ezxml: Slightly modified version of ezxml added

ezxml is a light xml parser, homepage: http://ezxml.sourceforge.net/

Removed xml write and modify functions, added cmake build system

5 files changed, 871 insertions, 0 deletions

diff --git a/libezxml/CMakeLists.txt b/libezxml/CMakeLists.txt
new file mode 100644
index 0000000..edf52c7
--- /dev/null
+++ b/libezxml/CMakeLists.txt
@@ -0,0 +1,7 @@
+cmake_minimum_required (VERSION 2.4)
+
+set (PACKAGE "libezxml")
+project (${PACKAGE} C)
+
+add_subdirectory (src)
+
diff --git a/libezxml/COPYING b/libezxml/COPYING
new file mode 100644
index 0000000..80e4e88
--- /dev/null
+++ b/libezxml/COPYING
@@ -0,0 +1,20 @@
+Copyright 2004-2006 Aaron Voisine <aaron@voisine.org>
+
+Permission is hereby granted, free of charge, to any person obtaining
+a copy of this software and associated documentation files (the
+"Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, sublicense, and/or sell copies of the Software, and to
+permit persons to whom the Software is furnished to do so, subject to
+the following conditions:
+
+The above copyright notice and this permission notice shall be included
+in all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
diff --git a/libezxml/src/CMakeLists.txt b/libezxml/src/CMakeLists.txt
new file mode 100644
index 0000000..d1cf76d
--- /dev/null
+++ b/libezxml/src/CMakeLists.txt
@@ -0,0 +1,4 @@
+set (CMAKE_C_FLAGS -Wall)
+
+add_library (ezxml STATIC ezxml.c)
+target_link_libraries (ezxml)
diff --git a/libezxml/src/ezxml.c b/libezxml/src/ezxml.c
new file mode 100644
index 0000000..17c8997
--- /dev/null
+++ b/libezxml/src/ezxml.c
@@ -0,0 +1,712 @@
+/* ezxml.c
+ *
+ * Copyright 2004-2006 Aaron Voisine <aaron@voisine.org>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining
+ * a copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sublicense, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+ * IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+ * CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+ * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+ * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#include <stdlib.h>
+#include <stdio.h>
+#include <stdarg.h>
+#include <string.h>
+#include <ctype.h>
+#include <unistd.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include "ezxml.h"
+
+#define EZXML_WS   "\t\r\n "  // whitespace
+#define EZXML_ERRL 128        // maximum error string length
+
+typedef struct ezxml_root *ezxml_root_t;
+struct ezxml_root {       // additional data for the root tag
+    struct ezxml xml;     // is a super-struct built on top of ezxml struct
+    ezxml_t cur;          // current xml tree insertion point
+    char *m;              // original xml string
+    size_t len;           // length of allocated memory for mmap, -1 for malloc
+    char *u;              // UTF-8 conversion of string if original was UTF-16
+    char *s;              // start of work area
+    char *e;              // end of work area
+    char **ent;           // general entities (ampersand sequences)
+    char ***attr;         // default attributes
+    char ***pi;           // processing instructions
+    short standalone;     // non-zero if <?xml standalone="yes"?>
+    char err[EZXML_ERRL]; // error string
+};
+
+char *EZXML_NIL[] = { NULL }; // empty, null terminated array of strings
+
+// returns the first child tag with the given name or NULL if not found
+ezxml_t ezxml_child(ezxml_t xml, const char *name)
+{
+    xml = (xml) ? xml->child : NULL;
+    while (xml && strcmp(name, xml->name)) xml = xml->sibling;
+    return xml;
+}
+
+// returns the Nth tag with the same name in the same subsection or NULL if not
+// found
+ezxml_t ezxml_idx(ezxml_t xml, int idx)
+{
+    for (; xml && idx; idx--) xml = xml->next;
+    return xml;
+}
+
+// returns the value of the requested tag attribute or NULL if not found
+const char *ezxml_attr(ezxml_t xml, const char *attr)
+{
+    int i = 0, j = 1;
+    ezxml_root_t root = (ezxml_root_t)xml;
+
+    if (! xml || ! xml->attr) return NULL;
+    while (xml->attr[i] && strcmp(attr, xml->attr[i])) i += 2;
+    if (xml->attr[i]) return xml->attr[i + 1]; // found attribute
+
+    while (root->xml.parent) root = (ezxml_root_t)root->xml.parent; // root tag
+    for (i = 0; root->attr[i] && strcmp(xml->name, root->attr[i][0]); i++);
+    if (! root->attr[i]) return NULL; // no matching default attributes
+    while (root->attr[i][j] && strcmp(attr, root->attr[i][j])) j += 3;
+    return (root->attr[i][j]) ? root->attr[i][j + 1] : NULL; // found default
+}
+
+// same as ezxml_get but takes an already initialized va_list
+ezxml_t ezxml_vget(ezxml_t xml, va_list ap)
+{
+    char *name = va_arg(ap, char *);
+    int idx = -1;
+
+    if (name && *name) {
+        idx = va_arg(ap, int);    
+        xml = ezxml_child(xml, name);
+    }
+    return (idx < 0) ? xml : ezxml_vget(ezxml_idx(xml, idx), ap);
+}
+
+// Traverses the xml tree to retrieve a specific subtag. Takes a variable
+// length list of tag names and indexes. The argument list must be terminated
+// by either an index of -1 or an empty string tag name. Example: 
+// title = ezxml_get(library, "shelf", 0, "book", 2, "title", -1);
+// This retrieves the title of the 3rd book on the 1st shelf of library.
