Implement Pl_RunLength pipeline

TODO

@@ -33,9 +33,9 @@ Soon
    prioritized so that we can poll all registered filters to see
    whether they are capable of filtering a particular stream.
 
- * If possible, consider adding RLE, CCITT3, CCITT4, or any other easy
-   filters.  For some reference code that we probably can't use but
-   may be handy anyway, see
+ * If possible, consider adding CCITT3, CCITT4, or any other easy
+   filters. For some reference code that we probably can't use but may
+   be handy anyway, see
    http://partners.adobe.com/public/developer/ps/sdk/index_archive.html
 
  * If possible, support the following types of broken files:

include/qpdf/Pl_RunLength.hh

@@ -0,0 +1,42 @@
+// Copyright (c) 2005-2015 Jay Berkenbilt
+//
+// This file is part of qpdf.  This software may be distributed under
+// the terms of version 2 of the Artistic License which may be found
+// in the source distribution.  It is provided "as is" without express
+// or implied warranty.
+
+#ifndef __PL_RUNLENGTH_HH__
+#define __PL_RUNLENGTH_HH__
+
+#include <qpdf/Pipeline.hh>
+
+class Pl_RunLength: public Pipeline
+{
+  public:
+    enum action_e { a_encode, a_decode };
+
+    QPDF_DLL
+    Pl_RunLength(char const* identifier, Pipeline* next,
+                 action_e action);
+    QPDF_DLL
+    virtual ~Pl_RunLength();
+
+    QPDF_DLL
+    virtual void write(unsigned char* data, size_t len);
+    QPDF_DLL
+    virtual void finish();
+
+  private:
+    void encode(unsigned char* data, size_t len);
+    void decode(unsigned char* data, size_t len);
+    void flush_encode();
+
+    enum state_e { st_top, st_copying, st_run };
+
+    action_e action;
+    state_e state;
+    unsigned char buf[128];
+    unsigned int length;
+};
+
+#endif // __PL_RUNLENGTH_HH__

libqpdf/Pl_RunLength.cc

@@ -0,0 +1,171 @@
+#include <qpdf/Pl_RunLength.hh>
+
+#include <qpdf/QUtil.hh>
+#include <qpdf/QTC.hh>
+
+Pl_RunLength::Pl_RunLength(char const* identifier, Pipeline* next,
+                           action_e action) :
+    Pipeline(identifier, next),
+    action(action),
+    state(st_top),
+    length(0)
+{
+}
+
+Pl_RunLength::~Pl_RunLength()
+{
+}
+
+void
+Pl_RunLength::write(unsigned char* data, size_t len)
+{
+    if (this->action == a_encode)
+    {
+        encode(data, len);
+    }
+    else
+    {
+        decode(data, len);
+    }
+}
+
+void
+Pl_RunLength::encode(unsigned char* data, size_t len)
+{
+    for (size_t i = 0; i < len; ++i)
+    {
+        if ((this->state == st_top) != (this->length <= 1))
+        {
+            throw std::logic_error(
+                "Pl_RunLength::encode: state/length inconsistency");
+        }
+        unsigned char ch = data[i];
+        if ((this->length > 0) &&
+            ((this->state == st_copying) || (this->length < 128)) &&
+            (ch == this->buf[this->length-1]))
+        {
+            QTC::TC("libtests", "Pl_RunLength: switch to run",
+                    (this->length == 128) ? 0 : 1);
+            if (this->state == st_copying)
+            {
+                --this->length;
+                flush_encode();
+                this->buf[0] = ch;
+                this->length = 1;
+            }
