dependencies: Fix parsing of shebangs with spaces

While finding an external program, we should only split the shebang
once since that is what Linux and BSD also do. This is also why
everyone uses #!/usr/bin/env in their shebangs since that allows
you to run an interpreter in a path with spaces in it.

See `man execve` for more details, specifically the sections for
interpreter scripts.

5 files changed, 108 insertions, 1 deletions

diff --git a/mesonbuild/dependencies/base.py b/mesonbuild/dependencies/base.py
index 682182c69..a72023275 100644
--- a/mesonbuild/dependencies/base.py
+++ b/mesonbuild/dependencies/base.py
@@ -560,7 +560,11 @@ class ExternalProgram:
             with open(script) as f:
                 first_line = f.readline().strip()
             if first_line.startswith('#!'):
-                commands = first_line[2:].split('#')[0].strip().split()
+                # In a shebang, everything before the first space is assumed to
+                # be the command to run and everything after the first space is
+                # the single argument to pass to that command. So we must split
+                # exactly once.
+                commands = first_line[2:].split('#')[0].strip().split(maxsplit=1)
                 if mesonlib.is_windows():
                     # Windows does not have UNIX paths so remove them,
                     # but don't remove Windows paths
diff --git a/test cases/common/167 external program shebang parsing/input.txt b/test cases/common/167 external program shebang parsing/input.txt
new file mode 100644
index 000000000..40e30d4b2
--- /dev/null
+++ b/test cases/common/167 external program shebang parsing/input.txt
@@ -0,0 +1 @@
+some stuff here
diff --git a/test cases/common/167 external program shebang parsing/main.c b/test cases/common/167 external program shebang parsing/main.c
new file mode 100644
index 000000000..2ebe169ac
--- /dev/null
+++ b/test cases/common/167 external program shebang parsing/main.c
@@ -0,0 +1,79 @@
+#include <stdio.h>
+#include <fcntl.h>
+#include <errno.h>
+#include <string.h>
+#include <stdlib.h>
+#include <sys/types.h>
+
+#ifdef _WIN32
+ #include <io.h>
+ #include <windows.h>
+#else
+ #include <unistd.h>
+#endif
+
+static int
+intrp_copyfile (char * src, char * dest)
+{
+#ifdef _WIN32
+  if (!CopyFile (src, dest, FALSE))
+    return 1;
+  return 0;
+#else
+  return execlp ("cp", "copyfile", src, dest, NULL);
+#endif
+}
+
+static char*
+parser_get_line (FILE * f)
+{
+  ssize_t size;
+  size_t n = 0;
+  char *line = NULL;
+
+  size = getline (&line, &n, f);
+  if (size < 0) {
+    fprintf (stderr, "%s\n", strerror (errno));
+    free (line);
+    return NULL;
+  }
+  return line;
+}
+
+int
+main (int argc, char * argv[])
+{
+  FILE *f;
+  char *line = NULL;
+
+  if (argc != 4) {
+    fprintf (stderr, "Invalid number of arguments: %i\n", argc);
+    goto err;
+  }
+
+  if ((f = fopen (argv[1], "r")) == NULL) {
+    fprintf (stderr, "%s\n", strerror (errno));
+    goto err;
+  }
+
+  line = parser_get_line (f);
+
+  if (!line || line[0] != '#' || line[1] != '!') {
+    fprintf (stderr, "Invalid script\n");
+    goto err;
+  }
+
+  free (line);
+  line = parser_get_line (f);
+
+  if (!line || strncmp (line, "copy", 4) != 0) {
+    fprintf (stderr, "Syntax error\n");
+    goto err;
+  }
+
+  return intrp_copyfile (argv[2], argv[3]);
+
+err:
+  free (line);
+  return 1;
+}
diff --git a/test cases/common/167 external program shebang parsing/meson.build b/test cases/common/167 external program shebang parsing/meson.build
new file mode 100644
index 000000000..c1cc5af9d
--- /dev/null
+++ b/test cases/common/167 external program shebang parsing/meson.build
@@ -0,0 +1,21 @@
+project('shebang parsing', 'c')
+
+interpreter = executable('aninterp', 'main.c', native : true)
+
+cdata = configuration_data()
+cdata.set('INTRP', interpreter.full_path())
+
+f = configure_file(input : 'script.int.in',
+                   output : 'script.int',
+                   configuration : cdata)
+
+# Test that parsing a shebang with spaces works properly. See `man execve`,
+# specifically the section on "Interpreter scripts" and the one under "NOTES".
+script = find_program(f)
+
+custom_target('interpthis',
+  input : 'input.txt',
+  output : 'output.txt',
+  depends : interpreter,
+  command : [script, '@INPUT@', '@OUTPUT@'],
+  build_by_default : true)
diff --git a/test cases/common/167 external program shebang parsing/script.int.in b/test cases/common/167 external program shebang parsing/script.int.in
new file mode 100644
index 000000000..77ff90907
--- /dev/null
+++ b/test cases/common/167 external program shebang parsing/script.int.in
@@ -0,0 +1,2 @@
+#!/usr/bin/env @INTRP@
+copy