Merge pull request #27 from cxw42/vnext

v0.500.4

Author: cxw42

Changes

@@ -1,6 +1,12 @@
 Revision history for Text-PerlPP
 (Note: GH = GitHub issue; # = RT issue)
 
+0.500.4 2018/05/29
+        Updated tests
+
+0.500.3 2018/05/28
+        Fixes to tests; updated documentation
+
 0.500.2 2018/05/27
         Fixes to installation package
 

MANIFEST

@@ -18,6 +18,7 @@ t/04-include.t
 t/05-external-command.t
 t/06-macro.t
 t/07-invalid.t
+t/08-persistent-state.t
 t/a.txt
 t/b.txt
 t/c.txt

README

@@ -16,7 +16,7 @@ README.md.
 
 REQUIRES
 
-Perl 5.10+ and Getopt::Long 2.50.  There is a fatpacked version that
+Perl 5.10.1+ and Getopt::Long 2.50.  There is a fatpacked version that
 already includes Getopt::Long in the GitHub releases archive:
 https://github.com/interpreters/perlpp/releases
 
@@ -45,13 +45,46 @@ Yet another alternative way of installing
     - Copy lib/Text/PerlPP.pm to a directory in your @INC
     - Copy bin/perlpp to a directory in your PATH.
 
+DEVELOPING
+
+We welcome contributions through the normal GitHub pull-request (PR) workflow.
+The build system is straight ExtUtils::MakeMaker.
+
+Before developing, run
+
+    perl Makefile.PL
+	cpanm --installdeps .            # if you have cpanminus installed
+
+Then, to test your code directly from the lib/ directory, run
+
+    make testhere
+
+Before submitting a PR, please test with the normal
+
+    make ; make test
+
+sequence, as well as with
+
+    make testhere
+    make testpacked
+
+That last one tests the packed version made by pack.PL in blib/perlpp.
+We test on p5p Perl 5.10.1 (cygwin x86) and 5.26.1 (cygwin x64),
+and cperl 5.26.2c (cygwin x64).
+
 SUPPORT AND DOCUMENTATION
 
 After installing, you can find documentation for this module with the
 perldoc command.
 
+	perldoc perlpp
+
+describes syntax and invocation, while
+
     perldoc Text::PerlPP
 
+describes using perlpp within another program.
+
 You can also look for information at:
 
     GitHub (report bugs here)

README.md

@@ -4,7 +4,7 @@ PerlPP: Perl preprocessor
 Translates **Text+Perl** to **Text**.
 It can be used for any kind of text templating, e.g. code generation.
 No external modules are required, just a single file.
-Requires Perl 5.10+.
+Requires Perl 5.10.1+.
 
 PerlPP runs in two passes: it generates a Perl script from your input, and then
 it runs the generated script.  If you see `error at (eval ##)`
@@ -58,11 +58,11 @@ Perl code is included between `<?` and `?>` tags.
 There are several modes, indicated by the character after the `<?`:
 
 	<?	code mode: Perl code is between the tags.
-	<?=	echo mode: prints a Perl expression
-	<?:	internal-command mode: executed by PerlPP itself (see below)
+	<?=	echo mode: prints a Perl expression.
+	<?:	internal-command mode: executed by PerlPP itself.
 	<?/	code mode, beginning with printing a line break.
 	<?#	comment mode: everything in <?# ... ?> is ignored.
-	<?!	external mode: everything in <?! ... ?> is run as an external command
+	<?!	external mode: everything in <?! ... ?> is run as an external command.
 
 The code mode is started by `<?` followed by any number of whitespaces
 or line breaks.
@@ -82,11 +82,11 @@ The Generated Script
 
 The generated script:
 
-- is in its own package, named based on the input filename
-- `use`s `5.010`, `strict`, and `warnings`
+- is in its own package, named based on the input filename and a unique number
+- `use`s `5.010001`, `strict`, and `warnings`
 - provides constants `true` (=`!!1`) and `false` (=`!!0`) (with `use constant`)
-- Declares `my %D` and initializes `%D` based on any **-D** options you provide
-- Declares `my %S` and initializes `%S` based on any **-s** options you provide
+- declares `my %D` and initializes `%D` based on any **-D** options you provide
+- declares `my %S` and initializes `%S` based on any **-s** options you provide
 
 Other than that, everything in the script comes from your input file(s).
 Use the **-E** option to see the generated script.
@@ -159,7 +159,7 @@ So `<?/ ... ?>` is effectively a shorthand for `<? print "\n"; ... ?>`.
 
