Add QUtil methods for dealing with PDF timestamp strings

Jay Berkenbilt
1 parent bfbeec54
Showing 6 changed files with 189 additions and 1 deletions
ChangeLog
include/qpdf/QUtil.hh
libqpdf/QUtil.cc
libtests/qtest/qutil/qutil.out
libtests/qutil.cc
manual/qpdf-manual.xml
+2021-02-09  Jay Berkenbilt  <ejb@ql.org>
+
+	* Add methods to QUtil for working with PDF timestamp strings:
+	pdf_time_to_qpdf_time, qpdf_time_to_pdf_time,
+	get_current_qpdf_time.
+
 2021-02-07  Jay Berkenbilt  <ejb@ql.org>
 	* Add new functions QUtil::pipe_file and QUtil::file_provider for
@@ -158,7 +158,6 @@ namespace QUtil
     QPDF_DLL
     void setLineBuf(FILE*);
-
     // May modify argv0
     QPDF_DLL
     char* getWhoami(char* argv0);
@@ -172,6 +171,51 @@ namespace QUtil
     QPDF_DLL
     time_t get_current_time();
+    // Portable structure representing a point in time with second
+    // granularity and time zone offset
+    struct QPDFTime
+    {
+        QPDFTime() = default;
+        QPDFTime(QPDFTime const&) = default;
+        QPDFTime& operator=(QPDFTime const&) = default;
+        QPDFTime(int year, int month, int day, int hour,
+                 int minute, int second, int tz_delta) :
+            year(year),
+            month(month),
+            day(day),
+            hour(hour),
+            minute(minute),
+            second(second),
+            tz_delta(tz_delta)
+        {
+        }
+        int year;               // actual year, no 1900 stuff
+        int month;              // 1--12
+        int day;                // 1--31
+        int hour;
+        int minute;
+        int second;
+        int tz_delta;           // minutes before UTC
+    };
+
+    QPDF_DLL
+    QPDFTime get_current_qpdf_time();
+
+    // Convert a QPDFTime structure to a PDF timestamp string, which
+    // is "D:yyyymmddhhmmss<z>" where <z> is either "Z" for UTC or
+    // "-hh'mm'" or "+hh'mm'" for timezone offset. Examples:
+    // "D:20210207161528-05'00'", "D:20210207211528Z". See
+    // get_current_qpdf_time and the QPDFTime structure above.
+    QPDF_DLL
+    std::string qpdf_time_to_pdf_time(QPDFTime const&);
+
+    // Convert a PDF timestamp string to a QPDFTime. If syntactically
+    // valid, return true and fill in qtm. If not valid, return false,
+    // and do not modify qtm. If qtm is null, just check the validity
+    // of the string.
+    QPDF_DLL
+    bool pdf_time_to_qpdf_time(std::string const&, QPDFTime* qtm = nullptr);
+
     // Return a string containing the byte representation of the UTF-8
     // encoding for the unicode value passed in.
     QPDF_DLL
@@ -23,6 +23,7 @@
 #include <fcntl.h>
 #include <memory>
 #include <locale>
+#include <regex>
 #ifndef QPDF_NO_WCHAR_T
 # include <cwchar>
 #endif
@@ -823,6 +824,108 @@ QUtil::get_current_time()
 #endif
 }
+QUtil::QPDFTime
+QUtil::get_current_qpdf_time()
+{
+#ifdef _WIN32
+    SYSTEMTIME ltime;
+    GetLocalTime(&ltime);
+    TIME_ZONE_INFORMATION tzinfo;
+    GetTimeZoneInformation(&tzinfo);
+    return QPDFTime(static_cast<int>(ltime.wYear),
+                    static_cast<int>(ltime.wMonth),
+                    static_cast<int>(ltime.wDay),
+                    static_cast<int>(ltime.wHour),
+                    static_cast<int>(ltime.wMinute),
+                    static_cast<int>(ltime.wSecond),
+                    static_cast<int>(tzinfo.Bias));
+#else
+    struct tm ltime;
+    time_t now = time(0);
+    tzset();
+    localtime_r(&now, &ltime);
+    return QPDFTime(static_cast<int>(ltime.tm_year + 1900),
+                    static_cast<int>(ltime.tm_mon + 1),
+                    static_cast<int>(ltime.tm_mday),
+                    static_cast<int>(ltime.tm_hour),
+                    static_cast<int>(ltime.tm_min),
+                    static_cast<int>(ltime.tm_sec),
+                    static_cast<int>(timezone / 60));
+#endif
+}
+
+std::string
