Merge pull request #4 from GaloisInc/T3-fix-unicode-lexing

Lexer: Properly support Unicode 15.1.0

.gitignore

@@ -26,3 +26,7 @@ sample-sources/
 !sample-sources/statement-expressions.rs
 !sample-sources/statements.rs
 !sample-sources/types.rs
+
+# Unicode-related autogenerated files
+DerivedCoreProperties.txt
+UnicodeLexer.x

scripts/unicode.py

@@ -0,0 +1,167 @@
+#!/usr/bin/env python
+#
+# Copyright 2011-2015 The Rust Project Developers
+#           2024 Galois Inc.
+#
+# This script was originally created by the Rust Project Developers as part of
+# the `unicode-xid` crate:
+#
+#   https://github.com/unicode-rs/unicode-xid/blob/b3a2718b062da229c0a50d12281de0e5d8e8cff6/scripts/unicode.py
+#
+# See the COPYRIGHT file in the `unicode-xid` crate:
+#
+#  https://github.com/unicode-rs/unicode-xid/blob/b3a2718b062da229c0a50d12281de0e5d8e8cff6/COPYRIGHT
+#
+# Galois Inc. has modified the script to generate an `alex`-based lexer instead
+# of a Rust-based lexer.
+#
+# Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
+# http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
+# <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
+# option. This file may not be copied, modified, or distributed
+# except according to those terms.
+
+import fileinput, re, os, sys
+
+unicode_version = (15, 1, 0)
+
+preamble = '''-- NOTE: The following code was generated by "scripts/unicode.py", do not edit directly
+--
+-- If you need to update this code, perform the following steps:
+--
+-- 1. (Optional) Update the value of `unicode_version` in "scripts/unicode.py".
+-- 2. Run the "scripts/unicode.py" script.
+-- 3. Copy the code (including the comments) in the autogenerated `UnicodeLexer.x` file.
+-- 4. Replace the existing autogenerated code here.
+'''
+
+postamble = '''-- End of code generated by "scripts/unicode.py".
+'''
+
+def unicode_url(f):
+    return "http://www.unicode.org/Public/%s.%s.%s/ucd/%s" % (unicode_version + (f,))
+
+def fetch(f):
+    if not os.path.exists(os.path.basename(f)):
+        os.system("curl -O %s" % unicode_url(f))
+
+    if not os.path.exists(os.path.basename(f)):
+        sys.stderr.write("cannot load %s" % f)
+        exit(1)
+
+def group_cat(cat):
+    cat_out = []
+    letters = sorted(set(cat))
+    cur_start = letters.pop(0)
+    cur_end = cur_start
+    for letter in letters:
+        assert letter > cur_end, \
+            "cur_end: %s, letter: %s" % (hex(cur_end), hex(letter))
+        if letter == cur_end + 1:
+            cur_end = letter
+        else:
+            cat_out.append((cur_start, cur_end))
+            cur_start = cur_end = letter
+    cat_out.append((cur_start, cur_end))
+    return cat_out
+
+def ungroup_cat(cat):
+    cat_out = []
+    for (lo, hi) in cat:
+        while lo <= hi:
+            cat_out.append(lo)
+            lo += 1
+    return cat_out
+
+def format_table_content(f, content, indent):
+    line = ""
+    first = True
+    for chunk in content.split("|"):
+        line += " " * indent
+        if first:
+            line += "= " + chunk
+        else:
+            line += "| " + chunk
+        line += "\n"
+        first = False
+    f.write(line + '\n')
+
+def load_properties(f, interestingprops):
+    fetch(f)
+    props = {}
+    re1 = re.compile("^ *([0-9A-F]+) *; *(\w+)")
+    re2 = re.compile("^ *([0-9A-F]+)\.\.([0-9A-F]+) *; *(\w+)")
+
+    for line in fileinput.input(os.path.basename(f)):
+        prop = None
+        d_lo = 0
+        d_hi = 0
+        m = re1.match(line)
+        if m:
+            d_lo = m.group(1)
+            d_hi = m.group(1)
+            prop = m.group(2)
+        else:
+            m = re2.match(line)
+            if m:
+                d_lo = m.group(1)
+                d_hi = m.group(2)
+                prop = m.group(3)
+            else:
+                continue
+        if interestingprops and prop not in interestingprops:
+            continue
+        d_lo = int(d_lo, 16)
+        d_hi = int(d_hi, 16)
+        if prop not in props:
+            props[prop] = []
+        props[prop].append((d_lo, d_hi))
+
+    # optimize if possible
+    for prop in props:
+        props[prop] = group_cat(ungroup_cat(props[prop]))
+
+    return props
+
+def escape_char(c):
+    return "\\x%04x" % c
+
+def emit_table(f, name, t_data):
+    f.write("@%s\n" % name)
+    data = ""
+    first = True
+    for dat in t_data:
+        if not first:
+            data += "|"
+        first = False
+        if dat[0] == dat[1]:
+            data += "%s" % escape_char(dat[0])
+        else:
+            data += "[%s-%s]" % (escape_char(dat[0]), escape_char(dat[1]))
+    format_table_content(f, data, 2)
+
+def emit_property_module(f, mod, tbl, emit):
+    for cat in emit:
+        emit_table(f, cat, tbl[cat])
+
+if __name__ == "__main__":
+    r = "UnicodeLexer.x"
+    if os.path.exists(r):
+        os.remove(r)
+    with open(r, "w") as rf:
+        # write the file's preamble
+        rf.write(preamble)
+
+        # download and parse all the data
+        rf.write('''
+-- Based on Unicode %s.%s.%s, using the following Unicode table:
+-- %s
+
+''' % (unicode_version + (unicode_url("DerviedCoreProperties.txt"),)))
+
+        want_derived = ["XID_Start", "XID_Continue"]
+        derived = load_properties("DerivedCoreProperties.txt", want_derived)
+        emit_property_module(rf, "derived_property", derived, want_derived)
+
+        # write the file's postamble
+        rf.write(postamble)