sbang: add support for php (#18299)

PHP supports an initial shebang, but its comment syntax can't handle our 2-line
shebangs. So, we need to embed the 2nd-line shebang comment to look like a
PHP comment:

    <?php #!/path/to/php ?>

This adds patching support to the sbang hook and support for
instrumenting php shebangs.

This also patches `phar`, which is a tool used to create php packages.
`phar` itself has to add sbangs to those packages (as phar archives
apparently contain UTF-8, as well as binary blobs), and `phar` sets a
checksum based on the contents of the package.

Co-authored-by: Todd Gamblin <tgamblin@llnl.gov>

5 files changed, 105 insertions, 7 deletions

diff --git a/bin/sbang b/bin/sbang
index 5c9ad37acd..aba8658a7a 100755
--- a/bin/sbang
+++ b/bin/sbang
@@ -61,15 +61,23 @@
 # Obviously, for this to work, `sbang` needs to have a short enough
 # path that *it* will run without hitting OS limits.
 #
-# For Lua, scripts the second line can't start with #!, as # is not
-# the comment character in lua (even though lua ignores #! on the
-# *first* line of a script).  So, instrument a lua script like this,
-# using -- instead of # on the second line:
+# For Lua, node, and php scripts, the second line can't start with #!, as
+# # is not the comment character in these languages (though they all
+# ignore #! on the *first* line of a script). So, instrument such scripts
+# like this, using --, //, or <?php ... ?> instead of # on the second
+# line, e.g.:
 #
 #     1    #!/bin/bash /path/to/sbang
 #     2    --!/long/path/to/lua with arguments
-#     3
-#     4    print "success!"
+#     3    print "success!"
+#
+#     1    #!/bin/bash /path/to/sbang
+#     2    //!/long/path/to/node with arguments
+#     3    print "success!"
+#
+#     1    #!/bin/bash /path/to/sbang
+#     2    <?php #/long/path/to/php with arguments ?>
+#     3    <?php echo "success!\n"; ?>
 #
 # How it works
 # -----------------------------
@@ -90,6 +98,9 @@ while read line && ((lines < 2)) ; do
         interpreter="${line#//!}"
     elif [[ "$line" = '--!'*lua* ]]; then
         interpreter="${line#--!}"
+    elif [[ "$line" = '<?php #!'*php* ]]; then
+        interpreter="${line#<?php\ \#!}"
+        interpreter="${interpreter%\ ?>}"
     fi
     lines=$((lines+1))
 done < "$script"
@@ -98,7 +109,7 @@ done < "$script"
 # #!/<spack-long-path>/perl -w
 # this is the interpreter line with all the parameters as a vector
 interpreter_v=(${interpreter})
-# this is the single interpreter path 
+# this is the single interpreter path
 interpreter_f="${interpreter_v[0]}"
 
 # Invoke any interpreter found, or raise an error if none was found.
diff --git a/lib/spack/spack/hooks/sbang.py b/lib/spack/spack/hooks/sbang.py
index 324d5d4fbb..24d84db8af 100644
--- a/lib/spack/spack/hooks/sbang.py
+++ b/lib/spack/spack/hooks/sbang.py
@@ -64,6 +64,10 @@ def filter_shebang(path):
     if re.search(r'^#!(/[^/\n]*)*lua\b', original):
         original = re.sub(r'^#', '--', original)
 
+    # Use <?php #! instead of #! on second line for php.
+    if re.search(r'^#!(/[^/\n]*)*php\b', original):
+        original = re.sub(r'^#', '<?php #', original) + ' ?>'
+
     # Use //! instead of #! on second line for node.js.
     if re.search(r'^#!(/[^/\n]*)*node\b', original):
         original = re.sub(r'^#', '//', original)
diff --git a/lib/spack/spack/test/sbang.py b/lib/spack/spack/test/sbang.py
index eeb68f20ab..c2d0adf4d2 100644
--- a/lib/spack/spack/test/sbang.py
+++ b/lib/spack/spack/test/sbang.py
@@ -30,6 +30,11 @@ node_line         = "#!/this/" + ('x' * 200) + "/is/node\n"
 node_in_text      = ("line\n") * 100 + "lua\n" + ("line\n" * 100)
 node_line_patched = "//!/this/" + ('x' * 200) + "/is/node\n"
 sbang_line        = '#!/bin/bash %s/bin/sbang\n' % spack.store.layout.root
+php_line         = "#!/this/" + ('x' * 200) + "/is/php\n"
+php_in_text      = ("line\n") * 100 + "php\n" + ("line\n" * 100)
+php_line_patched = "<?php #!/this/" + ('x' * 200) + "/is/php\n"
+php_line_patched2 = "?>\n"
+sbang_line        = '#!/bin/bash %s/bin/sbang\n' % spack.store.layout.root
 last_line         = "last!\n"
 
