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

First plausible typeset MCR output

This commit is contained in:
John Whitington 2024-10-04 14:32:49 +01:00
parent 31dec76f4b
commit e3beeae2a0
1 changed files with 18 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
cpdftexttopdf.ml
View File

@ -154,8 +154,6 @@ let typeset ~process_struct_tree ?subformat ?title ~papersize ~font ~fontsize te
else if process_struct_tree || subformat = Some Cpdfua.PDFUA1 then else if process_struct_tree || subformat = Some Cpdfua.PDFUA1 then
begin begin
let str = Pdf.addobj pdf Pdf.Null in let str = Pdf.addobj pdf Pdf.Null in
let p = Pdf.addobj pdf Pdf.Null in
let parent_tree = Pdf.addobj pdf Pdf.Null in
let topks = let topks =
map map
(fun parts_of_para -> (fun parts_of_para ->
@ -165,7 +163,24 @@ let typeset ~process_struct_tree ?subformat ?title ~papersize ~font ~fontsize te
Pdf.Indirect (Pdf.addobj pdf (Pdf.Dictionary [("/K", Pdf.Array ks); ("/P", Pdf.Indirect str); ("/S", Pdf.Name "/P")]))) Pdf.Indirect (Pdf.addobj pdf (Pdf.Dictionary [("/K", Pdf.Array ks); ("/P", Pdf.Indirect str); ("/S", Pdf.Name "/P")])))
nodes nodes
in in
Pdf.addobj_given_num pdf (parent_tree, Pdf.Dictionary [("/Nums", Pdf.Array [Pdf.Integer 1; Pdf.Array [Pdf.Indirect p]])]); let parent_tree =
(* We are making a map of (page number (/StructParents entry) to list of /P nodes, one for each MCID on that page in order. *)
let pairs =
map
(fun pn ->
let this_page_triples =
keep (fun (para, page, mcid) -> page = pn) tagtriples
in
Printf.printf "Page %i\n" pn;
iter
(fun (para, page, mcid) -> Printf.printf "Para index %i, MCID %i\n" para mcid)
this_page_triples;
(string_of_int pn, Pdf.Array (map (function (para, _, _) -> (List.nth topks para)) this_page_triples)))
(indx0 pages)
in
Pdf.addobj pdf (Pdftree.build_name_tree true pdf pairs)
(*Pdf.addobj pdf (Pdf.Dictionary [("/Nums", Pdf.Array [Pdf.Integer 1; Pdf.Array [Pdf.Indirect p]])])*)
in
Pdf.addobj_given_num pdf (str, Pdf.Dictionary [("/Type", Pdf.Name "/StructTreeRoot"); ("/K", Pdf.Array topks); ("/ParentTree", Pdf.Indirect parent_tree)]); Pdf.addobj_given_num pdf (str, Pdf.Dictionary [("/Type", Pdf.Name "/StructTreeRoot"); ("/K", Pdf.Array topks); ("/ParentTree", Pdf.Indirect parent_tree)]);
Pdf.replace_chain pdf ["/Root"] ("/StructTreeRoot", (Pdf.Indirect str)) Pdf.replace_chain pdf ["/Root"] ("/StructTreeRoot", (Pdf.Indirect str))
end; end;