TCPDF + FPDI : 确保所有字体完全嵌入
TCPDF + FPDI : Ensure that all fonts are fully embedded
我正在使用 TCPDF + FPDI。我想完全嵌入所有字体。使用 addTTFfont() 和 SetFont() 方法时，字体嵌入效果很好。但是，当我尝试使用 FPDI 打开现有的 pdf 文件时，我无法获得已用字体的列表，也无法确定其中哪些已被嵌入。这是我正在尝试做的事情：
- 获取源pdf文件的字体信息（名称、类型等）；
- 确定哪些字体没有完全嵌入；
- 嵌入这些字体。
到目前为止，我尝试让 TCPDF class protected 成为成员 fonts public。所以，我更改了 tcpdf.php 文件：
/**
* Array of used fonts.
* @protected
*/
public $fonts = array();
在main.php
$pdf = new FPDI();
$pdf->setSourceFile('Simple.pdf');
$pdf->AddPage();
$tplIdx = $pdf->importPage(1);
$pdf->useTemplate($tplIdx, 10, 10, 200);
$pdf->setFontSubsetting(false);
var_dump($pdf->fonts);
输出：
array (size=2)
'helvetica' =>
array (size=17)
'fontkey' => string 'helvetica' (length=9)
'i' => int 1
'type' => string 'core' (length=4)
'name' => string 'Helvetica' (length=9)
'desc' =>
array (size=13)
'Flags' => int 32
'FontBBox' => string '[-166 -225 1000 931]' (length=20)
'ItalicAngle' => int 0
'Ascent' => int 931
'Descent' => int -225
'Leading' => int 0
'CapHeight' => int 718
'XHeight' => int 523
'StemV' => int 88
'StemH' => int 76
'AvgWidth' => int 513
'MaxWidth' => int 1015
'MissingWidth' => int 513
'up' => int -100
'ut' => int 50
'cw' =>
array (size=256)
0 => int 500
1 => int 500
2 => int 500
...
...
'cbbox' =>
array (size=0)
empty
'dw' => int 513
'enc' => string '' (length=0)
'cidinfo' =>
array (size=4)
'Registry' => string 'Adobe' (length=5)
'Ordering' => string 'Identity' (length=8)
'Supplement' => int 0
'uni2cid' =>
array (size=0)
empty
'file' => string '' (length=0)
'ctg' => string '' (length=0)
'subset' => boolean false
'subsetchars' =>
array (size=255)
0 => boolean true
1 => boolean true
2 => boolean true
3 => boolean true
...
...
'n' => int 3
'helveticaB' =>
array (size=17)
'fontkey' => string 'helveticaB' (length=10)
'i' => int 2
'type' => string 'core' (length=4)
'name' => string 'Helvetica-Bold' (length=14)
'desc' =>
array (size=13)
'Flags' => int 32
'FontBBox' => string '[-170 -228 1003 962]' (length=20)
'ItalicAngle' => int 0
'Ascent' => int 962
'Descent' => int -228
'Leading' => int 0
'CapHeight' => int 718
'XHeight' => int 532
'StemV' => int 140
'StemH' => int 118
'AvgWidth' => int 535
'MaxWidth' => int 1000
'MissingWidth' => int 535
'up' => int -100
'ut' => int 50
'cw' =>
array (size=256)
0 => int 278
1 => int 278
2 => int 278
3 => int 278
...
...
'cbbox' =>
array (size=0)
empty
'dw' => int 535
'enc' => string '' (length=0)
'cidinfo' =>
array (size=4)
'Registry' => string 'Adobe' (length=5)
'Ordering' => string 'Identity' (length=8)
'Supplement' => int 0
'uni2cid' =>
array (size=0)
empty
'file' => string '' (length=0)
'ctg' => string '' (length=0)
'subset' => boolean false
'subsetchars' =>
array (size=255)
0 => boolean true
1 => boolean true
2 => boolean true
3 => boolean true
...
...
'n' => int 5
有关于 2 种字体的信息：“helvetica”（根据 TCPDF 文档，这是默认字体）和“helveticaB”。但是源文件也包含其他字体。它们没有列出。如何列出它们？
提前致谢...
