src/vendor/unicode-segmentation/scripts/unicode.py

   1 #!/usr/bin/env python
   2 #
   3 # Copyright 2011-2015 The Rust Project Developers. See the COPYRIGHT
   4 # file at the top-level directory of this distribution and at
   5 # http://rust-lang.org/COPYRIGHT.
   6 #
   7 # Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
   8 # http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
   9 # <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
  10 # option. This file may not be copied, modified, or distributed
  11 # except according to those terms.
  12
  13 # This script uses the following Unicode tables:
  14 # - DerivedCoreProperties.txt
  15 # - auxiliary/GraphemeBreakProperty.txt
  16 # - auxiliary/WordBreakProperty.txt
  17 # - ReadMe.txt
  18 # - UnicodeData.txt
  19 #
  20 # Since this should not require frequent updates, we just store this
  21 # out-of-line and check the unicode.rs file into git.
  22
  23 import fileinput, re, os, sys, operator
  24
  25 preamble = '''// Copyright 2012-2015 The Rust Project Developers. See the COPYRIGHT
  26 // file at the top-level directory of this distribution and at
  27 // http://rust-lang.org/COPYRIGHT.
  28 //
  29 // Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
  30 // http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
  31 // <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
  32 // option. This file may not be copied, modified, or distributed
  33 // except according to those terms.
  34
  35 // NOTE: The following code was generated by "scripts/unicode.py", do not edit directly
  36
  37 #![allow(missing_docs, non_upper_case_globals, non_snake_case)]
  38 '''
  39
  40 # Mapping taken from Table 12 from:
  41 # http://www.unicode.org/reports/tr44/#General_Category_Values
  42 expanded_categories = {
  43     'Lu': ['LC', 'L'], 'Ll': ['LC', 'L'], 'Lt': ['LC', 'L'],
  44     'Lm': ['L'], 'Lo': ['L'],
  45     'Mn': ['M'], 'Mc': ['M'], 'Me': ['M'],
  46     'Nd': ['N'], 'Nl': ['N'], 'No': ['No'],
  47     'Pc': ['P'], 'Pd': ['P'], 'Ps': ['P'], 'Pe': ['P'],
  48     'Pi': ['P'], 'Pf': ['P'], 'Po': ['P'],
  49     'Sm': ['S'], 'Sc': ['S'], 'Sk': ['S'], 'So': ['S'],
  50     'Zs': ['Z'], 'Zl': ['Z'], 'Zp': ['Z'],
  51     'Cc': ['C'], 'Cf': ['C'], 'Cs': ['C'], 'Co': ['C'], 'Cn': ['C'],
  52 }
  53
  54 # these are the surrogate codepoints, which are not valid rust characters
  55 surrogate_codepoints = (0xd800, 0xdfff)
  56
  57 def is_surrogate(n):
  58     return surrogate_codepoints[0] <= n <= surrogate_codepoints[1]
  59
  60 def fetch(f):
  61     if not os.path.exists(os.path.basename(f)):
  62         os.system("curl -O http://www.unicode.org/Public/UNIDATA/%s"
  63                   % f)
  64
  65     if not os.path.exists(os.path.basename(f)):
  66         sys.stderr.write("cannot load %s" % f)
  67         exit(1)
  68
  69 def load_gencats(f):
  70     fetch(f)
  71     gencats = {}
  72
  73     udict = {};
  74     range_start = -1;
  75     for line in fileinput.input(f):
  76         data = line.split(';');
  77         if len(data) != 15:
  78             continue
  79         cp = int(data[0], 16);
  80         if is_surrogate(cp):
  81             continue
  82         if range_start >= 0:
  83             for i in xrange(range_start, cp):
  84                 udict[i] = data;
  85             range_start = -1;
  86         if data[1].endswith(", First>"):
  87             range_start = cp;
  88             continue;
  89         udict[cp] = data;
  90
  91     for code in udict:
  92         [code_org, name, gencat, combine, bidi,
  93          decomp, deci, digit, num, mirror,
  94          old, iso, upcase, lowcase, titlecase ] = udict[code];
  95
  96         # place letter in categories as appropriate
  97         for cat in [gencat, "Assigned"] + expanded_categories.get(gencat, []):
  98             if cat not in gencats:
  99                 gencats[cat] = []
 100             gencats[cat].append(code)
 101
 102     gencats = group_cats(gencats)
 103     return gencats
 104
 105 def group_cats(cats):
 106     cats_out = {}
 107     for cat in cats:
 108         cats_out[cat] = group_cat(cats[cat])
 109     return cats_out
 110
 111 def group_cat(cat):
 112     cat_out = []
 113     letters = sorted(set(cat))
 114     cur_start = letters.pop(0)
 115     cur_end = cur_start
 116     for letter in letters:
 117         assert letter > cur_end, \
 118             "cur_end: %s, letter: %s" % (hex(cur_end), hex(letter))
