@@ -1,5 +1,5 @@
1#include "hexml.h"
1#include "hexml.h"
2−#include <malloc.h>
2+#include <stdlib.h>
3#include <string.h>
3#include <string.h>
4#include <assert.h>
4#include <assert.h>
5#include <ctype.h>
5#include <ctype.h>
@@ -10,18 +10,18 @@ typedef int bool;
10/////////////////////////////////////////////////////////////////////
10/////////////////////////////////////////////////////////////////////
13−int static inline end(str s) { return s.start + s.length; }
13+static inline int end(str s) { return s.start + s.length; }
15−str static inline start_length(int32_t start, int32_t length)
15+static inline str start_length(int32_t start, int32_t length)
17− if (start < 0 || length < 0) assert(0);
17+ assert(start >= 0 && length >= 0);
19 res.start = start;
19 res.start = start;
20 res.length = length;
20 res.length = length;
21 return res;
21 return res;
24−str static inline start_end(int32_t start, int32_t end)
24+static inline str start_end(int32_t start, int32_t end)
26 return start_length(start, end - start);
26 return start_length(start, end - start);
@@ -40,7 +40,7 @@ typedef struct
40 // have a cursor at the end, which is stack scoped, so children write out, then I do
40 // have a cursor at the end, which is stack scoped, so children write out, then I do
41 // when you commit, you copy over from end to front
41 // when you commit, you copy over from end to front
43− // nodes
43+ // nodes
45 int used_front; // front entries, stored for good
45 int used_front; // front entries, stored for good
46 int used_back; // back entries, stack based, copied into front
46 int used_back; // back entries, stack based, copied into front
@@ -51,11 +51,11 @@ typedef struct
52struct document
52struct document
54− char* body; // pointer to initial argument, not owned by us
54+ const char* body; // pointer to initial argument, not owned by us
56 // things only used while parsing
56 // things only used while parsing
57− char* cursor; // pointer to where we are in body
57+ const char* cursor; // pointer to where we are in body
58− char* end; // pointer to one past the last char
58+ const char* end; // pointer to one past the last char
59 // if cursor is > end we have gone past the end
59 // if cursor is > end we have gone past the end
61 char* error_message;
61 char* error_message;
@@ -63,8 +63,8 @@ struct document
63 attr_buffer attrs;
63 attr_buffer attrs;
66−int static inline doc_length(document* d) { return (int) (d->end - d->body); }
66+static inline int doc_length(const document* d) { return (int) (d->end - d->body); }
67−int static inline doc_position(document* d) { return (int) (d->cursor - d->body); }
67+static inline int doc_position(const document* d) { return (int) (d->cursor - d->body); }
70/////////////////////////////////////////////////////////////////////
70/////////////////////////////////////////////////////////////////////
@@ -72,49 +72,45 @@ int static inline doc_position(document* d) { return (int) (d->cursor - d->body)
73typedef struct
73typedef struct
75− document* d;
75+ const document* d;
76 char* buffer;
76 char* buffer;
77 int length;
77 int length;
78 int cursor;
78 int cursor;
81−void static inline render_char(render* r, char c)
81+static inline void render_char(render* r, char c)
83 if (r->cursor < r->length)
83 if (r->cursor < r->length)
84 r->buffer[r->cursor] = c;
84 r->buffer[r->cursor] = c;
85 r->cursor++;
85 r->cursor++;
88−void static inline bound(char* msg, int index, int mn, int mx)
88+static inline void bound(int idx, int mn, int mx)
90− if (index < mn || index > mx)
92− //printf("Bounds checking failed %s, got %i, which should be in %i:%i\n", msg, index, mn, mx);
93− assert(0);
90+ assert(idx >= mn && idx <= mx);
97−void static inline bound_str(char* msg, str s, int mn, int mx)
93+static inline void bound_str(str s, int mn, int mx)
99− if (s.length < 0) assert(0);
95+ assert(s.length >= 0);
100− bound(msg, s.start, mn, mx);
96+ bound(s.start, mn, mx);
101− bound(msg, end(s), mn, mx);