 The example
 
-	<?!echo Howdy!?>
+	<?! echo Howdy! ?>
 
 produces the output
 
@@ -203,18 +203,21 @@ In this case words like `fooSomeWord` will become `barSomeWord`.
 
 will run `some_perl_code;` at the time of script generation.  Whatever output
 the perl code produces will be included verbatim in the script output.
+Within `some_perl_code`, the current PerlPP instance is available as `$PSelf`.
+
 This can be used to dynamically select which files you want to include,
-using the provided `Include()` function.  For example:
+using the provided `Include()` method.  For example:
 
-	<?:macro my $fn="some_name"; Include $fn; ?>
+	<?:macro my $fn="some_name"; $PSelf->Include($fn); ?>
 
 has the same effect as
 
 	<?:include some_name ?>
 
-but `$fn` can be determined programmatically.  Note that it is not currently
-possible to select the filename to `Include` based on defines set with **-D**,
-since those do not take effect until the script has been generated.
+but `$fn` can be determined programmatically.  Note that defines set with
+**-D** or **-s** do not take effect effect until after the script has been
+generated, which is after the macro code runs.  However, those are available
+as hashes `$PSelf->{Defs}` and `$PSelf->{Sets}` in macro code.
 
 Capturing
 ---------
@@ -298,15 +301,16 @@ Tests with `<?:if NAME ... ?>` and `<?:elsif NAME ... ?>` have two restrictions:
 
 For example, `<?:if FOO eq "something" ?>` (note the whitespace before `?>`!)
 will work fine.  However, if you want to test `(FOO+1)*3`, you will need
-to use the full Perl code.
+to use the full Perl code `<? if( (FOO+1)*3 == 42 ) { ... } ?>` instead of
+`<?:if ?>` and `<?:endif?>`.
 
 Other Features
 --------------
 
 ### Custom Preprocessors
 
 It's possible to create your own pre/post-processors in a `<?:macro ?>` block
-using `PerlPP::AddPreprocessor` and `PerlPP::AddPostprocessor`.
+using `$PSelf->AddPreprocessor` and `$PSelf->AddPostprocessor`.
 This feature is used in [BigBenBox](https://github.com/d-ash/BigBenBox) for
 generating code in the C programming language.
 
@@ -329,14 +333,6 @@ and create corresponding *~/.vim/after/syntax/FILETYPE.vim*
 
 FILETYPE can be determined with `:set ft?`
 
-## Developing PerlPP
-
-Perlpp should run on any Perl v5.10+.  However, it does require `Getopt::Long`
-v2.50 or higher, so you might have to grab that from CPAN.
-To run the tests, you also need to grab `IPC::Run3`.
-
-The `Makefile` just runs the tests in `t/`; there is no build step.
-
 ## Copyright
 
 Distributed under the MIT license --- see

lib/Text/PerlPP.pm

@@ -5,7 +5,7 @@ package Text::PerlPP;
 
 # Semantic versioning, packed per Perl rules.  Must always be at least one
 # digit left of the decimal, and six digits right of the decimal.
-our $VERSION = '0.500002';
+our $VERSION = '0.500004';
 
 use 5.010001;
 use strict;

t/02-basic.t

@@ -3,7 +3,6 @@
 use rlib 'lib';
 use PerlPPTest;
 
-use IPC::Run3;
 (my $whereami = __FILE__) =~ s/02-basic\.t$//;
 #diag join(' ', 'File is', __FILE__, 'whereami', $whereami);
 
