fix(frontends/lean): identifier size when using unicode

see issue #756

src/frontends/lean/parser.cpp

@@ -174,7 +174,7 @@ void parser::scan() {
         if (p.first == m_info_at_line) {
             if (curr_is_identifier()) {
                 name const & id = get_name_val();
-                if (p.second <= m_info_at_col && m_info_at_col < p.second + id.size()) {
+                if (p.second <= m_info_at_col && m_info_at_col < p.second + id.utf8_size()) {
                     print_lean_info_header(regular_stream().get_stream());
                     bool ok = true;
                     try {
@@ -190,7 +190,7 @@ void parser::scan() {
                 }
             } else if (curr_is_keyword()) {
                 name const & tk = get_token_info().token();
-                if (p.second <= m_info_at_col && m_info_at_col < p.second + tk.size()) {
+                if (p.second <= m_info_at_col && m_info_at_col < p.second + tk.utf8_size()) {
                     print_lean_info_header(regular_stream().get_stream());
                     try {
                         print_token_info(*this, tk);
@@ -200,7 +200,7 @@ void parser::scan() {
                 }
             } else if (curr_is_command()) {
                 name const & tk = get_token_info().token();
-                if (p.second <= m_info_at_col && m_info_at_col < p.second + tk.size()) {
+                if (p.second <= m_info_at_col && m_info_at_col < p.second + tk.utf8_size()) {
                     print_lean_info_header(regular_stream().get_stream());
                     regular_stream() << "'" << tk << "' is a command\n";
                     print_lean_info_footer(regular_stream().get_stream());

src/util/name.cpp

@@ -19,6 +19,7 @@ Author: Leonardo de Moura
 #include "util/hash.h"
 #include "util/trace.h"
 #include "util/ascii.h"
+#include "util/utf8.h"
 #include "util/object_serializer.h"
 #include "util/lua_list.h"
 
@@ -294,7 +295,7 @@ static unsigned num_digits(unsigned k) {
     return r;
 }
 
-size_t name::size() const {
+size_t name::size_core(bool unicode) const {
     if (m_ptr == nullptr) {
         return strlen(anonymous_str);
     } else {
@@ -303,7 +304,7 @@ size_t name::size() const {
         size_t r      = 0;
         while (true) {
             if (i->m_is_string) {
-                r += strlen(i->m_str);
+                r += unicode ? utf8_strlen(i->m_str) : strlen(i->m_str);
             } else {
                 r += num_digits(i->m_k);
             }
@@ -318,6 +319,14 @@ size_t name::size() const {
     }
 }
 
+size_t name::size() const {
+    return size_core(false);
+}
+
+size_t name::utf8_size() const {
+    return size_core(true);
+}
+
 unsigned name::hash() const {
     return m_ptr ? m_ptr->m_hash : 11;
 }

src/util/name.h

@@ -33,6 +33,8 @@ private:
     explicit name(unsigned k);
     // the parameter bool is only used to distinguish this constructor from the public one.
     name(name const & prefix, unsigned k, bool);
+
+    size_t size_core(bool unicode) const;
 public:
     name();
     name(char const * name);
@@ -100,6 +102,8 @@ public:
     std::string to_string(char const * sep = lean_name_separator) const;
     /** \brief Size of the this name (in characters). */
     size_t size() const;
+    /** \brief Size of the this name in unicode. */
+    size_t utf8_size() const;
     unsigned hash() const;
     /** \brief Return true iff the name contains only safe ASCII chars */
     bool is_safe_ascii() const;

src/util/utf8.cpp

@@ -4,6 +4,7 @@ Released under Apache 2.0 license as described in the file LICENSE.
 
 Author: Leonardo de Moura
 */
+#include <cstdlib>
 
 namespace lean {
 bool is_utf8_next(unsigned char c) { return (c & 0xC0) == 0x80; }
@@ -26,4 +27,14 @@ unsigned get_utf8_size(unsigned char c) {
     else
         return 0;
 }
+
+size_t utf8_strlen(char const * str) {
+    size_t r = 0;
+    while (*str != 0) {
+        unsigned sz = get_utf8_size(*str);
+        r++;
+        str += sz;
+    }
+    return r;
+}
 }

src/util/utf8.h

@@ -8,4 +8,5 @@ Author: Leonardo de Moura
 namespace lean {
 bool is_utf8_next(unsigned char c);
 unsigned get_utf8_size(unsigned char c);
+size_t utf8_strlen(char const * str);
 }

tests/lean/extra/print_info.21.0.expected.out

@@ -0,0 +1,4 @@
+print_info.lean:13:8: warning: using 'sorry'
+LEAN_INFORMATION
+parameter a₁ : ℕ
+END_LEAN_INFORMATION

tests/lean/extra/print_info.21.1.expected.out

@@ -0,0 +1,4 @@
+print_info.lean:13:8: warning: using 'sorry'
+LEAN_INFORMATION
+parameter a₁ : ℕ
+END_LEAN_INFORMATION

tests/lean/extra/print_info.21.3.expected.out

@@ -0,0 +1,6 @@
+print_info.lean:13:8: warning: using 'sorry'
+LEAN_INFORMATION
+_ `+`:65 _:65 :=
+  | nat.add #1 #0
+  | [priority 999] int.add #1 #0
+END_LEAN_INFORMATION

tests/lean/extra/print_info.lean

@@ -15,3 +15,9 @@ end
 
 example (n : nat) : n + n = n + n :=
 rfl
+
+example (a₁ a₂ a₃ : nat) : a₁ = 0 → a₂ = 0 → a₃ = 0 → a₁ + a₂ + a₃ = 0 :=
+assume h₁ h₂ h₃, calc
+a₁ + a₂ + a₃ = 0 + a₂ + a₃ : h₁
+    ...      = 0 + 0 + a₃  : h₂
+    ...      = 0 + 0 + 0   : h₃

tests/lean/extra/print_info.sh

@@ -7,8 +7,8 @@ fi
 LEAN=$1
 export LEAN_PATH=../../../library:.
 
-lines=('4'  '7'  '8'  '12' '12' '12' '17' '17');
+lines=('4'  '7'  '8'  '12' '12' '12' '17' '17' '21' '21' '21');
-cols=('16'  '18' '19' '19' '20' '30' '0'  '2');
+cols=('16'  '18' '19' '19' '20' '30' '0'  '2'  '0'  '1'  '3');
 size=${#lines[@]}
 
 i=0