Improved error messages for CLI commands and fixed shell testing issues

tools/shell/embedded_shell.cpp

@@ -226,6 +226,46 @@ void highlight(char* buffer, char* resultBuf, uint32_t renderWidth, uint32_t cur
     strncpy(resultBuf, highlightBuffer.str().c_str(), LINENOISE_MAX_LINE - 1);
 }
 
+int damerauLevenshteinDistance(const std::string& s1, const std::string& s2) {
+    const size_t m = s1.size(), n = s2.size();
+    std::vector<std::vector<size_t>> dp(m + 1, std::vector<size_t>(n + 1, 0));
+    for (size_t i = 0; i <= m; i++) {
+        dp[i][0] = i;
+    }
+    for (size_t j = 0; j <= n; j++) {
+        dp[0][j] = j;
+    }
+    for (size_t i = 1; i <= m; i++) {
+        for (size_t j = 1; j <= n; j++) {
+            if (s1[i - 1] == s2[j - 1]) {
+				dp[i][j] = dp[i - 1][j - 1];
+                if (i > 1 && j > 1 && s1[i - 1] == s2[j - 2] && s1[i - 2] == s2[j - 1]) {
+					dp[i][j] = std::min(dp[i][j], dp[i - 2][j - 2]);
+				}
+            } else {
+                dp[i][j] = 1 + std::min({dp[i - 1][j], dp[i][j - 1], dp[i - 1][j - 1]});
+                if (i > 1 && j > 1 && s1[i - 1] == s2[j - 2] && s1[i - 2] == s2[j - 1]) {
+					dp[i][j] = std::min(dp[i][j], dp[i - 2][j - 2] + 1);
+				}
+            }
+        }
+    }
+    return dp[m][n];
+}
+
+std::string findClosestCommand(std::string lineStr) {
+    std::string closestCommand = "";
+    int editDistance = INT_MAX;
+    for (auto& command : shellCommand.commandList) {
+        int distance = damerauLevenshteinDistance(command, lineStr);
+        if (distance < editDistance) {
+            editDistance = distance;
+            closestCommand = command;
+        }
+    }
+    return closestCommand;
+}
+
 int EmbeddedShell::processShellCommands(std::string lineStr) {
     if (lineStr == shellCommand.HELP) {
         printHelp();
@@ -239,6 +279,7 @@ int EmbeddedShell::processShellCommands(std::string lineStr) {
         setMaxWidth(lineStr.substr(strlen(shellCommand.MAXWIDTH)));
     } else {
         printf("Error: Unknown command: \"%s\". Enter \":help\" for help\n", lineStr.c_str());
+        printf("Did you mean: \"%s\"?\n", findClosestCommand(lineStr).c_str());
     }
     return 0;
 }
@@ -316,7 +357,14 @@ void EmbeddedShell::run() {
                 if (queryResult->isSuccess()) {
                     printExecutionResult(*queryResult);
                 } else {
-                    printf("Error: %s\n", queryResult->getErrorMessage().c_str());
+                    std::string lineStrTrimmed = lineStr;
+                    lineStrTrimmed = lineStrTrimmed.erase(0, lineStr.find_first_not_of(" \t\n\r\f\v"));
+                    if (lineStrTrimmed.find_first_of(" \t\n\r\f\v") == std::string::npos && lineStrTrimmed.length() > 1) {
+                        printf("Error: \"%s\" is not a valid Cypher query. Did you mean to issue a CLI command, e.g., \"%s\"?\n", lineStr.c_str(), findClosestCommand(lineStrTrimmed).c_str());
+                    }
+                    else {
+                        printf("Error: %s\n", queryResult->getErrorMessage().c_str());
+                    }
                 }
             }
         } else if (!lineStr.empty() && lineStr[0] != ':') {

tools/shell/test/conftest.py

@@ -167,7 +167,8 @@ def get_tmp_path(tmp_path):
 @pytest.fixture
 def history_path():
     path = os.path.join(KUZU_ROOT, 'tools', 'shell', 'test', 'files')
-    shutil.rmtree(os.path.join(path, "history.txt"), ignore_errors=True)
+    if os.path.exists(os.path.join(path, "history.txt")):
+        os.remove(os.path.join(path, "history.txt"))
     return path
 
 

tools/shell/test/test_shell_basics.py

@@ -1,5 +1,4 @@
 import pytest
-import shutil
 from test_helper import *
 from conftest import ShellTest
 
@@ -87,7 +86,8 @@ def test_multi_queries_one_line(temp_db):
     )
     result = test.run()
     result.check_stdout("databases rule")
-    result.check_stdout(["Error: Parser exception: Invalid input <      >: expected rule oC_Cypher (line: 1, offset: 6)"])
+    result.check_stdout(["Error: Parser exception: mismatched input '<EOF>' expecting {CALL, COMMENT, COPY, EXPORT, DROP, ALTER, BEGIN, COMMIT, COMMIT_SKIP_CHECKPOINT, ROLLBACK, ROLLBACK_SKIP_CHECKPOINT, INSTALL, LOAD, OPTIONAL, MATCH, UNWIND, CREATE, MERGE, SET, DETACH, DELETE, WITH, RETURN} (line: 1, offset: 6)", 
+                         '"      "'])
 
