Merge pull request #1345 from jberkenbilt/fix-qdf-output-file

Fix qdf output file

Merge pull request #1345 from jberkenbilt/fix-qdf-output-file
Fix qdf output file
Jay Berkenbilt · GitHub
2 parents a5cab082 a2fc5b52
Showing 6 changed files with 111 additions and 46 deletions
ChangeLog
manual/fix-qdf.1.in
manual/qdf.rst
manual/release-notes.rst
qpdf/fix-qdf.cc
qpdf/qtest/fix-qdf.test
+2025-02-02  Jay Berkenbilt  <ejb@ql.org>
+
+	* Have fix-qdf accept a second argument, interpreted as the output
+	file. Fixes #1330.
+
 2024-02-01  M Holger  <m.holger@qpdf.org>
 	* Bug fix: in qpdf CLI / QPDFJob throw a QPDFUsage exception if a
@@ -3,9 +3,11 @@
 fix-qdf \- repair PDF files in QDF form after editing
 .SH SYNOPSIS
 .B fix-qdf
-< \fIinfilename\fR > \fIoutfilename\fR
+[\fIinfilename\fR [\fIoutfilename\fR]]
 .SH DESCRIPTION
-The fix-qdf program is part of the qpdf package.
+The fix-qdf program is part of the qpdf package. With no arguments,
+fix-qdf reads from standard input and writes to standard output. With
+one argument, it reads from that file and writes to standard output.
 .PP
 The fix-qdf program reads a PDF file in QDF form and writes out
 the same file with stream lengths, cross-reference table entries, and
@@ -23,10 +23,18 @@ files are full of offset and length information that makes it hard to
 add or remove data. A QDF file is organized in a manner such that, if
 edits are kept within certain constraints, the
 :command:`fix-qdf` program, distributed with qpdf, is
-able to restore edited files to a correct state. The
-:command:`fix-qdf` program takes no command-line
-arguments. It reads a possibly edited QDF file from standard input and
-writes a repaired file to standard output.
+able to restore edited files to a correct state.
+
+.. code-block:: bash
+
+  fix-qdf [infilename [outfilename]]
+
+With no arguments, :command:`fix-qdf` reads the possibly-edited QDF
+file from standard input and writes a repaired file to standard
+output. You can also specify the input and output files as
+command-line arguments. With one argument, the argument is taken as an
+input file. With two arguments, the first argument is an input file,
+and the second is an output file.
 For another way to work with PDF files in an editor, see :ref:`json`.
 Using qpdf JSON format allows you to edit the PDF file semantically
@@ -38,6 +38,14 @@ Planned changes for future 12.x (subject to change):
 .. x.y.z: not yet released
+11.10.0: not yet released
+  - CLI Enhancements
+
+    - The :command:`fix-qdf` command now allows an output file to be
+      specified as an optional second argument. This is useful for
+      environments in which writing a binary file to standard output
+      doesn't work (such as PowerShell 5).
+
 11.9.1: June 7, 2024
   - Bug Fixes
@@ -4,23 +4,27 @@
 #include <qpdf/QUtil.hh>
 #include <cstdio>
 #include <cstring>
+#include <fstream>
 #include <iostream>
 #include <regex>
 #include <string_view>
+using namespace std::literals;
 static char const* whoami = nullptr;
 static void
 usage()
 {
-    std::cerr << "Usage: " << whoami << " [filename]" << std::endl;
-    exit(2);
+    std::cerr << "Usage: " << whoami << " [infilename [outfilename]]" << std::endl
+              << "infilename defaults to standard output" << std::endl
+              << "outfilename defaults to standard output" << std::endl;
 }
 class QdfFixer
 {
   public:
-    QdfFixer(std::string const& filename);
+    QdfFixer(std::string const& filename, std::ostream& out);
+    ~QdfFixer() = default;
     void processLines(std::string const& input);
   private:
@@ -31,6 +35,7 @@ class QdfFixer
     void writeBinary(unsigned long long val, size_t bytes);
     std::string filename;
+    std::ostream& out;
     enum {
         st_top,
         st_in_obj,
@@ -67,8 +72,9 @@ class QdfFixer
     std::string ostream_extends;
 };
-QdfFixer::QdfFixer(std::string const& filename) :
-    filename(filename)
+QdfFixer::QdfFixer(std::string const& filename, std::ostream& out) :
+    filename(filename),
+    out(out)
 {
 }
@@ -131,9 +137,9 @@ QdfFixer::processLines(std::string const&amp; input)
                 xref_offset = last_offset;
                 state = st_at_xref;
             }
-            std::cout << line;
+            out << line;
         } else if (state == st_in_obj) {
-            std::cout << line;
