1 /* gmarkup.c - Simple XML-like parser
3 * Copyright 2000 Red Hat, Inc.
5 * GLib is free software; you can redistribute it and/or modify it
6 * under the terms of the GNU Lesser General Public License as
7 * published by the Free Software Foundation; either version 2 of the
8 * License, or (at your option) any later version.
10 * GLib is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 * Lesser General Public License for more details.
15 * You should have received a copy of the GNU Lesser General Public
16 * License along with GLib; see the file COPYING.LIB. If not,
17 * write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
18 * Boston, MA 02111-1307, USA.
32 g_markup_error_quark ()
34 static GQuark error_quark = 0;
37 error_quark = g_quark_from_static_string ("g-markup-error-quark");
42 typedef struct _GMarkupAttribute GMarkupAttribute;
44 struct _GMarkupAttribute
50 static GMarkupAttribute*
51 attribute_new (const gchar *name, const gchar *value)
53 GMarkupAttribute *attr;
55 attr = g_new (GMarkupAttribute, 1);
57 /* name/value are allowed to be NULL */
58 attr->name = g_strdup (name);
59 attr->value = g_strdup (value);
65 attribute_free (GMarkupAttribute *attr)
75 STATE_AFTER_OPEN_ANGLE,
76 STATE_AFTER_CLOSE_ANGLE,
77 STATE_AFTER_ELISION_SLASH, /* the slash that obviates need for end element */
78 STATE_INSIDE_OPEN_TAG_NAME,
79 STATE_INSIDE_ATTRIBUTE_NAME,
80 STATE_BETWEEN_ATTRIBUTES,
81 STATE_AFTER_ATTRIBUTE_EQUALS_SIGN,
82 STATE_INSIDE_ATTRIBUTE_VALUE,
84 STATE_AFTER_CLOSE_TAG_SLASH,
85 STATE_INSIDE_CLOSE_TAG_NAME,
86 STATE_INSIDE_PASSTHROUGH,
90 struct _GMarkupParseContext
92 const GMarkupParser *parser;
94 GMarkupParseFlags flags;
100 GDestroyNotify dnotify;
102 /* A piece of character data or an element that
103 * hasn't "ended" yet so we haven't yet called
104 * the callback for it.
106 GString *partial_chunk;
108 GMarkupParseState state;
112 const gchar *current_text;
113 gint current_text_len;
114 const gchar *current_text_end;
116 GString *leftover_char_portion;
118 /* used to save the start of the last interesting thingy */
123 guint document_empty : 1;
127 GMarkupParseContext *
128 g_markup_parse_context_new (const GMarkupParser *parser,
129 GMarkupParseFlags flags,
131 GDestroyNotify user_data_dnotify)
133 GMarkupParseContext *context;
135 g_return_val_if_fail (parser != NULL, NULL);
137 context = g_new (GMarkupParseContext, 1);
139 context->parser = parser;
140 context->flags = flags;
141 context->user_data = user_data;
142 context->dnotify = user_data_dnotify;
144 context->line_number = 1;
145 context->char_number = 1;
147 context->partial_chunk = NULL;
149 context->state = STATE_START;
150 context->tag_stack = NULL;
151 context->attributes = NULL;
153 context->current_text = NULL;
154 context->current_text_len = -1;
155 context->current_text_end = NULL;
156 context->leftover_char_portion = NULL;
158 context->start = NULL;
159 context->iter = NULL;
161 context->document_empty = TRUE;
162 context->parsing = FALSE;
168 g_markup_parse_context_free (GMarkupParseContext *context)
170 g_return_if_fail (context != NULL);
171 g_return_if_fail (!context->parsing);
173 if (context->dnotify)
174 (* context->dnotify) (context->user_data);
176 g_slist_foreach (context->attributes, (GFunc)attribute_free, NULL);
177 g_slist_free (context->attributes);
179 g_slist_foreach (context->tag_stack, (GFunc)g_free, NULL);
180 g_slist_free (context->tag_stack);
182 if (context->partial_chunk)
183 g_string_free (context->partial_chunk, TRUE);
185 if (context->leftover_char_portion)
186 g_string_free (context->leftover_char_portion, TRUE);
192 attribute_list_to_arrays (GSList *attributes,
203 len = g_slist_length (attributes);
