static void sweepref(fz_obj *obj)
{
int num = fz_tonum(obj);
int gen = fz_togen(obj);
if (num < 0 || num >= xref->len)
return;
if (uselist[num])
return;
uselist[num] = 1;
/* Bake in /Length in stream objects */
if (pdf_isstream(xref, num, gen))
{
fz_obj *len = fz_dictgets(obj, "Length");
if (fz_isindirect(len))
{
uselist[fz_tonum(len)] = 0;
len = fz_resolveindirect(len);
fz_dictputs(obj, "Length", len);
}
}
sweepobj(fz_resolveindirect(obj));
}
static void removeduplicateobjs(void)
{
int num, other;
for (num = 1; num < xref->len; num++)
{
/* Only compare an object to objects preceeding it */
for (other = 1; other < num; other++)
{
fz_obj *a, *b;
if (num == other || !uselist[num] || !uselist[other])
continue;
/*
* Comparing stream objects data contents would take too long.
*
* pdf_isstream calls pdf_cacheobject and ensures
* that the xref table has the objects loaded.
*/
if (pdf_isstream(xref, num, 0) || pdf_isstream(xref, other, 0))
continue;
a = xref->table[num].obj;
b = xref->table[other].obj;
a = fz_resolveindirect(a);
b = fz_resolveindirect(b);
if (fz_objcmp(a, b))
continue;
/* Keep the lowest numbered object */
renumbermap[num] = MIN(num, other);
renumbermap[other] = MIN(num, other);
uselist[MAX(num, other)] = 0;
/* One duplicate was found, do not look for another */
break;
}
}
}
static void
printglobalinfo(void)
{
printf("\nPDF-%d.%d\n", xref->version / 10, xref->version % 10);
if (info->u.info.obj)
{
printf("Info object (%d %d R):\n", fz_tonum(info->ref), fz_togen(info->ref));
fz_debugobj(fz_resolveindirect(info->u.info.obj));
}
if (cryptinfo->u.crypt.obj)
{
printf("\nEncryption object (%d %d R):\n", fz_tonum(cryptinfo->ref), fz_togen(cryptinfo->ref));
fz_debugobj(cryptinfo->u.crypt.obj);
}
printf("\nPages: %d\n\n", pagecount);
}
static fz_error
loadiccbased(fz_colorspace **csp, pdf_xref *xref, fz_obj *ref)
{
fz_obj *dict;
int n;
pdf_logrsrc("load ICCBased\n");
dict = fz_resolveindirect(ref);
n = fz_toint(fz_dictgets(dict, "N"));
switch (n)
{
case 1: *csp = pdf_devicegray; return fz_okay;
case 3: *csp = pdf_devicergb; return fz_okay;
case 4: *csp = pdf_devicecmyk; return fz_okay;
}
return fz_throw("syntaxerror: ICCBased must have 1, 3 or 4 components");
}
Epdf_Document* epdf_document_new(const char* filename)
{
Epdf_Document* doc;
fz_error error;
doc = (Epdf_Document*)malloc(sizeof(Epdf_Document));
if(!doc)
return NULL;
doc->xref = NULL;
doc->outline = NULL;
doc->rast = NULL;
doc->pagecount = 0;
doc->zoom = 1.0;
doc->rotate = 0;
doc->locked = false;
error = fz_newrenderer(&doc->rast, pdf_devicergb, 0, 1024 * 512);
if(error)
return pdfdoc_error(doc);
fz_obj* obj;
// Open PDF and load xref table
doc->xref = pdf_newxref();
if(!doc->xref)
return pdfdoc_error(doc);
error = pdf_loadxref(doc->xref, (char*)filename);
if(error)
{
fz_catch(error, "trying to repair");
fprintf(stderr, "There was a problem with file \"%s\".\nIt may be corrupted or generated by faulty software.\nTrying to repair the file.\n", filename);
error = pdf_repairxref(doc->xref, (char*)filename);
if(error)
return pdfdoc_error(doc);
}
error = pdf_decryptxref(doc->xref);
if(error)
pdfdoc_error(doc);
// Handle encrypted PDF files
