From f68fcf40a658671507ea044b41a4eb8fbabcfb22 Mon Sep 17 00:00:00 2001
From: Enno Rehling <enno.rehling@gmail.com>
Date: Tue, 8 Nov 2016 22:54:47 +0100
Subject: [PATCH] custom function to make "valid" names. remove any
 non-printable characters from a string. TODO: test with utf-8 values. TODO:
 remove leading/trainling whitespace.

---
 src/util/unicode.c      | 32 ++++++++++++++++++++++++++++++++
 src/util/unicode.h      |  4 ++--
 src/util/unicode.test.c | 12 ++++++++++++
 3 files changed, 46 insertions(+), 2 deletions(-)

diff --git a/src/util/unicode.c b/src/util/unicode.c
index f7ea2653b..dd8e03814 100644
--- a/src/util/unicode.c
+++ b/src/util/unicode.c
@@ -14,6 +14,7 @@
 #include <errno.h>
 #include <string.h>
 #include <wctype.h>
+#include <ctype.h>
 
 #define B00000000 0x00
 #define B10000000 0x80
@@ -31,6 +32,37 @@
 #define B00000011 0x03
 #define B00000001 0x01
 
+int unicode_utf8_mkname(utf8_t * op, size_t outlen, const utf8_t * ip)
+{
+    while (*ip) {
+        ucs4_t ucs = *ip;
+        size_t size = 1;
+        bool isp = false;
+//        bool iss = false;
+        if (ucs & 0x80) {
+            int ret = unicode_utf8_to_ucs4(&ucs, ip, &size);
+            if (ret !=0) {
+                return ret;
+            }
+            isp = iswprint(ucs);
+//            iss = iswspace(ucs);
+        } else {
+            isp = isprint(ucs);
+//            iss = isspace(ucs);
+        }
+        if (size > outlen) {
+            return ENOMEM;
+        }
+        if (isp) {
+            memcpy(op, ip, size);
+            op += size;
+            outlen -= size;
+        }
+        ip += size;
+    }
+    return 0;
+}
+
 int unicode_utf8_tolower(utf8_t * op, size_t outlen, const utf8_t * ip)
 {
     while (*ip) {
diff --git a/src/util/unicode.h b/src/util/unicode.h
index df68ade02..c817a04e3 100644
--- a/src/util/unicode.h
+++ b/src/util/unicode.h
@@ -41,8 +41,8 @@ extern "C" {
     int unicode_utf8_strcasecmp(const utf8_t * a, const utf8_t * b);
     int unicode_latin1_to_utf8(utf8_t * out, size_t * outlen,
         const char *in, size_t * inlen);
-    int unicode_utf8_tolower(utf8_t * out, size_t outlen,
-        const utf8_t * in);
+    int unicode_utf8_tolower(utf8_t *op, size_t outlen, const utf8_t *ip);
+    int unicode_utf8_mkname(utf8_t *op, size_t outlen, const utf8_t *ip);
 
 #ifdef __cplusplus
 }
diff --git a/src/util/unicode.test.c b/src/util/unicode.test.c
index 410e8e68a..3cfa0615c 100644
--- a/src/util/unicode.test.c
+++ b/src/util/unicode.test.c
@@ -5,6 +5,17 @@
 #include <string.h>
 #include <errno.h>
 
+static void test_unicode_mkname(CuTest * tc)
+{
+    char buffer[32];
+    CuAssertIntEquals(tc, 0, unicode_utf8_mkname(buffer, sizeof(buffer), "HeLlO\nW0Rld"));
+    CuAssertStrEquals(tc, "HeLlOW0Rld", buffer);
+    memset(buffer, 0, sizeof(buffer));
+    buffer[5] = 'X';
+    CuAssertIntEquals(tc, ENOMEM, unicode_utf8_mkname(buffer, 5, "HeLl\n W0Rld"));
+    CuAssertStrEquals(tc, "HeLl X", buffer);
+}
+
 static void test_unicode_tolower(CuTest * tc)
 {
     char buffer[32];
@@ -66,6 +77,7 @@ CuSuite *get_unicode_suite(void)
 {
     CuSuite *suite = CuSuiteNew();
     SUITE_ADD_TEST(suite, test_unicode_tolower);
+    SUITE_ADD_TEST(suite, test_unicode_mkname);
     SUITE_ADD_TEST(suite, test_unicode_utf8_to_other);
     return suite;
 }