207 names = g_new (gchar*, len + 1);
215 values = g_new (gchar*, len + 1);
221 /* We want to reverse the list, since it's
222 * backward from the order the attributes appeared
226 tmp_list = attributes;
229 GMarkupAttribute *attr = tmp_list->data;
234 names[i] = g_strdup (attr->name);
237 values[i] = g_strdup (attr->value);
239 tmp_list = g_slist_next (tmp_list);
254 mark_error (GMarkupParseContext *context,
257 context->state = STATE_ERROR;
259 if (context->parser->error)
260 (*context->parser->error) (context, error, context->user_data);
264 set_error (GMarkupParseContext *context,
266 GMarkupErrorType code,
274 va_start (args, format);
275 s = g_strdup_vprintf (format, args);
278 tmp_error = g_error_new (G_MARKUP_ERROR,
280 _("Error on line %d char %d: %s"),
281 context->line_number,
282 context->char_number,
287 mark_error (context, tmp_error);
289 g_propagate_error (error, tmp_error);
293 is_name_start_char (gunichar c)
295 if (g_unichar_isalpha (c) ||
304 is_name_char (gunichar c)
306 if (g_unichar_isalnum (c) ||
318 char_str (gunichar c,
322 g_unichar_to_utf8 (c, buf);
327 utf8_str (const gchar *utf8,
330 char_str (g_utf8_get_char (utf8), buf);
335 set_unescape_error (GMarkupParseContext *context,
337 const gchar *remaining_text,
338 const gchar *remaining_text_end,
339 GMarkupErrorType code,
346 gint remaining_newlines;
349 remaining_newlines = 0;
351 while (p != remaining_text_end)
354 ++remaining_newlines;
358 va_start (args, format);
359 s = g_strdup_vprintf (format, args);
362 tmp_error = g_error_new (G_MARKUP_ERROR,
364 _("Error on line %d: %s"),
365 context->line_number - remaining_newlines,
370 mark_error (context, tmp_error);
372 g_propagate_error (error, tmp_error);
378 USTATE_AFTER_AMPERSAND,
379 USTATE_INSIDE_ENTITY_NAME,
380 USTATE_AFTER_CHARREF_HASH
384 unescape_text (GMarkupParseContext *context,
386 const gchar *text_end,
390 #define MAX_ENT_LEN 5
396 str = g_string_new ("");
398 state = USTATE_INSIDE_TEXT;
401 while (p != text_end && context->state != STATE_ERROR)
403 g_assert (p < text_end);
407 case USTATE_INSIDE_TEXT:
409 while (p != text_end && *p != '&')
410 p = g_utf8_next_char (p);
414 g_string_append_len (str, start, p - start);
419 if (p != text_end && *p == '&')
421 p = g_utf8_next_char (p);
422 state = USTATE_AFTER_AMPERSAND;
427 case USTATE_AFTER_AMPERSAND:
431 p = g_utf8_next_char (p);
434 state = USTATE_AFTER_CHARREF_HASH;
436 else if (!is_name_start_char (g_utf8_get_char (p)))
440 set_unescape_error (context, error,
442 G_MARKUP_ERROR_PARSE,
443 _("Empty entity '&;' seen; valid "
444 "entities are: & " < > '"));
450 set_unescape_error (context, error,
452 G_MARKUP_ERROR_PARSE,
453 _("Character '%s' is not valid at "
454 "the start of an entity name; "
455 "the & character begins an entity; "
456 "if this ampersand isn't supposed "
457 "to be an entity, escape it as "
465 state = USTATE_INSIDE_ENTITY_NAME;
471 case USTATE_INSIDE_ENTITY_NAME:
473 gchar buf[MAX_ENT_LEN+1] = {
474 '\0', '\0', '\0', '\0', '\0', '\0'
478 while (p != text_end)
482 else if (!is_name_char (*p))
486 set_unescape_error (context, error,
488 G_MARKUP_ERROR_PARSE,
489 _("Character '%s' is not valid "
490 "inside an entity name"),
495 p = g_utf8_next_char (p);
498 if (context->state != STATE_ERROR)
513 /* move to after semicolon */
514 p = g_utf8_next_char (p);
516 state = USTATE_INSIDE_TEXT;
518 if (strcmp (buf, "lt") == 0)
519 g_string_append_c (str, '<');
520 else if (strcmp (buf, "gt") == 0)
521 g_string_append_c (str, '>');
522 else if (strcmp (buf, "amp") == 0)
523 g_string_append_c (str, '&');
524 else if (strcmp (buf, "quot") == 0)
525 g_string_append_c (str, '"');
526 else if (strcmp (buf, "apos") == 0)
