select loops and bases in arithmetic contexts

ShellCheck/AST.hs

@@ -23,7 +23,7 @@ import qualified Text.Regex as Re
 
 data Id = Id Int deriving (Show, Eq, Ord)
 
-data Token = T_AND_IF Id | T_OR_IF Id | T_DSEMI Id | T_Semi Id | T_DLESS Id | T_DGREAT Id | T_LESSAND Id | T_GREATAND Id | T_LESSGREAT Id | T_DLESSDASH Id | T_CLOBBER Id | T_If Id | T_Then Id | T_Else Id | T_Elif Id | T_Fi Id | T_Do Id | T_Done Id | T_Case Id | T_Esac Id | T_While Id | T_Until Id | T_For Id | T_Lbrace Id | T_Rbrace Id | T_Lparen Id | T_Rparen Id | T_Bang Id | T_In  Id | T_NEWLINE Id | T_EOF Id | T_Less Id | T_Greater Id | T_SingleQuoted Id String | T_Literal Id String | T_NormalWord Id [Token] | T_DoubleQuoted Id [Token] | T_DollarExpansion Id [Token] | T_DollarBraced Id Token | T_DollarArithmetic Id Token | T_BraceExpansion Id String | T_IoFile Id Token Token | T_HereDoc Id Bool Bool String | T_HereString Id Token | T_FdRedirect Id String Token | T_Assignment Id String Token | T_Array Id [Token] | T_Redirecting Id [Token] Token | T_SimpleCommand Id [Token] [Token] | T_Pipeline Id [Token] | T_Banged Id Token | T_AndIf Id (Token) (Token) | T_OrIf Id (Token) (Token) | T_Backgrounded Id Token | T_IfExpression Id [([Token],[Token])] [Token] | T_Subshell Id [Token] | T_BraceGroup Id [Token] | T_WhileExpression Id [Token] [Token] | T_UntilExpression Id [Token] [Token] | T_ForIn Id String [Token] [Token] | T_CaseExpression Id Token [([Token],[Token])] | T_Function Id String Token | T_Arithmetic Id Token | T_Script Id String [Token] | T_Condition Id ConditionType Token | T_Extglob Id String [Token] | TC_And Id ConditionType String Token Token | TC_Or Id ConditionType String Token Token | TC_Group Id ConditionType Token | TC_Binary Id ConditionType String Token Token | TC_Unary Id ConditionType String Token | TC_Noary Id ConditionType Token | TA_Binary Id String Token Token | TA_Unary Id String Token | TA_Sequence Id [Token] | TA_Variable Id String | TA_Trinary Id Token Token Token | TA_Expansion Id Token | TA_Literal Id String | T_Backticked Id String | T_ProcSub Id String [Token] | T_Glob Id String | T_ForArithmetic Id Token Token Token [Token] | T_DollarSingleQuoted Id String | T_DollarDoubleQuoted Id [Token]
+data Token = T_AND_IF Id | T_OR_IF Id | T_DSEMI Id | T_Semi Id | T_DLESS Id | T_DGREAT Id | T_LESSAND Id | T_GREATAND Id | T_LESSGREAT Id | T_DLESSDASH Id | T_CLOBBER Id | T_If Id | T_Then Id | T_Else Id | T_Elif Id | T_Fi Id | T_Do Id | T_Done Id | T_Case Id | T_Esac Id | T_While Id | T_Until Id | T_For Id | T_Select Id | T_Lbrace Id | T_Rbrace Id | T_Lparen Id | T_Rparen Id | T_Bang Id | T_In  Id | T_NEWLINE Id | T_EOF Id | T_Less Id | T_Greater Id | T_SingleQuoted Id String | T_Literal Id String | T_NormalWord Id [Token] | T_DoubleQuoted Id [Token] | T_DollarExpansion Id [Token] | T_DollarBraced Id Token | T_DollarArithmetic Id Token | T_BraceExpansion Id String | T_IoFile Id Token Token | T_HereDoc Id Bool Bool String | T_HereString Id Token | T_FdRedirect Id String Token | T_Assignment Id String Token | T_Array Id [Token] | T_Redirecting Id [Token] Token | T_SimpleCommand Id [Token] [Token] | T_Pipeline Id [Token] | T_Banged Id Token | T_AndIf Id (Token) (Token) | T_OrIf Id (Token) (Token) | T_Backgrounded Id Token | T_IfExpression Id [([Token],[Token])] [Token] | T_Subshell Id [Token] | T_BraceGroup Id [Token] | T_WhileExpression Id [Token] [Token] | T_UntilExpression Id [Token] [Token] | T_ForIn Id String [Token] [Token] | T_SelectIn Id String [Token] [Token] | T_CaseExpression Id Token [([Token],[Token])] | T_Function Id String Token | T_Arithmetic Id Token | T_Script Id String [Token] | T_Condition Id ConditionType Token | T_Extglob Id String [Token] | TC_And Id ConditionType String Token Token | TC_Or Id ConditionType String Token Token | TC_Group Id ConditionType Token | TC_Binary Id ConditionType String Token Token | TC_Unary Id ConditionType String Token | TC_Noary Id ConditionType Token | TA_Binary Id String Token Token | TA_Unary Id String Token | TA_Sequence Id [Token] | TA_Variable Id String | TA_Trinary Id Token Token Token | TA_Expansion Id Token | TA_Literal Id String | T_Backticked Id String | T_ProcSub Id String [Token] | T_Glob Id String | T_ForArithmetic Id Token Token Token [Token] | T_DollarSingleQuoted Id String | T_DollarDoubleQuoted Id [Token] | TA_Base Id String Token
 
