ErrorCode juggler_add_pages_from_file(juggler_t *dest, juggler_t *src, int dest_index)
{
pdf_obj *dest_pages = pdf_dict_getp(dest->ctx,
pdf_trailer(dest->ctx, dest->pdf), "Root/Pages");
int dest_pages_index = pdf_array_len(dest->ctx,
pdf_dict_gets(dest->ctx, dest_pages, "Kids"));
/* be aware that this function does not change the two variables if the page
index is greater than the number of pages */
find_destination_pages(dest->ctx,
dest_pages, dest_index, &dest_pages, &dest_pages_index);
pdf_obj *dest_kids = pdf_dict_gets(dest->ctx, dest_pages, "Kids");
if(!pdf_is_indirect(dest->ctx, dest_pages) ||
!pdf_is_dict(dest->ctx, dest_pages) || !pdf_is_array(dest->ctx, dest_kids))
{
return(ERROR_INVALID_RANGE);
}
pdf_obj *pages_root = pdf_dict_getp(src->ctx, pdf_trailer(src->ctx, src->pdf), "Root/Pages");
if(!pdf_is_indirect(src->ctx, pages_root) || !pdf_is_dict(src->ctx, pages_root))
return(ERROR_NO_PAGES);
/* if we copy the root pages-node and it's referenced objects, we will copy
all pages and all objects those pages need */
pdf_obj *new_pages_ref = copy_object_single(dest->ctx, dest->pdf, src->ctx, src->pdf, pages_root);
/* insert new pages-node */
pdf_array_insert_drop(dest->ctx, dest_kids, new_pages_ref, dest_pages_index);
/* update the parent */
pdf_obj *new_pages_parent = pdf_new_indirect(dest->ctx, dest->pdf,
pdf_to_num(dest->ctx, dest_pages), pdf_to_gen(dest->ctx, dest_pages));
pdf_dict_puts_drop(dest->ctx, new_pages_ref, "Parent", new_pages_parent);
/* TODO: If dest_pages contains anything inheritable but not the new node
we need to insert empty items to prevent this inerhitance */
/* update count */
int new_count = pdf_to_int(dest->ctx,
pdf_dict_gets(dest->ctx, dest_pages, "Count")) + src->pagecount;
pdf_dict_puts_drop(dest->ctx, dest_pages, "Count",
pdf_new_int(dest->ctx, dest->pdf, new_count));
/* let MuPDF rebuild the page tree */
pdf_finish_edit(dest->ctx, dest->pdf);
dest->pdf->page_count = new_count;
/* update juggler's state */
juggler_page_tree_changed_due_to_insert(dest, dest_index, src->pagecount);
return(NoError);
}
static int
pdf_xobject_uses_blending(pdf_document *doc, pdf_obj *dict)
{
pdf_obj *obj = pdf_dict_gets(dict, "Resources");
/* cf. http://code.google.com/p/sumatrapdf/issues/detail?id=2540 */
if (!strcmp(pdf_to_name(pdf_dict_getp(dict, "Group/S")), "Transparency"))
return 1;
return pdf_resources_use_blending(doc, obj);
}
int
pdf_count_pages(pdf_document *doc)
{
if (doc->page_count == 0)
{
pdf_obj *count = pdf_dict_getp(pdf_trailer(doc), "Root/Pages/Count");
doc->page_count = pdf_to_int(count);
}
return doc->page_count;
}
static wchar_t *get_prop_str(fz_context* ctx, pdf_obj *obj, const char *name)
{
pdf_obj* p_obj = pdf_dict_getp(ctx, obj, name);
int len = pdf_to_str_len(ctx, p_obj) + 1;
wchar_t *buf = AllocArray<wchar_t>(len);
pdf_to_ucs2_buf(ctx, (unsigned short *)buf, p_obj);
return p_obj ? pdf_clean_string(buf) : nullptr;
}
pdfout_data *
pdfout_page_labels_get (fz_context *ctx, pdf_document *doc)
{
pdf_obj *trailer = pdf_trailer (ctx, doc);
pdf_obj *labels_obj = pdf_dict_getp (ctx, trailer, "Root/PageLabels");
pdf_obj *array = pdf_dict_gets (ctx, labels_obj, "Nums");
if (array && pdf_is_array (ctx, array) == false)
pdfout_throw (ctx, "Nums is not an array");
int length = pdf_array_len (ctx, array);
pdfout_data *labels = pdfout_data_array_new (ctx);
for (int i = 0; i < length / 2; ++i)
{
pdf_obj *object = pdf_array_get (ctx, array, 2 * i);
if (pdf_is_int (ctx, object) == false)
pdfout_throw (ctx, "key in number tree not an int");
pdfout_data *hash = pdfout_data_hash_new (ctx);
int page = pdf_to_int (ctx, object);
if (page < 0)
pdfout_throw (ctx, "key in number tree is < 0");
push_int_key (ctx, hash, "page", page);
pdf_obj *dict = pdf_array_get (ctx, array, 2 * i + 1);
if (pdf_is_dict (ctx, dict) == false)
pdfout_throw (ctx, "value in number tree not a dict");
parse_dict (ctx, dict, hash);
pdfout_data_array_push (ctx, labels, hash);
}
return labels;
}
/*
When resetting or submitting a form, the fields to act upon are defined
by an array of either field references or field names, plus a flag determining
whether to act upon the fields in the array, or all fields other than those in
the array. specified_fields interprets this information and produces the array
of fields to be acted upon.