527 g_string_append_c (str, '\'');
530 set_unescape_error (context, error,
532 G_MARKUP_ERROR_PARSE,
533 _("Entity name '%s' is not known"),
539 set_unescape_error (context, error,
540 /* give line number of the & */
542 G_MARKUP_ERROR_PARSE,
543 _("Entity did not end with a semicolon; "
544 "most likely you used an ampersand "
545 "character without intending to start "
546 "an entity - escape ampersand as &"));
552 case USTATE_AFTER_CHARREF_HASH:
554 gboolean is_hex = FALSE;
558 p = g_utf8_next_char (p);
562 while (p != text_end && *p != ';')
563 p = g_utf8_next_char (p);
567 g_assert (*p == ';');
569 /* digit is between start and p */
573 gchar *digit = g_strndup (start, p - start);
579 l = strtoul (digit, &end, 16);
581 l = strtoul (digit, &end, 10);
585 set_unescape_error (context, error,
587 G_MARKUP_ERROR_PARSE,
588 _("Failed to parse '%s', which "
589 "should have been a digit "
590 "inside a character reference "
591 "(ê for example) - perhaps "
592 "the digit is too large"),
597 /* characters XML permits */
601 (l >= 0x20 && l <= 0xD7FF) ||
602 (l >= 0xE000 && l <= 0xFFFD) ||
603 (l >= 0x10000 && l <= 0x10FFFF))
606 g_string_append (str,
611 set_unescape_error (context, error,
613 G_MARKUP_ERROR_PARSE,
614 _("Character reference '%s' does not encode a permitted character"),
621 /* Move to next state */
622 p = g_utf8_next_char (p); /* past semicolon */
624 state = USTATE_INSIDE_TEXT;
628 set_unescape_error (context, error,
630 G_MARKUP_ERROR_PARSE,
631 _("Empty character reference; "
632 "should include a digit such as "
638 set_unescape_error (context, error,
640 G_MARKUP_ERROR_PARSE,
641 _("Character reference did not end with a "
643 "most likely you used an ampersand "
644 "character without intending to start "
645 "an entity - escape ampersand as &"));
651 g_assert_not_reached ();
656 /* If no errors, we should have returned to USTATE_INSIDE_TEXT */
657 g_assert (context->state == STATE_ERROR ||
658 state == USTATE_INSIDE_TEXT);
660 if (context->state == STATE_ERROR)
662 g_string_free (str, TRUE);
668 *unescaped = g_string_free (str, FALSE);
676 advance_char (GMarkupParseContext *context)
679 context->iter = g_utf8_next_char (context->iter);
680 context->char_number += 1;
681 if (*context->iter == '\n')
683 context->line_number += 1;
684 context->char_number = 1;
687 return context->iter != context->current_text_end;
691 skip_spaces (GMarkupParseContext *context)
695 if (!g_unichar_isspace (g_utf8_get_char (context->iter)))
698 while (advance_char (context));
702 advance_to_name_end (GMarkupParseContext *context)
706 if (!is_name_char (g_utf8_get_char (context->iter)))
709 while (advance_char (context));
713 add_to_partial (GMarkupParseContext *context,
714 const gchar *text_start,
715 const gchar *text_end)
717 if (context->partial_chunk == NULL)
718 context->partial_chunk = g_string_new ("");
720 if (text_start != text_end)
721 g_string_append_len (context->partial_chunk, text_start,
722 text_end - text_start);
724 /* Invariant here that partial_chunk exists */
728 free_partial (GMarkupParseContext *context)
730 if (context->partial_chunk != NULL)
732 g_string_free (context->partial_chunk, TRUE);
733 context->partial_chunk = NULL;
738 current_element (GMarkupParseContext *context)
740 return context->tag_stack->data;
744 current_attribute (GMarkupParseContext *context)
746 return ((GMarkupAttribute*)context->attributes->data)->name;
750 find_current_text_end (GMarkupParseContext *context)
752 /* This function must be safe (non-segfaulting) on invalid UTF8 */
753 const gchar *end = context->current_text + context->current_text_len;
759 p = context->current_text;
760 next = g_utf8_find_next_char (p, end);
