static fz_css_rule *
html_load_css(fz_context *ctx, fz_archive *zip, const char *base_uri, fz_css_rule *css, fz_xml *root)
{
fz_xml *html, *head, *node;
fz_buffer *buf;
char path[2048];
fz_var(buf);
html = fz_xml_find(root, "html");
head = fz_xml_find_down(html, "head");
for (node = fz_xml_down(head); node; node = fz_xml_next(node))
{
if (fz_xml_is_tag(node, "link"))
{
char *rel = fz_xml_att(node, "rel");
if (rel && !fz_strcasecmp(rel, "stylesheet"))
{
char *type = fz_xml_att(node, "type");
if ((type && !strcmp(type, "text/css")) || !type)
{
char *href = fz_xml_att(node, "href");
if (href)
{
fz_strlcpy(path, base_uri, sizeof path);
fz_strlcat(path, "/", sizeof path);
fz_strlcat(path, href, sizeof path);
fz_urldecode(path);
fz_cleanname(path);
buf = NULL;
fz_try(ctx)
{
buf = fz_read_archive_entry(ctx, zip, path);
fz_write_buffer_byte(ctx, buf, 0);
css = fz_parse_css(ctx, css, (char*)buf->data, path);
}
fz_always(ctx)
fz_drop_buffer(ctx, buf);
fz_catch(ctx)
fz_warn(ctx, "ignoring stylesheet %s", path);
}
}
}
static void
epub_parse_header(fz_context *ctx, epub_document *doc)
{
fz_archive *zip = doc->zip;
fz_buffer *buf;
fz_xml *container_xml, *content_opf;
fz_xml *container, *rootfiles, *rootfile;
fz_xml *package, *manifest, *spine, *itemref, *metadata;
char base_uri[2048];
const char *full_path;
const char *version;
char ncx[2048], s[2048];
epub_chapter **tailp;
size_t len;
unsigned char *data;
if (fz_has_archive_entry(ctx, zip, "META-INF/rights.xml"))
fz_throw(ctx, FZ_ERROR_GENERIC, "EPUB is locked by DRM");
if (fz_has_archive_entry(ctx, zip, "META-INF/encryption.xml"))
fz_throw(ctx, FZ_ERROR_GENERIC, "EPUB is locked by DRM");
/* parse META-INF/container.xml to find OPF */
buf = fz_read_archive_entry(ctx, zip, "META-INF/container.xml");
fz_write_buffer_byte(ctx, buf, 0);
len = fz_buffer_storage(ctx, buf, &data);
container_xml = fz_parse_xml(ctx, data, len, 0);
fz_drop_buffer(ctx, buf);
container = fz_xml_find(container_xml, "container");
rootfiles = fz_xml_find_down(container, "rootfiles");
rootfile = fz_xml_find_down(rootfiles, "rootfile");
full_path = fz_xml_att(rootfile, "full-path");
if (!full_path)
fz_throw(ctx, FZ_ERROR_GENERIC, "cannot find root file in EPUB");
fz_dirname(base_uri, full_path, sizeof base_uri);
/* parse OPF to find NCX and spine */
buf = fz_read_archive_entry(ctx, zip, full_path);
fz_write_buffer_byte(ctx, buf, 0);
len = fz_buffer_storage(ctx, buf, &data);
content_opf = fz_parse_xml(ctx, data, len, 0);
fz_drop_buffer(ctx, buf);
package = fz_xml_find(content_opf, "package");
version = fz_xml_att(package, "version");
if (!version || strcmp(version, "2.0"))
fz_warn(ctx, "unknown epub version: %s", version ? version : "<none>");
metadata = fz_xml_find_down(package, "metadata");
if (metadata)
{
doc->dc_title = find_metadata(ctx, metadata, "title");
doc->dc_creator = find_metadata(ctx, metadata, "creator");
}
manifest = fz_xml_find_down(package, "manifest");
spine = fz_xml_find_down(package, "spine");
if (path_from_idref(ncx, manifest, base_uri, fz_xml_att(spine, "toc"), sizeof ncx))
{
epub_parse_ncx(ctx, doc, ncx);
}
doc->spine = NULL;
tailp = &doc->spine;
itemref = fz_xml_find_down(spine, "itemref");
while (itemref)
{
if (path_from_idref(s, manifest, base_uri, fz_xml_att(itemref, "idref"), sizeof s))
{
*tailp = epub_parse_chapter(ctx, doc, s);
tailp = &(*tailp)->next;
}
itemref = fz_xml_find_next(itemref, "itemref");
}
fz_drop_xml(ctx, container_xml);
fz_drop_xml(ctx, content_opf);
}
static void
epub_parse_header(fz_context *ctx, epub_document *doc)
{
fz_archive *zip = doc->zip;
fz_buffer *buf;
fz_xml *container_xml, *content_opf;
fz_xml *container, *rootfiles, *rootfile;
fz_xml *package, *manifest, *spine, *itemref;
char base_uri[2048];
const char *full_path;
const char *version;
char ncx[2048], s[2048];
epub_chapter *head, *tail;
if (fz_has_archive_entry(ctx, zip, "META-INF/rights.xml"))
fz_throw(ctx, FZ_ERROR_GENERIC, "EPUB is locked by DRM");
if (fz_has_archive_entry(ctx, zip, "META-INF/encryption.xml"))
fz_throw(ctx, FZ_ERROR_GENERIC, "EPUB is locked by DRM");
/* parse META-INF/container.xml to find OPF */
buf = fz_read_archive_entry(ctx, zip, "META-INF/container.xml");
fz_write_buffer_byte(ctx, buf, 0);
container_xml = fz_parse_xml(ctx, buf->data, buf->len, 0);
fz_drop_buffer(ctx, buf);
container = fz_xml_find(container_xml, "container");
rootfiles = fz_xml_find_down(container, "rootfiles");
rootfile = fz_xml_find_down(rootfiles, "rootfile");
full_path = fz_xml_att(rootfile, "full-path");
if (!full_path)
fz_throw(ctx, FZ_ERROR_GENERIC, "cannot find root file in EPUB");
printf("epub: found root: %s\n", full_path);
fz_dirname(base_uri, full_path, sizeof base_uri);
/* parse OPF to find NCX and spine */
buf = fz_read_archive_entry(ctx, zip, full_path);
fz_write_buffer_byte(ctx, buf, 0);
content_opf = fz_parse_xml(ctx, buf->data, buf->len, 0);
fz_drop_buffer(ctx, buf);
package = fz_xml_find(content_opf, "package");
version = fz_xml_att(package, "version");
if (!version || strcmp(version, "2.0"))
fz_warn(ctx, "unknown epub version: %s", version ? version : "<none>");
manifest = fz_xml_find_down(package, "manifest");
spine = fz_xml_find_down(package, "spine");
if (path_from_idref(ncx, manifest, base_uri, fz_xml_att(spine, "toc"), sizeof ncx))
{
/* TODO: parse NCX to create fz_outline */
printf("epub: found outline: %s\n", ncx);
}
head = tail = NULL;
itemref = fz_xml_find_down(spine, "itemref");
while (itemref)
{
if (path_from_idref(s, manifest, base_uri, fz_xml_att(itemref, "idref"), sizeof s))
{
printf("epub: found spine %s\n", s);
if (!head)
head = tail = epub_parse_chapter(ctx, doc, s);
else
tail = tail->next = epub_parse_chapter(ctx, doc, s);
}
itemref = fz_xml_find_next(itemref, "itemref");
}
doc->spine = head;
printf("epub: done.\n");
fz_drop_xml(ctx, container_xml);
fz_drop_xml(ctx, content_opf);
}
