ox 1.5.8 → 1.5.9

data/README.md

@@ -4,6 +4,10 @@ A fast XML parser and Object marshaller as a Ruby gem.
 ## <a name="installation">Installation</a>
     gem install ox
 
+## <a name="documentation">Documentation</a>
+
+*Documentation*: http://www.ohler.com/ox
+
 ## <a name="source">Source</a>
 
 *GitHub* *repo*: https://github.com/ohler55/ox
@@ -12,7 +16,7 @@ A fast XML parser and Object marshaller as a Ruby gem.
 
 ## <a name="follow">Follow @oxgem on Twitter</a>
 
-[Follow @oxgem on Twitter](http://twitter.com/#!/oxgem) for announcements and news about the Ox gem.
+[Follow @peterohler on Twitter](http://twitter.com/#!/peterohler) for announcements and news about the Ox gem.
 
 ## <a name="build_status">Build Status</a>
 
@@ -30,9 +34,13 @@ A fast XML parser and Object marshaller as a Ruby gem.
 
 ## <a name="release">Release Notes</a>
 
-### Release 1.5.8
+### Release 1.5.9
+
+ - Merged a fix by edanaher that allows special characters in attribute values
+
+ - Changed all comments to old C style to get code to compile with Ruby 2.0.0.
 
- - Added an option to not symbolize attribute keys in generic parsing.
+ - Other changes to time handling for 2.0.0.
 
 ## <a name="description">Description</a>
 

data/ext/ox/base64.c

@@ -35,8 +35,8 @@
 
 static char	digits[] = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
 
-// invalid or terminating characters are set to 'X' or \x58
-static u_char   s_digits[256] = "\
+/* invalid or terminating characters are set to 'X' or \x58 */
+static uchar	s_digits[256] = "\
 \x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\
 \x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\
 \x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x3E\x58\x58\x58\x3F\
@@ -55,20 +55,20 @@ static u_char   s_digits[256] = "\
 \x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58\x58";
 
 void
-to_base64(const u_char *src, int len, char *b64) {
-    const u_char	*end3;
-    int			len3 = len % 3;
-    u_char		b1, b2, b3;
+to_base64(const uchar *src, int len, char *b64) {
+    const uchar	*end3;
+    int		len3 = len % 3;
+    uchar	b1, b2, b3;
     
     end3 = src + (len - len3);
     while (src < end3) {
 	b1 = *src++;
 	b2 = *src++;
 	b3 = *src++;
-	*b64++ = digits[(u_char)(b1 >> 2)];
-	*b64++ = digits[(u_char)(((b1 & 0x03) << 4) | (b2 >> 4))];
-	*b64++ = digits[(u_char)(((b2 & 0x0F) << 2) | (b3 >> 6))];
-	*b64++ = digits[(u_char)(b3 & 0x3F)];
+	*b64++ = digits[(uchar)(b1 >> 2)];
+	*b64++ = digits[(uchar)(((b1 & 0x03) << 4) | (b2 >> 4))];
+	*b64++ = digits[(uchar)(((b2 & 0x0F) << 2) | (b3 >> 6))];
+	*b64++ = digits[(uchar)(b3 & 0x3F)];
     }
     if (1 == len3) {
 	b1 = *src++;
@@ -108,16 +108,16 @@ b64_orig_size(const char *text) {
 }
 
 void
-from_base64(const char *b64, u_char *str) {
-    u_char      b0, b1, b2, b3;
+from_base64(const char *b64, uchar *str) {
+    uchar	b0, b1, b2, b3;
     
     while (1) {
-        if ('X' == (b0 = s_digits[(u_char)*b64++])) { break; }
-        if ('X' == (b1 = s_digits[(u_char)*b64++])) { break; }
+        if ('X' == (b0 = s_digits[(uchar)*b64++])) { break; }
+        if ('X' == (b1 = s_digits[(uchar)*b64++])) { break; }
         *str++ = (b0 << 2) | ((b1 >> 4) & 0x03);
-        if ('X' == (b2 = s_digits[(u_char)*b64++])) { break; }
+        if ('X' == (b2 = s_digits[(uchar)*b64++])) { break; }
         *str++ = (b1 << 4) | ((b2 >> 2) & 0x0F);
-        if ('X' == (b3 = s_digits[(u_char)*b64++])) { break; }
+        if ('X' == (b3 = s_digits[(uchar)*b64++])) { break; }
         *str++ = (b2 << 6) | b3;
     }
     *str = '\0';

data/ext/ox/base64.h

@@ -31,14 +31,13 @@
 #ifndef __BASE64_H__
 #define __BASE64_H__
 
-#include <sys/types.h>
-
+typedef unsigned char	uchar;
 
