Implement --set-page-labels and supporting API (fixes #939)

Jay Berkenbilt
1 parent 543038c5
Showing 11 changed files with 226 additions and 11 deletions
ChangeLog
include/qpdf/Constants.h
include/qpdf/QPDFJob.hh
include/qpdf/QPDFPageLabelDocumentHelper.hh
libqpdf/QPDFJob.cc
libqpdf/QPDFJob_config.cc
libqpdf/QPDFPageLabelDocumentHelper.cc
manual/release-notes.rst
qpdf/qtest/page-labels.test
qpdf/qtest/qpdf/set-page-labels.json
qpdf/qtest/qpdf/set-page-labels.out
+2024-01-05  Jay Berkenbilt  <ejb@ql.org>
+
+	* Add --set-page-labels command-line argument and supporting API.
+	Fixes #939.
+	  - QPDFJob::Config::setPageLabels
+	  - pdf_page_label_e enumerated type
+	  - QPDFPageLabelDocumentHelper::pageLabelDict
+
 2024-01-01  Jay Berkenbilt  <ejb@ql.org>
 	* Support comma-separated numeric values with --collate to select
@@ -232,4 +232,14 @@ enum pdf_annotation_flag_e {
 /* Encryption/password status for QPDFJob */
 enum qpdf_encryption_status_e { qpdf_es_encrypted = 1 << 0, qpdf_es_password_incorrect = 1 << 1 };
+/* Page label types */
+enum qpdf_page_label_e {
+    pl_none,
+    pl_digits,
+    pl_alpha_lower,
+    pl_alpha_upper,
+    pl_roman_lower,
+    pl_roman_upper,
+};
+
 #endif /* QPDFCONSTANTS_H */
@@ -296,7 +296,8 @@ class QPDFJob
         Config* config;
     };
-    class PageLabelsConfig {
+    class PageLabelsConfig
+    {
         friend class QPDFJob;
         friend class Config;
@@ -458,6 +459,22 @@ class QPDFJob
         std::vector<int> repeat_pagenos;
     };
+    struct PageLabelSpec
+    {
+        PageLabelSpec(
+            int first_page, qpdf_page_label_e label_type, int start_num, std::string_view prefix) :
+            first_page(first_page),
+            label_type(label_type),
+            start_num(start_num),
+            prefix(prefix)
+        {
+        }
+        int first_page;
+        qpdf_page_label_e label_type;
+        int start_num{1};
+        std::string prefix;
+    };
+
     // Helper functions
     static void usage(std::string const& msg);
     static JSON json_schema(int json_version, std::set<std::string>* keys = nullptr);
@@ -694,7 +711,7 @@ class QPDFJob
         bool json_output{false};
         std::string update_from_json;
         bool report_mem_usage{false};
-        std::vector<std::string> page_label_specs;
+        std::vector<PageLabelSpec> page_label_specs;
     };
     std::shared_ptr<Members> m;
 };
@@ -49,6 +49,11 @@ class QPDFPageLabelDocumentHelper: public QPDFDocumentHelper
     QPDF_DLL
     bool hasPageLabels();
+    // Helper function to create a dictionary suitable for adding to the /PageLabels numbers tree.
+    QPDF_DLL
+    static QPDFObjectHandle
+    pageLabelDict(qpdf_page_label_e label_type, int start_num, std::string_view prefix);
+
     // Return a page label dictionary representing the page label for the given page. The page does
     // not need to appear explicitly in the page label dictionary. This method will adjust /St as
     // needed to produce a label that is suitable for the page.
