2
1
mirror of https://github.com/qpdf/qpdf.git synced 2024-10-31 19:02:30 +00:00

Add CLI flags for image optimization

This commit is contained in:
Jay Berkenbilt 2019-01-04 18:06:56 -05:00
parent 6f3b76b6c1
commit ee2aad4381
19 changed files with 1061 additions and 0 deletions

View File

@ -1,5 +1,13 @@
2019-01-04 Jay Berkenbilt <ejb@ql.org>
* Add new option --optimize-images, which recompresses every image
using DCT (JPEG) compression as long as the image is not already
compressed with lossy compression and recompressing the image
reduces its size. The additional options --oi-min-width,
--oi-min-height, and --oi-min-area prevent recompression of images
whose width, height, or pixel area (width * height) are below a
specified threshold.
* Add new option --collate. When specified, the semantics of
--pages change from concatenation to collation. See the manual for
a more detailed discussion. Fixes #259.

View File

@ -860,6 +860,7 @@ class QPDFObjectHandle
// do nothing. Objects read normally from the file have
// descriptions. See comments on setObjectDescription for
// additional details.
QPDF_DLL
void warnIfPossible(std::string const& warning,
bool throw_if_no_description = false);

View File

@ -11,6 +11,8 @@
#include <qpdf/FileInputSource.hh>
#include <qpdf/Pl_StdioFile.hh>
#include <qpdf/Pl_Discard.hh>
#include <qpdf/Pl_DCT.hh>
#include <qpdf/Pl_Count.hh>
#include <qpdf/PointerHolder.hh>
#include <qpdf/QPDF.hh>
@ -127,6 +129,10 @@ struct Options
collate(false),
json(false),
check(false),
optimize_images(false),
oi_min_width(128),
oi_min_height(128),
oi_min_area(16384),
require_outfile(true),
infilename(0),
outfilename(0)
@ -208,6 +214,10 @@ struct Options
std::set<std::string> json_keys;
std::set<std::string> json_objects;
bool check;
bool optimize_images;
size_t oi_min_width;
size_t oi_min_height;
size_t oi_min_area;
std::vector<PageSpec> page_specs;
std::map<std::string, RotationSpec> rotations;
bool require_outfile;
@ -602,6 +612,10 @@ class ArgParser
void argJsonKey(char* parameter);
void argJsonObject(char* parameter);
void argCheck();
void argOptimizeImages();
void argOiMinWidth(char* paramter);
void argOiMinHeight(char* paramter);
void argOiMinArea(char* paramter);
void arg40Print(char* parameter);
void arg40Modify(char* parameter);
void arg40Extract(char* parameter);
@ -816,6 +830,13 @@ ArgParser::initOptionTable()
(*t)["json-object"] = oe_requiredParameter(
&ArgParser::argJsonObject, "trailer|obj[,gen]");
(*t)["check"] = oe_bare(&ArgParser::argCheck);
(*t)["optimize-images"] = oe_bare(&ArgParser::argOptimizeImages);
(*t)["oi-min-width"] = oe_requiredParameter(
&ArgParser::argOiMinWidth, "minimum-width");
(*t)["oi-min-height"] = oe_requiredParameter(
&ArgParser::argOiMinHeight, "minimum-height");
(*t)["oi-min-area"] = oe_requiredParameter(
&ArgParser::argOiMinArea, "minimum-area");
t = &this->encrypt40_option_table;
(*t)["--"] = oe_bare(&ArgParser::argEndEncrypt);
@ -1416,6 +1437,30 @@ ArgParser::argCheck()
o.require_outfile = false;
}
void
ArgParser::argOptimizeImages()