     deriving (Show)
 
@@ -96,6 +96,7 @@ analyze f g i t =
     delve (T_WhileExpression id c l) = dll c l $ T_WhileExpression id
     delve (T_UntilExpression id c l) = dll c l $ T_UntilExpression id
     delve (T_ForIn id v w l) = dll w l $ T_ForIn id v
+    delve (T_SelectIn id v w l) = dll w l $ T_SelectIn id v
     delve (T_CaseExpression id word cases) = do
         newWord <- round word
         newCases <- mapM (\(c, t) -> do
@@ -134,6 +135,7 @@ analyze f g i t =
         c <- round t3
         return $ TA_Trinary id a b c
     delve (TA_Expansion id t) = d1 t $ TA_Expansion id
+    delve (TA_Base id b t) = d1 t $ TA_Base id b
     delve t = return t
 
 getId t = case t of
@@ -160,6 +162,7 @@ getId t = case t of
         T_While id  -> id
         T_Until id  -> id
         T_For id  -> id
+        T_Select id  -> id
         T_Lbrace id  -> id
         T_Rbrace id  -> id
         T_Lparen id  -> id
@@ -197,6 +200,7 @@ getId t = case t of
         T_WhileExpression id _ _  -> id
         T_UntilExpression id _ _  -> id
         T_ForIn id _ _ _  -> id
+        T_SelectIn id _ _ _  -> id
         T_CaseExpression id _ _ -> id
         T_Function id _ _  -> id
         T_Arithmetic id _  -> id
@@ -217,6 +221,7 @@ getId t = case t of
         TA_Trinary id _ _ _  -> id
         TA_Expansion id _  -> id
         TA_Literal id _ -> id
+        TA_Base id _ _ -> id
         T_ProcSub id _ _ -> id
         T_Glob id _ -> id
         T_ForArithmetic id _ _ _ _ -> id