您需要确保导入的 document/pages 已预先嵌入所有字体。无法保证导入页面中的字体是通过FPDI嵌入的。
您也不能 access/list 带有 FPDI 的字体。
在深入研究 .pdf 文件并阅读 pdf 文档的一些主题后，我设法以某种方式解决了问题。以下代码只是草稿，需要大量改进。我已经测试了几个 .pdf 文件并且它有效，但是有很多东西需要更好的方法。
在尝试解决问题时，我决定保留 TCPDF 和 FPDI 库不变。我没有使用过这些库的任何方法，因为我不知道这些库的内部功能。
P.S：不幸的是，我没有太多时间而且我的代码很长（而且不稳定）：
<?php
require_once('tcpdf.php');
require_once('fpdi.php');
class PDF extends FPDI
{
var $_tplIdx;
private $objects;
private $font_list;
private $raw_data;
private $raw_data_len;
private $font_read_offset;
private $embedded_font_addr;
private $is_descendant;
function __construct($filename)
{
$this->objects = array();
$this->font_list = array();
$this->raw_data = file_get_contents($filename);
$this->raw_data_len = strlen($this->raw_data);
$this->font_read_offset = 0;
$this->is_descendant = false;
parent::__construct();
}
private function EmbeddedFontRefs()
{
do
{
$f = $this->GetXObj('Subtype /CIDFontType2');
if($f == NULL)
return;
//var_dump($f);
$keys = array_keys($f);
$f = $f[$keys[0]];
$name = '';
$ref = '';
if(preg_match('/Subtype \/CIDFontType2/', $f))
{
$matches = array();
if(preg_match('/BaseFont \/\S+/', $f, $matches))
$name = substr($matches[0], 10);
if(preg_match('/FontDescriptor \d+ \d+ R+/', $f, $matches))
{
$ref = substr($matches[0], 15);
$ref = $this->GetObjFromRef($ref);
}
$this->embedded_font_addr[$name] = $ref;
}
}while($f !== NULL);
}
private function ParseObj($obj_search_string, $referrer_object_addr=NULL, $referrer_object_name=NULL)
{
$object_data = $this->GetObj($obj_search_string);
if($object_data == NULL)
return false;
$object_addr = array_keys($object_data);
$object_addr = $object_addr[0];
$object_raw_data = $object_data[$object_addr];
$embedded = false;
$subset = false;
if(!array_key_exists($object_addr, $this->objects))
{
$this->objects[$object_addr] = array();
$obj_content_items = preg_split('/\n\r?/', $object_raw_data);
for($i = 0; $i < count($obj_content_items); $i++)
{
$item = explode(' ', $obj_content_items[$i], 2);
if($item[0][0] == '/')
$item[0] = substr($item[0], 1);
if($item[1][0] == '/')
$item[1] = substr($item[1], 1);
if(($i + 1) < count($obj_content_items))
$item[1] = substr($item[1], 0, -1);
if(($item[0] == 'FontFile') || ($item[0] == 'FontFile2') || ($item[0] == 'FontFile3'))
$embedded = true;
else if(($item[0] == 'BaseFont') || ($item[0] == 'FontName'))
{
$item[1] = str_replace('#20', ' ', $item[1]);
if(preg_match('/[A-Z]{5}\+/', $item[1]))
$subset = true;
}
$this->objects[$object_addr][$item[0]] = $item[1]; // item_name => item_value
}
}
if($referrer_object_addr == NULL)
{
if(array_key_exists('Subtype', $this->objects[$object_addr]))
{
if($this->objects[$object_addr]['Subtype'] == 'Type0') //Composite Fonts
{
if(array_key_exists('DescendantFonts', $this->objects[$object_addr]))
{
$descendant_fonts_obj_addr = $this->GetObjFromRef($this->objects[$object_addr]['DescendantFonts']);
array_push($this->font_list, '');
$this->font_list[count($this->font_list)-1] = $this->objects[$object_addr];
$this->ParseObj($descendant_fonts_obj_addr, $object_addr, 'DescendantFonts');
}
}
}
if(array_key_exists('FontDescriptor', $this->objects[$object_addr]))
{
$font_desc_obj_addr = $this->GetObjFromRef($this->objects[$object_addr]['FontDescriptor']);
array_push($this->font_list, '');
$this->font_list[count($this->font_list)-1] = $this->objects[$object_addr];
$this->ParseObj($font_desc_obj_addr, $object_addr, 'FontDescriptor');
}
}
else
{
if(array_key_exists($referrer_object_addr, $this->objects))
{
if($referrer_object_name == 'DescendantFonts')
{
$this->font_list[count($this->font_list)-1]['DescendantFonts'] = array();
array_push($this->font_list[count($this->font_list)-1]['DescendantFonts'], $this->objects[$object_addr]);
$this->is_descendant = true;
if(array_key_exists('FontDescriptor', $this->objects[$object_addr]))
{
$font_desc_obj_addr = $this->GetObjFromRef($this->objects[$object_addr]['FontDescriptor']);
$this->ParseObj($font_desc_obj_addr, $object_addr, 'FontDescriptor');
}
}
else if($referrer_object_name == 'FontDescriptor')
{
if(array_key_exists('Type', $this->objects[$object_addr]))
unset($this->objects[$object_addr]['Type']);
//Embedded or not?
