From 0b1623d07db963ecf3789aba7163321812cba88e Mon Sep 17 00:00:00 2001
From: Jay Berkenbilt <ejb@ql.org>
Date: Thu, 18 Feb 2021 07:55:39 -0500
Subject: [PATCH] Add QUtil::path_basename

---
 ChangeLog                      |  2 ++
 include/qpdf/QUtil.hh          | 16 ++++++++++++----
 libqpdf/QUtil.cc               | 31 +++++++++++++++++++++++++++++++
 libtests/qtest/qutil/qutil.out |  7 +++++++
 libtests/qutil.cc              | 25 +++++++++++++++++++++++++
 manual/qpdf-manual.xml         |  6 ++++++
 6 files changed, 83 insertions(+), 4 deletions(-)

diff --git a/ChangeLog b/ChangeLog
index 2a3f1780..4a758b54 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,7 @@
 2021-02-18  Jay Berkenbilt  <ejb@ql.org>
 
+	* Add QUtil::path_basename to get last element of a path.
+
 	* Add examples/pdf-attach-file.cc to illustrate new file
 	attachment method and also new parse that takes indirect objects.
 
diff --git a/include/qpdf/QUtil.hh b/include/qpdf/QUtil.hh
index 769c5fd7..dd42fe1c 100644
--- a/include/qpdf/QUtil.hh
+++ b/include/qpdf/QUtil.hh
@@ -130,14 +130,22 @@ namespace QUtil
     // Write the contents of filename as a binary file to the
     // pipeline.
     QPDF_DLL
-    void
-    pipe_file(char const* filename, Pipeline* p);
+    void pipe_file(char const* filename, Pipeline* p);
 
     // Return a function that will send the contents of the given file
     // through the given pipeline as binary data.
     QPDF_DLL
-    std::function<void(Pipeline*)>
-    file_provider(std::string const& filename);
+    std::function<void(Pipeline*)> file_provider(std::string const& filename);
+
+    // Return the last path element. On Windows, either / or \ are
+    // path separators. Otherwise, only / is a path separator. Strip
+    // any trailing path separators. Then, if any path separators
+    // remain, return everything after the last path separator.
+    // Otherwise, return the whole string. As a special case, if a
+    // string consists entirely of path separators, the first
+    // character is returned.
+    QPDF_DLL
+    std::string path_basename(std::string const& filename);
 
     QPDF_DLL
     char* copy_string(std::string const&);
diff --git a/libqpdf/QUtil.cc b/libqpdf/QUtil.cc
index a3fa94f4..17b163f5 100644
--- a/libqpdf/QUtil.cc
+++ b/libqpdf/QUtil.cc
@@ -663,6 +663,37 @@ QUtil::file_provider(std::string const& filename)
     };
 }
 
+std::string
+QUtil::path_basename(std::string const& filename)
+{
+#ifdef _WIN32
+    char const* pathsep = "/\\";
+#else
+    char const* pathsep = "/";
+#endif
+    std::string last = filename;
+    auto len = last.length();
+    while (len > 1)
+    {
+        auto pos = last.find_last_of(pathsep);
+        if (pos == len - 1)
+        {
+            last.pop_back();
+            --len;
+        }
+        else if (pos == std::string::npos)
+        {
+            break;
+        }
+        else
+        {
+            last = last.substr(pos + 1);
+            break;
+        }
+    }
+    return last;
+}
+
 char*
 QUtil::copy_string(std::string const& str)
 {
diff --git a/libtests/qtest/qutil/qutil.out b/libtests/qtest/qutil/qutil.out
index 7dded2e3..90f1fd16 100644
--- a/libtests/qtest/qutil/qutil.out
+++ b/libtests/qtest/qutil/qutil.out
@@ -97,6 +97,13 @@ file1: -qutil.out-, file2: -other-file-; same: 0: PASS
 file1: -qutil.out-, file2: --; same: 0: PASS
 file1: -qutil.out-, file2: -(null)-; same: 0: PASS
 file1: --, file2: -qutil.out-; same: 0: PASS
+---- path
+//// -> /
+a/b/cdef -> cdef
+a/b/cdef/ -> cdef
+/ -> /
+ -> 
+quack -> quack
 ---- read from file
 This file is used for qutil testing.
 It has mixed newlines.
diff --git a/libtests/qutil.cc b/libtests/qutil.cc
index abe05f35..7b1a8788 100644
--- a/libtests/qutil.cc
+++ b/libtests/qutil.cc
@@ -439,6 +439,29 @@ void same_file_test()
     assert_same_file("", "qutil.out", false);
 }
 
+void path_test()
+{
+    auto check = [](bool print, std::string const& a, std::string const& b) {
+        auto result = QUtil::path_basename(a);
+        if (print)
+        {
+            std::cout << a << " -> " << result << std::endl;
+        }
+        assert(result == b);
+    };
+
+#ifdef _WIN32
+    check(false, "asdf\\qwer", "qwer");
+    check(false, "asdf\\qwer/\\", "qwer");
+#endif
+    check(true, "////", "/");
+    check(true, "a/b/cdef", "cdef");
+    check(true, "a/b/cdef/", "cdef");
+    check(true, "/", "/");
+    check(true, "", "");
+    check(true, "quack", "quack");
+}
+
 void read_from_file_test()
 {
     std::list<std::string> lines = QUtil::read_lines_from_file("other-file");
@@ -636,6 +659,8 @@ int main(int argc, char* argv[])
 	get_whoami_test();
 	std::cout << "---- file" << std::endl;
 	same_file_test();
+	std::cout << "---- path" << std::endl;
+	path_test();
 	std::cout << "---- read from file" << std::endl;
 	read_from_file_test();
 	std::cout << "---- hex encode/decode" << std::endl;
diff --git a/manual/qpdf-manual.xml b/manual/qpdf-manual.xml
index 4a294cb3..2c1c235a 100644
--- a/manual/qpdf-manual.xml
+++ b/manual/qpdf-manual.xml
@@ -5283,6 +5283,12 @@ print "\n";
           matrices.
          </para>
         </listitem>
+        <listitem>
+         <para>
+          Add <function>QUtil::path_basename</function> to return the
+          last element of a path.
+         </para>
+        </listitem>
        </itemizedlist>
       </listitem>
       <listitem>