+            this->state = st_run;
+            this->buf[this->length] = ch;
+            ++this->length;
+        }
+        else
+        {
+            if ((this->length == 128) || (this->state == st_run))
+            {
+                flush_encode();
+            }
+            else if (this->length > 0)
+            {
+                this->state = st_copying;
+            }
+            this->buf[this->length] = ch;
+            ++this->length;
+        }
+    }
+}
+
+void
+Pl_RunLength::decode(unsigned char* data, size_t len)
+{
+    for (size_t i = 0; i < len; ++i)
+    {
+        unsigned char ch = data[i];
+        switch (this->state)
+        {
+          case st_top:
+            if (ch < 128)
+            {
+                // length represents remaining number of bytes to copy
+                this->length = 1 + ch;
+                this->state = st_copying;
+            }
+            else if (ch > 128)
+            {
+                // length represents number of copies of next byte
+                this->length = 257 - ch;
+                this->state = st_run;
+            }
+            else // ch == 128
+            {
+                // EOD; stay in this state
+            }
+            break;
+
+          case st_copying:
+            this->getNext()->write(&ch, 1);
+            if (--this->length == 0)
+            {
+                this->state = st_top;
+            }
+            break;
+
+          case st_run:
+            for (unsigned int j = 0; j < this->length; ++j)
+            {
+                this->getNext()->write(&ch, 1);
+            }
+            this->state = st_top;
+            break;
+        }
+    }
+}
+
+void
+Pl_RunLength::flush_encode()
+{
+    if (this->length == 128)
+    {
+        QTC::TC("libtests", "Pl_RunLength flush full buffer",
+                (this->state == st_copying ? 0 :
+                 this->state == st_run ? 1 :
+                 -1));
+    }
+    if (this->length == 0)
+    {
+        QTC::TC("libtests", "Pl_RunLength flush empty buffer");
+    }
+    if (this->state == st_run)
+    {
+        if ((this->length < 2) || (this->length > 128))
+        {
+            throw std::logic_error(
+                "Pl_RunLength: invalid length in flush_encode for run");
+        }
+        unsigned char ch = static_cast<unsigned char>(257 - this->length);
+        this->getNext()->write(&ch, 1);
+        this->getNext()->write(&this->buf[0], 1);
+    }
+    else if (this->length > 0)
+    {
+        unsigned char ch = static_cast<unsigned char>(this->length - 1);
+        this->getNext()->write(&ch, 1);
+        this->getNext()->write(this->buf, this->length);
+    }
+    this->state = st_top;
+    this->length = 0;
+}
+
+void
+Pl_RunLength::finish()
+{
+    // When decoding, we might have read a length byte not followed by
+    // data, which means the stream was terminated early, but we will
+    // just ignore this case since this is the only sensible thing to
+    // do.
+    if (this->action == a_encode)
+    {
+        flush_encode();
+        unsigned char ch = 128;
+        this->getNext()->write(&ch, 1);
+    }
+    this->getNext()->finish();
+}

