Move non-RegEx lexer hacks from Scanner to Parser.

[cycript.git] / Library.cpp

diff --git a/Library.cpp b/Library.cpp
index 4b7f115aa313ebb53be6d3e4bf15d9d419a41522..720322564e1fb181f6ac78407f6116164ffb8c8e 100644 (file)
--- a/Library.cpp
+++ b/Library.cpp
@@ -1,106 +1,73 @@
 /* Cycript - Optimizing JavaScript Compiler/Runtime
 /* Cycript - Optimizing JavaScript Compiler/Runtime

- * Copyright (C) 2009-2013  Jay Freeman (saurik)
+ * Copyright (C) 2009-2015  Jay Freeman (saurik)

 */
 
 */
 

-/* GNU General Public License, Version 3 {{{ */
+/* GNU Affero General Public License, Version 3 {{{ */

 /*
 /*

- * Cycript is free software: you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published
- * by the Free Software Foundation, either version 3 of the License,
- * or (at your option) any later version.
- *
- * Cycript is distributed in the hope that it will be useful, but
- * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU Affero General Public License as published by
+ * the Free Software Foundation, either version 3 of the License, or
+ * (at your option) any later version.
+
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of

  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the

- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with Cycript.  If not, see <http://www.gnu.org/licenses/>.
+ * GNU Affero General Public License for more details.
+
+ * You should have received a copy of the GNU Affero General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.

 **/
 /* }}} */
 
 **/
 /* }}} */
 

-#include <dlfcn.h>
-#include <iconv.h>
-

 #include "cycript.hpp"
 
 #include "cycript.hpp"
 

-#include "Pooling.hpp"
-
-#include <sys/mman.h>
-

 #include <iostream>
 #include <iostream>

-#include <ext/stdio_filebuf.h>

 #include <set>
 #include <map>
 #include <iomanip>
 #include <sstream>
 #include <cmath>
 
 #include <set>
 #include <map>
 #include <iomanip>
 #include <sstream>
 #include <cmath>
 

-#include "Parser.hpp"
-#include "Cycript.tab.hh"
+#include <dlfcn.h>
+
+#include <sys/mman.h>

 
 

+#include "ConvertUTF.h"
+#include "Driver.hpp"

 #include "Error.hpp"
 #include "Error.hpp"

-#include "String.hpp"

 #include "Execute.hpp"
 #include "Execute.hpp"

-#include "JavaScript.hpp"
-
-/* C Strings {{{ */
-template <typename Type_>
-_finline size_t iconv_(size_t (*iconv)(iconv_t, Type_, size_t *, char **, size_t *), iconv_t cd, char **inbuf, size_t *inbytesleft, char **outbuf, size_t *outbytesleft) {
-    return iconv(cd, const_cast<Type_>(inbuf), inbytesleft, outbuf, outbytesleft);
-}
+#include "Pooling.hpp"
+#include "String.hpp"
+#include "Syntax.hpp"

 
 

-#ifdef __GLIBC__
-#define UCS_2_INTERNAL "UCS-2"
-#else
-#define UCS_2_INTERNAL "UCS-2-INTERNAL"
-#endif
+template <>
+::pthread_key_t CYLocal<CYPool>::key_ = Key_();

 
 

+/* C Strings {{{ */

 CYUTF8String CYPoolUTF8String(CYPool &pool, CYUTF16String utf16) {
 CYUTF8String CYPoolUTF8String(CYPool &pool, CYUTF16String utf16) {

-    _assert(pool != NULL);
-
-    const char *in(reinterpret_cast<const char *>(utf16.data));
-
-    iconv_t conversion(_syscall(iconv_open("UTF-8", UCS_2_INTERNAL)));
-

     // XXX: this is wrong
     size_t size(utf16.size * 5);
     // XXX: this is wrong
     size_t size(utf16.size * 5);

-    char *out(new(pool) char[size]);
-    CYUTF8String utf8(out, size);
-
-    size = utf16.size * 2;
-    _syscall(iconv_(&iconv, conversion, const_cast<char **>(&in), &size, &out, &utf8.size));
+    char *temp(new(pool) char[size]);

 
 

-    *out = '\0';
-    utf8.size = out - utf8.data;
+    const uint16_t *lhs(utf16.data);
+    uint8_t *rhs(reinterpret_cast<uint8_t *>(temp));
+    _assert(ConvertUTF16toUTF8(&lhs, lhs + utf16.size, &rhs, rhs + size, lenientConversion) == conversionOK);

 
 

-    _syscall(iconv_close(conversion));
-
-    return utf8;
+    *rhs = 0;
+    return CYUTF8String(temp, reinterpret_cast<char *>(rhs) - temp);