$keys = array_keys($this->objects[$object_addr]);
foreach($keys as $key)
{
if(($key == 'FontFile') || ($key == 'FontFile2') || ($key == 'FontFile3'))
{
$embedded = true;
break;
}
}
if(!$this->is_descendant)
{
$this->font_list[count($this->font_list)-1]['Embedded'] = $embedded;
$this->font_list[count($this->font_list)-1]['Subset'] = $subset;
$this->font_list[count($this->font_list)-1]['FontDescriptor'] = $this->objects[$object_addr];
}
else
{
$this->font_list[count($this->font_list)-1]['DescendantFonts'][0]['Embedded'] = $embedded;
$this->font_list[count($this->font_list)-1]['DescendantFonts'][0]['Subset'] = $subset;
$this->font_list[count($this->font_list)-1]['DescendantFonts'][0]['FontDescriptor'] = $this->objects[$object_addr];
$this->is_descendant = false;
}
}
}
}
return true;
}
private function GetObj($obj_search_string)
{
$obj_offset = strpos($this->raw_data, $obj_search_string, $this->font_read_offset);
if($obj_offset == false)
{
return NULL;
}
$obj_start = 0;
$obj_end = 0;
$object_content_start = 0;
$object_content_end = 0;
$obj_start = strrpos($this->raw_data, 'endobj', $obj_offset - $this->raw_data_len) + 8; // for 'endobj\n\r';
$obj_end = strpos($this->raw_data, 'endobj', $obj_start);
$this->font_read_offset = $obj_end;
$object_content_start = strpos($this->raw_data, '<<', $obj_start) + 2; //for '<<'
$object_content_end = strpos($this->raw_data, '>>', $object_content_start) - 2; //for /n/r;
$object_addr = substr($this->raw_data, $obj_start, $object_content_start - $obj_start - 4); // -2 for /n/r;
$object_raw_data = substr($this->raw_data, $object_content_start, $object_content_end - $object_content_start);
return array($object_addr => $object_raw_data);
}
private function GetXObj($obj_search_string)
{
$obj_offset = strpos($this->raw_data, $obj_search_string, $this->font_read_offset);
if($obj_offset == false)
return NULL;
$obj_start = 0;
$obj_end = 0;
$object_content_start = 0;
$object_content_end = 0;
$obj_start = strrpos($this->raw_data, 'endobj', $obj_offset - $this->raw_data_len) + 8; // for 'endobj\n\r';
$obj_end = strpos($this->raw_data, 'endobj', $obj_start);
$this->font_read_offset = $obj_end;
$object_content_start = strpos($this->raw_data, '<<', $obj_start) + 2; //for '<<'
$object_content_end = $obj_end - 1;
$object_addr = substr($this->raw_data, $obj_start, $object_content_start - $obj_start - 3); // -2 for /n/r;
$object_raw_data = substr($this->raw_data, $object_content_start, $object_content_end - $object_content_start);
return array($object_addr => $object_raw_data);
}
private function GetFontDescriptor($font_object_addr)
{
if(array_key_exists('FontDescriptor', $this->objects[$font_object_addr]))
$font_desc_ref = $this->objects[$font_object_addr]['FontDescriptor'];
$font_desc_obj_addr = $this->GetObjFromRef($font_desc_ref);
$this->ParseObj($font_desc_obj_addr, $font_object_addr);
}
}
private function GetObjFromRef($obj_ref)
{
$obj_ref_num = explode(' ', $obj_ref);
$obj = '';
if(end($obj_ref_num) == 'R')
{
foreach($obj_ref_num as $num)
{
if($num != 'R')
$obj .= (string)$num. ' ';
else
$obj .= 'obj';
}
}
else if(end($obj_ref_num) == 'obj')
$obj = $obj_ref;
return $obj;
}
public function EnsureFonts($filename)
{
$this->raw_data = file_get_contents($filename);
$this->raw_data_len = strlen($this->raw_data);
$this->font_read_offset = 0;
$this->EmbeddedFontRefs();
$this->font_read_offset = 0;
$font_objs = array();
do
{
$xobj = $this->GetXObj('/Type /XObject');
if($xobj !== NULL)
{
$xobj_addr = array_keys($xobj);
$xobj_addr = $xobj_addr[0];
$matches = array();
if(preg_match('/Resources \d+ \d+ R/', $xobj[$xobj_addr], $matches))
{
array_push($font_objs, $this->GetObjFromRef(substr($matches[0], 10))); //Resources+' '
}
}
}while($xobj !== NULL);
$fxobjs = array();
