[src/trunk]: src/usr.bin/xlint/lint1 lint: document how reading tokens from t...

[rillig <rillig%NetBSD.org@localhost>]

details:   https://anonhg.NetBSD.org/src/rev/9a7816dd598c
branches:  trunk
changeset: 372978:9a7816dd598c
user:      rillig <rillig%NetBSD.org@localhost>
date:      Sun Jan 08 22:46:00 2023 +0000

description:
lint: document how reading tokens from the input works

Rename inpc to read_byte, to make the name more expressive.  Since C99,
lint is no longer required to use overly short identifiers.

No functional change.

diffstat:

 usr.bin/xlint/lint1/lex.c  |  38 ++++++++++++++++++++------------------
 usr.bin/xlint/lint1/scan.l |  10 ++++++++--
 2 files changed, 28 insertions(+), 20 deletions(-)

diffs (177 lines):

diff -r 7b0d49ab871c -r 9a7816dd598c usr.bin/xlint/lint1/lex.c
--- a/usr.bin/xlint/lint1/lex.c Sun Jan 08 21:11:45 2023 +0000
+++ b/usr.bin/xlint/lint1/lex.c Sun Jan 08 22:46:00 2023 +0000
@@ -1,4 +1,4 @@
-/* $NetBSD: lex.c,v 1.134 2022/10/01 10:04:06 rillig Exp $ */
+/* $NetBSD: lex.c,v 1.135 2023/01/08 22:46:00 rillig Exp $ */
 
 /*
  * Copyright (c) 1996 Christopher G. Demetriou.  All Rights Reserved.
@@ -38,7 +38,7 @@
 
 #include <sys/cdefs.h>
 #if defined(__RCSID)
-__RCSID("$NetBSD: lex.c,v 1.134 2022/10/01 10:04:06 rillig Exp $");
+__RCSID("$NetBSD: lex.c,v 1.135 2023/01/08 22:46:00 rillig Exp $");
 #endif
 
 #include <ctype.h>
@@ -398,11 +398,13 @@
 }
 
 /*
- * Read a character and ensure that it is positive (except EOF).
- * Increment line count(s) if necessary.
+ * When scanning the remainder of a long token (see lex_input), read a byte
+ * and return it as an unsigned char or as EOF.
+ *
+ * Increment the line counts if necessary.
  */
 static int
-inpc(void)
+read_byte(void)
 {
        int     c;
 
@@ -846,7 +848,7 @@
        int     n, c, v;
 
        if (pbc == -1) {
-               c = inpc();
+               c = read_byte();
        } else {
                c = pbc;
                pbc = -1;
@@ -868,7 +870,7 @@
        case EOF:
                return -2;
        case '\\':
-               switch (c = inpc()) {
+               switch (c = read_byte()) {
                case '"':
                        if (!allow_c90 && delim == '\'')
                                /* \" inside character constants undef... */
@@ -913,7 +915,7 @@
                        v = 0;
                        do {
                                v = (v << 3) + (c - '0');
-                               c = inpc();
+                               c = read_byte();
                        } while (--n > 0 && '0' <= c && c <= '7');
                        pbc = c;
                        if (v > TARG_UCHAR_MAX) {
@@ -928,7 +930,7 @@
                                warning(82);
                        v = 0;
                        n = 0;
-                       while (c = inpc(), isxdigit(c)) {
+                       while (c = read_byte(), isxdigit(c)) {
                                c = isdigit(c) ?
                                    c - '0' : toupper(c) - 'A' + 10;
                                v = (v << 4) + c;
@@ -1048,7 +1050,7 @@
                /* empty string means stdin */
                if (fnl == 0) {
                        fn = "{standard input}";
-                       fnl = 16;                       /* strlen (fn) */
+                       fnl = 16;       /* strlen (fn) */
                }
                curr_pos.p_file = record_filename(fn, fnl);
                /*
@@ -1118,7 +1120,7 @@
        eoc = false;
 
        /* Skip whitespace after the start of the comment */
-       while (c = inpc(), isspace(c))
+       while (c = read_byte(), isspace(c))
                continue;
 
        /* Read the potential keyword to keywd */
@@ -1128,7 +1130,7 @@
                if (islower(c) && l > 0 && ch_isupper(keywd[0]))
                        break;
                keywd[l++] = (char)c;
-               c = inpc();
+               c = read_byte();
        }
        while (l > 0 && ch_isspace(keywd[l - 1]))
                l--;
@@ -1144,14 +1146,14 @@
 
        /* skip whitespace after the keyword */
        while (isspace(c))
-               c = inpc();
+               c = read_byte();
 
        /* read the argument, if the keyword accepts one and there is one */
        l = 0;
        if (keywtab[i].arg) {
                while (isdigit(c) && l < sizeof(arg) - 1) {
                        arg[l++] = (char)c;
-                       c = inpc();
+                       c = read_byte();
                }
        }
        arg[l] = '\0';
@@ -1159,9 +1161,9 @@
 
        /* skip whitespace after the argument */
        while (isspace(c))
-               c = inpc();
+               c = read_byte();
 
-       if (c != '*' || (c = inpc()) != '/') {
+       if (c != '*' || (c = read_byte()) != '/') {
                if (keywtab[i].func != linted)
                        /* extra characters in lint comment */
                        warning(257);
@@ -1179,7 +1181,7 @@
 skip_rest:
        while (!eoc) {
                lc = c;
-               if ((c = inpc()) == EOF) {
+               if ((c = read_byte()) == EOF) {
                        /* unterminated comment */
                        error(256);
                        break;
@@ -1201,7 +1203,7 @@
                /* %s does not support // comments */
                gnuism(312, allow_c90 ? "C90" : "traditional C");
 
-       while ((c = inpc()) != EOF && c != '\n')
+       while ((c = read_byte()) != EOF && c != '\n')
                continue;
 }
 
diff -r 7b0d49ab871c -r 9a7816dd598c usr.bin/xlint/lint1/scan.l
--- a/usr.bin/xlint/lint1/scan.l        Sun Jan 08 21:11:45 2023 +0000
+++ b/usr.bin/xlint/lint1/scan.l        Sun Jan 08 22:46:00 2023 +0000
@@ -1,5 +1,5 @@
 %{
-/* $NetBSD: scan.l,v 1.136 2022/05/20 21:18:55 rillig Exp $ */
+/* $NetBSD: scan.l,v 1.137 2023/01/08 22:46:00 rillig Exp $ */
 
 /*
  * Copyright (c) 1996 Christopher G. Demetriou.  All Rights Reserved.
@@ -35,7 +35,7 @@
 
 #include <sys/cdefs.h>
 #if defined(__RCSID)
-__RCSID("$NetBSD: scan.l,v 1.136 2022/05/20 21:18:55 rillig Exp $");
+__RCSID("$NetBSD: scan.l,v 1.137 2023/01/08 22:46:00 rillig Exp $");
 #endif
 
 #include "lint1.h"
@@ -127,6 +127,12 @@
 
 %%
 
+/*
+ * In the above list of regular expressions, the tokens for character
+ * constants, string literals and comments are incomplete; they only match
+ * a prefix.  The remainder of these tokens is scanned by reading bytes
+ * directly from the input stream.
+ */
 int
 lex_input(void)
 {