 #define b64_size(len) ((len + 2) / 3 * 4)
 
 extern unsigned long    b64_orig_size(const char *text);
 
-extern void	        to_base64(const u_char *src, int len, char *b64);
-extern void             from_base64(const char *b64, u_char *str);
+extern void	        to_base64(const uchar *src, int len, char *b64);
+extern void             from_base64(const char *b64, uchar *str);
 
 #endif /* __BASE64_H__ */

data/ext/ox/cache.c

@@ -32,18 +32,28 @@
 #include <errno.h>
 #include <stdio.h>
 #include <string.h>
+#include <strings.h>
 #include <stdarg.h>
 
 #include "cache.h"
 
 struct _Cache {
-    char                *key; // only set if the node has a value, and it is not an exact match
+    char                *key; /* only set if the node has a value, and it is not an exact match */
     VALUE               value;
     struct _Cache       *slots[16];
 };
 
 static void     slot_print(Cache cache, unsigned int depth);
 
+static char* dupstr(const char *s) {
+    size_t	len = strlen(s);
+    char	*d = ALLOC_N(char, len + 1);
+
+    memcpy(d, s, len);
+
+    return d;
+}
+
 void
 ox_cache_new(Cache *cache) {
     *cache = ALLOC(struct _Cache);
@@ -58,16 +68,16 @@ ox_cache_get(Cache cache, const char *key, VALUE **slot) {
     Cache               *cp;
 
     for (; '\0' != *k; k++) {
-        cp = cache->slots + (unsigned int)(*k >> 4); // upper 4 bits
+        cp = cache->slots + (unsigned int)(*k >> 4); /* upper 4 bits */
         if (0 == *cp) {
             ox_cache_new(cp);
         }
         cache = *cp;
-        cp = cache->slots + (unsigned int)(*k & 0x0F); // lower 4 bits
+        cp = cache->slots + (unsigned int)(*k & 0x0F); /* lower 4 bits */
         if (0 == *cp) {
             ox_cache_new(cp);
             cache = *cp;
-            cache->key = ('\0' == *(k + 1)) ? 0 : strdup(key);
+            cache->key = ('\0' == *(k + 1)) ? 0 : dupstr(key);
             break;
         } else {
             cache = *cp;
@@ -78,8 +88,8 @@ ox_cache_get(Cache cache, const char *key, VALUE **slot) {
                     break;
                 } else {
                     Cache     *cp2 = cp;
-
-                    // if value was set along with the key then there are no slots filled yet
+		    
+                    /* if value was set along with the key then there are no slots filled yet */
                     cp2 = (*cp2)->slots + (*ck >> 4);
                     ox_cache_new(cp2);
                     cp2 = (*cp2)->slots + (*ck & 0x0F);
@@ -103,7 +113,7 @@ ox_cache_get(Cache cache, const char *key, VALUE **slot) {
 
 void
 ox_cache_print(Cache cache) {
-    //printf("-------------------------------------------\n");
+    /*printf("-------------------------------------------\n");*/
     slot_print(cache, 0);
 }
 
@@ -120,7 +130,7 @@ slot_print(Cache c, unsigned int depth) {
     indent[depth] = '\0';
     for (i = 0, cp = c->slots; i < 16; i++, cp++) {
         if (0 == *cp) {
-            //printf("%s%02u:\n", indent, i);
+            /*printf("%s%02u:\n", indent, i);*/
         } else {
             if (0 == (*cp)->key && Qundef == (*cp)->value) {
                 printf("%s%02u:\n", indent, i);

data/ext/ox/cache8.c

@@ -75,7 +75,7 @@ ox_cache8_get(Cache8 cache, VALUE key, slot_t **slot) {
 
 void
 ox_cache8_print(Cache8 cache) {
-    //printf("-------------------------------------------\n");
+    /*printf("-------------------------------------------\n"); */
     slot_print(cache, 0, 0);
 }
 