foreach($font_objs as $font_obj)
{
$fobj = $this->GetXObj($font_obj);
if($fobj !== NULL)
{
$xobj_addr = array_keys($fobj);
$xobj_addr = $xobj_addr[0];
$matches = array();
if(preg_match('/Font \d+ \d+ R/', $fobj[$xobj_addr], $matches))
{
array_push($fxobjs, $this->GetObjFromRef(substr($matches[0], 10))); //Resources+' '
}
}
}
$actual_font_objs = array();
foreach($fxobjs as $fxobj)
{
$f_obj = $this->GetXObj($fxobj);
if($f_obj !== NULL)
{
$xobj_addr = array_keys($f_obj);
$xobj_addr = $xobj_addr[0];
$matches = array();
if(preg_match('/F\d+/', $f_obj[$xobj_addr], $matches))
{
array_push($actual_font_objs, $this->GetObjFromRef(substr($matches[0], 10))); //Resources+' '
}
}
}
foreach($actual_font_objs as $fo)
{
$f = $this->GetXObj($fxobj);
if($f !== NULL)
{
$xobj_addr = array_keys($f);
$xobj_addr = $xobj_addr[0];
$matches = array();
$f_name = '';
if(preg_match('/BaseFont \/\S+/', $f[$xobj_addr], $matches))
{
$f_name = substr($matches[0], 10); // BaseFont+' '+'/'
}
if(preg_match('/FontDescriptor \d+ \d+ R/', $f[$xobj_addr], $matches))
{
$fd_replace_start_pos = strpos($this->raw_data, $matches[0]) + 15; //FontDescriptor+' ';
$fd_replace_end_pos = strpos($this->raw_data, 'R', $fd_replace_start_pos) + 1; //For 'R';
$s = substr($this->raw_data, 0, $fd_replace_start_pos);
$r = substr($this->embedded_font_addr[$f_name], 0, -3).'R'; //-4 for obj.
$e = substr($this->raw_data, $fd_replace_end_pos);
$this->raw_data = $s . $r . $e;
}
}
}
file_put_contents($filename, $this->raw_data);
}
public function GetFontInfo()
{
do
{
$this->ParseObj('/Type /Font');
}while($this->ParseObj('/Type /Font') == true);
return $this->GetNotEmbeddedFonts();
}
private function GetNotEmbeddedFonts()
{
$result = array();
foreach($this->font_list as $font)
{
if(array_key_exists('DescendantFonts', $font))
{
foreach($font['DescendantFonts'] as $desc_font)
{
if((array_key_exists('Embedded', $desc_font)) && (array_key_exists('BaseFont', $desc_font)))
if(!$desc_font['Embedded'])
array_push($result, $desc_font['BaseFont']);
}
}
else
{
if((array_key_exists('Embedded', $font)) && (array_key_exists('BaseFont', $font)))
if(!$font['Embedded'])
array_push($result, $font['BaseFont']);
}
}
for($i= 0; $i<count($result); $i++)
{
if(strpos($result[$i], ' ') || strpos($result[$i], ',')) //Adobe Acrobat system font name list. Needs improvment..
{
switch($r)
{
case 'Times New Roman':
$result[$i] = 'TimesNewRomanPSMT';
break;
case 'Times New Roman,Bold':
$result[$i] = 'TimesNewRomanPS-BoldMT';
break;
case 'Times New Roman,Italic':
$result[$i] = 'TimesNewRomanPS-ItalicMT';
break;
case 'Times New Roman,BoldItalic':
$result[$i] = 'TimesNewRomanPS-BoldItalicMT';
break;
case 'Aial':
$result[$i] = 'ArialMT';
break;
}
}
}
return $result;
}
}
$pdf = new PDF($filename);
$pdf->setSourceFile($filename);
$pdf->AddPage();
$tplIdx = $pdf->importPage(1);
$pdf->useTemplate($tplIdx);
$pdf->SetMargins(PDF_MARGIN_LEFT, 40, PDF_MARGIN_RIGHT);
$pdf->SetAutoPageBreak(true, 40);
$pdf->setFontSubsetting(false);
$not_embedded_fonts = $pdf->GetFontInfo();
foreach($not_embedded_fonts as $f)
{
$font_name = strtolower($f);
$font_name = preg_replace('/[^a-z0-9_]/', '', $font_name);
$search = array('bold', 'oblique', 'italic', 'regular');
$replace = array('b', 'i', 'i', '');
$font_name = str_replace($search, $replace, $font_name);
if($pdf->AddFont($font_name) !== false)
$pdf->SetFont($font_name);
else
print('"' .$font_name. '" font not found!');
}
$pdf->Output();
?>
我正在使用 TCPDF + FPDI。我想完全嵌入所有字体。使用 addTTFfont() 和 SetFont() 方法时，字体嵌入效果很好。但是，当我尝试使用 FPDI 打开现有的 pdf 文件时，我无法获得已用字体的列表，也无法确定其中哪些已被嵌入。这是我正在尝试做的事情：
- 获取源pdf文件的字体信息（名称、类型等）；
- 确定哪些字体没有完全嵌入；
- 嵌入这些字体。
到目前为止，我尝试让 TCPDF class protected 成为成员 fonts public。所以，我更改了 tcpdf.php 文件：
/**
* Array of used fonts.