     # two failing queries
     test = (
@@ -96,10 +96,11 @@ def test_multi_queries_one_line(temp_db):
         .statement('RETURN "databases rule" S a;      ;')
     )
     result = test.run()
-    result.check_stdout(["Error: Parser exception: Invalid input < S>: expected rule oC_Cypher (line: 1, offset: 24)", 
+    result.check_stdout(["Error: Parser exception: Invalid input < S>: expected rule ku_Statements (line: 1, offset: 24)", 
                          '"RETURN "databases rule" S a"',
                          "                         ^",
-                         "Error: Parser exception: Invalid input <      >: expected rule oC_Cypher (line: 1, offset: 6)"])
+                         "Error: Parser exception: mismatched input '<EOF>' expecting {CALL, COMMENT, COPY, EXPORT, DROP, ALTER, BEGIN, COMMIT, COMMIT_SKIP_CHECKPOINT, ROLLBACK, ROLLBACK_SKIP_CHECKPOINT, INSTALL, LOAD, OPTIONAL, MATCH, UNWIND, CREATE, MERGE, SET, DETACH, DELETE, WITH, RETURN} (line: 1, offset: 6)",
+                         '"      "'])
 
 
 def test_row_truncation(temp_db, csv_path):
@@ -155,5 +156,6 @@ def test_history_consecutive_repeats(temp_db, history_path):
     assert f.readline() == ''
     f.close()
 
-    shutil.rmtree(os.path.join(history_path, "history.txt"), ignore_errors=True)
+    if os.path.exists(os.path.join(history_path, "history.txt")):
+        os.remove(os.path.join(history_path, "history.txt"))
 

tools/shell/test/test_shell_commands.py

@@ -132,4 +132,19 @@ def test_max_width(temp_db, csv_path):
     result.check_stdout("Node table: LANGUAGE_CODE has been created.")
     result.check_not_stdout("| ... |")
     result.check_stdout("(1 column)")
-
+
+
+    def test_bad_command(temp_db):
+        test = (
+            ShellTest()
+            .add_argument(temp_db)
+            .statement(":maxrows")
+            .statement(":quiy")
+            .statement("clearr;")
+        )
+        result = test.run()
+        result.check_stdout('Error: Unknown command: ":maxrows". Enter ":help" for help')
+        result.check_stdout('Did you mean: ":max_rows"?')
+        result.check_stdout('Error: Unknown command: ":quiy". Enter ":help" for help')
+        result.check_stdout('Did you mean: ":quit"?')
+        result.check_stdout('Error: "clearr;" is not a valid Cypher query. Did you mean to issue a CLI command, e.g., ":clear"?')

tools/shell/test/test_shell_control_edit.py

@@ -1,6 +1,5 @@
 import pytest
 import pexpect
-import shutil
 from test_helper import *
 from conftest import ShellTest
 
@@ -61,7 +60,8 @@ def test_enter(temp_db, key, history_path):
     f = open(os.path.join(history_path, "history.txt"))
     assert f.readline() == 'RETURN "databases rule" AS a;\n'
     f.close()
-    shutil.rmtree(os.path.join(history_path, "history.txt"), ignore_errors=True)
+    if os.path.exists(os.path.join(history_path, "history.txt")):
+        os.remove(os.path.join(history_path, "history.txt"))
 
 
 @pytest.mark.parametrize(
@@ -198,7 +198,8 @@ def test_search(temp_db, key, history_path):
     # enter should process last match
     test.send_finished_statement(KEY_ACTION.ENTER.value)
     assert test.shell_process.expect_exact(["| databases rule |", pexpect.EOF]) == 0
-    shutil.rmtree(os.path.join(history_path, "history.txt"), ignore_errors=True)
+    if os.path.exists(os.path.join(history_path, "history.txt")):
+        os.remove(os.path.join(history_path, "history.txt"))
 
     # test starting search with text inputted already
     test = ShellTest().add_argument(temp_db)

tools/shell/test/test_shell_control_search.py

@@ -1,6 +1,5 @@
 import pytest
 import pexpect
-import shutil
 from test_helper import *
 from conftest import ShellTest
 
@@ -60,7 +59,8 @@ def test_search_next(temp_db, key, history_path):
         test.send_control_statement(key)
     test.send_finished_statement(KEY_ACTION.ENTER.value)
     assert test.shell_process.expect_exact(["| databases rule |", pexpect.EOF]) == 0
-    shutil.rmtree(os.path.join(history_path, "history.txt"), ignore_errors=True)
+    if os.path.exists(os.path.join(history_path, "history.txt")):
+        os.remove(os.path.join(history_path, "history.txt"))
 
 
 @pytest.mark.parametrize(

tools/shell/test/test_shell_flags.py

@@ -1,5 +1,4 @@
 import pytest
-import shutil
 from test_helper import *
 from conftest import ShellTest
 import os
@@ -178,7 +177,8 @@ def test_history_path(temp_db, history_path):
     assert f.readline() == 'RETURN "kuzu is cool" AS b;\n'
     f.close()
 
-    shutil.rmtree(os.path.join(history_path, "history.txt"), ignore_errors=True)
+    if os.path.exists(os.path.join(history_path, "history.txt")):
+        os.remove(os.path.join(history_path, "history.txt"))
 
 
 @pytest.mark.parametrize(