Merge pull request #8193 from hercules-ci/dry-strings

Deduplicate string literal rendering, fix 4909

7 files changed, 158 insertions, 64 deletions

diff --git a/src/libcmd/repl.cc b/src/libcmd/repl.cc
index 57848a5d3..80c08bf1c 100644
--- a/src/libcmd/repl.cc
+++ b/src/libcmd/repl.cc
@@ -40,6 +40,7 @@ extern "C" {
 #include "markdown.hh"
 #include "local-fs-store.hh"
 #include "progress-bar.hh"
+#include "print.hh"
 
 #if HAVE_BOEHMGC
 #define GC_INCLUDE_NEW
@@ -425,6 +426,7 @@ StringSet NixRepl::completePrefix(const std::string & prefix)
 }
 
 
+// FIXME: DRY and match or use the parser
 static bool isVarName(std::string_view s)
 {
     if (s.size() == 0) return false;
@@ -894,17 +896,6 @@ std::ostream & NixRepl::printValue(std::ostream & str, Value & v, unsigned int m
 }
 
 
-std::ostream & printStringValue(std::ostream & str, const char * string) {
-    str << "\"";
-    for (const char * i = string; *i; i++)
-        if (*i == '\"' || *i == '\\') str << "\\" << *i;
-        else if (*i == '\n') str << "\\n";
-        else if (*i == '\r') str << "\\r";
-        else if (*i == '\t') str << "\\t";
-        else str << *i;
-    str << "\"";
-    return str;
-}
 
 
 // FIXME: lot of cut&paste from Nix's eval.cc.
@@ -922,12 +913,14 @@ std::ostream & NixRepl::printValue(std::ostream & str, Value & v, unsigned int m
         break;
 
     case nBool:
-        str << ANSI_CYAN << (v.boolean ? "true" : "false") << ANSI_NORMAL;
+        str << ANSI_CYAN;
+        printLiteralBool(str, v.boolean);
+        str << ANSI_NORMAL;
         break;
 
     case nString:
         str << ANSI_WARNING;
-        printStringValue(str, v.string.s);
+        printLiteralString(str, v.string.s);
         str << ANSI_NORMAL;
         break;
 
@@ -964,10 +957,7 @@ std::ostream & NixRepl::printValue(std::ostream & str, Value & v, unsigned int m
                 sorted.emplace(state->symbols[i.name], i.value);
 
             for (auto & i : sorted) {
-                if (isVarName(i.first))
-                    str << i.first;
-                else
-                    printStringValue(str, i.first.c_str());
+                printAttributeName(str, i.first);
                 str << " = ";
                 if (seen.count(i.second))
                     str << "«repeated»";
diff --git a/src/libexpr/eval.cc b/src/libexpr/eval.cc
index 18cfd9531..6668add8c 100644
--- a/src/libexpr/eval.cc
+++ b/src/libexpr/eval.cc
@@ -9,6 +9,7 @@
 #include "filetransfer.hh"
 #include "function-trace.hh"
 #include "profiles.hh"
+#include "print.hh"
 
 #include <algorithm>
 #include <chrono>
@@ -104,18 +105,10 @@ void Value::print(const SymbolTable & symbols, std::ostream & str,
         str << integer;
         break;
     case tBool:
-        str << (boolean ? "true" : "false");
+        printLiteralBool(str, boolean);
         break;
     case tString:
-        str << "\"";
-        for (const char * i = string.s; *i; i++)
-            if (*i == '\"' || *i == '\\') str << "\\" << *i;
-            else if (*i == '\n') str << "\\n";
-            else if (*i == '\r') str << "\\r";
-            else if (*i == '\t') str << "\\t";
-            else if (*i == '$' && *(i+1) == '{') str << "\\" << *i;
-            else str << *i;
-        str << "\"";
+        printLiteralString(str, string.s);
         break;
     case tPath:
         str << path; // !!! escaping?
diff --git a/src/libexpr/nixexpr.cc b/src/libexpr/nixexpr.cc
index eb6f062b4..1557cbbeb 100644
--- a/src/libexpr/nixexpr.cc
+++ b/src/libexpr/nixexpr.cc
@@ -3,6 +3,7 @@
 #include "eval.hh"
 #include "symbol-table.hh"
 #include "util.hh"
+#include "print.hh"
 
