I'm experiencing a real difficulty when trying to compute (tx,ty) position of text objects from a parsed PDF stream.
I have a following stream code:
BT
0.75 0.68 0.67 0.902 k
/GS0 gs
/TT0 1 Tf
-0.018 Tc 7.56 0 0 7.56 77.1871 528.3107 Tm
(Text line 1)Tj
-0.019 Tc 0 -1.917 TD
(Text line 2)Tj
-0.017 Tc 0 -2.917 TD
(HEADER)Tj
ET
q
43.167 489.881 7.56 7.56 re
W n
BT
/TT0 1 Tf
0 Tc 7.56 0 0 7.56 43.1671 491.7707 Tm
(INDEX)Tj
ET
When I open this PDF in some PDF reader, the HEADER and INDEX objects appear exactly next to each other, as they were in the same line.
However, when calculating HEADER's ty value from previous Tm (528,3107), I get 491,7657 which is 0,0050 lower than INDEX's ty (491,7707). In other parts of file the more text paragraph has the greater is this difference.
Basically what I do is multiply Tm's scaling factor (7,56) by TD's ty deltas. Obiously, I'am doing it wrong, but still - on the net there is little docs for dummies like myself...
So my question is - how to the other PDF readers interpret HEADER and INDEX ty values as equal, so they print it at the same ty?
Related
I have to write a multi lingual text a pdf using C++. I have unicode values as well as glyph id values with their advances and displacements for the string input.
But I need to know how to position the dependent glyph with the independent base glyph.
Suppose if I have a advance and displacement values using FreeType / HarfBuzz, how should I input these values into the pdf content stream along with the glyph ids in the input.
I have tried the output values of FreeType & HarfBuzz, which could print the individual glyphs properly, but the positioning of the glyphs with its base glyph is not proper still, even if i used the advance and displacement values given in their outputs.
I just need the logic of how to use the output values in the content stream to deliver a proper readable word/letter.
Example:
Text = tamil letter + hindi letter.
I need to print this output.proper output
But currently only I am able to print this. improper output
Tamil combined letter:
வ = U+0BB5 TAMIL LETTER VA = base glyph
ா = U+0BBE TAMIL VOWEL SIGN AA = dependent glyph
HarfBuzz run:
hb-shape.exe -O json -u u+0bb5,u+0bbe --no-glyph-names "C:\\Windows\\Fonts\\Nirmala.ttf"
gid output:
[{"g":2953,"cl":0,"dx":0,"dy":0,"ax":2111,"ay":0},{"g":2959,"cl":0,"dx":0,"dy":0,"ax":1453,"ay":0}]
Hindi combined letter:
म = U+092E DEVANAGARI LETTER MA = base glyph
ि = U+093F DEVANAGARI VOWEL SIGN I = dependent glyph
HarfBuzz run:
hb-shape.exe -O json -u u+092e,u+093f --no-glyph-names "C:\\Windows\\Fonts\\Nirmala.ttf"
gid output:
[{"g":302,"cl":0,"dx":0,"dy":0,"ax":532,"ay":0},{"g":273,"cl":0,"dx":0,"dy":0,"ax":1379,"ay":0}]
Subjecting these output values into the formula,
PDF doc formula
Assuming unity for all variables except width and advance,
by obtaining the width value using FreeType and computing them.
Glyph Advance values for four glyphs in order:
tx = 1769
tx = 1132
tx = 1586
tx = 1448
If I provide these values in the content stream in the order as
<glyph id 1> tx 1 <glyph id 2> tx 2 <glyph id 3> tx 3 <glyph id 4> tx 4
Content stream:
/OC /oc2 BDC q BT /FXF1 1 Tf 70.866142 0.000000 0.000000 70.866142 28.346457 141.732285 Tm[<0B89>-1769<0B8F>-1132<0111>-1586<012E>-1448]TJ ET Q EMC
PDF Doc says (+)ve value of advances will move the text towards left.
Is it other way...?
Or if the difference of the advances is to be obtained...?
Additional PDF objects:
Font descriptor object,Base font object,Font object.
I have tried using only advance values and only computed values also.
The only problem is the horizontal & vertical space within combined glyphs, which also affects the spacing between subsequent glyphs.
Any of these does not render the glyphs as legible, atleast in a generalised programmatic manner.
From my analysis of #mkl at various stack overflow places, I suspect the need for individual transformation matrix or Td for each glyph. But is it that complex...?
As per my thought, it must be easily be rendered.
If individual transformation matrix or Td is the need, then how to compute the values to be supplied in for them.
Any help & guidance is welcome and much appreciated.
Thank you.
It helps to work out pdf as plain text you can compile by save in notepad.