{
o.optimize_images = true;
}
void
ArgParser::argOiMinWidth(char* parameter)
{
o.oi_min_width = QUtil::string_to_int(parameter);
}
void
ArgParser::argOiMinHeight(char* parameter)
{
o.oi_min_height = QUtil::string_to_int(parameter);
}
void
ArgParser::argOiMinArea(char* parameter)
{
o.oi_min_area = QUtil::string_to_int(parameter);
}
void
ArgParser::arg40Print(char* parameter)
{
@ -1911,6 +1956,10 @@ familiar with the PDF file format or who are PDF developers.\n\
contents including those for interactive form\n\
fields; may also want --generate-appearances\n\
--generate-appearances generate appearance streams for form fields\n\
--optimize-images compress images with DCT (JPEG) when advantageous\n\
--oi-min-width=w do not optimize images whose width is below w\n\
--oi-min-height=h do not optimize images whose height is below h\n\
--oi-min-area=a do not optimize images whose pixel count is below a\n\
--qdf turns on \"QDF mode\" (below)\n\
--linearize-pass1=file write intermediate pass of linearized file\n\
for debugging\n\
@ -3403,9 +3452,202 @@ static void do_inspection(QPDF& pdf, Options& o)
}
}
class ImageOptimizer: public QPDFObjectHandle::StreamDataProvider
{
public:
ImageOptimizer(Options& o, QPDFObjectHandle& image);
virtual ~ImageOptimizer()
{
}
virtual void provideStreamData(int objid, int generation,
Pipeline* pipeline);
PointerHolder<Pipeline> makePipeline(
std::string const& description, Pipeline* next);
bool evaluate(std::string const& description);
private:
Options& o;
QPDFObjectHandle image;
};
ImageOptimizer::ImageOptimizer(Options& o, QPDFObjectHandle& image) :
o(o),
image(image)
{
}
PointerHolder<Pipeline>
ImageOptimizer::makePipeline(std::string const& description, Pipeline* next)
{
PointerHolder<Pipeline> result;
QPDFObjectHandle dict = image.getDict();
QPDFObjectHandle w_obj = dict.getKey("/Width");
QPDFObjectHandle h_obj = dict.getKey("/Height");
QPDFObjectHandle colorspace_obj = dict.getKey("/ColorSpace");
QPDFObjectHandle components_obj = dict.getKey("/BitsPerComponent");
if (! (w_obj.isInteger() &&
h_obj.isInteger() &&
colorspace_obj.isName() &&
components_obj.isInteger()))
{
if (o.verbose && (! description.empty()))
{
std::cout << whoami << ": " << description
<< ": not optimizing because image dictionary"
<< " is missing required keys" << std::endl;
}
return result;
}
JDIMENSION w = w_obj.getIntValue();
JDIMENSION h = h_obj.getIntValue();
std::string colorspace = colorspace_obj.getName();
int components = 0;
J_COLOR_SPACE cs = JCS_UNKNOWN;
if (colorspace == "/DeviceRGB")
{
components = 3;
cs = JCS_RGB;
}
else if (colorspace == "/DeviceGray")
{
components = 1;
cs = JCS_GRAYSCALE;
}
else if (colorspace == "/DeviceCMYK")
{
components = 4;
cs = JCS_CMYK;
}
else
{
if (o.verbose && (! description.empty()))
{
std::cout << whoami << ": " << description
<< ": not optimizing because of unsupported"
<< " image parameters" << std::endl;
}
return result;
}
if (((o.oi_min_width > 0) && (w <= o.oi_min_width)) ||
((o.oi_min_height > 0) && (h <= o.oi_min_height)) ||