@@ -88,7 +88,7 @@ slot_print(Cache8 c, VALUE key, unsigned int depth) {
     for (i = 0, cp = c->slots; i < SLOT_CNT; i++, cp++) {
         if (0 != *cp) {
             k = (key << BITS) | i;
-            //printf("*** key: 0x%016lx  depth: %u  i: %u\n", k, depth, i);
+            /*printf("*** key: 0x%016lx  depth: %u  i: %u\n", k, depth, i); */
             if (DEPTH - 1 == depth) {
                 printf("0x%016lx: %4lu\n", k, (unsigned long)*cp);
             } else {

data/ext/ox/cache8_test.c

@@ -63,7 +63,7 @@ ox_cache8_test() {
         } else {
             printf("*** get on 0x%016lx returned %lu\n", *d, v);
         }
-        //ox_cache8_print(c);
+        /*ox_cache8_print(c); */
     }
     ox_cache8_print(c);
 }

data/ext/ox/cache_test.c

@@ -52,9 +52,9 @@ ox_cache_test() {
         v = ox_cache_get(c, *d, &slot);
         if (Qundef == v) {
             if (0 == slot) {
-                //printf("*** failed to get a slot for %s\n", *d);
+                /*printf("*** failed to get a slot for %s\n", *d); */
             } else {
-                //printf("*** added '%s' to cache\n", *d);
+                /*printf("*** added '%s' to cache\n", *d); */
                 v = ID2SYM(rb_intern(*d));
                 *slot = v;
             }
@@ -63,7 +63,7 @@ ox_cache_test() {
 
             printf("*** get on '%s' returned '%s' (%s)\n", *d, StringValuePtr(rs), rb_class2name(rb_obj_class(v)));
         }
-        //ox_cache_print(c);
+        /*ox_cache_print(c); */
     }
     ox_cache_print(c);
 }

data/ext/ox/dump.c

@@ -51,7 +51,7 @@ typedef struct _Element {
     struct _Str         clas;
     struct _Str         attr;
     unsigned long       id;
-    int                 indent; // < 0 indicates no \n
+    int                 indent; /* < 0 indicates no \n */
     int                 closed;
     char                type;
 } *Element;
@@ -66,7 +66,7 @@ typedef struct _Out {
     Cache8              circ_cache;
     unsigned long       circ_cnt;
     int                 indent;
-    int                 depth; // used by dumpHash
+    int                 depth; /* used by dumpHash */
     Options             opts;
     VALUE		obj;
 } *Out;
@@ -97,7 +97,7 @@ static void     dump_time_thin(Out out, VALUE obj);
 static void     dump_time_xsd(Out out, VALUE obj);
 static int      dump_hash(VALUE key, VALUE value, Out out);
 
-static int      is_xml_friendly(const u_char *str, int len);
+static int      is_xml_friendly(const uchar *str, int len);
 
 static const char	hex_chars[17] = "0123456789abcdef";
 
@@ -112,7 +112,7 @@ static char     xml_friendly_chars[256] = "\
 11111111111111111111111111111111";
 
 inline static int
-is_xml_friendly(const u_char *str, int len) {
+is_xml_friendly(const uchar *str, int len) {
     for (; 0 < len; str++, len--) {
         if ('1' != xml_friendly_chars[*str]) {
             return 0;
@@ -122,7 +122,7 @@ is_xml_friendly(const u_char *str, int len) {
 }
 
 inline static size_t
-xml_str_len(const u_char *str, size_t len) {
+xml_str_len(const uchar *str, size_t len) {
     size_t	size = 0;
 
     for (; 0 < len; str++, len--) {
@@ -132,8 +132,8 @@ xml_str_len(const u_char *str, size_t len) {
 }
 
 inline static void
-dump_hex(u_char c, Out out) {
-    u_char	d = (c >> 4) & 0x0F;
+dump_hex(uchar c, Out out) {
+    uchar	d = (c >> 4) & 0x0F;
 
     *out->cur++ = hex_chars[d];
     d = c & 0x0F;
@@ -152,12 +152,12 @@ obj_class_code(VALUE obj) {
     case T_FALSE:          return FalseClassCode;
     case T_FIXNUM:         return FixnumCode;
     case T_FLOAT:          return FloatCode;
-    case T_STRING:         return (is_xml_friendly((u_char*)StringValuePtr(obj), (int)RSTRING_LEN(obj))) ? StringCode : String64Code;
+    case T_STRING:         return (is_xml_friendly((uchar*)StringValuePtr(obj), (int)RSTRING_LEN(obj))) ? StringCode : String64Code;
     case T_SYMBOL:
     {
         const char      *sym = rb_id2name(SYM2ID(obj));
 
