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

more

This commit is contained in:
John Whitington 2023-03-06 19:11:12 +00:00
parent 9f50046c52
commit d00f8571d7
1 changed files with 24 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

30
cpdfcommand.ml
View File

@ -3376,13 +3376,15 @@ let warn_prerotate range pdf =
let prerotate range pdf = let prerotate range pdf =
Cpdfpage.upright ~fast:args.fast range pdf Cpdfpage.upright ~fast:args.fast range pdf
let find_composition_fonts pdf i obj marked = let find_composition_images pdf i obj marked =
match Pdf.lookup_direct pdf "/Subtype" obj with match Pdf.lookup_direct pdf "/Subtype" obj with
| Some (Pdf.Name "/Image") -> | Some (Pdf.Name "/Image") ->
Hashtbl.add marked i (); Hashtbl.add marked i ();
String.length (Pdfwrite.string_of_pdf_including_data obj); String.length (Pdfwrite.string_of_pdf_including_data obj);
| _ -> 0 | _ -> 0
let find_composition_fonts pdf i obj marked = 0
let find_composition_content_streams pdf i obj marked = let find_composition_content_streams pdf i obj marked =
match Pdf.lookup_direct pdf "/Type" obj with match Pdf.lookup_direct pdf "/Type" obj with
| Some (Pdf.Name "/Page") -> | Some (Pdf.Name "/Page") ->
@ -3402,23 +3404,39 @@ let find_composition_content_streams pdf i obj marked =
!l !l
| _ -> 0 | _ -> 0
let find_composition_structure_info pdf i obj marked = 0
let find_composition_link_annotations pdf i obj marked = 0
let find_composition_embedded_files pdf i obj marked = 0
let find_composition pdf = let find_composition pdf =
let marked = null_hash () in let marked = null_hash () in
let images = ref 0 in let images = ref 0 in
let fonts = ref 0 in
let content_streams = ref 0 in let content_streams = ref 0 in
let structure_info = ref 0 in
let link_annotations = ref 0 in
let embedded_files = ref 0 in
Pdf.objiter Pdf.objiter
(fun i obj -> (fun i obj ->
match Hashtbl.find marked i with _ -> () | exception Not_found -> match Hashtbl.find marked i with _ -> () | exception Not_found ->
images += find_composition_fonts pdf i obj marked; images += find_composition_images pdf i obj marked;
content_streams += find_composition_content_streams pdf i obj marked) fonts += find_composition_fonts pdf i obj marked;
content_streams += find_composition_content_streams pdf i obj marked;
structure_info += find_composition_structure_info pdf i obj marked;
link_annotations += find_composition_link_annotations pdf i obj marked;
embedded_files += find_composition_embedded_files pdf i obj marked)
pdf; pdf;
(!images, !content_streams) (!images, !fonts, !content_streams, !structure_info, !link_annotations, !embedded_files)
(* First go: images, fonts, content streams, structure info, link annotations, embedded files *) (* First go: images, fonts, content streams, structure info, link annotations, embedded files *)
let show_composition_json filesize pdf = let show_composition_json filesize pdf =
let perc x = float_of_int x /. float_of_int filesize *. 100. in let perc x = float_of_int x /. float_of_int filesize *. 100. in
let images, content_streams = find_composition pdf in let images, fonts, content_streams, structure_info, link_annotations, embedded_files =
let r = images + content_streams in find_composition pdf
in
let r = images + fonts + content_streams + structure_info + link_annotations + embedded_files in
`List [`Tuple [`String "Images"; `Int images; `Float (perc images)]; `List [`Tuple [`String "Images"; `Int images; `Float (perc images)];
`Tuple [`String "Content streams"; `Int content_streams; `Float (perc content_streams)]; `Tuple [`String "Content streams"; `Int content_streams; `Float (perc content_streams)];
`Tuple [`String "Unclassified"; `Int (filesize - r); `Float (perc (filesize - r))]] `Tuple [`String "Unclassified"; `Int (filesize - r); `Float (perc (filesize - r))]]