 #include <cstdlib>
 
@@ -60,45 +61,12 @@ Pos::operator std::shared_ptr<AbstractPos>() const
     return pos;
 }
 
-/* Displaying abstract syntax trees. */
-
-static void showString(std::ostream & str, std::string_view s)
-{
-    str << '"';
-    for (auto c : s)
-        if (c == '"' || c == '\\' || c == '$') str << "\\" << c;
-        else if (c == '\n') str << "\\n";
-        else if (c == '\r') str << "\\r";
-        else if (c == '\t') str << "\\t";
-        else str << c;
-    str << '"';
-}
-
+// FIXME: remove, because *symbols* are abstract and do not have a single
+//        textual representation; see printIdentifier()
 std::ostream & operator <<(std::ostream & str, const SymbolStr & symbol)
 {
     std::string_view s = symbol;
-
-    if (s.empty())
-        str << "\"\"";
-    else if (s == "if") // FIXME: handle other keywords
-        str << '"' << s << '"';
-    else {
-        char c = s[0];
-        if (!((c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z') || c == '_')) {
-            showString(str, s);
-            return str;
-        }
-        for (auto c : s)
-            if (!((c >= 'a' && c <= 'z') ||
-                  (c >= 'A' && c <= 'Z') ||
-                  (c >= '0' && c <= '9') ||
-                  c == '_' || c == '\'' || c == '-')) {
-                showString(str, s);
-                return str;
-            }
-        str << s;
-    }
-    return str;
+    return printIdentifier(str, s);
 }
 
 void Expr::show(const SymbolTable & symbols, std::ostream & str) const
@@ -118,7 +86,7 @@ void ExprFloat::show(const SymbolTable & symbols, std::ostream & str) const
 
 void ExprString::show(const SymbolTable & symbols, std::ostream & str) const
 {
-    showString(str, s);
+    printLiteralString(str, s);
 }
 