*/
static pdf_obj *specified_fields(pdf_document *doc, pdf_obj *fields, int exclude)
{
fz_context *ctx = doc->ctx;
pdf_obj *form = pdf_dict_getp(pdf_trailer(doc), "Root/AcroForm/Fields");
int i, n;
pdf_obj *result = pdf_new_array(doc, 0);
pdf_obj *nil = NULL;
fz_var(nil);
fz_try(ctx)
{
/* The 'fields' array not being present signals that all fields
* should be acted upon, so handle it using the exclude case - excluding none */
if (exclude || !fields)
{
/* mark the fields we don't want to act upon */
nil = pdf_new_null(doc);
n = pdf_array_len(fields);
for (i = 0; i < n; i++)
{
pdf_obj *field = pdf_array_get(fields, i);
if (pdf_is_string(field))
field = pdf_lookup_field(form, pdf_to_str_buf(field));
if (field)
pdf_dict_puts(field, "Exclude", nil);
}
/* Act upon all unmarked fields */
n = pdf_array_len(form);
for (i = 0; i < n; i++)
add_field_hierarchy_to_array(result, pdf_array_get(form, i));
/* Unmark the marked fields */
n = pdf_array_len(fields);
for (i = 0; i < n; i++)
{
pdf_obj *field = pdf_array_get(fields, i);
if (pdf_is_string(field))
field = pdf_lookup_field(form, pdf_to_str_buf(field));
if (field)
pdf_dict_dels(field, "Exclude");
}
}
else
{
n = pdf_array_len(fields);
for (i = 0; i < n; i++)
{
pdf_obj *field = pdf_array_get(fields, i);
if (pdf_is_string(field))
field = pdf_lookup_field(form, pdf_to_str_buf(field));
if (field)
add_field_hierarchy_to_array(result, field);
}
}
}
fz_always(ctx)
{
pdf_drop_obj(nil);
}
fz_catch(ctx)
{
pdf_drop_obj(result);
fz_rethrow(ctx);
}
return result;
}
pdf_page *
pdf_load_page_by_obj(pdf_document *doc, int number, pdf_obj *pageref)
{
fz_context *ctx = doc->ctx;
pdf_page *page;
pdf_annot *annot;
pdf_obj *pageobj, *obj;
fz_rect mediabox, cropbox, realbox;
float userunit;
fz_matrix mat;
/* SumatraPDF: allow replacing potentially slow pdf_lookup_page_obj */
pageobj = pdf_resolve_indirect(pageref);
page = fz_malloc_struct(ctx, pdf_page);
page->resources = NULL;
page->contents = NULL;
page->transparency = 0;
page->links = NULL;
page->annots = NULL;
page->annot_tailp = &page->annots;
page->deleted_annots = NULL;
page->tmp_annots = NULL;
page->me = pdf_keep_obj(pageobj);
page->incomplete = 0;
obj = pdf_dict_gets(pageobj, "UserUnit");
if (pdf_is_real(obj))
userunit = pdf_to_real(obj);
else
userunit = 1;
pdf_to_rect(ctx, pdf_lookup_inherited_page_item(doc, pageobj, "MediaBox"), &mediabox);
if (fz_is_empty_rect(&mediabox))
{
fz_warn(ctx, "cannot find page size for page %d", number + 1);
mediabox.x0 = 0;
mediabox.y0 = 0;
mediabox.x1 = 612;
mediabox.y1 = 792;
}
pdf_to_rect(ctx, pdf_lookup_inherited_page_item(doc, pageobj, "CropBox"), &cropbox);
if (!fz_is_empty_rect(&cropbox))
fz_intersect_rect(&mediabox, &cropbox);
page->mediabox.x0 = fz_min(mediabox.x0, mediabox.x1) * userunit;
page->mediabox.y0 = fz_min(mediabox.y0, mediabox.y1) * userunit;
page->mediabox.x1 = fz_max(mediabox.x0, mediabox.x1) * userunit;