@@ -2172,6 +2172,37 @@ QPDFJob::handleTransformations(QPDF&amp; pdf)
     if (m->remove_page_labels) {
         pdf.getRoot().removeKey("/PageLabels");
     }
+    if (!m->page_label_specs.empty()) {
+        auto nums = QPDFObjectHandle::newArray();
+        auto n_pages = QIntC::to_int(dh.getAllPages().size());
+        int last_page_seen{0};
+        for (auto& spec: m->page_label_specs) {
+            if (spec.first_page < 0) {
+                spec.first_page = n_pages + 1 + spec.first_page;
+            }
+            if (last_page_seen == 0) {
+                if (spec.first_page != 1) {
+                    throw std::runtime_error(
+                        "the first page label specification must start with page 1");
+                }
+            } else if (spec.first_page <= last_page_seen) {
+                throw std::runtime_error(
+                    "page label specifications must be in order by first page");
+            }
+            if (spec.first_page > n_pages) {
+                throw std::runtime_error(
+                    "page label spec: page " + std::to_string(spec.first_page) +
+                    " is more than the total number of pages (" + std::to_string(n_pages) + ")");
+            }
+            last_page_seen = spec.first_page;
+            nums.appendItem(QPDFObjectHandle::newInteger(spec.first_page - 1));
+            nums.appendItem(QPDFPageLabelDocumentHelper::pageLabelDict(
+                spec.label_type, spec.start_num, spec.prefix));
+        }
+        auto page_labels = QPDFObjectHandle::newDictionary();
+        page_labels.replaceKey("/Nums", nums);
+        pdf.getRoot().replaceKey("/PageLabels", page_labels);
+    }
     if (!m->attachments_to_remove.empty()) {
         QPDFEmbeddedFileDocumentHelper efdh(pdf);
         for (auto const& key: m->attachments_to_remove) {
@@ -3019,10 +3050,9 @@ QPDFJob::writeOutfile(QPDF&amp; pdf)
             try {
                 QUtil::remove_file(backup.c_str());
             } catch (QPDFSystemError& e) {
-                *m->log->getError()
-                    << m->message_prefix << ": unable to delete original file (" << e.what() << ");"
-                    << " original file left in " << backup
-                    << ", but the input was successfully replaced\n";
+                *m->log->getError() << m->message_prefix << ": unable to delete original file ("
+                                    << e.what() << ");" << " original file left in " << backup
+                                    << ", but the input was successfully replaced\n";
             }
         }
     }
 #include <qpdf/QPDFJob.hh>
+#include <regex>
+
 #include <qpdf/QPDFLogger.hh>
 #include <qpdf/QTC.hh>
 #include <qpdf/QUtil.hh>
@@ -1062,11 +1064,50 @@ QPDFJob::Config::encrypt(
 QPDFJob::Config*
 QPDFJob::Config::setPageLabels(const std::vector<std::string>& specs)
 {
-    // XXX validate
-    for (auto const& xxx: specs) {
-        std::cout << "XXX config: spec: " << xxx << std::endl;
+    static std::regex page_label_re(R"(^(z|r?\d+):([DaArR])?(?:/(\d+)?(?:/(.+)?)?)?$)");
+    o.m->page_label_specs.clear();
+    for (auto const& spec: specs) {
+        std::smatch match;
+        if (!std::regex_match(spec, match, page_label_re)) {
+            usage("page label spec must be n:[D|a|A|r|R][/start[/prefix]]");
+        }
+        auto first_page_str = match[1].str();
+        int first_page;
+        if (first_page_str == "z") {
+            first_page = -1;
+        } else if (first_page_str.at(0) == 'r') {
+            first_page = -QUtil::string_to_int(first_page_str.substr(1).c_str());
+        } else {
+            first_page = QUtil::string_to_int(first_page_str.c_str());
+        }
+        auto label_type_ch = match[2].matched ? match[2].str().at(0) : '\0';
+        qpdf_page_label_e label_type;
+        switch (label_type_ch) {
+        case 'D':
+            label_type = pl_digits;
+            break;
+        case 'a':
+            label_type = pl_alpha_lower;
+            break;
+        case 'A':
+            label_type = pl_alpha_upper;
+            break;
+        case 'r':
+            label_type = pl_roman_lower;
+            break;
+        case 'R':
+            label_type = pl_roman_upper;
+            break;
+        default:
+            label_type = pl_none;
+        }
+
+        auto start_num = match[3].matched ? QUtil::string_to_int(match[3].str().c_str()) : 1;
+        auto prefix = match[4].matched ? match[4].str() : "";
+        // We can't check ordering until we know how many pages there are, so that is delayed until
+        // near the end.