97+ bound(end(s), mn, mx);
104−void render_str(render* r, str s)
100+static void render_str(render* r, str s)
106− bound_str("render_str", s, 0, doc_length(r->d));
102+ bound_str(s, 0, doc_length(r->d));
107 for (int i = 0; i < s.length; i++)
103 for (int i = 0; i < s.length; i++)
108 render_char(r, r->d->body[s.start + i]);
104 render_char(r, r->d->body[s.start + i]);
111−void render_tag(render* r, node* n);
107+static void render_tag(render* r, const node* n);
113−void render_content(render* r, node* n)
109+static void render_content(render* r, const node* n)
115− bound_str("render_conent inner", n->inner, 0, doc_length(r->d));
111+ bound_str(n->inner, 0, doc_length(r->d));
116− bound_str("render_conent nodes", n->nodes, 0, r->d->nodes.used_front);
112+ bound_str(n->nodes, 0, r->d->nodes.used_front);
117− bound_str("render_conent attrs", n->attrs, 0, r->d->attrs.used);
113+ bound_str(n->attrs, 0, r->d->attrs.used);
119 int done = n->inner.start;
115 int done = n->inner.start;
120 for (int i = 0; i < n->nodes.length; i++)
116 for (int i = 0; i < n->nodes.length; i++)
@@ -127,7 +123,7 @@ void render_content(render* r, node* n)
127 render_str(r, start_end(done, end(n->inner)));
123 render_str(r, start_end(done, end(n->inner)));
130−void render_tag(render* r, node* n)
126+static void render_tag(render* r, const node* n)
132 render_char(r, '<');
128 render_char(r, '<');
133 render_str(r, n->name);
129 render_str(r, n->name);
@@ -149,7 +145,7 @@ void render_tag(render* r, node* n)
149 render_char(r, '>');
145 render_char(r, '>');
152−int node_render(document* d, node* n, char* buffer, int length)
148+int hexml_node_render(const document* d, const node* n, char* buffer, int length)
154 render r;
150 render r;
@@ -168,10 +164,10 @@ int node_render(document* d, node* n, char* buffer, int length)
168/////////////////////////////////////////////////////////////////////
164/////////////////////////////////////////////////////////////////////
169// NOT THE PARSER
165// NOT THE PARSER
171−char* document_error(document* d){return d->error_message;}
167+char* hexml_document_error(const document* d){return d->error_message;}
172−node* document_node(document* d){return &d->nodes.nodes[0];}
168+node* hexml_document_node(const document* d){return &d->nodes.nodes[0];}
174−void document_free(document* d)
170+void hexml_document_free(document* d)
176 free(d->error_message);
172 free(d->error_message);
177 free(d->nodes.alloc);
173 free(d->nodes.alloc);
@@ -179,20 +175,20 @@ void document_free(document* d)
182−node* node_children(document* d, node* n, int* res)
178+node* hexml_node_children(const document* d, const node* n, int* res)
184 *res = n->nodes.length;
180 *res = n->nodes.length;
185 return &d->nodes.nodes[n->nodes.start];
181 return &d->nodes.nodes[n->nodes.start];
188−attr* node_attributes(document* d, node* n, int* res)
184+attr* hexml_node_attributes(const document* d, const node* n, int* res)
190 *res = n->attrs.length;
186 *res = n->attrs.length;
191 return &d->attrs.attrs[n->attrs.start];
187 return &d->attrs.attrs[n->attrs.start];
195−attr* node_attributeBy(document* d, node* n, char* s, int slen)
191+attr* hexml_node_attribute(const document* d, const node* n, const char* s, int slen)
197 if (slen == -1) slen = (int) strlen(s);
193 if (slen == -1) slen = (int) strlen(s);
198 const int limit = end(n->attrs);
194 const int limit = end(n->attrs);
@@ -206,7 +202,7 @@ attr* node_attributeBy(document* d, node* n, char* s, int slen)
208// Search for given strings within a node
204// Search for given strings within a node
209−node* node_childBy(document* d, node* parent, node* prev, char* s, int slen)
205+node* hexml_node_child(const document* d, const node* parent, const node* prev, const char* s, int slen)
211 if (slen == -1) slen = (int) strlen(s);
207 if (slen == -1) slen = (int) strlen(s);
212 int i = prev == NULL ? parent->nodes.start : (int) (prev + 1 - d->nodes.nodes);
208 int i = prev == NULL ? parent->nodes.start : (int) (prev + 1 - d->nodes.nodes);
