diff options
Diffstat (limited to 'libtxt')
| -rw-r--r-- | libtxt/Makefile | 2 | ||||
| -rw-r--r-- | libtxt/html.c | 124 | ||||
| -rw-r--r-- | libtxt/inc/html.h | 12 | ||||
| -rw-r--r-- | libtxt/xml.c | 4 |
4 files changed, 132 insertions, 10 deletions
diff --git a/libtxt/Makefile b/libtxt/Makefile index 3536250..087b54c 100644 --- a/libtxt/Makefile +++ b/libtxt/Makefile @@ -6,7 +6,7 @@ LD = ccache ../cross/opt/bin/i686-elf-ld AR = ccache ../cross/opt/bin/i686-elf-ar WARNINGS = -Wall -Wextra -pedantic-errors -Wshadow -Wpointer-arith -Wwrite-strings -Wredundant-decls -Wnested-externs -Wno-long-long -CFLAGS = $(WARNINGS) -nostdlib -nostdinc -fno-builtin -mgeneral-regs-only -std=c99 -m32 -Iinc/ -I../libc/inc/ -fPIE -Duserspace -Ofast +CFLAGS = $(WARNINGS) -nostdlib -nostdinc -fno-builtin -mgeneral-regs-only -std=c99 -m32 -Iinc/ -I../libc/inc/ -I../libgui/inc/ -fPIE -Duserspace -Ofast all: libtxt diff --git a/libtxt/html.c b/libtxt/html.c index 0c07323..171de78 100644 --- a/libtxt/html.c +++ b/libtxt/html.c @@ -1,14 +1,19 @@ // MIT License, Copyright (c) 2020 Marvin Borner // HTML parsing is mainly based on the XML parser +#include <assert.h> +#include <gui.h> +#include <html.h> +#include <list.h> +#include <mem.h> #include <print.h> #include <str.h> +#include <xml.h> -int html_self_closing(const char *tag) +static int is_self_closing(const char *tag) { - // TODO: Add 'meta'? - const char *void_elements[] = { "area", "base", "br", "col", "embed", "hr", "img", - "input", "link", "param", "source", "track", "wbr" }; + const char *void_elements[] = { "area", "base", "br", "col", "embed", "hr", "img", + "input", "link", "meta", "param", "source", "track", "wbr" }; for (u32 i = 0; i < sizeof(void_elements) / sizeof(void_elements[0]); ++i) { if (!strcmp(void_elements[i], tag)) @@ -16,3 +21,114 @@ int html_self_closing(const char *tag) } return 0; } + +static struct dom *new_object(const char *tag, struct dom *parent) +{ + struct dom *object = malloc(sizeof(*object)); + object->tag = strdup(tag); + object->parent = parent; + object->content = NULL; + object->children = list_new(); + return object; +} + +static void print_dom(struct dom *dom, u32 level) +{ + struct node *iterator = dom->children->head; + while (iterator != NULL) { + struct dom *obj = iterator->data; + for (u32 i = 0; i < level; i++) + print("\t"); + printf("'%s': '%s'\n", obj->tag, obj->content ? obj->content : ""); + if (obj->children->head) + print_dom(obj, level + 1); + iterator = iterator->next; + } +} + +static struct dom *generate_dom(char *data, u32 length) +{ + struct xml_token tokens[128]; + struct xml parser; + xml_init(&parser); + void *buffer = data; + enum xml_error err = xml_parse(&parser, buffer, length, tokens, 128); + + if (err != XML_SUCCESS && err != XML_ERROR_BUFFERDRY) { + printf("\nXML parse error: %d\n", err); + return 0; + } + + struct dom *root = new_object("root", NULL); + struct dom *current = root; + + char name[256] = { 0 }; + for (u32 i = 0; i < parser.ntokens; i++) { + const struct xml_token *token = tokens + i; + name[0] = '\0'; + switch (token->type) { + case XML_START_TAG: + memcpy(&name, (u8 *)buffer + token->start_pos, + token->end_pos - token->start_pos); + name[token->end_pos - token->start_pos] = '\0'; + current = new_object(name, current); + printf("Adding %s to %s\n", current->tag, current->parent->tag); + list_add(current->parent->children, current); + break; + case XML_END_TAG: + memcpy(&name, (u8 *)buffer + token->start_pos, + token->end_pos - token->start_pos); + name[token->end_pos - token->start_pos] = '\0'; + assert(current && !strcmp(name, current->tag)); + current = current->parent; + break; + case XML_CHARACTER: + if (!current) + continue; + + if (token->end_pos == token->start_pos + 2) { + const char *ptr = (char *)buffer + token->start_pos; + + if (ptr[0] == '\r' && ptr[1] == '\n') + continue; + } + memcpy(&name, (u8 *)buffer + token->start_pos, + token->end_pos - token->start_pos); + name[token->end_pos - token->start_pos] = '\0'; + char *clean_name = name; + for (u32 j = 0; j < strlen(name); j++) { + if (name[j] == ' ' || name[j] == '\n' || name[j] == '\r' || + name[j] == '\t') { + clean_name++; + } else { + break; + } + } + if (!strlen(clean_name)) + break; + current->content = strdup(clean_name); + break; + default: + break; + } + + i += token->size; + } + + print("GENERATED!\n"); + print_dom(root, 0); + return root; +} + +int html_render_dom(struct element *container, struct dom *dom) +{ + (void)container; + (void)dom; + return 1; +} + +int html_render(struct element *container, char *data, u32 length) +{ + struct dom *dom = generate_dom(data, length); + return dom && html_render_dom(container, dom); +} diff --git a/libtxt/inc/html.h b/libtxt/inc/html.h index dd2b59f..4eaca0b 100644 --- a/libtxt/inc/html.h +++ b/libtxt/inc/html.h @@ -4,6 +4,16 @@ #ifndef HTML_H #define HTML_H -int html_self_closing(const char *tag); +#include <def.h> +#include <list.h> + +struct dom { + char *tag; + char *content; + struct dom *parent; + struct list *children; +}; + +int html_render(struct element *container, char *data, u32 length); #endif diff --git a/libtxt/xml.c b/libtxt/xml.c index b92181b..f40b289 100644 --- a/libtxt/xml.c +++ b/libtxt/xml.c @@ -485,10 +485,6 @@ enum xml_error xml_parse(struct xml *state, const char *buffer, u32 buffer_lengt state_commit(state, &temp); } - // TODO: Only for self-closing tags - if (end - lt == 0) - break; - if (end - lt < TAG_MINSIZE) return XML_ERROR_BUFFERDRY; |