Here I am altering a batch.cmd (work in progress :-) to test my compiler handles the changes as text but you can use raw pdf in editor too. beware cut and paste may need a value or two changed Also unknown yet how you can easily reference non Latin fonts (next hurdle after images, which are almost done), so I used "symbol" font as illustrative of those positioning mods.
Note for specific queries #mkl is the expert I simply do programming by examples, that function not by the book.
%PDF-1.0
%µ¶µ¶
1 0 obj<</Type/Catalog/Pages 2 0 R>>endobj
2 0 obj<</Type/Pages/Count 1/Kids[3 0 R]>>endobj
3 0 obj<</Type/Page/Parent 2 0 R/MediaBox [0 0 594 792]/Resources<</Font<< /F1 4 0 R /F2 5 0 R>>>>/Contents 6 0 R>>endobj
4 0 obj<</Type/Font/Subtype/Type1/BaseFont/Helvetica>>endobj
5 0 obj<</Type/Font/Subtype/Type1/BaseFont/Symbol>>endobj
%Comment the following /Length 0999 is a dummy value it should be altered to equal decimal stream length, but most readers will ignore or work around invalid
6 0 obj<</Length 1326>>
stream
q
BT /F1 20 Tf 072 740 Td (20 units (default units usually = pts) high Headline) Tj ET
BT /F1 16 Tf 036 700 Td (All text is "Body" text. (no heads or tails)) Tj ET
BT /F1 10 Tf 004 780 Td (Text can be any order see "Body" text above. (Printed by Filename="C:\Users\K\Downloads\Programming\CMDaPDF\MAKE2PDF.cmd") spot the escape errors) Tj ET
BT /F1 12 Tf 036 675 Td (Here # 12 units high you must include just enough text for parts of a line. PDF has no page feeds no wrapping,) Tj 0 -20 Td (nor \\new line feed, no ¶aragraphs) Tj 86 -15 Td (nor carriage \r\\return. \n\r ) Tj 100 5 Td ( It is not \007\010\011\012\\tabular, each page is one row of multiple pages,) Tj 50 -15 Td (each page is one text column wide .[ ×] no yes check) Tj 0 -10 Td (each row is one text column wide .[x] no is yes) Tj 0 -10 Td (each row is one text column wide . · bullet point OK) Tj ET
BT +0.50 Tc -1.4 Tw 999 TL /F1 1 Tf 15 001 10. 30 200.000 440.000 Tm [(Jane A)600(usten)] TJ ET
BT +0.50 Tc 0.00 Tw 000 TL /F2 1 Tf 15 000 000 15 200.000 430.000 Tm [(Ja)-1000(ne Austen)] TJ ET
BT -1.20 Tc 0.00 Tw 999 TL /F2 1 Tf 15 000 000 15 200.000 420.000 Tm [(J)-1200(a)800(ne Austen)] TJ ET
BT +0.00 Tc 0.00 Tw 000 TL /F2 1 Tf 15 000 000 15 200.000 410.000 Tm [(Jane A)100(us)-500(ten)] TJ ET
Q
endstream
xref
0 7
0000000000 65535 f
0000000019 00000 n
0000000065 00000 n
0000000117 00000 n
0000000242 00000 n
0000000306 00000 n
0000000527 00000 n
trailer<</Size 7/Root 1 0 R>>
startxref
1903
%%EOF
I have this s variable string:
ID#9NAME#9VALUE
How this string look like in PDF?
(ID) Tj (NAME) Tj (VALUE) Tj
I have to convert the s variable to PDF string.
How can I change the #9 character to a working tabstop character?
I can change the #9 character to 7 pieces of #20, but it is not good for me, because I and W are different widths.
Is there a trick?
Like horizontal spacing in percent?
(ID) Tj
some code that spacing 100 horizontal pixels
(NAME) Tj
some code that spacing 100 horizontal pixels
(VALUE) Tj
Your #9 seems to be ^09 i.e. (HT)
That should be x09 in (Base 16 / hex.) or \011 (base 8) or \t in literal string
IF defined like that in a base font, then you should be able to insert that.
(ID\t\tNAME\t\tVALUE) TJ
or
(ID\011\011NAME\011\011VALUE) Tj
However as pointed out by #mkl those were traditional mechanical printer carriage stops that could be set ON at 4 or 8 characters from line left or anything the printer operator chose to place indents or columns. Thus in a Word Processor are highly variable in number and position. But in a PDF are usually ignored.
In PDF it is more conventional to set each block of characters at a new x,y position, where y is constant for each text block at that elevation.
