fenix
/
cpdf-source
mirror of https://github.com/johnwhitington/cpdf-source.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Avoid inserting image data of incorrect spec

This commit is contained in:
John Whitington 2024-02-20 14:59:14 +00:00
parent cb396215c3
commit 5321eb0c1f
1 changed files with 31 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

40
cpdfimage.ml
View File

@ -628,8 +628,19 @@ let lossless_to_jpeg pdf ~pixel_threshold ~length_threshold ~percentage_threshol
remove out; remove out;
remove out2 remove out2
let test_components pdf dict =
match suitable_num pdf dict with -1 | -2 -> 1 | x -> x
let test_bpc pdf dict =
match Pdf.lookup_direct pdf "/BitsPerComponent" dict with
| Some (Pdf.Integer i) -> i
| _ -> 0
let lossless_resample pdf ~pixel_threshold ~length_threshold ~factor ~interpolate ~path_to_convert s dict reference = let lossless_resample pdf ~pixel_threshold ~length_threshold ~factor ~interpolate ~path_to_convert s dict reference =
Printf.printf "***lossless_resample IN dictionary: %S\n" (Pdfwrite.string_of_pdf dict); (* Printf.printf "***lossless_resample IN dictionary: %S\n" (Pdfwrite.string_of_pdf dict); *)
let in_components = test_components pdf dict in
let in_bpc = test_bpc pdf dict in
Printf.printf "\n***IN components = %i, bpc = %i\n" in_components in_bpc;
match lossless_out pdf ~pixel_threshold ~length_threshold ".png" s dict reference with match lossless_out pdf ~pixel_threshold ~length_threshold ".png" s dict reference with
| None -> () | None -> ()
| Some (_, _, _, 4, _, _) -> Printf.printf "lossless resampling for CMYK not supported yet\n%!" | Some (_, _, _, 4, _, _) -> Printf.printf "lossless resampling for CMYK not supported yet\n%!"
@ -651,16 +662,27 @@ let lossless_resample pdf ~pixel_threshold ~length_threshold ~factor ~interpolat
let newsize = in_channel_length result in let newsize = in_channel_length result in
if newsize < size then if newsize < size then
begin begin
if !debug_image_processing then Printf.printf "lossless resample %i -> %i (%i%%)\n%!" size newsize (int_of_float (float newsize /. float size *. 100.));
reference := reference :=
(match fst (obj_of_png_data (Pdfio.bytes_of_input_channel result)) with (match fst (obj_of_png_data (Pdfio.bytes_of_input_channel result)) with
| Pdf.Stream {contents = Pdf.Dictionary d, data} -> | Pdf.Stream {contents = Pdf.Dictionary d, data} as s ->
(* Find components of resultant colourspace, and bits per component. If differing from input, abandon *) let out_components = test_components pdf s in
(* We will then test all the files we have available, and make sure all results correct *) let out_bpc = test_bpc pdf s in
(* Then we will see about how to poke convert to do the correct thing - e.g not use PNG, force different output etc. *) Printf.printf "***OUT components = %i, bpc = %i\n" out_components out_bpc;
let d' = fold_right (fun (k, v) d -> if k <> "/ColorSpace" then add k v d else d) d (match dict with Pdf.Dictionary x -> x | _ -> []) in if out_components <> in_components || in_bpc <> out_bpc then
Printf.printf "***lossless_resample OUT dictionary: %S\n" (Pdfwrite.string_of_pdf (Pdf.Dictionary d')); begin
(Pdf.Dictionary d', data) if !debug_image_processing then Printf.printf "wrong bpc / components returned. Skipping.\n%!";
!reference
end
else
begin
if !debug_image_processing then Printf.printf "lossless resample %i -> %i (%i%%)\n%!" size newsize (int_of_float (float newsize /. float size *. 100.));
(* We will then test all the files we have available, and make sure all results correct *)
(* Then we will see about how to poke convert to do the correct thing - e.g not use PNG, force different output etc. *)
let d' = fold_right (fun (k, v) d -> if k <> "/ColorSpace" then add k v d else d) d (match dict with Pdf.Dictionary x -> x | _ -> []) in
(*Printf.printf "***lossless_resample OUT dictionary: %S\n" (Pdfwrite.string_of_pdf (Pdf.Dictionary d')); *)
(Pdf.Dictionary d', data)
end
| _ -> assert false) | _ -> assert false)
end end
else else