if(pdf_needspassword(doc->xref))
doc->locked = true;
/*
* Load meta information
* TODO: move this into mupdf library
*/
obj = fz_dictgets(doc->xref->trailer, (char*)"Root");
doc->xref->root = fz_resolveindirect(obj);
if(!doc->xref->root)
{
fz_throw("syntaxerror: missing Root object");
pdfdoc_error(doc);
}
fz_keepobj(doc->xref->root);
obj = fz_dictgets(doc->xref->trailer, "Info");
doc->xref->info = fz_resolveindirect(obj);
if(!doc->xref->info)
fprintf(stderr, "Could not load PDF meta information.\n");
if(doc->xref->info)
fz_keepobj(doc->xref->info);
doc->outline = pdf_loadoutline(doc->xref);
doc->filename = strdup(filename);
if(doc->xref->info)
{
obj = fz_dictgets(doc->xref->info, "Title");
if(obj)
doc->doctitle = pdf_toutf8(obj);
}
/*
* Start at first page
*/
doc->pagecount = pdf_getpagecount(doc->xref);
if(doc->zoom < 0.1)
doc->zoom = 0.1;
if(doc->zoom > 3.0)
doc->zoom = 3.0;
return doc;
}
void pdfmoz_open(pdfmoz_t *moz, char *filename)
{
SCROLLINFO si;
fz_error error;
fz_obj *obj;
fz_irect bbox;
int rot;
int i;
strcpy(moz->error, "");
error = fz_newrenderer(&moz->rast, pdf_devicergb, 0, 1024 * 512);
if (error)
pdfmoz_error(moz, error);
/*
* Open PDF and load xref table
*/
moz->filename = filename;
moz->xref = pdf_newxref();
error = pdf_loadxref(moz->xref, filename);
if (error)
{
error = pdf_repairxref(moz->xref, filename);
if (error)
pdfmoz_error(moz, error);
}
/*
* Handle encrypted PDF files
*/
error = pdf_decryptxref(moz->xref);
if (error)
pdfmoz_error(moz, error);
if (pdf_needspassword(moz->xref))
{
pdfmoz_warn(moz, "PDF file is encrypted and needs a password.");
}
moz->pagecount = pdf_getpagecount(moz->xref);
moz->pages = fz_malloc(sizeof(page_t) * moz->pagecount);
for (i = 0; i < moz->pagecount; i++)
{
fz_obj *pageobj;
pageobj = pdf_getpageobject(moz->xref, i);
moz->pages[i].obj = fz_keepobj(pageobj);
moz->pages[i].page = nil;
moz->pages[i].image = nil;
obj = fz_dictgets(moz->pages[i].obj, "CropBox");
if (!obj)
obj = fz_dictgets(moz->pages[i].obj, "MediaBox");
bbox = fz_roundrect(pdf_torect(obj));
moz->pages[i].w = bbox.x1 - bbox.x0;
moz->pages[i].h = bbox.y1 - bbox.y0;
rot = fz_toint(fz_dictgets(moz->pages[i].obj, "Rotate"));
if ((rot / 90) % 2)
{
int t = moz->pages[i].w;
moz->pages[i].w = moz->pages[i].h;
moz->pages[i].h = t;
}
moz->pages[i].px = 1 + PAD;
}
/*
* Load meta information
* TODO: move this into mupdf library
*/
obj = fz_dictgets(moz->xref->trailer, "Root");
moz->xref->root = fz_resolveindirect(obj);
if (!moz->xref->root)
pdfmoz_error(moz, fz_throw("syntaxerror: missing Root object"));
if (moz->xref->root)
fz_keepobj(moz->xref->root);
obj = fz_dictgets(moz->xref->trailer, "Info");
moz->xref->info = fz_resolveindirect(obj);
if (moz->xref->info)
fz_keepobj(moz->xref->info);
moz->doctitle = filename;
if (strrchr(moz->doctitle, '\\'))
moz->doctitle = strrchr(moz->doctitle, '\\') + 1;
if (strrchr(moz->doctitle, '/'))
moz->doctitle = strrchr(moz->doctitle, '/') + 1;
if (moz->xref->info)
{
obj = fz_dictgets(moz->xref->info, "Title");
if (obj)
moz->doctitle = pdf_toutf8(obj);
}
/*
* Start at first page
*/
si.cbSize = sizeof(si);
si.fMask = SIF_POS | SIF_RANGE; // XXX | SIF_PAGE;
si.nPos = 0;