+            out << line;
             if (line.compare("stream\n"sv) == 0) {
                 state = st_in_stream;
                 stream_start = offset;
@@ -166,8 +172,8 @@ QdfFixer::processLines(std::string const&amp; input)
                 auto esize = 1 + xref_f1_nbytes + xref_f2_nbytes;
                 xref_size = 1 + xref.size();
                 auto length = xref_size * esize;
-                std::cout << "  /Length " << length << "\n"
-                          << "  /W [ 1 " << xref_f1_nbytes << " " << xref_f2_nbytes << " ]\n";
+                out << "  /Length " << length << "\n"
+                    << "  /W [ 1 " << xref_f1_nbytes << " " << xref_f2_nbytes << " ]\n";
                 state = st_in_xref_stream_dict;
             }
         } else if (state == st_in_ostream_dict) {
@@ -209,10 +215,10 @@ QdfFixer::processLines(std::string const&amp; input)
             if ((line.find("/Length"sv) != line.npos) || (line.find("/W"sv) != line.npos)) {
                 // already printed
             } else if (line.find("/Size"sv) != line.npos) {
-                auto xref_size = 1 + xref.size();
-                std::cout << "  /Size " << xref_size << "\n";
+                auto size = 1 + xref.size();
+                out << "  /Size " << size << "\n";
             } else {
-                std::cout << line;
+                out << line;
             }
             if (line.compare("stream\n"sv) == 0) {
                 writeBinary(0, 1);
@@ -232,9 +238,9 @@ QdfFixer::processLines(std::string const&amp; input)
                     writeBinary(f1, xref_f1_nbytes);
                     writeBinary(f2, xref_f2_nbytes);
                 }
-                std::cout << "\nendstream\nendobj\n\n"
-                          << "startxref\n"
-                          << xref_offset << "\n%%EOF\n";
+                out << "\nendstream\nendobj\n\n"
+                    << "startxref\n"
+                    << xref_offset << "\n%%EOF\n";
                 state = st_done;
             }
         } else if (state == st_in_stream) {
@@ -242,7 +248,7 @@ QdfFixer::processLines(std::string const&amp; input)
                 stream_length = QIntC::to_size(last_offset - stream_start);
                 state = st_after_stream;
             }
-            std::cout << line;
+            out << line;
         } else if (state == st_after_stream) {
             if (line.compare("%QDF: ignore_newline\n"sv) == 0) {
                 if (stream_length > 0) {
@@ -252,7 +258,7 @@ QdfFixer::processLines(std::string const&amp; input)
                 checkObjId(m[1].str());
                 state = st_in_length;
             }
-            std::cout << line;
+            out << line;
         } else if (state == st_in_length) {
             if (!matches(re_num)) {
                 fatal(filename + ":" + std::to_string(lineno) + ": expected integer");
@@ -260,29 +266,29 @@ QdfFixer::processLines(std::string const&amp; input)
             std::string new_length = std::to_string(stream_length) + "\n";
             offset -= QIntC::to_offset(line.length());
             offset += QIntC::to_offset(new_length.length());
-            std::cout << new_length;
+            out << new_length;
             state = st_top;
         } else if (state == st_at_xref) {
             auto n = xref.size();
-            std::cout << "0 " << 1 + n << "\n0000000000 65535 f \n";
+            out << "0 " << 1 + n << "\n0000000000 65535 f \n";
             for (auto const& e: xref) {
-                std::cout << QUtil::int_to_string(e.getOffset(), 10) << " 00000 n \n";
+                out << QUtil::int_to_string(e.getOffset(), 10) << " 00000 n \n";
             }
             state = st_before_trailer;
         } else if (state == st_before_trailer) {
             if (line.compare("trailer <<\n"sv) == 0) {
-                std::cout << line;
+                out << line;
                 state = st_in_trailer;
             }
             // no output
         } else if (state == st_in_trailer) {
             if (matches(re_size_n)) {
-                std::cout << "  /Size " << 1 + xref.size() << "\n";
+                out << "  /Size " << 1 + xref.size() << "\n";
             } else {
-                std::cout << line;
+                out << line;
             }
             if (line.compare(">>\n"sv) == 0) {
