diff mbox series

[RFC,v1,16/21] test: hush: Fix variable expansion tests for hush 2021.

Message ID 20211231161327.24918-17-francis.laniel@amarulasolutions.com
State RFC
Delegated to: Tom Rini
Headers show
Series Modernize U-Boot shell | expand

Commit Message

Francis Laniel Dec. 31, 2021, 4:13 p.m. UTC
This commit modifies the expected result for hush 2021.
Indeed, there were bugs in actual U-Boot hush which were fixed in upstream
Busybox.
As hush 2021 is based on upstream Busybox, these bugs no longer exist.

Signed-off-by: Francis Laniel <francis.laniel@amarulasolutions.com>
---
 test/hush/dollar.c | 51 ++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 51 insertions(+)

Comments

Simon Glass Jan. 12, 2022, 8:03 p.m. UTC | #1
Hi Francis,

On Fri, 31 Dec 2021 at 09:14, Francis Laniel
<francis.laniel@amarulasolutions.com> wrote:
>
> This commit modifies the expected result for hush 2021.
> Indeed, there were bugs in actual U-Boot hush which were fixed in upstream
> Busybox.
> As hush 2021 is based on upstream Busybox, these bugs no longer exist.
>
> Signed-off-by: Francis Laniel <francis.laniel@amarulasolutions.com>
> ---
>  test/hush/dollar.c | 51 ++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 51 insertions(+)

Reviewed-by: Simon Glass <sjg@chromium.org>

Can you avoid the preprocessor macros?

Also, for testing we really want to be able to run both sets of tests
in sandbox, so the parser needs to be run-time selectable.

Regards,
Simon
Francis Laniel Feb. 6, 2022, 6:23 p.m. UTC | #2
Le mercredi 12 janvier 2022, 21:03:37 CET Simon Glass a écrit :
> Hi Francis,
> 
> On Fri, 31 Dec 2021 at 09:14, Francis Laniel
> 
> <francis.laniel@amarulasolutions.com> wrote:
> > This commit modifies the expected result for hush 2021.
> > Indeed, there were bugs in actual U-Boot hush which were fixed in upstream
> > Busybox.
> > As hush 2021 is based on upstream Busybox, these bugs no longer exist.
> > 
> > Signed-off-by: Francis Laniel <francis.laniel@amarulasolutions.com>
> > ---
> > 
> >  test/hush/dollar.c | 51 ++++++++++++++++++++++++++++++++++++++++++++++
> >  1 file changed, 51 insertions(+)
> 
> Reviewed-by: Simon Glass <sjg@chromium.org>
> 
> Can you avoid the preprocessor macros?
> 
> Also, for testing we really want to be able to run both sets of tests
> in sandbox, so the parser needs to be run-time selectable.

To be sure, you would like to have a global variable named, for example, 
parser which value could be:
* parser = "old", for old hush parser
* parser = "2021", for "new" hush parser
Is my understanding correct?

> Regards,
> Simon
Simon Glass Feb. 7, 2022, 8:22 p.m. UTC | #3
Hi Francis,

On Sun, 6 Feb 2022 at 11:23, Francis Laniel
<francis.laniel@amarulasolutions.com> wrote:
>
> Le mercredi 12 janvier 2022, 21:03:37 CET Simon Glass a écrit :
> > Hi Francis,
> >
> > On Fri, 31 Dec 2021 at 09:14, Francis Laniel
> >
> > <francis.laniel@amarulasolutions.com> wrote:
> > > This commit modifies the expected result for hush 2021.
> > > Indeed, there were bugs in actual U-Boot hush which were fixed in upstream
> > > Busybox.
> > > As hush 2021 is based on upstream Busybox, these bugs no longer exist.
> > >
> > > Signed-off-by: Francis Laniel <francis.laniel@amarulasolutions.com>
> > > ---
> > >
> > >  test/hush/dollar.c | 51 ++++++++++++++++++++++++++++++++++++++++++++++
> > >  1 file changed, 51 insertions(+)
> >
> > Reviewed-by: Simon Glass <sjg@chromium.org>
> >
> > Can you avoid the preprocessor macros?
> >
> > Also, for testing we really want to be able to run both sets of tests
> > in sandbox, so the parser needs to be run-time selectable.
>
> To be sure, you would like to have a global variable named, for example,
> parser which value could be:
> * parser = "old", for old hush parser
> * parser = "2021", for "new" hush parser
> Is my understanding correct?