+// Returns NULL if not found.
+ezxml_t ezxml_get(ezxml_t xml, ...)
+{
+    va_list ap;
+    ezxml_t r;
+
+    va_start(ap, xml);
+    r = ezxml_vget(xml, ap);
+    va_end(ap);
+    return r;
+}
+
+// set an error string and return root
+ezxml_t ezxml_err(ezxml_root_t root, char *s, const char *err, ...)
+{
+    va_list ap;
+    int line = 1;
+    char *t, fmt[EZXML_ERRL];
+    
+    for (t = root->s; t < s; t++) if (*t == '\n') line++;
+    snprintf(fmt, EZXML_ERRL, "[error near line %d]: %s", line, err);
+
+    va_start(ap, err);
+    vsnprintf(root->err, EZXML_ERRL, fmt, ap);
+    va_end(ap);
+
+    return &root->xml;
+}
+
+// Recursively decodes entity and character references and normalizes new lines
+// ent is a null terminated array of alternating entity names and values. set t
+// to '&' for general entity decoding, '%' for parameter entity decoding, 'c'
+// for cdata sections, ' ' for attribute normalization, or '*' for non-cdata
+// attribute normalization. Returns s, or if the decoded string is longer than
+// s, returns a malloced string that must be freed.
+char *ezxml_decode(char *s, char **ent, char t)
+{
+    char *e, *r = s, *m = s;
+    long b, c, d, l;
+
+    for (; *s; s++) { // normalize line endings
+        while (*s == '\r') {
+            *(s++) = '\n';
+            if (*s == '\n') memmove(s, (s + 1), strlen(s));
+        }
+    }
+    
+    for (s = r; ; ) {
+        while (*s && *s != '&' && (*s != '%' || t != '%') && !isspace(*s)) s++;
+
+        if (! *s) break;
+        else if (t != 'c' && ! strncmp(s, "&#", 2)) { // character reference
+            if (s[2] == 'x') c = strtol(s + 3, &e, 16); // base 16
+            else c = strtol(s + 2, &e, 10); // base 10
+            if (! c || *e != ';') { s++; continue; } // not a character ref
+
+            if (c < 0x80) *(s++) = c; // US-ASCII subset
+            else { // multi-byte UTF-8 sequence
+                for (b = 0, d = c; d; d /= 2) b++; // number of bits in c
+                b = (b - 2) / 5; // number of bytes in payload
+                *(s++) = (0xFF << (7 - b)) | (c >> (6 * b)); // head
+                while (b) *(s++) = 0x80 | ((c >> (6 * --b)) & 0x3F); // payload
+            }
+
+            memmove(s, strchr(s, ';') + 1, strlen(strchr(s, ';')));
+        }
+        else if ((*s == '&' && (t == '&' || t == ' ' || t == '*')) ||
+                 (*s == '%' && t == '%')) { // entity reference
+            for (b = 0; ent[b] && strncmp(s + 1, ent[b], strlen(ent[b]));
+                 b += 2); // find entity in entity list
+
+            if (ent[b++]) { // found a match
+                if ((c = strlen(ent[b])) - 1 > (e = strchr(s, ';')) - s) {
+                    l = (d = (s - r)) + c + strlen(e); // new length
+                    r = (r == m) ? strcpy(malloc(l), r) : realloc(r, l);
+                    e = strchr((s = r + d), ';'); // fix up pointers
+                }
+
+                memmove(s + c, e + 1, strlen(e)); // shift rest of string
+                strncpy(s, ent[b], c); // copy in replacement text
+            }
+            else s++; // not a known entity
+        }
+        else if ((t == ' ' || t == '*') && isspace(*s)) *(s++) = ' ';
+        else s++; // no decoding needed
+    }
+
+    if (t == '*') { // normalize spaces for non-cdata attributes
+        for (s = r; *s; s++) {
+            if ((l = strspn(s, " "))) memmove(s, s + l, strlen(s + l) + 1);
+            while (*s && *s != ' ') s++;
+        }
+        if (--s >= r && *s == ' ') *s = '\0'; // trim any trailing space
+    }
+    return r;
+}
+
+// called when parser finds start of new tag
+void ezxml_open_tag(ezxml_root_t root, char *name, char **attr)
+{
+    ezxml_t xml = root->cur;
+    
+    if (xml->name) xml = ezxml_add_child(xml, name, strlen(xml->txt));
+    else xml->name = name; // first open tag
+
+    xml->attr = attr;
+    root->cur = xml; // update tag insertion point
+}
+
+// called when parser finds character content between open and closing tag
+void ezxml_char_content(ezxml_root_t root, char *s, size_t len, char t)
+{
+    ezxml_t xml = root->cur;