So for a tab stop approach with tabs at one inch (based on default 1 unit =1/72") try this
stream
q
BT
/F1 12 Tf
1 0 0 1 144 720 Tm
(ID) Tj
ET
BT
/F1 12 Tf
1 0 0 1 216 720 Tm
(NAME) Tj
ET
BT
/F1 12 Tf
1 0 0 1 288 720 Tm
(VALUE) Tj
ET
Q
endstream
Remember in PDF all whitespace is equal, but some is more so than others.
so here find id name value accepts the non existent tabs as a single white space:-
Finally to answer your query you can set fixed space from the start of a text to the start of another just like tab stops using Td. Note I had intentionally used TD and Td mixed to show it does not matter in such cases :-), however the Human readable convention is to use CAPS for the Object (Noun) and lowercase the "action" (Verb) so Td is better for debugging by eye.
This can be written as suggested by #mkl (with me adding a start point)
50 800 Td (ID) Tj 100 0 Td (NAME) Tj 100 0 Td (VALUE) Tj
In comments you asked about adding lines and the simplest, for line by line loop programming, is to use something like this. (In this case skipping 780) and contra my above comment BT and ET are usually both CAPS.
BT 50 800 Td (ID) Tj 100 0 Td (NAME) Tj 100 0 Td (VALUE) Tj ET
BT 50 760 Td (A1) Tj 100 0 Td (Example) Tj 100 0 Td (2000) Tj ET
BT 50 740 Td (B2) Tj 100 0 Td (Another) Tj 100 0 Td (1000) Tj ET
I'm trying to write a justified paragraph in a PDF with multiple formats (sizes, italics, bold, colors) and to achieve this I could use the Graphics State Stack to avoid repeating text operators, but it seems that the behavior of the Graphics State Stack depends on the PDF reader. Do I have to repeat the text operators every time I want to change text format? or there is a better way to achieve what I need?
I have the following PDF stream to test the Graphics State Stack of PDF:
BT
1 0 0 1 56.69 785.19 Tm
0 -12.1 Td
/F1 11 Tf
1.79 Tw
(rzo motáúe issjstózñ x vasreqyxñ ómfzzííh nohé hábúgíoújé úyz túit k wf ñxaóúgsri rcémwewá)Tj
0 -16.5 Td
5.1 Tw
(óaxhkd óáfythra)Tj
q
/F2bi 10 Tf
0.835 0.283 0.833 rg
4.1718 Tw
(olvéd cjtwymelgv stzr cc uxnugtqúic)Tj
q
/F3b 15 Tf
0.491 0.895 0.74 rg
15 Tw
(q hwvúñóál íu vpfíxht)Tj
0 -16.5 Td
(qfyébávrx vkámday)Tj
Q
Q
(cúprnfr úhwñ rá wdwñ óyxáumvpn nmrdó)Tj ET
In the Ubuntu PDF reader the q operator doesn't affect the Td operator.
In the Chrome PDF reader the q operator do affect the Td operator.
The save and restore graphics state operators are not allowed in text objects, i.e. q and Q between BT and ET is invalid.
Thus, your content stream is invalid and the behavior of pdf viewers attempting to display it is undefined.
I'm struggling with the calculation of pdf text position. I've read through internet, but without success.
I've read the pdf reference, but it's uhmmmm.
Now, I write a piece of pdf and I'll tell you my thinking about it...
/TT3 1 Tf
11.9951 0 0 12 71.94 420.9803 Tm //scale x=11.9951 scaley=12 x position=71.94 y position=420.9803
<0003>Tj
/TT2 1 Tf
1.6657 -1.22 TD //x position=71.94+1.6657 y position=420.9803-1.22
-.0016 Tc
(2\))Tj //x position=71.94+1.6657-0.0016 y position=420.9803-1.22
/TT6 1 Tf
.8203 0 TD //x position=71.94+1.6657-0.0016+0.8203
0 Tc
( )Tj
/TT3 1 Tf
10.016 0 0 10.02 71.94 237.6803 Tm //x position=71.94 y position=237.6803
<0003>Tj
/TT2 1 Tf
I'm sure that something is wrong, because with this method the sequence is not rebuilt correctly.
Thank you a lot.
First of all you shouldn't try to keep track of multiple scalars separately but instead of the whole current text matrix and text line matrix.
Then you ripped those instructions from the context. So we have to assume no relevant instructions preceded them.
Thus, we start with text matrix and text line matrix equal to the identity matrix.
/TT3 1 Tf
11.9951 0 0 12 71.94 420.9803 Tm
This sets both the text matrix and the text line matrix to
11.9951 0 0
0 12 0
71.94 420.9803 1
Then
<0003>Tj
advances the text matrix by the width of this string. As I don't know the metrics of TT3, I cannot tell the resulting text matrix.