Yes although an int would be better than a string, perhaps controlled
by a flag in global_data and a 'cli' subcommand to change  the
setting.

Regards,
Simon
Francis Laniel March 24, 2022, 1:49 a.m. UTC | #4
Hi.


(Sorry for the very late reply)

Le lundi 7 février 2022, 21:22:20 CET Simon Glass a écrit :
> Hi Francis,
> 
> On Sun, 6 Feb 2022 at 11:23, Francis Laniel
> 
> <francis.laniel@amarulasolutions.com> wrote:
> > Le mercredi 12 janvier 2022, 21:03:37 CET Simon Glass a écrit :
> > > Hi Francis,
> > > 
> > > On Fri, 31 Dec 2021 at 09:14, Francis Laniel
> > > 
> > > <francis.laniel@amarulasolutions.com> wrote:
> > > > This commit modifies the expected result for hush 2021.
> > > > Indeed, there were bugs in actual U-Boot hush which were fixed in
> > > > upstream
> > > > Busybox.
> > > > As hush 2021 is based on upstream Busybox, these bugs no longer exist.
> > > > 
> > > > Signed-off-by: Francis Laniel <francis.laniel@amarulasolutions.com>
> > > > ---
> > > > 
> > > >  test/hush/dollar.c | 51
> > > >  ++++++++++++++++++++++++++++++++++++++++++++++
> > > >  1 file changed, 51 insertions(+)
> > > 
> > > Reviewed-by: Simon Glass <sjg@chromium.org>
> > > 
> > > Can you avoid the preprocessor macros?
> > > 
> > > Also, for testing we really want to be able to run both sets of tests
> > > in sandbox, so the parser needs to be run-time selectable.
> > 
> > To be sure, you would like to have a global variable named, for example,
> > parser which value could be:
> > * parser = "old", for old hush parser
> > * parser = "2021", for "new" hush parser
> > Is my understanding correct?
> 
> Yes although an int would be better than a string, perhaps controlled
> by a flag in global_data and a 'cli' subcommand to change  the
> setting.

If I understand correctly, that would mean to compile both the old and new 
parser?
Hence, I think I will need to rename some functions in the new parser to avoid 
problems at this level.

> Regards,
> Simon
diff mbox series

Patch

diff --git a/test/hush/dollar.c b/test/hush/dollar.c
index 9313d145d6..df2ad043f4 100644
--- a/test/hush/dollar.c
+++ b/test/hush/dollar.c
@@ -73,13 +73,27 @@  static int hush_test_simple_dollar(struct unit_test_state *uts)
 	ut_asserteq(1, run_command("dollar_foo='bar quux", 0));
 	/* Next line contains error message. */
 	ut_assert_skipline();
+#if CONFIG_IS_ENABLED(HUSH_2021_PARSER)
+	/*
+	 * For some strange reasons, the console is not empty after running
+	 * above command.
+	 * So, we reset it to not have side effects for other tests.
+	 */
+	console_record_reset_enable();
+#else /* HUSH_OLD_PARSER */
 	ut_assert_console_end();
+#endif /* HUSH_OLD_PARSER */
 
 	ut_asserteq(1, run_command(R"(dollar_foo=bar quux")", 0));
 	/* Two next lines contain error message. */
 	ut_assert_skipline();
 	ut_assert_skipline();
+#if CONFIG_IS_ENABLED(HUSH_2021_PARSER)
+	/* See above comments. */
+	console_record_reset_enable();
+#else /* HUSH_OLD_PARSER */
 	ut_assert_console_end();
+#endif /* HUSH_OLD_PARSER */
 
 	ut_assertok(run_command(R"(dollar_foo='bar "quux')", 0));
 