+    char *m = s;
+    size_t l;
+
+    if (! xml || ! xml->name || ! len) return; // sanity check
+
+    s[len] = '\0'; // null terminate text (calling functions anticipate this)
+    len = strlen(s = ezxml_decode(s, root->ent, t)) + 1;
+
+    if (! *(xml->txt)) xml->txt = s; // initial character content
+    else { // allocate our own memory and make a copy
+        xml->txt = (xml->flags & EZXML_TXTM) // allocate some space
+                   ? realloc(xml->txt, (l = strlen(xml->txt)) + len)
+                   : strcpy(malloc((l = strlen(xml->txt)) + len), xml->txt);
+        strcpy(xml->txt + l, s); // add new char content
+        if (s != m) free(s); // free s if it was malloced by ezxml_decode()
+    }
+
+    if (xml->txt != m) ezxml_set_flag(xml, EZXML_TXTM);
+}
+
+// called when parser finds closing tag
+ezxml_t ezxml_close_tag(ezxml_root_t root, char *name, char *s)
+{
+    if (! root->cur || ! root->cur->name || strcmp(name, root->cur->name))
+        return ezxml_err(root, s, "unexpected closing tag </%s>", name);
+
+    root->cur = root->cur->parent;
+    return NULL;
+}
+
+// checks for circular entity references, returns non-zero if no circular
+// references are found, zero otherwise
+int ezxml_ent_ok(char *name, char *s, char **ent)
+{
+    int i;
+
+    for (; ; s++) {
+        while (*s && *s != '&') s++; // find next entity reference
+        if (! *s) return 1;
+        if (! strncmp(s + 1, name, strlen(name))) return 0; // circular ref.
+        for (i = 0; ent[i] && strncmp(ent[i], s + 1, strlen(ent[i])); i += 2);
+        if (ent[i] && ! ezxml_ent_ok(name, ent[i + 1], ent)) return 0;
+    }
+}
+
+// called when the parser finds a processing instruction
+void ezxml_proc_inst(ezxml_root_t root, char *s, size_t len)
+{
+    int i = 0, j = 1;
+    char *target = s;
+
+    s[len] = '\0'; // null terminate instruction
+    if (*(s += strcspn(s, EZXML_WS))) {
+        *s = '\0'; // null terminate target
+        s += strspn(s + 1, EZXML_WS) + 1; // skip whitespace after target
+    }
+
+    if (! strcmp(target, "xml")) { // <?xml ... ?>
+        if ((s = strstr(s, "standalone")) && ! strncmp(s + strspn(s + 10,
+            EZXML_WS "='\"") + 10, "yes", 3)) root->standalone = 1;
+        return;
+    }
+
+    if (! root->pi[0]) *(root->pi = malloc(sizeof(char **))) = NULL; //first pi
+
+    while (root->pi[i] && strcmp(target, root->pi[i][0])) i++; // find target
+    if (! root->pi[i]) { // new target
+        root->pi = realloc(root->pi, sizeof(char **) * (i + 2));
+        root->pi[i] = malloc(sizeof(char *) * 3);
+        root->pi[i][0] = target;
+        root->pi[i][1] = (char *)(root->pi[i + 1] = NULL); // terminate pi list
+        root->pi[i][2] = strdup(""); // empty document position list
+    }
+
+    while (root->pi[i][j]) j++; // find end of instruction list for this target
+    root->pi[i] = realloc(root->pi[i], sizeof(char *) * (j + 3));
+    root->pi[i][j + 2] = realloc(root->pi[i][j + 1], j + 1);
+    strcpy(root->pi[i][j + 2] + j - 1, (root->xml.name) ? ">" : "<");
+    root->pi[i][j + 1] = NULL; // null terminate pi list for this target
+    root->pi[i][j] = s; // set instruction
+}
+
+// called when the parser finds an internal doctype subset
+short ezxml_internal_dtd(ezxml_root_t root, char *s, size_t len)
+{
+    char q, *c, *t, *n = NULL, *v, **ent, **pe;
+    int i, j;
+    
+    pe = memcpy(malloc(sizeof(EZXML_NIL)), EZXML_NIL, sizeof(EZXML_NIL));
+
+    for (s[len] = '\0'; s; ) {
+        while (*s && *s != '<' && *s != '%') s++; // find next declaration
+
+        if (! *s) break;
+        else if (! strncmp(s, "<!ENTITY", 8)) { // parse entity definitions
+            c = s += strspn(s + 8, EZXML_WS) + 8; // skip white space separator
+            n = s + strspn(s, EZXML_WS "%"); // find name
+            *(s = n + strcspn(n, EZXML_WS)) = ';'; // append ; to name
+
+            v = s + strspn(s + 1, EZXML_WS) + 1; // find value
+            if ((q = *(v++)) != '"' && q != '\'') { // skip externals
+                s = strchr(s, '>');
+                continue;
+            }
+
