char *GetHtml() {
// first add the homepage
const char *index = doc->GetHomePath();
ScopedMem<WCHAR> url(doc->ToStr(index));
Visit(NULL, url, 0);
// then add all pages linked to from the table of contents
doc->ParseToc(this);
// finally add all the remaining HTML files
Vec<char *> *paths = doc->GetAllPaths();
for (size_t i = 0; i < paths->Count(); i++) {
char *path = paths->At(i);
if (str::EndsWithI(path, ".htm") || str::EndsWithI(path, ".html")) {
if (*path == '/')
path++;
url.Set(str::conv::FromUtf8(path));
Visit(NULL, url, -1);
}
}
FreeVecMembers(*paths);
delete paths;
return html.StealData();
}
ChmDoc* ChmDoc::CreateFromFile(const WCHAR* fileName) {
ChmDoc* doc = new ChmDoc();
if (!doc || !doc->Load(fileName)) {
delete doc;
return nullptr;
}
return doc;
}
DocTocItem *Chm2EngineImpl::GetTocTree()
{
EbookTocBuilder builder(this);
doc->ParseToc(&builder);
if (doc->HasIndex()) {
// TODO: ToC code doesn't work too well for displaying an index,
// so this should really become a tree of its own (which
// doesn't rely on entries being in the same order as pages)
builder.Visit(L"Index", NULL, 1);
builder.SetIsIndex(true);
doc->ParseIndex(&builder);
}
return builder.GetRoot();
}
virtual void Visit(const WCHAR *name, const WCHAR *url, int level) {
if (!url || IsExternalUrl(url))
return;
ScopedMem<WCHAR> plainUrl(str::ToPlainUrl(url));
if (added.FindI(plainUrl) != -1)
return;
ScopedMem<char> urlUtf8(str::conv::ToUtf8(plainUrl));
size_t pageHtmlLen;
ScopedMem<unsigned char> pageHtml(doc->GetData(urlUtf8, &pageHtmlLen));
if (!pageHtml)
return;
html.AppendFmt("<pagebreak page_path=\"%s\" page_marker />", urlUtf8);
html.AppendAndFree(doc->ToUtf8(pageHtml, ExtractHttpCharset((const char *)pageHtml.Get(), pageHtmlLen)));
added.Append(plainUrl.StealData());
}
void CreateThumbnail(HtmlWindow *hw) {
this->hw = hw;
homeUrl.Set(str::conv::FromAnsi(doc->GetHomePath()));
if (*homeUrl == '/')
homeUrl.Set(str::Dup(homeUrl + 1));
hw->NavigateToDataUrl(homeUrl);
}
virtual const unsigned char *GetDataForUrl(const WCHAR *url, size_t *len) {
ScopedCritSec scope(&docAccess);
ScopedMem<WCHAR> plainUrl(url::GetFullPath(url));
ScopedMem<char> urlUtf8(str::conv::ToUtf8(plainUrl));
data.Append(doc->GetData(urlUtf8, len));
return data.Last();
}
bool ChmEngineImpl::Load(const WCHAR *fileName)
{
this->fileName = str::Dup(fileName);
Timer t(true);
doc = ChmDoc::CreateFromFile(fileName);
dbglog::LogF("ChmDoc::CreateFromFile(): %.2f ms", t.GetTimeInMs());
if (!doc)
return false;
// always make the document's homepage page 1
pages.Append(str::conv::FromAnsi(doc->GetHomePath()));
// parse the ToC here, since page numbering depends on it
t.Start();
doc->ParseToc(&ChmTocBuilder(doc, &pages, &tocRoot));
dbglog::LogF("doc->ParseToc(): %.2f ms", t.GetTimeInMs());
CrashIf(pages.Count() == 0);
return pages.Count() > 0;
}
ImageData *GetImageData(const char *id, const char *pagePath) {
ScopedMem<char> url(NormalizeURL(id, pagePath));
str::UrlDecodeInPlace(url);
for (size_t i = 0; i < images.Count(); i++) {
if (str::Eq(images.At(i).id, url))
return &images.At(i).base;
}
ImageData2 data = { 0 };
data.base.data = (char *)doc->GetData(url, &data.base.len);
if (!data.base.data)
return NULL;
data.id = url.StealData();
images.Append(data);
return &images.Last().base;
}
// Load and cache data for a given url inside CHM file.
bool ChmEngineImpl::GetDataForUrl(const WCHAR *url, char **data, size_t *len)
{
ScopedMem<WCHAR> plainUrl(str::ToPlainUrl(url));
ChmCacheEntry *e = FindDataForUrl(plainUrl);
if (!e) {
e = new ChmCacheEntry(plainUrl);
ScopedMem<char> urlUtf8(str::conv::ToUtf8(plainUrl));
e->data = (char *)doc->GetData(urlUtf8, &e->size);
if (!e->data) {
delete e;
return false;
}
urlDataCache.Append(e);
}
*data = e->data;
*len = e->size;
return true;
}