page->mediabox.y1 = fz_max(mediabox.y0, mediabox.y1) * userunit;
if (page->mediabox.x1 - page->mediabox.x0 < 1 || page->mediabox.y1 - page->mediabox.y0 < 1)
{
fz_warn(ctx, "invalid page size in page %d", number + 1);
page->mediabox = fz_unit_rect;
}
page->rotate = pdf_to_int(pdf_lookup_inherited_page_item(doc, pageobj, "Rotate"));
/* Snap page->rotate to 0, 90, 180 or 270 */
if (page->rotate < 0)
page->rotate = 360 - ((-page->rotate) % 360);
if (page->rotate >= 360)
page->rotate = page->rotate % 360;
page->rotate = 90*((page->rotate + 45)/90);
if (page->rotate > 360)
page->rotate = 0;
fz_pre_rotate(fz_scale(&page->ctm, 1, -1), -page->rotate);
realbox = page->mediabox;
fz_transform_rect(&realbox, &page->ctm);
fz_pre_scale(fz_translate(&mat, -realbox.x0, -realbox.y0), userunit, userunit);
fz_concat(&page->ctm, &page->ctm, &mat);
fz_try(ctx)
{
obj = pdf_dict_gets(pageobj, "Annots");
if (obj)
{
page->links = pdf_load_link_annots(doc, obj, &page->ctm);
pdf_load_annots(doc, page, obj);
}
}
fz_catch(ctx)
{
if (fz_caught(ctx) != FZ_ERROR_TRYLATER)
/* SumatraPDF: ignore annotations in case of unexpected errors */
fz_warn(ctx, "unexpectedly failed to load page annotations");
page->incomplete |= PDF_PAGE_INCOMPLETE_ANNOTS;
}
page->duration = pdf_to_real(pdf_dict_gets(pageobj, "Dur"));
obj = pdf_dict_gets(pageobj, "Trans");
page->transition_present = (obj != NULL);
if (obj)
{
pdf_load_transition(doc, page, obj);
}
// TODO: inherit
page->resources = pdf_lookup_inherited_page_item(doc, pageobj, "Resources");
if (page->resources)
pdf_keep_obj(page->resources);
obj = pdf_dict_gets(pageobj, "Contents");
fz_try(ctx)
{
page->contents = pdf_keep_obj(obj);
if (pdf_resources_use_blending(doc, page->resources))
page->transparency = 1;
/* cf. http://code.google.com/p/sumatrapdf/issues/detail?id=2107 */
else if (!strcmp(pdf_to_name(pdf_dict_getp(pageobj, "Group/S")), "Transparency"))
page->transparency = 1;
for (annot = page->annots; annot && !page->transparency; annot = annot->next)
if (annot->ap && pdf_resources_use_blending(doc, annot->ap->resources))
page->transparency = 1;
}
fz_catch(ctx)
{
if (fz_caught(ctx) != FZ_ERROR_TRYLATER)
{
pdf_free_page(doc, page);
fz_rethrow_message(ctx, "cannot load page %d contents (%d 0 R)", number + 1, pdf_to_num(pageref));
}
page->incomplete |= PDF_PAGE_INCOMPLETE_CONTENTS;
}
return page;
}
char *
pdf_parse_link_action(fz_context *ctx, pdf_document *doc, pdf_obj *action, int pagenum)
{
pdf_obj *obj, *dest, *file_spec;
if (!action)
return NULL;
obj = pdf_dict_get(ctx, action, PDF_NAME_S);
if (pdf_name_eq(ctx, PDF_NAME_GoTo, obj))
{
dest = pdf_dict_get(ctx, action, PDF_NAME_D);
return pdf_parse_link_dest(ctx, doc, dest);
}
else if (pdf_name_eq(ctx, PDF_NAME_URI, obj))
{
/* URI entries are ASCII strings */
const char *uri = pdf_to_str_buf(ctx, pdf_dict_get(ctx, action, PDF_NAME_URI));
if (!fz_is_external_link(ctx, uri))
{