((o.oi_min_area > 0) && ((w * h) <= o.oi_min_area)))
{
QTC::TC("qpdf", "qpdf image optimize too small");
if (o.verbose && (! description.empty()))
{
std::cout << whoami << ": " << description
<< ": not optimizing because of image"
<< " is smaller than requested minimum dimensions"
<< std::endl;
}
return result;
}
result = new Pl_DCT("jpg", next, w, h, components, cs);
return result;
}
bool
ImageOptimizer::evaluate(std::string const& description)
{
Pl_Discard d;
Pl_Count c("count", &d);
PointerHolder<Pipeline> p = makePipeline(description, &c);
if (p.getPointer() == 0)
{
// message issued by makePipeline
return false;
}
if (! image.pipeStreamData(p.getPointer(), 0, qpdf_dl_specialized,
true, false))
{
QTC::TC("qpdf", "qpdf image optimize no pipeline");
if (o.verbose)
{
std::cout << whoami << ": " << description
<< ": not optimizing because unable to decode data"
<< " or data already uses DCT"
<< std::endl;
}
return false;
}
long long orig_length = image.getDict().getKey("/Length").getIntValue();
if (c.getCount() >= orig_length)
{
QTC::TC("qpdf", "qpdf image optimize no shink");
if (o.verbose)
{
std::cout << whoami << ": " << description
<< ": not optimizing because DCT compression does not"
<< " reduce image size" << std::endl;
}
return false;
}
if (o.verbose)
{
std::cout << whoami << ": " << description
<< ": optimizing image reduces size from "
<< orig_length << " to " << c.getCount()
<< std::endl;
}
return true;
}
void
ImageOptimizer::provideStreamData(int, int, Pipeline* pipeline)
{
PointerHolder<Pipeline> p = makePipeline("", pipeline);
if (p.getPointer() == 0)
{
// Should not be possible
image.warnIfPossible("unable to create pipeline after previous"
" success; image data will be lost");
pipeline->finish();
return;
}
image.pipeStreamData(p.getPointer(), 0, qpdf_dl_specialized,
false, false);
}
static void handle_transformations(QPDF& pdf, Options& o)
{
QPDFPageDocumentHelper dh(pdf);
if (o.optimize_images)
{
int pageno = 0;
std::vector<QPDFPageObjectHelper> pages = dh.getAllPages();
for (std::vector<QPDFPageObjectHelper>::iterator iter = pages.begin();
iter != pages.end(); ++iter)
{
++pageno;
QPDFPageObjectHelper& ph(*iter);
QPDFObjectHandle page = ph.getObjectHandle();
std::map<std::string, QPDFObjectHandle> images =
ph.getPageImages();
for (std::map<std::string, QPDFObjectHandle>::iterator iter =
images.begin();
iter != images.end(); ++iter)
{
std::string name = (*iter).first;
QPDFObjectHandle& image = (*iter).second;
ImageOptimizer* io = new ImageOptimizer(o, image);
PointerHolder<QPDFObjectHandle::StreamDataProvider> sdp(io);
if (io->evaluate("image " + name + " on page " +
QUtil::int_to_string(pageno)))
{
QPDFObjectHandle new_image =
QPDFObjectHandle::newStream(&pdf);
new_image.replaceDict(image.getDict().shallowCopy());
new_image.replaceStreamData(
sdp,
QPDFObjectHandle::newName("/DCTDecode"),
QPDFObjectHandle::newNull());
page.getKey("/Resources").getKey("/XObject").replaceKey(
name, new_image);
}
}
}
}
if (o.generate_appearances)
{
QPDFAcroFormDocumentHelper afdh(pdf);