-        return (is_xml_friendly((u_char*)sym, (int)strlen(sym))) ? SymbolCode : Symbol64Code;
+        return (is_xml_friendly((uchar*)sym, (int)strlen(sym))) ? SymbolCode : Symbol64Code;
     }
     case T_DATA:           return (rb_cTime == clas) ? TimeCode : ((ox_date_class == clas) ? DateCode : 0);
     case T_STRUCT:         return (rb_cRange == clas) ? RangeCode : StructCode;
@@ -258,7 +258,7 @@ grow(Out out, size_t len) {
     if (size <= len * 2 + pos) {
         size += len;
     }
-    REALLOC_N(out->buf, char, size + 10); // 10 extra for terminator character plus extra (paranoid)
+    REALLOC_N(out->buf, char, size + 10); /* 10 extra for terminator character plus extra (paranoid) */
     out->end = out->buf + size;
     out->cur = out->buf + pos;
 }
@@ -267,14 +267,14 @@ static void
 dump_start(Out out, Element e) {
     size_t      size = e->indent + 4;
 
-    if (0 < e->attr.len) { // a="attr"
+    if (0 < e->attr.len) { /* a="attr" */
         size += e->attr.len + 5;
     }
-    if (0 < e->clas.len) { // c="class"
+    if (0 < e->clas.len) { /* c="class" */
         size += e->clas.len + 5;
     }
-    if (0 < e->id) { // i="id"
-        size += 24; // over estimate, 19 digits
+    if (0 < e->id) { /* i="id" */
+        size += 24; /* over estimate, 19 digits */
     }
     if (out->end - out->cur <= (long)size) {
         grow(out, size);
@@ -337,13 +337,13 @@ dump_value(Out out, const char *value, size_t size) {
 
 inline static void
 dump_str_value(Out out, const char *value, size_t size) {
-    size_t	xsize = xml_str_len((const u_char*)value, size);
+    size_t	xsize = xml_str_len((const uchar*)value, size);
 
     if (out->end - out->cur <= (long)xsize) {
         grow(out, xsize);
     }
     for (; '\0' != *value; value++) {
-	if ('1' == xml_friendly_chars[(u_char)*value]) {
+	if ('1' == xml_friendly_chars[(uchar)*value]) {
 	    *out->cur++ = *value;
 	} else {
 	    *out->cur++ = '&';
@@ -501,8 +501,9 @@ dump_time_xsd(Out out, VALUE obj) {
     if (out->end - out->cur <= 33) {
         grow(out, 33);
     }
-    // 2010-07-09T10:47:45.895826+09:00
+    /* 2010-07-09T10:47:45.895826+09:00 */
     tm = localtime(&sec);
+#if HAS_TM_GMTOFF
     if (0 > tm->tm_gmtoff) {
         tzsign = '-';
         tzhour = (int)(tm->tm_gmtoff / -3600);
@@ -511,7 +512,11 @@ dump_time_xsd(Out out, VALUE obj) {
         tzhour = (int)(tm->tm_gmtoff / 3600);
         tzmin = (int)(tm->tm_gmtoff / 60) - (tzhour * 60);
     }
-    // TBD replace with more efficient printer
+#else
+    tzhour = 0;
+    tzmin = 0;
+#endif
+    /* TBD replace with more efficient printer */
     out->cur += sprintf(out->cur, "%04d-%02d-%02dT%02d:%02d:%02d.%06ld%c%02d:%02d",
 			tm->tm_year + 1900, tm->tm_mon + 1, tm->tm_mday,
 			tm->tm_hour, tm->tm_min, tm->tm_sec, nsec / 1000,
@@ -555,7 +560,7 @@ dump_obj(ID aid, VALUE obj, unsigned int depth, Out out) {
 
     out->obj = obj;
     if (0 == aid) {
-        //e.attr.str = 0;
+        /*e.attr.str = 0; */
         e.attr.len = 0;
     } else {
         e.attr.str = rb_id2name(aid);
@@ -635,7 +640,7 @@ dump_obj(ID aid, VALUE obj, unsigned int depth, Out out) {
         break;
     case T_FLOAT:
         e.type = FloatCode;
-        cnt = sprintf(value_buf, "%0.16g", rb_num2dbl(obj)); // used sprintf due to bug in snprintf
+        cnt = sprintf(value_buf, "%0.16g", rb_num2dbl(obj)); /* used sprintf due to bug in snprintf */
         out->w_start(out, &e);
         dump_value(out, value_buf, cnt);
         e.indent = -1;
@@ -651,7 +656,7 @@ dump_obj(ID aid, VALUE obj, unsigned int depth, Out out) {
         str = StringValuePtr(obj);
         cnt = (int)RSTRING_LEN(obj);
 #if USE_B64
-        if (is_xml_friendly((u_char*)str, cnt)) {
+        if (is_xml_friendly((uchar*)str, cnt)) {
             e.type = StringCode;
             out->w_start(out, &e);
             dump_str_value(out, str, cnt);
@@ -662,7 +667,7 @@ dump_obj(ID aid, VALUE obj, unsigned int depth, Out out) {
             char        *b64 = ALLOCA_N(char, size + 1);
 