765 next = g_utf8_find_next_char (p, end);
768 /* p is now the start of the last character or character portion. */
770 next = g_utf8_next_char (p); /* this only touches *p, nothing beyond */
774 /* whole character */
775 context->current_text_end = end;
780 context->leftover_char_portion = g_string_new_len (p, end - p);
781 context->current_text_len -= (end - p);
782 context->current_text_end = p;
787 g_markup_parse_context_parse (GMarkupParseContext *context,
792 const gchar *first_invalid;
794 g_return_val_if_fail (context != NULL, FALSE);
795 g_return_val_if_fail (text != NULL, FALSE);
796 g_return_val_if_fail (context->state != STATE_ERROR, FALSE);
797 g_return_val_if_fail (!context->parsing, FALSE);
800 text_len = strlen (text);
805 context->parsing = TRUE;
807 if (context->leftover_char_portion)
809 const gchar *first_char;
811 if ((*text & 0xc0) != 0x80)
814 first_char = g_utf8_find_next_char (text, text + text_len);
818 /* leftover_char_portion was completed. Parse it. */
819 GString *portion = context->leftover_char_portion;
821 g_string_append_len (context->leftover_char_portion,
822 text, first_char - text);
824 /* hacks to allow recursion */
825 context->parsing = FALSE;
826 context->leftover_char_portion = NULL;
828 if (!g_markup_parse_context_parse (context,
829 portion->str, portion->len,
832 g_assert (context->state == STATE_ERROR);
835 g_string_free (portion, TRUE);
836 context->parsing = TRUE;
838 /* Skip the fraction of char that was in this text */
839 text_len -= (first_char - text);
844 /* another little chunk of the leftover char; geez
845 * someone is inefficient.
847 g_string_append_len (context->leftover_char_portion,
854 context->current_text = text;
855 context->current_text_len = text_len;
856 context->iter = context->current_text;
857 context->start = context->iter;
859 /* Nothing left after finishing the leftover char, or nothing
860 * passed in to begin with.
862 if (context->current_text_len == 0)
865 /* find_current_text_end () assumes the string starts at
866 * a character start, so we need to validate at least
867 * that much. It doesn't assume any following bytes
870 if ((*context->current_text & 0xc0) == 0x80) /* not a char start */
874 G_MARKUP_ERROR_BAD_UTF8,
875 _("Invalid UTF-8 encoded text"));
879 /* Initialize context->current_text_end, possibly adjusting
880 * current_text_len, and add any leftover char portion
882 find_current_text_end (context);
884 /* Validate UTF8 (must be done after we find the end, since
885 * we could have a trailing incomplete char)
887 if (!g_utf8_validate (context->current_text,
888 context->current_text_len,
893 p = context->current_text;
894 while (p != context->current_text_end)
901 context->line_number += newlines;
905 G_MARKUP_ERROR_BAD_UTF8,
906 _("Invalid UTF-8 encoded text"));
910 while (context->iter != context->current_text_end)
912 switch (context->state)
915 /* Possible next state: AFTER_OPEN_ANGLE */
917 g_assert (context->tag_stack == NULL);
919 /* whitespace is ignored outside of any elements */
920 skip_spaces (context);
922 if (context->iter != context->current_text_end)
924 if (*context->iter == '<')
926 /* Move after the open angle */
927 advance_char (context);
929 context->state = STATE_AFTER_OPEN_ANGLE;
931 /* this could start a passthrough */
932 context->start = context->iter;
934 /* document is now non-empty */
935 context->document_empty = FALSE;
941 G_MARKUP_ERROR_PARSE,
942 _("Document must begin with an element (e.g. <book>)"));
947 case STATE_AFTER_OPEN_ANGLE:
948 /* Possible next states: INSIDE_OPEN_TAG_NAME,
949 * AFTER_CLOSE_TAG_SLASH, INSIDE_PASSTHROUGH
951 if (*context->iter == '?' ||
952 *context->iter == '!')
