diff --git a/src/source_inspector/strings_xgettext.py b/src/source_inspector/strings_xgettext.py
index 2cf98b5..41ed1b1 100644
--- a/src/source_inspector/strings_xgettext.py
+++ b/src/source_inspector/strings_xgettext.py
@@ -9,6 +9,7 @@
 #
 
 import logging
+import string
 
 import attr
 from commoncode import command
@@ -55,13 +56,13 @@ def get_source_strings(location, **kwargs):
     """
     Return a mapping of strings for a source file at ``location``.
     """
-    return dict(source_strings=list(collect_strings(location=location, strip=True)))
+    return dict(source_strings=list(collect_strings(location=location, clean=True)))
 
 
-def collect_strings(location, strip=False):
+def collect_strings(location, clean=True):
     """
     Yield mappings of strings collected from file at location.
-    Strip strings if ``strip`` is True.
+    Clean strings if ``clean`` is True.
     """
     if not is_xgettext_installed():
         return
@@ -82,13 +83,13 @@ def collect_strings(location, strip=False):
     if rc != 0:
         raise Exception(open(err).read())
 
-    yield from parse_po_text(po_text=result, strip=strip)
+    yield from parse_po_text(po_text=result, clean=clean)
 
 
-def parse_po_text(po_text, strip=False):
+def parse_po_text(po_text, clean=True):
     """
     Yield mappings of strings collected from the ``po_text`` string.
-    Strip strings if ``strip`` is True.
+    Clean strings if ``clean`` is True.
 
     The po text lines looks like this:
     - Blocks sperated by 2 lines.
@@ -143,12 +144,57 @@ def parse_po_text(po_text, strip=False):
             elif line.startswith('"'):
                 strings.append(line)
 
-        strings = [l.strip('"').replace("\\n", "\n") for l in strings]
+        strings = [l.strip('"') for l in strings]
         string = "".join(strings)
-        if strip:
-            string = string.strip()
+        if clean:
+            string = clean_string(string)
+        if string:
+            yield dict(line_numbers=line_numbers, string=string)
 
-        yield dict(line_numbers=line_numbers, string=string)
+
+def clean_string(s):
+    """
+    Return a cleaned and normalized string or None.
+    """
+    s = s.strip('"')
+    s = s.replace("\\n", "\n")
+    s = s.strip()
+    non_printables = {
+        "\\a": "\a",
+        "\\b": "\b",
+        "\\v": "\v",
+        "\\f": "\f",
+        "\\x01": "\x01",
+        "\\x02": "\x02",
+        "\\x03": "\x03",
+        "\\x04": "\x04",
+        "\\x05": "\x05",
+        "\\x06": "\x06",
+        "\\x0e": "\x0e",
+        "\\x0f": "\x0f",
+        "\\x10": "\x10",
+        "\\x11": "\x11",
+        "\\x12": "\x12",
+        "\\x13": "\x13",
+        "\\x14": "\x14",
+        "\\x15": "\x15",
+        "\\x16": "\x16",
+        "\\x17": "\x17",
+        "\\x18": "\x18",
+        "\\x19": "\x19",
+        "\\x1a": "\x1a",
+        "\\x1b": "\x1b",
+        "\\x1c": "\x1c",
+        "\\x1d": "\x1d",
+        "\\x1e": "\x1e",
+        "\\x1f": "\x1f",
+        "\\x7f": "\x7f",
+    }
+
+    for plain, encoded in non_printables.items():
+        s = s.replace(plain, "")
+        s = s.replace(encoded, "")
+    return s
 
 
 _IS_XGETTEXT_INSTALLED = None
diff --git a/tests/data/strings_xgettext/lineedit.c-expected.json b/tests/data/strings_xgettext/lineedit.c-expected.json
index 0e28c53..790f512 100644
--- a/tests/data/strings_xgettext/lineedit.c-expected.json
+++ b/tests/data/strings_xgettext/lineedit.c-expected.json
@@ -4,12 +4,6 @@
       "path": "lineedit.c",
       "type": "file",
       "source_strings": [
-        {
-          "line_numbers": [
-            126
-          ],
-          "string": "\u001b"
-        },
         {
           "line_numbers": [
             128,
@@ -24,14 +18,6 @@
           ],
           "string": "HOME"
         },
-        {
-          "line_numbers": [
-            275,
-            1166,
-            2858
-          ],
-          "string": ""
-        },
         {
           "line_numbers": [
             454,
@@ -79,12 +65,6 @@
           ],
           "string": "."
         },
-        {
-          "line_numbers": [
-            905
-          ],
-          "string": ""
-        },
         {
           "line_numbers": [
             1000
@@ -200,7 +180,7 @@
           "line_numbers": [
             3052
           ],
-          "string": "\\\\[\\\\033[32;1m\\\\]\\\\u@\\\\[\\\\x1b[33;1m\\\\]\\\\h:\\\\[\\\\033[34;1m\\\\]\\\\w\\\\[\\\\033[35;1m\\\\] \\\\!\\\\[\\\\e[36;1m\\\\]\\\\$ \\\\[\\\\E[m\\\\]"
+          "string": "\\\\[\\\\033[32;1m\\\\]\\\\u@\\\\[\\[33;1m\\\\]\\\\h:\\\\[\\\\033[34;1m\\\\]\\\\w\\\\[\\\\033[35;1m\\\\] \\\\!\\\\[\\\\e[36;1m\\\\]\\\\$ \\\\[\\\\E[m\\\\]"
         },
         {
           "line_numbers": [
diff --git a/tests/test_symbols_xgettext.py b/tests/test_symbols_xgettext.py
index 01fad31..1e952ac 100644
--- a/tests/test_symbols_xgettext.py
+++ b/tests/test_symbols_xgettext.py
@@ -50,7 +50,7 @@ def test_parse_po_text(self):
 msgid "Collect source symbols using Universal ctags."
 msgstr ""
 """
-        results = list(parse_po_text(test))
+        results = list(parse_po_text(test, clean=False))
         expected = [
             {
                 "line_numbers": [
@@ -80,7 +80,7 @@ def test_parse_po_text(self):
 
         assert results == expected
 
-        results = list(parse_po_text(test, strip=True))
+        results = list(parse_po_text(test, clean=True))
         expected = [
             {
                 "line_numbers": [