View File

@ -402,3 +402,6 @@ QPDFFormFieldObjectHelper list found 0
QPDFFormFieldObjectHelper list first too low 0
QPDFFormFieldObjectHelper list last too high 0
QPDF detected dangling ref 0
qpdf image optimize no pipeline 0
qpdf image optimize no shink 0
qpdf image optimize too small 0

View File

@ -1849,6 +1849,43 @@ $td->runtest("get data",
{$td->FILE => "bad-jpeg-show.out", $td->EXIT_STATUS => 3},
$td->NORMALIZE_NEWLINES);
show_ntests();
# ----------
$td->notify("--- Image Optimization ---");
my @image_opt = (
['image-streams', 'image-streams', ''],
['small-images', 'defaults', ''],
['small-images', 'min-width',
'--oi-min-width=150 --oi-min-height=0 --oi-min-area=0'],
['small-images', 'min-height',
'--oi-min-width=0 --oi-min-height=150 --oi-min-area=0'],
['small-images', 'min-area',
'--oi-min-width=0 --oi-min-height=0 --oi-min-area=30000'],
['small-images', 'min-area-all',
'--oi-min-width=0 --oi-min-height=0 --oi-min-area=30000'],
);
$n_tests += 2 * scalar(@image_opt);
foreach my $d (@image_opt)
{
my ($f, $description, $args) = @$d;
$td->runtest("optimize images: $description",
{$td->COMMAND =>
"qpdf --static-id --optimize-images --verbose" .
" $args $f.pdf a.pdf",
$td->FILTER => "perl filter-optimize-images.pl"},
{$td->FILE => "optimize-images-$description.out",
$td->EXIT_STATUS => 0},
$td->NORMALIZE_NEWLINES);
$td->runtest("check json: $description",
{$td->COMMAND => "qpdf --json --json-key=pages a.pdf"},
{$td->FILE => "optimize-images-$description-json.out",
$td->EXIT_STATUS => 0},
$td->NORMALIZE_NEWLINES);
}
show_ntests();
# ----------
$td->notify("--- Preserve unreferenced objects ---");

View File

@ -0,0 +1,8 @@
use strict;
use warnings;
while (<>)
{
s/(reduces size from \d+ to )\d+/$1.../;
print;
}

View File

@ -0,0 +1,64 @@
{
"pages": [
{
"contents": [
"4 0 R"
],
"images": [
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/FlateDecode"
],
"filterable": true,
"height": 200,
"name": "/Im1",
"object": "6 0 R",
"width": 100
},
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/FlateDecode"
],
"filterable": true,
"height": 100,
"name": "/Im2",
"object": "7 0 R",
"width": 200
},
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/FlateDecode"
],
"filterable": true,
"height": 200,
"name": "/Im3",
"object": "8 0 R",
"width": 200
}
],
"label": null,
"object": "3 0 R",
"outlines": [],
"pageposfrom1": 1
}
],
"parameters": {
"decodelevel": "generalized"
},
"version": 1
}

View File

@ -0,0 +1,4 @@
qpdf: image /Im1 on page 1: not optimizing because of image is smaller than requested minimum dimensions
qpdf: image /Im2 on page 1: not optimizing because of image is smaller than requested minimum dimensions
qpdf: image /Im3 on page 1: not optimizing because DCT compression does not reduce image size
qpdf: wrote file a.pdf