954 /* include < in the passthrough */
955 const gchar *openangle = "<";
956 add_to_partial (context, openangle, openangle + 1);
957 context->start = context->iter;
958 context->state = STATE_INSIDE_PASSTHROUGH;
960 else if (*context->iter == '/')
963 advance_char (context);
965 context->state = STATE_AFTER_CLOSE_TAG_SLASH;
967 else if (is_name_start_char (g_utf8_get_char (context->iter)))
969 context->state = STATE_INSIDE_OPEN_TAG_NAME;
971 /* start of tag name */
972 context->start = context->iter;
979 G_MARKUP_ERROR_PARSE,
980 _("'%s' is not a valid character following "
981 "a '<' character; it may not begin an "
983 utf8_str (context->iter, buf));
987 /* The AFTER_CLOSE_ANGLE state is actually sort of
988 * broken, because it doesn't correspond to a range
989 * of characters in the input stream as the others do,
990 * and thus makes things harder to conceptualize
992 case STATE_AFTER_CLOSE_ANGLE:
993 /* Possible next states: INSIDE_TEXT, STATE_START */
994 if (context->tag_stack == NULL)
996 context->start = NULL;
997 context->state = STATE_START;
1001 context->start = context->iter;
1002 context->state = STATE_INSIDE_TEXT;
1006 case STATE_AFTER_ELISION_SLASH:
1007 /* Possible next state: AFTER_CLOSE_ANGLE */
1010 /* We need to pop the tag stack and call the end_element
1011 * function, since this is the close tag
1013 GError *tmp_error = NULL;
1015 g_assert (context->tag_stack != NULL);
1018 if (context->parser->end_element)
1019 (* context->parser->end_element) (context,
1020 context->tag_stack->data,
1024 g_free (context->tag_stack->data);
1025 context->tag_stack = g_slist_delete_link (context->tag_stack,
1026 context->tag_stack);
1030 mark_error (context, tmp_error);
1031 g_propagate_error (error, tmp_error);
1035 if (*context->iter == '>')
1037 /* move after the close angle */
1038 advance_char (context);
1039 context->state = STATE_AFTER_CLOSE_ANGLE;
1046 G_MARKUP_ERROR_PARSE,
1047 _("Odd character '%s', expected a '>' character "
1048 "to end the start tag of element '%s'"),
1049 utf8_str (context->iter, buf),
1050 current_element (context));
1056 case STATE_INSIDE_OPEN_TAG_NAME:
1057 /* Possible next states: BETWEEN_ATTRIBUTES */
1059 /* if there's a partial chunk then it's the first part of the
1060 * tag name. If there's a context->start then it's the start
1061 * of the tag name in current_text, the partial chunk goes
1062 * before that start though.
1064 advance_to_name_end (context);
1066 if (context->iter == context->current_text_end)
1068 /* The name hasn't necessarily ended. Merge with
1069 * partial chunk, leave state unchanged.
1071 add_to_partial (context, context->start, context->iter);
1075 /* The name has ended. Combine it with the partial chunk
1076 * if any; push it on the stack; enter next state.
1078 add_to_partial (context, context->start, context->iter);
1079 context->tag_stack =
1080 g_slist_prepend (context->tag_stack,
1081 g_string_free (context->partial_chunk,
1084 context->partial_chunk = NULL;
1086 context->state = STATE_BETWEEN_ATTRIBUTES;
1087 context->start = NULL;
1091 case STATE_INSIDE_ATTRIBUTE_NAME:
1092 /* Possible next states: AFTER_ATTRIBUTE_EQUALS_SIGN */
1094 /* read the full name, if we enter the equals sign state
1095 * then add the attribute to the list (without the value),
1096 * otherwise store a partial chunk to be prepended later.
1098 advance_to_name_end (context);
1100 if (context->iter == context->current_text_end)
1102 /* The name hasn't necessarily ended. Merge with
1103 * partial chunk, leave state unchanged.
1105 add_to_partial (context, context->start, context->iter);
1109 /* The name has ended. Combine it with the partial chunk
1110 * if any; push it on the stack; enter next state.