@@ -224,49 +220,49 @@ node* node_childBy(document* d, node* parent, node* prev, char* s, int slen)
224/////////////////////////////////////////////////////////////////////
220/////////////////////////////////////////////////////////////////////
225// PARSE TABLE
221// PARSE TABLE
227−const char tag_name1 = 0x1;
223+static const char tag_name1 = 0x1;
228−const char tag_name = 0x2;
224+static const char tag_name = 0x2;
229−const char tag_space = 0x4;
225+static const char tag_space = 0x4;
231−char parse_table[256];
227+static char parse_table[256];
233−void init_parse_table()
229+static void init_parse_table()
235 static bool done = 0;
231 static bool done = 0;
236 if (done) return;
232 if (done) return;
237 for (int i = 0; i < 256; i++)
233 for (int i = 0; i < 256; i++)
239 bool name1 = i == ':' || i == '_' || (i >= 'a' && i <= 'z') || (i >= 'A' && i <= 'Z');
235 bool name1 = i == ':' || i == '_' || (i >= 'a' && i <= 'z') || (i >= 'A' && i <= 'Z');
240− bool name = name1 || i == '-' || (i >= '0' && i <= '9');
236+ bool name = name1 || i == '-' || i == '.' || (i >= '0' && i <= '9');
241 bool space = i == ' ' || i == '\t' || i == '\r' || i == '\n';
237 bool space = i == ' ' || i == '\t' || i == '\r' || i == '\n';
242 parse_table[i] = (name1 ? tag_name1 : 0) | (name ? tag_name : 0) | (space ? tag_space : 0);
238 parse_table[i] = (name1 ? tag_name1 : 0) | (name ? tag_name : 0) | (space ? tag_space : 0);
244 done = 1;
240 done = 1;
247−bool static inline is(char c, char tag) { return parse_table[c] & tag; }
243+static inline bool is(char c, char tag) { return parse_table[(unsigned char) c] & tag; }
248−bool static inline is_name1(char c) { return is(c, tag_name1); }
244+static inline bool is_name1(char c) { return is(c, tag_name1); }
249−bool static inline is_name(char c) { return is(c, tag_name); }
245+static inline bool is_name(char c) { return is(c, tag_name); }
250−bool static inline is_space(char c) { return is(c, tag_space); }
246+static inline bool is_space(char c) { return is(c, tag_space); }
253/////////////////////////////////////////////////////////////////////
249/////////////////////////////////////////////////////////////////////
254// PARSER COMBINATORS
250// PARSER COMBINATORS
256−char static inline peekAt(document* d, int i) { return d->cursor[i]; }
252+static inline char peek_at(const document* d, int i) { return d->cursor[i]; }
253+static inline char peek(const document* d) { return peek_at(d, 0); }
257−void static inline skip(document* d, int i) { d->cursor += i; }
254+static inline void skip(document* d, int i) { d->cursor += i; }
258−char static inline peek(document* d) { return peekAt(d, 0); }
259−char static inline get(document* d) { char c = peek(d); skip(d, 1); return c; }
255+static inline char get(document* d) { char c = peek(d); skip(d, 1); return c; }
261// Remove whitespace characters from the cursor while they are still whitespace
257// Remove whitespace characters from the cursor while they are still whitespace
262−void static inline trim(document* d)
258+static inline void trim(document* d)
264 while (is_space(peek(d)))
260 while (is_space(peek(d)))
265 skip(d, 1);
261 skip(d, 1);
268// Find this character form the cursor onwards, if true adjust the cursor to that char, otherwise leave it at the end
264// Find this character form the cursor onwards, if true adjust the cursor to that char, otherwise leave it at the end
269−bool find(document* d, char c)
265+static bool find(document* d, char c)
271 char* x = memchr(d->cursor, c, d->end - d->cursor);
267 char* x = memchr(d->cursor, c, d->end - d->cursor);
272 if (x == NULL)
268 if (x == NULL)
@@ -285,7 +281,7 @@ bool find(document* d, char c)
285/////////////////////////////////////////////////////////////////////
281/////////////////////////////////////////////////////////////////////
286// PARSING CODE
282// PARSING CODE
288−void static inline node_alloc(node_buffer* b, int ask)