ShellCheck/Analytics.hs

@@ -38,7 +38,6 @@ checks = concat [
 
 runAllAnalytics = checkList checks
 checkList l t m = foldl (\x f -> f t x) m l
-checkList l t m = foldl (\x f -> f t x) m l
 
 runBasicAnalysis f t m = snd $ runState (doAnalysis f t) m
 basicChecks = [
@@ -58,6 +57,7 @@ basicChecks = [
     ,checkForDecimals
     ,checkDivBeforeMult
     ,checkArithmeticDeref
+    ,checkArithmeticBadOctal
     ,checkComparisonAgainstGlob
     ,checkPrintfVar
     ,checkCommarrays
@@ -211,7 +211,7 @@ checkPipePitfalls (T_Pipeline id commands) = do
     for [["grep"], ["sed"]] $  \id -> style id "You don't need grep | sed, sed can filter lines by itself."
     for [["grep"], ["awk"]] $  \id -> style id "You don't need grep | awk, awk can filter lines by itself."
     for [["ls"], ["?"]] $      \id -> warn id "Don't parse ls output; it mangles filenames."
-    for [["ls"], ["grep"]] $   \id -> warn id "Don't use ls | grep. Use a for loop with a condition in it."
+    for [["ls"], ["grep"]] $   \id -> warn id "Don't use ls | grep. Use a glob or a for loop with a condition."
     for [["ls"], ["xargs"]] $  \id -> warn id "Don't use ls | xargs. Use find -exec .. +"
     for [["find"], ["xargs"]]$ \id -> warn id "Don't use find | xargs cmd. find -exec cmd {} + handles whitespace."
     for [["?"], ["echo"]] $    \id -> info id "echo doesn't read from stdin, are you sure you should be piping to it?"
@@ -263,6 +263,7 @@ checkUndeclaredBash t@(T_Script id sb _) m =
     bashism (T_DollarDoubleQuoted id _) = warnMsg id "$\"..\""
     bashism (T_ForArithmetic id _ _ _ _) = warnMsg id "arithmetic for loops"
     bashism (T_Arithmetic id _) = warnMsg id "((..))"
+    bashism (T_SelectIn id _ _ _) = warnMsg id "select loops"
     bashism _ = return()
 
 prop_checkForInQuoted = verify checkForInQuoted "for f in \"$(ls)\"; do echo foo; done"
@@ -509,6 +510,12 @@ checkArithmeticDeref (TA_Expansion _ (T_DollarBraced id l)) | not . excepting $
     excepting s = (any (`elem` "/.:#%?*@") s) || (isDigit $ head s)
 checkArithmeticDeref _ = return ()
 
+prop_checkArithmeticBadOctal1 = verify checkArithmeticBadOctal "(( 0192 ))"
+prop_checkArithmeticBadOctal2 = verifyNot checkArithmeticBadOctal "(( 0x192 ))"
+prop_checkArithmeticBadOctal3 = verifyNot checkArithmeticBadOctal "(( 1 ^ 0777 ))"
+checkArithmeticBadOctal (TA_Base id "0" (TA_Literal _ str)) | '9' `elem` str || '8' `elem` str =
+    err id $ "Numbers with leading 0 are considered octal."
+checkArithmeticBadOctal _ = return ()
 
 prop_checkComparisonAgainstGlob = verify checkComparisonAgainstGlob "[[ $cow == $bar ]]"
 prop_checkComparisonAgainstGlob2 = verifyNot checkComparisonAgainstGlob "[[ $cow == \"$bar\" ]]"
@@ -782,7 +789,8 @@ getModifiedVariablesWithType spacefulF t =
                                                     else []
 