1112 GMarkupAttribute *attr;
1113 add_to_partial (context, context->start, context->iter);
1115 attr = attribute_new (NULL, NULL);
1117 attr->name = g_string_free (context->partial_chunk,
1120 context->partial_chunk = NULL;
1121 context->start = NULL;
1123 context->attributes =
1124 g_slist_prepend (context->attributes, attr);
1126 if (*context->iter == '=')
1128 advance_char (context);
1129 context->state = STATE_AFTER_ATTRIBUTE_EQUALS_SIGN;
1136 G_MARKUP_ERROR_PARSE,
1137 _("Odd character '%s', expected a '=' after "
1138 "attribute name '%s' of element '%s'"),
1139 utf8_str (context->iter, buf),
1141 current_element (context));
1147 case STATE_BETWEEN_ATTRIBUTES:
1148 /* Possible next states: AFTER_CLOSE_ANGLE,
1149 * AFTER_ELISION_SLASH, INSIDE_ATTRIBUTE_NAME
1151 skip_spaces (context);
1153 if (context->iter != context->current_text_end)
1155 if (*context->iter == '/')
1157 advance_char (context);
1158 context->state = STATE_AFTER_ELISION_SLASH;
1160 else if (*context->iter == '>')
1163 advance_char (context);
1164 context->state = STATE_AFTER_CLOSE_ANGLE;
1166 else if (is_name_start_char (g_utf8_get_char (context->iter)))
1168 context->state = STATE_INSIDE_ATTRIBUTE_NAME;
1169 /* start of attribute name */
1170 context->start = context->iter;
1177 G_MARKUP_ERROR_PARSE,
1178 _("Odd character '%s', expected a '>' or '/' "
1179 "character to end the start tag of "
1180 "element '%s', or optionally an attribute; "
1181 "perhaps you used an invalid character in "
1182 "an attribute name"),
1183 utf8_str (context->iter, buf),
1184 current_element (context));
1187 /* If we're done with attributes, invoke
1188 * the start_element callback
1190 if (context->state == STATE_AFTER_ELISION_SLASH ||
1191 context->state == STATE_AFTER_CLOSE_ANGLE)
1193 const gchar *start_name;
1194 gchar **attr_names = NULL;
1195 gchar **attr_values = NULL;
1198 /* Call user callback for element start */
1199 start_name = current_element (context);
1201 /* this gratuituously copies the attr names/values
1204 attribute_list_to_arrays (context->attributes,
1210 if (context->parser->start_element)
1211 (* context->parser->start_element) (context,
1218 g_strfreev (attr_names);
1219 g_strfreev (attr_values);
1221 /* Go ahead and free this. */
1222 g_slist_foreach (context->attributes, (GFunc)attribute_free,
1224 g_slist_free (context->attributes);
1225 context->attributes = NULL;
1227 if (tmp_error != NULL)
1229 mark_error (context, tmp_error);
1230 g_propagate_error (error, tmp_error);
1236 case STATE_AFTER_ATTRIBUTE_EQUALS_SIGN:
1237 /* Possible next state: INSIDE_ATTRIBUTE_VALUE */
1238 if (*context->iter == '"')
1240 advance_char (context);
1241 context->state = STATE_INSIDE_ATTRIBUTE_VALUE;
1242 context->start = context->iter;
1249 G_MARKUP_ERROR_PARSE,
1250 _("Odd character '%s', expected an open quote mark "
1251 "after the equals sign when giving value for "
1252 "attribute '%s' of element '%s'"),
1253 utf8_str (context->iter, buf),
1254 current_attribute (context),
1255 current_element (context));
1259 case STATE_INSIDE_ATTRIBUTE_VALUE:
1260 /* Possible next states: BETWEEN_ATTRIBUTES */
1263 if (*context->iter == '"')
1266 while (advance_char (context));
1268 if (context->iter == context->current_text_end)
1270 /* The value hasn't necessarily ended. Merge with
1271 * partial chunk, leave state unchanged.
1273 add_to_partial (context, context->start, context->iter);
1277 /* The value has ended at the quote mark. Combine it
1278 * with the partial chunk if any; set it for the current
1281 GMarkupAttribute *attr;
1283 add_to_partial (context, context->start, context->iter);
1285 attr = context->attributes->data;
1287 if (unescape_text (context,
1288 context->partial_chunk->str,
1289 context->partial_chunk->str +
1290 context->partial_chunk->len,
1294 /* success, advance past quote and set state. */
1295 advance_char (context);
1296 context->state = STATE_BETWEEN_ATTRIBUTES;
1297 context->start = NULL;
1300 free_partial (context);
1304 case STATE_INSIDE_TEXT:
1305 /* Possible next states: AFTER_OPEN_ANGLE */
1308 if (*context->iter == '<')
1311 while (advance_char (context));
1313 /* The text hasn't necessarily ended. Merge with
1314 * partial chunk, leave state unchanged.