284+static inline void node_alloc(node_buffer* b, int ask)
290 int space = b->size - b->used_back - b->used_front;
286 int space = b->size - b->used_back - b->used_front;
291 if (space >= ask) return;
287 if (space >= ask) return;
@@ -300,7 +296,7 @@ void static inline node_alloc(node_buffer* b, int ask)
300 b->alloc = buf2;
296 b->alloc = buf2;
303−void static inline attr_alloc(attr_buffer* b, int ask)
299+static inline void attr_alloc(attr_buffer* b, int ask)
305 int space = b->size - b->used;
301 int space = b->size - b->used;
306 if (space >= ask) return;
302 if (space >= ask) return;
@@ -314,7 +310,7 @@ void static inline attr_alloc(attr_buffer* b, int ask)
314 b->alloc = buf2;
310 b->alloc = buf2;
317−void set_error(document* d, char* msg)
313+static void set_error(document* d, const char* msg)
319 if (d->error_message != NULL) return; // keep the first error message
315 if (d->error_message != NULL) return; // keep the first error message
320 d->error_message = malloc(strlen(msg)+1);
316 d->error_message = malloc(strlen(msg)+1);
@@ -324,7 +320,7 @@ void set_error(document* d, char* msg)
325// you now expect a name, perhaps preceeded by whitespace
321// you now expect a name, perhaps preceeded by whitespace
326// the name may be empty
322// the name may be empty
327−str static inline parse_name(document* d)
323+static inline str parse_name(document* d)
329 int start = doc_position(d);
325 int start = doc_position(d);
330 if (!is_name1(peek(d)))
326 if (!is_name1(peek(d)))
@@ -335,14 +331,15 @@ str static inline parse_name(document* d)
335 return start_end(start, doc_position(d));
331 return start_end(start, doc_position(d));
338−str static inline parse_attrval(document* d)
334+static inline str parse_attrval(document* d)
341− if (get(d) != '=')
337+ if (peek(d) != '=')
343 set_error(d, "Expected = in attribute, but missing");
339 set_error(d, "Expected = in attribute, but missing");
344 return start_length(0, 0);
340 return start_length(0, 0);
347 char c = peek(d);
344 char c = peek(d);
348 if (c == '\"' || c == '\'')
345 if (c == '\"' || c == '\'')
@@ -368,10 +365,10 @@ str static inline parse_attrval(document* d)
369// seen a tag name, now looking for attributes terminated by >
366// seen a tag name, now looking for attributes terminated by >
370// puts the attributes it finds in the attribute buffer
367// puts the attributes it finds in the attribute buffer
371−str static inline parse_attributes(document* d)
368+static inline str parse_attributes(document* d)
373 int start = d->attrs.used;
370 int start = d->attrs.used;
374− for (int i = 0; ; i++)
371+ while (d->error_message == NULL)
377 str name = parse_name(d);
374 str name = parse_name(d);
@@ -384,11 +381,11 @@ str static inline parse_attributes(document* d)
384 return start_end(start, d->attrs.used);
381 return start_end(start, d->attrs.used);
387−str parse_content(document* d);
384+static str parse_content(document* d);
390// Add a new entry into tag, am at a '<'
387// Add a new entry into tag, am at a '<'
391−void static inline parse_tag(document* d)
388+static inline void parse_tag(document* d)
393 node_alloc(&d->nodes, 1);
390 node_alloc(&d->nodes, 1);
394 d->nodes.used_back++;
391 d->nodes.used_back++;
@@ -399,12 +396,18 @@ void static inline parse_tag(document* d)
399 assert(c == '<');
396 assert(c == '<');
400 if (peek(d) == '?') skip(d, 1);
397 if (peek(d) == '?') skip(d, 1);
401 me->name = parse_name(d);
398 me->name = parse_name(d);
399+ if (me->name.length == 0)
401+ set_error(d, "Missing tag name");
402 me->attrs = parse_attributes(d);
404 me->attrs = parse_attributes(d);
405+ if (d->error_message != NULL) return;
404− c = get(d);
407+ c = peek(d);
405− if ((c == '/' || c == '?') && peek(d) == '>')
408+ if ((c == '/' || c == '?') && peek_at(d, 1) == '>')
407− skip(d, 1);
410+ skip(d, 2);
408 me->nodes = start_length(0, 0);
411 me->nodes = start_length(0, 0);
409 me->outer.length = start_end(me->outer.start, doc_position(d)).length;