+            for (i = 0, ent = (*c == '%') ? pe : root->ent; ent[i]; i++);
+            ent = realloc(ent, (i + 3) * sizeof(char *)); // space for next ent
+            if (*c == '%') pe = ent;
+            else root->ent = ent;
+
+            *(++s) = '\0'; // null terminate name
+            if ((s = strchr(v, q))) *(s++) = '\0'; // null terminate value
+            ent[i + 1] = ezxml_decode(v, pe, '%'); // set value
+            ent[i + 2] = NULL; // null terminate entity list
+            if (! ezxml_ent_ok(n, ent[i + 1], ent)) { // circular reference
+                if (ent[i + 1] != v) free(ent[i + 1]);
+                ezxml_err(root, v, "circular entity declaration &%s", n);
+                break;
+            }
+            else ent[i] = n; // set entity name
+        }
+        else if (! strncmp(s, "<!ATTLIST", 9)) { // parse default attributes
+            t = s + strspn(s + 9, EZXML_WS) + 9; // skip whitespace separator
+            if (! *t) { ezxml_err(root, t, "unclosed <!ATTLIST"); break; }
+            if (*(s = t + strcspn(t, EZXML_WS ">")) == '>') continue;
+            else *s = '\0'; // null terminate tag name
+            for (i = 0; root->attr[i] && strcmp(n, root->attr[i][0]); i++);
+
+            while (*(n = ++s + strspn(s, EZXML_WS)) && *n != '>') {
+                if (*(s = n + strcspn(n, EZXML_WS))) *s = '\0'; // attr name
+                else { ezxml_err(root, t, "malformed <!ATTLIST"); break; }
+
+                s += strspn(s + 1, EZXML_WS) + 1; // find next token
+                c = (strncmp(s, "CDATA", 5)) ? "*" : " "; // is it cdata?
+                if (! strncmp(s, "NOTATION", 8))
+                    s += strspn(s + 8, EZXML_WS) + 8;
+                s = (*s == '(') ? strchr(s, ')') : s + strcspn(s, EZXML_WS);
+                if (! s) { ezxml_err(root, t, "malformed <!ATTLIST"); break; }
+
+                s += strspn(s, EZXML_WS ")"); // skip white space separator
+                if (! strncmp(s, "#FIXED", 6))
+                    s += strspn(s + 6, EZXML_WS) + 6;
+                if (*s == '#') { // no default value
+                    s += strcspn(s, EZXML_WS ">") - 1;
+                    if (*c == ' ') continue; // cdata is default, nothing to do
+                    v = NULL;
+                }
+                else if ((*s == '"' || *s == '\'')  &&  // default value
+                         (s = strchr(v = s + 1, *s))) *s = '\0';
+                else { ezxml_err(root, t, "malformed <!ATTLIST"); break; }
+
+                if (! root->attr[i]) { // new tag name
+                    root->attr = (! i) ? malloc(2 * sizeof(char **))
+                                       : realloc(root->attr,
+                                                 (i + 2) * sizeof(char **));
+                    root->attr[i] = malloc(2 * sizeof(char *));
+                    root->attr[i][0] = t; // set tag name
+                    root->attr[i][1] = (char *)(root->attr[i + 1] = NULL);
+                }
+
+                for (j = 1; root->attr[i][j]; j += 3); // find end of list
+                root->attr[i] = realloc(root->attr[i],
+                                        (j + 4) * sizeof(char *));
+
+                root->attr[i][j + 3] = NULL; // null terminate list
+                root->attr[i][j + 2] = c; // is it cdata?
+                root->attr[i][j + 1] = (v) ? ezxml_decode(v, root->ent, *c)
+                                           : NULL;
+                root->attr[i][j] = n; // attribute name 
+            }
+        }
+        else if (! strncmp(s, "<!--", 4)) s = strstr(s + 4, "-->"); // comments
+        else if (! strncmp(s, "<?", 2)) { // processing instructions
+            if ((s = strstr(c = s + 2, "?>")))
+                ezxml_proc_inst(root, c, s++ - c);
+        }
+        else if (*s == '<') s = strchr(s, '>'); // skip other declarations
+        else if (*(s++) == '%' && ! root->standalone) break;
+    }
+
+    free(pe);
+    return ! *root->err;
+}
+
+// Converts a UTF-16 string to UTF-8. Returns a new string that must be freed
+// or NULL if no conversion was needed.