1317 add_to_partial (context, context->start, context->iter);
1319 if (context->iter != context->current_text_end)
1321 gchar *unescaped = NULL;
1323 /* The text has ended at the open angle. Call the text
1327 if (unescape_text (context,
1328 context->partial_chunk->str,
1329 context->partial_chunk->str +
1330 context->partial_chunk->len,
1334 GError *tmp_error = NULL;
1336 if (context->parser->text)
1337 (*context->parser->text) (context,
1345 if (tmp_error == NULL)
1347 /* advance past open angle and set state. */
1348 advance_char (context);
1349 context->state = STATE_AFTER_OPEN_ANGLE;
1350 /* could begin a passthrough */
1351 context->start = context->iter;
1355 mark_error (context, tmp_error);
1356 g_propagate_error (error, tmp_error);
1360 free_partial (context);
1364 case STATE_AFTER_CLOSE_TAG_SLASH:
1365 /* Possible next state: INSIDE_CLOSE_TAG_NAME */
1366 if (is_name_start_char (g_utf8_get_char (context->iter)))
1368 context->state = STATE_INSIDE_CLOSE_TAG_NAME;
1370 /* start of tag name */
1371 context->start = context->iter;
1378 G_MARKUP_ERROR_PARSE,
1379 _("'%s' is not a valid character following "
1380 "the characters '</'; '%s' may not begin an "
1382 utf8_str (context->iter, buf),
1383 utf8_str (context->iter, buf));
1387 case STATE_INSIDE_CLOSE_TAG_NAME:
1388 /* Possible next state: AFTER_CLOSE_ANGLE */
1389 advance_to_name_end (context);
1391 if (context->iter == context->current_text_end)
1393 /* The name hasn't necessarily ended. Merge with
1394 * partial chunk, leave state unchanged.
1396 add_to_partial (context, context->start, context->iter);
1400 /* The name has ended. Combine it with the partial chunk
1401 * if any; check that it matches stack top and pop
1402 * stack; invoke proper callback; enter next state.
1406 add_to_partial (context, context->start, context->iter);
1408 close_name = g_string_free (context->partial_chunk, FALSE);
1409 context->partial_chunk = NULL;
1411 if (context->tag_stack == NULL)
1415 G_MARKUP_ERROR_PARSE,
1416 _("Element '%s' was closed, no element "
1417 "is currently open"),
1420 else if (strcmp (close_name, current_element (context)) != 0)
1424 G_MARKUP_ERROR_PARSE,
1425 _("Element '%s' was closed, but the currently "
1426 "open element is '%s'"),
1428 current_element (context));
1430 else if (*context->iter != '>')
1435 G_MARKUP_ERROR_PARSE,
1436 _("'%s' is not a valid character following "
1437 "the close element name '%s'; the allowed "
1438 "character is '>'"),
1439 utf8_str (context->iter, buf),
1445 advance_char (context);
1446 context->state = STATE_AFTER_CLOSE_ANGLE;
1447 context->start = NULL;
1449 /* call the end_element callback */
1451 if (context->parser->end_element)
1452 (* context->parser->end_element) (context,
1458 /* Pop the tag stack */
1459 g_free (context->tag_stack->data);
1460 context->tag_stack = g_slist_delete_link (context->tag_stack,
1461 context->tag_stack);
1465 mark_error (context, tmp_error);
1466 g_propagate_error (error, tmp_error);
1470 g_free (close_name);
1474 case STATE_INSIDE_PASSTHROUGH:
1475 /* Possible next state: AFTER_CLOSE_ANGLE */
1478 if (*context->iter == '>')
1481 while (advance_char (context));
1483 if (context->iter == context->current_text_end)
1485 /* The passthrough hasn't necessarily ended. Merge with
1486 * partial chunk, leave state unchanged.
1488 add_to_partial (context, context->start, context->iter);
1492 /* The passthrough has ended at the close angle. Combine
1493 * it with the partial chunk if any. Call the passthrough
1494 * callback. Note that the open/close angles are
1495 * included in the text of the passthrough.