* @protected
*/
public $fonts = array();
在main.php
$pdf = new FPDI();
$pdf->setSourceFile('Simple.pdf');
$pdf->AddPage();
$tplIdx = $pdf->importPage(1);
$pdf->useTemplate($tplIdx, 10, 10, 200);
$pdf->setFontSubsetting(false);
var_dump($pdf->fonts);
输出：
array (size=2)
'helvetica' =>
array (size=17)
'fontkey' => string 'helvetica' (length=9)
'i' => int 1
'type' => string 'core' (length=4)
'name' => string 'Helvetica' (length=9)
'desc' =>
array (size=13)
'Flags' => int 32
'FontBBox' => string '[-166 -225 1000 931]' (length=20)
'ItalicAngle' => int 0
'Ascent' => int 931
'Descent' => int -225
'Leading' => int 0
'CapHeight' => int 718
'XHeight' => int 523
'StemV' => int 88
'StemH' => int 76
'AvgWidth' => int 513
'MaxWidth' => int 1015
'MissingWidth' => int 513
'up' => int -100
'ut' => int 50
'cw' =>
array (size=256)
0 => int 500
1 => int 500
2 => int 500
...
...
'cbbox' =>
array (size=0)
empty
'dw' => int 513
'enc' => string '' (length=0)
'cidinfo' =>
array (size=4)
'Registry' => string 'Adobe' (length=5)
'Ordering' => string 'Identity' (length=8)
'Supplement' => int 0
'uni2cid' =>
array (size=0)
empty
'file' => string '' (length=0)
'ctg' => string '' (length=0)
'subset' => boolean false
'subsetchars' =>
array (size=255)
0 => boolean true
1 => boolean true
2 => boolean true
3 => boolean true
...
...
'n' => int 3
'helveticaB' =>
array (size=17)
'fontkey' => string 'helveticaB' (length=10)
'i' => int 2
'type' => string 'core' (length=4)
'name' => string 'Helvetica-Bold' (length=14)
'desc' =>
array (size=13)
'Flags' => int 32
'FontBBox' => string '[-170 -228 1003 962]' (length=20)
'ItalicAngle' => int 0
'Ascent' => int 962
'Descent' => int -228
'Leading' => int 0
'CapHeight' => int 718
'XHeight' => int 532
'StemV' => int 140
'StemH' => int 118
'AvgWidth' => int 535
'MaxWidth' => int 1000
'MissingWidth' => int 535
'up' => int -100
'ut' => int 50
'cw' =>
array (size=256)
0 => int 278
1 => int 278
2 => int 278
3 => int 278
...
...
'cbbox' =>
array (size=0)
empty
'dw' => int 535
'enc' => string '' (length=0)
'cidinfo' =>
array (size=4)
'Registry' => string 'Adobe' (length=5)
'Ordering' => string 'Identity' (length=8)
'Supplement' => int 0
'uni2cid' =>
array (size=0)
empty
'file' => string '' (length=0)
'ctg' => string '' (length=0)
'subset' => boolean false
'subsetchars' =>
array (size=255)
0 => boolean true
1 => boolean true
2 => boolean true
3 => boolean true
...
...
'n' => int 5
有关于 2 种字体的信息：“helvetica”（根据 TCPDF 文档，这是默认字体）和“helveticaB”。但是源文件也包含其他字体。它们没有列出。如何列出它们？
提前致谢...