View File

@ -0,0 +1,242 @@
{
"pages": [
{
"contents": [
"12 0 R"
],
"images": [
{
"bitspercomponent": 8,
"colorspace": "/DeviceCMYK",
"decodeparms": [
null
],
"filter": [
"/DCTDecode"
],
"filterable": false,
"height": 480,
"name": "/Im1",
"object": "14 0 R",
"width": 400
}
],
"label": null,
"object": "3 0 R",
"outlines": [],
"pageposfrom1": 1
},
{
"contents": [
"15 0 R"
],
"images": [
{
"bitspercomponent": 8,
"colorspace": "/DeviceCMYK",
"decodeparms": [
null
],
"filter": [
"/DCTDecode"
],
"filterable": false,
"height": 480,
"name": "/Im1",
"object": "16 0 R",
"width": 400
}
],
"label": null,
"object": "4 0 R",
"outlines": [],
"pageposfrom1": 2
},
{
"contents": [
"17 0 R"
],
"images": [
{
"bitspercomponent": 8,
"colorspace": "/DeviceCMYK",
"decodeparms": [
null
],
"filter": [
"/DCTDecode"
],
"filterable": false,
"height": 480,
"name": "/Im1",
"object": "18 0 R",
"width": 400
}
],
"label": null,
"object": "5 0 R",
"outlines": [],
"pageposfrom1": 3
},
{
"contents": [
"19 0 R"
],
"images": [
{
"bitspercomponent": 8,
"colorspace": "/DeviceRGB",
"decodeparms": [
null
],
"filter": [
"/DCTDecode"
],
"filterable": false,
"height": 480,
"name": "/Im1",
"object": "20 0 R",
"width": 400
}
],
"label": null,
"object": "6 0 R",
"outlines": [],
"pageposfrom1": 4
},
{
"contents": [
"21 0 R"
],
"images": [
{
"bitspercomponent": 8,
"colorspace": "/DeviceRGB",
"decodeparms": [
null
],
"filter": [
"/DCTDecode"
],
"filterable": false,
"height": 480,
"name": "/Im1",
"object": "22 0 R",
"width": 400
}
],
"label": null,
"object": "7 0 R",
"outlines": [],
"pageposfrom1": 5
},
{
"contents": [
"23 0 R"
],
"images": [
{
"bitspercomponent": 8,
"colorspace": "/DeviceRGB",
"decodeparms": [
null
],
"filter": [
"/DCTDecode"
],
"filterable": false,
"height": 480,
"name": "/Im1",
"object": "24 0 R",
"width": 400
}
],
"label": null,
"object": "8 0 R",
"outlines": [],
"pageposfrom1": 6
},
{
"contents": [
"25 0 R"
],
"images": [
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/DCTDecode"
],
"filterable": false,
"height": 480,
"name": "/Im1",
"object": "26 0 R",
"width": 400
}
],
"label": null,
"object": "9 0 R",
"outlines": [],
"pageposfrom1": 7
},
{
"contents": [
"27 0 R"
],
"images": [
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/DCTDecode"
],
"filterable": false,
"height": 480,
"name": "/Im1",
"object": "28 0 R",
"width": 400
}
],
"label": null,
"object": "10 0 R",
"outlines": [],
"pageposfrom1": 8
},
{
"contents": [
"29 0 R"
],
"images": [
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/DCTDecode"
],
"filterable": false,
"height": 480,
"name": "/Im1",
"object": "30 0 R",
"width": 400
}
],
"label": null,
"object": "11 0 R",
"outlines": [],
"pageposfrom1": 9
}
],
"parameters": {
"decodelevel": "generalized"
},
"version": 1
}

View File

@ -0,0 +1,10 @@
qpdf: image /Im1 on page 1: optimizing image reduces size from 768000 to ...
qpdf: image /Im1 on page 2: not optimizing because unable to decode data or data already uses DCT
qpdf: image /Im1 on page 3: optimizing image reduces size from 768998 to ...
qpdf: image /Im1 on page 4: optimizing image reduces size from 576000 to ...
qpdf: image /Im1 on page 5: not optimizing because unable to decode data or data already uses DCT
qpdf: image /Im1 on page 6: optimizing image reduces size from 641497 to ...
qpdf: image /Im1 on page 7: optimizing image reduces size from 192000 to ...
qpdf: image /Im1 on page 8: not optimizing because unable to decode data or data already uses DCT
qpdf: image /Im1 on page 9: optimizing image reduces size from 3001 to ...
qpdf: wrote file a.pdf

View File

@ -0,0 +1,64 @@
{
"pages": [
{
"contents": [
"4 0 R"
],
"images": [
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/FlateDecode"
],
"filterable": true,
"height": 200,
"name": "/Im1",
"object": "6 0 R",
"width": 100
},
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/FlateDecode"
],
"filterable": true,
"height": 100,
"name": "/Im2",
"object": "7 0 R",
"width": 200
},
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/FlateDecode"
],
"filterable": true,
"height": 200,
"name": "/Im3",
"object": "8 0 R",
"width": 200
}
],
"label": null,
"object": "3 0 R",
"outlines": [],
"pageposfrom1": 1
}
],
"parameters": {
"decodelevel": "generalized"
},
"version": 1
}