1497 GError *tmp_error = NULL;
1499 advance_char (context); /* advance past close angle */
1500 add_to_partial (context, context->start, context->iter);
1502 if (context->parser->passthrough)
1503 (*context->parser->passthrough) (context,
1504 context->partial_chunk->str,
1505 context->partial_chunk->len,
1509 free_partial (context);
1511 if (tmp_error == NULL)
1513 context->state = STATE_AFTER_CLOSE_ANGLE;
1514 context->start = context->iter; /* could begin text */
1518 mark_error (context, tmp_error);
1519 g_propagate_error (error, tmp_error);
1529 g_assert_not_reached ();
1535 context->parsing = FALSE;
1537 return context->state != STATE_ERROR;
1541 g_markup_parse_context_end_parse (GMarkupParseContext *context,
1544 g_return_val_if_fail (context != NULL, FALSE);
1545 g_return_val_if_fail (!context->parsing, FALSE);
1546 g_return_val_if_fail (context->state != STATE_ERROR, FALSE);
1548 if (context->document_empty)
1550 set_error (context, error, G_MARKUP_ERROR_EMPTY,
1551 _("Document was empty or contained only whitespace"));
1555 context->parsing = TRUE;
1557 switch (context->state)
1563 case STATE_AFTER_OPEN_ANGLE:
1564 set_error (context, error, G_MARKUP_ERROR_PARSE,
1565 _("Document ended unexpectedly just after an open angle bracket '<'"));
1568 case STATE_AFTER_CLOSE_ANGLE:
1569 if (context->tag_stack != NULL)
1571 /* Error message the same as for INSIDE_TEXT */
1572 set_error (context, error, G_MARKUP_ERROR_PARSE,
1573 _("Document ended unexpectedly with elements still open - "
1574 "'%s' was the last element opened"),
1575 current_element (context));
1579 case STATE_AFTER_ELISION_SLASH:
1580 set_error (context, error, G_MARKUP_ERROR_PARSE,
1581 _("Document ended unexpectedly, expected to see a close angle "
1582 "bracket ending the tag <%s/>"), current_element (context));
1585 case STATE_INSIDE_OPEN_TAG_NAME:
1586 set_error (context, error, G_MARKUP_ERROR_PARSE,
1587 _("Document ended unexpectedly inside an element name"));
1590 case STATE_INSIDE_ATTRIBUTE_NAME:
1591 set_error (context, error, G_MARKUP_ERROR_PARSE,
1592 _("Document ended unexpectedly inside an attribute name"));
1595 case STATE_BETWEEN_ATTRIBUTES:
1596 set_error (context, error, G_MARKUP_ERROR_PARSE,
1597 _("Document ended unexpectedly inside an element-opening "
1601 case STATE_AFTER_ATTRIBUTE_EQUALS_SIGN:
1602 set_error (context, error, G_MARKUP_ERROR_PARSE,
1603 _("Document ended unexpectedly after the equals sign "
1604 "following an attribute name; no attribute value"));
1607 case STATE_INSIDE_ATTRIBUTE_VALUE:
1608 set_error (context, error, G_MARKUP_ERROR_PARSE,
1609 _("Document ended unexpectedly while inside an attribute "
1613 case STATE_INSIDE_TEXT:
1614 g_assert (context->tag_stack != NULL);
1615 set_error (context, error, G_MARKUP_ERROR_PARSE,
1616 _("Document ended unexpectedly with elements still open - "
1617 "'%s' was the last element opened"),
1618 current_element (context));
1621 case STATE_AFTER_CLOSE_TAG_SLASH:
1622 case STATE_INSIDE_CLOSE_TAG_NAME:
1623 set_error (context, error, G_MARKUP_ERROR_PARSE,
1624 _("Document ended unexpectedly inside the close tag for"
1625 "element '%s'"), current_element);
1628 case STATE_INSIDE_PASSTHROUGH:
1629 set_error (context, error, G_MARKUP_ERROR_PARSE,
1630 _("Document ended unexpectedly inside a comment or "
1631 "processing instruction"));
1636 g_assert_not_reached ();
1640 context->parsing = FALSE;
1642 return context->state != STATE_ERROR;
1646 g_markup_parse_context_get_position (GMarkupParseContext *context,
1650 g_return_if_fail (context != NULL);
1653 *line_number = context->line_number;
1656 *char_number = context->char_number;
1660 append_escaped_text (GString *str,
1668 end = text + length;
1673 next = g_utf8_next_char (p);
1678 g_string_append (str, "&");
1682 g_string_append (str, "<");
1686 g_string_append (str, ">");
1690 g_string_append (str, "'");
1694 g_string_append (str, """);
1698 g_string_append_len (str, p, next - p);
1707 g_markup_escape_text (const gchar *text,
1712 g_return_val_if_fail (text != NULL, NULL);
1715 length = strlen (text);
1717 str = g_string_new ("");
1718 append_escaped_text (str, text, length);
1720 return g_string_free (str, FALSE);