             e.type = String64Code;
-            to_base64((u_char*)str, cnt, b64);
+            to_base64((uchar*)str, cnt, b64);
             out->w_start(out, &e);
             dump_value(out, b64, size);
             e.indent = -1;
@@ -683,7 +688,7 @@ dump_obj(ID aid, VALUE obj, unsigned int depth, Out out) {
 
         cnt = (int)strlen(sym);
 #if USE_B64
-        if (is_xml_friendly((u_char*)sym, cnt)) {
+        if (is_xml_friendly((uchar*)sym, cnt)) {
             e.type = SymbolCode;
             out->w_start(out, &e);
             dump_str_value(out, sym, cnt);
@@ -694,7 +699,7 @@ dump_obj(ID aid, VALUE obj, unsigned int depth, Out out) {
             char        *b64 = ALLOCA_N(char, size + 1);
 
             e.type = Symbol64Code;
-            to_base64((u_char*)sym, cnt, b64);
+            to_base64((uchar*)sym, cnt, b64);
             out->w_start(out, &e);
             dump_value(out, b64, size);
             e.indent = -1;
@@ -805,7 +810,7 @@ dump_obj(ID aid, VALUE obj, unsigned int depth, Out out) {
             out->w_start(out, &e);
             dump_gen_element(obj, depth + 1, out);
             out->w_end(out, &e);
-        } else { // Object
+        } else { /* Object */
 #if HAS_IVAR_HELPERS
             e.type = (Qtrue == rb_obj_is_kind_of(obj, rb_eException)) ? ExceptionCode : ObjectCode;
             cnt = (int)rb_ivar_count(obj);
@@ -820,10 +825,10 @@ dump_obj(ID aid, VALUE obj, unsigned int depth, Out out) {
                 out->w_end(out, &e);
             }
 #else
-            //VALUE       vars = rb_obj_instance_variables(obj);
-	    //#else
+            /*VALUE       vars = rb_obj_instance_variables(obj); */
+	    /*#else */
             VALUE       vars = rb_funcall2(obj, rb_intern("instance_variables"), 0, 0);
-	    //#endif            
+	    /*#endif             */
             e.type = (Qtrue == rb_obj_is_kind_of(obj, rb_eException)) ? ExceptionCode : ObjectCode;
             cnt = (int)RARRAY_LEN(vars);
             e.closed = (0 >= cnt);
@@ -862,14 +867,14 @@ dump_obj(ID aid, VALUE obj, unsigned int depth, Out out) {
         e.type = RegexpCode;
         out->w_start(out, &e);
 #if USE_B64
-        if (is_xml_friendly((u_char*)s, cnt)) {
-            //dump_value(out, "/", 1);
+        if (is_xml_friendly((uchar*)s, cnt)) {
+            /*dump_value(out, "/", 1); */
             dump_str_value(out, s, cnt);
         } else {
             ulong       size = b64_size(cnt);
             char        *b64 = ALLOCA_N(char, size + 1);
 
-            to_base64((u_char*)s, cnt, b64);
+            to_base64((uchar*)s, cnt, b64);
             dump_value(out, b64, size);
         }
 #else
@@ -1103,7 +1108,7 @@ dump_gen_attr(VALUE key, VALUE value, Out out) {
     fill_value(out, ks, klen);
     *out->cur++ = '=';
     *out->cur++ = '"';
-    fill_value(out, StringValuePtr(value), RSTRING_LEN(value));
+    dump_str_value(out, StringValuePtr(value), RSTRING_LEN(value));
     *out->cur++ = '"';
 
     return ST_CONTINUE;
@@ -1148,7 +1153,7 @@ dump_obj_to_xml(VALUE obj, Options copts, Out out) {
 
     out->w_time = (Yes == copts->xsd_date) ? dump_time_xsd : dump_time_thin;
     out->buf = ALLOC_N(char, 65336);
-    out->end = out->buf + 65325; // 10 less than end plus extra for possible errors
+    out->end = out->buf + 65325; /* 10 less than end plus extra for possible errors */
     out->cur = out->buf;
     out->circ_cache = 0;
     out->circ_cnt = 0;