pdf_obj *uri_base_obj = pdf_dict_getp(ctx, pdf_trailer(ctx, doc), "Root/URI/Base");
const char *uri_base = uri_base_obj ? pdf_to_str_buf(ctx, uri_base_obj) : "file://";
char *new_uri = fz_malloc(ctx, strlen(uri_base) + strlen(uri) + 1);
strcpy(new_uri, uri_base);
strcat(new_uri, uri);
return new_uri;
}
return fz_strdup(ctx, uri);
}
else if (pdf_name_eq(ctx, PDF_NAME_Launch, obj))
{
file_spec = pdf_dict_get(ctx, action, PDF_NAME_F);
return pdf_parse_file_spec(ctx, doc, file_spec, NULL);
}
else if (pdf_name_eq(ctx, PDF_NAME_GoToR, obj))
{
dest = pdf_dict_get(ctx, action, PDF_NAME_D);
file_spec = pdf_dict_get(ctx, action, PDF_NAME_F);
return pdf_parse_file_spec(ctx, doc, file_spec, dest);
}
else if (pdf_name_eq(ctx, PDF_NAME_Named, obj))
{
dest = pdf_dict_get(ctx, action, PDF_NAME_N);
if (pdf_name_eq(ctx, PDF_NAME_FirstPage, dest))
pagenum = 0;
else if (pdf_name_eq(ctx, PDF_NAME_LastPage, dest))
pagenum = pdf_count_pages(ctx, doc) - 1;
else if (pdf_name_eq(ctx, PDF_NAME_PrevPage, dest) && pagenum >= 0)
{
if (pagenum > 0)
pagenum--;
}
else if (pdf_name_eq(ctx, PDF_NAME_NextPage, dest) && pagenum >= 0)
{
if (pagenum < pdf_count_pages(ctx, doc) - 1)
pagenum++;
}
else
return NULL;
return fz_asprintf(ctx, "#%d", pagenum + 1);
}
return NULL;
}
char *pdf_annot_contents(pdf_document *doc, pdf_annot *annot)
{
return pdf_to_str_buf(pdf_dict_getp(annot->obj, "Contents"));
}
static fz_image *
pdf_load_image_imp(pdf_document *xref, pdf_obj *rdb, pdf_obj *dict, fz_stream *cstm, int forcemask)
{
fz_stream *stm = NULL;
fz_image *image = NULL;
pdf_obj *obj, *res;
int w, h, bpc, n;
int imagemask;
int interpolate;
int indexed;
fz_image *mask = NULL; /* explicit mask/soft mask image */
int usecolorkey = 0;
fz_colorspace *colorspace = NULL;
float decode[FZ_MAX_COLORS * 2];
int colorkey[FZ_MAX_COLORS * 2];
int i;
fz_context *ctx = xref->ctx;
fz_var(stm);
fz_var(mask);
fz_var(image);
fz_var(colorspace);
fz_try(ctx)
{
/* special case for JPEG2000 images */
if (pdf_is_jpx_image(ctx, dict))
{
image = pdf_load_jpx(xref, dict, forcemask);
if (forcemask)
{
fz_pixmap *mask_pixmap;
if (image->n != 2)
{
/* SumatraPDF: ignore invalid JPX softmasks */
fz_warn(ctx, "soft mask must be grayscale");
mask_pixmap = fz_new_pixmap(ctx, NULL, image->tile->w, image->tile->h);
fz_clear_pixmap_with_value(ctx, mask_pixmap, 255);
}
else
mask_pixmap = fz_alpha_from_gray(ctx, image->tile, 1);
fz_drop_pixmap(ctx, image->tile);
image->tile = mask_pixmap;
}
break; /* Out of fz_try */
}
w = pdf_to_int(pdf_dict_getsa(dict, "Width", "W"));
h = pdf_to_int(pdf_dict_getsa(dict, "Height", "H"));
bpc = pdf_to_int(pdf_dict_getsa(dict, "BitsPerComponent", "BPC"));
if (bpc == 0)
bpc = 8;
imagemask = pdf_to_bool(pdf_dict_getsa(dict, "ImageMask", "IM"));
interpolate = pdf_to_bool(pdf_dict_getsa(dict, "Interpolate", "I"));
indexed = 0;
usecolorkey = 0;
mask = NULL;
if (imagemask)
bpc = 1;
if (w <= 0)
fz_throw(ctx, "image width is zero (or less)");
if (h <= 0)