您需要确保导入的 document/pages 已预先嵌入所有字体。无法保证导入页面中的字体是通过FPDI嵌入的。
您也不能 access/list 带有 FPDI 的字体。
在深入研究 .pdf 文件并阅读 pdf 文档的一些主题后，我设法以某种方式解决了问题。以下代码只是草稿，需要大量改进。我已经测试了几个 .pdf 文件并且它有效，但是有很多东西需要更好的方法。
在尝试解决问题时，我决定保留 TCPDF 和 FPDI 库不变。我没有使用过这些库的任何方法，因为我不知道这些库的内部功能。
P.S：不幸的是，我没有太多时间而且我的代码很长（而且不稳定）：
<?php
require_once('tcpdf.php');
require_once('fpdi.php');
class PDF extends FPDI
{
var $_tplIdx;
private $objects;
private $font_list;
private $raw_data;
private $raw_data_len;
private $font_read_offset;
private $embedded_font_addr;
private $is_descendant;
function __construct($filename)
{
$this->objects = array();
$this->font_list = array();
$this->raw_data = file_get_contents($filename);
$this->raw_data_len = strlen($this->raw_data);
$this->font_read_offset = 0;
$this->is_descendant = false;
parent::__construct();
}
private function EmbeddedFontRefs()
{
do
{
$f = $this->GetXObj('Subtype /CIDFontType2');
if($f == NULL)
return;
//var_dump($f);
$keys = array_keys($f);
$f = $f[$keys[0]];
$name = '';
$ref = '';
if(preg_match('/Subtype \/CIDFontType2/', $f))
{
$matches = array();
if(preg_match('/BaseFont \/\S+/', $f, $matches))
$name = substr($matches[0], 10);
if(preg_match('/FontDescriptor \d+ \d+ R+/', $f, $matches))
{
$ref = substr($matches[0], 15);
$ref = $this->GetObjFromRef($ref);
}
$this->embedded_font_addr[$name] = $ref;
}
}while($f !== NULL);
}
private function ParseObj($obj_search_string, $referrer_object_addr=NULL, $referrer_object_name=NULL)
{
$object_data = $this->GetObj($obj_search_string);
if($object_data == NULL)
return false;
$object_addr = array_keys($object_data);
$object_addr = $object_addr[0];
$object_raw_data = $object_data[$object_addr];
$embedded = false;
$subset = false;
if(!array_key_exists($object_addr, $this->objects))
{
$this->objects[$object_addr] = array();
$obj_content_items = preg_split('/\n\r?/', $object_raw_data);
for($i = 0; $i < count($obj_content_items); $i++)
{
$item = explode(' ', $obj_content_items[$i], 2);
if($item[0][0] == '/')
$item[0] = substr($item[0], 1);
if($item[1][0] == '/')
$item[1] = substr($item[1], 1);
if(($i + 1) < count($obj_content_items))
$item[1] = substr($item[1], 0, -1);
if(($item[0] == 'FontFile') || ($item[0] == 'FontFile2') || ($item[0] == 'FontFile3'))
$embedded = true;
else if(($item[0] == 'BaseFont') || ($item[0] == 'FontName'))
{
$item[1] = str_replace('#20', ' ', $item[1]);
if(preg_match('/[A-Z]{5}\+/', $item[1]))
$subset = true;
}
$this->objects[$object_addr][$item[0]] = $item[1]; // item_name => item_value
}
}
if($referrer_object_addr == NULL)
{
if(array_key_exists('Subtype', $this->objects[$object_addr]))
{
if($this->objects[$object_addr]['Subtype'] == 'Type0') //Composite Fonts
{
if(array_key_exists('DescendantFonts', $this->objects[$object_addr]))
{
$descendant_fonts_obj_addr = $this->GetObjFromRef($this->objects[$object_addr]['DescendantFonts']);
array_push($this->font_list, '');
$this->font_list[count($this->font_list)-1] = $this->objects[$object_addr];
$this->ParseObj($descendant_fonts_obj_addr, $object_addr, 'DescendantFonts');
}
}
}
if(array_key_exists('FontDescriptor', $this->objects[$object_addr]))
{
$font_desc_obj_addr = $this->GetObjFromRef($this->objects[$object_addr]['FontDescriptor']);
array_push($this->font_list, '');
$this->font_list[count($this->font_list)-1] = $this->objects[$object_addr];
$this->ParseObj($font_desc_obj_addr, $object_addr, 'FontDescriptor');
}
}
else
{
if(array_key_exists($referrer_object_addr, $this->objects))
{
if($referrer_object_name == 'DescendantFonts')
{
$this->font_list[count($this->font_list)-1]['DescendantFonts'] = array();
array_push($this->font_list[count($this->font_list)-1]['DescendantFonts'], $this->objects[$object_addr]);
$this->is_descendant = true;
if(array_key_exists('FontDescriptor', $this->objects[$object_addr]))
{
$font_desc_obj_addr = $this->GetObjFromRef($this->objects[$object_addr]['FontDescriptor']);
$this->ParseObj($font_desc_obj_addr, $object_addr, 'FontDescriptor');
}
}
else if($referrer_object_name == 'FontDescriptor')
{
if(array_key_exists('Type', $this->objects[$object_addr]))
unset($this->objects[$object_addr]['Type']);
//Embedded or not?
