From c03fb02f44a06b8d30203f3567290f0bd70e3c73 Mon Sep 17 00:00:00 2001
From: simon987 <me@simon987.net>
Date: Wed, 15 Apr 2020 19:58:22 -0400
Subject: [PATCH] More text tests simon987/sist2#44

---
 libscan/text/text.c |  4 +++-
 test/main.cpp       | 19 +++++++++++++++++++
 2 files changed, 22 insertions(+), 1 deletion(-)

diff --git a/libscan/text/text.c b/libscan/text/text.c
index 1d3f1db..62ae442 100644
--- a/libscan/text/text.c
+++ b/libscan/text/text.c
@@ -30,7 +30,7 @@ scan_code_t parse_markup(scan_text_ctx_t *ctx, vfile_t *f, document_t *doc) {
 
     int to_read = MIN(MAX_MARKUP_SIZE, f->info.st_size);
 
-    char *buf = malloc(to_read);
+    char *buf = malloc(to_read + 1);
     int ret = f->read(f, buf, to_read);
     if (ret < 0) {
         CTX_LOG_ERRORF(doc->filepath, "read() returned error code: [%d]", ret)
@@ -38,6 +38,8 @@ scan_code_t parse_markup(scan_text_ctx_t *ctx, vfile_t *f, document_t *doc) {
         return SCAN_ERR_READ;
     }
 
+    *(buf + to_read) = '\0';
+
     text_buffer_t tex = text_buffer_create(ctx->content_size);
     text_buffer_append_markup(&tex, buf);
     text_buffer_terminate_string(&tex);
diff --git a/test/main.cpp b/test/main.cpp
index d325c30..90d4143 100644
--- a/test/main.cpp
+++ b/test/main.cpp
@@ -99,6 +99,25 @@ TEST(Text, MemWhitespace) {
     cleanup(&doc, &f);
 }
 
+TEST(Text, MemNoise) {
+    char content[600];
+
+    for (char &i : content) {
+        int x = rand();
+        i = x == 0 ? 1 : x;
+    }
+    content[599] = '\0';
+
+    vfile_t f;
+    document_t doc;
+    load_doc_mem((void *) content, strlen(content), &f, &doc);
+
+    parse_text(&text_500_ctx, &f, &doc);
+
+    ASSERT_TRUE(utf8valid(get_meta(&doc, MetaContent)->str_val) == 0);
+    cleanup(&doc, &f);
+}
+
 TEST(TextMarkup, Mem1) {
     const char *content = "<<a<aa<<<>test<aaaa><>test test    <>";
     vfile_t f;