fz_throw(ctx, "image height is zero (or less)");
if (bpc <= 0)
fz_throw(ctx, "image depth is zero (or less)");
if (bpc > 16)
fz_throw(ctx, "image depth is too large: %d", bpc);
if (w > (1 << 16))
fz_throw(ctx, "image is too wide");
if (h > (1 << 16))
fz_throw(ctx, "image is too high");
obj = pdf_dict_getsa(dict, "ColorSpace", "CS");
if (obj && !imagemask && !forcemask)
{
/* colorspace resource lookup is only done for inline images */
if (pdf_is_name(obj))
{
res = pdf_dict_get(pdf_dict_gets(rdb, "ColorSpace"), obj);
if (res)
obj = res;
}
colorspace = pdf_load_colorspace(xref, obj);
if (!strcmp(colorspace->name, "Indexed"))
indexed = 1;
n = colorspace->n;
}
else
{
n = 1;
}
obj = pdf_dict_getsa(dict, "Decode", "D");
if (obj)
{
for (i = 0; i < n * 2; i++)
decode[i] = pdf_to_real(pdf_array_get(obj, i));
}
else
{
float maxval = indexed ? (1 << bpc) - 1 : 1;
for (i = 0; i < n * 2; i++)
decode[i] = i & 1 ? maxval : 0;
}
obj = pdf_dict_getsa(dict, "SMask", "Mask");
if (pdf_is_dict(obj))
{
/* Not allowed for inline images or soft masks */
if (cstm)
fz_warn(ctx, "Ignoring invalid inline image soft mask");
else if (forcemask)
fz_warn(ctx, "Ignoring recursive image soft mask");
else
mask = pdf_load_image_imp(xref, rdb, obj, NULL, 1);
}
else if (pdf_is_array(obj))
{
usecolorkey = 1;
for (i = 0; i < n * 2; i++)
{
if (!pdf_is_int(pdf_array_get(obj, i)))
{
fz_warn(ctx, "invalid value in color key mask");
usecolorkey = 0;
}
colorkey[i] = pdf_to_int(pdf_array_get(obj, i));
}
}
/* Now, do we load a ref, or do we load the actual thing? */
if (!cstm)
{
/* Just load the compressed image data now and we can
* decode it on demand. */
int num = pdf_to_num(dict);
int gen = pdf_to_gen(dict);
fz_compressed_buffer *buffer = pdf_load_compressed_stream(xref, num, gen);
image = fz_new_image(ctx, w, h, bpc, colorspace, 96, 96, interpolate, imagemask, decode, usecolorkey ? colorkey : NULL, buffer, mask);
break; /* Out of fz_try */
}
/* We need to decompress the image now */
if (cstm)
{
int stride = (w * n * bpc + 7) / 8;
stm = pdf_open_inline_stream(xref, dict, stride * h, cstm, NULL);
}
else
{
stm = pdf_open_stream(xref, pdf_to_num(dict), pdf_to_gen(dict));
}
image = fz_new_image(ctx, w, h, bpc, colorspace, 96, 96, interpolate, imagemask, decode, usecolorkey ? colorkey : NULL, NULL, mask);
image->tile = fz_decomp_image_from_stream(ctx, stm, image, cstm != NULL, indexed, 0, 0);
}
fz_catch(ctx)
{
/* SumatraPDF: fix memory leak */
if (!image)
fz_drop_colorspace(ctx, colorspace);
else
fz_drop_image(ctx, image);
fz_rethrow(ctx);
}
/* cf. http://bugs.ghostscript.com/show_bug.cgi?id=693517 */
fz_try(ctx)
{
obj = pdf_dict_getp(dict, "SMask/Matte");
if (pdf_is_array(obj) && image->mask)
{
assert(!image->usecolorkey);
image->usecolorkey = 2;
for (i = 0; i < n; i++)
image->colorkey[i] = pdf_to_int(pdf_array_get(obj, i));
}
}
fz_catch(ctx)
{
fz_drop_image(ctx, image);
fz_rethrow(ctx);
}
return image;
}