 
@@ -79,6 +84,19 @@ class ScriptDirectory(object):
             f.write(node_in_text)
             f.write(last_line)
 
+        # php script with long shebang
+        self.php_shebang = os.path.join(self.tempdir, 'php')
+        with open(self.php_shebang, 'w') as f:
+            f.write(php_line)
+            f.write(last_line)
+
+        # php script with long shebang
+        self.php_textbang = os.path.join(self.tempdir, 'php_in_text')
+        with open(self.php_textbang, 'w') as f:
+            f.write(short_line)
+            f.write(php_in_text)
+            f.write(last_line)
+
         # Script already using sbang.
         self.has_sbang = os.path.join(self.tempdir, 'shebang')
         with open(self.has_sbang, 'w') as f:
diff --git a/var/spack/repos/builtin/packages/php/package.py b/var/spack/repos/builtin/packages/php/package.py
index 714bfb0e8f..d5da45eb94 100644
--- a/var/spack/repos/builtin/packages/php/package.py
+++ b/var/spack/repos/builtin/packages/php/package.py
@@ -5,6 +5,8 @@
 
 from spack import *
 
+import spack.hooks.sbang as sbang
+
 
 class Php(AutotoolsPackage):
     """
@@ -36,6 +38,27 @@ class Php(AutotoolsPackage):
     depends_on('libxml2')
     depends_on('sqlite')
 
+    patch('sbang.patch')
+
+    def patch(self):
+        """
+        phar sbang is added before build phase.
+        Because phar is php script with binary data
+        (Not UTF-8 text file) And phar is embeded own sha1 checksum.
+        """
+        shebang_limit = 127
+
+        if len(self.prefix.bin.php) + 2 <= shebang_limit:
+            return
+
+        new_sbang_line = '#!/bin/bash %s' % sbang.sbang_install_path()
+        original_bang = '-b "$(PHP_PHARCMD_BANG)"'
+        makefile = join_path('ext', 'phar', 'Makefile.frag')
+        filter_file(
+            original_bang,
+            original_bang + ' -z "{0}"'.format(new_sbang_line),
+            makefile, string=True)
+
     def autoreconf(self, spec, prefix):
         bash = which('bash')
         bash('./buildconf', '--force')
diff --git a/var/spack/repos/builtin/packages/php/sbang.patch b/var/spack/repos/builtin/packages/php/sbang.patch
new file mode 100755
index 0000000000..aca17c8ff3
--- /dev/null
+++ b/var/spack/repos/builtin/packages/php/sbang.patch
@@ -0,0 +1,42 @@
+--- spack-src/ext/phar/phar/pharcommand.inc.org	2019-12-18 01:35:53.000000000 +0900
++++ spack-src/ext/phar/phar/pharcommand.inc	2020-08-20 12:26:16.207347572 +0900
+@@ -68,6 +68,12 @@
+ 				'inf' => '<bang>   Hash-bang line to start the archive (e.g. #!/usr/bin/php). The hash '
+ 						 .'         mark itself \'#!\' and the newline character are optional.'
+ 			),
++			'z' => array(
++				'typ' => 'any',
++				'val' => NULL,
++				'inf' => '<bang>   Hash-bang line to start the archive for spack. The hash '
++						 .'         mark itself \'#!\' and the newline character are optional.'
++			),
+ 			'c' => array(
+ 				'typ' => 'compalg',
+ 				'val' => NULL,
+@@ -455,7 +461,7 @@
+ 	 */
+ 	static function cli_cmd_arg_pack()
+ 	{
+-		$args = self::phar_args('abcFhilpsxy', 'pharnew');
++		$args = self::phar_args('azbcFhilpsxy', 'pharnew');
+
+ 		$args[''] = array(
+ 			'typ'     => 'any',
+@@ -560,6 +566,7 @@
+ 		}
+
+ 		$alias    = $this->args['a']['val'];
++		$spack_hb = $this->args['z']['val'];
+ 		$hashbang = $this->args['b']['val'];
+ 		$archive  = $this->args['f']['val'];
+ 		$hash     = $this->args['h']['val'];
+@@ -571,6 +578,9 @@
+ 		$invregex = $this->args['x']['val'];
+ 		$input    = $this->args['']['val'];
+
++		if (isset($spack_hb)) {
++			$hashbang = "$spack_hb\n<?php #!$hashbang ?>";
++		}
+ 		$hash = self::phar_check_hash($hash, $privkey);
+
+ 		$phar  = new Phar($archive, 0, $alias);