From 2485ec45d8a8a762cea1623fd46821e1a6d55630 Mon Sep 17 00:00:00 2001
From: Ben Pfaff <blp@cs.stanford.edu>
Date: Sun, 2 Aug 2020 21:26:31 +0000
Subject: [PATCH] tests: Avoid dealing with character encoding issues on
 command line.

Unix-like systems don't have a command-line encoding, so we can pass
whatever bytes we want.  mingw uses some particular encoding for
command line arguments, so this screws up trying to pass UTF-8 in.  This
commit side-steps the problem by passing only ASCII characters on the
command line and making the test program decode the bytes.

This fixes several test failures under mingw.
---
 tests/automake.mk         |  2 +-
 tests/libpspp/i18n-test.c | 35 ++++++++++++++++++++++++++++++++---
 tests/libpspp/i18n.at     |  9 +++------
 3 files changed, 36 insertions(+), 10 deletions(-)

diff --git a/tests/automake.mk b/tests/automake.mk
index f3e14e7e98..8897d9fbc2 100644
--- a/tests/automake.mk
+++ b/tests/automake.mk
@@ -100,7 +100,7 @@ tests_libpspp_hmapx_test_SOURCES = \
 tests_libpspp_hmapx_test_CPPFLAGS = $(AM_CPPFLAGS) -DASSERT_LEVEL=10
 
 tests_libpspp_i18n_test_SOURCES = tests/libpspp/i18n-test.c
-tests_libpspp_i18n_test_LDADD = src/libpspp-core.la
+tests_libpspp_i18n_test_LDADD = src/libpspp-core.la gl/libgl.la
 
 tests_libpspp_abt_test_SOURCES = \
 	src/libpspp/abt.c \
diff --git a/tests/libpspp/i18n-test.c b/tests/libpspp/i18n-test.c
index b0d027090e..a4993a8d02 100644
--- a/tests/libpspp/i18n-test.c
+++ b/tests/libpspp/i18n-test.c
@@ -22,9 +22,35 @@
 
 #include "libpspp/i18n.h"
 
+#include "gl/xalloc.h"
+
 #undef NDEBUG
 #include <assert.h>
 
+/* Returns a malloc()'d copy of INPUT with \ooo octal escapes replaced by their
+   values. */
+static char *
+backslash_decode (const char *input)
+{
+  char *output = xmalloc (strlen (input) + 1);
+  char *p = output;
+  for (; *input; input++)
+    {
+      if (*input == '\\' && input[1] >= '0' && input[1] <= '7')
+        {
+          uint8_t c = 0;
+          while (input[1] >= '0' && input[1] <= '7')
+            c = c * 8 + (*++input - '0');
+          *p++ = c;
+        }
+      else
+        *p++ = *input;
+    }
+  *p = '\0';
+
+  return output;
+}
+
 int
 main (int argc, char *argv[])
 {
@@ -48,16 +74,17 @@ main (int argc, char *argv[])
     {
       const char *from = argv[2];
       const char *to = argv[3];
-      const char *string = argv[4];
+      char *string = backslash_decode (argv[4]);
       char *result = recode_string (to, from, string, -1);
       puts (result);
       assert (strlen (result) == recode_string_len (to, from, string, -1));
+      free (string);
       free (result);
     }
   else if (argc == 6 && !strcmp (argv[1], "concat"))
     {
-      const char *head = argv[2];
-      const char *tail = argv[3];
+      char *head = backslash_decode (argv[2]);
+      char *tail = backslash_decode (argv[3]);
       const char *encoding = argv[4];
       int max_len = atoi (argv[5]);
       char *result;
@@ -78,6 +105,8 @@ main (int argc, char *argv[])
         }
 
       free (result);
+      free (head);
+      free (tail);
     }
   else
     {
diff --git a/tests/libpspp/i18n.at b/tests/libpspp/i18n.at
index 77216aac20..d0e87bb3b9 100644
--- a/tests/libpspp/i18n.at
+++ b/tests/libpspp/i18n.at
@@ -20,16 +20,14 @@ AT_BANNER([i18n recoding])
 #                   [FROM-TEXT], [TO-TEXT])
 #
 # Converts FROM-TEXT from FROM-CODING to TO-CODING and checks that the result
-# is TO-TEXT.  The "printf" program is applied to both FROM-TEXT and TO-TEXT to
-# allow for backslash-escapes.  (Hex escapes are not portable; use octal
-# escapes instead.)
+# is TO-TEXT.  Octal backslash-escapes are supported in FROM-TEXT and TO-TEXT.
 m4_define([CHECK_I18N_RECODE],
   [AT_SETUP([convert $1])
    AT_KEYWORDS([i18n])
 
    dnl Skip the test if this host doesn't know the source and target encodings.
    AT_CHECK([i18n-test supports_encodings '$2' '$3'])
-   AT_CHECK_UNQUOTED([i18n-test recode '$2' '$3' `printf '$4'`], [0], [`printf '$5'`
+   AT_CHECK_UNQUOTED([i18n-test recode '$2' '$3' '$4'], [0], [`printf '$5'`
 ])
    AT_CLEANUP])
 
@@ -85,8 +83,7 @@ m4_define([CHECK_I18N_CONCAT],
    dnl Skip the test if this host doesn't know the encoding.
    AT_CHECK([i18n-test supports_encodings '$3'])
    AT_CHECK_UNQUOTED(
-     [i18n-test concat "`printf '$1'`" "`printf '$2'`" '$3' '$4'], [0],
-     [`printf '$5'`
+     [i18n-test concat '$1' '$2' '$3' '$4'], [0], [`printf '$5'`
 ])
    AT_CLEANUP])
 
-- 
2.30.2