@@ -39,7 +38,7 @@ my @testcases=(
 
 ); #@testcases
 
-plan tests => count_tests(\@testcases, 1, 2);
+plan tests => scalar count_tests(\@testcases, 1, 2);
 
 for my $lrTest (@testcases) {
 	my ($testin, $refout, $referr) = @$lrTest;

t/02-readme.t

@@ -88,7 +88,7 @@ RESULT
 	[ '<?= "!" . "?>foo<?= 42 ?><?" . "bar" ?>', '!foo42bar' ],
 ); #@testcases
 
-plan tests => count_tests(\@testcases, 1, 2);
+plan tests => scalar count_tests(\@testcases, 1, 2);
 
 for my $lrTest (@testcases) {
 	my ($testin, $refout, $referr) = @$lrTest;

t/03-cmdline.t

@@ -101,7 +101,7 @@ $testcases->load('-v','',qr/\bversion\b/)->
 
 ; #$testcases
 
-plan tests => count_tests($testcases->arr, 3, 4);
+plan tests => scalar count_tests($testcases->arr, 3, 4);
 
 for my $lrTest (@{$testcases->arr}) {
 	my ($where, $opts, $testin, $out_re, $err_re) = @$lrTest;

t/04-include.t

@@ -29,7 +29,7 @@ my @testcases=(
 		"a\nb\nc\n"],
 ); #@testcases
 
-plan tests => count_tests(\@testcases, 2, 3);
+plan tests => scalar count_tests(\@testcases, 2, 3);
 
 for my $lrTest (@testcases) {
 	my ($lineno, $testin, $refout, $referr) = @$lrTest;

t/05-external-command.t

@@ -1,15 +1,18 @@
 #!/usr/bin/env perl
-# Tests of perlpp <?!...?> external commands
+# Tests of perlpp <?!...?> external commands.
 #
-# TODO: On non-Unix, test only `echo` with no parameters.
+# Note: On non-Unix, we test only `echo` with no parameters.
+# On non-Unix, non-Windows, we skip this because yours truly doesn't know
+# enough about what external commands are available! :)
 
 use rlib 'lib';
 use PerlPPTest qw(:DEFAULT quote_string);
 use List::Util 'any';
 
-if(any { $_ eq $^O } 'VMS', 'os390', 'os400', 'riscos', 'amigaos') {
-	plan skip_all => "I don't know how to run this test on $^O";
-	exit;
+my $os = $^O;	# so we can mock $^O to test this test code!
+
+if(any { $_ eq $os } 'VMS', 'os390', 'os400', 'riscos', 'amigaos') {
+	plan skip_all => "I don't know how to run this test on $os";
 }
 
 (my $whereami = __FILE__) =~ s/macro\.t$//;
@@ -38,26 +41,27 @@ do {
 	is($out, "howdy\n", "basic echo");
 };
 
-if (any { $_ eq $^O } 'dos', 'os2', 'MSWin') {
-	skip "I don't know how to run the rest of the tests on $^O", $ntests-1;
-	exit;
-}
+SKIP: {
+	if (any { $_ eq $os } 'dos', 'os2', 'MSWin32') {
+		skip "I don't know how to run the rest of the tests on $os", $ntests-1;
+	}
 
-for my $lrTest (@testcases) {
-	my ($opts, $testin, $out_re, $err_re) = @$lrTest;
-	my ($out, $err);
+	for my $lrTest (@testcases) {
+		my ($opts, $testin, $out_re, $err_re) = @$lrTest;
+		my ($out, $err);
 
-	#diag "perlpp $opts <<<@{[quote_string $testin]}";
-	run_perlpp $opts, \$testin, \$out, \$err;
+		#diag "perlpp $opts <<<@{[quote_string $testin]}";
+		run_perlpp $opts, \$testin, \$out, \$err;
 
-	if(defined $out_re) {
-		like($out, $out_re);
-	}
-	if(defined $err_re) {
-		like($err, $err_re);
-	}
+		if(defined $out_re) {
+			like($out, $out_re);
+		}
+		if(defined $err_re) {
+			like($err, $err_re);
+		}
 
-} # foreach test
+	} # foreach test
+} #SKIP
 
 # TODO test -o / --output, and processing input from files rather than stdin