Improve filtering of ANSI escape sequences in build logs

All ANSI sequences except color setting are now filtered out. In
particular, terminal resets (such as from NixOS VM tests) are filtered
out.

Also, fix the completely broken tab character handling.

3 files changed, 48 insertions, 31 deletions

diff --git a/src/libutil/logging.cc b/src/libutil/logging.cc
index 6924e0080..27a631a37 100644
--- a/src/libutil/logging.cc
+++ b/src/libutil/logging.cc
@@ -44,7 +44,7 @@ public:
             prefix = std::string("<") + c + ">";
         }
 
-        writeToStderr(prefix + (tty ? fs.s : filterANSIEscapes(fs.s)) + "\n");
+        writeToStderr(prefix + filterANSIEscapes(fs.s) + "\n");
     }
 
     void startActivity(ActivityId act, Verbosity lvl, ActivityType type,
diff --git a/src/libutil/util.cc b/src/libutil/util.cc
index 272997397..f7a12d21b 100644
--- a/src/libutil/util.cc
+++ b/src/libutil/util.cc
@@ -1178,36 +1178,51 @@ void ignoreException()
 }
 
 
-string filterANSIEscapes(const string & s, bool nixOnly)
-{
-    string t, r;
-    enum { stTop, stEscape, stCSI } state = stTop;
-    for (auto c : s) {
-        if (state == stTop) {
-            if (c == '\e') {
-                state = stEscape;
-                r = c;
-            } else
-                t += c;
-        } else if (state == stEscape) {
-            r += c;
-            if (c == '[')
-                state = stCSI;
-            else {
-                t += r;
-                state = stTop;
+std::string filterANSIEscapes(const std::string & s, unsigned int width)
+{
+    std::string t, e;
+    size_t w = 0;
+    auto i = s.begin();
+
+    while (w < (size_t) width && i != s.end()) {
+
+        if (*i == '\e') {
+            std::string e;
+            e += *i++;
+            char last = 0;
+
+            if (i != s.end() && *i == '[') {
+                e += *i++;
+                // eat parameter bytes
+                while (i != s.end() && *i >= 0x30 && *i <= 0x3f) e += *i++;
+                // eat intermediate bytes
+                while (i != s.end() && *i >= 0x20 && *i <= 0x2f) e += *i++;
+                // eat final byte
+                if (i != s.end() && *i >= 0x40 && *i <= 0x7e) e += last = *i++;
+            } else {
+                if (i != s.end() && *i >= 0x40 && *i <= 0x5f) e += *i++;
             }
-        } else {
-            r += c;
-            if (c >= 0x40 && c <= 0x7e) {
-                if (nixOnly && (c != 'p' && c != 'q' && c != 's' && c != 'a' && c != 'b'))
-                    t += r;
-                state = stTop;
-                r.clear();
+
+            if (last == 'm')
+                t += e;
+        }
+
+        else if (*i == '\t') {
+            i++; t += ' '; w++;
+            while (w < (size_t) width && w % 8) {
+                t += ' '; w++;
             }
         }
+
+        else if (*i == '\r')
+            // do nothing for now
+            ;
+
+        else {
+            t += *i++; w++;
+        }
     }
-    t += r;
+
     return t;
 }
 
diff --git a/src/libutil/util.hh b/src/libutil/util.hh
index 75eb97515..47e02bc89 100644
--- a/src/libutil/util.hh
+++ b/src/libutil/util.hh
@@ -388,10 +388,12 @@ void ignoreException();
 #define ANSI_BLUE "\e[34;1m"
 
 
-/* Filter out ANSI escape codes from the given string. If ‘nixOnly’ is
-   set, only filter escape codes generated by Nixpkgs' stdenv (used to
-   denote nesting etc.). */
-string filterANSIEscapes(const string & s, bool nixOnly = false);
+/* Truncate a string to 'width' printable characters. Certain ANSI
+   escape sequences (such as colour setting) are copied but not
+   included in the character count. Other ANSI escape sequences are
+   filtered. Also, tabs are expanded to spaces. */
+std::string filterANSIEscapes(const std::string & s,
+    unsigned int width = std::numeric_limits<unsigned int>::max());
 
 
 /* Base64 encoding/decoding. */