/TT2 1 Tf
1.6657 -1.22 TD
This multiplies
1 0 0
0 1 0
1.6657 -1.22 1
to the text line matrix from the left, resulting in the new text line matrix (slightly rounded)
11.9951 0 0
0 12 0
91.92 406.3403 1
Then the text matrix is set to this value, too.
(2\))Tj
This advances the text matrix by the width of this string. As I don't know the metrics of TT2, I cannot tell the resulting text matrix. I don't even know whether this string in TT2 represents one or two glyphs. Thus, I cannot tell how often the character spacing is applied.
/TT6 1 Tf
.8203 0 TD
This multiplies
1 0 0
0 1 0
0.8203 0 1
to the text line matrix from the left, resulting in the new text line matrix (slightly rounded)
11.9951 0 0
0 12 0
101.76 406.3403 1
Now
0 Tc
( )Tj
This advances the text matrix by the width of this string. As I don't know the metrics of TT6, I cannot tell the resulting text matrix.
/TT3 1 Tf
10.016 0 0 10.02 71.94 237.6803 Tm
This sets both the text matrix and the text line matrix to
10.016 0 0
0 10.02 0
71.94 237.6803 1
I am on ubuntu.
I have a pdf file with pages divided into a grid. Each block of the grid contains name/age/dob/photo of a candidate. some records have a watermark "disqualified"
I need to scrape his pdf, with disqualified candidates in a separate list.
Using pyPdf I was able to get individual records, but it also includes watermarked candidates.
How to detect the watermark? If I can get the coordinates of the watermark, how do I match it with the candidate?
I am open to solutions other than python pyPdf
(Actually this is not an answer but merely an analysis to bit for a comment.)
I don't know pyPdf (or any python PDF classes) myself, but here is how the watermark is created for a sample entry; based upon this, anyone knowing pyPDF well enough, may more easily advice.
The Roundup
Depending on how pyPDF (or other python PDF classes) allows access to the page content, there are two major basic approaches:
If the class returns information on content (text and image) in their order in the page content stream: The watermark image xobject is referred to right before the data of the entry. Thus, any entry preceded by the drawing of a xobject image is marked.
If otherwise the information are not given in the order indicated by the page content stream, coordinate comparison must be used which per se is quite straight forward. In that case it might be of interest that the images are inserted with a [0.1 0 0 0.1 0 0] transformation matrix in action while the text is drawn with an identity transformation matrix.
The Details
This is entry # 200; the other watermarked entry is constructed similarly:
Watermarking is done by means of an image xobject. There is but one image xobject defined for the page used by both watermarked entries:
4 0 obj
<</Type/Page/MediaBox [0 0 595 841]
/Rotate 0/Parent 3 0 R
/Resources<</ProcSet[/PDF /ImageC /ImageI /Text]
/ColorSpace 18 0 R
/ExtGState 19 0 R
/XObject 20 0 R
/Font 21 0 R
>>
/Contents 5 0 R
>>
endobj
20 0 obj
<</R17
17 0 R>>
endobj
17 0 obj
<</Subtype/Image
/ColorSpace 16 0 R
/Width 128
/Height 88
/BitsPerComponent 8
/Filter/FlateDecode/Length 463>>stream
[...]
endstream
endobj
In the content stream this xobject /R17 is inserted right before the data of the entry itself is drawn:
q 0.1 0 0 0.1 0 0 cm
[...]
q 1045 0 0 495 462.5 6510.5 cm
/R17 Do
Q
q
10 0 0 10 0 0 cm BT
0.000487366 Tc
/R10 8 Tf
1 0 0 1 86 650.75 Tm
(Sex : Male)Tj
0.000304794 Tc
-64 0 Td
(Age : 43)Tj
-0.000140686 Tc
-1 11.05 Td
(House No :)Tj
-0.00002085 Tc
1 31.95 Td
(Name :)Tj
0.00008575 Tc
/R12 7.15 Tf
25.5 17.8 Td
( 200 )Tj
ET
Q
1547.5 6475 485 535.5 re
S
q
10 0 0 10 0 0 cm BT
-0.000403137 Tc
/R14 8 Tf
1 0 0 1 145.1 708.5 Tm
(XVX0001081)Tj
0.000421651 Tc
/R14 7.05 Tf
-90.35 -14.95 Td
(Ramesh Kumar)Tj
0.000373332 Tc
/R10 7.05 Tf
-33 -12.75 Td
(Father's )Tj
0.000193787 Tc
7.3 TL
(Name)'
0.00037774 Tc
/R14 7.05 Tf
40.25 1.8 Td
(Ram Singh)Tj
0 Tc
2.5 -11.85 Td
(37)Tj
0.00137196 Tc
/R12 7.15 Tf
-5.25 13.35 Td
(:)Tj