+char *ezxml_str2utf8(char **s, size_t *len)
+{
+    char *u;
+    size_t l = 0, sl, max = *len;
+    long c, d;
+    int b, be = (**s == '\xFE') ? 1 : (**s == '\xFF') ? 0 : -1;
+
+    if (be == -1) return NULL; // not UTF-16
+
+    u = malloc(max);
+    for (sl = 2; sl < *len - 1; sl += 2) {
+        c = (be) ? (((*s)[sl] & 0xFF) << 8) | ((*s)[sl + 1] & 0xFF)  //UTF-16BE
+                 : (((*s)[sl + 1] & 0xFF) << 8) | ((*s)[sl] & 0xFF); //UTF-16LE
+        if (c >= 0xD800 && c <= 0xDFFF && (sl += 2) < *len - 1) { // high-half
+            d = (be) ? (((*s)[sl] & 0xFF) << 8) | ((*s)[sl + 1] & 0xFF)
+                     : (((*s)[sl + 1] & 0xFF) << 8) | ((*s)[sl] & 0xFF);
+            c = (((c & 0x3FF) << 10) | (d & 0x3FF)) + 0x10000;
+        }
+
+        while (l + 6 > max) u = realloc(u, max += EZXML_BUFSIZE);
+        if (c < 0x80) u[l++] = c; // US-ASCII subset
+        else { // multi-byte UTF-8 sequence
+            for (b = 0, d = c; d; d /= 2) b++; // bits in c
+            b = (b - 2) / 5; // bytes in payload
+            u[l++] = (0xFF << (7 - b)) | (c >> (6 * b)); // head
+            while (b) u[l++] = 0x80 | ((c >> (6 * --b)) & 0x3F); // payload
+        }
+    }
+    return *s = realloc(u, *len = l);
+}
+
+// frees a tag attribute list
+void ezxml_free_attr(char **attr) {
+    int i = 0;
+    char *m;
+    
+    if (! attr || attr == EZXML_NIL) return; // nothing to free
+    while (attr[i]) i += 2; // find end of attribute list
+    m = attr[i + 1]; // list of which names and values are malloced
+    for (i = 0; m[i]; i++) {
+        if (m[i] & EZXML_NAMEM) free(attr[i * 2]);
+        if (m[i] & EZXML_TXTM) free(attr[(i * 2) + 1]);
+    }
+    free(m);
+    free(attr);
+}
+
+// parse the given xml string and return an ezxml structure
+ezxml_t ezxml_parse_str(char *s, size_t len)
+{
+    ezxml_root_t root = (ezxml_root_t)ezxml_new(NULL);
+    char q, e, *d, **attr, **a = NULL; // initialize a to avoid compile warning
+    int l, i, j;
+
+    root->m = s;
+    if (! len) return ezxml_err(root, NULL, "root tag missing");
+    root->u = ezxml_str2utf8(&s, &len); // convert utf-16 to utf-8
+    root->e = (root->s = s) + len; // record start and end of work area
+    
+    e = s[len - 1]; // save end char
+    s[len - 1] = '\0'; // turn end char into null terminator
+
+    while (*s && *s != '<') s++; // find first tag
+    if (! *s) return ezxml_err(root, s, "root tag missing");
+
+    for (; ; ) {
+        attr = (char **)EZXML_NIL;
+        d = ++s;
+        
+        if (isalpha(*s) || *s == '_' || *s == ':' || *s < '\0') { // new tag
+            if (! root->cur)
+                return ezxml_err(root, d, "markup outside of root element");
+
+            s += strcspn(s, EZXML_WS "/>");
+            while (isspace(*s)) *(s++) = '\0'; // null terminate tag name
+  
+            if (*s && *s != '/' && *s != '>') // find tag in default attr list
+                for (i = 0; (a = root->attr[i]) && strcmp(a[0], d); i++);
+
+            for (l = 0; *s && *s != '/' && *s != '>'; l += 2) { // new attrib
+                attr = (l) ? realloc(attr, (l + 4) * sizeof(char *))
+                           : malloc(4 * sizeof(char *)); // allocate space
+                attr[l + 3] = (l) ? realloc(attr[l + 1], (l / 2) + 2)
+                                  : malloc(2); // mem for list of maloced vals
+                strcpy(attr[l + 3] + (l / 2), " "); // value is not malloced
+                attr[l + 2] = NULL; // null terminate list
+                attr[l + 1] = ""; // temporary attribute value
+                attr[l] = s; // set attribute name
+
+                s += strcspn(s, EZXML_WS "=/>");
+                if (*s == '=' || isspace(*s)) { 
+                    *(s++) = '\0'; // null terminate tag attribute name
+                    q = *(s += strspn(s, EZXML_WS "="));
+                    if (q == '"' || q == '\'') { // attribute value