libqpdf/build.mk

@@ -28,6 +28,7 @@ SRCS_libqpdf = \
 	libqpdf/Pl_PNGFilter.cc \
 	libqpdf/Pl_QPDFTokenizer.cc \
 	libqpdf/Pl_RC4.cc \
+	libqpdf/Pl_RunLength.cc \
 	libqpdf/Pl_SHA2.cc \
 	libqpdf/Pl_StdioFile.cc \
 	libqpdf/QPDF.cc \

libtests/build.mk

@@ -14,6 +14,7 @@ BINS_libtests = \
 	qutil \
 	random \
 	rc4 \
+	runlength \
 	sha2
 
 TARGETS_libtests = $(foreach B,$(BINS_libtests),libtests/$(OUTPUT_DIR)/$(call binname,$(B)))

libtests/libtests.testcov

@@ -24,3 +24,6 @@ InputSource start_chars matched but not check 0
 InputSource not enough bytes 0
 InputSource findLast found more than one 0
 InputSource found match at buf[0] 0
+Pl_RunLength: switch to run 1
+Pl_RunLength flush full buffer 1
+Pl_RunLength flush empty buffer 0

libtests/qtest/runlength.test

@@ -0,0 +1,75 @@
+#!/usr/bin/env perl
+require 5.008;
+use warnings;
+use strict;
+
+chdir("runlength") or die "chdir testdir failed: $!\n";
+
+require TestDriver;
+
+my $td = new TestDriver('runlength');
+
+cleanup();
+
+my @files = (
+    "01",                       # basic case, ends with copy
+    "02",                       # basic case, ends with run
+    "03",                       # long run run
+    "04",                       # ends with copy, length % 128 == 0
+    "05",                       # run starts at byte 128
+    "empty",                    # empty file
+    );
+
+# Create this rather than committing an empty file, which always looks
+# like an error.
+open(F, ">empty");
+close(F);
+
+foreach my $f (@files)
+{
+    $td->runtest("encode $f",
+                 {$td->COMMAND => "runlength -encode $f a"},
+                 {$td->STRING => "", $td->EXIT_STATUS => 0});
+    $td->runtest("check encoded output",
+                 {$td->FILE => "a"},
+                 {$td->FILE => "$f.encoded"});
+    $td->runtest("decode $f.encoded",
+                 {$td->COMMAND => "runlength -decode $f.encoded a"},
+                 {$td->STRING => "", $td->EXIT_STATUS => 0});
+    $td->runtest("check decoded output",
+                 {$td->FILE => "a"},
+                 {$td->FILE => "$f"});
+}
+
+concatenate("01.encoded", "02.encoded", "concat.encoded");
+concatenate("01", "02", "concat");
+
+$td->runtest("decode with embedded EOD",
+             {$td->COMMAND => "runlength -decode concat.encoded a"},
+             {$td->STRING => "", $td->EXIT_STATUS => 0});
+$td->runtest("check decoded output",
+             {$td->FILE => "a"},
+             {$td->FILE => "concat"});
+
+cleanup();
+
+$td->report(2 + (4 * scalar(@files)));
+
+sub cleanup
+{
+    system("rm -f a concat concat.encoded empty");
+}
+
+sub concatenate
+{
+    my ($a, $b, $out) = @_;
+    open(F, ">$out");
+    foreach my $f ($a, $b)
+    {
+        local $/ = undef;
+        open(G, "<$f");
+        print F <G>;
+        close(G);
+    }
+    close(F);
+}

libtests/qtest/runlength/01

@@ -0,0 +1 @@
+wwwwwwwwwwwwwwwwwwwwwwwwwqqqqqrstv

libtests/qtest/runlength/01.encoded

@@ -0,0 +1 @@
+鑧黴rstv€

libtests/qtest/runlength/02

@@ -0,0 +1 @@
+wwwwwwwwwwwwwwwwwwwwwwwwwqqqqqrstvxxxxxxxxxxxxxxxxxxxxxxx

libtests/qtest/runlength/02.encoded

@@ -0,0 +1 @@
+鑧黴rstv陎€

libtests/qtest/runlength/04

@@ -0,0 +1 @@
+wwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwabababababababababababababababababababababababababababababababababababababababababababababababababababababababababababababababab

libtests/qtest/runlength/04.encoded

@@ -0,0 +1 @@
+亀abababababababababababababababababababababababababababababababababababababababababababababababababababababababababababababababab€

libtests/qtest/runlength/05

@@ -0,0 +1 @@
+ababababababababababababababababababababababababababababababababababababababababababababababababababababababababababababababababbbbbb

libtests/qtest/runlength/05.encoded

@@ -0,0 +1 @@
+~abababababababababababababababababababababababababababababababababababababababababababababababababababababababababababababababa鸼€

libtests/qtest/runlength/empty.encoded

@@ -0,0 +1 @@
+<EFBFBD>

libtests/runlength.cc

@@ -0,0 +1,47 @@
+#include <qpdf/Pl_RunLength.hh>
+#include <qpdf/Pl_StdioFile.hh>
+#include <qpdf/QUtil.hh>
+
+#include <stdio.h>
+#include <string.h>
+#include <iostream>
+#include <stdlib.h>
+
+int main(int argc, char* argv[])
+{
+    if (argc != 4)
+    {
+	std::cerr << "Usage: runlength {-encode|-decode} infile outfile"
+                  << std::endl;
+	exit(2);
+    }
+
+    bool encode = (strcmp("-encode", argv[1]) == 0);
+    char* infilename = argv[2];
+    char* outfilename = argv[3];
+
+    FILE* infile = QUtil::safe_fopen(infilename, "rb");
+    FILE* outfile = QUtil::safe_fopen(outfilename, "wb");
+    Pl_StdioFile out("stdout", outfile);
+    unsigned char buf[100];
+    bool done = false;
+    Pl_RunLength rl(
+        "runlength", &out,
+        (encode ? Pl_RunLength::a_encode : Pl_RunLength::a_decode));
+    while (! done)
+    {
+	size_t len = fread(buf, 1, sizeof(buf), infile);
+	if (len <= 0)
+	{
+	    done = true;
+	}
+	else
+	{
+	    rl.write(buf, len);
+	}
+    }
+    rl.finish();
+    fclose(infile);
+    fclose(outfile);
+    return 0;
+}