$keys = array_keys($this->objects[$object_addr]);
foreach($keys as $key)
{
if(($key == 'FontFile') || ($key == 'FontFile2') || ($key == 'FontFile3'))
{
$embedded = true;
break;
}
}
if(!$this->is_descendant)
{
$this->font_list[count($this->font_list)-1]['Embedded'] = $embedded;
$this->font_list[count($this->font_list)-1]['Subset'] = $subset;
$this->font_list[count($this->font_list)-1]['FontDescriptor'] = $this->objects[$object_addr];
}
else
{
$this->font_list[count($this->font_list)-1]['DescendantFonts'][0]['Embedded'] = $embedded;
$this->font_list[count($this->font_list)-1]['DescendantFonts'][0]['Subset'] = $subset;
$this->font_list[count($this->font_list)-1]['DescendantFonts'][0]['FontDescriptor'] = $this->objects[$object_addr];
$this->is_descendant = false;
}
}
}
}
return true;
}
private function GetObj($obj_search_string)
{
$obj_offset = strpos($this->raw_data, $obj_search_string, $this->font_read_offset);
if($obj_offset == false)
{
return NULL;
}
$obj_start = 0;
$obj_end = 0;
$object_content_start = 0;
$object_content_end = 0;
$obj_start = strrpos($this->raw_data, 'endobj', $obj_offset - $this->raw_data_len) + 8; // for 'endobj\n\r';
$obj_end = strpos($this->raw_data, 'endobj', $obj_start);
$this->font_read_offset = $obj_end;
$object_content_start = strpos($this->raw_data, '<<', $obj_start) + 2; //for '<<'
$object_content_end = strpos($this->raw_data, '>>', $object_content_start) - 2; //for /n/r;
$object_addr = substr($this->raw_data, $obj_start, $object_content_start - $obj_start - 4); // -2 for /n/r;
$object_raw_data = substr($this->raw_data, $object_content_start, $object_content_end - $object_content_start);
return array($object_addr => $object_raw_data);
}
private function GetXObj($obj_search_string)
{
$obj_offset = strpos($this->raw_data, $obj_search_string, $this->font_read_offset);
if($obj_offset == false)
return NULL;
$obj_start = 0;
$obj_end = 0;
$object_content_start = 0;
$object_content_end = 0;
$obj_start = strrpos($this->raw_data, 'endobj', $obj_offset - $this->raw_data_len) + 8; // for 'endobj\n\r';
$obj_end = strpos($this->raw_data, 'endobj', $obj_start);
$this->font_read_offset = $obj_end;
$object_content_start = strpos($this->raw_data, '<<', $obj_start) + 2; //for '<<'
$object_content_end = $obj_end - 1;
$object_addr = substr($this->raw_data, $obj_start, $object_content_start - $obj_start - 3); // -2 for /n/r;
$object_raw_data = substr($this->raw_data, $object_content_start, $object_content_end - $object_content_start);
return array($object_addr => $object_raw_data);
}
private function GetFontDescriptor($font_object_addr)
{
if(array_key_exists('FontDescriptor', $this->objects[$font_object_addr]))
$font_desc_ref = $this->objects[$font_object_addr]['FontDescriptor'];
$font_desc_obj_addr = $this->GetObjFromRef($font_desc_ref);
$this->ParseObj($font_desc_obj_addr, $font_object_addr);
}
}
private function GetObjFromRef($obj_ref)
{
$obj_ref_num = explode(' ', $obj_ref);
$obj = '';
if(end($obj_ref_num) == 'R')
{
foreach($obj_ref_num as $num)
{
if($num != 'R')
$obj .= (string)$num. ' ';
else
$obj .= 'obj';
}
}
else if(end($obj_ref_num) == 'obj')
$obj = $obj_ref;
return $obj;
}
public function EnsureFonts($filename)
{
$this->raw_data = file_get_contents($filename);
$this->raw_data_len = strlen($this->raw_data);
$this->font_read_offset = 0;
$this->EmbeddedFontRefs();
$this->font_read_offset = 0;
$font_objs = array();
do
{
$xobj = $this->GetXObj('/Type /XObject');
if($xobj !== NULL)
{
$xobj_addr = array_keys($xobj);
$xobj_addr = $xobj_addr[0];
$matches = array();
if(preg_match('/Resources \d+ \d+ R/', $xobj[$xobj_addr], $matches))