         --Points to 'for' rather than variable
-        T_ForIn id str words _ -> [(id, str, if any (isSpaceful spacefulF) words then Spaceful else Spaceless)]
+        T_ForIn id str words _ -> [(id, str, if any (isSpaceful spacefulF) words || null words then Spaceful else Spaceless)]
+        T_SelectIn id str words _ -> [(id, str, if any (isSpaceful spacefulF) words || null words then Spaceful else Spaceless)]
         _ -> []
 
 isSpaceful :: (String -> Bool) -> Token -> Bool

ShellCheck/Parser.hs

@@ -343,6 +343,7 @@ prop_a7 = isOk readArithmeticContents "3*2**10"
 prop_a8 = isOk readArithmeticContents "3"
 prop_a9 = isOk readArithmeticContents "a^!-b"
 prop_aA = isOk readArithmeticContents "! $?"
+prop_aB = isOk readArithmeticContents "10#08 * 16#f"
 readArithmeticContents =
     readSequence
   where
@@ -381,9 +382,35 @@ readArithmeticContents =
     readNumber = do
         id <- getNextId
         num <- many1 $ oneOf "0123456789."
-        return $ TA_Literal id num
+        return $ TA_Literal id (num)
 
-    readArithTerm = readGroup <|> readExpansion <|> readNumber <|> readVar
+    readBased = getArbitrary <|> getHex <|> getOct
+      where
+        getThing prefix litchars = try $ do
+            id <- getNextId
+            x <- prefix
+            t <- readExpansion <|> (do
+                i <- getNextId
+                stuff <- many1 litchars
+                return $ TA_Literal i stuff)
+            return $ TA_Base id x t
+
+        getArbitrary = getThing arbitrary variableChars
+        getHex = getThing hex hexDigit
+        getOct = getThing oct digit
+
+        arbitrary = try $ do
+            b <- many1 digit
+            s <- char '#'
+            return (b ++ [s])
+        hex = try $ do
+            z <- char '0'
+            x <- oneOf "xX"
+            return (z:x:[])
+        oct = string "0"
+
+    readArithTerm = readBased <|> readArithTermUnit
+    readArithTermUnit = readGroup <|> readExpansion <|> readNumber <|> readVar
 
     readSequence = do
         spacing
@@ -1138,6 +1165,22 @@ readForClause = called "for loop" $ do
         values <- readInClause <|> (readSequentialSep >> return [])
         return $ \id group -> (return $ T_ForIn id name values group)
 
+prop_readSelectClause1 = isOk readSelectClause "select foo in *; do echo $foo; done"
+prop_readSelectClause2 = isOk readSelectClause "select foo; do echo $foo; done"
+readSelectClause = called "select loop" $ do
+    pos <- getPosition
+    (T_Select id) <- g_Select
+    spacing
+    typ <- readRegular
+    group <- readDoGroup pos
+    typ id group
+  where
+    readRegular = do
+        name <- readVariableName
+        spacing
+        values <- readInClause <|> (readSequentialSep >> return [])
+        return $ \id group -> (return $ T_SelectIn id name values group)
+
 readInClause = do
     g_In
     things <- (readCmdWord) `reluctantlyTill`
@@ -1232,7 +1275,7 @@ readPattern = (readNormalWord `thenSkip` spacing) `sepBy1` (char '|' `thenSkip`
 
 readCompoundCommand = do
     id <- getNextId
-    cmd <- choice [ readBraceGroup, readArithmeticExpression, readSubshell, readCondition, readWhileClause, readUntilClause, readIfClause, readForClause, readCaseClause, readFunctionDefinition]
+    cmd <- choice [ readBraceGroup, readArithmeticExpression, readSubshell, readCondition, readWhileClause, readUntilClause, readIfClause, readForClause, readSelectClause, readCaseClause, readFunctionDefinition]
     spacing
     redirs <- many readIoRedirect
     when (not . null $ redirs) $ optional $ do
@@ -1334,6 +1377,7 @@ g_Esac = tryWordToken "esac" T_Esac
 g_While = tryWordToken "while" T_While
 g_Until = tryWordToken "until" T_Until
 g_For = tryWordToken "for" T_For
+g_Select = tryWordToken "select" T_Select
 g_In = tryWordToken "in" T_In
 g_Lbrace = tryWordToken "{" T_Lbrace
 g_Rbrace = tryWordToken "}" T_Rbrace