 void ExprPath::show(const SymbolTable & symbols, std::ostream & str) const
diff --git a/src/libexpr/print.cc b/src/libexpr/print.cc
new file mode 100644
index 000000000..d08672cfc
--- /dev/null
+++ b/src/libexpr/print.cc
@@ -0,0 +1,78 @@
+#include "print.hh"
+
+namespace nix {
+
+std::ostream &
+printLiteralString(std::ostream & str, const std::string_view string)
+{
+    str << "\"";
+    for (auto i = string.begin(); i != string.end(); ++i) {
+        if (*i == '\"' || *i == '\\') str << "\\" << *i;
+        else if (*i == '\n') str << "\\n";
+        else if (*i == '\r') str << "\\r";
+        else if (*i == '\t') str << "\\t";
+        else if (*i == '$' && *(i+1) == '{') str << "\\" << *i;
+        else str << *i;
+    }
+    str << "\"";
+    return str;
+}
+
+std::ostream &
+printLiteralBool(std::ostream & str, bool boolean)
+{
+    str << (boolean ? "true" : "false");
+    return str;
+}
+
+std::ostream &
+printIdentifier(std::ostream & str, std::string_view s) {
+    if (s.empty())
+        str << "\"\"";
+    else if (s == "if") // FIXME: handle other keywords
+        str << '"' << s << '"';
+    else {
+        char c = s[0];
+        if (!((c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z') || c == '_')) {
+            printLiteralString(str, s);
+            return str;
+        }
+        for (auto c : s)
+            if (!((c >= 'a' && c <= 'z') ||
+                  (c >= 'A' && c <= 'Z') ||
+                  (c >= '0' && c <= '9') ||
+                  c == '_' || c == '\'' || c == '-')) {
+                printLiteralString(str, s);
+                return str;
+            }
+        str << s;
+    }
+    return str;
+}
+
+// FIXME: keywords
+static bool isVarName(std::string_view s)
+{
+    if (s.size() == 0) return false;
+    char c = s[0];
+    if ((c >= '0' && c <= '9') || c == '-' || c == '\'') return false;
+    for (auto & i : s)
+        if (!((i >= 'a' && i <= 'z') ||
+              (i >= 'A' && i <= 'Z') ||
+              (i >= '0' && i <= '9') ||
+              i == '_' || i == '-' || i == '\''))
+            return false;
+    return true;
+}
+
+std::ostream &
+printAttributeName(std::ostream & str, std::string_view name) {
+    if (isVarName(name))
+        str << name;
+    else
+        printLiteralString(str, name);
+    return str;
+}
+
+
+}
diff --git a/src/libexpr/print.hh b/src/libexpr/print.hh
new file mode 100644
index 000000000..f9cfc3964
--- /dev/null
+++ b/src/libexpr/print.hh
@@ -0,0 +1,48 @@
+#pragma once
+/**
+ * @file
+ * @brief Common printing functions for the Nix language
+ *
+ * While most types come with their own methods for printing, they share some
+ * functions that are placed here.
+ */
+
+#include <iostream>
+
+namespace nix {
+    /**
+     * Print a string as a Nix string literal.
+     *
+     * Quotes and fairly minimal escaping are added.
+     *
+     * @param s The logical string
+     */
+    std::ostream & printLiteralString(std::ostream & o, std::string_view s);
+    inline std::ostream & printLiteralString(std::ostream & o, const char * s) {
+        return printLiteralString(o, std::string_view(s));
+    }
+    inline std::ostream & printLiteralString(std::ostream & o, const std::string & s) {
+        return printLiteralString(o, std::string_view(s));
+    }
+
+    /** Print `true` or `false`. */
+    std::ostream & printLiteralBool(std::ostream & o, bool b);
+
+    /**
+     * Print a string as an attribute name in the Nix expression language syntax.
+     *
+     * Prints a quoted string if necessary.
+     */
+    std::ostream & printAttributeName(std::ostream & o, std::string_view s);
+
+    /**
+     * Print a string as an identifier in the Nix expression language syntax.
+     *
+     * FIXME: "identifier" is ambiguous. Identifiers do not have a single
+     *        textual representation. They can be used in variable references,
+     *        let bindings, left-hand sides or attribute names in a select
+     *        expression, or something else entirely, like JSON. Use one of the
+     *        `print*` functions instead.
+     */
+    std::ostream & printIdentifier(std::ostream & o, std::string_view s);
+}
diff --git a/src/libstore/derivations.cc b/src/libstore/derivations.cc
index abdfb1978..7eb5cd275 100644
--- a/src/libstore/derivations.cc
+++ b/src/libstore/derivations.cc
@@ -313,6 +313,15 @@ Derivation parseDerivation(const Store & store, std::string && s, std::string_vi
 }
 
 
+/**
+ * Print a derivation string literal to an `std::string`.
+ *
+ * This syntax does not generalize to the expression language, which needs to
+ * escape `$`.
+ *
+ * @param res Where to print to
+ * @param s Which logical string to print
+ */
 static void printString(std::string & res, std::string_view s)
 {
     boost::container::small_vector<char, 64 * 1024> buffer;
diff --git a/tests/repl.sh b/tests/repl.sh
index be8adb742..2b3789521 100644
--- a/tests/repl.sh
+++ b/tests/repl.sh
@@ -79,6 +79,14 @@ testReplResponse '
 "result: ${a}"
 ' "result: 2"
 
+# check dollar escaping https://github.com/NixOS/nix/issues/4909
+# note the escaped \,
+#    \\
+# because the second argument is a regex
+testReplResponse '
+"$" + "{hi}"
+' '"\\${hi}"'
+
 testReplResponse '
 drvPath
 ' '".*-simple.drv"' \