+QUtil::qpdf_time_to_pdf_time(QPDFTime const& qtm)
+{
+    std::string tz_offset;
+    int t = qtm.tz_delta;
+    if (t == 0)
+    {
+        tz_offset = "Z";
+    }
+    else
+    {
+        if (t < 0)
+        {
+            t = -t;
+            tz_offset += "+";
+        }
+        else
+        {
+            tz_offset += "-";
+        }
+        tz_offset +=
+            QUtil::int_to_string(t / 60, 2) + "'" +
+            QUtil::int_to_string(t % 60, 2) + "'";
+    }
+    return ("D:" +
+            QUtil::int_to_string(qtm.year, 4) +
+            QUtil::int_to_string(qtm.month, 2) +
+            QUtil::int_to_string(qtm.day, 2) +
+            QUtil::int_to_string(qtm.hour, 2) +
+            QUtil::int_to_string(qtm.minute, 2) +
+            QUtil::int_to_string(qtm.second, 2) +
+            tz_offset);
+}
+
+bool
+QUtil::pdf_time_to_qpdf_time(std::string const& str, QPDFTime* qtm)
+{
+    static std::regex pdf_date("^D:([0-9]{4})([0-9]{2})([0-9]{2})"
+                               "([0-9]{2})([0-9]{2})([0-9]{2})"
+                               "(?:(Z)|([\\+\\-])([0-9]{2})'([0-9]{2})')$");
+    std::smatch m;
+    if (! std::regex_match(str, m, pdf_date))
+    {
+        return false;
+    }
+    int tz_delta = 0;
+    auto to_i = [](std::string const& s) {
+        return QUtil::string_to_int(s.c_str());
+    };
+
+    if (m[7] == "")
+    {
+        tz_delta = ((to_i(m[9]) * 60) +
+                    to_i(m[10]));
+        if (m[8] == "+")
+        {
+            tz_delta = -tz_delta;
+        }
+    }
+    if (qtm)
+    {
+        *qtm = QPDFTime(to_i(m[1]),
+                        to_i(m[2]),
+                        to_i(m[3]),
+                        to_i(m[4]),
+                        to_i(m[5]),
+                        to_i(m[6]),
+                        tz_delta);
+    }
+    return true;
+}
+
 std::string
 QUtil::toUTF8(unsigned long uval)
 {
@@ -105,3 +105,7 @@ rename file
 create file
 rename over existing
 delete file
+---- timestamp
+D:20210209144925-05'00'
+D:20210210011925+05'30'
+D:20210209191925Z
@@ -581,6 +581,27 @@ void rename_delete_test()
     assert_no_file("old\xcf\x80.~tmp");
 }
+void timestamp_test()
+{
+    auto check = [](QUtil::QPDFTime const& t) {
+        std::string pdf = QUtil::qpdf_time_to_pdf_time(t);
+        std::cout << pdf << std::endl;
+        QUtil::QPDFTime t2;
+        assert(QUtil::pdf_time_to_qpdf_time(pdf, &t2));
+        assert(QUtil::qpdf_time_to_pdf_time(t2) == pdf);
+    };
+    check(QUtil::QPDFTime(2021, 2, 9, 14, 49, 25, 300));
+    check(QUtil::QPDFTime(2021, 2, 10, 1, 19, 25, -330));
+    check(QUtil::QPDFTime(2021, 2, 9, 19, 19, 25, 0));
+    assert(! QUtil::pdf_time_to_qpdf_time("potato"));
+    // Round trip on the current time without actually printing it.
+    // Manual testing was done to ensure that we are actually getting
+    // back the current time in various timezones.
+    assert(QUtil::pdf_time_to_qpdf_time(
+               QUtil::qpdf_time_to_pdf_time(
+                   QUtil::get_current_qpdf_time())));
+}
+
 int main(int argc, char* argv[])
 {
     try
@@ -611,6 +632,8 @@ int main(int argc, char* argv[])
 	hex_encode_decode_test();
 	std::cout << "---- rename/delete" << std::endl;
 	rename_delete_test();
+	std::cout << "---- timestamp" << std::endl;
+	timestamp_test();
     }
     catch (std::exception& e)
     {
@@ -4943,6 +4943,14 @@ print &quot;\n&quot;;
         </listitem>
         <listitem>
          <para>
+          Add <function>QUtil::get_current_qpdf_time</function>,
+          <function>QUtil::pdf_time_to_qpdf_time</function>, and
+          <function>QUtil::qpdf_time_to_pdf_time</function> for
+          working with PDF timestamp strings.
+         </para>
+        </listitem>
+        <listitem>
+         <para>
           Add <function>warn</function> to
           <classname>QPDF</classname>'s public API.
          </para>