summaryrefslogtreecommitdiff
path: root/src
diff options
context:
space:
mode:
authorBenoit Giannangeli <giann008@gmail.com>2017-03-28 13:18:13 +0200
committerBenoit Giannangeli <giann008@gmail.com>2017-03-28 13:53:19 +0200
commit36f3247d47c1ad854fa89aabf17f6d954a6a6657 (patch)
tree20d48d5d73aa81f86998c33d439b1d45bf1ea510 /src
parentca46ad67f398f219cf838feb10127cd0cecc72d8 (diff)
downloadfengari-36f3247d47c1ad854fa89aabf17f6d954a6a6657.tar.gz
fengari-36f3247d47c1ad854fa89aabf17f6d954a6a6657.tar.bz2
fengari-36f3247d47c1ad854fa89aabf17f6d954a6a6657.zip
luaO_utf8esc
Diffstat (limited to 'src')
-rw-r--r--src/lauxlib.js2
-rw-r--r--src/llex.js12
-rw-r--r--src/lobject.js22
3 files changed, 29 insertions, 7 deletions
diff --git a/src/lauxlib.js b/src/lauxlib.js
index e3974cb..8bda842 100644
--- a/src/lauxlib.js
+++ b/src/lauxlib.js
@@ -437,7 +437,7 @@ if (typeof require === "function") {
lf.pos += bytes;
}
if (bytes > 0)
- return lf.binary ? toDataView(lf.buff) : new lobject.TValue(0, lf.buff).jsstring(0, bytes); // TODO: Here reading utf8 only
+ return lf.binary ? toDataView(lf.buff) : lobject.jsstring(lf.buff, 0, bytes); // TODO: Here reading utf8 only
else return null;
};
diff --git a/src/llex.js b/src/llex.js
index 96e9e00..2d34b1f 100644
--- a/src/llex.js
+++ b/src/llex.js
@@ -152,7 +152,7 @@ const save = function(ls, c) {
if (b.buffer.length >= Number.MAX_SAFE_INTEGER/2)
lexerror(ls, "lexical element too long", 0);
}
- b.buffer[b.n++] = c;
+ b.buffer[b.n++] = c < 0 ? 255 + c + 1 : c;
};
const luaX_token2str = function(ls, token) {
@@ -407,9 +407,9 @@ const readutf8desc = function(ls) {
};
const utf8esc = function(ls) {
- let buff = new Array(lobject.UTF8BUFFSZ);
- let n = lobject.luaO_utf8esc(buff, readutf8desc(ls));
- for (; n > 0; n--) /* add 'buff' to string */
+ let u = lobject.luaO_utf8esc(readutf8desc(ls));
+ let buff = u.buff;
+ for (let n = u.n; n > 0; n--) /* add 'buff' to string */
save(ls, buff[lobject.UTF8BUFFSZ - n]);
};
@@ -450,12 +450,12 @@ const read_string = function(ls, del, seminfo) {
case 't': c = '\t'; will = 'read_save'; break;
case 'v': c = '\v'; will = 'read_save'; break;
case 'x': c = readhexaesc(ls); will = 'read_save'; break;
- case 'u': utf8esc(ls); will = 'read_save'; break;
+ case 'u': utf8esc(ls); will = 'no_save'; break;
case '\n': case '\r':
inclinenumber(ls); c = '\n'; will = 'only_save'; break;
case '\\': case '\"': case '\'':
c = ls.current; will = 'read_save'; break;
- case -1: will = 'read_save'; break; /* will raise an error next loop */
+ case -1: will = 'no_save'; break; /* will raise an error next loop */
case 'z': { /* zap following span of spaces */
ls.buff.n -= 1; /* remove '\\' */
next(ls); /* skip the 'z' */
diff --git a/src/lobject.js b/src/lobject.js
index fca259d..966bdd8 100644
--- a/src/lobject.js
+++ b/src/lobject.js
@@ -499,6 +499,27 @@ const luaO_str2num = function(s) {
}
};
+const luaO_utf8esc = function(x) {
+ let buff = [];
+ let n = 1; /* number of bytes put in buffer (backwards) */
+ assert(x <= 0x10ffff);
+ if (x < 0x80) /* ascii? */
+ buff[UTF8BUFFSZ - 1] = x;
+ else { /* need continuation bytes */
+ let mfb = 0x3f; /* maximum that fits in first byte */
+ do { /* add continuation bytes */
+ buff[UTF8BUFFSZ - (n++)] = 0x80 | (x & 0x3f);
+ x >>= 6; /* remove added bits */
+ mfb >>= 1; /* now there is one less bit available in first byte */
+ } while (x > mfb); /* still needs continuation byte? */
+ buff[UTF8BUFFSZ - n] = (~mfb << 1) | x; /* add first byte */
+ }
+ return {
+ buff: buff,
+ n: n
+ };
+};
+
/*
** converts an integer to a "floating point byte", represented as
** (eeeeexxx), where the real value is (1xxx) * 2^(eeeee - 1) if
@@ -564,4 +585,5 @@ module.exports.luaO_hexavalue = luaO_hexavalue;
module.exports.luaO_int2fb = luaO_int2fb;
module.exports.luaO_str2num = luaO_str2num;
module.exports.luaO_utf8desc = luaO_utf8desc;
+module.exports.luaO_utf8esc = luaO_utf8esc;
module.exports.numarith = numarith;