+                        attr[l + 1] = ++s;
+                        while (*s && *s != q) s++;
+                        if (*s) *(s++) = '\0'; // null terminate attribute val
+                        else {
+                            ezxml_free_attr(attr);
+                            return ezxml_err(root, d, "missing %c", q);
+                        }
+
+                        for (j = 1; a && a[j] && strcmp(a[j], attr[l]); j +=3);
+                        attr[l + 1] = ezxml_decode(attr[l + 1], root->ent, (a
+                                                   && a[j]) ? *a[j + 2] : ' ');
+                        if (attr[l + 1] < d || attr[l + 1] > s)
+                            attr[l + 3][l / 2] = EZXML_TXTM; // value malloced
+                    }
+                }
+                while (isspace(*s)) s++;
+            }
+
+            if (*s == '/') { // self closing tag
+                *(s++) = '\0';
+                if ((*s && *s != '>') || (! *s && e != '>')) {
+                    if (l) ezxml_free_attr(attr);
+                    return ezxml_err(root, d, "missing >");
+                }
+                ezxml_open_tag(root, d, attr);
+                ezxml_close_tag(root, d, s);
+            }
+            else if ((q = *s) == '>' || (! *s && e == '>')) { // open tag
+                *s = '\0'; // temporarily null terminate tag name
+                ezxml_open_tag(root, d, attr);
+                *s = q;
+            }
+            else {
+                if (l) ezxml_free_attr(attr);
+                return ezxml_err(root, d, "missing >"); 
+            }
+        }
+        else if (*s == '/') { // close tag
+            s += strcspn(d = s + 1, EZXML_WS ">") + 1;
+            if (! (q = *s) && e != '>') return ezxml_err(root, d, "missing >");
+            *s = '\0'; // temporarily null terminate tag name
+            if (ezxml_close_tag(root, d, s)) return &root->xml;
+            if (isspace(*s = q)) s += strspn(s, EZXML_WS);
+        }
+        else if (! strncmp(s, "!--", 3)) { // xml comment
+            if (! (s = strstr(s + 3, "--")) || (*(s += 2) != '>' && *s) ||
+                (! *s && e != '>')) return ezxml_err(root, d, "unclosed <!--");
+        }
+        else if (! strncmp(s, "![CDATA[", 8)) { // cdata
+            if ((s = strstr(s, "]]>")))
+                ezxml_char_content(root, d + 8, (s += 2) - d - 10, 'c');
+            else return ezxml_err(root, d, "unclosed <![CDATA[");
+        }
+        else if (! strncmp(s, "!DOCTYPE", 8)) { // dtd
+            for (l = 0; *s && ((! l && *s != '>') || (l && (*s != ']' || 
+                 *(s + strspn(s + 1, EZXML_WS) + 1) != '>')));
+                 l = (*s == '[') ? 1 : l) s += strcspn(s + 1, "[]>") + 1;
+            if (! *s && e != '>')
+                return ezxml_err(root, d, "unclosed <!DOCTYPE");
+            d = (l) ? strchr(d, '[') + 1 : d;
+            if (l && ! ezxml_internal_dtd(root, d, s++ - d)) return &root->xml;
+        }
+        else if (*s == '?') { // <?...?> processing instructions
+            do { s = strchr(s, '?'); } while (s && *(++s) && *s != '>');
+            if (! s || (! *s && e != '>')) 
+                return ezxml_err(root, d, "unclosed <?");
+            else ezxml_proc_inst(root, d + 1, s - d - 2);
+        }
+        else return ezxml_err(root, d, "unexpected <");
+        
+        if (! s || ! *s) break;
+        *s = '\0';
+        d = ++s;
+        if (*s && *s != '<') { // tag character content
+            while (*s && *s != '<') s++;
+            if (*s) ezxml_char_content(root, d, s - d, '&');
+            else break;
+        }
+        else if (! *s) break;
+    }
+
+    if (! root->cur) return &root->xml;
+    else if (! root->cur->name) return ezxml_err(root, d, "root tag missing");
+    else return ezxml_err(root, d, "unclosed tag <%s>", root->cur->name);
+}
+
+// free the memory allocated for the ezxml structure
+void ezxml_free(ezxml_t xml)