 119         if letter == cur_end + 1:
 120             cur_end = letter
 121         else:
 122             cat_out.append((cur_start, cur_end))
 123             cur_start = cur_end = letter
 124     cat_out.append((cur_start, cur_end))
 125     return cat_out
 126
 127 def ungroup_cat(cat):
 128     cat_out = []
 129     for (lo, hi) in cat:
 130         while lo <= hi:
 131             cat_out.append(lo)
 132             lo += 1
 133     return cat_out
 134
 135 def format_table_content(f, content, indent):
 136     line = " "*indent
 137     first = True
 138     for chunk in content.split(","):
 139         if len(line) + len(chunk) < 98:
 140             if first:
 141                 line += chunk
 142             else:
 143                 line += ", " + chunk
 144             first = False
 145         else:
 146             f.write(line + ",\n")
 147             line = " "*indent + chunk
 148     f.write(line)
 149
 150 def load_properties(f, interestingprops):
 151     fetch(f)
 152     props = {}
 153     re1 = re.compile("^ *([0-9A-F]+) *; *(\w+)")
 154     re2 = re.compile("^ *([0-9A-F]+)\.\.([0-9A-F]+) *; *(\w+)")
 155
 156     for line in fileinput.input(os.path.basename(f)):
 157         prop = None
 158         d_lo = 0
 159         d_hi = 0
 160         m = re1.match(line)
 161         if m:
 162             d_lo = m.group(1)
 163             d_hi = m.group(1)
 164             prop = m.group(2)
 165         else:
 166             m = re2.match(line)
 167             if m:
 168                 d_lo = m.group(1)
 169                 d_hi = m.group(2)
 170                 prop = m.group(3)
 171             else:
 172                 continue
 173         if interestingprops and prop not in interestingprops:
 174             continue
 175         d_lo = int(d_lo, 16)
 176         d_hi = int(d_hi, 16)
 177         if prop not in props:
 178             props[prop] = []
 179         props[prop].append((d_lo, d_hi))
 180
 181     # optimize if possible
 182     for prop in props:
 183         props[prop] = group_cat(ungroup_cat(props[prop]))
 184
 185     return props
 186
 187 def escape_char(c):
 188     return "'\\u{%x}'" % c
 189
 190 def emit_table(f, name, t_data, t_type = "&'static [(char, char)]", is_pub=True,
 191         pfun=lambda x: "(%s,%s)" % (escape_char(x[0]), escape_char(x[1])), is_const=True):
 192     pub_string = "const"
 193     if not is_const:
 194         pub_string = "let"
 195     if is_pub:
 196         pub_string = "pub " + pub_string
 197     f.write("    %s %s: %s = &[\n" % (pub_string, name, t_type))
 198     data = ""
 199     first = True
 200     for dat in t_data:
 201         if not first:
 202             data += ","
 203         first = False
 204         data += pfun(dat)
 205     format_table_content(f, data, 8)
 206     f.write("\n    ];\n\n")
 207
 208 def emit_util_mod(f):
 209     f.write("""
 210 pub mod util {
 211     #[inline]
 212     pub fn bsearch_range_table(c: char, r: &'static [(char,char)]) -> bool {
 213         use core::cmp::Ordering::{Equal, Less, Greater};
 214         r.binary_search_by(|&(lo,hi)| {
 215             if lo <= c && c <= hi { Equal }
 216             else if hi < c { Less }
 217             else { Greater }
 218         }).is_ok()
 219     }
 220
 221     #[inline]
 222     fn is_alphabetic(c: char) -> bool {
 223         match c {
 224             'a' ... 'z' | 'A' ... 'Z' => true,
 225             c if c > '\x7f' => super::derived_property::Alphabetic(c),
 226             _ => false,
 227         }
 228     }
 229
 230     #[inline]
 231     fn is_numeric(c: char) -> bool {
 232         match c {
 233             '0' ... '9' => true,
 234             c if c > '\x7f' => super::general_category::N(c),
 235             _ => false,
 236         }
 237     }
 238
 239     #[inline]
 240     pub fn is_alphanumeric(c: char) -> bool {
 241         is_alphabetic(c) || is_numeric(c)
 242     }
 243 }
 244
 245 """)
 246
 247 def emit_property_module(f, mod, tbl, emit):
 248     f.write("mod %s {\n" % mod)
 249     for cat in sorted(emit):
 250         emit_table(f, "%s_table" % cat, tbl[cat], is_pub=False)
 251         f.write("    #[inline]\n")
 252         f.write("    pub fn %s(c: char) -> bool {\n" % cat)
 253         f.write("        super::util::bsearch_range_table(c, %s_table)\n" % cat)
 254         f.write("    }\n\n")
 255     f.write("}\n\n")
 256
 257 def emit_break_module(f, break_table, break_cats, name):