{
array_push($font_objs, $this->GetObjFromRef(substr($matches[0], 10))); //Resources+' '
}
}
}while($xobj !== NULL);
$fxobjs = array();
foreach($font_objs as $font_obj)
{
$fobj = $this->GetXObj($font_obj);
if($fobj !== NULL)
{
$xobj_addr = array_keys($fobj);
$xobj_addr = $xobj_addr[0];
$matches = array();
if(preg_match('/Font \d+ \d+ R/', $fobj[$xobj_addr], $matches))
{
array_push($fxobjs, $this->GetObjFromRef(substr($matches[0], 10))); //Resources+' '
}
}
}
$actual_font_objs = array();
foreach($fxobjs as $fxobj)
{
$f_obj = $this->GetXObj($fxobj);
if($f_obj !== NULL)
{
$xobj_addr = array_keys($f_obj);
$xobj_addr = $xobj_addr[0];
$matches = array();
if(preg_match('/F\d+/', $f_obj[$xobj_addr], $matches))
{
array_push($actual_font_objs, $this->GetObjFromRef(substr($matches[0], 10))); //Resources+' '
}
}
}
foreach($actual_font_objs as $fo)
{
$f = $this->GetXObj($fxobj);
if($f !== NULL)
{
$xobj_addr = array_keys($f);
$xobj_addr = $xobj_addr[0];
$matches = array();
$f_name = '';
if(preg_match('/BaseFont \/\S+/', $f[$xobj_addr], $matches))
{
$f_name = substr($matches[0], 10); // BaseFont+' '+'/'
}
if(preg_match('/FontDescriptor \d+ \d+ R/', $f[$xobj_addr], $matches))
{
$fd_replace_start_pos = strpos($this->raw_data, $matches[0]) + 15; //FontDescriptor+' ';
$fd_replace_end_pos = strpos($this->raw_data, 'R', $fd_replace_start_pos) + 1; //For 'R';
$s = substr($this->raw_data, 0, $fd_replace_start_pos);
$r = substr($this->embedded_font_addr[$f_name], 0, -3).'R'; //-4 for obj.
$e = substr($this->raw_data, $fd_replace_end_pos);
$this->raw_data = $s . $r . $e;
}
}
}
file_put_contents($filename, $this->raw_data);
}
public function GetFontInfo()
{
do
{
$this->ParseObj('/Type /Font');
}while($this->ParseObj('/Type /Font') == true);
return $this->GetNotEmbeddedFonts();
}
private function GetNotEmbeddedFonts()
{
$result = array();
foreach($this->font_list as $font)
{
if(array_key_exists('DescendantFonts', $font))
{
foreach($font['DescendantFonts'] as $desc_font)
{
if((array_key_exists('Embedded', $desc_font)) && (array_key_exists('BaseFont', $desc_font)))
if(!$desc_font['Embedded'])
array_push($result, $desc_font['BaseFont']);
}
}
else
{
if((array_key_exists('Embedded', $font)) && (array_key_exists('BaseFont', $font)))
if(!$font['Embedded'])
array_push($result, $font['BaseFont']);
}
}
for($i= 0; $i<count($result); $i++)
{
if(strpos($result[$i], ' ') || strpos($result[$i], ',')) //Adobe Acrobat system font name list. Needs improvment..
{
switch($r)
{
case 'Times New Roman':
$result[$i] = 'TimesNewRomanPSMT';
break;
case 'Times New Roman,Bold':
$result[$i] = 'TimesNewRomanPS-BoldMT';
break;
case 'Times New Roman,Italic':
$result[$i] = 'TimesNewRomanPS-ItalicMT';
break;
case 'Times New Roman,BoldItalic':
$result[$i] = 'TimesNewRomanPS-BoldItalicMT';
break;
case 'Aial':
$result[$i] = 'ArialMT';
break;
}
}
}
return $result;
}
}
$pdf = new PDF($filename);
$pdf->setSourceFile($filename);
$pdf->AddPage();
$tplIdx = $pdf->importPage(1);
$pdf->useTemplate($tplIdx);
$pdf->SetMargins(PDF_MARGIN_LEFT, 40, PDF_MARGIN_RIGHT);
$pdf->SetAutoPageBreak(true, 40);
$pdf->setFontSubsetting(false);
$not_embedded_fonts = $pdf->GetFontInfo();
foreach($not_embedded_fonts as $f)
{
$font_name = strtolower($f);
$font_name = preg_replace('/[^a-z0-9_]/', '', $font_name);
$search = array('bold', 'oblique', 'italic', 'regular');
$replace = array('b', 'i', 'i', '');
$font_name = str_replace($search, $replace, $font_name);
if($pdf->AddFont($font_name) !== false)
$pdf->SetFont($font_name);
else
print('"' .$font_name. '" font not found!');
}
$pdf->Output();
?>