+{
+    ezxml_root_t root = (ezxml_root_t)xml;
+    int i, j;
+    char **a, *s;
+
+    if (! xml) return;
+    ezxml_free(xml->child);
+    ezxml_free(xml->ordered);
+
+    if (! xml->parent) { // free root tag allocations
+        for (i = 10; root->ent[i]; i += 2) // 0 - 9 are default entites (<>&"')
+            if ((s = root->ent[i + 1]) < root->s || s > root->e) free(s);
+        free(root->ent); // free list of general entities
+
+        for (i = 0; (a = root->attr[i]); i++) {
+            for (j = 1; a[j++]; j += 2) // free malloced attribute values
+                if (a[j] && (a[j] < root->s || a[j] > root->e)) free(a[j]);
+            free(a);
+        }
+        if (root->attr[0]) free(root->attr); // free default attribute list
+
+        for (i = 0; root->pi[i]; i++) {
+            for (j = 1; root->pi[i][j]; j++);
+            free(root->pi[i][j + 1]);
+            free(root->pi[i]);
+        }            
+        if (root->pi[0]) free(root->pi); // free processing instructions
+
+        if (root->len == -1) free(root->m); // malloced xml data
+        if (root->u) free(root->u); // utf8 conversion
+    }
+
+    ezxml_free_attr(xml->attr); // tag attributes
+    if ((xml->flags & EZXML_TXTM)) free(xml->txt); // character content
+    if ((xml->flags & EZXML_NAMEM)) free(xml->name); // tag name
+    free(xml);
+}
+
+// return parser error message or empty string if none
+const char *ezxml_error(ezxml_t xml)
+{
+    while (xml && xml->parent) xml = xml->parent; // find root tag
+    return (xml) ? ((ezxml_root_t)xml)->err : "";
+}
+
+// returns a new empty ezxml structure with the given root tag name
+ezxml_t ezxml_new(const char *name)
+{
+    static char *ent[] = { "lt;", "&#60;", "gt;", "&#62;", "quot;", "&#34;",
+                           "apos;", "&#39;", "amp;", "&#38;", NULL };
+    ezxml_root_t root = (ezxml_root_t)memset(malloc(sizeof(struct ezxml_root)), 
+                                             '\0', sizeof(struct ezxml_root));
+    root->xml.name = (char *)name;
+    root->cur = &root->xml;
+    strcpy(root->err, root->xml.txt = "");
+    root->ent = memcpy(malloc(sizeof(ent)), ent, sizeof(ent));
+    root->attr = root->pi = (char ***)(root->xml.attr = EZXML_NIL);
+    return &root->xml;
+}
+
+// inserts an existing tag into an ezxml structure
+ezxml_t ezxml_insert(ezxml_t xml, ezxml_t dest, size_t off)
+{
+    ezxml_t cur, prev, head;
+
+    xml->next = xml->sibling = xml->ordered = NULL;
+    xml->off = off;
+    xml->parent = dest;
+
+    if ((head = dest->child)) { // already have sub tags
+        if (head->off <= off) { // not first subtag
+            for (cur = head; cur->ordered && cur->ordered->off <= off;
+                 cur = cur->ordered);
+            xml->ordered = cur->ordered;
+            cur->ordered = xml;
+        }
+        else { // first subtag
+            xml->ordered = head;
+            dest->child = xml;
+        }
+
+        for (cur = head, prev = NULL; cur && strcmp(cur->name, xml->name);
+             prev = cur, cur = cur->sibling); // find tag type
+        if (cur && cur->off <= off) { // not first of type
+            while (cur->next && cur->next->off <= off) cur = cur->next;
+            xml->next = cur->next;
+            cur->next = xml;
+        }
+        else { // first tag of this type
+            if (prev && cur) prev->sibling = cur->sibling; // remove old first
+            xml->next = cur; // old first tag is now next
+            for (cur = head, prev = NULL; cur && cur->off <= off;
+                 prev = cur, cur = cur->sibling); // new sibling insert point
+            xml->sibling = cur;
+            if (prev) prev->sibling = xml;
+        }
+    }
+    else dest->child = xml; // only sub tag
+
+    return xml;
+}
+
+// Adds a child tag. off is the offset of the child tag relative to the start
+// of the parent tag's character content. Returns the child tag.