 258     Name = name.capitalize()
 259     f.write("""pub mod %s {
 260     use core::result::Result::{Ok, Err};
 261
 262     pub use self::%sCat::*;
 263
 264     #[allow(non_camel_case_types)]
 265     #[derive(Clone, Copy, PartialEq, Eq)]
 266     pub enum %sCat {
 267 """ % (name, Name, Name))
 268
 269     break_cats.append("Any")
 270     break_cats.sort()
 271     for cat in break_cats:
 272         f.write(("        %sC_" % Name[0]) + cat + ",\n")
 273     f.write("""    }
 274
 275     fn bsearch_range_value_table(c: char, r: &'static [(char, char, %sCat)]) -> %sCat {
 276         use core::cmp::Ordering::{Equal, Less, Greater};
 277         match r.binary_search_by(|&(lo, hi, _)| {
 278             if lo <= c && c <= hi { Equal }
 279             else if hi < c { Less }
 280             else { Greater }
 281         }) {
 282             Ok(idx) => {
 283                 let (_, _, cat) = r[idx];
 284                 cat
 285             }
 286             Err(_) => %sC_Any
 287         }
 288     }
 289
 290     pub fn %s_category(c: char) -> %sCat {
 291         bsearch_range_value_table(c, %s_cat_table)
 292     }
 293
 294 """ % (Name, Name, Name[0], name, Name, name))
 295
 296     emit_table(f, "%s_cat_table" % name, break_table, "&'static [(char, char, %sCat)]" % Name,
 297         pfun=lambda x: "(%s,%s,%sC_%s)" % (escape_char(x[0]), escape_char(x[1]), Name[0], x[2]),
 298         is_pub=False, is_const=True)
 299     f.write("}\n")
 300
 301 if __name__ == "__main__":
 302     r = "tables.rs"
 303     if os.path.exists(r):
 304         os.remove(r)
 305     with open(r, "w") as rf:
 306         # write the file's preamble
 307         rf.write(preamble)
 308
 309         # download and parse all the data
 310         fetch("ReadMe.txt")
 311         with open("ReadMe.txt") as readme:
 312             pattern = "for Version (\d+)\.(\d+)\.(\d+) of the Unicode"
 313             unicode_version = re.search(pattern, readme.read()).groups()
 314         rf.write("""
 315 /// The version of [Unicode](http://www.unicode.org/)
 316 /// that this version of unicode-segmentation is based on.
 317 pub const UNICODE_VERSION: (u64, u64, u64) = (%s, %s, %s);
 318 """ % unicode_version)
 319
 320         gencats = load_gencats("UnicodeData.txt")
 321         derived = load_properties("DerivedCoreProperties.txt", ["Alphabetic"])
 322
 323         emit_util_mod(rf)
 324         for (name, cat, pfuns) in ("general_category", gencats, ["N"]), \
 325                                   ("derived_property", derived, ["Alphabetic"]):
 326             emit_property_module(rf, name, cat, pfuns)
 327
 328         ### grapheme cluster module
 329         # from http://www.unicode.org/reports/tr29/#Grapheme_Cluster_Break_Property_Values
 330         grapheme_cats = load_properties("auxiliary/GraphemeBreakProperty.txt", [])
 331
 332         # Control
 333         #  Note 1:
 334         # This category also includes Cs (surrogate codepoints), but Rust's `char`s are
 335         # Unicode Scalar Values only, and surrogates are thus invalid `char`s.
 336         # Thus, we have to remove Cs from the Control category
 337         #  Note 2:
 338         # 0x0a and 0x0d (CR and LF) are not in the Control category for Graphemes.
 339         # However, the Graphemes iterator treats these as a special case, so they
 340         # should be included in grapheme_cats["Control"] for our implementation.
 341         grapheme_cats["Control"] = group_cat(list(
 342             (set(ungroup_cat(grapheme_cats["Control"]))
 343              | set(ungroup_cat(grapheme_cats["CR"]))
 344              | set(ungroup_cat(grapheme_cats["LF"])))
 345             - set(ungroup_cat([surrogate_codepoints]))))
 346         del(grapheme_cats["CR"])
 347         del(grapheme_cats["LF"])
 348
 349         grapheme_table = []
 350         for cat in grapheme_cats:
 351             grapheme_table.extend([(x, y, cat) for (x, y) in grapheme_cats[cat]])
 352         grapheme_table.sort(key=lambda w: w[0])
 353         emit_break_module(rf, grapheme_table, grapheme_cats.keys(), "grapheme")
 354         rf.write("\n")
 355
 356         word_cats = load_properties("auxiliary/WordBreakProperty.txt", [])
 357         word_table = []
 358         for cat in word_cats:
 359             word_table.extend([(x, y, cat) for (x, y) in word_cats[cat]])
 360         word_table.sort(key=lambda w: w[0])
 361         emit_break_module(rf, word_table, word_cats.keys(), "word")