {
3753 push @replacement, Range->new($start,
3754 $end,
3755 Value => $value,
3756 Type => $type);
3757 }
3758
3759 trace "Range at $j is $r->[$j]" if main::DEBUG && $to_trace && $j != $i;
3760 #trace "$end >=", $r->[$j]->start, " && $end <", $r->[$j]->end if main::DEBUG && $to_trace;
3761
3762 # And finally, if we're changing or deleting only a portion of the
3763 # highest affected range, it must be split, as the lowest one was.
3764 if (! $extends_above
3765 && $j >= 0 # Remember that j can be -1 if before first
3766 # current element
3767 && $end >= $r->[$j]->start
3768 && $end < $r->[$j]->end)
3769 {
3770 push @replacement,
3771 Range->new($end + 1,
3772 $r->[$j]->end,
3773 Value => $r->[$j]->value,
3774 Type => $r->[$j]->type);
3775 }
3776 }
3777
3778 # And do the splice, as calculated above
3779 if (main::DEBUG && $to_trace) {
3780 trace "replacing $length element(s) at $i with ";
3781 foreach my $replacement (@replacement) {
3782 trace " $replacement";
3783 }
3784 trace "Before splice:";
3785 trace 'i-2=[', $i-2, ']', $r->[$i-2] if $i >= 2;
3786 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i >= 1;
3787 trace "i =[", $i, "]", $r->[$i];
3788 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < @$r - 1;
3789 trace 'i+2=[', $i+2, ']', $r->[$i+2] if $i < @$r - 2;
3790 }
3791
3792 my @return = splice @$r, $splice_start, $length, @replacement;
3793
3794 if (main::DEBUG && $to_trace) {
3795 trace "After splice:";
3796 trace 'i-2=[', $i-2, ']', $r->[$i-2] if $i >= 2;
3797 trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i >= 1;
3798 trace "i =[", $i, "]", $r->[$i];
3799 trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < @$r - 1;
3800 trace 'i+2=[', $i+2, ']', $r->[$i+2] if $i < @$r - 2;
3801 trace "removed @return";
3802 }
3803
3804 # An actual deletion could have changed the maximum in the list.
3805 # There was no deletion if the splice didn't return something, but
3806 # otherwise recalculate it. This is done too rarely to worry about
3807 # performance.
3808 if ($operation eq '-' && @return) {
3809 $max{$addr} = $r->[-1]->end;
3810 }
3811 return @return;
3812 }
3813
3814 sub reset_each_range { # reset the iterator for each_range();
3815 my $self = shift;
3816 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3817
f998e60c 3818 no overloading;
051df77b 3819 undef $each_range_iterator{pack 'J', $self};
99870f4d
KW
3820 return;
3821 }
3822
3823 sub each_range {
3824 # Iterate over each range in a range list. Results are undefined if
3825 # the range list is changed during the iteration.
3826
3827 my $self = shift;
3828 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3829
ffe43484 3830 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3831
3832 return if $self->is_empty;
3833
3834 $each_range_iterator{$addr} = -1
3835 if ! defined $each_range_iterator{$addr};
3836 $each_range_iterator{$addr}++;
3837 return $ranges{$addr}->[$each_range_iterator{$addr}]
3838 if $each_range_iterator{$addr} < @{$ranges{$addr}};
3839 undef $each_range_iterator{$addr};
3840 return;
3841 }
3842
3843 sub count { # Returns count of code points in range list
3844 my $self = shift;
3845 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3846
ffe43484 3847 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3848
3849 my $count = 0;
3850 foreach my $range (@{$ranges{$addr}}) {
3851 $count += $range->end - $range->start + 1;
3852 }
3853 return $count;
3854 }
3855
3856 sub delete_range { # Delete a range
3857 my $self = shift;
3858 my $start = shift;
3859 my $end = shift;
3860
3861 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3862
3863 return $self->_add_delete('-', $start, $end, "");
3864 }
3865
3866 sub is_empty { # Returns boolean as to if a range list is empty
3867 my $self = shift;
3868 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3869
f998e60c 3870 no overloading;
051df77b 3871 return scalar @{$ranges{pack 'J', $self}} == 0;
99870f4d
KW
3872 }
3873
3874 sub hash {
3875 # Quickly returns a scalar suitable for separating tables into
3876 # buckets, i.e. it is a hash function of the contents of a table, so
3877 # there are relatively few conflicts.