+        o.m->page_label_specs.emplace_back(first_page, label_type, start_num, prefix);
     }
-    o.m->page_label_specs = specs;
     return this;
 }
@@ -99,3 +99,36 @@ QPDFPageLabelDocumentHelper::getLabelsForPageRange(
         }
     }
 }
+
+QPDFObjectHandle
+QPDFPageLabelDocumentHelper::pageLabelDict(
+    qpdf_page_label_e label_type, int start_num, std::string_view prefix)
+{
+    auto num = "<< /Type /PageLabel >>"_qpdf;
+    switch (label_type) {
+    case pl_none:
+        break;
+    case pl_digits:
+        num.replaceKey("/S", "/D"_qpdf);
+        break;
+    case pl_alpha_lower:
+        num.replaceKey("/S", "/a"_qpdf);
+        break;
+    case pl_alpha_upper:
+        num.replaceKey("/S", "/A"_qpdf);
+        break;
+    case pl_roman_lower:
+        num.replaceKey("/S", "/r"_qpdf);
+        break;
+    case pl_roman_upper:
+        num.replaceKey("/S", "/R"_qpdf);
+        break;
+    }
+    if (!prefix.empty()) {
+        num.replaceKey("/P", QPDFObjectHandle::newUnicodeString(std::string(prefix)));
+    }
+    if (start_num != 1) {
+        num.replaceKey("/St", QPDFObjectHandle::newInteger(start_num));
+    }
+    return num;
+}
@@ -38,7 +38,7 @@ Planned changes for future 12.x (subject to change):
 .. x.y.z: not yet released
-11.7.1: not yet released
+11.8.0: not yet released
   - Bug fixes:
     - When flattening annotations, preserve hyperlinks and other
@@ -54,6 +54,19 @@ Planned changes for future 12.x (subject to change):
       :qpdf:ref:`--collate` to select different numbers of pages from
       different groups.
+    - Add :qpdf:ref:`--set-page-labels` option to completely override
+      page labels in the output.
+
+  - Library Enhancements
+
+    - Add API to support :qpdf:ref:`--set-page-labels`:
+
+      - ``QPDFJob::Config::setPageLabels``
+
+      - ``pdf_page_label_e`` enumerated type
+
+      - ``QPDFPageLabelDocumentHelper::pageLabelDict``
+
 11.7.0: December 24, 2023
   - Bug fixes:
@@ -29,5 +29,44 @@ $td-&gt;runtest(&quot;no page labels&quot;,
              {$td->FILE => "no-page-labels.out", $td->EXIT_STATUS => 0},
              $td->NORMALIZE_NEWLINES);
+# --set-page-labels
+my @errors = (
+    ["quack", ".*page label spec must be.*"],
+    ["5:r 10:D", ".*the first page .*must start with page 1.*"],
+    ["1:r 10:D 31:A",
+     ".*page 31 is more than the total number of pages \\(30\\).*"],
+);
+$n_tests += scalar(@errors);
+
+foreach my $d (@errors)
+{
+    my ($specs, $err) = @$d;
+    $td->runtest("error ($specs)",
+                 {$td->COMMAND => "qpdf --set-page-labels $specs --" .
+                      " page-labels-num-tree.pdf a.pdf"},
+                 {$td->REGEXP => $err, $td->EXIT_STATUS => 2},
+                 $td->NORMALIZE_NEWLINES);
+}
+
+$n_tests += 4;
+$td->runtest("set page labels",
+             {$td->COMMAND => "qpdf page-labels-num-tree.pdf a.pdf" .
+                  " --set-page-labels 1:a 3:R/2 6:r//Z- 8:A/17 r10:D/3 28: z://end --"},
+             {$td->STRING => "", $td->EXIT_STATUS => 0},
+             $td->NORMALIZE_NEWLINES);
+$td->runtest("after set page labels",
+             {$td->COMMAND => "test_driver 47 a.pdf"},
+             {$td->FILE => "set-page-labels.out", $td->EXIT_STATUS => 0},
+             $td->NORMALIZE_NEWLINES);
+$td->runtest("set page labels (json)",
+             {$td->COMMAND => "qpdf page-labels-num-tree.pdf b.pdf" .
+                  " --job-json-file=set-page-labels.json"},
+             {$td->STRING => "", $td->EXIT_STATUS => 0},
+             $td->NORMALIZE_NEWLINES);
+$td->runtest("after set page labels",
+             {$td->COMMAND => "test_driver 47 b.pdf"},
+             {$td->FILE => "set-page-labels.out", $td->EXIT_STATUS => 0},
+             $td->NORMALIZE_NEWLINES);
+
 cleanup();
 $td->report($n_tests);
+{
+  "setPageLabels": [
+    "1:a",
+    "3:R/2",
+    "6:r//Z-",
+    "8:A/17",
+    "r10:D/3",
+    "28:",
+    "z://end"
+  ]
+}
+1 << /S /a /St 1 >>
+3 << /S /R /St 2 >>
+6 << /P (Z-) /S /r /St 1 >>
+8 << /S /A /St 17 >>
+21 << /S /D /St 3 >>
+28 << /St 1 >>
+30 << /P (end) /St 1 >>
+test 47 done