 }
 
 CYUTF16String CYPoolUTF16String(CYPool &pool, CYUTF8String utf8) {
 }
 
 CYUTF16String CYPoolUTF16String(CYPool &pool, CYUTF8String utf8) {

-    _assert(pool != NULL);
-
-    const char *in(utf8.data);
-
-    iconv_t conversion(_syscall(iconv_open(UCS_2_INTERNAL, "UTF-8")));
-

     // XXX: this is wrong
     size_t size(utf8.size * 5);
     uint16_t *temp(new (pool) uint16_t[size]);
     // XXX: this is wrong
     size_t size(utf8.size * 5);
     uint16_t *temp(new (pool) uint16_t[size]);

-    CYUTF16String utf16(temp, size * 2);
-    char *out(reinterpret_cast<char *>(temp));
-
-    size = utf8.size;
-    _syscall(iconv_(&iconv, conversion, const_cast<char **>(&in), &size, &out, &utf16.size));

 
 

-    utf16.size = reinterpret_cast<uint16_t *>(out) - utf16.data;
-    temp[utf16.size] = 0;
+    const uint8_t *lhs(reinterpret_cast<const uint8_t *>(utf8.data));
+    uint16_t *rhs(temp);
+    _assert(ConvertUTF8toUTF16(&lhs, lhs + utf8.size, &rhs, rhs + size, lenientConversion) == conversionOK);

 
 

-    _syscall(iconv_close(conversion));
-
-    return utf16;
+    *rhs = 0;
+    return CYUTF16String(temp, rhs - temp);

 }
 /* }}} */
 /* Index Offsets {{{ */
 }
 /* }}} */
 /* Index Offsets {{{ */


@@ -149,7 +116,7 @@ void CYStringify(std::ostringstream &str, const char *data, size_t size) {
     str << (single ? '\'' : '"');
 
     for (const char *value(data), *end(data + size); value != end; ++value)
     str << (single ? '\'' : '"');
 
     for (const char *value(data), *end(data + size); value != end; ++value)

-        switch (*value) {
+        switch (uint8_t next = *value) {

             case '\\': str << "\\\\"; break;
             case '\b': str << "\\b"; break;
             case '\f': str << "\\f"; break;
             case '\\': str << "\\\\"; break;
             case '\b': str << "\\b"; break;
             case '\f': str << "\\f"; break;


@@ -170,18 +137,48 @@ void CYStringify(std::ostringstream &str, const char *data, size_t size) {
                 else goto simple;
             break;
 
                 else goto simple;
             break;
 

+            case '\0':
+                if (value[1] >= '0' && value[1] <= '9')
+                    str << "\\x00";
+                else
+                    str << "\\0";
+            break;
+

             default:
             default:

-                // this test is designed to be "awesome", generating neither warnings nor incorrect results
-                if (*value < 0x20 || *value >= 0x7f)
-                    str << "\\x" << std::setbase(16) << std::setw(2) << std::setfill('0') << unsigned(uint8_t(*value));
-                else simple:
+                if (next >= 0x20 && next < 0x7f) simple:

                     str << *value;
                     str << *value;

+                else {
+                    unsigned levels(1);
+                    if ((next & 0x80) != 0)
+                        while ((next & 0x80 >> ++levels) != 0);
+
+                    unsigned point(next & 0xff >> levels);
+                    while (--levels != 0)
+                        point = point << 6 | uint8_t(*++value) & 0x3f;
+
+                    if (point < 0x100)
+                        str << "\\x" << std::setbase(16) << std::setw(2) << std::setfill('0') << point;
+                    else if (point < 0x10000)
+                        str << "\\u" << std::setbase(16) << std::setw(4) << std::setfill('0') << point;
+                    else {
+                        point -= 0x10000;
+                        str << "\\u" << std::setbase(16) << std::setw(4) << std::setfill('0') << (0xd800 | point >> 0x0a);
+                        str << "\\u" << std::setbase(16) << std::setw(4) << std::setfill('0') << (0xdc00 | point & 0x3ff);
+                    }
+                }

         }
 
     str << (single ? '\'' : '"');
 }
 
 void CYNumerify(std::ostringstream &str, double value) {
         }
 
     str << (single ? '\'' : '"');
 }
 
 void CYNumerify(std::ostringstream &str, double value) {

+    if (std::isinf(value)) {
+        if (value < 0)
+            str << '-';
+        str << "Infinity";
+        return;
+    }
+

     char string[32];
     // XXX: I want this to print 1e3 rather than 1000
     sprintf(string, "%.17g", value);
     char string[32];
     // XXX: I want this to print 1e3 rather than 1000
     sprintf(string, "%.17g", value);


