1PERLHACKTUT(1) Perl Programmers Reference Guide PERLHACKTUT(1)
2
6 perlhacktut - Walk through the creation of a simple C code patch
7
9 This document takes you through a simple patch example.
10 If you haven't read perlhack yet, go do that first! You might also want
12 to read through perlsource too.
13 Once you're done here, check out perlhacktips next.
15
17 Let's take a simple patch from start to finish.
18 Here's something Larry suggested: if a "U" is the first active format
20 during a "pack", (for example, "pack "U3C8", @stuff") then the
21 resulting string should be treated as UTF-8 encoded.
22 If you are working with a git clone of the Perl repository, you will
24 want to create a branch for your changes. This will make creating a
25 proper patch much simpler. See the perlgit for details on how to do
26 this.
27 Writing the patch
29 How do we prepare to fix this up? First we locate the code in question
30 - the "pack" happens at runtime, so it's going to be in one of the pp
31 files. Sure enough, "pp_pack" is in pp.c. Since we're going to be
32 altering this file, let's copy it to pp.c~.
33 [Well, it was in pp.c when this tutorial was written. It has now been
35 split off with "pp_unpack" to its own file, pp_pack.c]
36 Now let's look over "pp_pack": we take a pattern into "pat", and then
38 loop over the pattern, taking each format character in turn into
39 "datum_type". Then for each possible format character, we swallow up
40 the other arguments in the pattern (a field width, an asterisk, and so
41 on) and convert the next chunk input into the specified format, adding
42 it onto the output SV "cat".
43 How do we know if the "U" is the first format in the "pat"? Well, if we
45 have a pointer to the start of "pat" then, if we see a "U" we can test
46 whether we're still at the start of the string. So, here's where "pat"
47 is set up:
48 STRLEN fromlen;
50 char *pat = SvPVx(*++MARK, fromlen);
51 char *patend = pat + fromlen;
52 I32 len;
53 I32 datumtype;
54 SV *fromstr;
55 We'll have another string pointer in there:
57 STRLEN fromlen;
59 char *pat = SvPVx(*++MARK, fromlen);
60 char *patend = pat + fromlen;
61 + char *patcopy;
62 I32 len;
63 I32 datumtype;
64 SV *fromstr;
65 And just before we start the loop, we'll set "patcopy" to be the start
67 of "pat":
68 items = SP - MARK;
70 MARK++;
71 SvPVCLEAR(cat);
72 + patcopy = pat;
73 while (pat < patend) {
74 Now if we see a "U" which was at the start of the string, we turn on
76 the "UTF8" flag for the output SV, "cat":
77 + if (datumtype == 'U' && pat==patcopy+1)
79 + SvUTF8_on(cat);
80 if (datumtype == '#') {
81 while (pat < patend && *pat != '\n')
82 pat++;
83 Remember that it has to be "patcopy+1" because the first character of
85 the string is the "U" which has been swallowed into "datumtype!"
86 Oops, we forgot one thing: what if there are spaces at the start of the
88 pattern? "pack(" U*", @stuff)" will have "U" as the first active
89 character, even though it's not the first thing in the pattern. In this
90 case, we have to advance "patcopy" along with "pat" when we see spaces:
91 if (isSPACE(datumtype))
93 continue;
94 needs to become
96 if (isSPACE(datumtype)) {
98 patcopy++;
99 continue;
100 }
101 OK. That's the C part done. Now we must do two additional things before
103 this patch is ready to go: we've changed the behaviour of Perl, and so
104 we must document that change. We must also provide some more regression
105 tests to make sure our patch works and doesn't create a bug somewhere
106 else along the line.
107 Testing the patch
109 The regression tests for each operator live in t/op/, and so we make a
110 copy of t/op/pack.t to t/op/pack.t~. Now we can add our tests to the
111 end. First, we'll test that the "U" does indeed create Unicode strings.
112 t/op/pack.t has a sensible ok() function, but if it didn't we could use
114 the one from t/test.pl.
115 require './test.pl';
117 plan( tests => 159 );
118 so instead of this:
120 print 'not ' unless "1.20.300.4000" eq sprintf "%vd",
122 pack("U*",1,20,300,4000);
123 print "ok $test\n"; $test++;
124 we can write the more sensible (see Test::More for a full explanation
126 of is() and other testing functions).
127 is( "1.20.300.4000", sprintf "%vd", pack("U*",1,20,300,4000),
129 "U* produces Unicode" );
130 Now we'll test that we got that space-at-the-beginning business right:
132 is( "1.20.300.4000", sprintf "%vd", pack(" U*",1,20,300,4000),
134 " with spaces at the beginning" );
135 And finally we'll test that we don't make Unicode strings if "U" is not
137 the first active format:
138 isnt( v1.20.300.4000, sprintf "%vd", pack("C0U*",1,20,300,4000),
140 "U* not first isn't Unicode" );
141 Mustn't forget to change the number of tests which appears at the top,
143 or else the automated tester will get confused. This will either look
144 like this:
145 print "1..156\n";
147 or this:
149 plan( tests => 156 );
151 We now compile up Perl, and run it through the test suite. Our new
153 tests pass, hooray!
154 Documenting the patch
156 Finally, the documentation. The job is never done until the paperwork
157 is over, so let's describe the change we've just made. The relevant
158 place is pod/perlfunc.pod; again, we make a copy, and then we'll insert
159 this text in the description of "pack":
160 =item *
162 If the pattern begins with a C<U>, the resulting string will be treated
164 as UTF-8-encoded Unicode. You can force UTF-8 encoding on in a string
165 with an initial C<U0>, and the bytes that follow will be interpreted as
166 Unicode characters. If you don't want this to happen, you can begin
167 your pattern with C<C0> (or anything else) to force Perl not to UTF-8
168 encode your string, and then follow this with a C<U*> somewhere in your
169 pattern.
170 Submit
172 See perlhack for details on how to submit this patch.
173
175 This document was originally written by Nathan Torkington, and is
176 maintained by the perl5-porters mailing list.
177perl v5.32.1 2021-03-31 PERLHACKTUT(1)