412 me->outer.length = start_end(me->outer.start, doc_position(d)).length;
410 me->inner = start_length(doc_position(d), 0);
413 me->inner = start_length(doc_position(d), 0);
@@ -415,6 +418,8 @@ void static inline parse_tag(document* d)
415 set_error(d, "Gunk at the end of the tag");
418 set_error(d, "Gunk at the end of the tag");
418 me->inner.start = doc_position(d);
423 me->inner.start = doc_position(d);
419 str content = parse_content(d);
424 str content = parse_content(d);
@@ -424,7 +429,7 @@ void static inline parse_tag(document* d)
424 me->inner.length = start_end(me->inner.start, doc_position(d)).length;
429 me->inner.length = start_end(me->inner.start, doc_position(d)).length;
426 if (d->error_message != NULL) return;
431 if (d->error_message != NULL) return;
427− if (peek(d) == '<' && peekAt(d, 1) == '/')
432+ if (peek(d) == '<' && peek_at(d, 1) == '/')
429 skip(d, 2);
434 skip(d, 2);
430 if (d->end - d->cursor >= me->name.length &&
435 if (d->end - d->cursor >= me->name.length &&
@@ -446,7 +451,7 @@ void static inline parse_tag(document* d)
447// Parser until </, return the index of your node children
452// Parser until </, return the index of your node children
448// Not inline as it is recursive
453// Not inline as it is recursive
449−str parse_content(document* d)
454+static str parse_content(document* d)
451 int before = d->nodes.used_back;
456 int before = d->nodes.used_back;
452 while (d->error_message == NULL)
457 while (d->error_message == NULL)
@@ -456,22 +461,22 @@ str parse_content(document* d)
459− else if (peekAt(d, 1) == '/')
464+ else if (peek_at(d, 1) == '/')
461 // have found a </
466 // have found a </
464− else if (peekAt(d, 1) == '!' && peekAt(d, 2) == '-' && peekAt(d, 3) == '-')
469+ else if (peek_at(d, 1) == '!' && peek_at(d, 2) == '-' && peek_at(d, 3) == '-')
466 skip(d, 3);
471 skip(d, 3);
467 // you can't reuse the two '-' characters for the closing as well
472 // you can't reuse the two '-' characters for the closing as well
468− if (peekAt(d, 0) == '\0' || peekAt(d, 1) == '\0')
473+ if (peek_at(d, 0) == '\0' || peek_at(d, 1) == '\0')
470 set_error(d, "Didn't get a closing comment");
475 set_error(d, "Didn't get a closing comment");
471 return start_end(0, 0);
476 return start_end(0, 0);
473 skip(d, 2);
478 skip(d, 2);
474− while (1)
479+ for (;;)
476 if (!find(d, '>'))
481 if (!find(d, '>'))
@@ -479,7 +484,7 @@ str parse_content(document* d)
479 return start_end(0, 0);
484 return start_end(0, 0);
481 skip(d, 1);
486 skip(d, 1);
482− if (peekAt(d, -3) == '-' && peekAt(d, -2))
487+ if (peek_at(d, -3) == '-' && peek_at(d, -2) == '-')
@@ -507,7 +512,7 @@ typedef struct
507 node nodes[500];
512 node nodes[500];
510−document* document_parse(char* s, int slen)
515+document* hexml_document_parse(const char* s, int slen)
512 if (slen == -1) slen = (int) strlen(s);
517 if (slen == -1) slen = (int) strlen(s);
513 assert(s[slen] == 0);
518 assert(s[slen] == 0);
@@ -534,15 +539,13 @@ document* document_parse(char* s, int slen)
534 d->nodes.nodes[0].outer = start_length(0, slen);
539 d->nodes.nodes[0].outer = start_length(0, slen);
535 d->nodes.nodes[0].inner = start_length(0, slen);
540 d->nodes.nodes[0].inner = start_length(0, slen);
536 d->nodes.nodes[0].attrs = start_length(0, 0);
541 d->nodes.nodes[0].attrs = start_length(0, 0);
538 // Introduce an intermediate result, otherwise behaviour is undefined
543 // Introduce an intermediate result, otherwise behaviour is undefined
539 // because there is no guaranteed ordering between LHS and RHS evaluation
544 // because there is no guaranteed ordering between LHS and RHS evaluation
540 str content = parse_content(d);
545 str content = parse_content(d);
541 d->nodes.nodes[0].nodes = content;
546 d->nodes.nodes[0].nodes = content;
543− if (d->cursor < d->end && d->error_message == NULL)
548+ if (d->cursor < d->end)
545 set_error(d, "Trailing junk at the end of the document");
549 set_error(d, "Trailing junk at the end of the document");
547 return d;
550 return d;