@@ -223,73 +220,32 @@ double CYCastDouble(const char *value) {
     return CYCastDouble(value, strlen(value));
 }
 
     return CYCastDouble(value, strlen(value));
 }
 

-size_t CYArrayLength(JSContextRef context, JSObjectRef array) {
-    return CYCastDouble(context, CYGetProperty(context, array, length_s));
-}
-
-JSValueRef CYArrayGet(JSContextRef context, JSObjectRef array, size_t index) {
-    JSValueRef exception(NULL);
-    JSValueRef value(JSObjectGetPropertyAtIndex(context, array, index, &exception));
-    CYThrow(context, exception);
-    return value;
+_visible bool CYStartsWith(const CYUTF8String &haystack, const CYUTF8String &needle) {
+    return haystack.size >= needle.size && strncmp(haystack.data, needle.data, needle.size) == 0;

 }
 
 }
 

-void CYArrayPush(JSContextRef context, JSObjectRef array, JSValueRef value) {
-    JSValueRef exception(NULL);
-    JSValueRef arguments[1];
-    arguments[0] = value;
-    JSObjectRef Array(CYGetCachedObject(context, CYJSString("Array_prototype")));
-    JSObjectCallAsFunction(context, CYCastJSObject(context, CYGetProperty(context, Array, push_s)), array, 1, arguments, &exception);
-    CYThrow(context, exception);
-}
-
-extern "C" void CydgetMemoryParse(const uint16_t **data, size_t *size) {
+CYUTF8String CYPoolCode(CYPool &pool, std::istream &stream) {

     CYLocalPool local;
     CYLocalPool local;

-
-    CYUTF8String utf8(CYPoolUTF8String(local, CYUTF16String(*data, *size)));
-    CYStream stream(utf8.data, utf8.data + utf8.size);
-    CYDriver driver(stream);
-
-    cy::parser parser(driver);
-    if (parser.parse() != 0 || !driver.errors_.empty())
-        return;
+    CYDriver driver(local, stream);
+    _assert(!driver.Parse());
+    _assert(driver.errors_.empty());

 
     CYOptions options;
     CYContext context(options);
 
     CYOptions options;
     CYContext context(options);

-    driver.program_->Replace(context);
-    std::ostringstream str;
-    CYOutput out(str, options);
-    out << *driver.program_;
-    std::string code(str.str());
-
-    CYPool pool;
-    CYUTF16String utf16(CYPoolUTF16String(pool, CYUTF8String(code.c_str(), code.size())));
-
-    size_t bytes(utf16.size * sizeof(uint16_t));
-    uint16_t *copy(reinterpret_cast<uint16_t *>(malloc(bytes)));
-    memcpy(copy, utf16.data, bytes);
-
-    *data = copy;
-    *size = utf16.size;
-}
-
-static bool initialized_;
-
-void CYInitializeStatic() {
-    if (!initialized_)
-        initialized_ = true;
-    else return;
+    driver.script_->Replace(context);

 
 

-    _aprcall(apr_initialize());
+    std::stringbuf str;
+    CYOutput out(str, options);
+    out << *driver.script_;
+    return $pool.strdup(str.str().c_str());

 }
 
 CYPool &CYGetGlobalPool() {
 }
 
 CYPool &CYGetGlobalPool() {

-    CYInitializeStatic();

     static CYPool pool;
     return pool;
 }
 
     static CYPool pool;
     return pool;
 }
 

-void CYThrow(const char *format, ...) {
+_visible void CYThrow(const char *format, ...) {

     va_list args;
     va_start(args, format);
     throw CYPoolError(format, args);
     va_list args;
     va_start(args, format);
     throw CYPoolError(format, args);


@@ -301,13 +257,20 @@ const char *CYPoolError::PoolCString(CYPool &pool) const {
     return pool.strdup(message_);
 }
 
     return pool.strdup(message_);
 }
 

+CYPoolError::CYPoolError(const CYPoolError &rhs) :
+    message_(pool_.strdup(rhs.message_))
+{
+}
+

 CYPoolError::CYPoolError(const char *format, ...) {
     va_list args;
     va_start(args, format);
 CYPoolError::CYPoolError(const char *format, ...) {
     va_list args;
     va_start(args, format);

-    message_ = pool_.vsprintf(format, args);
+    // XXX: there might be a beter way to think about this
+    message_ = pool_.vsprintf(64, format, args);

     va_end(args);
 }
 
 CYPoolError::CYPoolError(const char *format, va_list args) {
     va_end(args);
 }
 
 CYPoolError::CYPoolError(const char *format, va_list args) {

-    message_ = pool_.vsprintf(format, args);
+    // XXX: there might be a beter way to think about this
+    message_ = pool_.vsprintf(64, format, args);

 }
 }