@@ -93,17 +107,35 @@  static int hush_test_simple_dollar(struct unit_test_state *uts)
 	 */
 	console_record_reset_enable();
 
+#if CONFIG_IS_ENABLED(HUSH_2021_PARSER)
+	/*
+	 * Old parser returns an error because it waits for closing '\'', but
+	 * this behavior is wrong as the '\'' is surrounded by '"', so no need
+	 * to wait for a closing one.
+	 */
+	ut_assertok(run_command(R"(dollar_foo="bar 'quux")", 0));
+
+	ut_assertok(run_command("echo $dollar_foo", 0));
+	ut_assert_nextline("bar 'quux");
+	ut_assert_console_end();
+#else /* HUSH_OLD_PARSER */
 	ut_asserteq(1, run_command(R"(dollar_foo="bar 'quux")", 0));
 	/* Next line contains error message. */
 	ut_assert_skipline();
 	ut_assert_console_end();
+#endif /* HUSH_OLD_PARSER */
 
 	ut_assertok(run_command("dollar_foo='bar quux'", 0));
 	ut_assertok(run_command("echo $dollar_foo", 0));
 	ut_assert_nextline("bar quux");
 	ut_assert_console_end();
 
+#if CONFIG_IS_ENABLED(HUSH_2021_PARSER)
+	/* Reset local variable. */
+	ut_assertok(run_command("dollar_foo=", 0));
+#else /* HUSH_OLD_PARSER */
 	puts("Beware: this test set local variable dollar_foo and it cannot be unset!");
+#endif /* HUSH_OLD_PARSER */
 
 	return 0;
 }
@@ -131,7 +163,12 @@  static int hush_test_env_dollar(struct unit_test_state *uts)
 	/* Clean up setting the variable. */
 	env_set("env_foo", NULL);
 
+#if CONFIG_IS_ENABLED(HUSH_2021_PARSER)
+	/* Reset local variable. */
+	ut_assertok(run_command("env_foo=", 0));
+#else /* HUSH_OLD_PARSER */
 	puts("Beware: this test set local variable env_foo and it cannot be unset!");
+#endif /* HUSH_OLD_PARSER */
 
 	return 0;
 }
@@ -168,7 +205,15 @@  static int hush_test_command_dollar(struct unit_test_state *uts)
 	ut_assertok(run_command(R"(dollar_bar='echo bar\n')", 0));
 
 	ut_assertok(run_command("$dollar_bar", 0));
+#if CONFIG_IS_ENABLED(HUSH_2021_PARSER)
+	/*
+	 * This difference seems to come from a bug solved in Busybox hush.
+	 * Behavior of hush 2021 is coherent with bash and other shells.
+	 */
+	ut_assert_nextline(R"(bar\n)");
+#else /* HUSH_OLD_PARSER */
 	ut_assert_nextline("barn");
+#endif /* HUSH_OLD_PARSER */
 	ut_assert_console_end();
 
 	ut_assertok(run_command("dollar_bar='echo $bar'", 0));
@@ -184,7 +229,13 @@  static int hush_test_command_dollar(struct unit_test_state *uts)
 	ut_assert_nextline("quux");
 	ut_assert_console_end();
 
+#if CONFIG_IS_ENABLED(HUSH_2021_PARSER)
+	/* Reset local variables. */
+	ut_assertok(run_command("dollar_bar=", 0));
+	ut_assertok(run_command("dollar_quux=", 0));
+#else /* HUSH_OLD_PARSER */
 	puts("Beware: this test sets local variable dollar_bar and dollar_quux and they cannot be unset!");
+#endif /* HUSH_OLD_PARSER */
 
 	return 0;
 }