3878
3879 my $self = shift;
3880 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3881
ffe43484 3882 my $addr = do { no overloading; pack 'J', $self; };
99870f4d
KW
3883
3884 # These are quickly computable. Return looks like 'min..max;count'
3885 return $self->min . "..$max{$addr};" . scalar @{$ranges{$addr}};
3886 }
3887} # End closure for _Range_List_Base
3888
3889package Range_List;
3890use base '_Range_List_Base';
3891
3892# A Range_List is a range list for match tables; i.e. the range values are
3893# not significant. Thus a number of operations can be safely added to it,
3894# such as inversion, intersection. Note that union is also an unsafe
3895# operation when range values are cared about, and that method is in the base
3896# class, not here. But things are set up so that that method is callable only
3897# during initialization. Only in this derived class, is there an operation
3898# that combines two tables. A Range_Map can thus be used to initialize a
3899# Range_List, and its mappings will be in the list, but are not significant to
3900# this class.
3901
3902sub trace { return main::trace(@_); }
3903
3904{ # Closure
3905
3906 use overload
3907 fallback => 0,
3908 '+' => sub { my $self = shift;
3909 my $other = shift;
3910
3911 return $self->_union($other)
3912 },
3913 '&' => sub { my $self = shift;
3914 my $other = shift;
3915
3916 return $self->_intersect($other, 0);
3917 },
3918 '~' => "_invert",
3919 '-' => "_subtract",
3920 ;
3921
3922 sub _invert {
3923 # Returns a new Range_List that gives all code points not in $self.
3924
3925 my $self = shift;
3926
3927 my $new = Range_List->new;
3928
3929 # Go through each range in the table, finding the gaps between them
3930 my $max = -1; # Set so no gap before range beginning at 0
3931 for my $range ($self->ranges) {
3932 my $start = $range->start;
3933 my $end = $range->end;
3934
3935 # If there is a gap before this range, the inverse will contain
3936 # that gap.
3937 if ($start > $max + 1) {
3938 $new->add_range($max + 1, $start - 1);
3939 }
3940 $max = $end;
3941 }
3942
3943 # And finally, add the gap from the end of the table to the max
3944 # possible code point
3945 if ($max < $LAST_UNICODE_CODEPOINT) {
3946 $new->add_range($max + 1, $LAST_UNICODE_CODEPOINT);
3947 }
3948 return $new;
3949 }
3950
3951 sub _subtract {
3952 # Returns a new Range_List with the argument deleted from it. The
3953 # argument can be a single code point, a range, or something that has
3954 # a range, with the _range_list() method on it returning them
3955
3956 my $self = shift;
3957 my $other = shift;
3958 my $reversed = shift;
3959 Carp::carp_extra_args(\@_) if main::DEBUG && @_;
3960
3961 if ($reversed) {
3962 Carp::my_carp_bug("Can't cope with a "
3963 . __PACKAGE__
3964 . " being the second parameter in a '-'. Subtraction ignored.");
3965 return $self;
3966 }
3967
3968 my $new = Range_List->new(Initialize => $self);
3969
3970 if (! ref $other) { # Single code point
3971 $new->delete_range($other, $other);
3972 }
3973 elsif ($other->isa('Range')) {
3974 $new->delete_range($other->start, $other->end);
3975 }
3976 elsif ($other->can('_range_list')) {
3977 foreach my $range ($other->_range_list->ranges) {
3978 $new->delete_range($range->start, $range->end);
3979 }
3980 }
3981 else {
3982 Carp::my_carp_bug("Can't cope with a "