-                std::cout << "startxref\n" << xref_offset << "\n%%EOF\n";
+                out << "startxref\n" << xref_offset << "\n%%EOF\n";
                 state = st_done;
             }
         } else if (state == st_done) {
@@ -332,9 +338,9 @@ QdfFixer::writeOstream()
     }
     dict_data += ">>\n";
     offset_adjust += QIntC::to_offset(dict_data.length());
-    std::cout << dict_data << "stream\n" << offsets;
+    out << dict_data << "stream\n" << offsets;
     for (auto const& o: ostream) {
-        std::cout << o;
+        out << o;
     }
     for (auto const& o: ostream_discarded) {
@@ -361,7 +367,7 @@ QdfFixer::writeBinary(unsigned long long val, size_t bytes)
         data[i - 1] = static_cast<char>(val & 0xff); // i.e. val % 256
         val >>= 8;                                   // i.e. val = val / 256
     }
-    std::cout << data;
+    out << data;
 }
 static int
@@ -370,27 +376,44 @@ realmain(int argc, char* argv[])
     whoami = QUtil::getWhoami(argv[0]);
     QUtil::setLineBuf(stdout);
     char const* filename = nullptr;
-    if (argc > 2) {
+    char const* outfilename = nullptr;
+    if (argc > 3) {
         usage();
     } else if ((argc > 1) && (strcmp(argv[1], "--version") == 0)) {
         std::cout << whoami << " from qpdf version " << QPDF::QPDFVersion() << std::endl;
         return 0;
     } else if ((argc > 1) && (strcmp(argv[1], "--help") == 0)) {
         usage();
-    } else if (argc == 2) {
+    } else if (argc >= 2) {
         filename = argv[1];
+        if (argc == 3) {
+            outfilename = argv[2];
+        }
     }
-    std::string input;
-    if (filename == nullptr) {
-        filename = "standard input";
-        QUtil::binary_stdin();
-        input = QUtil::read_file_into_string(stdin);
-    } else {
-        input = QUtil::read_file_into_string(filename);
+    try {
+        std::string input;
+        if (filename == nullptr) {
+            filename = "standard input";
+            QUtil::binary_stdin();
+            input = QUtil::read_file_into_string(stdin);
+        } else {
+            input = QUtil::read_file_into_string(filename);
+        }
+        std::unique_ptr<std::ofstream> out = nullptr;
+        if (outfilename) {
+            out = std::make_unique<std::ofstream>(outfilename, std::ios::binary);
+            if (out->fail()) {
+                QUtil::throw_system_error("open "s + outfilename);
+            }
+        } else {
+            QUtil::binary_stdout();
+        }
+        QdfFixer qf(filename, out ? *out : std::cout);
+        qf.processLines(input);
+    } catch (std::exception& e) {
+        std::cerr << whoami << ": error: " << e.what() << std::endl;
+        exit(qpdf_exit_error);
     }
-    QUtil::binary_stdout();
-    QdfFixer qf(filename);
-    qf.processLines(input);
     return 0;
 }
@@ -14,7 +14,7 @@ cleanup();
 my $td = new TestDriver('fix-qdf');
-my $n_tests = 5;
+my $n_tests = 11;
 for (my $n = 1; $n <= 2; ++$n)
 {
@@ -23,6 +23,15 @@ for (my $n = 1; $n &lt;= 2; ++$n)
                  {$td->FILE => "fix$n.qdf.out",
                   $td->EXIT_STATUS => 0});
+    $td->runtest("fix-qdf $n with named output",
+                 {$td->COMMAND => "fix-qdf fix$n.qdf a.pdf"},
+                 {$td->STRING => "",
+                  $td->EXIT_STATUS => 0});
+
+    $td->runtest("check fix-qdf $n output",
+                 {$td->FILE => "a.pdf"},
+                 {$td->FILE => "fix$n.qdf.out"});
+
     $td->runtest("identity fix-qdf $n",
                  {$td->COMMAND => "fix-qdf fix$n.qdf.out"},
                  {$td->FILE => "fix$n.qdf.out",
@@ -54,5 +63,15 @@ $td-&gt;runtest(&quot;fix-qdf with big object stream&quot;, # &gt; 255 objects in a stream
              {$td->FILE => "big-ostream.pdf",
               $td->EXIT_STATUS => 0});
+$td->runtest("fix-qdf error opening input",
+             {$td->COMMAND => "fix-qdf /does/not/exist/potato.pdf"},
+             {$td->REGEXP => "^fix-qdf: error: open .*/does/not/exist/potato.pdf: .*",
+              $td->EXIT_STATUS => 2});
+
+$td->runtest("fix-qdf error opening output", # > 255 objects in a stream
+             {$td->COMMAND => "fix-qdf fix1.qdf /does/not/exist/salad.pdf"},
+             {$td->REGEXP => "^fix-qdf: error: open .*/does/not/exist/salad.pdf: .*",
+              $td->EXIT_STATUS => 2});
+
 cleanup();
 $td->report($n_tests);