Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-08-11 09:13:57 -04:00
Code Issues Packages Projects Releases Wiki Activity

Add API to count images in a PDF file

Browse Source
This commit is contained in:
Kovid Goyal 2015-03-17 09:36:46 +05:30
parent f7da0bc178
commit bf152707b3
2 changed files with 36 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
src/calibre/utils/podofo/__init__.py
View File

@ -119,6 +119,14 @@ def get_xmp_metadata(path):
p.load(raw) p.load(raw)
return p.get_xmp_metadata() return p.get_xmp_metadata()
def get_image_count(path):
podofo = get_podofo()
p = podofo.PDFDoc()
with open(path, 'rb') as f:
raw = f.read()
p.load(raw)
return p.image_count()
def test_outline(src): def test_outline(src):
podofo = get_podofo() podofo = get_podofo()
p = podofo.PDFDoc() p = podofo.PDFDoc()

28
src/calibre/utils/podofo/doc.cpp
View File

@ -141,6 +141,31 @@ PDFDoc_page_count(PDFDoc *self, PyObject *args) {
return Py_BuildValue("i", count); return Py_BuildValue("i", count);
} // }}} } // }}}
// image_count() {{{
static PyObject *
PDFDoc_image_count(PDFDoc *self, PyObject *args) {
int count = 0;
const PdfObject* obj_type = NULL;
const PdfObject* obj_sub_type = NULL;
try {
TCIVecObjects it = self->doc->GetObjects().begin();
while( it != self->doc->GetObjects().end() ) {
if( (*it)->IsDictionary() ) {
obj_type = (*it)->GetDictionary().GetKey( PdfName::KeyType );
obj_sub_type = (*it)->GetDictionary().GetKey( PdfName::KeySubtype );
if( ( obj_type && obj_type->IsName() && ( obj_type->GetName().GetName() == "XObject" ) ) ||
( obj_sub_type && obj_sub_type->IsName() && ( obj_sub_type->GetName().GetName() == "Image" ) ) ) count++;
self->doc->FreeObjectMemory( *it );
}
it++;
}
} catch(const PdfError & err) {
podofo_set_exception(err);
return NULL;
}
return Py_BuildValue("i", count);
} // }}}
// delete_page {{{ // delete_page {{{
static PyObject * static PyObject *
PDFDoc_delete_page(PDFDoc *self, PyObject *args) { PDFDoc_delete_page(PDFDoc *self, PyObject *args) {
@ -539,6 +564,9 @@ static PyMethodDef PDFDoc_methods[] = {
{"page_count", (PyCFunction)PDFDoc_page_count, METH_VARARGS, {"page_count", (PyCFunction)PDFDoc_page_count, METH_VARARGS,
"page_count() -> Number of pages in the PDF." "page_count() -> Number of pages in the PDF."
}, },
{"image_count", (PyCFunction)PDFDoc_image_count, METH_VARARGS,
"image_count() -> Number of images in the PDF."
},
{"delete_page", (PyCFunction)PDFDoc_delete_page, METH_VARARGS, {"delete_page", (PyCFunction)PDFDoc_delete_page, METH_VARARGS,
"delete_page(page_num) -> Delete the specified page from the pdf (0 is the first page)." "delete_page(page_num) -> Delete the specified page from the pdf (0 is the first page)."
}, },