diff --git a/src/PDFonts.jl b/src/PDFonts.jl
index 94399df..043590b 100644
--- a/src/PDFonts.jl
+++ b/src/PDFonts.jl
@@ -22,6 +22,14 @@ const endbfrange = b"endbfrange"
 const begincodespacerange = b"begincodespacerange"
 const endcodespacerange = b"endcodespacerange"
 
+# EXTRACTION_MODE determines how text extraction of textboxes is handled
+# - `:spaces` (default)
+#   all white spaces are handled as a single space character
+# - `:tabs`
+#   non-space white spaces are handled as tab characters
+# - `:boxes`
+#   text is split into several textboxes with respective coordinates
+const EXTRACTION_MODE = Ref(:spaces) # :spaces, :tabs, :boxes
 
 mutable struct CMap
     code_space::IntervalTree{UInt8,
@@ -571,13 +579,26 @@ function get_TextBox(ss::Vector{Union{CosXString, CosLiteralString,
     totalw = 0f0
     tj = 0f0
     text = ""
+    offset = 0f0
+    params = Tuple{String, Float32, Float32, Float32}[]
     for s in ss
         if s isa CosXString || s isa CosLiteralString
             prev_char = INIT_CODE(pdfont.widths)
             t = String(get_encoded_string(s, pdfont))
-            if (-tj) > 180 && length(t) > 0 && t[1] != ' ' &&
-                length(text) > 0 && text[end] != ' '
-                text *= " "
+            if (-tj) > 180
+                if EXTRACTION_MODE[] == :spaces
+                    if length(t) > 0 && t[1] != ' ' && length(text) > 0 && text[end] != ' '
+                        text *=  " "
+                    end
+                elseif EXTRACTION_MODE[] == :tabs
+                    text *= "\t"
+                elseif EXTRACTION_MODE[] == :boxes
+                    push!(params, (text, totalw * th, tfs, offset * th))
+                    offset += totalw - tj * tfs / 1000f0
+                    text = ""
+                    tj = 0f0
+                    totalw = 0f0
+                end
             end
             text *= t
             barr = Vector{UInt8}(s)
@@ -588,8 +609,8 @@ function get_TextBox(ss::Vector{Union{CosXString, CosLiteralString,
             tj = s |> get |> Float32
         end
     end
-    totalw *= th
-    return text, totalw, tfs
+    push!(params, (text, totalw * th, tfs, offset * th))
+    return params
 end
 
 function get_character_width(cid::UInt16, w::CIDWidth)
diff --git a/src/PDPageElement.jl b/src/PDPageElement.jl
index d042345..ca9d5f6 100644
--- a/src/PDPageElement.jl
+++ b/src/PDPageElement.jl
@@ -693,20 +693,27 @@ end
     fontname, font = eval_unicode_mapping(tr, state)
     
     heap = get(state, :text_layout, Vector{TextLayout})
-    text, w, h = get_TextBox(tr.ss, font, tfs, tc, tw, th)
+    boxparams = get_TextBox(tr.ss, font, tfs, tc, tw, th)
+    h = boxparams[1][3]
 
     d = get(state, :h_profile, Dict{Int, Int})
     ih = round(Int, h*10)
-    d[ih] = get(d, ih, 0) + length(text)
-
-    tb = [0f0 0f0 1f0; w 0f0 1f0; w h 1f0; 0f0 h 1f0]*trm
+    
     if !get(state, :in_artifact, false)
-        tl = TextLayout(tb[1,1], tb[1,2], tb[2,1], tb[2,2],
-                        tb[3,1], tb[3,2], tb[4,1], tb[4,2],
-                        text, fontname, font.flags)
-        push!(heap, tl)
+        for params in boxparams
+            text = params[1]
+            w = params[2]
+            offset = params[4]
+            d[ih] = get(d, ih, 0) + length(text)
+            tb = [offset 0f0 1f0; offset + w 0f0 1f0; offset + w h 1f0; offset h 1f0]*trm
+            tl = TextLayout(tb[1,1], tb[1,2], tb[2,1], tb[2,2],
+                            tb[3,1], tb[3,2], tb[4,1], tb[4,2],
+                            text, fontname, font.flags)
+            push!(heap, tl)
+        end
     end
-    offset_text_pos!(w, 0f0, state)    
+    totalw = boxparams[end][4] + boxparams[end][2]
+    offset_text_pos!(totalw, 0f0, state)    
     return state
 end