[git:edid-decode/master] edid-decode: improve extract_string()

[Hans Verkuil]

This is an automatic generated email to let you know that the following patch 
were queued:

Subject: edid-decode: improve extract_string()
Author:  Hans Verkuil <hverkuil-ci...@xs4all.nl>
Date:    Wed Jul 31 15:18:46 2024 +0200

This function assumed regular 7-bit ASCII, but it really is full
8-bit CP437. So don't stop at the first unprintable character, but
just continue.

While at it, improve the error/warning messages.

Signed-off-by: Hans Verkuil <hverkuil-ci...@xs4all.nl>

 parse-base-block.cpp | 26 ++++++++++++++++++++------
 1 file changed, 20 insertions(+), 6 deletions(-)

---

diff --git a/parse-base-block.cpp b/parse-base-block.cpp
index 5644e72999c9..6565932f24a7 100644
--- a/parse-base-block.cpp
+++ b/parse-base-block.cpp
@@ -499,27 +499,41 @@ char *extract_string(const unsigned char *x, unsigned len)
        memset(s, 0, sizeof(s));
 
        for (i = 0; i < len; i++) {
+               // The encoding is cp437, so any character is allowed,
+               // but in practice it is unwise to use a non-ASCII character.
+               bool non_ascii = (x[i] >= 1 && x[i] < 0x20 && x[i] != 0x0a) || 
x[i] >= 0x7f;
+
                if (seen_newline) {
                        if (x[i] != 0x20) {
                                fail("Non-space after newline.\n");
                                return s;
                        }
-               } else if (isgraph(x[i]) || x[i] == 0x20) {
-                       s[i] = x[i];
                } else if (x[i] == 0x0a) {
                        seen_newline = true;
                        if (!i)
                                fail("Empty string.\n");
                        else if (s[i - 1] == 0x20)
-                               fail("One or more trailing spaces.\n");
-               } else {
-                       fail("Non-printable character.\n");
+                               fail("One or more trailing spaces before 
newline.\n");
+               } else if (!x[i]) {
+                       // While incorrect, a \0 is often used to end the string
+                       fail("NUL byte at position %u.\n", i);
                        return s;
+               } else if (x[i] == 0xff) {
+                       // 0xff is sometimes (incorrectly) used to pad the 
remainder
+                       // of the string
+                       fail("0xff byte at position %u.\n", i);
+                       return s;
+               } else if (!non_ascii) {
+                       s[i] = x[i];
+               } else {
+                       warn("Non-ASCII character 0x%02x at position %u, can 
cause problems.\n",
+                            x[i], i);
+                       s[i] = '.';
                }
        }
        /* Does the string end with a space? */
        if (!seen_newline && s[len - 1] == 0x20)
-               fail("One or more trailing spaces.\n");
+               fail("No newline, but one or more trailing spaces.\n");
 
        return s;
 }