si.nMin = 0;
si.nMax = 1;
si.nPage = 1;
SetScrollInfo(moz->hwnd, SB_VERT, &si, TRUE);
moz->scrollpage = 0;
moz->scrollyofs = 0;
InvalidateRect(moz->hwnd, NULL, FALSE);
}
/*
* Load CMap stream in PDF file
*/
fz_error
pdf_loadembeddedcmap(pdf_cmap **cmapp, pdf_xref *xref, fz_obj *stmref)
{
fz_error error = fz_okay;
fz_obj *stmobj;
fz_stream *file = nil;
pdf_cmap *cmap = nil;
pdf_cmap *usecmap;
fz_obj *wmode;
fz_obj *obj;
if ((*cmapp = pdf_finditem(xref->store, PDF_KCMAP, stmref)))
{
pdf_keepcmap(*cmapp);
return fz_okay;
}
pdf_logfont("load embedded cmap (%d %d R) {\n", fz_tonum(stmref), fz_togen(stmref));
stmobj = fz_resolveindirect(stmref);
error = pdf_openstream(&file, xref, fz_tonum(stmref), fz_togen(stmref));
if (error)
{
error = fz_rethrow(error, "cannot open cmap stream");
goto cleanup;
}
error = pdf_parsecmap(&cmap, file);
if (error)
{
error = fz_rethrow(error, "cannot parse cmap stream");
goto cleanup;
}
fz_dropstream(file);
wmode = fz_dictgets(stmobj, "WMode");
if (fz_isint(wmode))
{
pdf_logfont("wmode %d\n", wmode);
pdf_setwmode(cmap, fz_toint(wmode));
}
obj = fz_dictgets(stmobj, "UseCMap");
if (fz_isname(obj))
{
pdf_logfont("usecmap /%s\n", fz_toname(obj));
error = pdf_loadsystemcmap(&usecmap, fz_toname(obj));
if (error)
{
error = fz_rethrow(error, "cannot load system usecmap '%s'", fz_toname(obj));
goto cleanup;
}
pdf_setusecmap(cmap, usecmap);
pdf_dropcmap(usecmap);
}
else if (fz_isindirect(obj))
{
pdf_logfont("usecmap (%d %d R)\n", fz_tonum(obj), fz_togen(obj));
error = pdf_loadembeddedcmap(&usecmap, xref, obj);
if (error)
{
error = fz_rethrow(error, "cannot load embedded usecmap");
goto cleanup;
}
pdf_setusecmap(cmap, usecmap);
pdf_dropcmap(usecmap);
}
pdf_logfont("}\n");
error = pdf_storeitem(xref->store, PDF_KCMAP, stmref, cmap);
if (error)
{
error = fz_rethrow(error, "cannot store cmap resource");
goto cleanup;
}
*cmapp = cmap;
return fz_okay;
cleanup:
if (file)
fz_dropstream(file);
if (cmap)
pdf_dropcmap(cmap);
return error; /* already rethrown */
}
void
pdf_loadpagetreenode(pdf_xref *xref, fz_obj *node, struct info info)
{
fz_obj *dict, *kids, *count;
fz_obj *obj, *tmp;
int i, n;
/* prevent infinite recursion */
if (fz_dictgets(node, ".seen"))
return;
kids = fz_dictgets(node, "Kids");
count = fz_dictgets(node, "Count");
if (fz_isarray(kids) && fz_isint(count))
{
obj = fz_dictgets(node, "Resources");
if (obj)
info.resources = obj;
obj = fz_dictgets(node, "MediaBox");
if (obj)
info.mediabox = obj;
obj = fz_dictgets(node, "CropBox");
if (obj)
info.cropbox = obj;
obj = fz_dictgets(node, "Rotate");
if (obj)
info.rotate = obj;
tmp = fz_newnull();
fz_dictputs(node, ".seen", tmp);
fz_dropobj(tmp);
n = fz_arraylen(kids);
for (i = 0; i < n; i++)
{
obj = fz_arrayget(kids, i);
pdf_loadpagetreenode(xref, obj, info);
}
fz_dictdels(node, ".seen");
}
else
{
dict = fz_resolveindirect(node);
if (info.resources && !fz_dictgets(dict, "Resources"))
fz_dictputs(dict, "Resources", info.resources);
if (info.mediabox && !fz_dictgets(dict, "MediaBox"))
fz_dictputs(dict, "MediaBox", info.mediabox);
if (info.cropbox && !fz_dictgets(dict, "CropBox"))
fz_dictputs(dict, "CropBox", info.cropbox);