+ezxml_t ezxml_add_child(ezxml_t xml, const char *name, size_t off)
+{
+    ezxml_t child;
+
+    if (! xml) return NULL;
+    child = (ezxml_t)memset(malloc(sizeof(struct ezxml)), '\0',
+                            sizeof(struct ezxml));
+    child->name = (char *)name;
+    child->attr = EZXML_NIL;
+    child->txt = "";
+
+    return ezxml_insert(child, xml, off);
+}
+
+// sets a flag for the given tag and returns the tag
+ezxml_t ezxml_set_flag(ezxml_t xml, short flag)
+{
+    if (xml) xml->flags |= flag;
+    return xml;
+}
+
diff --git a/libezxml/src/ezxml.h b/libezxml/src/ezxml.h
new file mode 100644
index 0000000..edd0810
--- /dev/null
+++ b/libezxml/src/ezxml.h
@@ -0,0 +1,128 @@
+/* ezxml.h
+ *
+ * Copyright 2004-2006 Aaron Voisine <aaron@voisine.org>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining
+ * a copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sublicense, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+ * IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+ * CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+ * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+ * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#ifndef _EZXML_H
+#define _EZXML_H
+
+#include <stdlib.h>
+#include <stdio.h>
+#include <stdarg.h>
+#include <fcntl.h>
+
+#define EZXML_BUFSIZE 1024 // size of internal memory buffers
+#define EZXML_NAMEM   0x80 // name is malloced
+#define EZXML_TXTM    0x40 // txt is malloced
+#define EZXML_DUP     0x20 // attribute name and value are strduped
+
+typedef struct ezxml *ezxml_t;
+struct ezxml {
+    char *name;      // tag name
+    char **attr;     // tag attributes { name, value, name, value, ... NULL }
+    char *txt;       // tag character content, empty string if none
+    size_t off;      // tag offset from start of parent tag character content
+    ezxml_t next;    // next tag with same name in this section at this depth
+    ezxml_t sibling; // next tag with different name in same section and depth
+    ezxml_t ordered; // next tag, same section and depth, in original order
+    ezxml_t child;   // head of sub tag list, NULL if none
+    ezxml_t parent;  // parent tag, NULL if current tag is root tag
+    short flags;     // additional information
+};
+
+// Given a string of xml data and its length, parses it and creates an ezxml
+// structure. For efficiency, modifies the data by adding null terminators
+// and decoding ampersand sequences. If you don't want this, copy the data and
+// pass in the copy. Returns NULL on failure.
+ezxml_t ezxml_parse_str(char *s, size_t len);
+
+// returns the first child tag (one level deeper) with the given name or NULL
+// if not found
+ezxml_t ezxml_child(ezxml_t xml, const char *name);
+
+// returns the next tag of the same name in the same section and depth or NULL
+// if not found
+#define ezxml_next(xml) ((xml) ? xml->next : NULL)
+
+// Returns the Nth tag with the same name in the same section at the same depth
+// or NULL if not found. An index of 0 returns the tag given.
+ezxml_t ezxml_idx(ezxml_t xml, int idx);
+
+// returns the name of the given tag
+#define ezxml_name(xml) ((xml) ? xml->name : NULL)
+
+// returns the given tag's character content or empty string if none
+#define ezxml_txt(xml) ((xml) ? xml->txt : "")
+
+// returns the value of the requested tag attribute, or NULL if not found
+const char *ezxml_attr(ezxml_t xml, const char *attr);
+
+// Traverses the ezxml sturcture to retrieve a specific subtag. Takes a
+// variable length list of tag names and indexes. The argument list must be
+// terminated by either an index of -1 or an empty string tag name. Example: 
+// title = ezxml_get(library, "shelf", 0, "book", 2, "title", -1);
+// This retrieves the title of the 3rd book on the 1st shelf of library.
+// Returns NULL if not found.
+ezxml_t ezxml_get(ezxml_t xml, ...);
+
+// frees the memory allocated for an ezxml structure
+void ezxml_free(ezxml_t xml);
+    
+// returns parser error message or empty string if none
+const char *ezxml_error(ezxml_t xml);
+
+// returns a new empty ezxml structure with the given root tag name
+ezxml_t ezxml_new(const char *name);
+
+// wrapper for ezxml_new() that strdup()s name
+#define ezxml_new_d(name) ezxml_set_flag(ezxml_new(strdup(name)), EZXML_NAMEM)
+
+// Adds a child tag. off is the offset of the child tag relative to the start
+// of the parent tag's character content. Returns the child tag.
+ezxml_t ezxml_add_child(ezxml_t xml, const char *name, size_t off);
+
+// wrapper for ezxml_add_child() that strdup()s name
+#define ezxml_add_child_d(xml, name, off) \
+    ezxml_set_flag(ezxml_add_child(xml, strdup(name), off), EZXML_NAMEM)
+
+// sets the character content for the given tag and returns the tag
+ezxml_t ezxml_set_txt(ezxml_t xml, const char *txt);
+
+// wrapper for ezxml_set_txt() that strdup()s txt
+#define ezxml_set_txt_d(xml, txt) \
+    ezxml_set_flag(ezxml_set_txt(xml, strdup(txt)), EZXML_TXTM)
+
+// Sets the given tag attribute or adds a new attribute if not found. A value
+// of NULL will remove the specified attribute. Returns the tag given.
+ezxml_t ezxml_set_attr(ezxml_t xml, const char *name, const char *value);
+
+// Wrapper for ezxml_set_attr() that strdup()s name/value. Value cannot be NULL
+#define ezxml_set_attr_d(xml, name, value) \
+    ezxml_set_attr(ezxml_set_flag(xml, EZXML_DUP), strdup(name), strdup(value))
+
+// sets a flag for the given tag and returns the tag
+ezxml_t ezxml_set_flag(ezxml_t xml, short flag);
+
+// inserts an existing tag into an ezxml structure
+ezxml_t ezxml_insert(ezxml_t xml, ezxml_t dest, size_t off);
+
+#endif // _EZXML_H