View File

@ -0,0 +1,4 @@
qpdf: image /Im1 on page 1: not optimizing because of image is smaller than requested minimum dimensions
qpdf: image /Im2 on page 1: not optimizing because of image is smaller than requested minimum dimensions
qpdf: image /Im3 on page 1: not optimizing because DCT compression does not reduce image size
qpdf: wrote file a.pdf

View File

@ -0,0 +1,64 @@
{
"pages": [
{
"contents": [
"4 0 R"
],
"images": [
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/FlateDecode"
],
"filterable": true,
"height": 200,
"name": "/Im1",
"object": "6 0 R",
"width": 100
},
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/FlateDecode"
],
"filterable": true,
"height": 100,
"name": "/Im2",
"object": "7 0 R",
"width": 200
},
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/FlateDecode"
],
"filterable": true,
"height": 200,
"name": "/Im3",
"object": "8 0 R",
"width": 200
}
],
"label": null,
"object": "3 0 R",
"outlines": [],
"pageposfrom1": 1
}
],
"parameters": {
"decodelevel": "generalized"
},
"version": 1
}

View File

@ -0,0 +1,4 @@
qpdf: image /Im1 on page 1: not optimizing because of image is smaller than requested minimum dimensions
qpdf: image /Im2 on page 1: not optimizing because of image is smaller than requested minimum dimensions
qpdf: image /Im3 on page 1: not optimizing because DCT compression does not reduce image size
qpdf: wrote file a.pdf

View File

@ -0,0 +1,64 @@
{
"pages": [
{
"contents": [
"4 0 R"
],
"images": [
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/DCTDecode"
],
"filterable": false,
"height": 200,
"name": "/Im1",
"object": "6 0 R",
"width": 100
},
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/FlateDecode"
],
"filterable": true,
"height": 100,
"name": "/Im2",
"object": "7 0 R",
"width": 200
},
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/FlateDecode"
],
"filterable": true,
"height": 200,
"name": "/Im3",
"object": "8 0 R",
"width": 200
}
],
"label": null,
"object": "3 0 R",
"outlines": [],
"pageposfrom1": 1
}
],
"parameters": {
"decodelevel": "generalized"
},
"version": 1
}

View File

@ -0,0 +1,4 @@
qpdf: image /Im1 on page 1: optimizing image reduces size from 20000 to ...
qpdf: image /Im2 on page 1: not optimizing because of image is smaller than requested minimum dimensions
qpdf: image /Im3 on page 1: not optimizing because DCT compression does not reduce image size
qpdf: wrote file a.pdf

View File

@ -0,0 +1,64 @@
{
"pages": [
{
"contents": [
"4 0 R"
],
"images": [
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/FlateDecode"
],
"filterable": true,
"height": 200,
"name": "/Im1",
"object": "6 0 R",
"width": 100
},
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/DCTDecode"
],
"filterable": false,
"height": 100,
"name": "/Im2",
"object": "7 0 R",
"width": 200
},
{
"bitspercomponent": 8,
"colorspace": "/DeviceGray",
"decodeparms": [
null
],
"filter": [
"/FlateDecode"
],
"filterable": true,
"height": 200,
"name": "/Im3",
"object": "8 0 R",
"width": 200
}
],
"label": null,
"object": "3 0 R",
"outlines": [],
"pageposfrom1": 1
}
],
"parameters": {
"decodelevel": "generalized"
},
"version": 1
}

View File

@ -0,0 +1,4 @@
qpdf: image /Im1 on page 1: not optimizing because of image is smaller than requested minimum dimensions
qpdf: image /Im2 on page 1: optimizing image reduces size from 20000 to ...
qpdf: image /Im3 on page 1: not optimizing because DCT compression does not reduce image size
qpdf: wrote file a.pdf

File diff suppressed because one or more lines are too long