if (info.rotate && !fz_dictgets(dict, "Rotate"))
fz_dictputs(dict, "Rotate", info.rotate);
if (xref->pagelen == xref->pagecap)
{
fz_warn("found more pages than expected");
xref->pagecap ++;
xref->pagerefs = fz_realloc(xref->pagerefs, sizeof(fz_obj*) * xref->pagecap);
xref->pageobjs = fz_realloc(xref->pageobjs, sizeof(fz_obj*) * xref->pagecap);
}
xref->pagerefs[xref->pagelen] = fz_keepobj(node);
xref->pageobjs[xref->pagelen] = fz_keepobj(dict);
xref->pagelen ++;
}
}
void pdfapp_open(pdfapp_t *app, char *filename)
{
#if 0
fz_error error;
fz_obj *obj;
char *password = "";
#else
fz_error error;
fz_stream *file;
char *password = "";
fz_obj *obj;
fz_obj *info;
int fd;
fd = open(filename, O_BINARY | O_RDONLY, 0666);
if (fd < 0)
fprintf(stderr, "error, file %s does not exist\n", filename);
#endif
/*
* Open PDF and load xref table
*/
app->filename = filename;
#if 0
app->xref = pdf_newxref();
error = pdf_loadxref(app->xref, filename);
if (error)
{
fz_catch(error, "trying to air");
error = pdf_repairxref(app->xref, filename);
if (error)
pdfapp_error(app, error);
}
error = pdf_decryptxref(app->xref);
if (error)
pdfapp_error(app, error);
#else
file = fz_open_fd(fd);
error = pdf_open_xref_with_stream(&app->xref, file, NULL);
if (error)
pdfapp_error(app, fz_rethrow(error, "cannot open document '%s'", filename));
fz_close(file);
#endif
/*
* Handle encrypted PDF files
*/
if (pdf_needs_password(app->xref))
{
int okay = pdf_authenticate_password(app->xref, password);
while (!okay)
{
//password = winpassword(app, filename);
if (!password)
exit(1);
okay = pdf_authenticate_password(app->xref, password);
if (!okay)
pdfapp_warn(app, "Invalid password.");
}
}
/*
* Load meta information
* TODO: move this into mupdf library
*/
#if 0
obj = fz_dictgets(app->xref->trailer, "Root");
app->xref->root = fz_resolveindirect(obj);
if (!app->xref->root)
pdfapp_error(app, fz_throw("syntaxerror: missing Root object"));
fz_keepobj(app->xref->root);
obj = fz_dictgets(app->xref->trailer, "Info");
app->xref->info = fz_resolveindirect(obj);
if (!app->xref->info)
pdfapp_warn(app, "Could not load PDF meta information.");
if (app->xref->info)
fz_keepobj(app->xref->info);
/*app->outline = pdf_loadoutline(app->xref);*/
app->doctitle = filename;
if (strrchr(app->doctitle, '\\'))
app->doctitle = strrchr(app->doctitle, '\\') + 1;
if (strrchr(app->doctitle, '/'))
app->doctitle = strrchr(app->doctitle, '/') + 1;
if (app->xref->info)
{
obj = fz_dictgets(app->xref->info, "Title");
if (obj)
{
app->doctitle = pdf_toutf8(obj);
}
}
#else
app->outline = pdf_load_outline(app->xref);
app->doctitle = filename;
if (strrchr(app->doctitle, '\\'))
app->doctitle = strrchr(app->doctitle, '\\') + 1;
if (strrchr(app->doctitle, '/'))
app->doctitle = strrchr(app->doctitle, '/') + 1;
info = fz_dict_gets(app->xref->trailer, "Info");
if (info)
{
obj = fz_dict_gets(info, "Title");
if (obj)
app->doctitle = pdf_to_utf8(obj);
}
#endif
/*
* Start at first page
*/
#if 0
app->pagecount = pdf_getpagecount(app->xref);
#else
error = pdf_load_page_tree(app->xref);
if (error)
pdfapp_error(app, fz_rethrow(error, "cannot load page tree"));
app->pagecount = pdf_count_pages(app->xref);
#endif
app->rotate = 0;
}
