Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-09 03:04:10 -04:00
Code Issues Packages Projects Releases Wiki Activity

Detect font usage inside XObjects as well

Browse Source
This commit is contained in:
Kovid Goyal 2019-07-28 14:12:47 +05:30
parent c842a82a0f
commit bf602f48ed
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
1 changed files with 29 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

43
src/calibre/utils/podofo/fonts.cpp
View File

@ -78,14 +78,17 @@ replace_font_references(PDFDoc *self, const std::unordered_map<uint64_t, uint64_
} }
} }
static bool static void
used_fonts_in_page(PdfPage *page, unordered_reference_set &ans) { used_fonts_in_canvas(PdfCanvas *canvas, unordered_reference_set &ans) {
PdfContentsTokenizer tokenizer(page); PdfContentsTokenizer tokenizer(canvas);
bool in_text_block = false; bool in_text_block = false;
const char* token = NULL; const char* token = NULL;
EPdfContentsType contents_type; EPdfContentsType contents_type;
PdfVariant var; PdfVariant var;
std::stack<PdfVariant> stack; std::stack<PdfVariant> stack;
const PdfDictionary &resources = canvas->GetResources()->GetDictionary();
if (!resources.HasKey("Font")) return;
const PdfDictionary &fonts_dict = resources.GetKey("Font")->GetDictionary();
while (tokenizer.ReadNext(contents_type, token, var)) { while (tokenizer.ReadNext(contents_type, token, var)) {
if (contents_type == ePdfContentsType_Variant) stack.push(var); if (contents_type == ePdfContentsType_Variant) stack.push(var);
@ -102,12 +105,13 @@ used_fonts_in_page(PdfPage *page, unordered_reference_set &ans) {
stack.pop(); stack.pop();
if (stack.size() > 0 && stack.top().IsName()) { if (stack.size() > 0 && stack.top().IsName()) {
const PdfName &reference_name = stack.top().GetName(); const PdfName &reference_name = stack.top().GetName();
PdfObject* font = page->GetFromResources("Font", reference_name); if (fonts_dict.HasKey(reference_name)) {
if (font) ans.insert(font->Reference()); ans.insert(fonts_dict.GetKey(reference_name)->GetReference());
}
} }
} }
} }
return true; return;
} }
static PyObject* static PyObject*
@ -238,24 +242,35 @@ static PyObject*
remove_unused_fonts(PDFDoc *self, PyObject *args) { remove_unused_fonts(PDFDoc *self, PyObject *args) {
unsigned long count = 0; unsigned long count = 0;
unordered_reference_set used_fonts; unordered_reference_set used_fonts;
// Look in Pages
for (int i = 0; i < self->doc->GetPageCount(); i++) { for (int i = 0; i < self->doc->GetPageCount(); i++) {
PdfPage *page = self->doc->GetPage(i); PdfPage *page = self->doc->GetPage(i);
if (page) used_fonts_in_page(page, used_fonts); if (page) used_fonts_in_canvas(page, used_fonts);
}
// Look in XObjects
PdfVecObjects &objects = self->doc->GetObjects();
for (auto &k : objects) {
if (k->IsDictionary()) {
const PdfDictionary &dict = k->GetDictionary();
if (dictionary_has_key_name(dict, PdfName::KeyType, "XObject") && dictionary_has_key_name(dict, PdfName::KeySubtype, "Form")) {
PdfXObject xo(k);
used_fonts_in_canvas(&xo, used_fonts);
}
}
} }
unordered_reference_set all_fonts; unordered_reference_set all_fonts;
unordered_reference_set type3_fonts; unordered_reference_set type3_fonts;
charprocs_usage_map charprocs_usage; charprocs_usage_map charprocs_usage;
PdfVecObjects &objects = self->doc->GetObjects(); for (auto &k : objects) {
for (TCIVecObjects it = objects.begin(); it != objects.end(); it++) { if (k->IsDictionary()) {
if ((*it)->IsDictionary()) { const PdfDictionary &dict = k->GetDictionary();
const PdfDictionary &dict = (*it)->GetDictionary();
if (dictionary_has_key_name(dict, PdfName::KeyType, "Font")) { if (dictionary_has_key_name(dict, PdfName::KeyType, "Font")) {
const std::string &font_type = dict.GetKey(PdfName::KeySubtype)->GetName().GetName(); const std::string &font_type = dict.GetKey(PdfName::KeySubtype)->GetName().GetName();
if (font_type == "Type0") { if (font_type == "Type0") {
all_fonts.insert((*it)->Reference()); all_fonts.insert(k->Reference());
} else if (font_type == "Type3") { } else if (font_type == "Type3") {
all_fonts.insert((*it)->Reference()); all_fonts.insert(k->Reference());
type3_fonts.insert((*it)->Reference()); type3_fonts.insert(k->Reference());
for (auto &x : dict.GetKey("CharProcs")->GetDictionary().GetKeys()) { for (auto &x : dict.GetKey("CharProcs")->GetDictionary().GetKeys()) {
const PdfReference &ref = x.second->GetReference(); const PdfReference &ref = x.second->GetReference();
if (charprocs_usage.find(ref) == charprocs_usage.end()) charprocs_usage[ref] = 1; if (charprocs_usage.find(ref) == charprocs_usage.end()) charprocs_usage[ref] = 1;